diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,129645 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 18516,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00016202203499675956,
+      "grad_norm": 9.86984634399414,
+      "learning_rate": 8.992805755395684e-09,
+      "loss": 0.5557,
+      "step": 1
+    },
+    {
+      "epoch": 0.0003240440699935191,
+      "grad_norm": 8.46564769744873,
+      "learning_rate": 1.798561151079137e-08,
+      "loss": 0.5138,
+      "step": 2
+    },
+    {
+      "epoch": 0.0004860661049902787,
+      "grad_norm": 8.7678861618042,
+      "learning_rate": 2.6978417266187054e-08,
+      "loss": 0.5389,
+      "step": 3
+    },
+    {
+      "epoch": 0.0006480881399870382,
+      "grad_norm": 8.85426139831543,
+      "learning_rate": 3.597122302158274e-08,
+      "loss": 0.5323,
+      "step": 4
+    },
+    {
+      "epoch": 0.0008101101749837978,
+      "grad_norm": 8.901697158813477,
+      "learning_rate": 4.496402877697842e-08,
+      "loss": 0.5217,
+      "step": 5
+    },
+    {
+      "epoch": 0.0009721322099805574,
+      "grad_norm": 8.482805252075195,
+      "learning_rate": 5.395683453237411e-08,
+      "loss": 0.5235,
+      "step": 6
+    },
+    {
+      "epoch": 0.001134154244977317,
+      "grad_norm": 8.641008377075195,
+      "learning_rate": 6.294964028776979e-08,
+      "loss": 0.5277,
+      "step": 7
+    },
+    {
+      "epoch": 0.0012961762799740765,
+      "grad_norm": 9.918697357177734,
+      "learning_rate": 7.194244604316547e-08,
+      "loss": 0.5525,
+      "step": 8
+    },
+    {
+      "epoch": 0.001458198314970836,
+      "grad_norm": 8.959720611572266,
+      "learning_rate": 8.093525179856116e-08,
+      "loss": 0.5073,
+      "step": 9
+    },
+    {
+      "epoch": 0.0016202203499675956,
+      "grad_norm": 8.834793090820312,
+      "learning_rate": 8.992805755395684e-08,
+      "loss": 0.529,
+      "step": 10
+    },
+    {
+      "epoch": 0.0017822423849643552,
+      "grad_norm": 8.80196762084961,
+      "learning_rate": 9.892086330935252e-08,
+      "loss": 0.5193,
+      "step": 11
+    },
+    {
+      "epoch": 0.0019442644199611147,
+      "grad_norm": 8.514031410217285,
+      "learning_rate": 1.0791366906474822e-07,
+      "loss": 0.5291,
+      "step": 12
+    },
+    {
+      "epoch": 0.002106286454957874,
+      "grad_norm": 8.85981559753418,
+      "learning_rate": 1.169064748201439e-07,
+      "loss": 0.5294,
+      "step": 13
+    },
+    {
+      "epoch": 0.002268308489954634,
+      "grad_norm": 8.460282325744629,
+      "learning_rate": 1.2589928057553958e-07,
+      "loss": 0.5154,
+      "step": 14
+    },
+    {
+      "epoch": 0.002430330524951393,
+      "grad_norm": 9.40372371673584,
+      "learning_rate": 1.3489208633093525e-07,
+      "loss": 0.5348,
+      "step": 15
+    },
+    {
+      "epoch": 0.002592352559948153,
+      "grad_norm": 8.342942237854004,
+      "learning_rate": 1.4388489208633095e-07,
+      "loss": 0.5128,
+      "step": 16
+    },
+    {
+      "epoch": 0.0027543745949449123,
+      "grad_norm": 8.372694969177246,
+      "learning_rate": 1.5287769784172664e-07,
+      "loss": 0.5387,
+      "step": 17
+    },
+    {
+      "epoch": 0.002916396629941672,
+      "grad_norm": 7.830298900604248,
+      "learning_rate": 1.618705035971223e-07,
+      "loss": 0.4864,
+      "step": 18
+    },
+    {
+      "epoch": 0.0030784186649384314,
+      "grad_norm": 7.605618000030518,
+      "learning_rate": 1.70863309352518e-07,
+      "loss": 0.5016,
+      "step": 19
+    },
+    {
+      "epoch": 0.0032404406999351912,
+      "grad_norm": 7.3603997230529785,
+      "learning_rate": 1.7985611510791368e-07,
+      "loss": 0.4962,
+      "step": 20
+    },
+    {
+      "epoch": 0.0034024627349319506,
+      "grad_norm": 8.310635566711426,
+      "learning_rate": 1.8884892086330937e-07,
+      "loss": 0.4981,
+      "step": 21
+    },
+    {
+      "epoch": 0.0035644847699287103,
+      "grad_norm": 7.44785737991333,
+      "learning_rate": 1.9784172661870504e-07,
+      "loss": 0.4864,
+      "step": 22
+    },
+    {
+      "epoch": 0.0037265068049254697,
+      "grad_norm": 6.7629008293151855,
+      "learning_rate": 2.0683453237410074e-07,
+      "loss": 0.4597,
+      "step": 23
+    },
+    {
+      "epoch": 0.0038885288399222295,
+      "grad_norm": 6.164111137390137,
+      "learning_rate": 2.1582733812949643e-07,
+      "loss": 0.4475,
+      "step": 24
+    },
+    {
+      "epoch": 0.004050550874918989,
+      "grad_norm": 6.299708843231201,
+      "learning_rate": 2.248201438848921e-07,
+      "loss": 0.4362,
+      "step": 25
+    },
+    {
+      "epoch": 0.004212572909915748,
+      "grad_norm": 6.691930294036865,
+      "learning_rate": 2.338129496402878e-07,
+      "loss": 0.4647,
+      "step": 26
+    },
+    {
+      "epoch": 0.004374594944912508,
+      "grad_norm": 6.3759846687316895,
+      "learning_rate": 2.428057553956835e-07,
+      "loss": 0.4502,
+      "step": 27
+    },
+    {
+      "epoch": 0.004536616979909268,
+      "grad_norm": 6.641443729400635,
+      "learning_rate": 2.5179856115107916e-07,
+      "loss": 0.4535,
+      "step": 28
+    },
+    {
+      "epoch": 0.004698639014906027,
+      "grad_norm": 6.435218811035156,
+      "learning_rate": 2.6079136690647483e-07,
+      "loss": 0.4315,
+      "step": 29
+    },
+    {
+      "epoch": 0.004860661049902786,
+      "grad_norm": 5.86301851272583,
+      "learning_rate": 2.697841726618705e-07,
+      "loss": 0.4106,
+      "step": 30
+    },
+    {
+      "epoch": 0.005022683084899547,
+      "grad_norm": 5.353550910949707,
+      "learning_rate": 2.787769784172662e-07,
+      "loss": 0.3607,
+      "step": 31
+    },
+    {
+      "epoch": 0.005184705119896306,
+      "grad_norm": 4.887706756591797,
+      "learning_rate": 2.877697841726619e-07,
+      "loss": 0.36,
+      "step": 32
+    },
+    {
+      "epoch": 0.005346727154893065,
+      "grad_norm": 4.950960159301758,
+      "learning_rate": 2.9676258992805756e-07,
+      "loss": 0.3418,
+      "step": 33
+    },
+    {
+      "epoch": 0.005508749189889825,
+      "grad_norm": 5.3128581047058105,
+      "learning_rate": 3.057553956834533e-07,
+      "loss": 0.3229,
+      "step": 34
+    },
+    {
+      "epoch": 0.005670771224886585,
+      "grad_norm": 4.85807466506958,
+      "learning_rate": 3.1474820143884896e-07,
+      "loss": 0.2946,
+      "step": 35
+    },
+    {
+      "epoch": 0.005832793259883344,
+      "grad_norm": 4.823163986206055,
+      "learning_rate": 3.237410071942446e-07,
+      "loss": 0.304,
+      "step": 36
+    },
+    {
+      "epoch": 0.0059948152948801035,
+      "grad_norm": 4.9618239402771,
+      "learning_rate": 3.3273381294964035e-07,
+      "loss": 0.3032,
+      "step": 37
+    },
+    {
+      "epoch": 0.006156837329876863,
+      "grad_norm": 4.626596927642822,
+      "learning_rate": 3.41726618705036e-07,
+      "loss": 0.272,
+      "step": 38
+    },
+    {
+      "epoch": 0.006318859364873623,
+      "grad_norm": 3.9511983394622803,
+      "learning_rate": 3.5071942446043163e-07,
+      "loss": 0.2683,
+      "step": 39
+    },
+    {
+      "epoch": 0.0064808813998703824,
+      "grad_norm": 3.7039501667022705,
+      "learning_rate": 3.5971223021582736e-07,
+      "loss": 0.2829,
+      "step": 40
+    },
+    {
+      "epoch": 0.006642903434867142,
+      "grad_norm": 3.3785510063171387,
+      "learning_rate": 3.68705035971223e-07,
+      "loss": 0.2598,
+      "step": 41
+    },
+    {
+      "epoch": 0.006804925469863901,
+      "grad_norm": 3.1743481159210205,
+      "learning_rate": 3.7769784172661875e-07,
+      "loss": 0.2307,
+      "step": 42
+    },
+    {
+      "epoch": 0.006966947504860661,
+      "grad_norm": 3.37454891204834,
+      "learning_rate": 3.8669064748201447e-07,
+      "loss": 0.2391,
+      "step": 43
+    },
+    {
+      "epoch": 0.007128969539857421,
+      "grad_norm": 3.0130465030670166,
+      "learning_rate": 3.956834532374101e-07,
+      "loss": 0.2258,
+      "step": 44
+    },
+    {
+      "epoch": 0.00729099157485418,
+      "grad_norm": 2.937732458114624,
+      "learning_rate": 4.0467625899280576e-07,
+      "loss": 0.22,
+      "step": 45
+    },
+    {
+      "epoch": 0.007453013609850939,
+      "grad_norm": 3.2253363132476807,
+      "learning_rate": 4.136690647482015e-07,
+      "loss": 0.2128,
+      "step": 46
+    },
+    {
+      "epoch": 0.0076150356448477,
+      "grad_norm": 2.944216012954712,
+      "learning_rate": 4.2266187050359715e-07,
+      "loss": 0.2199,
+      "step": 47
+    },
+    {
+      "epoch": 0.007777057679844459,
+      "grad_norm": 2.8392646312713623,
+      "learning_rate": 4.3165467625899287e-07,
+      "loss": 0.2185,
+      "step": 48
+    },
+    {
+      "epoch": 0.007939079714841219,
+      "grad_norm": 2.9585957527160645,
+      "learning_rate": 4.406474820143885e-07,
+      "loss": 0.2416,
+      "step": 49
+    },
+    {
+      "epoch": 0.008101101749837978,
+      "grad_norm": 2.977691173553467,
+      "learning_rate": 4.496402877697842e-07,
+      "loss": 0.2219,
+      "step": 50
+    },
+    {
+      "epoch": 0.008263123784834738,
+      "grad_norm": 2.6376028060913086,
+      "learning_rate": 4.586330935251799e-07,
+      "loss": 0.2271,
+      "step": 51
+    },
+    {
+      "epoch": 0.008425145819831496,
+      "grad_norm": 2.8397135734558105,
+      "learning_rate": 4.676258992805756e-07,
+      "loss": 0.2227,
+      "step": 52
+    },
+    {
+      "epoch": 0.008587167854828257,
+      "grad_norm": 2.7388408184051514,
+      "learning_rate": 4.7661870503597127e-07,
+      "loss": 0.1965,
+      "step": 53
+    },
+    {
+      "epoch": 0.008749189889825017,
+      "grad_norm": 2.7062389850616455,
+      "learning_rate": 4.85611510791367e-07,
+      "loss": 0.1991,
+      "step": 54
+    },
+    {
+      "epoch": 0.008911211924821775,
+      "grad_norm": 2.688417911529541,
+      "learning_rate": 4.946043165467626e-07,
+      "loss": 0.2121,
+      "step": 55
+    },
+    {
+      "epoch": 0.009073233959818535,
+      "grad_norm": 2.418215751647949,
+      "learning_rate": 5.035971223021583e-07,
+      "loss": 0.2104,
+      "step": 56
+    },
+    {
+      "epoch": 0.009235255994815296,
+      "grad_norm": 2.9777352809906006,
+      "learning_rate": 5.12589928057554e-07,
+      "loss": 0.2145,
+      "step": 57
+    },
+    {
+      "epoch": 0.009397278029812054,
+      "grad_norm": 2.8849077224731445,
+      "learning_rate": 5.215827338129497e-07,
+      "loss": 0.1984,
+      "step": 58
+    },
+    {
+      "epoch": 0.009559300064808814,
+      "grad_norm": 2.7535386085510254,
+      "learning_rate": 5.305755395683454e-07,
+      "loss": 0.2159,
+      "step": 59
+    },
+    {
+      "epoch": 0.009721322099805573,
+      "grad_norm": 3.017183780670166,
+      "learning_rate": 5.39568345323741e-07,
+      "loss": 0.2485,
+      "step": 60
+    },
+    {
+      "epoch": 0.009883344134802333,
+      "grad_norm": 3.307050943374634,
+      "learning_rate": 5.485611510791367e-07,
+      "loss": 0.21,
+      "step": 61
+    },
+    {
+      "epoch": 0.010045366169799093,
+      "grad_norm": 2.6310977935791016,
+      "learning_rate": 5.575539568345325e-07,
+      "loss": 0.1888,
+      "step": 62
+    },
+    {
+      "epoch": 0.010207388204795852,
+      "grad_norm": 2.3625879287719727,
+      "learning_rate": 5.665467625899281e-07,
+      "loss": 0.1938,
+      "step": 63
+    },
+    {
+      "epoch": 0.010369410239792612,
+      "grad_norm": 3.2340619564056396,
+      "learning_rate": 5.755395683453238e-07,
+      "loss": 0.1812,
+      "step": 64
+    },
+    {
+      "epoch": 0.010531432274789372,
+      "grad_norm": 2.1760573387145996,
+      "learning_rate": 5.845323741007194e-07,
+      "loss": 0.18,
+      "step": 65
+    },
+    {
+      "epoch": 0.01069345430978613,
+      "grad_norm": 2.495835542678833,
+      "learning_rate": 5.935251798561151e-07,
+      "loss": 0.2233,
+      "step": 66
+    },
+    {
+      "epoch": 0.01085547634478289,
+      "grad_norm": 2.425354242324829,
+      "learning_rate": 6.025179856115109e-07,
+      "loss": 0.2017,
+      "step": 67
+    },
+    {
+      "epoch": 0.01101749837977965,
+      "grad_norm": 2.386017322540283,
+      "learning_rate": 6.115107913669066e-07,
+      "loss": 0.1897,
+      "step": 68
+    },
+    {
+      "epoch": 0.01117952041477641,
+      "grad_norm": 2.476449728012085,
+      "learning_rate": 6.205035971223022e-07,
+      "loss": 0.2163,
+      "step": 69
+    },
+    {
+      "epoch": 0.01134154244977317,
+      "grad_norm": 2.3602700233459473,
+      "learning_rate": 6.294964028776979e-07,
+      "loss": 0.1795,
+      "step": 70
+    },
+    {
+      "epoch": 0.011503564484769928,
+      "grad_norm": 2.5833189487457275,
+      "learning_rate": 6.384892086330936e-07,
+      "loss": 0.215,
+      "step": 71
+    },
+    {
+      "epoch": 0.011665586519766688,
+      "grad_norm": 2.3865418434143066,
+      "learning_rate": 6.474820143884893e-07,
+      "loss": 0.171,
+      "step": 72
+    },
+    {
+      "epoch": 0.011827608554763449,
+      "grad_norm": 2.407026529312134,
+      "learning_rate": 6.564748201438849e-07,
+      "loss": 0.1766,
+      "step": 73
+    },
+    {
+      "epoch": 0.011989630589760207,
+      "grad_norm": 2.3541834354400635,
+      "learning_rate": 6.654676258992807e-07,
+      "loss": 0.1703,
+      "step": 74
+    },
+    {
+      "epoch": 0.012151652624756967,
+      "grad_norm": 2.7565481662750244,
+      "learning_rate": 6.744604316546763e-07,
+      "loss": 0.1833,
+      "step": 75
+    },
+    {
+      "epoch": 0.012313674659753726,
+      "grad_norm": 2.53861665725708,
+      "learning_rate": 6.83453237410072e-07,
+      "loss": 0.1967,
+      "step": 76
+    },
+    {
+      "epoch": 0.012475696694750486,
+      "grad_norm": 2.920691728591919,
+      "learning_rate": 6.924460431654677e-07,
+      "loss": 0.1945,
+      "step": 77
+    },
+    {
+      "epoch": 0.012637718729747246,
+      "grad_norm": 2.714219093322754,
+      "learning_rate": 7.014388489208633e-07,
+      "loss": 0.1961,
+      "step": 78
+    },
+    {
+      "epoch": 0.012799740764744005,
+      "grad_norm": 2.2343974113464355,
+      "learning_rate": 7.104316546762591e-07,
+      "loss": 0.1868,
+      "step": 79
+    },
+    {
+      "epoch": 0.012961762799740765,
+      "grad_norm": 2.333791971206665,
+      "learning_rate": 7.194244604316547e-07,
+      "loss": 0.1705,
+      "step": 80
+    },
+    {
+      "epoch": 0.013123784834737525,
+      "grad_norm": 2.554150104522705,
+      "learning_rate": 7.284172661870504e-07,
+      "loss": 0.1822,
+      "step": 81
+    },
+    {
+      "epoch": 0.013285806869734284,
+      "grad_norm": 2.51990008354187,
+      "learning_rate": 7.37410071942446e-07,
+      "loss": 0.1889,
+      "step": 82
+    },
+    {
+      "epoch": 0.013447828904731044,
+      "grad_norm": 2.2917540073394775,
+      "learning_rate": 7.464028776978418e-07,
+      "loss": 0.1788,
+      "step": 83
+    },
+    {
+      "epoch": 0.013609850939727802,
+      "grad_norm": 2.376124620437622,
+      "learning_rate": 7.553956834532375e-07,
+      "loss": 0.1995,
+      "step": 84
+    },
+    {
+      "epoch": 0.013771872974724562,
+      "grad_norm": 2.521143674850464,
+      "learning_rate": 7.643884892086331e-07,
+      "loss": 0.1691,
+      "step": 85
+    },
+    {
+      "epoch": 0.013933895009721323,
+      "grad_norm": 2.2666220664978027,
+      "learning_rate": 7.733812949640289e-07,
+      "loss": 0.1831,
+      "step": 86
+    },
+    {
+      "epoch": 0.014095917044718081,
+      "grad_norm": 2.2533257007598877,
+      "learning_rate": 7.823741007194246e-07,
+      "loss": 0.1726,
+      "step": 87
+    },
+    {
+      "epoch": 0.014257939079714841,
+      "grad_norm": 2.371137857437134,
+      "learning_rate": 7.913669064748202e-07,
+      "loss": 0.182,
+      "step": 88
+    },
+    {
+      "epoch": 0.014419961114711602,
+      "grad_norm": 2.547879934310913,
+      "learning_rate": 8.003597122302159e-07,
+      "loss": 0.1908,
+      "step": 89
+    },
+    {
+      "epoch": 0.01458198314970836,
+      "grad_norm": 2.433431625366211,
+      "learning_rate": 8.093525179856115e-07,
+      "loss": 0.1951,
+      "step": 90
+    },
+    {
+      "epoch": 0.01474400518470512,
+      "grad_norm": 2.1704154014587402,
+      "learning_rate": 8.183453237410073e-07,
+      "loss": 0.1545,
+      "step": 91
+    },
+    {
+      "epoch": 0.014906027219701879,
+      "grad_norm": 2.624864339828491,
+      "learning_rate": 8.27338129496403e-07,
+      "loss": 0.1969,
+      "step": 92
+    },
+    {
+      "epoch": 0.015068049254698639,
+      "grad_norm": 2.3828327655792236,
+      "learning_rate": 8.363309352517986e-07,
+      "loss": 0.165,
+      "step": 93
+    },
+    {
+      "epoch": 0.0152300712896954,
+      "grad_norm": 2.242358922958374,
+      "learning_rate": 8.453237410071943e-07,
+      "loss": 0.1761,
+      "step": 94
+    },
+    {
+      "epoch": 0.015392093324692158,
+      "grad_norm": 2.126227378845215,
+      "learning_rate": 8.543165467625899e-07,
+      "loss": 0.1626,
+      "step": 95
+    },
+    {
+      "epoch": 0.015554115359688918,
+      "grad_norm": 2.2046821117401123,
+      "learning_rate": 8.633093525179857e-07,
+      "loss": 0.1867,
+      "step": 96
+    },
+    {
+      "epoch": 0.015716137394685678,
+      "grad_norm": 2.2523837089538574,
+      "learning_rate": 8.723021582733814e-07,
+      "loss": 0.1755,
+      "step": 97
+    },
+    {
+      "epoch": 0.015878159429682438,
+      "grad_norm": 2.326197862625122,
+      "learning_rate": 8.81294964028777e-07,
+      "loss": 0.1811,
+      "step": 98
+    },
+    {
+      "epoch": 0.016040181464679195,
+      "grad_norm": 2.4861881732940674,
+      "learning_rate": 8.902877697841728e-07,
+      "loss": 0.2073,
+      "step": 99
+    },
+    {
+      "epoch": 0.016202203499675955,
+      "grad_norm": 2.3390986919403076,
+      "learning_rate": 8.992805755395684e-07,
+      "loss": 0.1749,
+      "step": 100
+    },
+    {
+      "epoch": 0.016364225534672715,
+      "grad_norm": 2.372002124786377,
+      "learning_rate": 9.082733812949641e-07,
+      "loss": 0.1778,
+      "step": 101
+    },
+    {
+      "epoch": 0.016526247569669476,
+      "grad_norm": 2.4759466648101807,
+      "learning_rate": 9.172661870503598e-07,
+      "loss": 0.1649,
+      "step": 102
+    },
+    {
+      "epoch": 0.016688269604666236,
+      "grad_norm": 2.2869863510131836,
+      "learning_rate": 9.262589928057554e-07,
+      "loss": 0.159,
+      "step": 103
+    },
+    {
+      "epoch": 0.016850291639662993,
+      "grad_norm": 2.3882856369018555,
+      "learning_rate": 9.352517985611512e-07,
+      "loss": 0.1893,
+      "step": 104
+    },
+    {
+      "epoch": 0.017012313674659753,
+      "grad_norm": 2.316922187805176,
+      "learning_rate": 9.442446043165468e-07,
+      "loss": 0.1737,
+      "step": 105
+    },
+    {
+      "epoch": 0.017174335709656513,
+      "grad_norm": 2.4775216579437256,
+      "learning_rate": 9.532374100719425e-07,
+      "loss": 0.204,
+      "step": 106
+    },
+    {
+      "epoch": 0.017336357744653273,
+      "grad_norm": 2.44256329536438,
+      "learning_rate": 9.622302158273383e-07,
+      "loss": 0.203,
+      "step": 107
+    },
+    {
+      "epoch": 0.017498379779650033,
+      "grad_norm": 2.3133935928344727,
+      "learning_rate": 9.71223021582734e-07,
+      "loss": 0.1842,
+      "step": 108
+    },
+    {
+      "epoch": 0.017660401814646794,
+      "grad_norm": 2.498054265975952,
+      "learning_rate": 9.802158273381295e-07,
+      "loss": 0.169,
+      "step": 109
+    },
+    {
+      "epoch": 0.01782242384964355,
+      "grad_norm": 2.51688814163208,
+      "learning_rate": 9.892086330935252e-07,
+      "loss": 0.1845,
+      "step": 110
+    },
+    {
+      "epoch": 0.01798444588464031,
+      "grad_norm": 2.356818437576294,
+      "learning_rate": 9.98201438848921e-07,
+      "loss": 0.1835,
+      "step": 111
+    },
+    {
+      "epoch": 0.01814646791963707,
+      "grad_norm": 2.1603775024414062,
+      "learning_rate": 1.0071942446043167e-06,
+      "loss": 0.155,
+      "step": 112
+    },
+    {
+      "epoch": 0.01830848995463383,
+      "grad_norm": 2.509432554244995,
+      "learning_rate": 1.0161870503597124e-06,
+      "loss": 0.2281,
+      "step": 113
+    },
+    {
+      "epoch": 0.01847051198963059,
+      "grad_norm": 2.5537493228912354,
+      "learning_rate": 1.025179856115108e-06,
+      "loss": 0.1793,
+      "step": 114
+    },
+    {
+      "epoch": 0.018632534024627348,
+      "grad_norm": 2.246959924697876,
+      "learning_rate": 1.0341726618705036e-06,
+      "loss": 0.1801,
+      "step": 115
+    },
+    {
+      "epoch": 0.018794556059624108,
+      "grad_norm": 2.19378662109375,
+      "learning_rate": 1.0431654676258993e-06,
+      "loss": 0.1637,
+      "step": 116
+    },
+    {
+      "epoch": 0.01895657809462087,
+      "grad_norm": 2.3158321380615234,
+      "learning_rate": 1.052158273381295e-06,
+      "loss": 0.1617,
+      "step": 117
+    },
+    {
+      "epoch": 0.01911860012961763,
+      "grad_norm": 2.4354169368743896,
+      "learning_rate": 1.0611510791366908e-06,
+      "loss": 0.1779,
+      "step": 118
+    },
+    {
+      "epoch": 0.01928062216461439,
+      "grad_norm": 2.5059328079223633,
+      "learning_rate": 1.0701438848920865e-06,
+      "loss": 0.1832,
+      "step": 119
+    },
+    {
+      "epoch": 0.019442644199611146,
+      "grad_norm": 2.449944019317627,
+      "learning_rate": 1.079136690647482e-06,
+      "loss": 0.2071,
+      "step": 120
+    },
+    {
+      "epoch": 0.019604666234607906,
+      "grad_norm": 2.3459267616271973,
+      "learning_rate": 1.0881294964028777e-06,
+      "loss": 0.1757,
+      "step": 121
+    },
+    {
+      "epoch": 0.019766688269604666,
+      "grad_norm": 2.456517457962036,
+      "learning_rate": 1.0971223021582735e-06,
+      "loss": 0.1804,
+      "step": 122
+    },
+    {
+      "epoch": 0.019928710304601426,
+      "grad_norm": 2.350886344909668,
+      "learning_rate": 1.1061151079136692e-06,
+      "loss": 0.1675,
+      "step": 123
+    },
+    {
+      "epoch": 0.020090732339598186,
+      "grad_norm": 2.2044262886047363,
+      "learning_rate": 1.115107913669065e-06,
+      "loss": 0.1666,
+      "step": 124
+    },
+    {
+      "epoch": 0.020252754374594947,
+      "grad_norm": 2.2976362705230713,
+      "learning_rate": 1.1241007194244604e-06,
+      "loss": 0.1797,
+      "step": 125
+    },
+    {
+      "epoch": 0.020414776409591703,
+      "grad_norm": 2.1294548511505127,
+      "learning_rate": 1.1330935251798561e-06,
+      "loss": 0.1529,
+      "step": 126
+    },
+    {
+      "epoch": 0.020576798444588464,
+      "grad_norm": 2.205601453781128,
+      "learning_rate": 1.1420863309352519e-06,
+      "loss": 0.1641,
+      "step": 127
+    },
+    {
+      "epoch": 0.020738820479585224,
+      "grad_norm": 2.2161824703216553,
+      "learning_rate": 1.1510791366906476e-06,
+      "loss": 0.1617,
+      "step": 128
+    },
+    {
+      "epoch": 0.020900842514581984,
+      "grad_norm": 2.266568422317505,
+      "learning_rate": 1.1600719424460433e-06,
+      "loss": 0.1746,
+      "step": 129
+    },
+    {
+      "epoch": 0.021062864549578744,
+      "grad_norm": 2.3664584159851074,
+      "learning_rate": 1.1690647482014388e-06,
+      "loss": 0.1777,
+      "step": 130
+    },
+    {
+      "epoch": 0.0212248865845755,
+      "grad_norm": 2.2521674633026123,
+      "learning_rate": 1.1780575539568347e-06,
+      "loss": 0.1707,
+      "step": 131
+    },
+    {
+      "epoch": 0.02138690861957226,
+      "grad_norm": 2.394850254058838,
+      "learning_rate": 1.1870503597122303e-06,
+      "loss": 0.1856,
+      "step": 132
+    },
+    {
+      "epoch": 0.02154893065456902,
+      "grad_norm": 2.332915782928467,
+      "learning_rate": 1.196043165467626e-06,
+      "loss": 0.1818,
+      "step": 133
+    },
+    {
+      "epoch": 0.02171095268956578,
+      "grad_norm": 2.307866334915161,
+      "learning_rate": 1.2050359712230217e-06,
+      "loss": 0.1799,
+      "step": 134
+    },
+    {
+      "epoch": 0.021872974724562542,
+      "grad_norm": 2.2809898853302,
+      "learning_rate": 1.2140287769784172e-06,
+      "loss": 0.1762,
+      "step": 135
+    },
+    {
+      "epoch": 0.0220349967595593,
+      "grad_norm": 2.1759355068206787,
+      "learning_rate": 1.2230215827338131e-06,
+      "loss": 0.1522,
+      "step": 136
+    },
+    {
+      "epoch": 0.02219701879455606,
+      "grad_norm": 2.172473907470703,
+      "learning_rate": 1.2320143884892087e-06,
+      "loss": 0.1693,
+      "step": 137
+    },
+    {
+      "epoch": 0.02235904082955282,
+      "grad_norm": 2.1808273792266846,
+      "learning_rate": 1.2410071942446044e-06,
+      "loss": 0.1549,
+      "step": 138
+    },
+    {
+      "epoch": 0.02252106286454958,
+      "grad_norm": 2.164473533630371,
+      "learning_rate": 1.25e-06,
+      "loss": 0.1798,
+      "step": 139
+    },
+    {
+      "epoch": 0.02268308489954634,
+      "grad_norm": 2.2718453407287598,
+      "learning_rate": 1.2589928057553958e-06,
+      "loss": 0.1606,
+      "step": 140
+    },
+    {
+      "epoch": 0.022845106934543096,
+      "grad_norm": 2.1400561332702637,
+      "learning_rate": 1.2679856115107913e-06,
+      "loss": 0.1428,
+      "step": 141
+    },
+    {
+      "epoch": 0.023007128969539856,
+      "grad_norm": 2.193063497543335,
+      "learning_rate": 1.2769784172661873e-06,
+      "loss": 0.1729,
+      "step": 142
+    },
+    {
+      "epoch": 0.023169151004536617,
+      "grad_norm": 2.2718231678009033,
+      "learning_rate": 1.285971223021583e-06,
+      "loss": 0.1803,
+      "step": 143
+    },
+    {
+      "epoch": 0.023331173039533377,
+      "grad_norm": 2.3830406665802,
+      "learning_rate": 1.2949640287769785e-06,
+      "loss": 0.1688,
+      "step": 144
+    },
+    {
+      "epoch": 0.023493195074530137,
+      "grad_norm": 2.2217133045196533,
+      "learning_rate": 1.3039568345323742e-06,
+      "loss": 0.172,
+      "step": 145
+    },
+    {
+      "epoch": 0.023655217109526897,
+      "grad_norm": 2.429953098297119,
+      "learning_rate": 1.3129496402877697e-06,
+      "loss": 0.1761,
+      "step": 146
+    },
+    {
+      "epoch": 0.023817239144523654,
+      "grad_norm": 2.481562614440918,
+      "learning_rate": 1.3219424460431657e-06,
+      "loss": 0.1777,
+      "step": 147
+    },
+    {
+      "epoch": 0.023979261179520414,
+      "grad_norm": 2.3909571170806885,
+      "learning_rate": 1.3309352517985614e-06,
+      "loss": 0.1814,
+      "step": 148
+    },
+    {
+      "epoch": 0.024141283214517174,
+      "grad_norm": 2.3645970821380615,
+      "learning_rate": 1.339928057553957e-06,
+      "loss": 0.179,
+      "step": 149
+    },
+    {
+      "epoch": 0.024303305249513935,
+      "grad_norm": 2.5163588523864746,
+      "learning_rate": 1.3489208633093526e-06,
+      "loss": 0.2024,
+      "step": 150
+    },
+    {
+      "epoch": 0.024465327284510695,
+      "grad_norm": 2.2522170543670654,
+      "learning_rate": 1.3579136690647481e-06,
+      "loss": 0.1844,
+      "step": 151
+    },
+    {
+      "epoch": 0.02462734931950745,
+      "grad_norm": 2.276763677597046,
+      "learning_rate": 1.366906474820144e-06,
+      "loss": 0.1934,
+      "step": 152
+    },
+    {
+      "epoch": 0.02478937135450421,
+      "grad_norm": 2.1778862476348877,
+      "learning_rate": 1.3758992805755398e-06,
+      "loss": 0.166,
+      "step": 153
+    },
+    {
+      "epoch": 0.024951393389500972,
+      "grad_norm": 2.162675619125366,
+      "learning_rate": 1.3848920863309353e-06,
+      "loss": 0.1823,
+      "step": 154
+    },
+    {
+      "epoch": 0.025113415424497732,
+      "grad_norm": 2.212240695953369,
+      "learning_rate": 1.393884892086331e-06,
+      "loss": 0.1814,
+      "step": 155
+    },
+    {
+      "epoch": 0.025275437459494492,
+      "grad_norm": 2.283370018005371,
+      "learning_rate": 1.4028776978417265e-06,
+      "loss": 0.1821,
+      "step": 156
+    },
+    {
+      "epoch": 0.02543745949449125,
+      "grad_norm": 2.1461737155914307,
+      "learning_rate": 1.4118705035971225e-06,
+      "loss": 0.1689,
+      "step": 157
+    },
+    {
+      "epoch": 0.02559948152948801,
+      "grad_norm": 2.1343040466308594,
+      "learning_rate": 1.4208633093525182e-06,
+      "loss": 0.1754,
+      "step": 158
+    },
+    {
+      "epoch": 0.02576150356448477,
+      "grad_norm": 2.200503349304199,
+      "learning_rate": 1.4298561151079137e-06,
+      "loss": 0.1824,
+      "step": 159
+    },
+    {
+      "epoch": 0.02592352559948153,
+      "grad_norm": 2.3421995639801025,
+      "learning_rate": 1.4388489208633094e-06,
+      "loss": 0.1768,
+      "step": 160
+    },
+    {
+      "epoch": 0.02608554763447829,
+      "grad_norm": 2.1087827682495117,
+      "learning_rate": 1.447841726618705e-06,
+      "loss": 0.1929,
+      "step": 161
+    },
+    {
+      "epoch": 0.02624756966947505,
+      "grad_norm": 2.4311277866363525,
+      "learning_rate": 1.4568345323741009e-06,
+      "loss": 0.1845,
+      "step": 162
+    },
+    {
+      "epoch": 0.026409591704471807,
+      "grad_norm": 2.0372049808502197,
+      "learning_rate": 1.4658273381294966e-06,
+      "loss": 0.163,
+      "step": 163
+    },
+    {
+      "epoch": 0.026571613739468567,
+      "grad_norm": 2.1894679069519043,
+      "learning_rate": 1.474820143884892e-06,
+      "loss": 0.1511,
+      "step": 164
+    },
+    {
+      "epoch": 0.026733635774465327,
+      "grad_norm": 2.3050854206085205,
+      "learning_rate": 1.4838129496402878e-06,
+      "loss": 0.1867,
+      "step": 165
+    },
+    {
+      "epoch": 0.026895657809462088,
+      "grad_norm": 2.4033379554748535,
+      "learning_rate": 1.4928057553956835e-06,
+      "loss": 0.1739,
+      "step": 166
+    },
+    {
+      "epoch": 0.027057679844458848,
+      "grad_norm": 2.3603515625,
+      "learning_rate": 1.5017985611510793e-06,
+      "loss": 0.162,
+      "step": 167
+    },
+    {
+      "epoch": 0.027219701879455604,
+      "grad_norm": 1.9225950241088867,
+      "learning_rate": 1.510791366906475e-06,
+      "loss": 0.1444,
+      "step": 168
+    },
+    {
+      "epoch": 0.027381723914452365,
+      "grad_norm": 2.3583033084869385,
+      "learning_rate": 1.5197841726618707e-06,
+      "loss": 0.1942,
+      "step": 169
+    },
+    {
+      "epoch": 0.027543745949449125,
+      "grad_norm": 2.416661262512207,
+      "learning_rate": 1.5287769784172662e-06,
+      "loss": 0.1778,
+      "step": 170
+    },
+    {
+      "epoch": 0.027705767984445885,
+      "grad_norm": 2.0838210582733154,
+      "learning_rate": 1.537769784172662e-06,
+      "loss": 0.1581,
+      "step": 171
+    },
+    {
+      "epoch": 0.027867790019442645,
+      "grad_norm": 2.3810360431671143,
+      "learning_rate": 1.5467625899280579e-06,
+      "loss": 0.174,
+      "step": 172
+    },
+    {
+      "epoch": 0.028029812054439402,
+      "grad_norm": 2.243232011795044,
+      "learning_rate": 1.5557553956834534e-06,
+      "loss": 0.1563,
+      "step": 173
+    },
+    {
+      "epoch": 0.028191834089436162,
+      "grad_norm": 2.2563958168029785,
+      "learning_rate": 1.5647482014388491e-06,
+      "loss": 0.1719,
+      "step": 174
+    },
+    {
+      "epoch": 0.028353856124432922,
+      "grad_norm": 2.283281087875366,
+      "learning_rate": 1.5737410071942446e-06,
+      "loss": 0.1607,
+      "step": 175
+    },
+    {
+      "epoch": 0.028515878159429683,
+      "grad_norm": 2.166625499725342,
+      "learning_rate": 1.5827338129496403e-06,
+      "loss": 0.1615,
+      "step": 176
+    },
+    {
+      "epoch": 0.028677900194426443,
+      "grad_norm": 2.2912237644195557,
+      "learning_rate": 1.5917266187050363e-06,
+      "loss": 0.1731,
+      "step": 177
+    },
+    {
+      "epoch": 0.028839922229423203,
+      "grad_norm": 2.196793794631958,
+      "learning_rate": 1.6007194244604318e-06,
+      "loss": 0.1644,
+      "step": 178
+    },
+    {
+      "epoch": 0.02900194426441996,
+      "grad_norm": 2.193145513534546,
+      "learning_rate": 1.6097122302158275e-06,
+      "loss": 0.1632,
+      "step": 179
+    },
+    {
+      "epoch": 0.02916396629941672,
+      "grad_norm": 2.420989990234375,
+      "learning_rate": 1.618705035971223e-06,
+      "loss": 0.1953,
+      "step": 180
+    },
+    {
+      "epoch": 0.02932598833441348,
+      "grad_norm": 2.0741348266601562,
+      "learning_rate": 1.6276978417266187e-06,
+      "loss": 0.1573,
+      "step": 181
+    },
+    {
+      "epoch": 0.02948801036941024,
+      "grad_norm": 2.2281997203826904,
+      "learning_rate": 1.6366906474820147e-06,
+      "loss": 0.1897,
+      "step": 182
+    },
+    {
+      "epoch": 0.029650032404407,
+      "grad_norm": 2.163466453552246,
+      "learning_rate": 1.6456834532374102e-06,
+      "loss": 0.1742,
+      "step": 183
+    },
+    {
+      "epoch": 0.029812054439403757,
+      "grad_norm": 2.054077625274658,
+      "learning_rate": 1.654676258992806e-06,
+      "loss": 0.1741,
+      "step": 184
+    },
+    {
+      "epoch": 0.029974076474400518,
+      "grad_norm": 2.1084864139556885,
+      "learning_rate": 1.6636690647482014e-06,
+      "loss": 0.1887,
+      "step": 185
+    },
+    {
+      "epoch": 0.030136098509397278,
+      "grad_norm": 2.160194158554077,
+      "learning_rate": 1.6726618705035971e-06,
+      "loss": 0.162,
+      "step": 186
+    },
+    {
+      "epoch": 0.030298120544394038,
+      "grad_norm": 2.227296829223633,
+      "learning_rate": 1.681654676258993e-06,
+      "loss": 0.1717,
+      "step": 187
+    },
+    {
+      "epoch": 0.0304601425793908,
+      "grad_norm": 2.0688178539276123,
+      "learning_rate": 1.6906474820143886e-06,
+      "loss": 0.1705,
+      "step": 188
+    },
+    {
+      "epoch": 0.030622164614387555,
+      "grad_norm": 2.081521511077881,
+      "learning_rate": 1.6996402877697843e-06,
+      "loss": 0.1759,
+      "step": 189
+    },
+    {
+      "epoch": 0.030784186649384315,
+      "grad_norm": 2.2949635982513428,
+      "learning_rate": 1.7086330935251798e-06,
+      "loss": 0.1693,
+      "step": 190
+    },
+    {
+      "epoch": 0.030946208684381075,
+      "grad_norm": 2.3229682445526123,
+      "learning_rate": 1.7176258992805755e-06,
+      "loss": 0.186,
+      "step": 191
+    },
+    {
+      "epoch": 0.031108230719377836,
+      "grad_norm": 2.3723998069763184,
+      "learning_rate": 1.7266187050359715e-06,
+      "loss": 0.1902,
+      "step": 192
+    },
+    {
+      "epoch": 0.031270252754374596,
+      "grad_norm": 2.150329828262329,
+      "learning_rate": 1.735611510791367e-06,
+      "loss": 0.1723,
+      "step": 193
+    },
+    {
+      "epoch": 0.031432274789371356,
+      "grad_norm": 2.097501277923584,
+      "learning_rate": 1.7446043165467627e-06,
+      "loss": 0.165,
+      "step": 194
+    },
+    {
+      "epoch": 0.031594296824368116,
+      "grad_norm": 2.10372257232666,
+      "learning_rate": 1.7535971223021584e-06,
+      "loss": 0.1578,
+      "step": 195
+    },
+    {
+      "epoch": 0.031756318859364877,
+      "grad_norm": 2.1476476192474365,
+      "learning_rate": 1.762589928057554e-06,
+      "loss": 0.1662,
+      "step": 196
+    },
+    {
+      "epoch": 0.03191834089436163,
+      "grad_norm": 2.1310970783233643,
+      "learning_rate": 1.7715827338129499e-06,
+      "loss": 0.1655,
+      "step": 197
+    },
+    {
+      "epoch": 0.03208036292935839,
+      "grad_norm": 2.2113521099090576,
+      "learning_rate": 1.7805755395683456e-06,
+      "loss": 0.1758,
+      "step": 198
+    },
+    {
+      "epoch": 0.03224238496435515,
+      "grad_norm": 2.0751709938049316,
+      "learning_rate": 1.7895683453237411e-06,
+      "loss": 0.1585,
+      "step": 199
+    },
+    {
+      "epoch": 0.03240440699935191,
+      "grad_norm": 1.9745714664459229,
+      "learning_rate": 1.7985611510791368e-06,
+      "loss": 0.1666,
+      "step": 200
+    },
+    {
+      "epoch": 0.03256642903434867,
+      "grad_norm": 2.1305060386657715,
+      "learning_rate": 1.8075539568345323e-06,
+      "loss": 0.1843,
+      "step": 201
+    },
+    {
+      "epoch": 0.03272845106934543,
+      "grad_norm": 2.0196950435638428,
+      "learning_rate": 1.8165467625899283e-06,
+      "loss": 0.1768,
+      "step": 202
+    },
+    {
+      "epoch": 0.03289047310434219,
+      "grad_norm": 2.2363412380218506,
+      "learning_rate": 1.825539568345324e-06,
+      "loss": 0.169,
+      "step": 203
+    },
+    {
+      "epoch": 0.03305249513933895,
+      "grad_norm": 2.118345260620117,
+      "learning_rate": 1.8345323741007195e-06,
+      "loss": 0.1486,
+      "step": 204
+    },
+    {
+      "epoch": 0.03321451717433571,
+      "grad_norm": 2.0574629306793213,
+      "learning_rate": 1.8435251798561152e-06,
+      "loss": 0.1794,
+      "step": 205
+    },
+    {
+      "epoch": 0.03337653920933247,
+      "grad_norm": 1.975506067276001,
+      "learning_rate": 1.8525179856115107e-06,
+      "loss": 0.1608,
+      "step": 206
+    },
+    {
+      "epoch": 0.03353856124432923,
+      "grad_norm": 2.1187586784362793,
+      "learning_rate": 1.8615107913669067e-06,
+      "loss": 0.1514,
+      "step": 207
+    },
+    {
+      "epoch": 0.033700583279325985,
+      "grad_norm": 2.031742811203003,
+      "learning_rate": 1.8705035971223024e-06,
+      "loss": 0.1567,
+      "step": 208
+    },
+    {
+      "epoch": 0.033862605314322745,
+      "grad_norm": 2.198672294616699,
+      "learning_rate": 1.879496402877698e-06,
+      "loss": 0.1709,
+      "step": 209
+    },
+    {
+      "epoch": 0.034024627349319506,
+      "grad_norm": 1.9215576648712158,
+      "learning_rate": 1.8884892086330936e-06,
+      "loss": 0.1466,
+      "step": 210
+    },
+    {
+      "epoch": 0.034186649384316266,
+      "grad_norm": 2.0358407497406006,
+      "learning_rate": 1.8974820143884896e-06,
+      "loss": 0.1641,
+      "step": 211
+    },
+    {
+      "epoch": 0.034348671419313026,
+      "grad_norm": 2.2609496116638184,
+      "learning_rate": 1.906474820143885e-06,
+      "loss": 0.1607,
+      "step": 212
+    },
+    {
+      "epoch": 0.034510693454309786,
+      "grad_norm": 2.0340802669525146,
+      "learning_rate": 1.915467625899281e-06,
+      "loss": 0.1605,
+      "step": 213
+    },
+    {
+      "epoch": 0.034672715489306546,
+      "grad_norm": 2.190073013305664,
+      "learning_rate": 1.9244604316546765e-06,
+      "loss": 0.1893,
+      "step": 214
+    },
+    {
+      "epoch": 0.03483473752430331,
+      "grad_norm": 2.129892587661743,
+      "learning_rate": 1.933453237410072e-06,
+      "loss": 0.1657,
+      "step": 215
+    },
+    {
+      "epoch": 0.03499675955930007,
+      "grad_norm": 1.9545798301696777,
+      "learning_rate": 1.942446043165468e-06,
+      "loss": 0.1698,
+      "step": 216
+    },
+    {
+      "epoch": 0.03515878159429683,
+      "grad_norm": 1.9465980529785156,
+      "learning_rate": 1.9514388489208637e-06,
+      "loss": 0.1571,
+      "step": 217
+    },
+    {
+      "epoch": 0.03532080362929359,
+      "grad_norm": 2.116178512573242,
+      "learning_rate": 1.960431654676259e-06,
+      "loss": 0.1581,
+      "step": 218
+    },
+    {
+      "epoch": 0.03548282566429034,
+      "grad_norm": 2.0396151542663574,
+      "learning_rate": 1.9694244604316547e-06,
+      "loss": 0.1637,
+      "step": 219
+    },
+    {
+      "epoch": 0.0356448476992871,
+      "grad_norm": 2.1016218662261963,
+      "learning_rate": 1.9784172661870504e-06,
+      "loss": 0.1707,
+      "step": 220
+    },
+    {
+      "epoch": 0.03580686973428386,
+      "grad_norm": 2.122370719909668,
+      "learning_rate": 1.987410071942446e-06,
+      "loss": 0.1701,
+      "step": 221
+    },
+    {
+      "epoch": 0.03596889176928062,
+      "grad_norm": 2.2020812034606934,
+      "learning_rate": 1.996402877697842e-06,
+      "loss": 0.1755,
+      "step": 222
+    },
+    {
+      "epoch": 0.03613091380427738,
+      "grad_norm": 2.160726547241211,
+      "learning_rate": 2.0053956834532376e-06,
+      "loss": 0.1704,
+      "step": 223
+    },
+    {
+      "epoch": 0.03629293583927414,
+      "grad_norm": 2.121961832046509,
+      "learning_rate": 2.0143884892086333e-06,
+      "loss": 0.1631,
+      "step": 224
+    },
+    {
+      "epoch": 0.0364549578742709,
+      "grad_norm": 2.045151472091675,
+      "learning_rate": 2.023381294964029e-06,
+      "loss": 0.1794,
+      "step": 225
+    },
+    {
+      "epoch": 0.03661697990926766,
+      "grad_norm": 1.982482671737671,
+      "learning_rate": 2.0323741007194248e-06,
+      "loss": 0.1649,
+      "step": 226
+    },
+    {
+      "epoch": 0.03677900194426442,
+      "grad_norm": 2.0936903953552246,
+      "learning_rate": 2.0413669064748205e-06,
+      "loss": 0.1894,
+      "step": 227
+    },
+    {
+      "epoch": 0.03694102397926118,
+      "grad_norm": 1.9307271242141724,
+      "learning_rate": 2.050359712230216e-06,
+      "loss": 0.1658,
+      "step": 228
+    },
+    {
+      "epoch": 0.037103046014257936,
+      "grad_norm": 2.0920655727386475,
+      "learning_rate": 2.0593525179856115e-06,
+      "loss": 0.1823,
+      "step": 229
+    },
+    {
+      "epoch": 0.037265068049254696,
+      "grad_norm": 2.037468194961548,
+      "learning_rate": 2.0683453237410072e-06,
+      "loss": 0.1854,
+      "step": 230
+    },
+    {
+      "epoch": 0.037427090084251456,
+      "grad_norm": 2.0791573524475098,
+      "learning_rate": 2.0773381294964034e-06,
+      "loss": 0.1668,
+      "step": 231
+    },
+    {
+      "epoch": 0.037589112119248216,
+      "grad_norm": 2.020608425140381,
+      "learning_rate": 2.0863309352517987e-06,
+      "loss": 0.1664,
+      "step": 232
+    },
+    {
+      "epoch": 0.03775113415424498,
+      "grad_norm": 2.0509209632873535,
+      "learning_rate": 2.0953237410071944e-06,
+      "loss": 0.1655,
+      "step": 233
+    },
+    {
+      "epoch": 0.03791315618924174,
+      "grad_norm": 1.895331621170044,
+      "learning_rate": 2.10431654676259e-06,
+      "loss": 0.1671,
+      "step": 234
+    },
+    {
+      "epoch": 0.0380751782242385,
+      "grad_norm": 1.8676968812942505,
+      "learning_rate": 2.113309352517986e-06,
+      "loss": 0.1499,
+      "step": 235
+    },
+    {
+      "epoch": 0.03823720025923526,
+      "grad_norm": 1.9968868494033813,
+      "learning_rate": 2.1223021582733816e-06,
+      "loss": 0.1704,
+      "step": 236
+    },
+    {
+      "epoch": 0.03839922229423202,
+      "grad_norm": 2.107897996902466,
+      "learning_rate": 2.1312949640287773e-06,
+      "loss": 0.1738,
+      "step": 237
+    },
+    {
+      "epoch": 0.03856124432922878,
+      "grad_norm": 2.0445058345794678,
+      "learning_rate": 2.140287769784173e-06,
+      "loss": 0.1809,
+      "step": 238
+    },
+    {
+      "epoch": 0.03872326636422554,
+      "grad_norm": 2.110241413116455,
+      "learning_rate": 2.1492805755395683e-06,
+      "loss": 0.1731,
+      "step": 239
+    },
+    {
+      "epoch": 0.03888528839922229,
+      "grad_norm": 2.0265743732452393,
+      "learning_rate": 2.158273381294964e-06,
+      "loss": 0.1581,
+      "step": 240
+    },
+    {
+      "epoch": 0.03904731043421905,
+      "grad_norm": 1.8976444005966187,
+      "learning_rate": 2.16726618705036e-06,
+      "loss": 0.1622,
+      "step": 241
+    },
+    {
+      "epoch": 0.03920933246921581,
+      "grad_norm": 1.967790126800537,
+      "learning_rate": 2.1762589928057555e-06,
+      "loss": 0.1639,
+      "step": 242
+    },
+    {
+      "epoch": 0.03937135450421257,
+      "grad_norm": 2.0550568103790283,
+      "learning_rate": 2.185251798561151e-06,
+      "loss": 0.1786,
+      "step": 243
+    },
+    {
+      "epoch": 0.03953337653920933,
+      "grad_norm": 1.9519095420837402,
+      "learning_rate": 2.194244604316547e-06,
+      "loss": 0.1772,
+      "step": 244
+    },
+    {
+      "epoch": 0.03969539857420609,
+      "grad_norm": 2.0229897499084473,
+      "learning_rate": 2.2032374100719426e-06,
+      "loss": 0.1746,
+      "step": 245
+    },
+    {
+      "epoch": 0.03985742060920285,
+      "grad_norm": 2.133192777633667,
+      "learning_rate": 2.2122302158273384e-06,
+      "loss": 0.172,
+      "step": 246
+    },
+    {
+      "epoch": 0.04001944264419961,
+      "grad_norm": 2.323986053466797,
+      "learning_rate": 2.221223021582734e-06,
+      "loss": 0.1845,
+      "step": 247
+    },
+    {
+      "epoch": 0.04018146467919637,
+      "grad_norm": 2.0216121673583984,
+      "learning_rate": 2.23021582733813e-06,
+      "loss": 0.1844,
+      "step": 248
+    },
+    {
+      "epoch": 0.04034348671419313,
+      "grad_norm": 1.8960579633712769,
+      "learning_rate": 2.2392086330935255e-06,
+      "loss": 0.1468,
+      "step": 249
+    },
+    {
+      "epoch": 0.04050550874918989,
+      "grad_norm": 1.971975564956665,
+      "learning_rate": 2.248201438848921e-06,
+      "loss": 0.1822,
+      "step": 250
+    },
+    {
+      "epoch": 0.040667530784186647,
+      "grad_norm": 1.8761088848114014,
+      "learning_rate": 2.257194244604317e-06,
+      "loss": 0.1613,
+      "step": 251
+    },
+    {
+      "epoch": 0.04082955281918341,
+      "grad_norm": 1.8940060138702393,
+      "learning_rate": 2.2661870503597123e-06,
+      "loss": 0.1827,
+      "step": 252
+    },
+    {
+      "epoch": 0.04099157485418017,
+      "grad_norm": 1.880953311920166,
+      "learning_rate": 2.275179856115108e-06,
+      "loss": 0.1738,
+      "step": 253
+    },
+    {
+      "epoch": 0.04115359688917693,
+      "grad_norm": 2.1408498287200928,
+      "learning_rate": 2.2841726618705037e-06,
+      "loss": 0.2066,
+      "step": 254
+    },
+    {
+      "epoch": 0.04131561892417369,
+      "grad_norm": 2.4467737674713135,
+      "learning_rate": 2.2931654676258994e-06,
+      "loss": 0.1646,
+      "step": 255
+    },
+    {
+      "epoch": 0.04147764095917045,
+      "grad_norm": 1.8072450160980225,
+      "learning_rate": 2.302158273381295e-06,
+      "loss": 0.1475,
+      "step": 256
+    },
+    {
+      "epoch": 0.04163966299416721,
+      "grad_norm": 2.035788059234619,
+      "learning_rate": 2.311151079136691e-06,
+      "loss": 0.1742,
+      "step": 257
+    },
+    {
+      "epoch": 0.04180168502916397,
+      "grad_norm": 2.0450069904327393,
+      "learning_rate": 2.3201438848920866e-06,
+      "loss": 0.1733,
+      "step": 258
+    },
+    {
+      "epoch": 0.04196370706416073,
+      "grad_norm": 2.0818724632263184,
+      "learning_rate": 2.3291366906474823e-06,
+      "loss": 0.1743,
+      "step": 259
+    },
+    {
+      "epoch": 0.04212572909915749,
+      "grad_norm": 2.11169695854187,
+      "learning_rate": 2.3381294964028776e-06,
+      "loss": 0.1632,
+      "step": 260
+    },
+    {
+      "epoch": 0.04228775113415424,
+      "grad_norm": 2.079584836959839,
+      "learning_rate": 2.3471223021582738e-06,
+      "loss": 0.1868,
+      "step": 261
+    },
+    {
+      "epoch": 0.042449773169151,
+      "grad_norm": 2.0339434146881104,
+      "learning_rate": 2.3561151079136695e-06,
+      "loss": 0.1536,
+      "step": 262
+    },
+    {
+      "epoch": 0.04261179520414776,
+      "grad_norm": 2.0371828079223633,
+      "learning_rate": 2.365107913669065e-06,
+      "loss": 0.174,
+      "step": 263
+    },
+    {
+      "epoch": 0.04277381723914452,
+      "grad_norm": 2.0146055221557617,
+      "learning_rate": 2.3741007194244605e-06,
+      "loss": 0.1599,
+      "step": 264
+    },
+    {
+      "epoch": 0.04293583927414128,
+      "grad_norm": 2.0458879470825195,
+      "learning_rate": 2.3830935251798562e-06,
+      "loss": 0.1722,
+      "step": 265
+    },
+    {
+      "epoch": 0.04309786130913804,
+      "grad_norm": 1.9001797437667847,
+      "learning_rate": 2.392086330935252e-06,
+      "loss": 0.1519,
+      "step": 266
+    },
+    {
+      "epoch": 0.0432598833441348,
+      "grad_norm": 1.9941308498382568,
+      "learning_rate": 2.4010791366906477e-06,
+      "loss": 0.1796,
+      "step": 267
+    },
+    {
+      "epoch": 0.04342190537913156,
+      "grad_norm": 2.0200061798095703,
+      "learning_rate": 2.4100719424460434e-06,
+      "loss": 0.1746,
+      "step": 268
+    },
+    {
+      "epoch": 0.04358392741412832,
+      "grad_norm": 2.166887044906616,
+      "learning_rate": 2.419064748201439e-06,
+      "loss": 0.1712,
+      "step": 269
+    },
+    {
+      "epoch": 0.043745949449125084,
+      "grad_norm": 2.011035680770874,
+      "learning_rate": 2.4280575539568344e-06,
+      "loss": 0.1726,
+      "step": 270
+    },
+    {
+      "epoch": 0.043907971484121844,
+      "grad_norm": 2.070662021636963,
+      "learning_rate": 2.4370503597122306e-06,
+      "loss": 0.1877,
+      "step": 271
+    },
+    {
+      "epoch": 0.0440699935191186,
+      "grad_norm": 1.8755451440811157,
+      "learning_rate": 2.4460431654676263e-06,
+      "loss": 0.153,
+      "step": 272
+    },
+    {
+      "epoch": 0.04423201555411536,
+      "grad_norm": 2.0150034427642822,
+      "learning_rate": 2.4550359712230216e-06,
+      "loss": 0.1857,
+      "step": 273
+    },
+    {
+      "epoch": 0.04439403758911212,
+      "grad_norm": 1.969179391860962,
+      "learning_rate": 2.4640287769784173e-06,
+      "loss": 0.1758,
+      "step": 274
+    },
+    {
+      "epoch": 0.04455605962410888,
+      "grad_norm": 1.8629865646362305,
+      "learning_rate": 2.473021582733813e-06,
+      "loss": 0.1643,
+      "step": 275
+    },
+    {
+      "epoch": 0.04471808165910564,
+      "grad_norm": 1.8298242092132568,
+      "learning_rate": 2.4820143884892088e-06,
+      "loss": 0.1666,
+      "step": 276
+    },
+    {
+      "epoch": 0.0448801036941024,
+      "grad_norm": 1.870679259300232,
+      "learning_rate": 2.4910071942446045e-06,
+      "loss": 0.1613,
+      "step": 277
+    },
+    {
+      "epoch": 0.04504212572909916,
+      "grad_norm": 2.0010008811950684,
+      "learning_rate": 2.5e-06,
+      "loss": 0.171,
+      "step": 278
+    },
+    {
+      "epoch": 0.04520414776409592,
+      "grad_norm": 2.0288240909576416,
+      "learning_rate": 2.508992805755396e-06,
+      "loss": 0.1759,
+      "step": 279
+    },
+    {
+      "epoch": 0.04536616979909268,
+      "grad_norm": 1.911431908607483,
+      "learning_rate": 2.5179856115107916e-06,
+      "loss": 0.1709,
+      "step": 280
+    },
+    {
+      "epoch": 0.04552819183408944,
+      "grad_norm": 2.0871028900146484,
+      "learning_rate": 2.526978417266187e-06,
+      "loss": 0.1817,
+      "step": 281
+    },
+    {
+      "epoch": 0.04569021386908619,
+      "grad_norm": 1.9303252696990967,
+      "learning_rate": 2.5359712230215827e-06,
+      "loss": 0.189,
+      "step": 282
+    },
+    {
+      "epoch": 0.04585223590408295,
+      "grad_norm": 2.0121030807495117,
+      "learning_rate": 2.544964028776979e-06,
+      "loss": 0.1956,
+      "step": 283
+    },
+    {
+      "epoch": 0.04601425793907971,
+      "grad_norm": 1.8766732215881348,
+      "learning_rate": 2.5539568345323745e-06,
+      "loss": 0.1731,
+      "step": 284
+    },
+    {
+      "epoch": 0.04617627997407647,
+      "grad_norm": 1.7872287034988403,
+      "learning_rate": 2.5629496402877703e-06,
+      "loss": 0.1672,
+      "step": 285
+    },
+    {
+      "epoch": 0.04633830200907323,
+      "grad_norm": 2.2016873359680176,
+      "learning_rate": 2.571942446043166e-06,
+      "loss": 0.1942,
+      "step": 286
+    },
+    {
+      "epoch": 0.04650032404406999,
+      "grad_norm": 2.001664400100708,
+      "learning_rate": 2.5809352517985613e-06,
+      "loss": 0.1779,
+      "step": 287
+    },
+    {
+      "epoch": 0.046662346079066754,
+      "grad_norm": 1.881262183189392,
+      "learning_rate": 2.589928057553957e-06,
+      "loss": 0.1621,
+      "step": 288
+    },
+    {
+      "epoch": 0.046824368114063514,
+      "grad_norm": 2.1064798831939697,
+      "learning_rate": 2.5989208633093527e-06,
+      "loss": 0.1788,
+      "step": 289
+    },
+    {
+      "epoch": 0.046986390149060274,
+      "grad_norm": 1.8403074741363525,
+      "learning_rate": 2.6079136690647484e-06,
+      "loss": 0.1627,
+      "step": 290
+    },
+    {
+      "epoch": 0.047148412184057034,
+      "grad_norm": 2.0622360706329346,
+      "learning_rate": 2.6169064748201437e-06,
+      "loss": 0.1742,
+      "step": 291
+    },
+    {
+      "epoch": 0.047310434219053794,
+      "grad_norm": 1.878718376159668,
+      "learning_rate": 2.6258992805755395e-06,
+      "loss": 0.1509,
+      "step": 292
+    },
+    {
+      "epoch": 0.04747245625405055,
+      "grad_norm": 2.0211925506591797,
+      "learning_rate": 2.6348920863309356e-06,
+      "loss": 0.1923,
+      "step": 293
+    },
+    {
+      "epoch": 0.04763447828904731,
+      "grad_norm": 2.1039490699768066,
+      "learning_rate": 2.6438848920863313e-06,
+      "loss": 0.1733,
+      "step": 294
+    },
+    {
+      "epoch": 0.04779650032404407,
+      "grad_norm": 2.026010751724243,
+      "learning_rate": 2.652877697841727e-06,
+      "loss": 0.1759,
+      "step": 295
+    },
+    {
+      "epoch": 0.04795852235904083,
+      "grad_norm": 2.0125303268432617,
+      "learning_rate": 2.6618705035971228e-06,
+      "loss": 0.1844,
+      "step": 296
+    },
+    {
+      "epoch": 0.04812054439403759,
+      "grad_norm": 1.8597307205200195,
+      "learning_rate": 2.670863309352518e-06,
+      "loss": 0.1637,
+      "step": 297
+    },
+    {
+      "epoch": 0.04828256642903435,
+      "grad_norm": 1.8322033882141113,
+      "learning_rate": 2.679856115107914e-06,
+      "loss": 0.1638,
+      "step": 298
+    },
+    {
+      "epoch": 0.04844458846403111,
+      "grad_norm": 1.9159802198410034,
+      "learning_rate": 2.6888489208633095e-06,
+      "loss": 0.1668,
+      "step": 299
+    },
+    {
+      "epoch": 0.04860661049902787,
+      "grad_norm": 1.653609037399292,
+      "learning_rate": 2.6978417266187052e-06,
+      "loss": 0.1523,
+      "step": 300
+    },
+    {
+      "epoch": 0.04876863253402463,
+      "grad_norm": 1.8605982065200806,
+      "learning_rate": 2.706834532374101e-06,
+      "loss": 0.1641,
+      "step": 301
+    },
+    {
+      "epoch": 0.04893065456902139,
+      "grad_norm": 1.9447520971298218,
+      "learning_rate": 2.7158273381294963e-06,
+      "loss": 0.1795,
+      "step": 302
+    },
+    {
+      "epoch": 0.04909267660401815,
+      "grad_norm": 2.1139204502105713,
+      "learning_rate": 2.7248201438848924e-06,
+      "loss": 0.1839,
+      "step": 303
+    },
+    {
+      "epoch": 0.0492546986390149,
+      "grad_norm": 2.0103182792663574,
+      "learning_rate": 2.733812949640288e-06,
+      "loss": 0.1707,
+      "step": 304
+    },
+    {
+      "epoch": 0.04941672067401166,
+      "grad_norm": 1.9638346433639526,
+      "learning_rate": 2.742805755395684e-06,
+      "loss": 0.1548,
+      "step": 305
+    },
+    {
+      "epoch": 0.04957874270900842,
+      "grad_norm": 2.0801048278808594,
+      "learning_rate": 2.7517985611510796e-06,
+      "loss": 0.1813,
+      "step": 306
+    },
+    {
+      "epoch": 0.049740764744005184,
+      "grad_norm": 1.848359227180481,
+      "learning_rate": 2.760791366906475e-06,
+      "loss": 0.1655,
+      "step": 307
+    },
+    {
+      "epoch": 0.049902786779001944,
+      "grad_norm": 1.85602605342865,
+      "learning_rate": 2.7697841726618706e-06,
+      "loss": 0.1789,
+      "step": 308
+    },
+    {
+      "epoch": 0.050064808813998704,
+      "grad_norm": 1.8783005475997925,
+      "learning_rate": 2.7787769784172663e-06,
+      "loss": 0.1753,
+      "step": 309
+    },
+    {
+      "epoch": 0.050226830848995464,
+      "grad_norm": 1.981858253479004,
+      "learning_rate": 2.787769784172662e-06,
+      "loss": 0.1798,
+      "step": 310
+    },
+    {
+      "epoch": 0.050388852883992225,
+      "grad_norm": 2.0939548015594482,
+      "learning_rate": 2.7967625899280578e-06,
+      "loss": 0.1923,
+      "step": 311
+    },
+    {
+      "epoch": 0.050550874918988985,
+      "grad_norm": 2.0538747310638428,
+      "learning_rate": 2.805755395683453e-06,
+      "loss": 0.1869,
+      "step": 312
+    },
+    {
+      "epoch": 0.050712896953985745,
+      "grad_norm": 1.807354211807251,
+      "learning_rate": 2.8147482014388492e-06,
+      "loss": 0.1719,
+      "step": 313
+    },
+    {
+      "epoch": 0.0508749189889825,
+      "grad_norm": 2.057577610015869,
+      "learning_rate": 2.823741007194245e-06,
+      "loss": 0.1939,
+      "step": 314
+    },
+    {
+      "epoch": 0.05103694102397926,
+      "grad_norm": 1.821679949760437,
+      "learning_rate": 2.8327338129496407e-06,
+      "loss": 0.1564,
+      "step": 315
+    },
+    {
+      "epoch": 0.05119896305897602,
+      "grad_norm": 1.8081183433532715,
+      "learning_rate": 2.8417266187050364e-06,
+      "loss": 0.1717,
+      "step": 316
+    },
+    {
+      "epoch": 0.05136098509397278,
+      "grad_norm": 1.8699404001235962,
+      "learning_rate": 2.850719424460432e-06,
+      "loss": 0.173,
+      "step": 317
+    },
+    {
+      "epoch": 0.05152300712896954,
+      "grad_norm": 2.0023691654205322,
+      "learning_rate": 2.8597122302158274e-06,
+      "loss": 0.1764,
+      "step": 318
+    },
+    {
+      "epoch": 0.0516850291639663,
+      "grad_norm": 2.0363411903381348,
+      "learning_rate": 2.868705035971223e-06,
+      "loss": 0.1744,
+      "step": 319
+    },
+    {
+      "epoch": 0.05184705119896306,
+      "grad_norm": 1.8756717443466187,
+      "learning_rate": 2.877697841726619e-06,
+      "loss": 0.176,
+      "step": 320
+    },
+    {
+      "epoch": 0.05200907323395982,
+      "grad_norm": 2.1065685749053955,
+      "learning_rate": 2.8866906474820146e-06,
+      "loss": 0.1713,
+      "step": 321
+    },
+    {
+      "epoch": 0.05217109526895658,
+      "grad_norm": 1.929309606552124,
+      "learning_rate": 2.89568345323741e-06,
+      "loss": 0.1705,
+      "step": 322
+    },
+    {
+      "epoch": 0.05233311730395334,
+      "grad_norm": 1.8109393119812012,
+      "learning_rate": 2.9046762589928064e-06,
+      "loss": 0.1771,
+      "step": 323
+    },
+    {
+      "epoch": 0.0524951393389501,
+      "grad_norm": 1.9056837558746338,
+      "learning_rate": 2.9136690647482017e-06,
+      "loss": 0.1714,
+      "step": 324
+    },
+    {
+      "epoch": 0.052657161373946854,
+      "grad_norm": 1.8842450380325317,
+      "learning_rate": 2.9226618705035975e-06,
+      "loss": 0.1938,
+      "step": 325
+    },
+    {
+      "epoch": 0.052819183408943614,
+      "grad_norm": 1.9185986518859863,
+      "learning_rate": 2.931654676258993e-06,
+      "loss": 0.1771,
+      "step": 326
+    },
+    {
+      "epoch": 0.052981205443940374,
+      "grad_norm": 1.8425815105438232,
+      "learning_rate": 2.940647482014389e-06,
+      "loss": 0.19,
+      "step": 327
+    },
+    {
+      "epoch": 0.053143227478937134,
+      "grad_norm": 1.9350950717926025,
+      "learning_rate": 2.949640287769784e-06,
+      "loss": 0.187,
+      "step": 328
+    },
+    {
+      "epoch": 0.053305249513933894,
+      "grad_norm": 1.8640246391296387,
+      "learning_rate": 2.95863309352518e-06,
+      "loss": 0.1677,
+      "step": 329
+    },
+    {
+      "epoch": 0.053467271548930655,
+      "grad_norm": 1.8789522647857666,
+      "learning_rate": 2.9676258992805756e-06,
+      "loss": 0.1823,
+      "step": 330
+    },
+    {
+      "epoch": 0.053629293583927415,
+      "grad_norm": 1.8093454837799072,
+      "learning_rate": 2.9766187050359714e-06,
+      "loss": 0.1587,
+      "step": 331
+    },
+    {
+      "epoch": 0.053791315618924175,
+      "grad_norm": 1.8029776811599731,
+      "learning_rate": 2.985611510791367e-06,
+      "loss": 0.1723,
+      "step": 332
+    },
+    {
+      "epoch": 0.053953337653920935,
+      "grad_norm": 1.8550293445587158,
+      "learning_rate": 2.9946043165467632e-06,
+      "loss": 0.1948,
+      "step": 333
+    },
+    {
+      "epoch": 0.054115359688917695,
+      "grad_norm": 1.8107119798660278,
+      "learning_rate": 3.0035971223021585e-06,
+      "loss": 0.172,
+      "step": 334
+    },
+    {
+      "epoch": 0.054277381723914456,
+      "grad_norm": 1.7598873376846313,
+      "learning_rate": 3.0125899280575543e-06,
+      "loss": 0.1709,
+      "step": 335
+    },
+    {
+      "epoch": 0.05443940375891121,
+      "grad_norm": 1.8725013732910156,
+      "learning_rate": 3.02158273381295e-06,
+      "loss": 0.2019,
+      "step": 336
+    },
+    {
+      "epoch": 0.05460142579390797,
+      "grad_norm": 1.7645277976989746,
+      "learning_rate": 3.0305755395683457e-06,
+      "loss": 0.1677,
+      "step": 337
+    },
+    {
+      "epoch": 0.05476344782890473,
+      "grad_norm": 1.922553300857544,
+      "learning_rate": 3.0395683453237414e-06,
+      "loss": 0.1788,
+      "step": 338
+    },
+    {
+      "epoch": 0.05492546986390149,
+      "grad_norm": 1.7409249544143677,
+      "learning_rate": 3.0485611510791367e-06,
+      "loss": 0.1629,
+      "step": 339
+    },
+    {
+      "epoch": 0.05508749189889825,
+      "grad_norm": 1.914994478225708,
+      "learning_rate": 3.0575539568345324e-06,
+      "loss": 0.1757,
+      "step": 340
+    },
+    {
+      "epoch": 0.05524951393389501,
+      "grad_norm": 1.8456170558929443,
+      "learning_rate": 3.066546762589928e-06,
+      "loss": 0.1906,
+      "step": 341
+    },
+    {
+      "epoch": 0.05541153596889177,
+      "grad_norm": 1.949918270111084,
+      "learning_rate": 3.075539568345324e-06,
+      "loss": 0.2056,
+      "step": 342
+    },
+    {
+      "epoch": 0.05557355800388853,
+      "grad_norm": 1.8012248277664185,
+      "learning_rate": 3.08453237410072e-06,
+      "loss": 0.1719,
+      "step": 343
+    },
+    {
+      "epoch": 0.05573558003888529,
+      "grad_norm": 1.942472219467163,
+      "learning_rate": 3.0935251798561158e-06,
+      "loss": 0.1834,
+      "step": 344
+    },
+    {
+      "epoch": 0.05589760207388205,
+      "grad_norm": 1.8514517545700073,
+      "learning_rate": 3.102517985611511e-06,
+      "loss": 0.1911,
+      "step": 345
+    },
+    {
+      "epoch": 0.056059624108878804,
+      "grad_norm": 1.7124977111816406,
+      "learning_rate": 3.1115107913669068e-06,
+      "loss": 0.1656,
+      "step": 346
+    },
+    {
+      "epoch": 0.056221646143875564,
+      "grad_norm": 1.6776280403137207,
+      "learning_rate": 3.1205035971223025e-06,
+      "loss": 0.1741,
+      "step": 347
+    },
+    {
+      "epoch": 0.056383668178872325,
+      "grad_norm": 2.2318480014801025,
+      "learning_rate": 3.1294964028776982e-06,
+      "loss": 0.219,
+      "step": 348
+    },
+    {
+      "epoch": 0.056545690213869085,
+      "grad_norm": 1.7956312894821167,
+      "learning_rate": 3.1384892086330935e-06,
+      "loss": 0.1731,
+      "step": 349
+    },
+    {
+      "epoch": 0.056707712248865845,
+      "grad_norm": 1.813264012336731,
+      "learning_rate": 3.1474820143884892e-06,
+      "loss": 0.1798,
+      "step": 350
+    },
+    {
+      "epoch": 0.056869734283862605,
+      "grad_norm": 1.6817779541015625,
+      "learning_rate": 3.156474820143885e-06,
+      "loss": 0.1615,
+      "step": 351
+    },
+    {
+      "epoch": 0.057031756318859365,
+      "grad_norm": 1.8656824827194214,
+      "learning_rate": 3.1654676258992807e-06,
+      "loss": 0.1682,
+      "step": 352
+    },
+    {
+      "epoch": 0.057193778353856126,
+      "grad_norm": 1.817460536956787,
+      "learning_rate": 3.174460431654677e-06,
+      "loss": 0.1608,
+      "step": 353
+    },
+    {
+      "epoch": 0.057355800388852886,
+      "grad_norm": 1.9450856447219849,
+      "learning_rate": 3.1834532374100726e-06,
+      "loss": 0.1892,
+      "step": 354
+    },
+    {
+      "epoch": 0.057517822423849646,
+      "grad_norm": 1.7994742393493652,
+      "learning_rate": 3.192446043165468e-06,
+      "loss": 0.1634,
+      "step": 355
+    },
+    {
+      "epoch": 0.057679844458846406,
+      "grad_norm": 1.8007848262786865,
+      "learning_rate": 3.2014388489208636e-06,
+      "loss": 0.1731,
+      "step": 356
+    },
+    {
+      "epoch": 0.05784186649384316,
+      "grad_norm": 1.920350432395935,
+      "learning_rate": 3.2104316546762593e-06,
+      "loss": 0.174,
+      "step": 357
+    },
+    {
+      "epoch": 0.05800388852883992,
+      "grad_norm": 1.810903549194336,
+      "learning_rate": 3.219424460431655e-06,
+      "loss": 0.1828,
+      "step": 358
+    },
+    {
+      "epoch": 0.05816591056383668,
+      "grad_norm": 1.9278273582458496,
+      "learning_rate": 3.2284172661870507e-06,
+      "loss": 0.1881,
+      "step": 359
+    },
+    {
+      "epoch": 0.05832793259883344,
+      "grad_norm": 1.7868634462356567,
+      "learning_rate": 3.237410071942446e-06,
+      "loss": 0.1719,
+      "step": 360
+    },
+    {
+      "epoch": 0.0584899546338302,
+      "grad_norm": 1.6021710634231567,
+      "learning_rate": 3.2464028776978418e-06,
+      "loss": 0.1478,
+      "step": 361
+    },
+    {
+      "epoch": 0.05865197666882696,
+      "grad_norm": 1.6602813005447388,
+      "learning_rate": 3.2553956834532375e-06,
+      "loss": 0.1527,
+      "step": 362
+    },
+    {
+      "epoch": 0.05881399870382372,
+      "grad_norm": 1.8697694540023804,
+      "learning_rate": 3.2643884892086336e-06,
+      "loss": 0.1809,
+      "step": 363
+    },
+    {
+      "epoch": 0.05897602073882048,
+      "grad_norm": 1.9650862216949463,
+      "learning_rate": 3.2733812949640294e-06,
+      "loss": 0.1869,
+      "step": 364
+    },
+    {
+      "epoch": 0.05913804277381724,
+      "grad_norm": 1.7278015613555908,
+      "learning_rate": 3.2823741007194247e-06,
+      "loss": 0.1734,
+      "step": 365
+    },
+    {
+      "epoch": 0.059300064808814,
+      "grad_norm": 1.995110273361206,
+      "learning_rate": 3.2913669064748204e-06,
+      "loss": 0.1962,
+      "step": 366
+    },
+    {
+      "epoch": 0.05946208684381076,
+      "grad_norm": 1.8341810703277588,
+      "learning_rate": 3.300359712230216e-06,
+      "loss": 0.1917,
+      "step": 367
+    },
+    {
+      "epoch": 0.059624108878807515,
+      "grad_norm": 1.8125361204147339,
+      "learning_rate": 3.309352517985612e-06,
+      "loss": 0.1794,
+      "step": 368
+    },
+    {
+      "epoch": 0.059786130913804275,
+      "grad_norm": 1.6951332092285156,
+      "learning_rate": 3.3183453237410075e-06,
+      "loss": 0.1623,
+      "step": 369
+    },
+    {
+      "epoch": 0.059948152948801035,
+      "grad_norm": 1.753805160522461,
+      "learning_rate": 3.327338129496403e-06,
+      "loss": 0.1757,
+      "step": 370
+    },
+    {
+      "epoch": 0.060110174983797796,
+      "grad_norm": 1.876604676246643,
+      "learning_rate": 3.3363309352517986e-06,
+      "loss": 0.2007,
+      "step": 371
+    },
+    {
+      "epoch": 0.060272197018794556,
+      "grad_norm": 1.8833791017532349,
+      "learning_rate": 3.3453237410071943e-06,
+      "loss": 0.19,
+      "step": 372
+    },
+    {
+      "epoch": 0.060434219053791316,
+      "grad_norm": 1.769794225692749,
+      "learning_rate": 3.3543165467625904e-06,
+      "loss": 0.1679,
+      "step": 373
+    },
+    {
+      "epoch": 0.060596241088788076,
+      "grad_norm": 1.955523133277893,
+      "learning_rate": 3.363309352517986e-06,
+      "loss": 0.1937,
+      "step": 374
+    },
+    {
+      "epoch": 0.060758263123784836,
+      "grad_norm": 1.8921654224395752,
+      "learning_rate": 3.372302158273382e-06,
+      "loss": 0.1607,
+      "step": 375
+    },
+    {
+      "epoch": 0.0609202851587816,
+      "grad_norm": 1.7015023231506348,
+      "learning_rate": 3.381294964028777e-06,
+      "loss": 0.1699,
+      "step": 376
+    },
+    {
+      "epoch": 0.06108230719377836,
+      "grad_norm": 1.8324452638626099,
+      "learning_rate": 3.390287769784173e-06,
+      "loss": 0.1949,
+      "step": 377
+    },
+    {
+      "epoch": 0.06124432922877511,
+      "grad_norm": 1.611916422843933,
+      "learning_rate": 3.3992805755395686e-06,
+      "loss": 0.1381,
+      "step": 378
+    },
+    {
+      "epoch": 0.06140635126377187,
+      "grad_norm": 1.8987873792648315,
+      "learning_rate": 3.4082733812949643e-06,
+      "loss": 0.1813,
+      "step": 379
+    },
+    {
+      "epoch": 0.06156837329876863,
+      "grad_norm": 1.7634623050689697,
+      "learning_rate": 3.4172661870503596e-06,
+      "loss": 0.1739,
+      "step": 380
+    },
+    {
+      "epoch": 0.06173039533376539,
+      "grad_norm": 1.7649253606796265,
+      "learning_rate": 3.4262589928057554e-06,
+      "loss": 0.1654,
+      "step": 381
+    },
+    {
+      "epoch": 0.06189241736876215,
+      "grad_norm": 1.7400226593017578,
+      "learning_rate": 3.435251798561151e-06,
+      "loss": 0.1646,
+      "step": 382
+    },
+    {
+      "epoch": 0.06205443940375891,
+      "grad_norm": 1.8828028440475464,
+      "learning_rate": 3.4442446043165472e-06,
+      "loss": 0.1785,
+      "step": 383
+    },
+    {
+      "epoch": 0.06221646143875567,
+      "grad_norm": 1.9339855909347534,
+      "learning_rate": 3.453237410071943e-06,
+      "loss": 0.1811,
+      "step": 384
+    },
+    {
+      "epoch": 0.06237848347375243,
+      "grad_norm": 1.6471163034439087,
+      "learning_rate": 3.4622302158273387e-06,
+      "loss": 0.1637,
+      "step": 385
+    },
+    {
+      "epoch": 0.06254050550874919,
+      "grad_norm": 1.9495569467544556,
+      "learning_rate": 3.471223021582734e-06,
+      "loss": 0.1963,
+      "step": 386
+    },
+    {
+      "epoch": 0.06270252754374595,
+      "grad_norm": 1.7704451084136963,
+      "learning_rate": 3.4802158273381297e-06,
+      "loss": 0.1774,
+      "step": 387
+    },
+    {
+      "epoch": 0.06286454957874271,
+      "grad_norm": 1.7431086301803589,
+      "learning_rate": 3.4892086330935254e-06,
+      "loss": 0.1647,
+      "step": 388
+    },
+    {
+      "epoch": 0.06302657161373947,
+      "grad_norm": 2.1561217308044434,
+      "learning_rate": 3.498201438848921e-06,
+      "loss": 0.1727,
+      "step": 389
+    },
+    {
+      "epoch": 0.06318859364873623,
+      "grad_norm": 1.6407549381256104,
+      "learning_rate": 3.507194244604317e-06,
+      "loss": 0.1625,
+      "step": 390
+    },
+    {
+      "epoch": 0.06335061568373299,
+      "grad_norm": 1.7267616987228394,
+      "learning_rate": 3.516187050359712e-06,
+      "loss": 0.1668,
+      "step": 391
+    },
+    {
+      "epoch": 0.06351263771872975,
+      "grad_norm": 1.9636503458023071,
+      "learning_rate": 3.525179856115108e-06,
+      "loss": 0.2013,
+      "step": 392
+    },
+    {
+      "epoch": 0.0636746597537265,
+      "grad_norm": 1.6342748403549194,
+      "learning_rate": 3.534172661870504e-06,
+      "loss": 0.1827,
+      "step": 393
+    },
+    {
+      "epoch": 0.06383668178872326,
+      "grad_norm": 1.8311208486557007,
+      "learning_rate": 3.5431654676258998e-06,
+      "loss": 0.1811,
+      "step": 394
+    },
+    {
+      "epoch": 0.06399870382372003,
+      "grad_norm": 1.904611587524414,
+      "learning_rate": 3.5521582733812955e-06,
+      "loss": 0.2114,
+      "step": 395
+    },
+    {
+      "epoch": 0.06416072585871678,
+      "grad_norm": 1.7322005033493042,
+      "learning_rate": 3.561151079136691e-06,
+      "loss": 0.1734,
+      "step": 396
+    },
+    {
+      "epoch": 0.06432274789371355,
+      "grad_norm": 1.7229621410369873,
+      "learning_rate": 3.5701438848920865e-06,
+      "loss": 0.1606,
+      "step": 397
+    },
+    {
+      "epoch": 0.0644847699287103,
+      "grad_norm": 1.8113038539886475,
+      "learning_rate": 3.5791366906474822e-06,
+      "loss": 0.191,
+      "step": 398
+    },
+    {
+      "epoch": 0.06464679196370707,
+      "grad_norm": 1.7715518474578857,
+      "learning_rate": 3.588129496402878e-06,
+      "loss": 0.1914,
+      "step": 399
+    },
+    {
+      "epoch": 0.06480881399870382,
+      "grad_norm": 1.8682818412780762,
+      "learning_rate": 3.5971223021582737e-06,
+      "loss": 0.1771,
+      "step": 400
+    },
+    {
+      "epoch": 0.06497083603370059,
+      "grad_norm": 1.6986286640167236,
+      "learning_rate": 3.606115107913669e-06,
+      "loss": 0.1693,
+      "step": 401
+    },
+    {
+      "epoch": 0.06513285806869734,
+      "grad_norm": 1.7627536058425903,
+      "learning_rate": 3.6151079136690647e-06,
+      "loss": 0.1719,
+      "step": 402
+    },
+    {
+      "epoch": 0.06529488010369411,
+      "grad_norm": 1.707191824913025,
+      "learning_rate": 3.624100719424461e-06,
+      "loss": 0.1713,
+      "step": 403
+    },
+    {
+      "epoch": 0.06545690213869086,
+      "grad_norm": 14.93448543548584,
+      "learning_rate": 3.6330935251798566e-06,
+      "loss": 0.1574,
+      "step": 404
+    },
+    {
+      "epoch": 0.06561892417368761,
+      "grad_norm": 2.0957789421081543,
+      "learning_rate": 3.6420863309352523e-06,
+      "loss": 0.2031,
+      "step": 405
+    },
+    {
+      "epoch": 0.06578094620868438,
+      "grad_norm": 1.8755544424057007,
+      "learning_rate": 3.651079136690648e-06,
+      "loss": 0.1766,
+      "step": 406
+    },
+    {
+      "epoch": 0.06594296824368114,
+      "grad_norm": 1.6951358318328857,
+      "learning_rate": 3.6600719424460433e-06,
+      "loss": 0.1615,
+      "step": 407
+    },
+    {
+      "epoch": 0.0661049902786779,
+      "grad_norm": 1.621053695678711,
+      "learning_rate": 3.669064748201439e-06,
+      "loss": 0.1667,
+      "step": 408
+    },
+    {
+      "epoch": 0.06626701231367466,
+      "grad_norm": 1.7325719594955444,
+      "learning_rate": 3.6780575539568347e-06,
+      "loss": 0.1811,
+      "step": 409
+    },
+    {
+      "epoch": 0.06642903434867142,
+      "grad_norm": 1.7083348035812378,
+      "learning_rate": 3.6870503597122305e-06,
+      "loss": 0.1689,
+      "step": 410
+    },
+    {
+      "epoch": 0.06659105638366818,
+      "grad_norm": 1.5969374179840088,
+      "learning_rate": 3.696043165467626e-06,
+      "loss": 0.1649,
+      "step": 411
+    },
+    {
+      "epoch": 0.06675307841866494,
+      "grad_norm": 1.688307285308838,
+      "learning_rate": 3.7050359712230215e-06,
+      "loss": 0.1681,
+      "step": 412
+    },
+    {
+      "epoch": 0.0669151004536617,
+      "grad_norm": 1.976590871810913,
+      "learning_rate": 3.7140287769784176e-06,
+      "loss": 0.2102,
+      "step": 413
+    },
+    {
+      "epoch": 0.06707712248865846,
+      "grad_norm": 1.8404573202133179,
+      "learning_rate": 3.7230215827338134e-06,
+      "loss": 0.1617,
+      "step": 414
+    },
+    {
+      "epoch": 0.06723914452365522,
+      "grad_norm": 1.8087718486785889,
+      "learning_rate": 3.732014388489209e-06,
+      "loss": 0.1652,
+      "step": 415
+    },
+    {
+      "epoch": 0.06740116655865197,
+      "grad_norm": 1.8954182863235474,
+      "learning_rate": 3.741007194244605e-06,
+      "loss": 0.1803,
+      "step": 416
+    },
+    {
+      "epoch": 0.06756318859364874,
+      "grad_norm": 1.7707210779190063,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.1833,
+      "step": 417
+    },
+    {
+      "epoch": 0.06772521062864549,
+      "grad_norm": 1.5861374139785767,
+      "learning_rate": 3.758992805755396e-06,
+      "loss": 0.1485,
+      "step": 418
+    },
+    {
+      "epoch": 0.06788723266364226,
+      "grad_norm": 1.7552915811538696,
+      "learning_rate": 3.7679856115107915e-06,
+      "loss": 0.1783,
+      "step": 419
+    },
+    {
+      "epoch": 0.06804925469863901,
+      "grad_norm": 1.6234557628631592,
+      "learning_rate": 3.7769784172661873e-06,
+      "loss": 0.1765,
+      "step": 420
+    },
+    {
+      "epoch": 0.06821127673363578,
+      "grad_norm": 1.9008007049560547,
+      "learning_rate": 3.785971223021583e-06,
+      "loss": 0.1839,
+      "step": 421
+    },
+    {
+      "epoch": 0.06837329876863253,
+      "grad_norm": 1.8210786581039429,
+      "learning_rate": 3.794964028776979e-06,
+      "loss": 0.1912,
+      "step": 422
+    },
+    {
+      "epoch": 0.0685353208036293,
+      "grad_norm": 1.6907942295074463,
+      "learning_rate": 3.8039568345323744e-06,
+      "loss": 0.1651,
+      "step": 423
+    },
+    {
+      "epoch": 0.06869734283862605,
+      "grad_norm": 1.691689372062683,
+      "learning_rate": 3.81294964028777e-06,
+      "loss": 0.176,
+      "step": 424
+    },
+    {
+      "epoch": 0.06885936487362282,
+      "grad_norm": 1.6513557434082031,
+      "learning_rate": 3.821942446043166e-06,
+      "loss": 0.173,
+      "step": 425
+    },
+    {
+      "epoch": 0.06902138690861957,
+      "grad_norm": 1.5421315431594849,
+      "learning_rate": 3.830935251798562e-06,
+      "loss": 0.1508,
+      "step": 426
+    },
+    {
+      "epoch": 0.06918340894361633,
+      "grad_norm": 1.7500141859054565,
+      "learning_rate": 3.839928057553957e-06,
+      "loss": 0.195,
+      "step": 427
+    },
+    {
+      "epoch": 0.06934543097861309,
+      "grad_norm": 1.5468772649765015,
+      "learning_rate": 3.848920863309353e-06,
+      "loss": 0.142,
+      "step": 428
+    },
+    {
+      "epoch": 0.06950745301360985,
+      "grad_norm": 1.614610195159912,
+      "learning_rate": 3.857913669064748e-06,
+      "loss": 0.1648,
+      "step": 429
+    },
+    {
+      "epoch": 0.06966947504860661,
+      "grad_norm": 1.802154541015625,
+      "learning_rate": 3.866906474820144e-06,
+      "loss": 0.185,
+      "step": 430
+    },
+    {
+      "epoch": 0.06983149708360337,
+      "grad_norm": 1.7224853038787842,
+      "learning_rate": 3.87589928057554e-06,
+      "loss": 0.1658,
+      "step": 431
+    },
+    {
+      "epoch": 0.06999351911860013,
+      "grad_norm": 1.7054455280303955,
+      "learning_rate": 3.884892086330936e-06,
+      "loss": 0.166,
+      "step": 432
+    },
+    {
+      "epoch": 0.07015554115359689,
+      "grad_norm": 1.712424397468567,
+      "learning_rate": 3.893884892086331e-06,
+      "loss": 0.1694,
+      "step": 433
+    },
+    {
+      "epoch": 0.07031756318859365,
+      "grad_norm": 1.892596960067749,
+      "learning_rate": 3.902877697841727e-06,
+      "loss": 0.1748,
+      "step": 434
+    },
+    {
+      "epoch": 0.07047958522359041,
+      "grad_norm": 1.7858827114105225,
+      "learning_rate": 3.911870503597123e-06,
+      "loss": 0.1781,
+      "step": 435
+    },
+    {
+      "epoch": 0.07064160725858717,
+      "grad_norm": 1.9324816465377808,
+      "learning_rate": 3.920863309352518e-06,
+      "loss": 0.1967,
+      "step": 436
+    },
+    {
+      "epoch": 0.07080362929358393,
+      "grad_norm": 1.74705970287323,
+      "learning_rate": 3.929856115107914e-06,
+      "loss": 0.1813,
+      "step": 437
+    },
+    {
+      "epoch": 0.07096565132858068,
+      "grad_norm": 1.7450978755950928,
+      "learning_rate": 3.938848920863309e-06,
+      "loss": 0.181,
+      "step": 438
+    },
+    {
+      "epoch": 0.07112767336357745,
+      "grad_norm": 1.758083701133728,
+      "learning_rate": 3.9478417266187056e-06,
+      "loss": 0.1903,
+      "step": 439
+    },
+    {
+      "epoch": 0.0712896953985742,
+      "grad_norm": 1.6639132499694824,
+      "learning_rate": 3.956834532374101e-06,
+      "loss": 0.1826,
+      "step": 440
+    },
+    {
+      "epoch": 0.07145171743357097,
+      "grad_norm": 1.6857917308807373,
+      "learning_rate": 3.965827338129496e-06,
+      "loss": 0.1705,
+      "step": 441
+    },
+    {
+      "epoch": 0.07161373946856772,
+      "grad_norm": 1.6692376136779785,
+      "learning_rate": 3.974820143884892e-06,
+      "loss": 0.1877,
+      "step": 442
+    },
+    {
+      "epoch": 0.07177576150356449,
+      "grad_norm": 1.5959900617599487,
+      "learning_rate": 3.9838129496402885e-06,
+      "loss": 0.1659,
+      "step": 443
+    },
+    {
+      "epoch": 0.07193778353856124,
+      "grad_norm": 1.768189549446106,
+      "learning_rate": 3.992805755395684e-06,
+      "loss": 0.1933,
+      "step": 444
+    },
+    {
+      "epoch": 0.07209980557355801,
+      "grad_norm": 1.590326189994812,
+      "learning_rate": 4.00179856115108e-06,
+      "loss": 0.1662,
+      "step": 445
+    },
+    {
+      "epoch": 0.07226182760855476,
+      "grad_norm": 1.768247365951538,
+      "learning_rate": 4.010791366906475e-06,
+      "loss": 0.1964,
+      "step": 446
+    },
+    {
+      "epoch": 0.07242384964355152,
+      "grad_norm": 1.639091968536377,
+      "learning_rate": 4.0197841726618705e-06,
+      "loss": 0.1858,
+      "step": 447
+    },
+    {
+      "epoch": 0.07258587167854828,
+      "grad_norm": 1.7374097108840942,
+      "learning_rate": 4.028776978417267e-06,
+      "loss": 0.1829,
+      "step": 448
+    },
+    {
+      "epoch": 0.07274789371354504,
+      "grad_norm": 1.716292381286621,
+      "learning_rate": 4.037769784172662e-06,
+      "loss": 0.1936,
+      "step": 449
+    },
+    {
+      "epoch": 0.0729099157485418,
+      "grad_norm": 1.5779844522476196,
+      "learning_rate": 4.046762589928058e-06,
+      "loss": 0.1685,
+      "step": 450
+    },
+    {
+      "epoch": 0.07307193778353856,
+      "grad_norm": 1.6292905807495117,
+      "learning_rate": 4.055755395683453e-06,
+      "loss": 0.1623,
+      "step": 451
+    },
+    {
+      "epoch": 0.07323395981853532,
+      "grad_norm": 1.4410508871078491,
+      "learning_rate": 4.0647482014388495e-06,
+      "loss": 0.1431,
+      "step": 452
+    },
+    {
+      "epoch": 0.07339598185353208,
+      "grad_norm": 1.6054296493530273,
+      "learning_rate": 4.073741007194245e-06,
+      "loss": 0.1682,
+      "step": 453
+    },
+    {
+      "epoch": 0.07355800388852884,
+      "grad_norm": 1.6179120540618896,
+      "learning_rate": 4.082733812949641e-06,
+      "loss": 0.1837,
+      "step": 454
+    },
+    {
+      "epoch": 0.0737200259235256,
+      "grad_norm": 1.683802604675293,
+      "learning_rate": 4.091726618705036e-06,
+      "loss": 0.1766,
+      "step": 455
+    },
+    {
+      "epoch": 0.07388204795852236,
+      "grad_norm": 1.5937223434448242,
+      "learning_rate": 4.100719424460432e-06,
+      "loss": 0.1603,
+      "step": 456
+    },
+    {
+      "epoch": 0.07404406999351912,
+      "grad_norm": 1.7861487865447998,
+      "learning_rate": 4.109712230215828e-06,
+      "loss": 0.1868,
+      "step": 457
+    },
+    {
+      "epoch": 0.07420609202851587,
+      "grad_norm": 1.786029577255249,
+      "learning_rate": 4.118705035971223e-06,
+      "loss": 0.1933,
+      "step": 458
+    },
+    {
+      "epoch": 0.07436811406351264,
+      "grad_norm": 1.6774612665176392,
+      "learning_rate": 4.127697841726619e-06,
+      "loss": 0.1834,
+      "step": 459
+    },
+    {
+      "epoch": 0.07453013609850939,
+      "grad_norm": 1.8457497358322144,
+      "learning_rate": 4.1366906474820145e-06,
+      "loss": 0.1785,
+      "step": 460
+    },
+    {
+      "epoch": 0.07469215813350616,
+      "grad_norm": 1.6226444244384766,
+      "learning_rate": 4.14568345323741e-06,
+      "loss": 0.1689,
+      "step": 461
+    },
+    {
+      "epoch": 0.07485418016850291,
+      "grad_norm": 1.7107115983963013,
+      "learning_rate": 4.154676258992807e-06,
+      "loss": 0.1993,
+      "step": 462
+    },
+    {
+      "epoch": 0.07501620220349968,
+      "grad_norm": 1.7487287521362305,
+      "learning_rate": 4.163669064748202e-06,
+      "loss": 0.2032,
+      "step": 463
+    },
+    {
+      "epoch": 0.07517822423849643,
+      "grad_norm": 1.8493058681488037,
+      "learning_rate": 4.172661870503597e-06,
+      "loss": 0.1937,
+      "step": 464
+    },
+    {
+      "epoch": 0.0753402462734932,
+      "grad_norm": 1.7110384702682495,
+      "learning_rate": 4.1816546762589935e-06,
+      "loss": 0.1755,
+      "step": 465
+    },
+    {
+      "epoch": 0.07550226830848995,
+      "grad_norm": 1.6556897163391113,
+      "learning_rate": 4.190647482014389e-06,
+      "loss": 0.1961,
+      "step": 466
+    },
+    {
+      "epoch": 0.07566429034348672,
+      "grad_norm": 1.5604274272918701,
+      "learning_rate": 4.199640287769784e-06,
+      "loss": 0.1725,
+      "step": 467
+    },
+    {
+      "epoch": 0.07582631237848347,
+      "grad_norm": 1.639374017715454,
+      "learning_rate": 4.20863309352518e-06,
+      "loss": 0.1795,
+      "step": 468
+    },
+    {
+      "epoch": 0.07598833441348023,
+      "grad_norm": 1.582411289215088,
+      "learning_rate": 4.2176258992805755e-06,
+      "loss": 0.1786,
+      "step": 469
+    },
+    {
+      "epoch": 0.076150356448477,
+      "grad_norm": 1.62992262840271,
+      "learning_rate": 4.226618705035972e-06,
+      "loss": 0.1732,
+      "step": 470
+    },
+    {
+      "epoch": 0.07631237848347375,
+      "grad_norm": 1.8388803005218506,
+      "learning_rate": 4.235611510791367e-06,
+      "loss": 0.208,
+      "step": 471
+    },
+    {
+      "epoch": 0.07647440051847051,
+      "grad_norm": 1.8594791889190674,
+      "learning_rate": 4.244604316546763e-06,
+      "loss": 0.2148,
+      "step": 472
+    },
+    {
+      "epoch": 0.07663642255346727,
+      "grad_norm": 1.814803957939148,
+      "learning_rate": 4.2535971223021584e-06,
+      "loss": 0.2031,
+      "step": 473
+    },
+    {
+      "epoch": 0.07679844458846403,
+      "grad_norm": 1.6284650564193726,
+      "learning_rate": 4.2625899280575546e-06,
+      "loss": 0.1865,
+      "step": 474
+    },
+    {
+      "epoch": 0.07696046662346079,
+      "grad_norm": 1.8454334735870361,
+      "learning_rate": 4.27158273381295e-06,
+      "loss": 0.1818,
+      "step": 475
+    },
+    {
+      "epoch": 0.07712248865845756,
+      "grad_norm": 1.8718208074569702,
+      "learning_rate": 4.280575539568346e-06,
+      "loss": 0.1912,
+      "step": 476
+    },
+    {
+      "epoch": 0.07728451069345431,
+      "grad_norm": 1.4540497064590454,
+      "learning_rate": 4.289568345323741e-06,
+      "loss": 0.1595,
+      "step": 477
+    },
+    {
+      "epoch": 0.07744653272845108,
+      "grad_norm": 1.749086618423462,
+      "learning_rate": 4.298561151079137e-06,
+      "loss": 0.2013,
+      "step": 478
+    },
+    {
+      "epoch": 0.07760855476344783,
+      "grad_norm": 1.8641357421875,
+      "learning_rate": 4.307553956834533e-06,
+      "loss": 0.202,
+      "step": 479
+    },
+    {
+      "epoch": 0.07777057679844458,
+      "grad_norm": 1.670167326927185,
+      "learning_rate": 4.316546762589928e-06,
+      "loss": 0.1815,
+      "step": 480
+    },
+    {
+      "epoch": 0.07793259883344135,
+      "grad_norm": 1.800274133682251,
+      "learning_rate": 4.325539568345324e-06,
+      "loss": 0.2006,
+      "step": 481
+    },
+    {
+      "epoch": 0.0780946208684381,
+      "grad_norm": 2.035557985305786,
+      "learning_rate": 4.33453237410072e-06,
+      "loss": 0.1903,
+      "step": 482
+    },
+    {
+      "epoch": 0.07825664290343487,
+      "grad_norm": 1.7821377515792847,
+      "learning_rate": 4.343525179856116e-06,
+      "loss": 0.2089,
+      "step": 483
+    },
+    {
+      "epoch": 0.07841866493843162,
+      "grad_norm": 1.605576515197754,
+      "learning_rate": 4.352517985611511e-06,
+      "loss": 0.1625,
+      "step": 484
+    },
+    {
+      "epoch": 0.07858068697342839,
+      "grad_norm": 1.7494524717330933,
+      "learning_rate": 4.361510791366907e-06,
+      "loss": 0.1915,
+      "step": 485
+    },
+    {
+      "epoch": 0.07874270900842514,
+      "grad_norm": 1.4881764650344849,
+      "learning_rate": 4.370503597122302e-06,
+      "loss": 0.1512,
+      "step": 486
+    },
+    {
+      "epoch": 0.07890473104342191,
+      "grad_norm": 1.9464272260665894,
+      "learning_rate": 4.3794964028776985e-06,
+      "loss": 0.1752,
+      "step": 487
+    },
+    {
+      "epoch": 0.07906675307841866,
+      "grad_norm": 1.8107142448425293,
+      "learning_rate": 4.388489208633094e-06,
+      "loss": 0.1953,
+      "step": 488
+    },
+    {
+      "epoch": 0.07922877511341543,
+      "grad_norm": 1.7034633159637451,
+      "learning_rate": 4.397482014388489e-06,
+      "loss": 0.1875,
+      "step": 489
+    },
+    {
+      "epoch": 0.07939079714841218,
+      "grad_norm": 1.7011704444885254,
+      "learning_rate": 4.406474820143885e-06,
+      "loss": 0.1942,
+      "step": 490
+    },
+    {
+      "epoch": 0.07955281918340894,
+      "grad_norm": 1.567685604095459,
+      "learning_rate": 4.415467625899281e-06,
+      "loss": 0.1869,
+      "step": 491
+    },
+    {
+      "epoch": 0.0797148412184057,
+      "grad_norm": 1.5715806484222412,
+      "learning_rate": 4.424460431654677e-06,
+      "loss": 0.1618,
+      "step": 492
+    },
+    {
+      "epoch": 0.07987686325340246,
+      "grad_norm": 1.6158721446990967,
+      "learning_rate": 4.433453237410073e-06,
+      "loss": 0.1903,
+      "step": 493
+    },
+    {
+      "epoch": 0.08003888528839923,
+      "grad_norm": 1.7132471799850464,
+      "learning_rate": 4.442446043165468e-06,
+      "loss": 0.1761,
+      "step": 494
+    },
+    {
+      "epoch": 0.08020090732339598,
+      "grad_norm": 1.561064600944519,
+      "learning_rate": 4.4514388489208635e-06,
+      "loss": 0.1644,
+      "step": 495
+    },
+    {
+      "epoch": 0.08036292935839275,
+      "grad_norm": 1.6423125267028809,
+      "learning_rate": 4.46043165467626e-06,
+      "loss": 0.1724,
+      "step": 496
+    },
+    {
+      "epoch": 0.0805249513933895,
+      "grad_norm": 1.5064512491226196,
+      "learning_rate": 4.469424460431655e-06,
+      "loss": 0.1421,
+      "step": 497
+    },
+    {
+      "epoch": 0.08068697342838627,
+      "grad_norm": 1.592776894569397,
+      "learning_rate": 4.478417266187051e-06,
+      "loss": 0.1831,
+      "step": 498
+    },
+    {
+      "epoch": 0.08084899546338302,
+      "grad_norm": 1.378753423690796,
+      "learning_rate": 4.487410071942446e-06,
+      "loss": 0.1469,
+      "step": 499
+    },
+    {
+      "epoch": 0.08101101749837979,
+      "grad_norm": 1.6191986799240112,
+      "learning_rate": 4.496402877697842e-06,
+      "loss": 0.1797,
+      "step": 500
+    },
+    {
+      "epoch": 0.08117303953337654,
+      "grad_norm": 1.6818901300430298,
+      "learning_rate": 4.505395683453238e-06,
+      "loss": 0.1666,
+      "step": 501
+    },
+    {
+      "epoch": 0.08133506156837329,
+      "grad_norm": 1.6711454391479492,
+      "learning_rate": 4.514388489208634e-06,
+      "loss": 0.1883,
+      "step": 502
+    },
+    {
+      "epoch": 0.08149708360337006,
+      "grad_norm": 1.5778876543045044,
+      "learning_rate": 4.523381294964029e-06,
+      "loss": 0.1556,
+      "step": 503
+    },
+    {
+      "epoch": 0.08165910563836681,
+      "grad_norm": 1.751618504524231,
+      "learning_rate": 4.5323741007194245e-06,
+      "loss": 0.1773,
+      "step": 504
+    },
+    {
+      "epoch": 0.08182112767336358,
+      "grad_norm": 1.6356712579727173,
+      "learning_rate": 4.541366906474821e-06,
+      "loss": 0.1674,
+      "step": 505
+    },
+    {
+      "epoch": 0.08198314970836033,
+      "grad_norm": 1.717413306236267,
+      "learning_rate": 4.550359712230216e-06,
+      "loss": 0.2015,
+      "step": 506
+    },
+    {
+      "epoch": 0.0821451717433571,
+      "grad_norm": 1.6795759201049805,
+      "learning_rate": 4.559352517985612e-06,
+      "loss": 0.1858,
+      "step": 507
+    },
+    {
+      "epoch": 0.08230719377835385,
+      "grad_norm": 1.4658695459365845,
+      "learning_rate": 4.5683453237410074e-06,
+      "loss": 0.1719,
+      "step": 508
+    },
+    {
+      "epoch": 0.08246921581335062,
+      "grad_norm": 1.865307331085205,
+      "learning_rate": 4.577338129496403e-06,
+      "loss": 0.1978,
+      "step": 509
+    },
+    {
+      "epoch": 0.08263123784834737,
+      "grad_norm": 1.8304200172424316,
+      "learning_rate": 4.586330935251799e-06,
+      "loss": 0.1906,
+      "step": 510
+    },
+    {
+      "epoch": 0.08279325988334413,
+      "grad_norm": 1.795590877532959,
+      "learning_rate": 4.595323741007194e-06,
+      "loss": 0.1827,
+      "step": 511
+    },
+    {
+      "epoch": 0.0829552819183409,
+      "grad_norm": 1.6314994096755981,
+      "learning_rate": 4.60431654676259e-06,
+      "loss": 0.1725,
+      "step": 512
+    },
+    {
+      "epoch": 0.08311730395333765,
+      "grad_norm": 1.6502032279968262,
+      "learning_rate": 4.6133093525179865e-06,
+      "loss": 0.1895,
+      "step": 513
+    },
+    {
+      "epoch": 0.08327932598833442,
+      "grad_norm": 1.5785597562789917,
+      "learning_rate": 4.622302158273382e-06,
+      "loss": 0.1602,
+      "step": 514
+    },
+    {
+      "epoch": 0.08344134802333117,
+      "grad_norm": 1.8144105672836304,
+      "learning_rate": 4.631294964028777e-06,
+      "loss": 0.174,
+      "step": 515
+    },
+    {
+      "epoch": 0.08360337005832794,
+      "grad_norm": 1.689868688583374,
+      "learning_rate": 4.640287769784173e-06,
+      "loss": 0.176,
+      "step": 516
+    },
+    {
+      "epoch": 0.08376539209332469,
+      "grad_norm": 1.7384047508239746,
+      "learning_rate": 4.6492805755395685e-06,
+      "loss": 0.1739,
+      "step": 517
+    },
+    {
+      "epoch": 0.08392741412832146,
+      "grad_norm": 1.7412197589874268,
+      "learning_rate": 4.658273381294965e-06,
+      "loss": 0.1957,
+      "step": 518
+    },
+    {
+      "epoch": 0.08408943616331821,
+      "grad_norm": 1.6650651693344116,
+      "learning_rate": 4.66726618705036e-06,
+      "loss": 0.184,
+      "step": 519
+    },
+    {
+      "epoch": 0.08425145819831498,
+      "grad_norm": 1.7459502220153809,
+      "learning_rate": 4.676258992805755e-06,
+      "loss": 0.1772,
+      "step": 520
+    },
+    {
+      "epoch": 0.08441348023331173,
+      "grad_norm": 1.6877442598342896,
+      "learning_rate": 4.685251798561151e-06,
+      "loss": 0.1872,
+      "step": 521
+    },
+    {
+      "epoch": 0.08457550226830848,
+      "grad_norm": 1.6881842613220215,
+      "learning_rate": 4.6942446043165475e-06,
+      "loss": 0.1841,
+      "step": 522
+    },
+    {
+      "epoch": 0.08473752430330525,
+      "grad_norm": 1.4388700723648071,
+      "learning_rate": 4.703237410071943e-06,
+      "loss": 0.1652,
+      "step": 523
+    },
+    {
+      "epoch": 0.084899546338302,
+      "grad_norm": 1.5876359939575195,
+      "learning_rate": 4.712230215827339e-06,
+      "loss": 0.2057,
+      "step": 524
+    },
+    {
+      "epoch": 0.08506156837329877,
+      "grad_norm": 1.647616982460022,
+      "learning_rate": 4.721223021582734e-06,
+      "loss": 0.186,
+      "step": 525
+    },
+    {
+      "epoch": 0.08522359040829552,
+      "grad_norm": 1.5112206935882568,
+      "learning_rate": 4.73021582733813e-06,
+      "loss": 0.1697,
+      "step": 526
+    },
+    {
+      "epoch": 0.08538561244329229,
+      "grad_norm": 1.553218126296997,
+      "learning_rate": 4.739208633093526e-06,
+      "loss": 0.1871,
+      "step": 527
+    },
+    {
+      "epoch": 0.08554763447828904,
+      "grad_norm": 1.5007338523864746,
+      "learning_rate": 4.748201438848921e-06,
+      "loss": 0.1555,
+      "step": 528
+    },
+    {
+      "epoch": 0.08570965651328581,
+      "grad_norm": 1.6641744375228882,
+      "learning_rate": 4.757194244604317e-06,
+      "loss": 0.1909,
+      "step": 529
+    },
+    {
+      "epoch": 0.08587167854828257,
+      "grad_norm": 1.8256597518920898,
+      "learning_rate": 4.7661870503597125e-06,
+      "loss": 0.1872,
+      "step": 530
+    },
+    {
+      "epoch": 0.08603370058327933,
+      "grad_norm": 1.818845272064209,
+      "learning_rate": 4.775179856115108e-06,
+      "loss": 0.1928,
+      "step": 531
+    },
+    {
+      "epoch": 0.08619572261827609,
+      "grad_norm": 1.664866328239441,
+      "learning_rate": 4.784172661870504e-06,
+      "loss": 0.167,
+      "step": 532
+    },
+    {
+      "epoch": 0.08635774465327284,
+      "grad_norm": 1.511879563331604,
+      "learning_rate": 4.7931654676259e-06,
+      "loss": 0.1692,
+      "step": 533
+    },
+    {
+      "epoch": 0.0865197666882696,
+      "grad_norm": 1.680817723274231,
+      "learning_rate": 4.802158273381295e-06,
+      "loss": 0.1897,
+      "step": 534
+    },
+    {
+      "epoch": 0.08668178872326636,
+      "grad_norm": 1.5544551610946655,
+      "learning_rate": 4.8111510791366915e-06,
+      "loss": 0.1597,
+      "step": 535
+    },
+    {
+      "epoch": 0.08684381075826313,
+      "grad_norm": 1.759607195854187,
+      "learning_rate": 4.820143884892087e-06,
+      "loss": 0.1925,
+      "step": 536
+    },
+    {
+      "epoch": 0.08700583279325988,
+      "grad_norm": 1.6217458248138428,
+      "learning_rate": 4.829136690647482e-06,
+      "loss": 0.1897,
+      "step": 537
+    },
+    {
+      "epoch": 0.08716785482825665,
+      "grad_norm": 1.574110507965088,
+      "learning_rate": 4.838129496402878e-06,
+      "loss": 0.1898,
+      "step": 538
+    },
+    {
+      "epoch": 0.0873298768632534,
+      "grad_norm": 1.4586173295974731,
+      "learning_rate": 4.8471223021582736e-06,
+      "loss": 0.1767,
+      "step": 539
+    },
+    {
+      "epoch": 0.08749189889825017,
+      "grad_norm": 1.533091425895691,
+      "learning_rate": 4.856115107913669e-06,
+      "loss": 0.1819,
+      "step": 540
+    },
+    {
+      "epoch": 0.08765392093324692,
+      "grad_norm": 1.4116888046264648,
+      "learning_rate": 4.865107913669065e-06,
+      "loss": 0.1611,
+      "step": 541
+    },
+    {
+      "epoch": 0.08781594296824369,
+      "grad_norm": 1.6544139385223389,
+      "learning_rate": 4.874100719424461e-06,
+      "loss": 0.1787,
+      "step": 542
+    },
+    {
+      "epoch": 0.08797796500324044,
+      "grad_norm": 1.6531710624694824,
+      "learning_rate": 4.8830935251798564e-06,
+      "loss": 0.1961,
+      "step": 543
+    },
+    {
+      "epoch": 0.0881399870382372,
+      "grad_norm": 1.4757072925567627,
+      "learning_rate": 4.892086330935253e-06,
+      "loss": 0.1713,
+      "step": 544
+    },
+    {
+      "epoch": 0.08830200907323396,
+      "grad_norm": 1.6193426847457886,
+      "learning_rate": 4.901079136690648e-06,
+      "loss": 0.1845,
+      "step": 545
+    },
+    {
+      "epoch": 0.08846403110823071,
+      "grad_norm": 1.7032020092010498,
+      "learning_rate": 4.910071942446043e-06,
+      "loss": 0.189,
+      "step": 546
+    },
+    {
+      "epoch": 0.08862605314322748,
+      "grad_norm": 1.4911900758743286,
+      "learning_rate": 4.919064748201439e-06,
+      "loss": 0.1592,
+      "step": 547
+    },
+    {
+      "epoch": 0.08878807517822424,
+      "grad_norm": 1.6090925931930542,
+      "learning_rate": 4.928057553956835e-06,
+      "loss": 0.1983,
+      "step": 548
+    },
+    {
+      "epoch": 0.088950097213221,
+      "grad_norm": 1.876451015472412,
+      "learning_rate": 4.937050359712231e-06,
+      "loss": 0.2078,
+      "step": 549
+    },
+    {
+      "epoch": 0.08911211924821776,
+      "grad_norm": 1.5491797924041748,
+      "learning_rate": 4.946043165467626e-06,
+      "loss": 0.178,
+      "step": 550
+    },
+    {
+      "epoch": 0.08927414128321452,
+      "grad_norm": 1.6761187314987183,
+      "learning_rate": 4.955035971223021e-06,
+      "loss": 0.2046,
+      "step": 551
+    },
+    {
+      "epoch": 0.08943616331821128,
+      "grad_norm": 1.6212235689163208,
+      "learning_rate": 4.9640287769784175e-06,
+      "loss": 0.2055,
+      "step": 552
+    },
+    {
+      "epoch": 0.08959818535320804,
+      "grad_norm": 1.6363742351531982,
+      "learning_rate": 4.973021582733814e-06,
+      "loss": 0.1966,
+      "step": 553
+    },
+    {
+      "epoch": 0.0897602073882048,
+      "grad_norm": 1.580815076828003,
+      "learning_rate": 4.982014388489209e-06,
+      "loss": 0.2054,
+      "step": 554
+    },
+    {
+      "epoch": 0.08992222942320155,
+      "grad_norm": 1.9822163581848145,
+      "learning_rate": 4.991007194244605e-06,
+      "loss": 0.155,
+      "step": 555
+    },
+    {
+      "epoch": 0.09008425145819832,
+      "grad_norm": 1.7965415716171265,
+      "learning_rate": 5e-06,
+      "loss": 0.1808,
+      "step": 556
+    },
+    {
+      "epoch": 0.09024627349319507,
+      "grad_norm": 1.4814484119415283,
+      "learning_rate": 4.999999961753026e-06,
+      "loss": 0.1716,
+      "step": 557
+    },
+    {
+      "epoch": 0.09040829552819184,
+      "grad_norm": 1.5439770221710205,
+      "learning_rate": 4.999999847012101e-06,
+      "loss": 0.1672,
+      "step": 558
+    },
+    {
+      "epoch": 0.09057031756318859,
+      "grad_norm": 1.5221511125564575,
+      "learning_rate": 4.999999655777232e-06,
+      "loss": 0.1829,
+      "step": 559
+    },
+    {
+      "epoch": 0.09073233959818536,
+      "grad_norm": 1.8477576971054077,
+      "learning_rate": 4.9999993880484235e-06,
+      "loss": 0.1827,
+      "step": 560
+    },
+    {
+      "epoch": 0.09089436163318211,
+      "grad_norm": 1.5962644815444946,
+      "learning_rate": 4.999999043825682e-06,
+      "loss": 0.1675,
+      "step": 561
+    },
+    {
+      "epoch": 0.09105638366817888,
+      "grad_norm": 1.5648473501205444,
+      "learning_rate": 4.999998623109022e-06,
+      "loss": 0.1645,
+      "step": 562
+    },
+    {
+      "epoch": 0.09121840570317563,
+      "grad_norm": 1.4602282047271729,
+      "learning_rate": 4.999998125898452e-06,
+      "loss": 0.1595,
+      "step": 563
+    },
+    {
+      "epoch": 0.09138042773817238,
+      "grad_norm": 2.079158306121826,
+      "learning_rate": 4.99999755219399e-06,
+      "loss": 0.2142,
+      "step": 564
+    },
+    {
+      "epoch": 0.09154244977316915,
+      "grad_norm": 1.9085109233856201,
+      "learning_rate": 4.9999969019956526e-06,
+      "loss": 0.1923,
+      "step": 565
+    },
+    {
+      "epoch": 0.0917044718081659,
+      "grad_norm": 1.8776715993881226,
+      "learning_rate": 4.9999961753034595e-06,
+      "loss": 0.1935,
+      "step": 566
+    },
+    {
+      "epoch": 0.09186649384316267,
+      "grad_norm": 1.6320695877075195,
+      "learning_rate": 4.9999953721174345e-06,
+      "loss": 0.1881,
+      "step": 567
+    },
+    {
+      "epoch": 0.09202851587815943,
+      "grad_norm": 1.6475578546524048,
+      "learning_rate": 4.9999944924376e-06,
+      "loss": 0.191,
+      "step": 568
+    },
+    {
+      "epoch": 0.09219053791315619,
+      "grad_norm": 1.4152305126190186,
+      "learning_rate": 4.9999935362639844e-06,
+      "loss": 0.1728,
+      "step": 569
+    },
+    {
+      "epoch": 0.09235255994815295,
+      "grad_norm": 1.6144766807556152,
+      "learning_rate": 4.999992503596616e-06,
+      "loss": 0.1727,
+      "step": 570
+    },
+    {
+      "epoch": 0.09251458198314971,
+      "grad_norm": 1.455117106437683,
+      "learning_rate": 4.999991394435527e-06,
+      "loss": 0.1658,
+      "step": 571
+    },
+    {
+      "epoch": 0.09267660401814647,
+      "grad_norm": 1.536605715751648,
+      "learning_rate": 4.999990208780751e-06,
+      "loss": 0.1875,
+      "step": 572
+    },
+    {
+      "epoch": 0.09283862605314323,
+      "grad_norm": 1.4941188097000122,
+      "learning_rate": 4.999988946632326e-06,
+      "loss": 0.1732,
+      "step": 573
+    },
+    {
+      "epoch": 0.09300064808813999,
+      "grad_norm": 1.8162550926208496,
+      "learning_rate": 4.999987607990287e-06,
+      "loss": 0.2139,
+      "step": 574
+    },
+    {
+      "epoch": 0.09316267012313674,
+      "grad_norm": 1.8714874982833862,
+      "learning_rate": 4.9999861928546786e-06,
+      "loss": 0.2113,
+      "step": 575
+    },
+    {
+      "epoch": 0.09332469215813351,
+      "grad_norm": 1.6756885051727295,
+      "learning_rate": 4.999984701225542e-06,
+      "loss": 0.1855,
+      "step": 576
+    },
+    {
+      "epoch": 0.09348671419313026,
+      "grad_norm": 1.7077648639678955,
+      "learning_rate": 4.999983133102923e-06,
+      "loss": 0.184,
+      "step": 577
+    },
+    {
+      "epoch": 0.09364873622812703,
+      "grad_norm": 1.6993731260299683,
+      "learning_rate": 4.9999814884868705e-06,
+      "loss": 0.2088,
+      "step": 578
+    },
+    {
+      "epoch": 0.09381075826312378,
+      "grad_norm": 1.6294496059417725,
+      "learning_rate": 4.999979767377434e-06,
+      "loss": 0.1909,
+      "step": 579
+    },
+    {
+      "epoch": 0.09397278029812055,
+      "grad_norm": 1.4724843502044678,
+      "learning_rate": 4.999977969774666e-06,
+      "loss": 0.1786,
+      "step": 580
+    },
+    {
+      "epoch": 0.0941348023331173,
+      "grad_norm": 1.5849987268447876,
+      "learning_rate": 4.999976095678622e-06,
+      "loss": 0.1942,
+      "step": 581
+    },
+    {
+      "epoch": 0.09429682436811407,
+      "grad_norm": 1.393203854560852,
+      "learning_rate": 4.99997414508936e-06,
+      "loss": 0.157,
+      "step": 582
+    },
+    {
+      "epoch": 0.09445884640311082,
+      "grad_norm": 1.5272077322006226,
+      "learning_rate": 4.999972118006939e-06,
+      "loss": 0.1809,
+      "step": 583
+    },
+    {
+      "epoch": 0.09462086843810759,
+      "grad_norm": 1.4402399063110352,
+      "learning_rate": 4.999970014431421e-06,
+      "loss": 0.187,
+      "step": 584
+    },
+    {
+      "epoch": 0.09478289047310434,
+      "grad_norm": 1.433086633682251,
+      "learning_rate": 4.99996783436287e-06,
+      "loss": 0.179,
+      "step": 585
+    },
+    {
+      "epoch": 0.0949449125081011,
+      "grad_norm": 1.463783860206604,
+      "learning_rate": 4.999965577801354e-06,
+      "loss": 0.1703,
+      "step": 586
+    },
+    {
+      "epoch": 0.09510693454309786,
+      "grad_norm": 1.4071027040481567,
+      "learning_rate": 4.9999632447469395e-06,
+      "loss": 0.1708,
+      "step": 587
+    },
+    {
+      "epoch": 0.09526895657809462,
+      "grad_norm": 1.5153921842575073,
+      "learning_rate": 4.999960835199701e-06,
+      "loss": 0.1786,
+      "step": 588
+    },
+    {
+      "epoch": 0.09543097861309138,
+      "grad_norm": 1.577370047569275,
+      "learning_rate": 4.999958349159709e-06,
+      "loss": 0.178,
+      "step": 589
+    },
+    {
+      "epoch": 0.09559300064808814,
+      "grad_norm": 1.562628984451294,
+      "learning_rate": 4.999955786627042e-06,
+      "loss": 0.1793,
+      "step": 590
+    },
+    {
+      "epoch": 0.0957550226830849,
+      "grad_norm": 1.4154744148254395,
+      "learning_rate": 4.999953147601779e-06,
+      "loss": 0.1573,
+      "step": 591
+    },
+    {
+      "epoch": 0.09591704471808166,
+      "grad_norm": 1.473583459854126,
+      "learning_rate": 4.999950432083998e-06,
+      "loss": 0.1735,
+      "step": 592
+    },
+    {
+      "epoch": 0.09607906675307842,
+      "grad_norm": 1.5967344045639038,
+      "learning_rate": 4.999947640073784e-06,
+      "loss": 0.1798,
+      "step": 593
+    },
+    {
+      "epoch": 0.09624108878807518,
+      "grad_norm": 1.5491890907287598,
+      "learning_rate": 4.999944771571222e-06,
+      "loss": 0.1899,
+      "step": 594
+    },
+    {
+      "epoch": 0.09640311082307194,
+      "grad_norm": 1.5898832082748413,
+      "learning_rate": 4.9999418265764e-06,
+      "loss": 0.1825,
+      "step": 595
+    },
+    {
+      "epoch": 0.0965651328580687,
+      "grad_norm": 1.4132360219955444,
+      "learning_rate": 4.999938805089407e-06,
+      "loss": 0.1663,
+      "step": 596
+    },
+    {
+      "epoch": 0.09672715489306545,
+      "grad_norm": 1.5513783693313599,
+      "learning_rate": 4.999935707110337e-06,
+      "loss": 0.2075,
+      "step": 597
+    },
+    {
+      "epoch": 0.09688917692806222,
+      "grad_norm": 1.444082498550415,
+      "learning_rate": 4.999932532639285e-06,
+      "loss": 0.1501,
+      "step": 598
+    },
+    {
+      "epoch": 0.09705119896305897,
+      "grad_norm": 1.5398008823394775,
+      "learning_rate": 4.999929281676346e-06,
+      "loss": 0.1837,
+      "step": 599
+    },
+    {
+      "epoch": 0.09721322099805574,
+      "grad_norm": 1.4911479949951172,
+      "learning_rate": 4.99992595422162e-06,
+      "loss": 0.1966,
+      "step": 600
+    },
+    {
+      "epoch": 0.09737524303305249,
+      "grad_norm": 1.8352231979370117,
+      "learning_rate": 4.99992255027521e-06,
+      "loss": 0.201,
+      "step": 601
+    },
+    {
+      "epoch": 0.09753726506804926,
+      "grad_norm": 1.491248369216919,
+      "learning_rate": 4.9999190698372216e-06,
+      "loss": 0.187,
+      "step": 602
+    },
+    {
+      "epoch": 0.09769928710304601,
+      "grad_norm": 1.4229164123535156,
+      "learning_rate": 4.999915512907757e-06,
+      "loss": 0.1733,
+      "step": 603
+    },
+    {
+      "epoch": 0.09786130913804278,
+      "grad_norm": 1.624070167541504,
+      "learning_rate": 4.9999118794869285e-06,
+      "loss": 0.1749,
+      "step": 604
+    },
+    {
+      "epoch": 0.09802333117303953,
+      "grad_norm": 1.3550504446029663,
+      "learning_rate": 4.999908169574846e-06,
+      "loss": 0.1683,
+      "step": 605
+    },
+    {
+      "epoch": 0.0981853532080363,
+      "grad_norm": 1.5067567825317383,
+      "learning_rate": 4.999904383171623e-06,
+      "loss": 0.1891,
+      "step": 606
+    },
+    {
+      "epoch": 0.09834737524303305,
+      "grad_norm": 1.5590304136276245,
+      "learning_rate": 4.999900520277376e-06,
+      "loss": 0.1831,
+      "step": 607
+    },
+    {
+      "epoch": 0.0985093972780298,
+      "grad_norm": 1.4621655941009521,
+      "learning_rate": 4.999896580892221e-06,
+      "loss": 0.1648,
+      "step": 608
+    },
+    {
+      "epoch": 0.09867141931302657,
+      "grad_norm": 1.5438154935836792,
+      "learning_rate": 4.999892565016282e-06,
+      "loss": 0.1728,
+      "step": 609
+    },
+    {
+      "epoch": 0.09883344134802333,
+      "grad_norm": 1.5466684103012085,
+      "learning_rate": 4.99988847264968e-06,
+      "loss": 0.1915,
+      "step": 610
+    },
+    {
+      "epoch": 0.0989954633830201,
+      "grad_norm": 1.430253267288208,
+      "learning_rate": 4.99988430379254e-06,
+      "loss": 0.1577,
+      "step": 611
+    },
+    {
+      "epoch": 0.09915748541801685,
+      "grad_norm": 1.4443409442901611,
+      "learning_rate": 4.99988005844499e-06,
+      "loss": 0.1668,
+      "step": 612
+    },
+    {
+      "epoch": 0.09931950745301361,
+      "grad_norm": 1.5595300197601318,
+      "learning_rate": 4.999875736607159e-06,
+      "loss": 0.1686,
+      "step": 613
+    },
+    {
+      "epoch": 0.09948152948801037,
+      "grad_norm": 1.4867337942123413,
+      "learning_rate": 4.999871338279181e-06,
+      "loss": 0.1906,
+      "step": 614
+    },
+    {
+      "epoch": 0.09964355152300713,
+      "grad_norm": 1.4922622442245483,
+      "learning_rate": 4.99986686346119e-06,
+      "loss": 0.1727,
+      "step": 615
+    },
+    {
+      "epoch": 0.09980557355800389,
+      "grad_norm": 1.4832661151885986,
+      "learning_rate": 4.999862312153322e-06,
+      "loss": 0.1703,
+      "step": 616
+    },
+    {
+      "epoch": 0.09996759559300065,
+      "grad_norm": 1.4716413021087646,
+      "learning_rate": 4.999857684355716e-06,
+      "loss": 0.1704,
+      "step": 617
+    },
+    {
+      "epoch": 0.10012961762799741,
+      "grad_norm": 1.4800324440002441,
+      "learning_rate": 4.999852980068516e-06,
+      "loss": 0.1604,
+      "step": 618
+    },
+    {
+      "epoch": 0.10029163966299416,
+      "grad_norm": 1.4337764978408813,
+      "learning_rate": 4.999848199291863e-06,
+      "loss": 0.1785,
+      "step": 619
+    },
+    {
+      "epoch": 0.10045366169799093,
+      "grad_norm": 1.373763918876648,
+      "learning_rate": 4.9998433420259055e-06,
+      "loss": 0.1532,
+      "step": 620
+    },
+    {
+      "epoch": 0.10061568373298768,
+      "grad_norm": 1.4618251323699951,
+      "learning_rate": 4.999838408270791e-06,
+      "loss": 0.1849,
+      "step": 621
+    },
+    {
+      "epoch": 0.10077770576798445,
+      "grad_norm": 1.356284499168396,
+      "learning_rate": 4.99983339802667e-06,
+      "loss": 0.1555,
+      "step": 622
+    },
+    {
+      "epoch": 0.1009397278029812,
+      "grad_norm": 1.548274040222168,
+      "learning_rate": 4.999828311293697e-06,
+      "loss": 0.1913,
+      "step": 623
+    },
+    {
+      "epoch": 0.10110174983797797,
+      "grad_norm": 1.549202799797058,
+      "learning_rate": 4.999823148072027e-06,
+      "loss": 0.1909,
+      "step": 624
+    },
+    {
+      "epoch": 0.10126377187297472,
+      "grad_norm": 1.5712025165557861,
+      "learning_rate": 4.999817908361818e-06,
+      "loss": 0.1814,
+      "step": 625
+    },
+    {
+      "epoch": 0.10142579390797149,
+      "grad_norm": 1.3882803916931152,
+      "learning_rate": 4.999812592163232e-06,
+      "loss": 0.1725,
+      "step": 626
+    },
+    {
+      "epoch": 0.10158781594296824,
+      "grad_norm": 1.4438263177871704,
+      "learning_rate": 4.999807199476428e-06,
+      "loss": 0.1663,
+      "step": 627
+    },
+    {
+      "epoch": 0.101749837977965,
+      "grad_norm": 1.430675745010376,
+      "learning_rate": 4.9998017303015735e-06,
+      "loss": 0.1685,
+      "step": 628
+    },
+    {
+      "epoch": 0.10191186001296176,
+      "grad_norm": 1.514258623123169,
+      "learning_rate": 4.999796184638836e-06,
+      "loss": 0.2044,
+      "step": 629
+    },
+    {
+      "epoch": 0.10207388204795852,
+      "grad_norm": 1.3589094877243042,
+      "learning_rate": 4.999790562488385e-06,
+      "loss": 0.1636,
+      "step": 630
+    },
+    {
+      "epoch": 0.10223590408295528,
+      "grad_norm": 1.56080961227417,
+      "learning_rate": 4.999784863850391e-06,
+      "loss": 0.1925,
+      "step": 631
+    },
+    {
+      "epoch": 0.10239792611795204,
+      "grad_norm": 1.3032417297363281,
+      "learning_rate": 4.999779088725031e-06,
+      "loss": 0.1394,
+      "step": 632
+    },
+    {
+      "epoch": 0.1025599481529488,
+      "grad_norm": 1.3875499963760376,
+      "learning_rate": 4.999773237112479e-06,
+      "loss": 0.1593,
+      "step": 633
+    },
+    {
+      "epoch": 0.10272197018794556,
+      "grad_norm": 1.6035758256912231,
+      "learning_rate": 4.999767309012916e-06,
+      "loss": 0.1822,
+      "step": 634
+    },
+    {
+      "epoch": 0.10288399222294232,
+      "grad_norm": 1.519217610359192,
+      "learning_rate": 4.999761304426523e-06,
+      "loss": 0.187,
+      "step": 635
+    },
+    {
+      "epoch": 0.10304601425793908,
+      "grad_norm": 1.6218081712722778,
+      "learning_rate": 4.999755223353483e-06,
+      "loss": 0.1963,
+      "step": 636
+    },
+    {
+      "epoch": 0.10320803629293585,
+      "grad_norm": 1.4830374717712402,
+      "learning_rate": 4.999749065793982e-06,
+      "loss": 0.1656,
+      "step": 637
+    },
+    {
+      "epoch": 0.1033700583279326,
+      "grad_norm": 1.5597121715545654,
+      "learning_rate": 4.9997428317482086e-06,
+      "loss": 0.1817,
+      "step": 638
+    },
+    {
+      "epoch": 0.10353208036292935,
+      "grad_norm": 1.3938210010528564,
+      "learning_rate": 4.999736521216355e-06,
+      "loss": 0.1594,
+      "step": 639
+    },
+    {
+      "epoch": 0.10369410239792612,
+      "grad_norm": 1.3650377988815308,
+      "learning_rate": 4.999730134198612e-06,
+      "loss": 0.1661,
+      "step": 640
+    },
+    {
+      "epoch": 0.10385612443292287,
+      "grad_norm": 1.493725061416626,
+      "learning_rate": 4.999723670695177e-06,
+      "loss": 0.173,
+      "step": 641
+    },
+    {
+      "epoch": 0.10401814646791964,
+      "grad_norm": 1.5311143398284912,
+      "learning_rate": 4.999717130706247e-06,
+      "loss": 0.1933,
+      "step": 642
+    },
+    {
+      "epoch": 0.10418016850291639,
+      "grad_norm": 1.5196541547775269,
+      "learning_rate": 4.9997105142320205e-06,
+      "loss": 0.1797,
+      "step": 643
+    },
+    {
+      "epoch": 0.10434219053791316,
+      "grad_norm": 1.541266679763794,
+      "learning_rate": 4.999703821272702e-06,
+      "loss": 0.1813,
+      "step": 644
+    },
+    {
+      "epoch": 0.10450421257290991,
+      "grad_norm": 1.65276038646698,
+      "learning_rate": 4.999697051828497e-06,
+      "loss": 0.2055,
+      "step": 645
+    },
+    {
+      "epoch": 0.10466623460790668,
+      "grad_norm": 1.6365846395492554,
+      "learning_rate": 4.99969020589961e-06,
+      "loss": 0.2004,
+      "step": 646
+    },
+    {
+      "epoch": 0.10482825664290343,
+      "grad_norm": 1.675693154335022,
+      "learning_rate": 4.999683283486252e-06,
+      "loss": 0.1797,
+      "step": 647
+    },
+    {
+      "epoch": 0.1049902786779002,
+      "grad_norm": 1.5289831161499023,
+      "learning_rate": 4.999676284588635e-06,
+      "loss": 0.2035,
+      "step": 648
+    },
+    {
+      "epoch": 0.10515230071289695,
+      "grad_norm": 1.4459340572357178,
+      "learning_rate": 4.9996692092069735e-06,
+      "loss": 0.1682,
+      "step": 649
+    },
+    {
+      "epoch": 0.10531432274789371,
+      "grad_norm": 1.7202799320220947,
+      "learning_rate": 4.999662057341482e-06,
+      "loss": 0.2091,
+      "step": 650
+    },
+    {
+      "epoch": 0.10547634478289047,
+      "grad_norm": 1.4932163953781128,
+      "learning_rate": 4.999654828992382e-06,
+      "loss": 0.1856,
+      "step": 651
+    },
+    {
+      "epoch": 0.10563836681788723,
+      "grad_norm": 1.3993346691131592,
+      "learning_rate": 4.999647524159892e-06,
+      "loss": 0.1592,
+      "step": 652
+    },
+    {
+      "epoch": 0.105800388852884,
+      "grad_norm": 1.7462083101272583,
+      "learning_rate": 4.999640142844237e-06,
+      "loss": 0.1681,
+      "step": 653
+    },
+    {
+      "epoch": 0.10596241088788075,
+      "grad_norm": 1.5770518779754639,
+      "learning_rate": 4.9996326850456435e-06,
+      "loss": 0.1914,
+      "step": 654
+    },
+    {
+      "epoch": 0.10612443292287752,
+      "grad_norm": 1.3480045795440674,
+      "learning_rate": 4.9996251507643375e-06,
+      "loss": 0.1521,
+      "step": 655
+    },
+    {
+      "epoch": 0.10628645495787427,
+      "grad_norm": 1.5072104930877686,
+      "learning_rate": 4.999617540000552e-06,
+      "loss": 0.1811,
+      "step": 656
+    },
+    {
+      "epoch": 0.10644847699287104,
+      "grad_norm": 1.4140270948410034,
+      "learning_rate": 4.9996098527545184e-06,
+      "loss": 0.1599,
+      "step": 657
+    },
+    {
+      "epoch": 0.10661049902786779,
+      "grad_norm": 1.3946372270584106,
+      "learning_rate": 4.999602089026472e-06,
+      "loss": 0.1585,
+      "step": 658
+    },
+    {
+      "epoch": 0.10677252106286456,
+      "grad_norm": 1.7381147146224976,
+      "learning_rate": 4.9995942488166506e-06,
+      "loss": 0.2138,
+      "step": 659
+    },
+    {
+      "epoch": 0.10693454309786131,
+      "grad_norm": 1.5091066360473633,
+      "learning_rate": 4.999586332125294e-06,
+      "loss": 0.1769,
+      "step": 660
+    },
+    {
+      "epoch": 0.10709656513285806,
+      "grad_norm": 1.4993563890457153,
+      "learning_rate": 4.999578338952646e-06,
+      "loss": 0.1819,
+      "step": 661
+    },
+    {
+      "epoch": 0.10725858716785483,
+      "grad_norm": 1.6799975633621216,
+      "learning_rate": 4.9995702692989476e-06,
+      "loss": 0.2105,
+      "step": 662
+    },
+    {
+      "epoch": 0.10742060920285158,
+      "grad_norm": 1.5190070867538452,
+      "learning_rate": 4.999562123164448e-06,
+      "loss": 0.1842,
+      "step": 663
+    },
+    {
+      "epoch": 0.10758263123784835,
+      "grad_norm": 1.4410516023635864,
+      "learning_rate": 4.999553900549398e-06,
+      "loss": 0.1664,
+      "step": 664
+    },
+    {
+      "epoch": 0.1077446532728451,
+      "grad_norm": 1.4741100072860718,
+      "learning_rate": 4.999545601454046e-06,
+      "loss": 0.1726,
+      "step": 665
+    },
+    {
+      "epoch": 0.10790667530784187,
+      "grad_norm": 1.4260085821151733,
+      "learning_rate": 4.999537225878648e-06,
+      "loss": 0.1803,
+      "step": 666
+    },
+    {
+      "epoch": 0.10806869734283862,
+      "grad_norm": 1.4949939250946045,
+      "learning_rate": 4.999528773823459e-06,
+      "loss": 0.1779,
+      "step": 667
+    },
+    {
+      "epoch": 0.10823071937783539,
+      "grad_norm": 1.3150126934051514,
+      "learning_rate": 4.999520245288739e-06,
+      "loss": 0.1637,
+      "step": 668
+    },
+    {
+      "epoch": 0.10839274141283214,
+      "grad_norm": 1.3088539838790894,
+      "learning_rate": 4.999511640274748e-06,
+      "loss": 0.157,
+      "step": 669
+    },
+    {
+      "epoch": 0.10855476344782891,
+      "grad_norm": 1.4893115758895874,
+      "learning_rate": 4.999502958781749e-06,
+      "loss": 0.1793,
+      "step": 670
+    },
+    {
+      "epoch": 0.10871678548282566,
+      "grad_norm": 1.5916615724563599,
+      "learning_rate": 4.999494200810009e-06,
+      "loss": 0.1923,
+      "step": 671
+    },
+    {
+      "epoch": 0.10887880751782242,
+      "grad_norm": 1.4635965824127197,
+      "learning_rate": 4.999485366359794e-06,
+      "loss": 0.1867,
+      "step": 672
+    },
+    {
+      "epoch": 0.10904082955281919,
+      "grad_norm": 1.5369232892990112,
+      "learning_rate": 4.999476455431377e-06,
+      "loss": 0.175,
+      "step": 673
+    },
+    {
+      "epoch": 0.10920285158781594,
+      "grad_norm": 1.5748496055603027,
+      "learning_rate": 4.999467468025028e-06,
+      "loss": 0.2062,
+      "step": 674
+    },
+    {
+      "epoch": 0.1093648736228127,
+      "grad_norm": 1.5589791536331177,
+      "learning_rate": 4.999458404141023e-06,
+      "loss": 0.1822,
+      "step": 675
+    },
+    {
+      "epoch": 0.10952689565780946,
+      "grad_norm": 1.4582257270812988,
+      "learning_rate": 4.99944926377964e-06,
+      "loss": 0.1925,
+      "step": 676
+    },
+    {
+      "epoch": 0.10968891769280623,
+      "grad_norm": 1.5575858354568481,
+      "learning_rate": 4.9994400469411575e-06,
+      "loss": 0.1785,
+      "step": 677
+    },
+    {
+      "epoch": 0.10985093972780298,
+      "grad_norm": 1.3714617490768433,
+      "learning_rate": 4.999430753625858e-06,
+      "loss": 0.1631,
+      "step": 678
+    },
+    {
+      "epoch": 0.11001296176279975,
+      "grad_norm": 1.4028851985931396,
+      "learning_rate": 4.999421383834027e-06,
+      "loss": 0.1758,
+      "step": 679
+    },
+    {
+      "epoch": 0.1101749837977965,
+      "grad_norm": 1.395218014717102,
+      "learning_rate": 4.999411937565949e-06,
+      "loss": 0.1711,
+      "step": 680
+    },
+    {
+      "epoch": 0.11033700583279327,
+      "grad_norm": 1.4894384145736694,
+      "learning_rate": 4.999402414821915e-06,
+      "loss": 0.197,
+      "step": 681
+    },
+    {
+      "epoch": 0.11049902786779002,
+      "grad_norm": 1.489985466003418,
+      "learning_rate": 4.999392815602214e-06,
+      "loss": 0.1664,
+      "step": 682
+    },
+    {
+      "epoch": 0.11066104990278677,
+      "grad_norm": 1.4633452892303467,
+      "learning_rate": 4.9993831399071425e-06,
+      "loss": 0.1793,
+      "step": 683
+    },
+    {
+      "epoch": 0.11082307193778354,
+      "grad_norm": 1.4508930444717407,
+      "learning_rate": 4.999373387736996e-06,
+      "loss": 0.1914,
+      "step": 684
+    },
+    {
+      "epoch": 0.1109850939727803,
+      "grad_norm": 1.3373337984085083,
+      "learning_rate": 4.999363559092071e-06,
+      "loss": 0.1516,
+      "step": 685
+    },
+    {
+      "epoch": 0.11114711600777706,
+      "grad_norm": 1.4252420663833618,
+      "learning_rate": 4.999353653972669e-06,
+      "loss": 0.1849,
+      "step": 686
+    },
+    {
+      "epoch": 0.11130913804277381,
+      "grad_norm": 1.383408546447754,
+      "learning_rate": 4.999343672379095e-06,
+      "loss": 0.162,
+      "step": 687
+    },
+    {
+      "epoch": 0.11147116007777058,
+      "grad_norm": 1.477158546447754,
+      "learning_rate": 4.999333614311652e-06,
+      "loss": 0.1832,
+      "step": 688
+    },
+    {
+      "epoch": 0.11163318211276733,
+      "grad_norm": 1.374411940574646,
+      "learning_rate": 4.999323479770649e-06,
+      "loss": 0.1662,
+      "step": 689
+    },
+    {
+      "epoch": 0.1117952041477641,
+      "grad_norm": 1.3238674402236938,
+      "learning_rate": 4.999313268756396e-06,
+      "loss": 0.156,
+      "step": 690
+    },
+    {
+      "epoch": 0.11195722618276086,
+      "grad_norm": 1.4178396463394165,
+      "learning_rate": 4.999302981269204e-06,
+      "loss": 0.165,
+      "step": 691
+    },
+    {
+      "epoch": 0.11211924821775761,
+      "grad_norm": 1.5902644395828247,
+      "learning_rate": 4.99929261730939e-06,
+      "loss": 0.1938,
+      "step": 692
+    },
+    {
+      "epoch": 0.11228127025275438,
+      "grad_norm": 1.420102596282959,
+      "learning_rate": 4.999282176877271e-06,
+      "loss": 0.1818,
+      "step": 693
+    },
+    {
+      "epoch": 0.11244329228775113,
+      "grad_norm": 1.4864150285720825,
+      "learning_rate": 4.999271659973164e-06,
+      "loss": 0.1992,
+      "step": 694
+    },
+    {
+      "epoch": 0.1126053143227479,
+      "grad_norm": 1.5267049074172974,
+      "learning_rate": 4.999261066597393e-06,
+      "loss": 0.1995,
+      "step": 695
+    },
+    {
+      "epoch": 0.11276733635774465,
+      "grad_norm": 1.4048455953598022,
+      "learning_rate": 4.999250396750281e-06,
+      "loss": 0.1719,
+      "step": 696
+    },
+    {
+      "epoch": 0.11292935839274142,
+      "grad_norm": 1.4921525716781616,
+      "learning_rate": 4.999239650432155e-06,
+      "loss": 0.191,
+      "step": 697
+    },
+    {
+      "epoch": 0.11309138042773817,
+      "grad_norm": 1.41100013256073,
+      "learning_rate": 4.999228827643344e-06,
+      "loss": 0.1837,
+      "step": 698
+    },
+    {
+      "epoch": 0.11325340246273494,
+      "grad_norm": 1.3732647895812988,
+      "learning_rate": 4.999217928384179e-06,
+      "loss": 0.1936,
+      "step": 699
+    },
+    {
+      "epoch": 0.11341542449773169,
+      "grad_norm": 1.556138277053833,
+      "learning_rate": 4.999206952654993e-06,
+      "loss": 0.1966,
+      "step": 700
+    },
+    {
+      "epoch": 0.11357744653272846,
+      "grad_norm": 1.5490831136703491,
+      "learning_rate": 4.9991959004561225e-06,
+      "loss": 0.1914,
+      "step": 701
+    },
+    {
+      "epoch": 0.11373946856772521,
+      "grad_norm": 1.3613611459732056,
+      "learning_rate": 4.999184771787905e-06,
+      "loss": 0.1692,
+      "step": 702
+    },
+    {
+      "epoch": 0.11390149060272196,
+      "grad_norm": 1.4281145334243774,
+      "learning_rate": 4.999173566650682e-06,
+      "loss": 0.1671,
+      "step": 703
+    },
+    {
+      "epoch": 0.11406351263771873,
+      "grad_norm": 1.629227638244629,
+      "learning_rate": 4.999162285044795e-06,
+      "loss": 0.1739,
+      "step": 704
+    },
+    {
+      "epoch": 0.11422553467271548,
+      "grad_norm": 1.5268644094467163,
+      "learning_rate": 4.999150926970591e-06,
+      "loss": 0.1809,
+      "step": 705
+    },
+    {
+      "epoch": 0.11438755670771225,
+      "grad_norm": 1.445254921913147,
+      "learning_rate": 4.9991394924284155e-06,
+      "loss": 0.1684,
+      "step": 706
+    },
+    {
+      "epoch": 0.114549578742709,
+      "grad_norm": 1.533254861831665,
+      "learning_rate": 4.99912798141862e-06,
+      "loss": 0.1923,
+      "step": 707
+    },
+    {
+      "epoch": 0.11471160077770577,
+      "grad_norm": 1.3670967817306519,
+      "learning_rate": 4.999116393941556e-06,
+      "loss": 0.1548,
+      "step": 708
+    },
+    {
+      "epoch": 0.11487362281270252,
+      "grad_norm": 1.3976548910140991,
+      "learning_rate": 4.999104729997577e-06,
+      "loss": 0.1821,
+      "step": 709
+    },
+    {
+      "epoch": 0.11503564484769929,
+      "grad_norm": 1.4151368141174316,
+      "learning_rate": 4.999092989587042e-06,
+      "loss": 0.188,
+      "step": 710
+    },
+    {
+      "epoch": 0.11519766688269605,
+      "grad_norm": 1.2826131582260132,
+      "learning_rate": 4.999081172710309e-06,
+      "loss": 0.152,
+      "step": 711
+    },
+    {
+      "epoch": 0.11535968891769281,
+      "grad_norm": 1.463115930557251,
+      "learning_rate": 4.9990692793677395e-06,
+      "loss": 0.1712,
+      "step": 712
+    },
+    {
+      "epoch": 0.11552171095268957,
+      "grad_norm": 1.4324864149093628,
+      "learning_rate": 4.999057309559698e-06,
+      "loss": 0.1749,
+      "step": 713
+    },
+    {
+      "epoch": 0.11568373298768632,
+      "grad_norm": 1.2447808980941772,
+      "learning_rate": 4.999045263286551e-06,
+      "loss": 0.1539,
+      "step": 714
+    },
+    {
+      "epoch": 0.11584575502268309,
+      "grad_norm": 1.3643749952316284,
+      "learning_rate": 4.999033140548666e-06,
+      "loss": 0.165,
+      "step": 715
+    },
+    {
+      "epoch": 0.11600777705767984,
+      "grad_norm": 1.491412878036499,
+      "learning_rate": 4.999020941346414e-06,
+      "loss": 0.1752,
+      "step": 716
+    },
+    {
+      "epoch": 0.1161697990926766,
+      "grad_norm": 1.363407850265503,
+      "learning_rate": 4.999008665680169e-06,
+      "loss": 0.1652,
+      "step": 717
+    },
+    {
+      "epoch": 0.11633182112767336,
+      "grad_norm": 1.3879029750823975,
+      "learning_rate": 4.998996313550306e-06,
+      "loss": 0.1635,
+      "step": 718
+    },
+    {
+      "epoch": 0.11649384316267013,
+      "grad_norm": 1.5233573913574219,
+      "learning_rate": 4.9989838849572035e-06,
+      "loss": 0.175,
+      "step": 719
+    },
+    {
+      "epoch": 0.11665586519766688,
+      "grad_norm": 1.3929238319396973,
+      "learning_rate": 4.998971379901242e-06,
+      "loss": 0.1714,
+      "step": 720
+    },
+    {
+      "epoch": 0.11681788723266365,
+      "grad_norm": 1.4243627786636353,
+      "learning_rate": 4.9989587983828036e-06,
+      "loss": 0.1931,
+      "step": 721
+    },
+    {
+      "epoch": 0.1169799092676604,
+      "grad_norm": 1.6758102178573608,
+      "learning_rate": 4.998946140402273e-06,
+      "loss": 0.1698,
+      "step": 722
+    },
+    {
+      "epoch": 0.11714193130265717,
+      "grad_norm": 1.4985467195510864,
+      "learning_rate": 4.998933405960038e-06,
+      "loss": 0.194,
+      "step": 723
+    },
+    {
+      "epoch": 0.11730395333765392,
+      "grad_norm": 1.500422716140747,
+      "learning_rate": 4.998920595056488e-06,
+      "loss": 0.1828,
+      "step": 724
+    },
+    {
+      "epoch": 0.11746597537265067,
+      "grad_norm": 1.4340198040008545,
+      "learning_rate": 4.998907707692015e-06,
+      "loss": 0.1903,
+      "step": 725
+    },
+    {
+      "epoch": 0.11762799740764744,
+      "grad_norm": 1.3363415002822876,
+      "learning_rate": 4.998894743867013e-06,
+      "loss": 0.1711,
+      "step": 726
+    },
+    {
+      "epoch": 0.1177900194426442,
+      "grad_norm": 1.4468626976013184,
+      "learning_rate": 4.998881703581879e-06,
+      "loss": 0.17,
+      "step": 727
+    },
+    {
+      "epoch": 0.11795204147764096,
+      "grad_norm": 1.3693926334381104,
+      "learning_rate": 4.998868586837013e-06,
+      "loss": 0.183,
+      "step": 728
+    },
+    {
+      "epoch": 0.11811406351263772,
+      "grad_norm": 1.2546919584274292,
+      "learning_rate": 4.998855393632815e-06,
+      "loss": 0.1606,
+      "step": 729
+    },
+    {
+      "epoch": 0.11827608554763448,
+      "grad_norm": 1.3889665603637695,
+      "learning_rate": 4.998842123969689e-06,
+      "loss": 0.1593,
+      "step": 730
+    },
+    {
+      "epoch": 0.11843810758263124,
+      "grad_norm": 1.3932158946990967,
+      "learning_rate": 4.998828777848041e-06,
+      "loss": 0.1779,
+      "step": 731
+    },
+    {
+      "epoch": 0.118600129617628,
+      "grad_norm": 1.3757567405700684,
+      "learning_rate": 4.998815355268279e-06,
+      "loss": 0.1707,
+      "step": 732
+    },
+    {
+      "epoch": 0.11876215165262476,
+      "grad_norm": 1.5059839487075806,
+      "learning_rate": 4.998801856230815e-06,
+      "loss": 0.1737,
+      "step": 733
+    },
+    {
+      "epoch": 0.11892417368762152,
+      "grad_norm": 1.3769277334213257,
+      "learning_rate": 4.998788280736061e-06,
+      "loss": 0.1745,
+      "step": 734
+    },
+    {
+      "epoch": 0.11908619572261828,
+      "grad_norm": 1.4347563982009888,
+      "learning_rate": 4.998774628784432e-06,
+      "loss": 0.1948,
+      "step": 735
+    },
+    {
+      "epoch": 0.11924821775761503,
+      "grad_norm": 1.40513277053833,
+      "learning_rate": 4.998760900376347e-06,
+      "loss": 0.1795,
+      "step": 736
+    },
+    {
+      "epoch": 0.1194102397926118,
+      "grad_norm": 1.2432270050048828,
+      "learning_rate": 4.998747095512225e-06,
+      "loss": 0.1416,
+      "step": 737
+    },
+    {
+      "epoch": 0.11957226182760855,
+      "grad_norm": 1.4356383085250854,
+      "learning_rate": 4.99873321419249e-06,
+      "loss": 0.1925,
+      "step": 738
+    },
+    {
+      "epoch": 0.11973428386260532,
+      "grad_norm": 1.1989301443099976,
+      "learning_rate": 4.998719256417563e-06,
+      "loss": 0.1551,
+      "step": 739
+    },
+    {
+      "epoch": 0.11989630589760207,
+      "grad_norm": 1.339026689529419,
+      "learning_rate": 4.998705222187875e-06,
+      "loss": 0.1952,
+      "step": 740
+    },
+    {
+      "epoch": 0.12005832793259884,
+      "grad_norm": 1.408982276916504,
+      "learning_rate": 4.998691111503854e-06,
+      "loss": 0.1869,
+      "step": 741
+    },
+    {
+      "epoch": 0.12022034996759559,
+      "grad_norm": 1.28006112575531,
+      "learning_rate": 4.998676924365931e-06,
+      "loss": 0.1713,
+      "step": 742
+    },
+    {
+      "epoch": 0.12038237200259236,
+      "grad_norm": 1.423261046409607,
+      "learning_rate": 4.998662660774541e-06,
+      "loss": 0.1713,
+      "step": 743
+    },
+    {
+      "epoch": 0.12054439403758911,
+      "grad_norm": 1.3350955247879028,
+      "learning_rate": 4.998648320730121e-06,
+      "loss": 0.156,
+      "step": 744
+    },
+    {
+      "epoch": 0.12070641607258586,
+      "grad_norm": 1.3964903354644775,
+      "learning_rate": 4.998633904233108e-06,
+      "loss": 0.1827,
+      "step": 745
+    },
+    {
+      "epoch": 0.12086843810758263,
+      "grad_norm": 1.4213593006134033,
+      "learning_rate": 4.998619411283945e-06,
+      "loss": 0.1742,
+      "step": 746
+    },
+    {
+      "epoch": 0.12103046014257939,
+      "grad_norm": 1.2968708276748657,
+      "learning_rate": 4.998604841883073e-06,
+      "loss": 0.1623,
+      "step": 747
+    },
+    {
+      "epoch": 0.12119248217757615,
+      "grad_norm": 1.5439331531524658,
+      "learning_rate": 4.998590196030942e-06,
+      "loss": 0.2048,
+      "step": 748
+    },
+    {
+      "epoch": 0.1213545042125729,
+      "grad_norm": 1.4168784618377686,
+      "learning_rate": 4.998575473727995e-06,
+      "loss": 0.1814,
+      "step": 749
+    },
+    {
+      "epoch": 0.12151652624756967,
+      "grad_norm": 1.568098545074463,
+      "learning_rate": 4.998560674974686e-06,
+      "loss": 0.184,
+      "step": 750
+    },
+    {
+      "epoch": 0.12167854828256643,
+      "grad_norm": 1.4985617399215698,
+      "learning_rate": 4.998545799771466e-06,
+      "loss": 0.1938,
+      "step": 751
+    },
+    {
+      "epoch": 0.1218405703175632,
+      "grad_norm": 1.5311156511306763,
+      "learning_rate": 4.998530848118792e-06,
+      "loss": 0.1802,
+      "step": 752
+    },
+    {
+      "epoch": 0.12200259235255995,
+      "grad_norm": 1.499812126159668,
+      "learning_rate": 4.99851582001712e-06,
+      "loss": 0.1726,
+      "step": 753
+    },
+    {
+      "epoch": 0.12216461438755671,
+      "grad_norm": 1.4239556789398193,
+      "learning_rate": 4.99850071546691e-06,
+      "loss": 0.2014,
+      "step": 754
+    },
+    {
+      "epoch": 0.12232663642255347,
+      "grad_norm": 1.3375072479248047,
+      "learning_rate": 4.998485534468624e-06,
+      "loss": 0.176,
+      "step": 755
+    },
+    {
+      "epoch": 0.12248865845755022,
+      "grad_norm": 1.356996774673462,
+      "learning_rate": 4.998470277022728e-06,
+      "loss": 0.1734,
+      "step": 756
+    },
+    {
+      "epoch": 0.12265068049254699,
+      "grad_norm": 1.4485342502593994,
+      "learning_rate": 4.998454943129687e-06,
+      "loss": 0.1824,
+      "step": 757
+    },
+    {
+      "epoch": 0.12281270252754374,
+      "grad_norm": 1.4348751306533813,
+      "learning_rate": 4.99843953278997e-06,
+      "loss": 0.1857,
+      "step": 758
+    },
+    {
+      "epoch": 0.12297472456254051,
+      "grad_norm": 1.4200366735458374,
+      "learning_rate": 4.998424046004051e-06,
+      "loss": 0.1772,
+      "step": 759
+    },
+    {
+      "epoch": 0.12313674659753726,
+      "grad_norm": 1.4113013744354248,
+      "learning_rate": 4.998408482772401e-06,
+      "loss": 0.1823,
+      "step": 760
+    },
+    {
+      "epoch": 0.12329876863253403,
+      "grad_norm": 1.5233389139175415,
+      "learning_rate": 4.9983928430954986e-06,
+      "loss": 0.1756,
+      "step": 761
+    },
+    {
+      "epoch": 0.12346079066753078,
+      "grad_norm": 1.6090972423553467,
+      "learning_rate": 4.99837712697382e-06,
+      "loss": 0.1848,
+      "step": 762
+    },
+    {
+      "epoch": 0.12362281270252755,
+      "grad_norm": 1.318171501159668,
+      "learning_rate": 4.998361334407849e-06,
+      "loss": 0.1513,
+      "step": 763
+    },
+    {
+      "epoch": 0.1237848347375243,
+      "grad_norm": 1.4973655939102173,
+      "learning_rate": 4.998345465398066e-06,
+      "loss": 0.1755,
+      "step": 764
+    },
+    {
+      "epoch": 0.12394685677252107,
+      "grad_norm": 1.4733107089996338,
+      "learning_rate": 4.998329519944957e-06,
+      "loss": 0.1767,
+      "step": 765
+    },
+    {
+      "epoch": 0.12410887880751782,
+      "grad_norm": 1.373856544494629,
+      "learning_rate": 4.998313498049011e-06,
+      "loss": 0.1895,
+      "step": 766
+    },
+    {
+      "epoch": 0.12427090084251458,
+      "grad_norm": 1.2825859785079956,
+      "learning_rate": 4.998297399710718e-06,
+      "loss": 0.1642,
+      "step": 767
+    },
+    {
+      "epoch": 0.12443292287751134,
+      "grad_norm": 1.416802167892456,
+      "learning_rate": 4.9982812249305704e-06,
+      "loss": 0.1892,
+      "step": 768
+    },
+    {
+      "epoch": 0.1245949449125081,
+      "grad_norm": 1.3358246088027954,
+      "learning_rate": 4.998264973709063e-06,
+      "loss": 0.1655,
+      "step": 769
+    },
+    {
+      "epoch": 0.12475696694750486,
+      "grad_norm": 1.3493046760559082,
+      "learning_rate": 4.998248646046693e-06,
+      "loss": 0.1602,
+      "step": 770
+    },
+    {
+      "epoch": 0.12491898898250162,
+      "grad_norm": 1.3179484605789185,
+      "learning_rate": 4.99823224194396e-06,
+      "loss": 0.1788,
+      "step": 771
+    },
+    {
+      "epoch": 0.12508101101749838,
+      "grad_norm": 1.3762329816818237,
+      "learning_rate": 4.998215761401366e-06,
+      "loss": 0.1727,
+      "step": 772
+    },
+    {
+      "epoch": 0.12524303305249515,
+      "grad_norm": 1.5525070428848267,
+      "learning_rate": 4.998199204419415e-06,
+      "loss": 0.1876,
+      "step": 773
+    },
+    {
+      "epoch": 0.1254050550874919,
+      "grad_norm": 1.4245859384536743,
+      "learning_rate": 4.9981825709986145e-06,
+      "loss": 0.1737,
+      "step": 774
+    },
+    {
+      "epoch": 0.12556707712248866,
+      "grad_norm": 1.5142040252685547,
+      "learning_rate": 4.998165861139472e-06,
+      "loss": 0.1775,
+      "step": 775
+    },
+    {
+      "epoch": 0.12572909915748542,
+      "grad_norm": 1.4016090631484985,
+      "learning_rate": 4.9981490748425e-06,
+      "loss": 0.1676,
+      "step": 776
+    },
+    {
+      "epoch": 0.12589112119248216,
+      "grad_norm": 1.5504176616668701,
+      "learning_rate": 4.998132212108212e-06,
+      "loss": 0.1861,
+      "step": 777
+    },
+    {
+      "epoch": 0.12605314322747893,
+      "grad_norm": 1.3723235130310059,
+      "learning_rate": 4.998115272937123e-06,
+      "loss": 0.1657,
+      "step": 778
+    },
+    {
+      "epoch": 0.1262151652624757,
+      "grad_norm": 1.4091416597366333,
+      "learning_rate": 4.998098257329753e-06,
+      "loss": 0.1698,
+      "step": 779
+    },
+    {
+      "epoch": 0.12637718729747247,
+      "grad_norm": 1.3922086954116821,
+      "learning_rate": 4.998081165286621e-06,
+      "loss": 0.1828,
+      "step": 780
+    },
+    {
+      "epoch": 0.1265392093324692,
+      "grad_norm": 1.3854937553405762,
+      "learning_rate": 4.998063996808251e-06,
+      "loss": 0.1805,
+      "step": 781
+    },
+    {
+      "epoch": 0.12670123136746597,
+      "grad_norm": 1.359626293182373,
+      "learning_rate": 4.9980467518951666e-06,
+      "loss": 0.1816,
+      "step": 782
+    },
+    {
+      "epoch": 0.12686325340246274,
+      "grad_norm": 1.2333866357803345,
+      "learning_rate": 4.998029430547898e-06,
+      "loss": 0.1541,
+      "step": 783
+    },
+    {
+      "epoch": 0.1270252754374595,
+      "grad_norm": 1.4757193326950073,
+      "learning_rate": 4.998012032766974e-06,
+      "loss": 0.1925,
+      "step": 784
+    },
+    {
+      "epoch": 0.12718729747245625,
+      "grad_norm": 1.434784173965454,
+      "learning_rate": 4.997994558552926e-06,
+      "loss": 0.1815,
+      "step": 785
+    },
+    {
+      "epoch": 0.127349319507453,
+      "grad_norm": 1.4703211784362793,
+      "learning_rate": 4.997977007906291e-06,
+      "loss": 0.1768,
+      "step": 786
+    },
+    {
+      "epoch": 0.12751134154244978,
+      "grad_norm": 1.414862871170044,
+      "learning_rate": 4.997959380827603e-06,
+      "loss": 0.1674,
+      "step": 787
+    },
+    {
+      "epoch": 0.12767336357744652,
+      "grad_norm": 1.548573613166809,
+      "learning_rate": 4.997941677317403e-06,
+      "loss": 0.197,
+      "step": 788
+    },
+    {
+      "epoch": 0.1278353856124433,
+      "grad_norm": 1.4064699411392212,
+      "learning_rate": 4.997923897376233e-06,
+      "loss": 0.1857,
+      "step": 789
+    },
+    {
+      "epoch": 0.12799740764744005,
+      "grad_norm": 1.2528257369995117,
+      "learning_rate": 4.997906041004637e-06,
+      "loss": 0.1574,
+      "step": 790
+    },
+    {
+      "epoch": 0.12815942968243682,
+      "grad_norm": 1.3743071556091309,
+      "learning_rate": 4.9978881082031605e-06,
+      "loss": 0.1833,
+      "step": 791
+    },
+    {
+      "epoch": 0.12832145171743356,
+      "grad_norm": 1.437005639076233,
+      "learning_rate": 4.997870098972353e-06,
+      "loss": 0.1712,
+      "step": 792
+    },
+    {
+      "epoch": 0.12848347375243033,
+      "grad_norm": 1.3489969968795776,
+      "learning_rate": 4.997852013312765e-06,
+      "loss": 0.1772,
+      "step": 793
+    },
+    {
+      "epoch": 0.1286454957874271,
+      "grad_norm": 1.6645337343215942,
+      "learning_rate": 4.99783385122495e-06,
+      "loss": 0.1913,
+      "step": 794
+    },
+    {
+      "epoch": 0.12880751782242386,
+      "grad_norm": 1.3913012742996216,
+      "learning_rate": 4.997815612709463e-06,
+      "loss": 0.1605,
+      "step": 795
+    },
+    {
+      "epoch": 0.1289695398574206,
+      "grad_norm": 1.3065690994262695,
+      "learning_rate": 4.997797297766864e-06,
+      "loss": 0.1604,
+      "step": 796
+    },
+    {
+      "epoch": 0.12913156189241737,
+      "grad_norm": 1.439260721206665,
+      "learning_rate": 4.997778906397713e-06,
+      "loss": 0.1689,
+      "step": 797
+    },
+    {
+      "epoch": 0.12929358392741414,
+      "grad_norm": 1.3989150524139404,
+      "learning_rate": 4.9977604386025704e-06,
+      "loss": 0.1724,
+      "step": 798
+    },
+    {
+      "epoch": 0.12945560596241087,
+      "grad_norm": 1.5063525438308716,
+      "learning_rate": 4.9977418943820036e-06,
+      "loss": 0.1865,
+      "step": 799
+    },
+    {
+      "epoch": 0.12961762799740764,
+      "grad_norm": 1.3701797723770142,
+      "learning_rate": 4.997723273736579e-06,
+      "loss": 0.1755,
+      "step": 800
+    },
+    {
+      "epoch": 0.1297796500324044,
+      "grad_norm": 2.1302490234375,
+      "learning_rate": 4.997704576666867e-06,
+      "loss": 0.1756,
+      "step": 801
+    },
+    {
+      "epoch": 0.12994167206740118,
+      "grad_norm": 1.360653042793274,
+      "learning_rate": 4.9976858031734375e-06,
+      "loss": 0.1637,
+      "step": 802
+    },
+    {
+      "epoch": 0.13010369410239792,
+      "grad_norm": 1.3926098346710205,
+      "learning_rate": 4.997666953256869e-06,
+      "loss": 0.1664,
+      "step": 803
+    },
+    {
+      "epoch": 0.13026571613739468,
+      "grad_norm": 1.2852853536605835,
+      "learning_rate": 4.9976480269177345e-06,
+      "loss": 0.1714,
+      "step": 804
+    },
+    {
+      "epoch": 0.13042773817239145,
+      "grad_norm": 1.3100239038467407,
+      "learning_rate": 4.997629024156615e-06,
+      "loss": 0.167,
+      "step": 805
+    },
+    {
+      "epoch": 0.13058976020738822,
+      "grad_norm": 1.3834192752838135,
+      "learning_rate": 4.997609944974092e-06,
+      "loss": 0.1626,
+      "step": 806
+    },
+    {
+      "epoch": 0.13075178224238496,
+      "grad_norm": 1.1783868074417114,
+      "learning_rate": 4.997590789370749e-06,
+      "loss": 0.1597,
+      "step": 807
+    },
+    {
+      "epoch": 0.13091380427738172,
+      "grad_norm": 1.45823073387146,
+      "learning_rate": 4.99757155734717e-06,
+      "loss": 0.1965,
+      "step": 808
+    },
+    {
+      "epoch": 0.1310758263123785,
+      "grad_norm": 1.4000074863433838,
+      "learning_rate": 4.9975522489039474e-06,
+      "loss": 0.1707,
+      "step": 809
+    },
+    {
+      "epoch": 0.13123784834737523,
+      "grad_norm": 1.4006067514419556,
+      "learning_rate": 4.997532864041669e-06,
+      "loss": 0.1683,
+      "step": 810
+    },
+    {
+      "epoch": 0.131399870382372,
+      "grad_norm": 1.3444918394088745,
+      "learning_rate": 4.99751340276093e-06,
+      "loss": 0.1902,
+      "step": 811
+    },
+    {
+      "epoch": 0.13156189241736876,
+      "grad_norm": 1.2597421407699585,
+      "learning_rate": 4.997493865062323e-06,
+      "loss": 0.1666,
+      "step": 812
+    },
+    {
+      "epoch": 0.13172391445236553,
+      "grad_norm": 1.7309461832046509,
+      "learning_rate": 4.997474250946448e-06,
+      "loss": 0.1891,
+      "step": 813
+    },
+    {
+      "epoch": 0.13188593648736227,
+      "grad_norm": 1.5406413078308105,
+      "learning_rate": 4.9974545604139055e-06,
+      "loss": 0.1793,
+      "step": 814
+    },
+    {
+      "epoch": 0.13204795852235904,
+      "grad_norm": 1.402835726737976,
+      "learning_rate": 4.9974347934652965e-06,
+      "loss": 0.1846,
+      "step": 815
+    },
+    {
+      "epoch": 0.1322099805573558,
+      "grad_norm": 1.3662683963775635,
+      "learning_rate": 4.997414950101227e-06,
+      "loss": 0.1669,
+      "step": 816
+    },
+    {
+      "epoch": 0.13237200259235257,
+      "grad_norm": 1.3234673738479614,
+      "learning_rate": 4.9973950303223026e-06,
+      "loss": 0.1713,
+      "step": 817
+    },
+    {
+      "epoch": 0.1325340246273493,
+      "grad_norm": 1.6133899688720703,
+      "learning_rate": 4.997375034129135e-06,
+      "loss": 0.1928,
+      "step": 818
+    },
+    {
+      "epoch": 0.13269604666234608,
+      "grad_norm": 1.3357648849487305,
+      "learning_rate": 4.997354961522335e-06,
+      "loss": 0.1872,
+      "step": 819
+    },
+    {
+      "epoch": 0.13285806869734285,
+      "grad_norm": 1.4020048379898071,
+      "learning_rate": 4.997334812502516e-06,
+      "loss": 0.1831,
+      "step": 820
+    },
+    {
+      "epoch": 0.13302009073233959,
+      "grad_norm": 1.350357174873352,
+      "learning_rate": 4.997314587070295e-06,
+      "loss": 0.166,
+      "step": 821
+    },
+    {
+      "epoch": 0.13318211276733635,
+      "grad_norm": 1.5430233478546143,
+      "learning_rate": 4.9972942852262915e-06,
+      "loss": 0.1769,
+      "step": 822
+    },
+    {
+      "epoch": 0.13334413480233312,
+      "grad_norm": 1.4370311498641968,
+      "learning_rate": 4.997273906971126e-06,
+      "loss": 0.1876,
+      "step": 823
+    },
+    {
+      "epoch": 0.1335061568373299,
+      "grad_norm": 1.2412606477737427,
+      "learning_rate": 4.997253452305423e-06,
+      "loss": 0.1494,
+      "step": 824
+    },
+    {
+      "epoch": 0.13366817887232663,
+      "grad_norm": 1.3403682708740234,
+      "learning_rate": 4.9972329212298065e-06,
+      "loss": 0.18,
+      "step": 825
+    },
+    {
+      "epoch": 0.1338302009073234,
+      "grad_norm": 1.2844048738479614,
+      "learning_rate": 4.9972123137449065e-06,
+      "loss": 0.1618,
+      "step": 826
+    },
+    {
+      "epoch": 0.13399222294232016,
+      "grad_norm": 1.5786724090576172,
+      "learning_rate": 4.997191629851352e-06,
+      "loss": 0.1946,
+      "step": 827
+    },
+    {
+      "epoch": 0.13415424497731693,
+      "grad_norm": 1.384547472000122,
+      "learning_rate": 4.997170869549778e-06,
+      "loss": 0.1725,
+      "step": 828
+    },
+    {
+      "epoch": 0.13431626701231367,
+      "grad_norm": 1.4201945066452026,
+      "learning_rate": 4.997150032840818e-06,
+      "loss": 0.1766,
+      "step": 829
+    },
+    {
+      "epoch": 0.13447828904731043,
+      "grad_norm": 1.369523048400879,
+      "learning_rate": 4.99712911972511e-06,
+      "loss": 0.1735,
+      "step": 830
+    },
+    {
+      "epoch": 0.1346403110823072,
+      "grad_norm": 1.2436349391937256,
+      "learning_rate": 4.997108130203293e-06,
+      "loss": 0.1599,
+      "step": 831
+    },
+    {
+      "epoch": 0.13480233311730394,
+      "grad_norm": 1.3385329246520996,
+      "learning_rate": 4.99708706427601e-06,
+      "loss": 0.1676,
+      "step": 832
+    },
+    {
+      "epoch": 0.1349643551523007,
+      "grad_norm": 1.412265419960022,
+      "learning_rate": 4.997065921943907e-06,
+      "loss": 0.1745,
+      "step": 833
+    },
+    {
+      "epoch": 0.13512637718729748,
+      "grad_norm": 1.3985753059387207,
+      "learning_rate": 4.997044703207629e-06,
+      "loss": 0.154,
+      "step": 834
+    },
+    {
+      "epoch": 0.13528839922229424,
+      "grad_norm": 1.3470317125320435,
+      "learning_rate": 4.9970234080678244e-06,
+      "loss": 0.1806,
+      "step": 835
+    },
+    {
+      "epoch": 0.13545042125729098,
+      "grad_norm": 1.3285126686096191,
+      "learning_rate": 4.9970020365251485e-06,
+      "loss": 0.179,
+      "step": 836
+    },
+    {
+      "epoch": 0.13561244329228775,
+      "grad_norm": 1.4892990589141846,
+      "learning_rate": 4.9969805885802515e-06,
+      "loss": 0.1902,
+      "step": 837
+    },
+    {
+      "epoch": 0.13577446532728452,
+      "grad_norm": 1.373530626296997,
+      "learning_rate": 4.996959064233792e-06,
+      "loss": 0.1539,
+      "step": 838
+    },
+    {
+      "epoch": 0.13593648736228128,
+      "grad_norm": 1.3297441005706787,
+      "learning_rate": 4.996937463486427e-06,
+      "loss": 0.1704,
+      "step": 839
+    },
+    {
+      "epoch": 0.13609850939727802,
+      "grad_norm": 1.343121886253357,
+      "learning_rate": 4.996915786338818e-06,
+      "loss": 0.1592,
+      "step": 840
+    },
+    {
+      "epoch": 0.1362605314322748,
+      "grad_norm": 1.4662681818008423,
+      "learning_rate": 4.9968940327916284e-06,
+      "loss": 0.1863,
+      "step": 841
+    },
+    {
+      "epoch": 0.13642255346727156,
+      "grad_norm": 1.4304791688919067,
+      "learning_rate": 4.9968722028455245e-06,
+      "loss": 0.1775,
+      "step": 842
+    },
+    {
+      "epoch": 0.1365845755022683,
+      "grad_norm": 1.3115490674972534,
+      "learning_rate": 4.996850296501172e-06,
+      "loss": 0.163,
+      "step": 843
+    },
+    {
+      "epoch": 0.13674659753726506,
+      "grad_norm": 1.5014792680740356,
+      "learning_rate": 4.996828313759245e-06,
+      "loss": 0.1856,
+      "step": 844
+    },
+    {
+      "epoch": 0.13690861957226183,
+      "grad_norm": 1.2926596403121948,
+      "learning_rate": 4.996806254620411e-06,
+      "loss": 0.1664,
+      "step": 845
+    },
+    {
+      "epoch": 0.1370706416072586,
+      "grad_norm": 1.501706600189209,
+      "learning_rate": 4.99678411908535e-06,
+      "loss": 0.1922,
+      "step": 846
+    },
+    {
+      "epoch": 0.13723266364225534,
+      "grad_norm": 1.2282168865203857,
+      "learning_rate": 4.996761907154736e-06,
+      "loss": 0.1679,
+      "step": 847
+    },
+    {
+      "epoch": 0.1373946856772521,
+      "grad_norm": 1.315477967262268,
+      "learning_rate": 4.996739618829251e-06,
+      "loss": 0.1789,
+      "step": 848
+    },
+    {
+      "epoch": 0.13755670771224887,
+      "grad_norm": 1.2631007432937622,
+      "learning_rate": 4.996717254109574e-06,
+      "loss": 0.1654,
+      "step": 849
+    },
+    {
+      "epoch": 0.13771872974724564,
+      "grad_norm": 1.4962035417556763,
+      "learning_rate": 4.996694812996391e-06,
+      "loss": 0.2064,
+      "step": 850
+    },
+    {
+      "epoch": 0.13788075178224238,
+      "grad_norm": 1.2344715595245361,
+      "learning_rate": 4.99667229549039e-06,
+      "loss": 0.163,
+      "step": 851
+    },
+    {
+      "epoch": 0.13804277381723914,
+      "grad_norm": 1.3445398807525635,
+      "learning_rate": 4.996649701592258e-06,
+      "loss": 0.1601,
+      "step": 852
+    },
+    {
+      "epoch": 0.1382047958522359,
+      "grad_norm": 1.3226261138916016,
+      "learning_rate": 4.996627031302686e-06,
+      "loss": 0.1853,
+      "step": 853
+    },
+    {
+      "epoch": 0.13836681788723265,
+      "grad_norm": 1.4674776792526245,
+      "learning_rate": 4.99660428462237e-06,
+      "loss": 0.2106,
+      "step": 854
+    },
+    {
+      "epoch": 0.13852883992222942,
+      "grad_norm": 1.2431912422180176,
+      "learning_rate": 4.996581461552003e-06,
+      "loss": 0.1622,
+      "step": 855
+    },
+    {
+      "epoch": 0.13869086195722619,
+      "grad_norm": 1.3208400011062622,
+      "learning_rate": 4.996558562092286e-06,
+      "loss": 0.1614,
+      "step": 856
+    },
+    {
+      "epoch": 0.13885288399222295,
+      "grad_norm": 1.3402066230773926,
+      "learning_rate": 4.996535586243918e-06,
+      "loss": 0.1581,
+      "step": 857
+    },
+    {
+      "epoch": 0.1390149060272197,
+      "grad_norm": 1.3583393096923828,
+      "learning_rate": 4.996512534007602e-06,
+      "loss": 0.1659,
+      "step": 858
+    },
+    {
+      "epoch": 0.13917692806221646,
+      "grad_norm": 1.4143537282943726,
+      "learning_rate": 4.9964894053840455e-06,
+      "loss": 0.1879,
+      "step": 859
+    },
+    {
+      "epoch": 0.13933895009721323,
+      "grad_norm": 1.3818471431732178,
+      "learning_rate": 4.996466200373954e-06,
+      "loss": 0.182,
+      "step": 860
+    },
+    {
+      "epoch": 0.13950097213221,
+      "grad_norm": 1.3754279613494873,
+      "learning_rate": 4.996442918978038e-06,
+      "loss": 0.1816,
+      "step": 861
+    },
+    {
+      "epoch": 0.13966299416720673,
+      "grad_norm": 1.2652978897094727,
+      "learning_rate": 4.99641956119701e-06,
+      "loss": 0.1495,
+      "step": 862
+    },
+    {
+      "epoch": 0.1398250162022035,
+      "grad_norm": 1.4501652717590332,
+      "learning_rate": 4.996396127031584e-06,
+      "loss": 0.1854,
+      "step": 863
+    },
+    {
+      "epoch": 0.13998703823720027,
+      "grad_norm": 1.2763866186141968,
+      "learning_rate": 4.996372616482478e-06,
+      "loss": 0.1654,
+      "step": 864
+    },
+    {
+      "epoch": 0.140149060272197,
+      "grad_norm": 1.335204839706421,
+      "learning_rate": 4.996349029550411e-06,
+      "loss": 0.1533,
+      "step": 865
+    },
+    {
+      "epoch": 0.14031108230719377,
+      "grad_norm": 1.3689930438995361,
+      "learning_rate": 4.996325366236105e-06,
+      "loss": 0.1826,
+      "step": 866
+    },
+    {
+      "epoch": 0.14047310434219054,
+      "grad_norm": 1.2703019380569458,
+      "learning_rate": 4.996301626540284e-06,
+      "loss": 0.1686,
+      "step": 867
+    },
+    {
+      "epoch": 0.1406351263771873,
+      "grad_norm": 1.4817618131637573,
+      "learning_rate": 4.996277810463675e-06,
+      "loss": 0.2053,
+      "step": 868
+    },
+    {
+      "epoch": 0.14079714841218405,
+      "grad_norm": 1.4522895812988281,
+      "learning_rate": 4.996253918007004e-06,
+      "loss": 0.1929,
+      "step": 869
+    },
+    {
+      "epoch": 0.14095917044718081,
+      "grad_norm": 1.3986607789993286,
+      "learning_rate": 4.996229949171004e-06,
+      "loss": 0.1761,
+      "step": 870
+    },
+    {
+      "epoch": 0.14112119248217758,
+      "grad_norm": 1.342969298362732,
+      "learning_rate": 4.996205903956409e-06,
+      "loss": 0.177,
+      "step": 871
+    },
+    {
+      "epoch": 0.14128321451717435,
+      "grad_norm": 1.3436371088027954,
+      "learning_rate": 4.996181782363955e-06,
+      "loss": 0.1776,
+      "step": 872
+    },
+    {
+      "epoch": 0.1414452365521711,
+      "grad_norm": 1.4952949285507202,
+      "learning_rate": 4.996157584394378e-06,
+      "loss": 0.1983,
+      "step": 873
+    },
+    {
+      "epoch": 0.14160725858716786,
+      "grad_norm": 1.2056357860565186,
+      "learning_rate": 4.99613331004842e-06,
+      "loss": 0.1559,
+      "step": 874
+    },
+    {
+      "epoch": 0.14176928062216462,
+      "grad_norm": 1.2493481636047363,
+      "learning_rate": 4.996108959326823e-06,
+      "loss": 0.1677,
+      "step": 875
+    },
+    {
+      "epoch": 0.14193130265716136,
+      "grad_norm": 1.348042607307434,
+      "learning_rate": 4.996084532230332e-06,
+      "loss": 0.1691,
+      "step": 876
+    },
+    {
+      "epoch": 0.14209332469215813,
+      "grad_norm": 1.3877140283584595,
+      "learning_rate": 4.996060028759695e-06,
+      "loss": 0.1829,
+      "step": 877
+    },
+    {
+      "epoch": 0.1422553467271549,
+      "grad_norm": 1.3358656167984009,
+      "learning_rate": 4.996035448915661e-06,
+      "loss": 0.1798,
+      "step": 878
+    },
+    {
+      "epoch": 0.14241736876215166,
+      "grad_norm": 1.5458526611328125,
+      "learning_rate": 4.996010792698983e-06,
+      "loss": 0.2146,
+      "step": 879
+    },
+    {
+      "epoch": 0.1425793907971484,
+      "grad_norm": 1.238255500793457,
+      "learning_rate": 4.995986060110415e-06,
+      "loss": 0.1554,
+      "step": 880
+    },
+    {
+      "epoch": 0.14274141283214517,
+      "grad_norm": 1.332350730895996,
+      "learning_rate": 4.995961251150714e-06,
+      "loss": 0.1563,
+      "step": 881
+    },
+    {
+      "epoch": 0.14290343486714194,
+      "grad_norm": 1.4522160291671753,
+      "learning_rate": 4.995936365820638e-06,
+      "loss": 0.194,
+      "step": 882
+    },
+    {
+      "epoch": 0.1430654569021387,
+      "grad_norm": 1.2477561235427856,
+      "learning_rate": 4.99591140412095e-06,
+      "loss": 0.1582,
+      "step": 883
+    },
+    {
+      "epoch": 0.14322747893713544,
+      "grad_norm": 1.5550086498260498,
+      "learning_rate": 4.9958863660524125e-06,
+      "loss": 0.1858,
+      "step": 884
+    },
+    {
+      "epoch": 0.1433895009721322,
+      "grad_norm": 1.1642705202102661,
+      "learning_rate": 4.995861251615792e-06,
+      "loss": 0.1353,
+      "step": 885
+    },
+    {
+      "epoch": 0.14355152300712898,
+      "grad_norm": 1.1860448122024536,
+      "learning_rate": 4.995836060811859e-06,
+      "loss": 0.1516,
+      "step": 886
+    },
+    {
+      "epoch": 0.14371354504212572,
+      "grad_norm": 1.2857884168624878,
+      "learning_rate": 4.99581079364138e-06,
+      "loss": 0.165,
+      "step": 887
+    },
+    {
+      "epoch": 0.14387556707712248,
+      "grad_norm": 1.3009867668151855,
+      "learning_rate": 4.995785450105131e-06,
+      "loss": 0.1873,
+      "step": 888
+    },
+    {
+      "epoch": 0.14403758911211925,
+      "grad_norm": 1.2586079835891724,
+      "learning_rate": 4.995760030203888e-06,
+      "loss": 0.1622,
+      "step": 889
+    },
+    {
+      "epoch": 0.14419961114711602,
+      "grad_norm": 1.3150060176849365,
+      "learning_rate": 4.995734533938427e-06,
+      "loss": 0.1756,
+      "step": 890
+    },
+    {
+      "epoch": 0.14436163318211276,
+      "grad_norm": 1.2379282712936401,
+      "learning_rate": 4.995708961309528e-06,
+      "loss": 0.1527,
+      "step": 891
+    },
+    {
+      "epoch": 0.14452365521710953,
+      "grad_norm": 1.3145521879196167,
+      "learning_rate": 4.995683312317975e-06,
+      "loss": 0.1756,
+      "step": 892
+    },
+    {
+      "epoch": 0.1446856772521063,
+      "grad_norm": 1.420900583267212,
+      "learning_rate": 4.9956575869645515e-06,
+      "loss": 0.1738,
+      "step": 893
+    },
+    {
+      "epoch": 0.14484769928710303,
+      "grad_norm": 1.2677338123321533,
+      "learning_rate": 4.995631785250046e-06,
+      "loss": 0.1759,
+      "step": 894
+    },
+    {
+      "epoch": 0.1450097213220998,
+      "grad_norm": 1.1973481178283691,
+      "learning_rate": 4.995605907175247e-06,
+      "loss": 0.1592,
+      "step": 895
+    },
+    {
+      "epoch": 0.14517174335709657,
+      "grad_norm": 1.342222809791565,
+      "learning_rate": 4.9955799527409465e-06,
+      "loss": 0.1733,
+      "step": 896
+    },
+    {
+      "epoch": 0.14533376539209333,
+      "grad_norm": 1.3735847473144531,
+      "learning_rate": 4.995553921947938e-06,
+      "loss": 0.1729,
+      "step": 897
+    },
+    {
+      "epoch": 0.14549578742709007,
+      "grad_norm": 1.2551465034484863,
+      "learning_rate": 4.99552781479702e-06,
+      "loss": 0.1819,
+      "step": 898
+    },
+    {
+      "epoch": 0.14565780946208684,
+      "grad_norm": 1.5449222326278687,
+      "learning_rate": 4.995501631288989e-06,
+      "loss": 0.1913,
+      "step": 899
+    },
+    {
+      "epoch": 0.1458198314970836,
+      "grad_norm": 1.166682243347168,
+      "learning_rate": 4.995475371424648e-06,
+      "loss": 0.1512,
+      "step": 900
+    },
+    {
+      "epoch": 0.14598185353208037,
+      "grad_norm": 1.2635172605514526,
+      "learning_rate": 4.995449035204798e-06,
+      "loss": 0.1575,
+      "step": 901
+    },
+    {
+      "epoch": 0.14614387556707711,
+      "grad_norm": 1.3577876091003418,
+      "learning_rate": 4.995422622630247e-06,
+      "loss": 0.1916,
+      "step": 902
+    },
+    {
+      "epoch": 0.14630589760207388,
+      "grad_norm": 1.3075361251831055,
+      "learning_rate": 4.995396133701803e-06,
+      "loss": 0.1454,
+      "step": 903
+    },
+    {
+      "epoch": 0.14646791963707065,
+      "grad_norm": 1.339972972869873,
+      "learning_rate": 4.995369568420276e-06,
+      "loss": 0.1741,
+      "step": 904
+    },
+    {
+      "epoch": 0.1466299416720674,
+      "grad_norm": 1.459546685218811,
+      "learning_rate": 4.995342926786478e-06,
+      "loss": 0.1882,
+      "step": 905
+    },
+    {
+      "epoch": 0.14679196370706415,
+      "grad_norm": 1.2182546854019165,
+      "learning_rate": 4.995316208801226e-06,
+      "loss": 0.1348,
+      "step": 906
+    },
+    {
+      "epoch": 0.14695398574206092,
+      "grad_norm": 1.376412272453308,
+      "learning_rate": 4.995289414465337e-06,
+      "loss": 0.1798,
+      "step": 907
+    },
+    {
+      "epoch": 0.1471160077770577,
+      "grad_norm": 1.489270806312561,
+      "learning_rate": 4.99526254377963e-06,
+      "loss": 0.1869,
+      "step": 908
+    },
+    {
+      "epoch": 0.14727802981205443,
+      "grad_norm": 1.320319652557373,
+      "learning_rate": 4.9952355967449265e-06,
+      "loss": 0.1671,
+      "step": 909
+    },
+    {
+      "epoch": 0.1474400518470512,
+      "grad_norm": 1.327093243598938,
+      "learning_rate": 4.995208573362053e-06,
+      "loss": 0.1765,
+      "step": 910
+    },
+    {
+      "epoch": 0.14760207388204796,
+      "grad_norm": 1.3886643648147583,
+      "learning_rate": 4.995181473631836e-06,
+      "loss": 0.1717,
+      "step": 911
+    },
+    {
+      "epoch": 0.14776409591704473,
+      "grad_norm": 1.3381421566009521,
+      "learning_rate": 4.995154297555103e-06,
+      "loss": 0.1716,
+      "step": 912
+    },
+    {
+      "epoch": 0.14792611795204147,
+      "grad_norm": 1.1352941989898682,
+      "learning_rate": 4.995127045132687e-06,
+      "loss": 0.1547,
+      "step": 913
+    },
+    {
+      "epoch": 0.14808813998703824,
+      "grad_norm": 1.4500410556793213,
+      "learning_rate": 4.995099716365421e-06,
+      "loss": 0.1913,
+      "step": 914
+    },
+    {
+      "epoch": 0.148250162022035,
+      "grad_norm": 1.28703773021698,
+      "learning_rate": 4.995072311254144e-06,
+      "loss": 0.18,
+      "step": 915
+    },
+    {
+      "epoch": 0.14841218405703174,
+      "grad_norm": 1.2766846418380737,
+      "learning_rate": 4.995044829799689e-06,
+      "loss": 0.1769,
+      "step": 916
+    },
+    {
+      "epoch": 0.1485742060920285,
+      "grad_norm": 1.2691501379013062,
+      "learning_rate": 4.995017272002902e-06,
+      "loss": 0.1684,
+      "step": 917
+    },
+    {
+      "epoch": 0.14873622812702528,
+      "grad_norm": 1.4125738143920898,
+      "learning_rate": 4.994989637864624e-06,
+      "loss": 0.1809,
+      "step": 918
+    },
+    {
+      "epoch": 0.14889825016202204,
+      "grad_norm": 1.3057345151901245,
+      "learning_rate": 4.994961927385701e-06,
+      "loss": 0.1608,
+      "step": 919
+    },
+    {
+      "epoch": 0.14906027219701878,
+      "grad_norm": 1.4931777715682983,
+      "learning_rate": 4.99493414056698e-06,
+      "loss": 0.1883,
+      "step": 920
+    },
+    {
+      "epoch": 0.14922229423201555,
+      "grad_norm": 1.4381946325302124,
+      "learning_rate": 4.994906277409313e-06,
+      "loss": 0.1773,
+      "step": 921
+    },
+    {
+      "epoch": 0.14938431626701232,
+      "grad_norm": 1.3123935461044312,
+      "learning_rate": 4.99487833791355e-06,
+      "loss": 0.159,
+      "step": 922
+    },
+    {
+      "epoch": 0.14954633830200909,
+      "grad_norm": 1.396132469177246,
+      "learning_rate": 4.994850322080549e-06,
+      "loss": 0.1704,
+      "step": 923
+    },
+    {
+      "epoch": 0.14970836033700582,
+      "grad_norm": 1.267357349395752,
+      "learning_rate": 4.9948222299111644e-06,
+      "loss": 0.1605,
+      "step": 924
+    },
+    {
+      "epoch": 0.1498703823720026,
+      "grad_norm": 1.2908340692520142,
+      "learning_rate": 4.994794061406258e-06,
+      "loss": 0.1625,
+      "step": 925
+    },
+    {
+      "epoch": 0.15003240440699936,
+      "grad_norm": 1.3178255558013916,
+      "learning_rate": 4.994765816566689e-06,
+      "loss": 0.176,
+      "step": 926
+    },
+    {
+      "epoch": 0.1501944264419961,
+      "grad_norm": 1.2020256519317627,
+      "learning_rate": 4.994737495393325e-06,
+      "loss": 0.1565,
+      "step": 927
+    },
+    {
+      "epoch": 0.15035644847699287,
+      "grad_norm": 1.346628189086914,
+      "learning_rate": 4.994709097887029e-06,
+      "loss": 0.1868,
+      "step": 928
+    },
+    {
+      "epoch": 0.15051847051198963,
+      "grad_norm": 1.440252661705017,
+      "learning_rate": 4.994680624048674e-06,
+      "loss": 0.184,
+      "step": 929
+    },
+    {
+      "epoch": 0.1506804925469864,
+      "grad_norm": 1.4660731554031372,
+      "learning_rate": 4.994652073879127e-06,
+      "loss": 0.192,
+      "step": 930
+    },
+    {
+      "epoch": 0.15084251458198314,
+      "grad_norm": 1.4855326414108276,
+      "learning_rate": 4.9946234473792645e-06,
+      "loss": 0.1839,
+      "step": 931
+    },
+    {
+      "epoch": 0.1510045366169799,
+      "grad_norm": 1.2965097427368164,
+      "learning_rate": 4.994594744549961e-06,
+      "loss": 0.1671,
+      "step": 932
+    },
+    {
+      "epoch": 0.15116655865197667,
+      "grad_norm": 1.3417192697525024,
+      "learning_rate": 4.994565965392094e-06,
+      "loss": 0.1841,
+      "step": 933
+    },
+    {
+      "epoch": 0.15132858068697344,
+      "grad_norm": 1.2108187675476074,
+      "learning_rate": 4.994537109906546e-06,
+      "loss": 0.1594,
+      "step": 934
+    },
+    {
+      "epoch": 0.15149060272197018,
+      "grad_norm": 1.4936498403549194,
+      "learning_rate": 4.994508178094199e-06,
+      "loss": 0.1838,
+      "step": 935
+    },
+    {
+      "epoch": 0.15165262475696695,
+      "grad_norm": 1.2515209913253784,
+      "learning_rate": 4.99447916995594e-06,
+      "loss": 0.1531,
+      "step": 936
+    },
+    {
+      "epoch": 0.15181464679196371,
+      "grad_norm": 1.2355049848556519,
+      "learning_rate": 4.994450085492653e-06,
+      "loss": 0.1466,
+      "step": 937
+    },
+    {
+      "epoch": 0.15197666882696045,
+      "grad_norm": 1.317088007926941,
+      "learning_rate": 4.99442092470523e-06,
+      "loss": 0.1851,
+      "step": 938
+    },
+    {
+      "epoch": 0.15213869086195722,
+      "grad_norm": 1.262994408607483,
+      "learning_rate": 4.994391687594564e-06,
+      "loss": 0.1656,
+      "step": 939
+    },
+    {
+      "epoch": 0.152300712896954,
+      "grad_norm": 1.1988916397094727,
+      "learning_rate": 4.994362374161548e-06,
+      "loss": 0.1457,
+      "step": 940
+    },
+    {
+      "epoch": 0.15246273493195076,
+      "grad_norm": 1.2624685764312744,
+      "learning_rate": 4.99433298440708e-06,
+      "loss": 0.1759,
+      "step": 941
+    },
+    {
+      "epoch": 0.1526247569669475,
+      "grad_norm": 1.4372855424880981,
+      "learning_rate": 4.994303518332059e-06,
+      "loss": 0.1792,
+      "step": 942
+    },
+    {
+      "epoch": 0.15278677900194426,
+      "grad_norm": 1.3524651527404785,
+      "learning_rate": 4.994273975937386e-06,
+      "loss": 0.1956,
+      "step": 943
+    },
+    {
+      "epoch": 0.15294880103694103,
+      "grad_norm": 1.2189857959747314,
+      "learning_rate": 4.994244357223965e-06,
+      "loss": 0.1646,
+      "step": 944
+    },
+    {
+      "epoch": 0.1531108230719378,
+      "grad_norm": 1.5901044607162476,
+      "learning_rate": 4.994214662192704e-06,
+      "loss": 0.1821,
+      "step": 945
+    },
+    {
+      "epoch": 0.15327284510693454,
+      "grad_norm": 1.2014747858047485,
+      "learning_rate": 4.994184890844509e-06,
+      "loss": 0.1585,
+      "step": 946
+    },
+    {
+      "epoch": 0.1534348671419313,
+      "grad_norm": 1.2449418306350708,
+      "learning_rate": 4.994155043180292e-06,
+      "loss": 0.1766,
+      "step": 947
+    },
+    {
+      "epoch": 0.15359688917692807,
+      "grad_norm": 1.1856237649917603,
+      "learning_rate": 4.9941251192009665e-06,
+      "loss": 0.1561,
+      "step": 948
+    },
+    {
+      "epoch": 0.1537589112119248,
+      "grad_norm": 1.3287744522094727,
+      "learning_rate": 4.994095118907449e-06,
+      "loss": 0.18,
+      "step": 949
+    },
+    {
+      "epoch": 0.15392093324692158,
+      "grad_norm": 1.4006623029708862,
+      "learning_rate": 4.994065042300655e-06,
+      "loss": 0.1852,
+      "step": 950
+    },
+    {
+      "epoch": 0.15408295528191834,
+      "grad_norm": 1.3355299234390259,
+      "learning_rate": 4.994034889381508e-06,
+      "loss": 0.1653,
+      "step": 951
+    },
+    {
+      "epoch": 0.1542449773169151,
+      "grad_norm": 1.5296586751937866,
+      "learning_rate": 4.994004660150927e-06,
+      "loss": 0.2171,
+      "step": 952
+    },
+    {
+      "epoch": 0.15440699935191185,
+      "grad_norm": 1.447704553604126,
+      "learning_rate": 4.99397435460984e-06,
+      "loss": 0.188,
+      "step": 953
+    },
+    {
+      "epoch": 0.15456902138690862,
+      "grad_norm": 1.272879958152771,
+      "learning_rate": 4.993943972759173e-06,
+      "loss": 0.1728,
+      "step": 954
+    },
+    {
+      "epoch": 0.15473104342190538,
+      "grad_norm": 1.1389333009719849,
+      "learning_rate": 4.9939135145998554e-06,
+      "loss": 0.1572,
+      "step": 955
+    },
+    {
+      "epoch": 0.15489306545690215,
+      "grad_norm": 1.3290883302688599,
+      "learning_rate": 4.993882980132819e-06,
+      "loss": 0.1665,
+      "step": 956
+    },
+    {
+      "epoch": 0.1550550874918989,
+      "grad_norm": 1.3863415718078613,
+      "learning_rate": 4.993852369358999e-06,
+      "loss": 0.1757,
+      "step": 957
+    },
+    {
+      "epoch": 0.15521710952689566,
+      "grad_norm": 1.211903691291809,
+      "learning_rate": 4.993821682279332e-06,
+      "loss": 0.1507,
+      "step": 958
+    },
+    {
+      "epoch": 0.15537913156189243,
+      "grad_norm": 1.2453619241714478,
+      "learning_rate": 4.9937909188947555e-06,
+      "loss": 0.1578,
+      "step": 959
+    },
+    {
+      "epoch": 0.15554115359688916,
+      "grad_norm": 1.4014263153076172,
+      "learning_rate": 4.993760079206212e-06,
+      "loss": 0.1949,
+      "step": 960
+    },
+    {
+      "epoch": 0.15570317563188593,
+      "grad_norm": 1.3202704191207886,
+      "learning_rate": 4.993729163214644e-06,
+      "loss": 0.1744,
+      "step": 961
+    },
+    {
+      "epoch": 0.1558651976668827,
+      "grad_norm": 1.3204492330551147,
+      "learning_rate": 4.993698170920999e-06,
+      "loss": 0.1833,
+      "step": 962
+    },
+    {
+      "epoch": 0.15602721970187947,
+      "grad_norm": 1.3719549179077148,
+      "learning_rate": 4.993667102326226e-06,
+      "loss": 0.1733,
+      "step": 963
+    },
+    {
+      "epoch": 0.1561892417368762,
+      "grad_norm": 1.2820831537246704,
+      "learning_rate": 4.993635957431273e-06,
+      "loss": 0.18,
+      "step": 964
+    },
+    {
+      "epoch": 0.15635126377187297,
+      "grad_norm": 1.2792859077453613,
+      "learning_rate": 4.993604736237094e-06,
+      "loss": 0.1643,
+      "step": 965
+    },
+    {
+      "epoch": 0.15651328580686974,
+      "grad_norm": 1.3071776628494263,
+      "learning_rate": 4.993573438744645e-06,
+      "loss": 0.1543,
+      "step": 966
+    },
+    {
+      "epoch": 0.1566753078418665,
+      "grad_norm": 1.3182742595672607,
+      "learning_rate": 4.993542064954883e-06,
+      "loss": 0.1719,
+      "step": 967
+    },
+    {
+      "epoch": 0.15683732987686325,
+      "grad_norm": 1.3087525367736816,
+      "learning_rate": 4.993510614868767e-06,
+      "loss": 0.1889,
+      "step": 968
+    },
+    {
+      "epoch": 0.15699935191186,
+      "grad_norm": 1.2305779457092285,
+      "learning_rate": 4.993479088487262e-06,
+      "loss": 0.1651,
+      "step": 969
+    },
+    {
+      "epoch": 0.15716137394685678,
+      "grad_norm": 1.1964472532272339,
+      "learning_rate": 4.99344748581133e-06,
+      "loss": 0.1523,
+      "step": 970
+    },
+    {
+      "epoch": 0.15732339598185352,
+      "grad_norm": 1.23350989818573,
+      "learning_rate": 4.993415806841939e-06,
+      "loss": 0.16,
+      "step": 971
+    },
+    {
+      "epoch": 0.1574854180168503,
+      "grad_norm": 1.3825608491897583,
+      "learning_rate": 4.993384051580059e-06,
+      "loss": 0.1884,
+      "step": 972
+    },
+    {
+      "epoch": 0.15764744005184705,
+      "grad_norm": 1.4116489887237549,
+      "learning_rate": 4.993352220026661e-06,
+      "loss": 0.1763,
+      "step": 973
+    },
+    {
+      "epoch": 0.15780946208684382,
+      "grad_norm": 1.418979287147522,
+      "learning_rate": 4.993320312182718e-06,
+      "loss": 0.1982,
+      "step": 974
+    },
+    {
+      "epoch": 0.15797148412184056,
+      "grad_norm": 1.3567852973937988,
+      "learning_rate": 4.993288328049208e-06,
+      "loss": 0.1845,
+      "step": 975
+    },
+    {
+      "epoch": 0.15813350615683733,
+      "grad_norm": 1.257607340812683,
+      "learning_rate": 4.993256267627108e-06,
+      "loss": 0.1626,
+      "step": 976
+    },
+    {
+      "epoch": 0.1582955281918341,
+      "grad_norm": 1.2353687286376953,
+      "learning_rate": 4.993224130917399e-06,
+      "loss": 0.1695,
+      "step": 977
+    },
+    {
+      "epoch": 0.15845755022683086,
+      "grad_norm": 1.4237691164016724,
+      "learning_rate": 4.993191917921066e-06,
+      "loss": 0.1841,
+      "step": 978
+    },
+    {
+      "epoch": 0.1586195722618276,
+      "grad_norm": 1.2207448482513428,
+      "learning_rate": 4.9931596286390935e-06,
+      "loss": 0.1518,
+      "step": 979
+    },
+    {
+      "epoch": 0.15878159429682437,
+      "grad_norm": 1.366822600364685,
+      "learning_rate": 4.9931272630724704e-06,
+      "loss": 0.1751,
+      "step": 980
+    },
+    {
+      "epoch": 0.15894361633182114,
+      "grad_norm": 1.3193942308425903,
+      "learning_rate": 4.993094821222186e-06,
+      "loss": 0.1615,
+      "step": 981
+    },
+    {
+      "epoch": 0.15910563836681788,
+      "grad_norm": 1.361041784286499,
+      "learning_rate": 4.993062303089233e-06,
+      "loss": 0.1669,
+      "step": 982
+    },
+    {
+      "epoch": 0.15926766040181464,
+      "grad_norm": 1.4017630815505981,
+      "learning_rate": 4.993029708674607e-06,
+      "loss": 0.1801,
+      "step": 983
+    },
+    {
+      "epoch": 0.1594296824368114,
+      "grad_norm": 1.3844646215438843,
+      "learning_rate": 4.992997037979304e-06,
+      "loss": 0.1771,
+      "step": 984
+    },
+    {
+      "epoch": 0.15959170447180818,
+      "grad_norm": 1.2512445449829102,
+      "learning_rate": 4.992964291004326e-06,
+      "loss": 0.1692,
+      "step": 985
+    },
+    {
+      "epoch": 0.15975372650680492,
+      "grad_norm": 1.4162150621414185,
+      "learning_rate": 4.992931467750673e-06,
+      "loss": 0.1696,
+      "step": 986
+    },
+    {
+      "epoch": 0.15991574854180168,
+      "grad_norm": 1.2711294889450073,
+      "learning_rate": 4.99289856821935e-06,
+      "loss": 0.1628,
+      "step": 987
+    },
+    {
+      "epoch": 0.16007777057679845,
+      "grad_norm": 1.318920373916626,
+      "learning_rate": 4.992865592411362e-06,
+      "loss": 0.1504,
+      "step": 988
+    },
+    {
+      "epoch": 0.16023979261179522,
+      "grad_norm": 1.3679457902908325,
+      "learning_rate": 4.992832540327721e-06,
+      "loss": 0.1738,
+      "step": 989
+    },
+    {
+      "epoch": 0.16040181464679196,
+      "grad_norm": 1.2336037158966064,
+      "learning_rate": 4.992799411969436e-06,
+      "loss": 0.1659,
+      "step": 990
+    },
+    {
+      "epoch": 0.16056383668178872,
+      "grad_norm": 1.1132322549819946,
+      "learning_rate": 4.992766207337523e-06,
+      "loss": 0.1369,
+      "step": 991
+    },
+    {
+      "epoch": 0.1607258587167855,
+      "grad_norm": 1.2798656225204468,
+      "learning_rate": 4.992732926432995e-06,
+      "loss": 0.1702,
+      "step": 992
+    },
+    {
+      "epoch": 0.16088788075178223,
+      "grad_norm": 1.3788708448410034,
+      "learning_rate": 4.992699569256872e-06,
+      "loss": 0.1978,
+      "step": 993
+    },
+    {
+      "epoch": 0.161049902786779,
+      "grad_norm": 1.2766364812850952,
+      "learning_rate": 4.9926661358101745e-06,
+      "loss": 0.1856,
+      "step": 994
+    },
+    {
+      "epoch": 0.16121192482177576,
+      "grad_norm": 1.371200680732727,
+      "learning_rate": 4.992632626093926e-06,
+      "loss": 0.1999,
+      "step": 995
+    },
+    {
+      "epoch": 0.16137394685677253,
+      "grad_norm": 1.2448960542678833,
+      "learning_rate": 4.9925990401091505e-06,
+      "loss": 0.1574,
+      "step": 996
+    },
+    {
+      "epoch": 0.16153596889176927,
+      "grad_norm": 1.2014307975769043,
+      "learning_rate": 4.992565377856876e-06,
+      "loss": 0.1572,
+      "step": 997
+    },
+    {
+      "epoch": 0.16169799092676604,
+      "grad_norm": 1.3043875694274902,
+      "learning_rate": 4.992531639338133e-06,
+      "loss": 0.1746,
+      "step": 998
+    },
+    {
+      "epoch": 0.1618600129617628,
+      "grad_norm": 1.3030551671981812,
+      "learning_rate": 4.992497824553954e-06,
+      "loss": 0.172,
+      "step": 999
+    },
+    {
+      "epoch": 0.16202203499675957,
+      "grad_norm": 1.3272135257720947,
+      "learning_rate": 4.992463933505374e-06,
+      "loss": 0.1755,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1621840570317563,
+      "grad_norm": 1.2103444337844849,
+      "learning_rate": 4.992429966193428e-06,
+      "loss": 0.1522,
+      "step": 1001
+    },
+    {
+      "epoch": 0.16234607906675308,
+      "grad_norm": 1.353193759918213,
+      "learning_rate": 4.9923959226191574e-06,
+      "loss": 0.1796,
+      "step": 1002
+    },
+    {
+      "epoch": 0.16250810110174985,
+      "grad_norm": 1.339188814163208,
+      "learning_rate": 4.992361802783603e-06,
+      "loss": 0.1926,
+      "step": 1003
+    },
+    {
+      "epoch": 0.16267012313674659,
+      "grad_norm": 1.269451379776001,
+      "learning_rate": 4.992327606687808e-06,
+      "loss": 0.1597,
+      "step": 1004
+    },
+    {
+      "epoch": 0.16283214517174335,
+      "grad_norm": 1.325924277305603,
+      "learning_rate": 4.992293334332821e-06,
+      "loss": 0.1785,
+      "step": 1005
+    },
+    {
+      "epoch": 0.16299416720674012,
+      "grad_norm": 1.2406764030456543,
+      "learning_rate": 4.992258985719688e-06,
+      "loss": 0.1658,
+      "step": 1006
+    },
+    {
+      "epoch": 0.1631561892417369,
+      "grad_norm": 1.3176758289337158,
+      "learning_rate": 4.992224560849461e-06,
+      "loss": 0.1802,
+      "step": 1007
+    },
+    {
+      "epoch": 0.16331821127673363,
+      "grad_norm": 1.3824093341827393,
+      "learning_rate": 4.992190059723194e-06,
+      "loss": 0.1961,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1634802333117304,
+      "grad_norm": 1.223380446434021,
+      "learning_rate": 4.9921554823419424e-06,
+      "loss": 0.1567,
+      "step": 1009
+    },
+    {
+      "epoch": 0.16364225534672716,
+      "grad_norm": 1.3019189834594727,
+      "learning_rate": 4.992120828706763e-06,
+      "loss": 0.1695,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1638042773817239,
+      "grad_norm": 1.3338146209716797,
+      "learning_rate": 4.9920860988187185e-06,
+      "loss": 0.1876,
+      "step": 1011
+    },
+    {
+      "epoch": 0.16396629941672067,
+      "grad_norm": 1.167526364326477,
+      "learning_rate": 4.99205129267887e-06,
+      "loss": 0.1609,
+      "step": 1012
+    },
+    {
+      "epoch": 0.16412832145171743,
+      "grad_norm": 1.1966924667358398,
+      "learning_rate": 4.9920164102882816e-06,
+      "loss": 0.1576,
+      "step": 1013
+    },
+    {
+      "epoch": 0.1642903434867142,
+      "grad_norm": 1.1444998979568481,
+      "learning_rate": 4.991981451648022e-06,
+      "loss": 0.1646,
+      "step": 1014
+    },
+    {
+      "epoch": 0.16445236552171094,
+      "grad_norm": 1.1742830276489258,
+      "learning_rate": 4.99194641675916e-06,
+      "loss": 0.1607,
+      "step": 1015
+    },
+    {
+      "epoch": 0.1646143875567077,
+      "grad_norm": 1.199351191520691,
+      "learning_rate": 4.9919113056227685e-06,
+      "loss": 0.1506,
+      "step": 1016
+    },
+    {
+      "epoch": 0.16477640959170448,
+      "grad_norm": 1.2436610460281372,
+      "learning_rate": 4.991876118239922e-06,
+      "loss": 0.1683,
+      "step": 1017
+    },
+    {
+      "epoch": 0.16493843162670124,
+      "grad_norm": 1.322636365890503,
+      "learning_rate": 4.991840854611696e-06,
+      "loss": 0.1757,
+      "step": 1018
+    },
+    {
+      "epoch": 0.16510045366169798,
+      "grad_norm": 1.06990647315979,
+      "learning_rate": 4.99180551473917e-06,
+      "loss": 0.1369,
+      "step": 1019
+    },
+    {
+      "epoch": 0.16526247569669475,
+      "grad_norm": 1.193261981010437,
+      "learning_rate": 4.991770098623425e-06,
+      "loss": 0.1608,
+      "step": 1020
+    },
+    {
+      "epoch": 0.16542449773169152,
+      "grad_norm": 1.2813093662261963,
+      "learning_rate": 4.991734606265544e-06,
+      "loss": 0.1531,
+      "step": 1021
+    },
+    {
+      "epoch": 0.16558651976668826,
+      "grad_norm": 1.2579412460327148,
+      "learning_rate": 4.9916990376666156e-06,
+      "loss": 0.1675,
+      "step": 1022
+    },
+    {
+      "epoch": 0.16574854180168502,
+      "grad_norm": 1.3288764953613281,
+      "learning_rate": 4.991663392827726e-06,
+      "loss": 0.1501,
+      "step": 1023
+    },
+    {
+      "epoch": 0.1659105638366818,
+      "grad_norm": 1.4220918416976929,
+      "learning_rate": 4.991627671749966e-06,
+      "loss": 0.1591,
+      "step": 1024
+    },
+    {
+      "epoch": 0.16607258587167856,
+      "grad_norm": 1.4772206544876099,
+      "learning_rate": 4.991591874434429e-06,
+      "loss": 0.1787,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1662346079066753,
+      "grad_norm": 1.2587242126464844,
+      "learning_rate": 4.9915560008822105e-06,
+      "loss": 0.164,
+      "step": 1026
+    },
+    {
+      "epoch": 0.16639662994167206,
+      "grad_norm": 1.2085176706314087,
+      "learning_rate": 4.991520051094407e-06,
+      "loss": 0.1701,
+      "step": 1027
+    },
+    {
+      "epoch": 0.16655865197666883,
+      "grad_norm": 1.2200427055358887,
+      "learning_rate": 4.99148402507212e-06,
+      "loss": 0.1538,
+      "step": 1028
+    },
+    {
+      "epoch": 0.1667206740116656,
+      "grad_norm": 1.2087466716766357,
+      "learning_rate": 4.991447922816451e-06,
+      "loss": 0.1672,
+      "step": 1029
+    },
+    {
+      "epoch": 0.16688269604666234,
+      "grad_norm": 1.2888163328170776,
+      "learning_rate": 4.991411744328505e-06,
+      "loss": 0.1578,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1670447180816591,
+      "grad_norm": 1.478442668914795,
+      "learning_rate": 4.991375489609388e-06,
+      "loss": 0.1984,
+      "step": 1031
+    },
+    {
+      "epoch": 0.16720674011665587,
+      "grad_norm": 1.3319884538650513,
+      "learning_rate": 4.991339158660211e-06,
+      "loss": 0.1584,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1673687621516526,
+      "grad_norm": 1.1861016750335693,
+      "learning_rate": 4.991302751482084e-06,
+      "loss": 0.1581,
+      "step": 1033
+    },
+    {
+      "epoch": 0.16753078418664938,
+      "grad_norm": 1.2733880281448364,
+      "learning_rate": 4.991266268076121e-06,
+      "loss": 0.1757,
+      "step": 1034
+    },
+    {
+      "epoch": 0.16769280622164615,
+      "grad_norm": 1.2338616847991943,
+      "learning_rate": 4.99122970844344e-06,
+      "loss": 0.1733,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1678548282566429,
+      "grad_norm": 1.305107831954956,
+      "learning_rate": 4.991193072585158e-06,
+      "loss": 0.1722,
+      "step": 1036
+    },
+    {
+      "epoch": 0.16801685029163965,
+      "grad_norm": 1.245059847831726,
+      "learning_rate": 4.991156360502397e-06,
+      "loss": 0.1636,
+      "step": 1037
+    },
+    {
+      "epoch": 0.16817887232663642,
+      "grad_norm": 1.260838270187378,
+      "learning_rate": 4.99111957219628e-06,
+      "loss": 0.1777,
+      "step": 1038
+    },
+    {
+      "epoch": 0.1683408943616332,
+      "grad_norm": 1.13496994972229,
+      "learning_rate": 4.9910827076679325e-06,
+      "loss": 0.1579,
+      "step": 1039
+    },
+    {
+      "epoch": 0.16850291639662995,
+      "grad_norm": 1.2134323120117188,
+      "learning_rate": 4.991045766918482e-06,
+      "loss": 0.1666,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1686649384316267,
+      "grad_norm": 1.0484683513641357,
+      "learning_rate": 4.9910087499490585e-06,
+      "loss": 0.1457,
+      "step": 1041
+    },
+    {
+      "epoch": 0.16882696046662346,
+      "grad_norm": 1.2500617504119873,
+      "learning_rate": 4.990971656760797e-06,
+      "loss": 0.1704,
+      "step": 1042
+    },
+    {
+      "epoch": 0.16898898250162023,
+      "grad_norm": 1.296579360961914,
+      "learning_rate": 4.990934487354831e-06,
+      "loss": 0.1739,
+      "step": 1043
+    },
+    {
+      "epoch": 0.16915100453661697,
+      "grad_norm": 1.2237029075622559,
+      "learning_rate": 4.990897241732296e-06,
+      "loss": 0.1624,
+      "step": 1044
+    },
+    {
+      "epoch": 0.16931302657161373,
+      "grad_norm": 1.361970067024231,
+      "learning_rate": 4.9908599198943346e-06,
+      "loss": 0.193,
+      "step": 1045
+    },
+    {
+      "epoch": 0.1694750486066105,
+      "grad_norm": 1.2445952892303467,
+      "learning_rate": 4.990822521842086e-06,
+      "loss": 0.1788,
+      "step": 1046
+    },
+    {
+      "epoch": 0.16963707064160727,
+      "grad_norm": 1.3412961959838867,
+      "learning_rate": 4.990785047576697e-06,
+      "loss": 0.1853,
+      "step": 1047
+    },
+    {
+      "epoch": 0.169799092676604,
+      "grad_norm": 1.2915765047073364,
+      "learning_rate": 4.990747497099312e-06,
+      "loss": 0.182,
+      "step": 1048
+    },
+    {
+      "epoch": 0.16996111471160077,
+      "grad_norm": 1.2325940132141113,
+      "learning_rate": 4.990709870411082e-06,
+      "loss": 0.1495,
+      "step": 1049
+    },
+    {
+      "epoch": 0.17012313674659754,
+      "grad_norm": 1.2738209962844849,
+      "learning_rate": 4.990672167513158e-06,
+      "loss": 0.1584,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1702851587815943,
+      "grad_norm": 1.3789118528366089,
+      "learning_rate": 4.990634388406692e-06,
+      "loss": 0.1932,
+      "step": 1051
+    },
+    {
+      "epoch": 0.17044718081659105,
+      "grad_norm": 1.3370823860168457,
+      "learning_rate": 4.990596533092841e-06,
+      "loss": 0.1788,
+      "step": 1052
+    },
+    {
+      "epoch": 0.17060920285158782,
+      "grad_norm": 1.381735920906067,
+      "learning_rate": 4.990558601572764e-06,
+      "loss": 0.189,
+      "step": 1053
+    },
+    {
+      "epoch": 0.17077122488658458,
+      "grad_norm": 1.2468198537826538,
+      "learning_rate": 4.9905205938476195e-06,
+      "loss": 0.1635,
+      "step": 1054
+    },
+    {
+      "epoch": 0.17093324692158132,
+      "grad_norm": 1.2447727918624878,
+      "learning_rate": 4.990482509918572e-06,
+      "loss": 0.1759,
+      "step": 1055
+    },
+    {
+      "epoch": 0.1710952689565781,
+      "grad_norm": 1.266304612159729,
+      "learning_rate": 4.990444349786788e-06,
+      "loss": 0.186,
+      "step": 1056
+    },
+    {
+      "epoch": 0.17125729099157486,
+      "grad_norm": 1.2319304943084717,
+      "learning_rate": 4.990406113453433e-06,
+      "loss": 0.1781,
+      "step": 1057
+    },
+    {
+      "epoch": 0.17141931302657162,
+      "grad_norm": 1.234866738319397,
+      "learning_rate": 4.990367800919677e-06,
+      "loss": 0.1719,
+      "step": 1058
+    },
+    {
+      "epoch": 0.17158133506156836,
+      "grad_norm": 1.0335838794708252,
+      "learning_rate": 4.990329412186693e-06,
+      "loss": 0.1407,
+      "step": 1059
+    },
+    {
+      "epoch": 0.17174335709656513,
+      "grad_norm": 1.2708618640899658,
+      "learning_rate": 4.990290947255656e-06,
+      "loss": 0.1635,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1719053791315619,
+      "grad_norm": 1.3228340148925781,
+      "learning_rate": 4.990252406127742e-06,
+      "loss": 0.186,
+      "step": 1061
+    },
+    {
+      "epoch": 0.17206740116655866,
+      "grad_norm": 1.2613164186477661,
+      "learning_rate": 4.9902137888041304e-06,
+      "loss": 0.1698,
+      "step": 1062
+    },
+    {
+      "epoch": 0.1722294232015554,
+      "grad_norm": 1.2189115285873413,
+      "learning_rate": 4.990175095286003e-06,
+      "loss": 0.1622,
+      "step": 1063
+    },
+    {
+      "epoch": 0.17239144523655217,
+      "grad_norm": 1.113931655883789,
+      "learning_rate": 4.990136325574545e-06,
+      "loss": 0.1433,
+      "step": 1064
+    },
+    {
+      "epoch": 0.17255346727154894,
+      "grad_norm": 1.1805305480957031,
+      "learning_rate": 4.9900974796709405e-06,
+      "loss": 0.1345,
+      "step": 1065
+    },
+    {
+      "epoch": 0.17271548930654568,
+      "grad_norm": 1.3531560897827148,
+      "learning_rate": 4.990058557576379e-06,
+      "loss": 0.175,
+      "step": 1066
+    },
+    {
+      "epoch": 0.17287751134154244,
+      "grad_norm": 1.403969168663025,
+      "learning_rate": 4.990019559292052e-06,
+      "loss": 0.1674,
+      "step": 1067
+    },
+    {
+      "epoch": 0.1730395333765392,
+      "grad_norm": 1.4083645343780518,
+      "learning_rate": 4.989980484819152e-06,
+      "loss": 0.1722,
+      "step": 1068
+    },
+    {
+      "epoch": 0.17320155541153598,
+      "grad_norm": 1.2746031284332275,
+      "learning_rate": 4.989941334158874e-06,
+      "loss": 0.1638,
+      "step": 1069
+    },
+    {
+      "epoch": 0.17336357744653272,
+      "grad_norm": 1.2782847881317139,
+      "learning_rate": 4.9899021073124175e-06,
+      "loss": 0.1733,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17352559948152949,
+      "grad_norm": 1.4043595790863037,
+      "learning_rate": 4.989862804280982e-06,
+      "loss": 0.1756,
+      "step": 1071
+    },
+    {
+      "epoch": 0.17368762151652625,
+      "grad_norm": 1.3001036643981934,
+      "learning_rate": 4.989823425065769e-06,
+      "loss": 0.1839,
+      "step": 1072
+    },
+    {
+      "epoch": 0.17384964355152302,
+      "grad_norm": 1.2596933841705322,
+      "learning_rate": 4.989783969667986e-06,
+      "loss": 0.175,
+      "step": 1073
+    },
+    {
+      "epoch": 0.17401166558651976,
+      "grad_norm": 1.2950873374938965,
+      "learning_rate": 4.989744438088838e-06,
+      "loss": 0.1684,
+      "step": 1074
+    },
+    {
+      "epoch": 0.17417368762151653,
+      "grad_norm": 1.1493442058563232,
+      "learning_rate": 4.989704830329535e-06,
+      "loss": 0.1672,
+      "step": 1075
+    },
+    {
+      "epoch": 0.1743357096565133,
+      "grad_norm": 1.1817610263824463,
+      "learning_rate": 4.98966514639129e-06,
+      "loss": 0.1617,
+      "step": 1076
+    },
+    {
+      "epoch": 0.17449773169151003,
+      "grad_norm": 1.269439697265625,
+      "learning_rate": 4.989625386275315e-06,
+      "loss": 0.1821,
+      "step": 1077
+    },
+    {
+      "epoch": 0.1746597537265068,
+      "grad_norm": 1.2774385213851929,
+      "learning_rate": 4.98958554998283e-06,
+      "loss": 0.1772,
+      "step": 1078
+    },
+    {
+      "epoch": 0.17482177576150357,
+      "grad_norm": 1.3013243675231934,
+      "learning_rate": 4.98954563751505e-06,
+      "loss": 0.1776,
+      "step": 1079
+    },
+    {
+      "epoch": 0.17498379779650033,
+      "grad_norm": 1.1634021997451782,
+      "learning_rate": 4.989505648873198e-06,
+      "loss": 0.164,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17514581983149707,
+      "grad_norm": 1.296128273010254,
+      "learning_rate": 4.989465584058499e-06,
+      "loss": 0.1689,
+      "step": 1081
+    },
+    {
+      "epoch": 0.17530784186649384,
+      "grad_norm": 1.4235268831253052,
+      "learning_rate": 4.989425443072177e-06,
+      "loss": 0.2038,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1754698639014906,
+      "grad_norm": 1.384175181388855,
+      "learning_rate": 4.989385225915461e-06,
+      "loss": 0.1739,
+      "step": 1083
+    },
+    {
+      "epoch": 0.17563188593648738,
+      "grad_norm": 1.1942464113235474,
+      "learning_rate": 4.9893449325895804e-06,
+      "loss": 0.1599,
+      "step": 1084
+    },
+    {
+      "epoch": 0.17579390797148411,
+      "grad_norm": 1.3841326236724854,
+      "learning_rate": 4.989304563095769e-06,
+      "loss": 0.1975,
+      "step": 1085
+    },
+    {
+      "epoch": 0.17595593000648088,
+      "grad_norm": 1.2877038717269897,
+      "learning_rate": 4.989264117435263e-06,
+      "loss": 0.1825,
+      "step": 1086
+    },
+    {
+      "epoch": 0.17611795204147765,
+      "grad_norm": 1.3684138059616089,
+      "learning_rate": 4.9892235956092985e-06,
+      "loss": 0.1802,
+      "step": 1087
+    },
+    {
+      "epoch": 0.1762799740764744,
+      "grad_norm": 1.2257750034332275,
+      "learning_rate": 4.9891829976191155e-06,
+      "loss": 0.169,
+      "step": 1088
+    },
+    {
+      "epoch": 0.17644199611147116,
+      "grad_norm": 1.281955599784851,
+      "learning_rate": 4.989142323465957e-06,
+      "loss": 0.1884,
+      "step": 1089
+    },
+    {
+      "epoch": 0.17660401814646792,
+      "grad_norm": 1.2138808965682983,
+      "learning_rate": 4.9891015731510665e-06,
+      "loss": 0.1738,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1767660401814647,
+      "grad_norm": 1.185698390007019,
+      "learning_rate": 4.989060746675691e-06,
+      "loss": 0.1527,
+      "step": 1091
+    },
+    {
+      "epoch": 0.17692806221646143,
+      "grad_norm": 1.1625752449035645,
+      "learning_rate": 4.989019844041081e-06,
+      "loss": 0.1573,
+      "step": 1092
+    },
+    {
+      "epoch": 0.1770900842514582,
+      "grad_norm": 1.1147958040237427,
+      "learning_rate": 4.988978865248486e-06,
+      "loss": 0.1471,
+      "step": 1093
+    },
+    {
+      "epoch": 0.17725210628645496,
+      "grad_norm": 1.0958008766174316,
+      "learning_rate": 4.988937810299161e-06,
+      "loss": 0.1473,
+      "step": 1094
+    },
+    {
+      "epoch": 0.17741412832145173,
+      "grad_norm": 1.2946974039077759,
+      "learning_rate": 4.988896679194363e-06,
+      "loss": 0.169,
+      "step": 1095
+    },
+    {
+      "epoch": 0.17757615035644847,
+      "grad_norm": 1.2533652782440186,
+      "learning_rate": 4.98885547193535e-06,
+      "loss": 0.1622,
+      "step": 1096
+    },
+    {
+      "epoch": 0.17773817239144524,
+      "grad_norm": 1.2509846687316895,
+      "learning_rate": 4.988814188523381e-06,
+      "loss": 0.1657,
+      "step": 1097
+    },
+    {
+      "epoch": 0.177900194426442,
+      "grad_norm": 1.2330974340438843,
+      "learning_rate": 4.988772828959722e-06,
+      "loss": 0.171,
+      "step": 1098
+    },
+    {
+      "epoch": 0.17806221646143874,
+      "grad_norm": 1.1308923959732056,
+      "learning_rate": 4.988731393245636e-06,
+      "loss": 0.1486,
+      "step": 1099
+    },
+    {
+      "epoch": 0.1782242384964355,
+      "grad_norm": 1.2318450212478638,
+      "learning_rate": 4.988689881382392e-06,
+      "loss": 0.1681,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17838626053143228,
+      "grad_norm": 1.292179822921753,
+      "learning_rate": 4.988648293371262e-06,
+      "loss": 0.1751,
+      "step": 1101
+    },
+    {
+      "epoch": 0.17854828256642905,
+      "grad_norm": 1.2476286888122559,
+      "learning_rate": 4.988606629213515e-06,
+      "loss": 0.1792,
+      "step": 1102
+    },
+    {
+      "epoch": 0.17871030460142578,
+      "grad_norm": 1.4195069074630737,
+      "learning_rate": 4.988564888910428e-06,
+      "loss": 0.1311,
+      "step": 1103
+    },
+    {
+      "epoch": 0.17887232663642255,
+      "grad_norm": 1.1690012216567993,
+      "learning_rate": 4.9885230724632775e-06,
+      "loss": 0.1554,
+      "step": 1104
+    },
+    {
+      "epoch": 0.17903434867141932,
+      "grad_norm": 1.2006351947784424,
+      "learning_rate": 4.988481179873342e-06,
+      "loss": 0.1589,
+      "step": 1105
+    },
+    {
+      "epoch": 0.17919637070641609,
+      "grad_norm": 1.189113974571228,
+      "learning_rate": 4.9884392111419056e-06,
+      "loss": 0.1631,
+      "step": 1106
+    },
+    {
+      "epoch": 0.17935839274141283,
+      "grad_norm": 1.1603453159332275,
+      "learning_rate": 4.9883971662702514e-06,
+      "loss": 0.1682,
+      "step": 1107
+    },
+    {
+      "epoch": 0.1795204147764096,
+      "grad_norm": 1.2347701787948608,
+      "learning_rate": 4.988355045259665e-06,
+      "loss": 0.1641,
+      "step": 1108
+    },
+    {
+      "epoch": 0.17968243681140636,
+      "grad_norm": 1.1546489000320435,
+      "learning_rate": 4.988312848111436e-06,
+      "loss": 0.1586,
+      "step": 1109
+    },
+    {
+      "epoch": 0.1798444588464031,
+      "grad_norm": 1.1792272329330444,
+      "learning_rate": 4.988270574826857e-06,
+      "loss": 0.1579,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18000648088139987,
+      "grad_norm": 1.2593525648117065,
+      "learning_rate": 4.988228225407218e-06,
+      "loss": 0.1624,
+      "step": 1111
+    },
+    {
+      "epoch": 0.18016850291639663,
+      "grad_norm": 1.1456053256988525,
+      "learning_rate": 4.9881857998538175e-06,
+      "loss": 0.1591,
+      "step": 1112
+    },
+    {
+      "epoch": 0.1803305249513934,
+      "grad_norm": 1.169472336769104,
+      "learning_rate": 4.988143298167952e-06,
+      "loss": 0.1534,
+      "step": 1113
+    },
+    {
+      "epoch": 0.18049254698639014,
+      "grad_norm": 1.1957638263702393,
+      "learning_rate": 4.988100720350924e-06,
+      "loss": 0.1684,
+      "step": 1114
+    },
+    {
+      "epoch": 0.1806545690213869,
+      "grad_norm": 1.2542282342910767,
+      "learning_rate": 4.988058066404035e-06,
+      "loss": 0.1566,
+      "step": 1115
+    },
+    {
+      "epoch": 0.18081659105638367,
+      "grad_norm": 1.421115517616272,
+      "learning_rate": 4.988015336328589e-06,
+      "loss": 0.1874,
+      "step": 1116
+    },
+    {
+      "epoch": 0.18097861309138044,
+      "grad_norm": 1.3274586200714111,
+      "learning_rate": 4.987972530125895e-06,
+      "loss": 0.19,
+      "step": 1117
+    },
+    {
+      "epoch": 0.18114063512637718,
+      "grad_norm": 1.2618639469146729,
+      "learning_rate": 4.987929647797263e-06,
+      "loss": 0.1596,
+      "step": 1118
+    },
+    {
+      "epoch": 0.18130265716137395,
+      "grad_norm": 1.172289490699768,
+      "learning_rate": 4.987886689344003e-06,
+      "loss": 0.1407,
+      "step": 1119
+    },
+    {
+      "epoch": 0.18146467919637072,
+      "grad_norm": 1.2448173761367798,
+      "learning_rate": 4.987843654767432e-06,
+      "loss": 0.1548,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18162670123136745,
+      "grad_norm": 1.3023630380630493,
+      "learning_rate": 4.987800544068866e-06,
+      "loss": 0.1638,
+      "step": 1121
+    },
+    {
+      "epoch": 0.18178872326636422,
+      "grad_norm": 1.1977190971374512,
+      "learning_rate": 4.987757357249623e-06,
+      "loss": 0.1713,
+      "step": 1122
+    },
+    {
+      "epoch": 0.181950745301361,
+      "grad_norm": 1.3710788488388062,
+      "learning_rate": 4.987714094311026e-06,
+      "loss": 0.1915,
+      "step": 1123
+    },
+    {
+      "epoch": 0.18211276733635776,
+      "grad_norm": 1.228145956993103,
+      "learning_rate": 4.987670755254397e-06,
+      "loss": 0.1672,
+      "step": 1124
+    },
+    {
+      "epoch": 0.1822747893713545,
+      "grad_norm": 1.3164801597595215,
+      "learning_rate": 4.9876273400810636e-06,
+      "loss": 0.2021,
+      "step": 1125
+    },
+    {
+      "epoch": 0.18243681140635126,
+      "grad_norm": 1.2684440612792969,
+      "learning_rate": 4.987583848792353e-06,
+      "loss": 0.1668,
+      "step": 1126
+    },
+    {
+      "epoch": 0.18259883344134803,
+      "grad_norm": 1.2884535789489746,
+      "learning_rate": 4.987540281389596e-06,
+      "loss": 0.1717,
+      "step": 1127
+    },
+    {
+      "epoch": 0.18276085547634477,
+      "grad_norm": 1.170680284500122,
+      "learning_rate": 4.987496637874127e-06,
+      "loss": 0.1524,
+      "step": 1128
+    },
+    {
+      "epoch": 0.18292287751134154,
+      "grad_norm": 1.2426053285598755,
+      "learning_rate": 4.98745291824728e-06,
+      "loss": 0.167,
+      "step": 1129
+    },
+    {
+      "epoch": 0.1830848995463383,
+      "grad_norm": 1.2498632669448853,
+      "learning_rate": 4.987409122510394e-06,
+      "loss": 0.1702,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18324692158133507,
+      "grad_norm": 1.2569835186004639,
+      "learning_rate": 4.987365250664807e-06,
+      "loss": 0.1658,
+      "step": 1131
+    },
+    {
+      "epoch": 0.1834089436163318,
+      "grad_norm": 1.327354073524475,
+      "learning_rate": 4.9873213027118635e-06,
+      "loss": 0.1838,
+      "step": 1132
+    },
+    {
+      "epoch": 0.18357096565132858,
+      "grad_norm": 1.0752174854278564,
+      "learning_rate": 4.987277278652907e-06,
+      "loss": 0.1393,
+      "step": 1133
+    },
+    {
+      "epoch": 0.18373298768632534,
+      "grad_norm": 1.2634954452514648,
+      "learning_rate": 4.987233178489285e-06,
+      "loss": 0.1603,
+      "step": 1134
+    },
+    {
+      "epoch": 0.1838950097213221,
+      "grad_norm": 1.1911259889602661,
+      "learning_rate": 4.987189002222347e-06,
+      "loss": 0.1575,
+      "step": 1135
+    },
+    {
+      "epoch": 0.18405703175631885,
+      "grad_norm": 1.2841031551361084,
+      "learning_rate": 4.987144749853444e-06,
+      "loss": 0.1747,
+      "step": 1136
+    },
+    {
+      "epoch": 0.18421905379131562,
+      "grad_norm": 1.2230887413024902,
+      "learning_rate": 4.987100421383931e-06,
+      "loss": 0.1785,
+      "step": 1137
+    },
+    {
+      "epoch": 0.18438107582631239,
+      "grad_norm": 1.2518471479415894,
+      "learning_rate": 4.987056016815163e-06,
+      "loss": 0.1764,
+      "step": 1138
+    },
+    {
+      "epoch": 0.18454309786130912,
+      "grad_norm": 1.0959880352020264,
+      "learning_rate": 4.9870115361485e-06,
+      "loss": 0.1398,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1847051198963059,
+      "grad_norm": 1.3243485689163208,
+      "learning_rate": 4.986966979385302e-06,
+      "loss": 0.1744,
+      "step": 1140
+    },
+    {
+      "epoch": 0.18486714193130266,
+      "grad_norm": 1.2652283906936646,
+      "learning_rate": 4.986922346526933e-06,
+      "loss": 0.1814,
+      "step": 1141
+    },
+    {
+      "epoch": 0.18502916396629943,
+      "grad_norm": 1.2872614860534668,
+      "learning_rate": 4.986877637574758e-06,
+      "loss": 0.1784,
+      "step": 1142
+    },
+    {
+      "epoch": 0.18519118600129617,
+      "grad_norm": 1.1612954139709473,
+      "learning_rate": 4.9868328525301465e-06,
+      "loss": 0.1584,
+      "step": 1143
+    },
+    {
+      "epoch": 0.18535320803629293,
+      "grad_norm": 1.3280037641525269,
+      "learning_rate": 4.986787991394467e-06,
+      "loss": 0.1737,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1855152300712897,
+      "grad_norm": 1.0931044816970825,
+      "learning_rate": 4.986743054169093e-06,
+      "loss": 0.1389,
+      "step": 1145
+    },
+    {
+      "epoch": 0.18567725210628647,
+      "grad_norm": 1.2529737949371338,
+      "learning_rate": 4.9866980408554e-06,
+      "loss": 0.1537,
+      "step": 1146
+    },
+    {
+      "epoch": 0.1858392741412832,
+      "grad_norm": 1.2056132555007935,
+      "learning_rate": 4.986652951454764e-06,
+      "loss": 0.1531,
+      "step": 1147
+    },
+    {
+      "epoch": 0.18600129617627997,
+      "grad_norm": 1.2318496704101562,
+      "learning_rate": 4.986607785968565e-06,
+      "loss": 0.1839,
+      "step": 1148
+    },
+    {
+      "epoch": 0.18616331821127674,
+      "grad_norm": 1.3150043487548828,
+      "learning_rate": 4.9865625443981854e-06,
+      "loss": 0.1764,
+      "step": 1149
+    },
+    {
+      "epoch": 0.18632534024627348,
+      "grad_norm": 1.3862760066986084,
+      "learning_rate": 4.986517226745009e-06,
+      "loss": 0.1665,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18648736228127025,
+      "grad_norm": 1.3067452907562256,
+      "learning_rate": 4.986471833010423e-06,
+      "loss": 0.1704,
+      "step": 1151
+    },
+    {
+      "epoch": 0.18664938431626701,
+      "grad_norm": 1.1937528848648071,
+      "learning_rate": 4.9864263631958165e-06,
+      "loss": 0.1661,
+      "step": 1152
+    },
+    {
+      "epoch": 0.18681140635126378,
+      "grad_norm": 1.2431349754333496,
+      "learning_rate": 4.986380817302579e-06,
+      "loss": 0.1721,
+      "step": 1153
+    },
+    {
+      "epoch": 0.18697342838626052,
+      "grad_norm": 1.0870198011398315,
+      "learning_rate": 4.986335195332107e-06,
+      "loss": 0.1445,
+      "step": 1154
+    },
+    {
+      "epoch": 0.1871354504212573,
+      "grad_norm": 1.366522192955017,
+      "learning_rate": 4.986289497285794e-06,
+      "loss": 0.1665,
+      "step": 1155
+    },
+    {
+      "epoch": 0.18729747245625405,
+      "grad_norm": 1.1426249742507935,
+      "learning_rate": 4.986243723165039e-06,
+      "loss": 0.1526,
+      "step": 1156
+    },
+    {
+      "epoch": 0.18745949449125082,
+      "grad_norm": 1.2970033884048462,
+      "learning_rate": 4.986197872971244e-06,
+      "loss": 0.18,
+      "step": 1157
+    },
+    {
+      "epoch": 0.18762151652624756,
+      "grad_norm": 1.2247062921524048,
+      "learning_rate": 4.9861519467058094e-06,
+      "loss": 0.1613,
+      "step": 1158
+    },
+    {
+      "epoch": 0.18778353856124433,
+      "grad_norm": 1.169364333152771,
+      "learning_rate": 4.986105944370142e-06,
+      "loss": 0.1621,
+      "step": 1159
+    },
+    {
+      "epoch": 0.1879455605962411,
+      "grad_norm": 1.2808866500854492,
+      "learning_rate": 4.986059865965649e-06,
+      "loss": 0.1978,
+      "step": 1160
+    },
+    {
+      "epoch": 0.18810758263123784,
+      "grad_norm": 1.209956169128418,
+      "learning_rate": 4.986013711493739e-06,
+      "loss": 0.1637,
+      "step": 1161
+    },
+    {
+      "epoch": 0.1882696046662346,
+      "grad_norm": 1.3208413124084473,
+      "learning_rate": 4.985967480955827e-06,
+      "loss": 0.181,
+      "step": 1162
+    },
+    {
+      "epoch": 0.18843162670123137,
+      "grad_norm": 1.1594798564910889,
+      "learning_rate": 4.985921174353325e-06,
+      "loss": 0.1576,
+      "step": 1163
+    },
+    {
+      "epoch": 0.18859364873622814,
+      "grad_norm": 1.7212146520614624,
+      "learning_rate": 4.9858747916876515e-06,
+      "loss": 0.1835,
+      "step": 1164
+    },
+    {
+      "epoch": 0.18875567077122488,
+      "grad_norm": 1.0957105159759521,
+      "learning_rate": 4.985828332960225e-06,
+      "loss": 0.1462,
+      "step": 1165
+    },
+    {
+      "epoch": 0.18891769280622164,
+      "grad_norm": 1.3084644079208374,
+      "learning_rate": 4.985781798172467e-06,
+      "loss": 0.1791,
+      "step": 1166
+    },
+    {
+      "epoch": 0.1890797148412184,
+      "grad_norm": 1.1999437808990479,
+      "learning_rate": 4.985735187325802e-06,
+      "loss": 0.1615,
+      "step": 1167
+    },
+    {
+      "epoch": 0.18924173687621518,
+      "grad_norm": 1.319326400756836,
+      "learning_rate": 4.9856885004216545e-06,
+      "loss": 0.1777,
+      "step": 1168
+    },
+    {
+      "epoch": 0.18940375891121192,
+      "grad_norm": 1.5149074792861938,
+      "learning_rate": 4.985641737461455e-06,
+      "loss": 0.1957,
+      "step": 1169
+    },
+    {
+      "epoch": 0.18956578094620868,
+      "grad_norm": 1.3002344369888306,
+      "learning_rate": 4.985594898446633e-06,
+      "loss": 0.1655,
+      "step": 1170
+    },
+    {
+      "epoch": 0.18972780298120545,
+      "grad_norm": 1.3267287015914917,
+      "learning_rate": 4.985547983378622e-06,
+      "loss": 0.1698,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1898898250162022,
+      "grad_norm": 1.2608460187911987,
+      "learning_rate": 4.9855009922588585e-06,
+      "loss": 0.1638,
+      "step": 1172
+    },
+    {
+      "epoch": 0.19005184705119896,
+      "grad_norm": 1.1155411005020142,
+      "learning_rate": 4.985453925088779e-06,
+      "loss": 0.1507,
+      "step": 1173
+    },
+    {
+      "epoch": 0.19021386908619572,
+      "grad_norm": 1.2303216457366943,
+      "learning_rate": 4.985406781869824e-06,
+      "loss": 0.1799,
+      "step": 1174
+    },
+    {
+      "epoch": 0.1903758911211925,
+      "grad_norm": 1.1471197605133057,
+      "learning_rate": 4.985359562603436e-06,
+      "loss": 0.1512,
+      "step": 1175
+    },
+    {
+      "epoch": 0.19053791315618923,
+      "grad_norm": 1.1257221698760986,
+      "learning_rate": 4.98531226729106e-06,
+      "loss": 0.1563,
+      "step": 1176
+    },
+    {
+      "epoch": 0.190699935191186,
+      "grad_norm": 1.333609700202942,
+      "learning_rate": 4.985264895934142e-06,
+      "loss": 0.2002,
+      "step": 1177
+    },
+    {
+      "epoch": 0.19086195722618277,
+      "grad_norm": 1.2045459747314453,
+      "learning_rate": 4.985217448534134e-06,
+      "loss": 0.1613,
+      "step": 1178
+    },
+    {
+      "epoch": 0.19102397926117953,
+      "grad_norm": 1.1349177360534668,
+      "learning_rate": 4.985169925092485e-06,
+      "loss": 0.1533,
+      "step": 1179
+    },
+    {
+      "epoch": 0.19118600129617627,
+      "grad_norm": 1.0895230770111084,
+      "learning_rate": 4.985122325610651e-06,
+      "loss": 0.1478,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19134802333117304,
+      "grad_norm": 1.375070571899414,
+      "learning_rate": 4.985074650090087e-06,
+      "loss": 0.1824,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1915100453661698,
+      "grad_norm": 1.3039453029632568,
+      "learning_rate": 4.985026898532253e-06,
+      "loss": 0.1727,
+      "step": 1182
+    },
+    {
+      "epoch": 0.19167206740116655,
+      "grad_norm": 1.2111849784851074,
+      "learning_rate": 4.984979070938609e-06,
+      "loss": 0.1557,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1918340894361633,
+      "grad_norm": 1.2889275550842285,
+      "learning_rate": 4.98493116731062e-06,
+      "loss": 0.1659,
+      "step": 1184
+    },
+    {
+      "epoch": 0.19199611147116008,
+      "grad_norm": 1.2837518453598022,
+      "learning_rate": 4.98488318764975e-06,
+      "loss": 0.1584,
+      "step": 1185
+    },
+    {
+      "epoch": 0.19215813350615685,
+      "grad_norm": 1.1638296842575073,
+      "learning_rate": 4.984835131957468e-06,
+      "loss": 0.1552,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1923201555411536,
+      "grad_norm": 1.2172795534133911,
+      "learning_rate": 4.9847870002352435e-06,
+      "loss": 0.1586,
+      "step": 1187
+    },
+    {
+      "epoch": 0.19248217757615035,
+      "grad_norm": 1.3987014293670654,
+      "learning_rate": 4.98473879248455e-06,
+      "loss": 0.1872,
+      "step": 1188
+    },
+    {
+      "epoch": 0.19264419961114712,
+      "grad_norm": 1.2352248430252075,
+      "learning_rate": 4.984690508706863e-06,
+      "loss": 0.1545,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1928062216461439,
+      "grad_norm": 1.2587782144546509,
+      "learning_rate": 4.984642148903659e-06,
+      "loss": 0.1632,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19296824368114063,
+      "grad_norm": 1.3047451972961426,
+      "learning_rate": 4.9845937130764185e-06,
+      "loss": 0.1715,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1931302657161374,
+      "grad_norm": 1.393811583518982,
+      "learning_rate": 4.984545201226623e-06,
+      "loss": 0.1814,
+      "step": 1192
+    },
+    {
+      "epoch": 0.19329228775113416,
+      "grad_norm": 1.2711546421051025,
+      "learning_rate": 4.984496613355756e-06,
+      "loss": 0.1545,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1934543097861309,
+      "grad_norm": 1.268141269683838,
+      "learning_rate": 4.984447949465305e-06,
+      "loss": 0.1825,
+      "step": 1194
+    },
+    {
+      "epoch": 0.19361633182112767,
+      "grad_norm": 1.3329243659973145,
+      "learning_rate": 4.984399209556759e-06,
+      "loss": 0.2102,
+      "step": 1195
+    },
+    {
+      "epoch": 0.19377835385612444,
+      "grad_norm": 1.380758285522461,
+      "learning_rate": 4.98435039363161e-06,
+      "loss": 0.1462,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1939403758911212,
+      "grad_norm": 1.252043604850769,
+      "learning_rate": 4.98430150169135e-06,
+      "loss": 0.1578,
+      "step": 1197
+    },
+    {
+      "epoch": 0.19410239792611794,
+      "grad_norm": 1.2641617059707642,
+      "learning_rate": 4.984252533737477e-06,
+      "loss": 0.177,
+      "step": 1198
+    },
+    {
+      "epoch": 0.1942644199611147,
+      "grad_norm": 1.1985224485397339,
+      "learning_rate": 4.984203489771488e-06,
+      "loss": 0.1531,
+      "step": 1199
+    },
+    {
+      "epoch": 0.19442644199611148,
+      "grad_norm": 1.3280110359191895,
+      "learning_rate": 4.984154369794883e-06,
+      "loss": 0.181,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19458846403110824,
+      "grad_norm": 1.196886658668518,
+      "learning_rate": 4.9841051738091675e-06,
+      "loss": 0.1658,
+      "step": 1201
+    },
+    {
+      "epoch": 0.19475048606610498,
+      "grad_norm": 1.1714789867401123,
+      "learning_rate": 4.984055901815844e-06,
+      "loss": 0.1547,
+      "step": 1202
+    },
+    {
+      "epoch": 0.19491250810110175,
+      "grad_norm": 1.390263557434082,
+      "learning_rate": 4.984006553816421e-06,
+      "loss": 0.1754,
+      "step": 1203
+    },
+    {
+      "epoch": 0.19507453013609852,
+      "grad_norm": 1.1799697875976562,
+      "learning_rate": 4.983957129812409e-06,
+      "loss": 0.147,
+      "step": 1204
+    },
+    {
+      "epoch": 0.19523655217109526,
+      "grad_norm": 1.1202303171157837,
+      "learning_rate": 4.983907629805319e-06,
+      "loss": 0.1609,
+      "step": 1205
+    },
+    {
+      "epoch": 0.19539857420609202,
+      "grad_norm": 1.2552509307861328,
+      "learning_rate": 4.9838580537966676e-06,
+      "loss": 0.1653,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1955605962410888,
+      "grad_norm": 1.516097068786621,
+      "learning_rate": 4.98380840178797e-06,
+      "loss": 0.1858,
+      "step": 1207
+    },
+    {
+      "epoch": 0.19572261827608556,
+      "grad_norm": 1.360250473022461,
+      "learning_rate": 4.983758673780747e-06,
+      "loss": 0.1848,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1958846403110823,
+      "grad_norm": 1.223142385482788,
+      "learning_rate": 4.983708869776518e-06,
+      "loss": 0.1666,
+      "step": 1209
+    },
+    {
+      "epoch": 0.19604666234607906,
+      "grad_norm": 1.1533637046813965,
+      "learning_rate": 4.9836589897768084e-06,
+      "loss": 0.1673,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19620868438107583,
+      "grad_norm": 1.2063921689987183,
+      "learning_rate": 4.983609033783144e-06,
+      "loss": 0.165,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1963707064160726,
+      "grad_norm": 1.1458781957626343,
+      "learning_rate": 4.983559001797054e-06,
+      "loss": 0.1422,
+      "step": 1212
+    },
+    {
+      "epoch": 0.19653272845106934,
+      "grad_norm": 1.3407319784164429,
+      "learning_rate": 4.9835088938200674e-06,
+      "loss": 0.2003,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1966947504860661,
+      "grad_norm": 1.2016749382019043,
+      "learning_rate": 4.983458709853719e-06,
+      "loss": 0.1774,
+      "step": 1214
+    },
+    {
+      "epoch": 0.19685677252106287,
+      "grad_norm": 1.1205006837844849,
+      "learning_rate": 4.983408449899545e-06,
+      "loss": 0.1567,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1970187945560596,
+      "grad_norm": 1.1485540866851807,
+      "learning_rate": 4.9833581139590814e-06,
+      "loss": 0.1703,
+      "step": 1216
+    },
+    {
+      "epoch": 0.19718081659105638,
+      "grad_norm": 1.086588978767395,
+      "learning_rate": 4.983307702033869e-06,
+      "loss": 0.1514,
+      "step": 1217
+    },
+    {
+      "epoch": 0.19734283862605315,
+      "grad_norm": 1.2290334701538086,
+      "learning_rate": 4.983257214125451e-06,
+      "loss": 0.1737,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1975048606610499,
+      "grad_norm": 1.3126085996627808,
+      "learning_rate": 4.98320665023537e-06,
+      "loss": 0.1807,
+      "step": 1219
+    },
+    {
+      "epoch": 0.19766688269604665,
+      "grad_norm": 1.2392646074295044,
+      "learning_rate": 4.9831560103651765e-06,
+      "loss": 0.184,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19782890473104342,
+      "grad_norm": 1.257860779762268,
+      "learning_rate": 4.983105294516418e-06,
+      "loss": 0.1823,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1979909267660402,
+      "grad_norm": 1.3138630390167236,
+      "learning_rate": 4.983054502690646e-06,
+      "loss": 0.1887,
+      "step": 1222
+    },
+    {
+      "epoch": 0.19815294880103695,
+      "grad_norm": 1.1415594816207886,
+      "learning_rate": 4.983003634889415e-06,
+      "loss": 0.1509,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1983149708360337,
+      "grad_norm": 1.1062023639678955,
+      "learning_rate": 4.9829526911142825e-06,
+      "loss": 0.1551,
+      "step": 1224
+    },
+    {
+      "epoch": 0.19847699287103046,
+      "grad_norm": 1.069726824760437,
+      "learning_rate": 4.982901671366805e-06,
+      "loss": 0.1436,
+      "step": 1225
+    },
+    {
+      "epoch": 0.19863901490602723,
+      "grad_norm": 1.1246789693832397,
+      "learning_rate": 4.982850575648545e-06,
+      "loss": 0.1428,
+      "step": 1226
+    },
+    {
+      "epoch": 0.19880103694102397,
+      "grad_norm": 1.2873611450195312,
+      "learning_rate": 4.982799403961067e-06,
+      "loss": 0.1658,
+      "step": 1227
+    },
+    {
+      "epoch": 0.19896305897602073,
+      "grad_norm": 1.4319779872894287,
+      "learning_rate": 4.982748156305934e-06,
+      "loss": 0.1705,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1991250810110175,
+      "grad_norm": 1.3176896572113037,
+      "learning_rate": 4.982696832684716e-06,
+      "loss": 0.1681,
+      "step": 1229
+    },
+    {
+      "epoch": 0.19928710304601427,
+      "grad_norm": 1.3779314756393433,
+      "learning_rate": 4.982645433098984e-06,
+      "loss": 0.1747,
+      "step": 1230
+    },
+    {
+      "epoch": 0.199449125081011,
+      "grad_norm": 1.322016954421997,
+      "learning_rate": 4.982593957550308e-06,
+      "loss": 0.1696,
+      "step": 1231
+    },
+    {
+      "epoch": 0.19961114711600778,
+      "grad_norm": 1.2837717533111572,
+      "learning_rate": 4.982542406040266e-06,
+      "loss": 0.1536,
+      "step": 1232
+    },
+    {
+      "epoch": 0.19977316915100454,
+      "grad_norm": 1.2116730213165283,
+      "learning_rate": 4.982490778570434e-06,
+      "loss": 0.174,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1999351911860013,
+      "grad_norm": 1.3564155101776123,
+      "learning_rate": 4.98243907514239e-06,
+      "loss": 0.1875,
+      "step": 1234
+    },
+    {
+      "epoch": 0.20009721322099805,
+      "grad_norm": 1.15635085105896,
+      "learning_rate": 4.982387295757719e-06,
+      "loss": 0.1603,
+      "step": 1235
+    },
+    {
+      "epoch": 0.20025923525599482,
+      "grad_norm": 1.2285187244415283,
+      "learning_rate": 4.982335440418004e-06,
+      "loss": 0.1765,
+      "step": 1236
+    },
+    {
+      "epoch": 0.20042125729099158,
+      "grad_norm": 1.1799185276031494,
+      "learning_rate": 4.982283509124831e-06,
+      "loss": 0.1598,
+      "step": 1237
+    },
+    {
+      "epoch": 0.20058327932598832,
+      "grad_norm": 1.2755019664764404,
+      "learning_rate": 4.98223150187979e-06,
+      "loss": 0.1781,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2007453013609851,
+      "grad_norm": 1.2312284708023071,
+      "learning_rate": 4.982179418684473e-06,
+      "loss": 0.1673,
+      "step": 1239
+    },
+    {
+      "epoch": 0.20090732339598186,
+      "grad_norm": 1.0840716361999512,
+      "learning_rate": 4.982127259540471e-06,
+      "loss": 0.1442,
+      "step": 1240
+    },
+    {
+      "epoch": 0.20106934543097862,
+      "grad_norm": 1.2539925575256348,
+      "learning_rate": 4.9820750244493825e-06,
+      "loss": 0.157,
+      "step": 1241
+    },
+    {
+      "epoch": 0.20123136746597536,
+      "grad_norm": 1.1770260334014893,
+      "learning_rate": 4.9820227134128045e-06,
+      "loss": 0.1677,
+      "step": 1242
+    },
+    {
+      "epoch": 0.20139338950097213,
+      "grad_norm": 1.3164162635803223,
+      "learning_rate": 4.9819703264323375e-06,
+      "loss": 0.163,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2015554115359689,
+      "grad_norm": 1.240263819694519,
+      "learning_rate": 4.981917863509585e-06,
+      "loss": 0.1804,
+      "step": 1244
+    },
+    {
+      "epoch": 0.20171743357096567,
+      "grad_norm": 1.2952622175216675,
+      "learning_rate": 4.981865324646152e-06,
+      "loss": 0.1781,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2018794556059624,
+      "grad_norm": 1.1874738931655884,
+      "learning_rate": 4.981812709843646e-06,
+      "loss": 0.1665,
+      "step": 1246
+    },
+    {
+      "epoch": 0.20204147764095917,
+      "grad_norm": 1.251424789428711,
+      "learning_rate": 4.981760019103677e-06,
+      "loss": 0.1743,
+      "step": 1247
+    },
+    {
+      "epoch": 0.20220349967595594,
+      "grad_norm": 1.243334174156189,
+      "learning_rate": 4.981707252427857e-06,
+      "loss": 0.166,
+      "step": 1248
+    },
+    {
+      "epoch": 0.20236552171095268,
+      "grad_norm": 1.208656668663025,
+      "learning_rate": 4.981654409817801e-06,
+      "loss": 0.1696,
+      "step": 1249
+    },
+    {
+      "epoch": 0.20252754374594945,
+      "grad_norm": 1.227049708366394,
+      "learning_rate": 4.981601491275125e-06,
+      "loss": 0.1641,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2026895657809462,
+      "grad_norm": 1.4524271488189697,
+      "learning_rate": 4.981548496801449e-06,
+      "loss": 0.1648,
+      "step": 1251
+    },
+    {
+      "epoch": 0.20285158781594298,
+      "grad_norm": 1.2255809307098389,
+      "learning_rate": 4.981495426398395e-06,
+      "loss": 0.1596,
+      "step": 1252
+    },
+    {
+      "epoch": 0.20301360985093972,
+      "grad_norm": 1.2606754302978516,
+      "learning_rate": 4.981442280067585e-06,
+      "loss": 0.1745,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2031756318859365,
+      "grad_norm": 1.145128846168518,
+      "learning_rate": 4.981389057810647e-06,
+      "loss": 0.1593,
+      "step": 1254
+    },
+    {
+      "epoch": 0.20333765392093325,
+      "grad_norm": 1.2478601932525635,
+      "learning_rate": 4.981335759629208e-06,
+      "loss": 0.1837,
+      "step": 1255
+    },
+    {
+      "epoch": 0.20349967595593,
+      "grad_norm": 1.1817132234573364,
+      "learning_rate": 4.9812823855248996e-06,
+      "loss": 0.152,
+      "step": 1256
+    },
+    {
+      "epoch": 0.20366169799092676,
+      "grad_norm": 1.2523393630981445,
+      "learning_rate": 4.981228935499355e-06,
+      "loss": 0.1685,
+      "step": 1257
+    },
+    {
+      "epoch": 0.20382372002592353,
+      "grad_norm": 1.1938599348068237,
+      "learning_rate": 4.98117540955421e-06,
+      "loss": 0.1645,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2039857420609203,
+      "grad_norm": 1.2206389904022217,
+      "learning_rate": 4.981121807691101e-06,
+      "loss": 0.1622,
+      "step": 1259
+    },
+    {
+      "epoch": 0.20414776409591703,
+      "grad_norm": 1.2022802829742432,
+      "learning_rate": 4.981068129911669e-06,
+      "loss": 0.1563,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2043097861309138,
+      "grad_norm": 1.244292974472046,
+      "learning_rate": 4.981014376217556e-06,
+      "loss": 0.1593,
+      "step": 1261
+    },
+    {
+      "epoch": 0.20447180816591057,
+      "grad_norm": 1.2590563297271729,
+      "learning_rate": 4.980960546610408e-06,
+      "loss": 0.1762,
+      "step": 1262
+    },
+    {
+      "epoch": 0.20463383020090734,
+      "grad_norm": 1.3502979278564453,
+      "learning_rate": 4.98090664109187e-06,
+      "loss": 0.1802,
+      "step": 1263
+    },
+    {
+      "epoch": 0.20479585223590407,
+      "grad_norm": 1.1320891380310059,
+      "learning_rate": 4.980852659663593e-06,
+      "loss": 0.1658,
+      "step": 1264
+    },
+    {
+      "epoch": 0.20495787427090084,
+      "grad_norm": 1.092449426651001,
+      "learning_rate": 4.980798602327228e-06,
+      "loss": 0.1452,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2051198963058976,
+      "grad_norm": 1.2534480094909668,
+      "learning_rate": 4.9807444690844296e-06,
+      "loss": 0.1628,
+      "step": 1266
+    },
+    {
+      "epoch": 0.20528191834089435,
+      "grad_norm": 1.4840904474258423,
+      "learning_rate": 4.980690259936853e-06,
+      "loss": 0.1727,
+      "step": 1267
+    },
+    {
+      "epoch": 0.20544394037589112,
+      "grad_norm": 1.1676522493362427,
+      "learning_rate": 4.980635974886158e-06,
+      "loss": 0.1553,
+      "step": 1268
+    },
+    {
+      "epoch": 0.20560596241088788,
+      "grad_norm": 1.1992971897125244,
+      "learning_rate": 4.980581613934005e-06,
+      "loss": 0.1611,
+      "step": 1269
+    },
+    {
+      "epoch": 0.20576798444588465,
+      "grad_norm": 1.3434457778930664,
+      "learning_rate": 4.980527177082058e-06,
+      "loss": 0.2013,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2059300064808814,
+      "grad_norm": 1.306654691696167,
+      "learning_rate": 4.980472664331982e-06,
+      "loss": 0.196,
+      "step": 1271
+    },
+    {
+      "epoch": 0.20609202851587816,
+      "grad_norm": 1.187699317932129,
+      "learning_rate": 4.980418075685445e-06,
+      "loss": 0.1684,
+      "step": 1272
+    },
+    {
+      "epoch": 0.20625405055087492,
+      "grad_norm": 1.3280266523361206,
+      "learning_rate": 4.980363411144117e-06,
+      "loss": 0.1887,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2064160725858717,
+      "grad_norm": 1.3123143911361694,
+      "learning_rate": 4.980308670709671e-06,
+      "loss": 0.1653,
+      "step": 1274
+    },
+    {
+      "epoch": 0.20657809462086843,
+      "grad_norm": 1.1524012088775635,
+      "learning_rate": 4.980253854383782e-06,
+      "loss": 0.1647,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2067401166558652,
+      "grad_norm": 1.745810866355896,
+      "learning_rate": 4.980198962168128e-06,
+      "loss": 0.1887,
+      "step": 1276
+    },
+    {
+      "epoch": 0.20690213869086196,
+      "grad_norm": 1.2560205459594727,
+      "learning_rate": 4.980143994064387e-06,
+      "loss": 0.1829,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2070641607258587,
+      "grad_norm": 1.1559257507324219,
+      "learning_rate": 4.9800889500742415e-06,
+      "loss": 0.1694,
+      "step": 1278
+    },
+    {
+      "epoch": 0.20722618276085547,
+      "grad_norm": 1.3864670991897583,
+      "learning_rate": 4.980033830199376e-06,
+      "loss": 0.1933,
+      "step": 1279
+    },
+    {
+      "epoch": 0.20738820479585224,
+      "grad_norm": 1.1323646306991577,
+      "learning_rate": 4.979978634441477e-06,
+      "loss": 0.1587,
+      "step": 1280
+    },
+    {
+      "epoch": 0.207550226830849,
+      "grad_norm": 1.1979477405548096,
+      "learning_rate": 4.979923362802233e-06,
+      "loss": 0.1691,
+      "step": 1281
+    },
+    {
+      "epoch": 0.20771224886584574,
+      "grad_norm": 1.0671766996383667,
+      "learning_rate": 4.979868015283336e-06,
+      "loss": 0.1499,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2078742709008425,
+      "grad_norm": 1.2722365856170654,
+      "learning_rate": 4.979812591886478e-06,
+      "loss": 0.1702,
+      "step": 1283
+    },
+    {
+      "epoch": 0.20803629293583928,
+      "grad_norm": 1.3761050701141357,
+      "learning_rate": 4.979757092613357e-06,
+      "loss": 0.1916,
+      "step": 1284
+    },
+    {
+      "epoch": 0.20819831497083605,
+      "grad_norm": 1.2674570083618164,
+      "learning_rate": 4.9797015174656685e-06,
+      "loss": 0.1792,
+      "step": 1285
+    },
+    {
+      "epoch": 0.20836033700583279,
+      "grad_norm": 1.072959065437317,
+      "learning_rate": 4.979645866445114e-06,
+      "loss": 0.1454,
+      "step": 1286
+    },
+    {
+      "epoch": 0.20852235904082955,
+      "grad_norm": 1.4338332414627075,
+      "learning_rate": 4.979590139553398e-06,
+      "loss": 0.1976,
+      "step": 1287
+    },
+    {
+      "epoch": 0.20868438107582632,
+      "grad_norm": 1.3137916326522827,
+      "learning_rate": 4.9795343367922235e-06,
+      "loss": 0.1822,
+      "step": 1288
+    },
+    {
+      "epoch": 0.20884640311082306,
+      "grad_norm": 1.1669158935546875,
+      "learning_rate": 4.9794784581632986e-06,
+      "loss": 0.1585,
+      "step": 1289
+    },
+    {
+      "epoch": 0.20900842514581983,
+      "grad_norm": 1.2235841751098633,
+      "learning_rate": 4.979422503668334e-06,
+      "loss": 0.1558,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2091704471808166,
+      "grad_norm": 1.178096890449524,
+      "learning_rate": 4.97936647330904e-06,
+      "loss": 0.1593,
+      "step": 1291
+    },
+    {
+      "epoch": 0.20933246921581336,
+      "grad_norm": 1.1539371013641357,
+      "learning_rate": 4.979310367087132e-06,
+      "loss": 0.1526,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2094944912508101,
+      "grad_norm": 1.1756463050842285,
+      "learning_rate": 4.979254185004327e-06,
+      "loss": 0.1486,
+      "step": 1293
+    },
+    {
+      "epoch": 0.20965651328580687,
+      "grad_norm": 1.4390196800231934,
+      "learning_rate": 4.979197927062343e-06,
+      "loss": 0.1937,
+      "step": 1294
+    },
+    {
+      "epoch": 0.20981853532080363,
+      "grad_norm": 1.195878505706787,
+      "learning_rate": 4.979141593262902e-06,
+      "loss": 0.1656,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2099805573558004,
+      "grad_norm": 1.1376088857650757,
+      "learning_rate": 4.979085183607728e-06,
+      "loss": 0.1668,
+      "step": 1296
+    },
+    {
+      "epoch": 0.21014257939079714,
+      "grad_norm": 1.0837515592575073,
+      "learning_rate": 4.979028698098546e-06,
+      "loss": 0.1448,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2103046014257939,
+      "grad_norm": 1.2839235067367554,
+      "learning_rate": 4.978972136737086e-06,
+      "loss": 0.1886,
+      "step": 1298
+    },
+    {
+      "epoch": 0.21046662346079067,
+      "grad_norm": 1.239888072013855,
+      "learning_rate": 4.978915499525077e-06,
+      "loss": 0.175,
+      "step": 1299
+    },
+    {
+      "epoch": 0.21062864549578741,
+      "grad_norm": 1.2794440984725952,
+      "learning_rate": 4.978858786464252e-06,
+      "loss": 0.1718,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21079066753078418,
+      "grad_norm": 1.2917277812957764,
+      "learning_rate": 4.978801997556348e-06,
+      "loss": 0.1761,
+      "step": 1301
+    },
+    {
+      "epoch": 0.21095268956578095,
+      "grad_norm": 1.2395626306533813,
+      "learning_rate": 4.978745132803101e-06,
+      "loss": 0.1647,
+      "step": 1302
+    },
+    {
+      "epoch": 0.21111471160077772,
+      "grad_norm": 1.2468279600143433,
+      "learning_rate": 4.9786881922062515e-06,
+      "loss": 0.1574,
+      "step": 1303
+    },
+    {
+      "epoch": 0.21127673363577446,
+      "grad_norm": 1.3926787376403809,
+      "learning_rate": 4.9786311757675425e-06,
+      "loss": 0.2085,
+      "step": 1304
+    },
+    {
+      "epoch": 0.21143875567077122,
+      "grad_norm": 1.1423304080963135,
+      "learning_rate": 4.978574083488716e-06,
+      "loss": 0.1471,
+      "step": 1305
+    },
+    {
+      "epoch": 0.211600777705768,
+      "grad_norm": 1.2583112716674805,
+      "learning_rate": 4.978516915371522e-06,
+      "loss": 0.1673,
+      "step": 1306
+    },
+    {
+      "epoch": 0.21176279974076476,
+      "grad_norm": 1.2514091730117798,
+      "learning_rate": 4.978459671417707e-06,
+      "loss": 0.1794,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2119248217757615,
+      "grad_norm": 1.5359858274459839,
+      "learning_rate": 4.978402351629024e-06,
+      "loss": 0.1582,
+      "step": 1308
+    },
+    {
+      "epoch": 0.21208684381075826,
+      "grad_norm": 1.161242127418518,
+      "learning_rate": 4.978344956007227e-06,
+      "loss": 0.1542,
+      "step": 1309
+    },
+    {
+      "epoch": 0.21224886584575503,
+      "grad_norm": 1.2749699354171753,
+      "learning_rate": 4.9782874845540715e-06,
+      "loss": 0.1607,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21241088788075177,
+      "grad_norm": 1.3189657926559448,
+      "learning_rate": 4.978229937271317e-06,
+      "loss": 0.1844,
+      "step": 1311
+    },
+    {
+      "epoch": 0.21257290991574854,
+      "grad_norm": 1.31024169921875,
+      "learning_rate": 4.978172314160724e-06,
+      "loss": 0.1992,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2127349319507453,
+      "grad_norm": 1.2073999643325806,
+      "learning_rate": 4.978114615224055e-06,
+      "loss": 0.1685,
+      "step": 1313
+    },
+    {
+      "epoch": 0.21289695398574207,
+      "grad_norm": 1.292680263519287,
+      "learning_rate": 4.9780568404630746e-06,
+      "loss": 0.1758,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2130589760207388,
+      "grad_norm": 1.2444231510162354,
+      "learning_rate": 4.977998989879552e-06,
+      "loss": 0.1771,
+      "step": 1315
+    },
+    {
+      "epoch": 0.21322099805573558,
+      "grad_norm": 1.1644980907440186,
+      "learning_rate": 4.977941063475258e-06,
+      "loss": 0.1637,
+      "step": 1316
+    },
+    {
+      "epoch": 0.21338302009073234,
+      "grad_norm": 1.2748589515686035,
+      "learning_rate": 4.977883061251962e-06,
+      "loss": 0.1813,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2135450421257291,
+      "grad_norm": 1.2317109107971191,
+      "learning_rate": 4.977824983211443e-06,
+      "loss": 0.1654,
+      "step": 1318
+    },
+    {
+      "epoch": 0.21370706416072585,
+      "grad_norm": 1.2727543115615845,
+      "learning_rate": 4.977766829355474e-06,
+      "loss": 0.1848,
+      "step": 1319
+    },
+    {
+      "epoch": 0.21386908619572262,
+      "grad_norm": 1.2605267763137817,
+      "learning_rate": 4.977708599685837e-06,
+      "loss": 0.1881,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21403110823071939,
+      "grad_norm": 1.2069200277328491,
+      "learning_rate": 4.977650294204313e-06,
+      "loss": 0.1601,
+      "step": 1321
+    },
+    {
+      "epoch": 0.21419313026571613,
+      "grad_norm": 1.2258086204528809,
+      "learning_rate": 4.977591912912685e-06,
+      "loss": 0.1709,
+      "step": 1322
+    },
+    {
+      "epoch": 0.2143551523007129,
+      "grad_norm": 1.255131721496582,
+      "learning_rate": 4.977533455812741e-06,
+      "loss": 0.173,
+      "step": 1323
+    },
+    {
+      "epoch": 0.21451717433570966,
+      "grad_norm": 1.2011017799377441,
+      "learning_rate": 4.977474922906268e-06,
+      "loss": 0.1784,
+      "step": 1324
+    },
+    {
+      "epoch": 0.21467919637070643,
+      "grad_norm": 1.1094778776168823,
+      "learning_rate": 4.977416314195058e-06,
+      "loss": 0.1653,
+      "step": 1325
+    },
+    {
+      "epoch": 0.21484121840570317,
+      "grad_norm": 1.3300212621688843,
+      "learning_rate": 4.977357629680903e-06,
+      "loss": 0.178,
+      "step": 1326
+    },
+    {
+      "epoch": 0.21500324044069993,
+      "grad_norm": 1.190017819404602,
+      "learning_rate": 4.977298869365601e-06,
+      "loss": 0.1671,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2151652624756967,
+      "grad_norm": 1.2279374599456787,
+      "learning_rate": 4.977240033250948e-06,
+      "loss": 0.1738,
+      "step": 1328
+    },
+    {
+      "epoch": 0.21532728451069347,
+      "grad_norm": 1.1839739084243774,
+      "learning_rate": 4.977181121338745e-06,
+      "loss": 0.1663,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2154893065456902,
+      "grad_norm": 1.1655634641647339,
+      "learning_rate": 4.977122133630795e-06,
+      "loss": 0.148,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21565132858068697,
+      "grad_norm": 1.151414394378662,
+      "learning_rate": 4.977063070128902e-06,
+      "loss": 0.1716,
+      "step": 1331
+    },
+    {
+      "epoch": 0.21581335061568374,
+      "grad_norm": 1.125185489654541,
+      "learning_rate": 4.9770039308348725e-06,
+      "loss": 0.1489,
+      "step": 1332
+    },
+    {
+      "epoch": 0.21597537265068048,
+      "grad_norm": 1.1280821561813354,
+      "learning_rate": 4.976944715750517e-06,
+      "loss": 0.1624,
+      "step": 1333
+    },
+    {
+      "epoch": 0.21613739468567725,
+      "grad_norm": 1.2168023586273193,
+      "learning_rate": 4.9768854248776475e-06,
+      "loss": 0.168,
+      "step": 1334
+    },
+    {
+      "epoch": 0.21629941672067401,
+      "grad_norm": 1.1852880716323853,
+      "learning_rate": 4.976826058218079e-06,
+      "loss": 0.1585,
+      "step": 1335
+    },
+    {
+      "epoch": 0.21646143875567078,
+      "grad_norm": 1.210832953453064,
+      "learning_rate": 4.976766615773626e-06,
+      "loss": 0.1665,
+      "step": 1336
+    },
+    {
+      "epoch": 0.21662346079066752,
+      "grad_norm": 1.44858980178833,
+      "learning_rate": 4.9767070975461075e-06,
+      "loss": 0.1918,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2167854828256643,
+      "grad_norm": 1.1260679960250854,
+      "learning_rate": 4.976647503537347e-06,
+      "loss": 0.1658,
+      "step": 1338
+    },
+    {
+      "epoch": 0.21694750486066106,
+      "grad_norm": 1.1670085191726685,
+      "learning_rate": 4.976587833749164e-06,
+      "loss": 0.1604,
+      "step": 1339
+    },
+    {
+      "epoch": 0.21710952689565782,
+      "grad_norm": 1.2441238164901733,
+      "learning_rate": 4.9765280881833885e-06,
+      "loss": 0.1674,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21727154893065456,
+      "grad_norm": 1.1147247552871704,
+      "learning_rate": 4.976468266841846e-06,
+      "loss": 0.1441,
+      "step": 1341
+    },
+    {
+      "epoch": 0.21743357096565133,
+      "grad_norm": 1.1170032024383545,
+      "learning_rate": 4.976408369726368e-06,
+      "loss": 0.153,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2175955930006481,
+      "grad_norm": 1.2686792612075806,
+      "learning_rate": 4.976348396838786e-06,
+      "loss": 0.1759,
+      "step": 1343
+    },
+    {
+      "epoch": 0.21775761503564484,
+      "grad_norm": 1.1871846914291382,
+      "learning_rate": 4.976288348180935e-06,
+      "loss": 0.1524,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2179196370706416,
+      "grad_norm": 1.2280350923538208,
+      "learning_rate": 4.976228223754654e-06,
+      "loss": 0.1729,
+      "step": 1345
+    },
+    {
+      "epoch": 0.21808165910563837,
+      "grad_norm": 1.1394003629684448,
+      "learning_rate": 4.976168023561782e-06,
+      "loss": 0.1569,
+      "step": 1346
+    },
+    {
+      "epoch": 0.21824368114063514,
+      "grad_norm": 1.1886025667190552,
+      "learning_rate": 4.976107747604161e-06,
+      "loss": 0.1539,
+      "step": 1347
+    },
+    {
+      "epoch": 0.21840570317563188,
+      "grad_norm": 1.199474811553955,
+      "learning_rate": 4.976047395883634e-06,
+      "loss": 0.1721,
+      "step": 1348
+    },
+    {
+      "epoch": 0.21856772521062864,
+      "grad_norm": 1.3025383949279785,
+      "learning_rate": 4.975986968402048e-06,
+      "loss": 0.1944,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2187297472456254,
+      "grad_norm": 1.2892481088638306,
+      "learning_rate": 4.975926465161254e-06,
+      "loss": 0.1628,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21889176928062218,
+      "grad_norm": 1.1720067262649536,
+      "learning_rate": 4.975865886163101e-06,
+      "loss": 0.1611,
+      "step": 1351
+    },
+    {
+      "epoch": 0.21905379131561892,
+      "grad_norm": 1.284433364868164,
+      "learning_rate": 4.975805231409444e-06,
+      "loss": 0.1986,
+      "step": 1352
+    },
+    {
+      "epoch": 0.21921581335061568,
+      "grad_norm": 1.3091256618499756,
+      "learning_rate": 4.975744500902138e-06,
+      "loss": 0.173,
+      "step": 1353
+    },
+    {
+      "epoch": 0.21937783538561245,
+      "grad_norm": 1.1260159015655518,
+      "learning_rate": 4.975683694643041e-06,
+      "loss": 0.1485,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2195398574206092,
+      "grad_norm": 1.1246238946914673,
+      "learning_rate": 4.975622812634014e-06,
+      "loss": 0.1565,
+      "step": 1355
+    },
+    {
+      "epoch": 0.21970187945560596,
+      "grad_norm": 1.2171014547348022,
+      "learning_rate": 4.97556185487692e-06,
+      "loss": 0.1701,
+      "step": 1356
+    },
+    {
+      "epoch": 0.21986390149060273,
+      "grad_norm": 1.202234148979187,
+      "learning_rate": 4.975500821373624e-06,
+      "loss": 0.146,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2200259235255995,
+      "grad_norm": 1.1403602361679077,
+      "learning_rate": 4.9754397121259935e-06,
+      "loss": 0.1585,
+      "step": 1358
+    },
+    {
+      "epoch": 0.22018794556059623,
+      "grad_norm": 1.2086261510849,
+      "learning_rate": 4.975378527135899e-06,
+      "loss": 0.166,
+      "step": 1359
+    },
+    {
+      "epoch": 0.220349967595593,
+      "grad_norm": 1.2435903549194336,
+      "learning_rate": 4.975317266405211e-06,
+      "loss": 0.1783,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22051198963058977,
+      "grad_norm": 1.0744258165359497,
+      "learning_rate": 4.975255929935805e-06,
+      "loss": 0.1591,
+      "step": 1361
+    },
+    {
+      "epoch": 0.22067401166558653,
+      "grad_norm": 1.2490445375442505,
+      "learning_rate": 4.975194517729557e-06,
+      "loss": 0.1816,
+      "step": 1362
+    },
+    {
+      "epoch": 0.22083603370058327,
+      "grad_norm": 1.1408127546310425,
+      "learning_rate": 4.975133029788347e-06,
+      "loss": 0.1566,
+      "step": 1363
+    },
+    {
+      "epoch": 0.22099805573558004,
+      "grad_norm": 1.2081525325775146,
+      "learning_rate": 4.975071466114057e-06,
+      "loss": 0.1567,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2211600777705768,
+      "grad_norm": 1.1004225015640259,
+      "learning_rate": 4.975009826708568e-06,
+      "loss": 0.1483,
+      "step": 1365
+    },
+    {
+      "epoch": 0.22132209980557355,
+      "grad_norm": 1.3118537664413452,
+      "learning_rate": 4.974948111573768e-06,
+      "loss": 0.1818,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2214841218405703,
+      "grad_norm": 1.1786134243011475,
+      "learning_rate": 4.974886320711546e-06,
+      "loss": 0.1716,
+      "step": 1367
+    },
+    {
+      "epoch": 0.22164614387556708,
+      "grad_norm": 1.2705552577972412,
+      "learning_rate": 4.9748244541237915e-06,
+      "loss": 0.1852,
+      "step": 1368
+    },
+    {
+      "epoch": 0.22180816591056385,
+      "grad_norm": 1.1799037456512451,
+      "learning_rate": 4.974762511812398e-06,
+      "loss": 0.1414,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2219701879455606,
+      "grad_norm": 1.1992617845535278,
+      "learning_rate": 4.97470049377926e-06,
+      "loss": 0.1662,
+      "step": 1370
+    },
+    {
+      "epoch": 0.22213220998055735,
+      "grad_norm": 1.1669738292694092,
+      "learning_rate": 4.974638400026275e-06,
+      "loss": 0.1544,
+      "step": 1371
+    },
+    {
+      "epoch": 0.22229423201555412,
+      "grad_norm": 1.2163314819335938,
+      "learning_rate": 4.974576230555344e-06,
+      "loss": 0.1643,
+      "step": 1372
+    },
+    {
+      "epoch": 0.22245625405055086,
+      "grad_norm": 1.3139532804489136,
+      "learning_rate": 4.9745139853683685e-06,
+      "loss": 0.1617,
+      "step": 1373
+    },
+    {
+      "epoch": 0.22261827608554763,
+      "grad_norm": 1.1525776386260986,
+      "learning_rate": 4.974451664467253e-06,
+      "loss": 0.1752,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2227802981205444,
+      "grad_norm": 1.110068917274475,
+      "learning_rate": 4.974389267853905e-06,
+      "loss": 0.1511,
+      "step": 1375
+    },
+    {
+      "epoch": 0.22294232015554116,
+      "grad_norm": 1.1273125410079956,
+      "learning_rate": 4.974326795530234e-06,
+      "loss": 0.1528,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2231043421905379,
+      "grad_norm": 1.3362386226654053,
+      "learning_rate": 4.97426424749815e-06,
+      "loss": 0.2061,
+      "step": 1377
+    },
+    {
+      "epoch": 0.22326636422553467,
+      "grad_norm": 1.0781415700912476,
+      "learning_rate": 4.974201623759568e-06,
+      "loss": 0.1522,
+      "step": 1378
+    },
+    {
+      "epoch": 0.22342838626053144,
+      "grad_norm": 1.1233181953430176,
+      "learning_rate": 4.974138924316403e-06,
+      "loss": 0.1601,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2235904082955282,
+      "grad_norm": 1.0733990669250488,
+      "learning_rate": 4.974076149170575e-06,
+      "loss": 0.1613,
+      "step": 1380
+    },
+    {
+      "epoch": 0.22375243033052494,
+      "grad_norm": 0.9839252233505249,
+      "learning_rate": 4.9740132983240036e-06,
+      "loss": 0.1275,
+      "step": 1381
+    },
+    {
+      "epoch": 0.2239144523655217,
+      "grad_norm": 1.2110671997070312,
+      "learning_rate": 4.973950371778612e-06,
+      "loss": 0.1764,
+      "step": 1382
+    },
+    {
+      "epoch": 0.22407647440051848,
+      "grad_norm": 1.2692327499389648,
+      "learning_rate": 4.973887369536326e-06,
+      "loss": 0.1741,
+      "step": 1383
+    },
+    {
+      "epoch": 0.22423849643551522,
+      "grad_norm": 1.226586937904358,
+      "learning_rate": 4.973824291599074e-06,
+      "loss": 0.1578,
+      "step": 1384
+    },
+    {
+      "epoch": 0.22440051847051198,
+      "grad_norm": 1.1944162845611572,
+      "learning_rate": 4.973761137968784e-06,
+      "loss": 0.1543,
+      "step": 1385
+    },
+    {
+      "epoch": 0.22456254050550875,
+      "grad_norm": 1.3723398447036743,
+      "learning_rate": 4.973697908647391e-06,
+      "loss": 0.1834,
+      "step": 1386
+    },
+    {
+      "epoch": 0.22472456254050552,
+      "grad_norm": 1.196521520614624,
+      "learning_rate": 4.973634603636828e-06,
+      "loss": 0.1519,
+      "step": 1387
+    },
+    {
+      "epoch": 0.22488658457550226,
+      "grad_norm": 1.104246735572815,
+      "learning_rate": 4.973571222939031e-06,
+      "loss": 0.1552,
+      "step": 1388
+    },
+    {
+      "epoch": 0.22504860661049902,
+      "grad_norm": 1.137224555015564,
+      "learning_rate": 4.973507766555941e-06,
+      "loss": 0.1628,
+      "step": 1389
+    },
+    {
+      "epoch": 0.2252106286454958,
+      "grad_norm": 1.2728477716445923,
+      "learning_rate": 4.973444234489499e-06,
+      "loss": 0.1687,
+      "step": 1390
+    },
+    {
+      "epoch": 0.22537265068049256,
+      "grad_norm": 1.2384727001190186,
+      "learning_rate": 4.97338062674165e-06,
+      "loss": 0.1916,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2255346727154893,
+      "grad_norm": 1.280981183052063,
+      "learning_rate": 4.973316943314338e-06,
+      "loss": 0.1612,
+      "step": 1392
+    },
+    {
+      "epoch": 0.22569669475048607,
+      "grad_norm": 1.1161082983016968,
+      "learning_rate": 4.9732531842095135e-06,
+      "loss": 0.1507,
+      "step": 1393
+    },
+    {
+      "epoch": 0.22585871678548283,
+      "grad_norm": 1.3085819482803345,
+      "learning_rate": 4.9731893494291275e-06,
+      "loss": 0.192,
+      "step": 1394
+    },
+    {
+      "epoch": 0.22602073882047957,
+      "grad_norm": 1.1078639030456543,
+      "learning_rate": 4.973125438975131e-06,
+      "loss": 0.1516,
+      "step": 1395
+    },
+    {
+      "epoch": 0.22618276085547634,
+      "grad_norm": 1.1671632528305054,
+      "learning_rate": 4.973061452849481e-06,
+      "loss": 0.1693,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2263447828904731,
+      "grad_norm": 1.103091835975647,
+      "learning_rate": 4.9729973910541365e-06,
+      "loss": 0.1415,
+      "step": 1397
+    },
+    {
+      "epoch": 0.22650680492546987,
+      "grad_norm": 1.1498920917510986,
+      "learning_rate": 4.972933253591056e-06,
+      "loss": 0.1562,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2266688269604666,
+      "grad_norm": 1.1860241889953613,
+      "learning_rate": 4.972869040462202e-06,
+      "loss": 0.1641,
+      "step": 1399
+    },
+    {
+      "epoch": 0.22683084899546338,
+      "grad_norm": 1.2506409883499146,
+      "learning_rate": 4.972804751669539e-06,
+      "loss": 0.1622,
+      "step": 1400
+    },
+    {
+      "epoch": 0.22699287103046015,
+      "grad_norm": 1.1482832431793213,
+      "learning_rate": 4.9727403872150345e-06,
+      "loss": 0.1506,
+      "step": 1401
+    },
+    {
+      "epoch": 0.22715489306545691,
+      "grad_norm": 1.3503156900405884,
+      "learning_rate": 4.972675947100659e-06,
+      "loss": 0.2058,
+      "step": 1402
+    },
+    {
+      "epoch": 0.22731691510045365,
+      "grad_norm": 1.2772489786148071,
+      "learning_rate": 4.972611431328384e-06,
+      "loss": 0.1633,
+      "step": 1403
+    },
+    {
+      "epoch": 0.22747893713545042,
+      "grad_norm": 1.2101973295211792,
+      "learning_rate": 4.972546839900181e-06,
+      "loss": 0.1793,
+      "step": 1404
+    },
+    {
+      "epoch": 0.2276409591704472,
+      "grad_norm": 1.225361943244934,
+      "learning_rate": 4.972482172818029e-06,
+      "loss": 0.174,
+      "step": 1405
+    },
+    {
+      "epoch": 0.22780298120544393,
+      "grad_norm": 1.211272120475769,
+      "learning_rate": 4.972417430083906e-06,
+      "loss": 0.1821,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2279650032404407,
+      "grad_norm": 1.1716586351394653,
+      "learning_rate": 4.9723526116997925e-06,
+      "loss": 0.1567,
+      "step": 1407
+    },
+    {
+      "epoch": 0.22812702527543746,
+      "grad_norm": 1.2331879138946533,
+      "learning_rate": 4.972287717667672e-06,
+      "loss": 0.1849,
+      "step": 1408
+    },
+    {
+      "epoch": 0.22828904731043423,
+      "grad_norm": 1.1068236827850342,
+      "learning_rate": 4.972222747989531e-06,
+      "loss": 0.1506,
+      "step": 1409
+    },
+    {
+      "epoch": 0.22845106934543097,
+      "grad_norm": 1.0929367542266846,
+      "learning_rate": 4.972157702667356e-06,
+      "loss": 0.1625,
+      "step": 1410
+    },
+    {
+      "epoch": 0.22861309138042774,
+      "grad_norm": 1.1465107202529907,
+      "learning_rate": 4.972092581703138e-06,
+      "loss": 0.1676,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2287751134154245,
+      "grad_norm": 1.1976622343063354,
+      "learning_rate": 4.972027385098868e-06,
+      "loss": 0.1546,
+      "step": 1412
+    },
+    {
+      "epoch": 0.22893713545042127,
+      "grad_norm": 1.201499104499817,
+      "learning_rate": 4.971962112856543e-06,
+      "loss": 0.1639,
+      "step": 1413
+    },
+    {
+      "epoch": 0.229099157485418,
+      "grad_norm": 1.1784822940826416,
+      "learning_rate": 4.97189676497816e-06,
+      "loss": 0.1709,
+      "step": 1414
+    },
+    {
+      "epoch": 0.22926117952041478,
+      "grad_norm": 1.2148628234863281,
+      "learning_rate": 4.971831341465718e-06,
+      "loss": 0.1641,
+      "step": 1415
+    },
+    {
+      "epoch": 0.22942320155541154,
+      "grad_norm": 1.1798665523529053,
+      "learning_rate": 4.971765842321218e-06,
+      "loss": 0.1761,
+      "step": 1416
+    },
+    {
+      "epoch": 0.22958522359040828,
+      "grad_norm": 1.0728332996368408,
+      "learning_rate": 4.9717002675466645e-06,
+      "loss": 0.1542,
+      "step": 1417
+    },
+    {
+      "epoch": 0.22974724562540505,
+      "grad_norm": 1.094524621963501,
+      "learning_rate": 4.971634617144065e-06,
+      "loss": 0.1603,
+      "step": 1418
+    },
+    {
+      "epoch": 0.22990926766040182,
+      "grad_norm": 1.1839401721954346,
+      "learning_rate": 4.971568891115427e-06,
+      "loss": 0.1517,
+      "step": 1419
+    },
+    {
+      "epoch": 0.23007128969539858,
+      "grad_norm": 1.2270963191986084,
+      "learning_rate": 4.971503089462762e-06,
+      "loss": 0.1751,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23023331173039532,
+      "grad_norm": 1.342659831047058,
+      "learning_rate": 4.971437212188084e-06,
+      "loss": 0.1662,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2303953337653921,
+      "grad_norm": 1.0489001274108887,
+      "learning_rate": 4.9713712592934075e-06,
+      "loss": 0.1384,
+      "step": 1422
+    },
+    {
+      "epoch": 0.23055735580038886,
+      "grad_norm": 1.2548742294311523,
+      "learning_rate": 4.971305230780751e-06,
+      "loss": 0.1829,
+      "step": 1423
+    },
+    {
+      "epoch": 0.23071937783538563,
+      "grad_norm": 1.1241710186004639,
+      "learning_rate": 4.971239126652135e-06,
+      "loss": 0.1583,
+      "step": 1424
+    },
+    {
+      "epoch": 0.23088139987038236,
+      "grad_norm": 1.1561769247055054,
+      "learning_rate": 4.971172946909582e-06,
+      "loss": 0.1747,
+      "step": 1425
+    },
+    {
+      "epoch": 0.23104342190537913,
+      "grad_norm": 1.1572365760803223,
+      "learning_rate": 4.971106691555116e-06,
+      "loss": 0.1447,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2312054439403759,
+      "grad_norm": 1.766800880432129,
+      "learning_rate": 4.971040360590767e-06,
+      "loss": 0.1769,
+      "step": 1427
+    },
+    {
+      "epoch": 0.23136746597537264,
+      "grad_norm": 1.1585261821746826,
+      "learning_rate": 4.9709739540185616e-06,
+      "loss": 0.1782,
+      "step": 1428
+    },
+    {
+      "epoch": 0.2315294880103694,
+      "grad_norm": 0.9941073656082153,
+      "learning_rate": 4.9709074718405335e-06,
+      "loss": 0.1433,
+      "step": 1429
+    },
+    {
+      "epoch": 0.23169151004536617,
+      "grad_norm": 1.3293631076812744,
+      "learning_rate": 4.970840914058716e-06,
+      "loss": 0.1575,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23185353208036294,
+      "grad_norm": 1.0774224996566772,
+      "learning_rate": 4.970774280675146e-06,
+      "loss": 0.1464,
+      "step": 1431
+    },
+    {
+      "epoch": 0.23201555411535968,
+      "grad_norm": 1.1900949478149414,
+      "learning_rate": 4.970707571691862e-06,
+      "loss": 0.1418,
+      "step": 1432
+    },
+    {
+      "epoch": 0.23217757615035645,
+      "grad_norm": 1.2267847061157227,
+      "learning_rate": 4.9706407871109056e-06,
+      "loss": 0.1644,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2323395981853532,
+      "grad_norm": 1.3006565570831299,
+      "learning_rate": 4.970573926934319e-06,
+      "loss": 0.1772,
+      "step": 1434
+    },
+    {
+      "epoch": 0.23250162022034998,
+      "grad_norm": 1.324775218963623,
+      "learning_rate": 4.97050699116415e-06,
+      "loss": 0.1835,
+      "step": 1435
+    },
+    {
+      "epoch": 0.23266364225534672,
+      "grad_norm": 1.1360862255096436,
+      "learning_rate": 4.970439979802445e-06,
+      "loss": 0.1578,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2328256642903435,
+      "grad_norm": 1.1616891622543335,
+      "learning_rate": 4.970372892851255e-06,
+      "loss": 0.1525,
+      "step": 1437
+    },
+    {
+      "epoch": 0.23298768632534025,
+      "grad_norm": 1.2094240188598633,
+      "learning_rate": 4.970305730312632e-06,
+      "loss": 0.1678,
+      "step": 1438
+    },
+    {
+      "epoch": 0.233149708360337,
+      "grad_norm": 1.3469047546386719,
+      "learning_rate": 4.970238492188633e-06,
+      "loss": 0.1741,
+      "step": 1439
+    },
+    {
+      "epoch": 0.23331173039533376,
+      "grad_norm": 1.2809699773788452,
+      "learning_rate": 4.9701711784813135e-06,
+      "loss": 0.1724,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23347375243033053,
+      "grad_norm": 1.1660102605819702,
+      "learning_rate": 4.970103789192734e-06,
+      "loss": 0.1461,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2336357744653273,
+      "grad_norm": 1.1002432107925415,
+      "learning_rate": 4.970036324324955e-06,
+      "loss": 0.1546,
+      "step": 1442
+    },
+    {
+      "epoch": 0.23379779650032403,
+      "grad_norm": 1.0596157312393188,
+      "learning_rate": 4.9699687838800425e-06,
+      "loss": 0.1511,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2339598185353208,
+      "grad_norm": 1.231737732887268,
+      "learning_rate": 4.969901167860063e-06,
+      "loss": 0.1683,
+      "step": 1444
+    },
+    {
+      "epoch": 0.23412184057031757,
+      "grad_norm": 1.239596962928772,
+      "learning_rate": 4.969833476267084e-06,
+      "loss": 0.1763,
+      "step": 1445
+    },
+    {
+      "epoch": 0.23428386260531434,
+      "grad_norm": 1.0822583436965942,
+      "learning_rate": 4.969765709103177e-06,
+      "loss": 0.1484,
+      "step": 1446
+    },
+    {
+      "epoch": 0.23444588464031108,
+      "grad_norm": 1.0738798379898071,
+      "learning_rate": 4.969697866370417e-06,
+      "loss": 0.1508,
+      "step": 1447
+    },
+    {
+      "epoch": 0.23460790667530784,
+      "grad_norm": 1.1625072956085205,
+      "learning_rate": 4.9696299480708785e-06,
+      "loss": 0.1582,
+      "step": 1448
+    },
+    {
+      "epoch": 0.2347699287103046,
+      "grad_norm": 1.0520604848861694,
+      "learning_rate": 4.969561954206641e-06,
+      "loss": 0.1381,
+      "step": 1449
+    },
+    {
+      "epoch": 0.23493195074530135,
+      "grad_norm": 1.2117533683776855,
+      "learning_rate": 4.969493884779783e-06,
+      "loss": 0.1823,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23509397278029812,
+      "grad_norm": 1.217532753944397,
+      "learning_rate": 4.969425739792388e-06,
+      "loss": 0.1778,
+      "step": 1451
+    },
+    {
+      "epoch": 0.23525599481529488,
+      "grad_norm": 1.1805583238601685,
+      "learning_rate": 4.969357519246542e-06,
+      "loss": 0.164,
+      "step": 1452
+    },
+    {
+      "epoch": 0.23541801685029165,
+      "grad_norm": 1.243613600730896,
+      "learning_rate": 4.96928922314433e-06,
+      "loss": 0.1634,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2355800388852884,
+      "grad_norm": 1.224203109741211,
+      "learning_rate": 4.9692208514878445e-06,
+      "loss": 0.1688,
+      "step": 1454
+    },
+    {
+      "epoch": 0.23574206092028516,
+      "grad_norm": 1.2090548276901245,
+      "learning_rate": 4.9691524042791765e-06,
+      "loss": 0.178,
+      "step": 1455
+    },
+    {
+      "epoch": 0.23590408295528192,
+      "grad_norm": 1.1282353401184082,
+      "learning_rate": 4.96908388152042e-06,
+      "loss": 0.1615,
+      "step": 1456
+    },
+    {
+      "epoch": 0.2360661049902787,
+      "grad_norm": 1.0806899070739746,
+      "learning_rate": 4.9690152832136715e-06,
+      "loss": 0.1528,
+      "step": 1457
+    },
+    {
+      "epoch": 0.23622812702527543,
+      "grad_norm": 1.3501883745193481,
+      "learning_rate": 4.968946609361031e-06,
+      "loss": 0.1745,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2363901490602722,
+      "grad_norm": 1.1516779661178589,
+      "learning_rate": 4.968877859964598e-06,
+      "loss": 0.1427,
+      "step": 1459
+    },
+    {
+      "epoch": 0.23655217109526896,
+      "grad_norm": 1.0786315202713013,
+      "learning_rate": 4.968809035026477e-06,
+      "loss": 0.1552,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2367141931302657,
+      "grad_norm": 1.1482332944869995,
+      "learning_rate": 4.968740134548775e-06,
+      "loss": 0.168,
+      "step": 1461
+    },
+    {
+      "epoch": 0.23687621516526247,
+      "grad_norm": 1.3289909362792969,
+      "learning_rate": 4.968671158533599e-06,
+      "loss": 0.1726,
+      "step": 1462
+    },
+    {
+      "epoch": 0.23703823720025924,
+      "grad_norm": 1.1058714389801025,
+      "learning_rate": 4.968602106983059e-06,
+      "loss": 0.1569,
+      "step": 1463
+    },
+    {
+      "epoch": 0.237200259235256,
+      "grad_norm": 1.113430142402649,
+      "learning_rate": 4.968532979899269e-06,
+      "loss": 0.1542,
+      "step": 1464
+    },
+    {
+      "epoch": 0.23736228127025275,
+      "grad_norm": 1.465322732925415,
+      "learning_rate": 4.968463777284342e-06,
+      "loss": 0.1846,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2375243033052495,
+      "grad_norm": 1.1434563398361206,
+      "learning_rate": 4.9683944991403985e-06,
+      "loss": 0.1574,
+      "step": 1466
+    },
+    {
+      "epoch": 0.23768632534024628,
+      "grad_norm": 1.3530895709991455,
+      "learning_rate": 4.9683251454695565e-06,
+      "loss": 0.1646,
+      "step": 1467
+    },
+    {
+      "epoch": 0.23784834737524305,
+      "grad_norm": 1.193445086479187,
+      "learning_rate": 4.968255716273938e-06,
+      "loss": 0.1696,
+      "step": 1468
+    },
+    {
+      "epoch": 0.23801036941023979,
+      "grad_norm": 1.2225266695022583,
+      "learning_rate": 4.968186211555668e-06,
+      "loss": 0.1669,
+      "step": 1469
+    },
+    {
+      "epoch": 0.23817239144523655,
+      "grad_norm": 1.0870074033737183,
+      "learning_rate": 4.968116631316873e-06,
+      "loss": 0.1507,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23833441348023332,
+      "grad_norm": 1.3210666179656982,
+      "learning_rate": 4.968046975559681e-06,
+      "loss": 0.1747,
+      "step": 1471
+    },
+    {
+      "epoch": 0.23849643551523006,
+      "grad_norm": 1.0983004570007324,
+      "learning_rate": 4.967977244286225e-06,
+      "loss": 0.1512,
+      "step": 1472
+    },
+    {
+      "epoch": 0.23865845755022683,
+      "grad_norm": 1.1378045082092285,
+      "learning_rate": 4.9679074374986365e-06,
+      "loss": 0.1599,
+      "step": 1473
+    },
+    {
+      "epoch": 0.2388204795852236,
+      "grad_norm": 0.968702495098114,
+      "learning_rate": 4.967837555199054e-06,
+      "loss": 0.1294,
+      "step": 1474
+    },
+    {
+      "epoch": 0.23898250162022036,
+      "grad_norm": 1.1442790031433105,
+      "learning_rate": 4.967767597389613e-06,
+      "loss": 0.1666,
+      "step": 1475
+    },
+    {
+      "epoch": 0.2391445236552171,
+      "grad_norm": 1.2723668813705444,
+      "learning_rate": 4.967697564072457e-06,
+      "loss": 0.1767,
+      "step": 1476
+    },
+    {
+      "epoch": 0.23930654569021387,
+      "grad_norm": 1.1569876670837402,
+      "learning_rate": 4.967627455249726e-06,
+      "loss": 0.1688,
+      "step": 1477
+    },
+    {
+      "epoch": 0.23946856772521063,
+      "grad_norm": 1.2262758016586304,
+      "learning_rate": 4.9675572709235665e-06,
+      "loss": 0.1873,
+      "step": 1478
+    },
+    {
+      "epoch": 0.2396305897602074,
+      "grad_norm": 1.2375082969665527,
+      "learning_rate": 4.967487011096126e-06,
+      "loss": 0.1713,
+      "step": 1479
+    },
+    {
+      "epoch": 0.23979261179520414,
+      "grad_norm": 1.2592830657958984,
+      "learning_rate": 4.967416675769555e-06,
+      "loss": 0.1847,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2399546338302009,
+      "grad_norm": 1.3204195499420166,
+      "learning_rate": 4.9673462649460045e-06,
+      "loss": 0.1656,
+      "step": 1481
+    },
+    {
+      "epoch": 0.24011665586519768,
+      "grad_norm": 1.1137760877609253,
+      "learning_rate": 4.967275778627628e-06,
+      "loss": 0.1659,
+      "step": 1482
+    },
+    {
+      "epoch": 0.24027867790019442,
+      "grad_norm": 1.2828960418701172,
+      "learning_rate": 4.967205216816584e-06,
+      "loss": 0.1922,
+      "step": 1483
+    },
+    {
+      "epoch": 0.24044069993519118,
+      "grad_norm": 1.2552210092544556,
+      "learning_rate": 4.967134579515032e-06,
+      "loss": 0.1616,
+      "step": 1484
+    },
+    {
+      "epoch": 0.24060272197018795,
+      "grad_norm": 1.251631259918213,
+      "learning_rate": 4.96706386672513e-06,
+      "loss": 0.1807,
+      "step": 1485
+    },
+    {
+      "epoch": 0.24076474400518472,
+      "grad_norm": 1.195192575454712,
+      "learning_rate": 4.966993078449046e-06,
+      "loss": 0.1639,
+      "step": 1486
+    },
+    {
+      "epoch": 0.24092676604018146,
+      "grad_norm": 1.0965896844863892,
+      "learning_rate": 4.966922214688943e-06,
+      "loss": 0.1566,
+      "step": 1487
+    },
+    {
+      "epoch": 0.24108878807517822,
+      "grad_norm": 1.3898707628250122,
+      "learning_rate": 4.96685127544699e-06,
+      "loss": 0.1726,
+      "step": 1488
+    },
+    {
+      "epoch": 0.241250810110175,
+      "grad_norm": 1.1594412326812744,
+      "learning_rate": 4.966780260725357e-06,
+      "loss": 0.1645,
+      "step": 1489
+    },
+    {
+      "epoch": 0.24141283214517173,
+      "grad_norm": 1.1280404329299927,
+      "learning_rate": 4.966709170526219e-06,
+      "loss": 0.1496,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2415748541801685,
+      "grad_norm": 1.0832349061965942,
+      "learning_rate": 4.966638004851748e-06,
+      "loss": 0.1577,
+      "step": 1491
+    },
+    {
+      "epoch": 0.24173687621516526,
+      "grad_norm": 1.2211941480636597,
+      "learning_rate": 4.966566763704124e-06,
+      "loss": 0.1856,
+      "step": 1492
+    },
+    {
+      "epoch": 0.24189889825016203,
+      "grad_norm": 1.1475859880447388,
+      "learning_rate": 4.9664954470855265e-06,
+      "loss": 0.1602,
+      "step": 1493
+    },
+    {
+      "epoch": 0.24206092028515877,
+      "grad_norm": 1.326402187347412,
+      "learning_rate": 4.966424054998137e-06,
+      "loss": 0.2013,
+      "step": 1494
+    },
+    {
+      "epoch": 0.24222294232015554,
+      "grad_norm": 1.3355193138122559,
+      "learning_rate": 4.96635258744414e-06,
+      "loss": 0.1864,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2423849643551523,
+      "grad_norm": 1.110781192779541,
+      "learning_rate": 4.966281044425722e-06,
+      "loss": 0.1592,
+      "step": 1496
+    },
+    {
+      "epoch": 0.24254698639014907,
+      "grad_norm": 1.1372196674346924,
+      "learning_rate": 4.966209425945072e-06,
+      "loss": 0.1666,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2427090084251458,
+      "grad_norm": 1.1487059593200684,
+      "learning_rate": 4.9661377320043815e-06,
+      "loss": 0.167,
+      "step": 1498
+    },
+    {
+      "epoch": 0.24287103046014258,
+      "grad_norm": 1.1017019748687744,
+      "learning_rate": 4.966065962605845e-06,
+      "loss": 0.1661,
+      "step": 1499
+    },
+    {
+      "epoch": 0.24303305249513935,
+      "grad_norm": 1.0528111457824707,
+      "learning_rate": 4.965994117751658e-06,
+      "loss": 0.1547,
+      "step": 1500
+    },
+    {
+      "epoch": 0.24319507453013609,
+      "grad_norm": 1.1754587888717651,
+      "learning_rate": 4.965922197444017e-06,
+      "loss": 0.1734,
+      "step": 1501
+    },
+    {
+      "epoch": 0.24335709656513285,
+      "grad_norm": 1.20737886428833,
+      "learning_rate": 4.965850201685126e-06,
+      "loss": 0.1809,
+      "step": 1502
+    },
+    {
+      "epoch": 0.24351911860012962,
+      "grad_norm": 1.1585618257522583,
+      "learning_rate": 4.965778130477185e-06,
+      "loss": 0.1683,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2436811406351264,
+      "grad_norm": 1.1678732633590698,
+      "learning_rate": 4.965705983822401e-06,
+      "loss": 0.1567,
+      "step": 1504
+    },
+    {
+      "epoch": 0.24384316267012313,
+      "grad_norm": 1.127969741821289,
+      "learning_rate": 4.965633761722981e-06,
+      "loss": 0.1636,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2440051847051199,
+      "grad_norm": 1.1678990125656128,
+      "learning_rate": 4.965561464181134e-06,
+      "loss": 0.1616,
+      "step": 1506
+    },
+    {
+      "epoch": 0.24416720674011666,
+      "grad_norm": 1.166166067123413,
+      "learning_rate": 4.965489091199073e-06,
+      "loss": 0.1483,
+      "step": 1507
+    },
+    {
+      "epoch": 0.24432922877511343,
+      "grad_norm": 1.092760443687439,
+      "learning_rate": 4.965416642779012e-06,
+      "loss": 0.1541,
+      "step": 1508
+    },
+    {
+      "epoch": 0.24449125081011017,
+      "grad_norm": 1.230621099472046,
+      "learning_rate": 4.965344118923168e-06,
+      "loss": 0.1603,
+      "step": 1509
+    },
+    {
+      "epoch": 0.24465327284510693,
+      "grad_norm": 1.213835597038269,
+      "learning_rate": 4.96527151963376e-06,
+      "loss": 0.1719,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2448152948801037,
+      "grad_norm": 1.032114863395691,
+      "learning_rate": 4.965198844913009e-06,
+      "loss": 0.1459,
+      "step": 1511
+    },
+    {
+      "epoch": 0.24497731691510044,
+      "grad_norm": 1.3471325635910034,
+      "learning_rate": 4.9651260947631395e-06,
+      "loss": 0.1784,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2451393389500972,
+      "grad_norm": 1.0477313995361328,
+      "learning_rate": 4.965053269186378e-06,
+      "loss": 0.1542,
+      "step": 1513
+    },
+    {
+      "epoch": 0.24530136098509397,
+      "grad_norm": 1.0779989957809448,
+      "learning_rate": 4.9649803681849495e-06,
+      "loss": 0.1431,
+      "step": 1514
+    },
+    {
+      "epoch": 0.24546338302009074,
+      "grad_norm": 1.123201847076416,
+      "learning_rate": 4.964907391761088e-06,
+      "loss": 0.1537,
+      "step": 1515
+    },
+    {
+      "epoch": 0.24562540505508748,
+      "grad_norm": 1.2395498752593994,
+      "learning_rate": 4.9648343399170254e-06,
+      "loss": 0.1803,
+      "step": 1516
+    },
+    {
+      "epoch": 0.24578742709008425,
+      "grad_norm": 1.1589041948318481,
+      "learning_rate": 4.964761212654997e-06,
+      "loss": 0.1604,
+      "step": 1517
+    },
+    {
+      "epoch": 0.24594944912508102,
+      "grad_norm": 1.2080649137496948,
+      "learning_rate": 4.964688009977239e-06,
+      "loss": 0.1731,
+      "step": 1518
+    },
+    {
+      "epoch": 0.24611147116007778,
+      "grad_norm": 1.0178292989730835,
+      "learning_rate": 4.964614731885994e-06,
+      "loss": 0.1335,
+      "step": 1519
+    },
+    {
+      "epoch": 0.24627349319507452,
+      "grad_norm": 1.047829508781433,
+      "learning_rate": 4.9645413783835006e-06,
+      "loss": 0.1516,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2464355152300713,
+      "grad_norm": 1.1087828874588013,
+      "learning_rate": 4.964467949472006e-06,
+      "loss": 0.1596,
+      "step": 1521
+    },
+    {
+      "epoch": 0.24659753726506806,
+      "grad_norm": 1.0549379587173462,
+      "learning_rate": 4.964394445153756e-06,
+      "loss": 0.1287,
+      "step": 1522
+    },
+    {
+      "epoch": 0.2467595593000648,
+      "grad_norm": 1.2687568664550781,
+      "learning_rate": 4.964320865431001e-06,
+      "loss": 0.1743,
+      "step": 1523
+    },
+    {
+      "epoch": 0.24692158133506156,
+      "grad_norm": 1.0614700317382812,
+      "learning_rate": 4.964247210305989e-06,
+      "loss": 0.1444,
+      "step": 1524
+    },
+    {
+      "epoch": 0.24708360337005833,
+      "grad_norm": 1.253021478652954,
+      "learning_rate": 4.964173479780976e-06,
+      "loss": 0.15,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2472456254050551,
+      "grad_norm": 1.270143747329712,
+      "learning_rate": 4.964099673858219e-06,
+      "loss": 0.1641,
+      "step": 1526
+    },
+    {
+      "epoch": 0.24740764744005184,
+      "grad_norm": 1.1180580854415894,
+      "learning_rate": 4.964025792539974e-06,
+      "loss": 0.1566,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2475696694750486,
+      "grad_norm": 1.1324844360351562,
+      "learning_rate": 4.963951835828503e-06,
+      "loss": 0.1518,
+      "step": 1528
+    },
+    {
+      "epoch": 0.24773169151004537,
+      "grad_norm": 1.225598931312561,
+      "learning_rate": 4.963877803726068e-06,
+      "loss": 0.1681,
+      "step": 1529
+    },
+    {
+      "epoch": 0.24789371354504214,
+      "grad_norm": 1.3392179012298584,
+      "learning_rate": 4.963803696234935e-06,
+      "loss": 0.1815,
+      "step": 1530
+    },
+    {
+      "epoch": 0.24805573558003888,
+      "grad_norm": 1.1699576377868652,
+      "learning_rate": 4.963729513357372e-06,
+      "loss": 0.164,
+      "step": 1531
+    },
+    {
+      "epoch": 0.24821775761503564,
+      "grad_norm": 1.1221762895584106,
+      "learning_rate": 4.9636552550956465e-06,
+      "loss": 0.1688,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2483797796500324,
+      "grad_norm": 1.1934576034545898,
+      "learning_rate": 4.963580921452033e-06,
+      "loss": 0.1814,
+      "step": 1533
+    },
+    {
+      "epoch": 0.24854180168502915,
+      "grad_norm": 1.0234527587890625,
+      "learning_rate": 4.963506512428804e-06,
+      "loss": 0.1457,
+      "step": 1534
+    },
+    {
+      "epoch": 0.24870382372002592,
+      "grad_norm": 1.1598873138427734,
+      "learning_rate": 4.963432028028239e-06,
+      "loss": 0.1653,
+      "step": 1535
+    },
+    {
+      "epoch": 0.24886584575502269,
+      "grad_norm": 1.1791259050369263,
+      "learning_rate": 4.963357468252614e-06,
+      "loss": 0.1833,
+      "step": 1536
+    },
+    {
+      "epoch": 0.24902786779001945,
+      "grad_norm": 1.2526721954345703,
+      "learning_rate": 4.9632828331042124e-06,
+      "loss": 0.1743,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2491898898250162,
+      "grad_norm": 1.1005678176879883,
+      "learning_rate": 4.9632081225853165e-06,
+      "loss": 0.179,
+      "step": 1538
+    },
+    {
+      "epoch": 0.24935191186001296,
+      "grad_norm": 1.1455602645874023,
+      "learning_rate": 4.963133336698214e-06,
+      "loss": 0.17,
+      "step": 1539
+    },
+    {
+      "epoch": 0.24951393389500973,
+      "grad_norm": 1.3969601392745972,
+      "learning_rate": 4.9630584754451906e-06,
+      "loss": 0.1942,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2496759559300065,
+      "grad_norm": 1.1257600784301758,
+      "learning_rate": 4.962983538828539e-06,
+      "loss": 0.1532,
+      "step": 1541
+    },
+    {
+      "epoch": 0.24983797796500323,
+      "grad_norm": 1.151532769203186,
+      "learning_rate": 4.962908526850552e-06,
+      "loss": 0.1781,
+      "step": 1542
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.063228964805603,
+      "learning_rate": 4.962833439513524e-06,
+      "loss": 0.157,
+      "step": 1543
+    },
+    {
+      "epoch": 0.25016202203499677,
+      "grad_norm": 1.101438283920288,
+      "learning_rate": 4.962758276819752e-06,
+      "loss": 0.155,
+      "step": 1544
+    },
+    {
+      "epoch": 0.25032404406999353,
+      "grad_norm": 1.0311859846115112,
+      "learning_rate": 4.9626830387715365e-06,
+      "loss": 0.1524,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2504860661049903,
+      "grad_norm": 1.2852880954742432,
+      "learning_rate": 4.9626077253711805e-06,
+      "loss": 0.1775,
+      "step": 1546
+    },
+    {
+      "epoch": 0.250648088139987,
+      "grad_norm": 1.198421835899353,
+      "learning_rate": 4.962532336620987e-06,
+      "loss": 0.1595,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2508101101749838,
+      "grad_norm": 1.1567708253860474,
+      "learning_rate": 4.962456872523263e-06,
+      "loss": 0.1514,
+      "step": 1548
+    },
+    {
+      "epoch": 0.25097213220998055,
+      "grad_norm": 1.1411212682724,
+      "learning_rate": 4.9623813330803174e-06,
+      "loss": 0.1753,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2511341542449773,
+      "grad_norm": 1.034947395324707,
+      "learning_rate": 4.962305718294462e-06,
+      "loss": 0.1484,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2512961762799741,
+      "grad_norm": 1.2145742177963257,
+      "learning_rate": 4.962230028168011e-06,
+      "loss": 0.1791,
+      "step": 1551
+    },
+    {
+      "epoch": 0.25145819831497085,
+      "grad_norm": 1.173612356185913,
+      "learning_rate": 4.96215426270328e-06,
+      "loss": 0.1555,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2516202203499676,
+      "grad_norm": 1.1682580709457397,
+      "learning_rate": 4.9620784219025855e-06,
+      "loss": 0.1707,
+      "step": 1553
+    },
+    {
+      "epoch": 0.2517822423849643,
+      "grad_norm": 1.1669775247573853,
+      "learning_rate": 4.962002505768251e-06,
+      "loss": 0.172,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2519442644199611,
+      "grad_norm": 1.1306649446487427,
+      "learning_rate": 4.961926514302597e-06,
+      "loss": 0.1706,
+      "step": 1555
+    },
+    {
+      "epoch": 0.25210628645495786,
+      "grad_norm": 1.0562387704849243,
+      "learning_rate": 4.961850447507948e-06,
+      "loss": 0.1599,
+      "step": 1556
+    },
+    {
+      "epoch": 0.25226830848995463,
+      "grad_norm": 1.000566005706787,
+      "learning_rate": 4.961774305386634e-06,
+      "loss": 0.142,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2524303305249514,
+      "grad_norm": 1.1344536542892456,
+      "learning_rate": 4.961698087940984e-06,
+      "loss": 0.1568,
+      "step": 1558
+    },
+    {
+      "epoch": 0.25259235255994816,
+      "grad_norm": 1.1511046886444092,
+      "learning_rate": 4.961621795173329e-06,
+      "loss": 0.1594,
+      "step": 1559
+    },
+    {
+      "epoch": 0.25275437459494493,
+      "grad_norm": 1.1231741905212402,
+      "learning_rate": 4.961545427086006e-06,
+      "loss": 0.1734,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2529163966299417,
+      "grad_norm": 1.2015700340270996,
+      "learning_rate": 4.961468983681347e-06,
+      "loss": 0.1664,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2530784186649384,
+      "grad_norm": 1.1765875816345215,
+      "learning_rate": 4.961392464961695e-06,
+      "loss": 0.1562,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2532404406999352,
+      "grad_norm": 1.1523048877716064,
+      "learning_rate": 4.96131587092939e-06,
+      "loss": 0.1658,
+      "step": 1563
+    },
+    {
+      "epoch": 0.25340246273493194,
+      "grad_norm": 1.113807201385498,
+      "learning_rate": 4.961239201586776e-06,
+      "loss": 0.1567,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2535644847699287,
+      "grad_norm": 1.1596148014068604,
+      "learning_rate": 4.961162456936199e-06,
+      "loss": 0.1667,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2537265068049255,
+      "grad_norm": 1.4176827669143677,
+      "learning_rate": 4.961085636980005e-06,
+      "loss": 0.2077,
+      "step": 1566
+    },
+    {
+      "epoch": 0.25388852883992225,
+      "grad_norm": 1.307410717010498,
+      "learning_rate": 4.961008741720546e-06,
+      "loss": 0.1839,
+      "step": 1567
+    },
+    {
+      "epoch": 0.254050550874919,
+      "grad_norm": 1.387787938117981,
+      "learning_rate": 4.960931771160177e-06,
+      "loss": 0.1534,
+      "step": 1568
+    },
+    {
+      "epoch": 0.2542125729099157,
+      "grad_norm": 1.2094498872756958,
+      "learning_rate": 4.96085472530125e-06,
+      "loss": 0.1716,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2543745949449125,
+      "grad_norm": 1.3138521909713745,
+      "learning_rate": 4.960777604146124e-06,
+      "loss": 0.1826,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25453661697990926,
+      "grad_norm": 1.0649223327636719,
+      "learning_rate": 4.960700407697158e-06,
+      "loss": 0.1628,
+      "step": 1571
+    },
+    {
+      "epoch": 0.254698639014906,
+      "grad_norm": 1.0295525789260864,
+      "learning_rate": 4.9606231359567146e-06,
+      "loss": 0.1441,
+      "step": 1572
+    },
+    {
+      "epoch": 0.2548606610499028,
+      "grad_norm": 1.0389995574951172,
+      "learning_rate": 4.960545788927158e-06,
+      "loss": 0.1559,
+      "step": 1573
+    },
+    {
+      "epoch": 0.25502268308489956,
+      "grad_norm": 1.1343345642089844,
+      "learning_rate": 4.960468366610854e-06,
+      "loss": 0.1656,
+      "step": 1574
+    },
+    {
+      "epoch": 0.2551847051198963,
+      "grad_norm": 1.055141568183899,
+      "learning_rate": 4.960390869010172e-06,
+      "loss": 0.1614,
+      "step": 1575
+    },
+    {
+      "epoch": 0.25534672715489304,
+      "grad_norm": 1.1201953887939453,
+      "learning_rate": 4.960313296127485e-06,
+      "loss": 0.1514,
+      "step": 1576
+    },
+    {
+      "epoch": 0.2555087491898898,
+      "grad_norm": 1.041771650314331,
+      "learning_rate": 4.960235647965163e-06,
+      "loss": 0.1388,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2556707712248866,
+      "grad_norm": 1.19291353225708,
+      "learning_rate": 4.960157924525585e-06,
+      "loss": 0.1508,
+      "step": 1578
+    },
+    {
+      "epoch": 0.25583279325988334,
+      "grad_norm": 1.1116414070129395,
+      "learning_rate": 4.960080125811127e-06,
+      "loss": 0.1493,
+      "step": 1579
+    },
+    {
+      "epoch": 0.2559948152948801,
+      "grad_norm": 1.1531809568405151,
+      "learning_rate": 4.960002251824172e-06,
+      "loss": 0.1732,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2561568373298769,
+      "grad_norm": 1.1807605028152466,
+      "learning_rate": 4.9599243025671e-06,
+      "loss": 0.1722,
+      "step": 1581
+    },
+    {
+      "epoch": 0.25631885936487364,
+      "grad_norm": 1.049898624420166,
+      "learning_rate": 4.959846278042298e-06,
+      "loss": 0.1554,
+      "step": 1582
+    },
+    {
+      "epoch": 0.2564808813998704,
+      "grad_norm": 1.3067052364349365,
+      "learning_rate": 4.959768178252152e-06,
+      "loss": 0.194,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2566429034348671,
+      "grad_norm": 1.1992570161819458,
+      "learning_rate": 4.959690003199052e-06,
+      "loss": 0.1739,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2568049254698639,
+      "grad_norm": 1.368794560432434,
+      "learning_rate": 4.959611752885392e-06,
+      "loss": 0.174,
+      "step": 1585
+    },
+    {
+      "epoch": 0.25696694750486065,
+      "grad_norm": 1.3984448909759521,
+      "learning_rate": 4.959533427313562e-06,
+      "loss": 0.1909,
+      "step": 1586
+    },
+    {
+      "epoch": 0.2571289695398574,
+      "grad_norm": 1.0769413709640503,
+      "learning_rate": 4.959455026485963e-06,
+      "loss": 0.1573,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2572909915748542,
+      "grad_norm": 1.1236642599105835,
+      "learning_rate": 4.95937655040499e-06,
+      "loss": 0.1559,
+      "step": 1588
+    },
+    {
+      "epoch": 0.25745301360985096,
+      "grad_norm": 1.0431153774261475,
+      "learning_rate": 4.959297999073048e-06,
+      "loss": 0.1562,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2576150356448477,
+      "grad_norm": 1.1195615530014038,
+      "learning_rate": 4.959219372492539e-06,
+      "loss": 0.1543,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25777705767984443,
+      "grad_norm": 1.070062518119812,
+      "learning_rate": 4.959140670665867e-06,
+      "loss": 0.1548,
+      "step": 1591
+    },
+    {
+      "epoch": 0.2579390797148412,
+      "grad_norm": 1.1161715984344482,
+      "learning_rate": 4.9590618935954415e-06,
+      "loss": 0.1677,
+      "step": 1592
+    },
+    {
+      "epoch": 0.25810110174983797,
+      "grad_norm": 1.1165508031845093,
+      "learning_rate": 4.958983041283674e-06,
+      "loss": 0.1623,
+      "step": 1593
+    },
+    {
+      "epoch": 0.25826312378483474,
+      "grad_norm": 1.2309761047363281,
+      "learning_rate": 4.958904113732975e-06,
+      "loss": 0.1937,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2584251458198315,
+      "grad_norm": 1.0280898809432983,
+      "learning_rate": 4.958825110945761e-06,
+      "loss": 0.145,
+      "step": 1595
+    },
+    {
+      "epoch": 0.25858716785482827,
+      "grad_norm": 1.1085617542266846,
+      "learning_rate": 4.958746032924449e-06,
+      "loss": 0.1624,
+      "step": 1596
+    },
+    {
+      "epoch": 0.25874918988982504,
+      "grad_norm": 1.162490963935852,
+      "learning_rate": 4.958666879671458e-06,
+      "loss": 0.1569,
+      "step": 1597
+    },
+    {
+      "epoch": 0.25891121192482175,
+      "grad_norm": 1.267038106918335,
+      "learning_rate": 4.95858765118921e-06,
+      "loss": 0.1804,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2590732339598185,
+      "grad_norm": 1.0841212272644043,
+      "learning_rate": 4.9585083474801296e-06,
+      "loss": 0.1546,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2592352559948153,
+      "grad_norm": 1.171931266784668,
+      "learning_rate": 4.9584289685466444e-06,
+      "loss": 0.1688,
+      "step": 1600
+    },
+    {
+      "epoch": 0.25939727802981205,
+      "grad_norm": 1.20844304561615,
+      "learning_rate": 4.9583495143911804e-06,
+      "loss": 0.1752,
+      "step": 1601
+    },
+    {
+      "epoch": 0.2595593000648088,
+      "grad_norm": 1.2079849243164062,
+      "learning_rate": 4.958269985016172e-06,
+      "loss": 0.1573,
+      "step": 1602
+    },
+    {
+      "epoch": 0.2597213220998056,
+      "grad_norm": 1.101150631904602,
+      "learning_rate": 4.95819038042405e-06,
+      "loss": 0.1602,
+      "step": 1603
+    },
+    {
+      "epoch": 0.25988334413480235,
+      "grad_norm": 1.0884321928024292,
+      "learning_rate": 4.958110700617251e-06,
+      "loss": 0.1451,
+      "step": 1604
+    },
+    {
+      "epoch": 0.2600453661697991,
+      "grad_norm": 1.1064330339431763,
+      "learning_rate": 4.958030945598213e-06,
+      "loss": 0.1497,
+      "step": 1605
+    },
+    {
+      "epoch": 0.26020738820479583,
+      "grad_norm": 1.055196762084961,
+      "learning_rate": 4.957951115369378e-06,
+      "loss": 0.1493,
+      "step": 1606
+    },
+    {
+      "epoch": 0.2603694102397926,
+      "grad_norm": 1.1254777908325195,
+      "learning_rate": 4.957871209933185e-06,
+      "loss": 0.156,
+      "step": 1607
+    },
+    {
+      "epoch": 0.26053143227478937,
+      "grad_norm": 1.1287553310394287,
+      "learning_rate": 4.957791229292082e-06,
+      "loss": 0.1537,
+      "step": 1608
+    },
+    {
+      "epoch": 0.26069345430978613,
+      "grad_norm": 1.1099610328674316,
+      "learning_rate": 4.957711173448515e-06,
+      "loss": 0.1447,
+      "step": 1609
+    },
+    {
+      "epoch": 0.2608554763447829,
+      "grad_norm": 1.1488322019577026,
+      "learning_rate": 4.957631042404934e-06,
+      "loss": 0.1814,
+      "step": 1610
+    },
+    {
+      "epoch": 0.26101749837977967,
+      "grad_norm": 1.169036626815796,
+      "learning_rate": 4.957550836163789e-06,
+      "loss": 0.1601,
+      "step": 1611
+    },
+    {
+      "epoch": 0.26117952041477643,
+      "grad_norm": 1.2050291299819946,
+      "learning_rate": 4.957470554727536e-06,
+      "loss": 0.1633,
+      "step": 1612
+    },
+    {
+      "epoch": 0.26134154244977315,
+      "grad_norm": 1.1404060125350952,
+      "learning_rate": 4.9573901980986315e-06,
+      "loss": 0.1632,
+      "step": 1613
+    },
+    {
+      "epoch": 0.2615035644847699,
+      "grad_norm": 1.3234813213348389,
+      "learning_rate": 4.9573097662795344e-06,
+      "loss": 0.1938,
+      "step": 1614
+    },
+    {
+      "epoch": 0.2616655865197667,
+      "grad_norm": 1.1954011917114258,
+      "learning_rate": 4.957229259272705e-06,
+      "loss": 0.1611,
+      "step": 1615
+    },
+    {
+      "epoch": 0.26182760855476345,
+      "grad_norm": 1.1279500722885132,
+      "learning_rate": 4.957148677080605e-06,
+      "loss": 0.1471,
+      "step": 1616
+    },
+    {
+      "epoch": 0.2619896305897602,
+      "grad_norm": 1.2045884132385254,
+      "learning_rate": 4.957068019705703e-06,
+      "loss": 0.1625,
+      "step": 1617
+    },
+    {
+      "epoch": 0.262151652624757,
+      "grad_norm": 1.2116997241973877,
+      "learning_rate": 4.956987287150465e-06,
+      "loss": 0.1822,
+      "step": 1618
+    },
+    {
+      "epoch": 0.26231367465975375,
+      "grad_norm": 1.1517181396484375,
+      "learning_rate": 4.956906479417361e-06,
+      "loss": 0.1566,
+      "step": 1619
+    },
+    {
+      "epoch": 0.26247569669475046,
+      "grad_norm": 1.1915812492370605,
+      "learning_rate": 4.956825596508867e-06,
+      "loss": 0.1808,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2626377187297472,
+      "grad_norm": 1.0590485334396362,
+      "learning_rate": 4.9567446384274525e-06,
+      "loss": 0.153,
+      "step": 1621
+    },
+    {
+      "epoch": 0.262799740764744,
+      "grad_norm": 1.147382140159607,
+      "learning_rate": 4.956663605175599e-06,
+      "loss": 0.1621,
+      "step": 1622
+    },
+    {
+      "epoch": 0.26296176279974076,
+      "grad_norm": 1.2645186185836792,
+      "learning_rate": 4.956582496755783e-06,
+      "loss": 0.1875,
+      "step": 1623
+    },
+    {
+      "epoch": 0.26312378483473753,
+      "grad_norm": 1.0937161445617676,
+      "learning_rate": 4.956501313170487e-06,
+      "loss": 0.1671,
+      "step": 1624
+    },
+    {
+      "epoch": 0.2632858068697343,
+      "grad_norm": 1.2296099662780762,
+      "learning_rate": 4.956420054422197e-06,
+      "loss": 0.1887,
+      "step": 1625
+    },
+    {
+      "epoch": 0.26344782890473106,
+      "grad_norm": 1.030358910560608,
+      "learning_rate": 4.956338720513397e-06,
+      "loss": 0.1567,
+      "step": 1626
+    },
+    {
+      "epoch": 0.26360985093972783,
+      "grad_norm": 1.3956369161605835,
+      "learning_rate": 4.956257311446576e-06,
+      "loss": 0.1818,
+      "step": 1627
+    },
+    {
+      "epoch": 0.26377187297472454,
+      "grad_norm": 1.2601412534713745,
+      "learning_rate": 4.956175827224226e-06,
+      "loss": 0.1668,
+      "step": 1628
+    },
+    {
+      "epoch": 0.2639338950097213,
+      "grad_norm": 1.13845694065094,
+      "learning_rate": 4.956094267848839e-06,
+      "loss": 0.1688,
+      "step": 1629
+    },
+    {
+      "epoch": 0.2640959170447181,
+      "grad_norm": 1.1298080682754517,
+      "learning_rate": 4.956012633322912e-06,
+      "loss": 0.1605,
+      "step": 1630
+    },
+    {
+      "epoch": 0.26425793907971484,
+      "grad_norm": 1.1088753938674927,
+      "learning_rate": 4.955930923648941e-06,
+      "loss": 0.1386,
+      "step": 1631
+    },
+    {
+      "epoch": 0.2644199611147116,
+      "grad_norm": 1.1304455995559692,
+      "learning_rate": 4.955849138829428e-06,
+      "loss": 0.1563,
+      "step": 1632
+    },
+    {
+      "epoch": 0.2645819831497084,
+      "grad_norm": 1.1293634176254272,
+      "learning_rate": 4.955767278866872e-06,
+      "loss": 0.1387,
+      "step": 1633
+    },
+    {
+      "epoch": 0.26474400518470514,
+      "grad_norm": 1.0126081705093384,
+      "learning_rate": 4.955685343763782e-06,
+      "loss": 0.1385,
+      "step": 1634
+    },
+    {
+      "epoch": 0.26490602721970186,
+      "grad_norm": 1.0980356931686401,
+      "learning_rate": 4.955603333522663e-06,
+      "loss": 0.1575,
+      "step": 1635
+    },
+    {
+      "epoch": 0.2650680492546986,
+      "grad_norm": 1.0100072622299194,
+      "learning_rate": 4.9555212481460245e-06,
+      "loss": 0.1492,
+      "step": 1636
+    },
+    {
+      "epoch": 0.2652300712896954,
+      "grad_norm": 1.275562047958374,
+      "learning_rate": 4.955439087636378e-06,
+      "loss": 0.1708,
+      "step": 1637
+    },
+    {
+      "epoch": 0.26539209332469216,
+      "grad_norm": 1.1762036085128784,
+      "learning_rate": 4.955356851996236e-06,
+      "loss": 0.1598,
+      "step": 1638
+    },
+    {
+      "epoch": 0.2655541153596889,
+      "grad_norm": 1.199687123298645,
+      "learning_rate": 4.9552745412281175e-06,
+      "loss": 0.1696,
+      "step": 1639
+    },
+    {
+      "epoch": 0.2657161373946857,
+      "grad_norm": 1.060361385345459,
+      "learning_rate": 4.955192155334539e-06,
+      "loss": 0.1561,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26587815942968246,
+      "grad_norm": 1.3154797554016113,
+      "learning_rate": 4.955109694318024e-06,
+      "loss": 0.1891,
+      "step": 1641
+    },
+    {
+      "epoch": 0.26604018146467917,
+      "grad_norm": 1.0718739032745361,
+      "learning_rate": 4.955027158181092e-06,
+      "loss": 0.1487,
+      "step": 1642
+    },
+    {
+      "epoch": 0.26620220349967594,
+      "grad_norm": 1.0659395456314087,
+      "learning_rate": 4.95494454692627e-06,
+      "loss": 0.15,
+      "step": 1643
+    },
+    {
+      "epoch": 0.2663642255346727,
+      "grad_norm": 1.4435391426086426,
+      "learning_rate": 4.9548618605560855e-06,
+      "loss": 0.1911,
+      "step": 1644
+    },
+    {
+      "epoch": 0.26652624756966947,
+      "grad_norm": 1.1285501718521118,
+      "learning_rate": 4.954779099073069e-06,
+      "loss": 0.1573,
+      "step": 1645
+    },
+    {
+      "epoch": 0.26668826960466624,
+      "grad_norm": 1.2026793956756592,
+      "learning_rate": 4.954696262479753e-06,
+      "loss": 0.1823,
+      "step": 1646
+    },
+    {
+      "epoch": 0.266850291639663,
+      "grad_norm": 1.1052802801132202,
+      "learning_rate": 4.954613350778671e-06,
+      "loss": 0.1497,
+      "step": 1647
+    },
+    {
+      "epoch": 0.2670123136746598,
+      "grad_norm": 1.0569524765014648,
+      "learning_rate": 4.954530363972361e-06,
+      "loss": 0.1615,
+      "step": 1648
+    },
+    {
+      "epoch": 0.26717433570965654,
+      "grad_norm": 1.3057336807250977,
+      "learning_rate": 4.954447302063362e-06,
+      "loss": 0.1856,
+      "step": 1649
+    },
+    {
+      "epoch": 0.26733635774465325,
+      "grad_norm": 1.0749009847640991,
+      "learning_rate": 4.954364165054214e-06,
+      "loss": 0.169,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26749837977965,
+      "grad_norm": 1.1229180097579956,
+      "learning_rate": 4.954280952947463e-06,
+      "loss": 0.1572,
+      "step": 1651
+    },
+    {
+      "epoch": 0.2676604018146468,
+      "grad_norm": 1.134561538696289,
+      "learning_rate": 4.9541976657456535e-06,
+      "loss": 0.1618,
+      "step": 1652
+    },
+    {
+      "epoch": 0.26782242384964355,
+      "grad_norm": 1.229466199874878,
+      "learning_rate": 4.954114303451335e-06,
+      "loss": 0.1516,
+      "step": 1653
+    },
+    {
+      "epoch": 0.2679844458846403,
+      "grad_norm": 1.3184914588928223,
+      "learning_rate": 4.954030866067057e-06,
+      "loss": 0.1565,
+      "step": 1654
+    },
+    {
+      "epoch": 0.2681464679196371,
+      "grad_norm": 1.1177244186401367,
+      "learning_rate": 4.953947353595374e-06,
+      "loss": 0.1492,
+      "step": 1655
+    },
+    {
+      "epoch": 0.26830848995463386,
+      "grad_norm": 1.136872410774231,
+      "learning_rate": 4.95386376603884e-06,
+      "loss": 0.1586,
+      "step": 1656
+    },
+    {
+      "epoch": 0.26847051198963057,
+      "grad_norm": 1.8951784372329712,
+      "learning_rate": 4.953780103400012e-06,
+      "loss": 0.1676,
+      "step": 1657
+    },
+    {
+      "epoch": 0.26863253402462733,
+      "grad_norm": 1.2328137159347534,
+      "learning_rate": 4.953696365681452e-06,
+      "loss": 0.1778,
+      "step": 1658
+    },
+    {
+      "epoch": 0.2687945560596241,
+      "grad_norm": 1.0323786735534668,
+      "learning_rate": 4.953612552885721e-06,
+      "loss": 0.1374,
+      "step": 1659
+    },
+    {
+      "epoch": 0.26895657809462087,
+      "grad_norm": 1.2623248100280762,
+      "learning_rate": 4.953528665015383e-06,
+      "loss": 0.1787,
+      "step": 1660
+    },
+    {
+      "epoch": 0.26911860012961764,
+      "grad_norm": 1.1066336631774902,
+      "learning_rate": 4.953444702073006e-06,
+      "loss": 0.1546,
+      "step": 1661
+    },
+    {
+      "epoch": 0.2692806221646144,
+      "grad_norm": 1.1255744695663452,
+      "learning_rate": 4.953360664061159e-06,
+      "loss": 0.1574,
+      "step": 1662
+    },
+    {
+      "epoch": 0.26944264419961117,
+      "grad_norm": 1.143039584159851,
+      "learning_rate": 4.953276550982411e-06,
+      "loss": 0.1585,
+      "step": 1663
+    },
+    {
+      "epoch": 0.2696046662346079,
+      "grad_norm": 1.1908190250396729,
+      "learning_rate": 4.953192362839338e-06,
+      "loss": 0.1744,
+      "step": 1664
+    },
+    {
+      "epoch": 0.26976668826960465,
+      "grad_norm": 1.1476013660430908,
+      "learning_rate": 4.953108099634516e-06,
+      "loss": 0.1698,
+      "step": 1665
+    },
+    {
+      "epoch": 0.2699287103046014,
+      "grad_norm": 1.0105767250061035,
+      "learning_rate": 4.953023761370521e-06,
+      "loss": 0.1454,
+      "step": 1666
+    },
+    {
+      "epoch": 0.2700907323395982,
+      "grad_norm": 0.9987475872039795,
+      "learning_rate": 4.9529393480499365e-06,
+      "loss": 0.1345,
+      "step": 1667
+    },
+    {
+      "epoch": 0.27025275437459495,
+      "grad_norm": 1.234141230583191,
+      "learning_rate": 4.952854859675343e-06,
+      "loss": 0.1642,
+      "step": 1668
+    },
+    {
+      "epoch": 0.2704147764095917,
+      "grad_norm": 1.2194494009017944,
+      "learning_rate": 4.952770296249328e-06,
+      "loss": 0.1766,
+      "step": 1669
+    },
+    {
+      "epoch": 0.2705767984445885,
+      "grad_norm": 1.22120201587677,
+      "learning_rate": 4.952685657774476e-06,
+      "loss": 0.1875,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2707388204795852,
+      "grad_norm": 1.0285940170288086,
+      "learning_rate": 4.952600944253379e-06,
+      "loss": 0.1544,
+      "step": 1671
+    },
+    {
+      "epoch": 0.27090084251458196,
+      "grad_norm": 1.1357102394104004,
+      "learning_rate": 4.952516155688628e-06,
+      "loss": 0.1668,
+      "step": 1672
+    },
+    {
+      "epoch": 0.27106286454957873,
+      "grad_norm": 1.0606294870376587,
+      "learning_rate": 4.952431292082818e-06,
+      "loss": 0.1639,
+      "step": 1673
+    },
+    {
+      "epoch": 0.2712248865845755,
+      "grad_norm": 1.2481400966644287,
+      "learning_rate": 4.9523463534385444e-06,
+      "loss": 0.1709,
+      "step": 1674
+    },
+    {
+      "epoch": 0.27138690861957226,
+      "grad_norm": 1.1794558763504028,
+      "learning_rate": 4.9522613397584075e-06,
+      "loss": 0.1749,
+      "step": 1675
+    },
+    {
+      "epoch": 0.27154893065456903,
+      "grad_norm": 1.0960800647735596,
+      "learning_rate": 4.952176251045008e-06,
+      "loss": 0.1427,
+      "step": 1676
+    },
+    {
+      "epoch": 0.2717109526895658,
+      "grad_norm": 1.2068989276885986,
+      "learning_rate": 4.95209108730095e-06,
+      "loss": 0.1591,
+      "step": 1677
+    },
+    {
+      "epoch": 0.27187297472456257,
+      "grad_norm": 1.1335726976394653,
+      "learning_rate": 4.952005848528838e-06,
+      "loss": 0.1549,
+      "step": 1678
+    },
+    {
+      "epoch": 0.2720349967595593,
+      "grad_norm": 1.113123893737793,
+      "learning_rate": 4.95192053473128e-06,
+      "loss": 0.1502,
+      "step": 1679
+    },
+    {
+      "epoch": 0.27219701879455604,
+      "grad_norm": 1.2119156122207642,
+      "learning_rate": 4.951835145910888e-06,
+      "loss": 0.1636,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2723590408295528,
+      "grad_norm": 1.0948169231414795,
+      "learning_rate": 4.951749682070274e-06,
+      "loss": 0.1543,
+      "step": 1681
+    },
+    {
+      "epoch": 0.2725210628645496,
+      "grad_norm": 1.4161688089370728,
+      "learning_rate": 4.951664143212053e-06,
+      "loss": 0.1859,
+      "step": 1682
+    },
+    {
+      "epoch": 0.27268308489954635,
+      "grad_norm": 1.244466781616211,
+      "learning_rate": 4.951578529338842e-06,
+      "loss": 0.1875,
+      "step": 1683
+    },
+    {
+      "epoch": 0.2728451069345431,
+      "grad_norm": 2.2999372482299805,
+      "learning_rate": 4.95149284045326e-06,
+      "loss": 0.1494,
+      "step": 1684
+    },
+    {
+      "epoch": 0.2730071289695399,
+      "grad_norm": 1.1153371334075928,
+      "learning_rate": 4.95140707655793e-06,
+      "loss": 0.1467,
+      "step": 1685
+    },
+    {
+      "epoch": 0.2731691510045366,
+      "grad_norm": 1.0886139869689941,
+      "learning_rate": 4.951321237655477e-06,
+      "loss": 0.148,
+      "step": 1686
+    },
+    {
+      "epoch": 0.27333117303953336,
+      "grad_norm": 1.16849684715271,
+      "learning_rate": 4.951235323748524e-06,
+      "loss": 0.1876,
+      "step": 1687
+    },
+    {
+      "epoch": 0.2734931950745301,
+      "grad_norm": 1.0493582487106323,
+      "learning_rate": 4.951149334839703e-06,
+      "loss": 0.1584,
+      "step": 1688
+    },
+    {
+      "epoch": 0.2736552171095269,
+      "grad_norm": 1.146984577178955,
+      "learning_rate": 4.951063270931644e-06,
+      "loss": 0.1778,
+      "step": 1689
+    },
+    {
+      "epoch": 0.27381723914452366,
+      "grad_norm": 1.1503607034683228,
+      "learning_rate": 4.950977132026981e-06,
+      "loss": 0.1555,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27397926117952043,
+      "grad_norm": 1.2447178363800049,
+      "learning_rate": 4.950890918128348e-06,
+      "loss": 0.1876,
+      "step": 1691
+    },
+    {
+      "epoch": 0.2741412832145172,
+      "grad_norm": 1.197851300239563,
+      "learning_rate": 4.9508046292383846e-06,
+      "loss": 0.1763,
+      "step": 1692
+    },
+    {
+      "epoch": 0.2743033052495139,
+      "grad_norm": 1.1843172311782837,
+      "learning_rate": 4.950718265359729e-06,
+      "loss": 0.166,
+      "step": 1693
+    },
+    {
+      "epoch": 0.2744653272845107,
+      "grad_norm": 1.0466846227645874,
+      "learning_rate": 4.950631826495027e-06,
+      "loss": 0.1472,
+      "step": 1694
+    },
+    {
+      "epoch": 0.27462734931950744,
+      "grad_norm": 1.1562830209732056,
+      "learning_rate": 4.950545312646921e-06,
+      "loss": 0.1596,
+      "step": 1695
+    },
+    {
+      "epoch": 0.2747893713545042,
+      "grad_norm": 1.1788415908813477,
+      "learning_rate": 4.950458723818058e-06,
+      "loss": 0.1586,
+      "step": 1696
+    },
+    {
+      "epoch": 0.274951393389501,
+      "grad_norm": 1.238731861114502,
+      "learning_rate": 4.9503720600110884e-06,
+      "loss": 0.1626,
+      "step": 1697
+    },
+    {
+      "epoch": 0.27511341542449774,
+      "grad_norm": 1.1631656885147095,
+      "learning_rate": 4.950285321228664e-06,
+      "loss": 0.1508,
+      "step": 1698
+    },
+    {
+      "epoch": 0.2752754374594945,
+      "grad_norm": 1.2769675254821777,
+      "learning_rate": 4.950198507473438e-06,
+      "loss": 0.1555,
+      "step": 1699
+    },
+    {
+      "epoch": 0.2754374594944913,
+      "grad_norm": 1.2280595302581787,
+      "learning_rate": 4.950111618748067e-06,
+      "loss": 0.1416,
+      "step": 1700
+    },
+    {
+      "epoch": 0.275599481529488,
+      "grad_norm": 1.3224934339523315,
+      "learning_rate": 4.95002465505521e-06,
+      "loss": 0.1673,
+      "step": 1701
+    },
+    {
+      "epoch": 0.27576150356448476,
+      "grad_norm": 1.2435288429260254,
+      "learning_rate": 4.949937616397527e-06,
+      "loss": 0.1776,
+      "step": 1702
+    },
+    {
+      "epoch": 0.2759235255994815,
+      "grad_norm": 1.2028788328170776,
+      "learning_rate": 4.949850502777681e-06,
+      "loss": 0.1704,
+      "step": 1703
+    },
+    {
+      "epoch": 0.2760855476344783,
+      "grad_norm": 1.0093039274215698,
+      "learning_rate": 4.949763314198339e-06,
+      "loss": 0.1424,
+      "step": 1704
+    },
+    {
+      "epoch": 0.27624756966947506,
+      "grad_norm": 1.1535528898239136,
+      "learning_rate": 4.949676050662169e-06,
+      "loss": 0.1564,
+      "step": 1705
+    },
+    {
+      "epoch": 0.2764095917044718,
+      "grad_norm": 1.1377063989639282,
+      "learning_rate": 4.949588712171838e-06,
+      "loss": 0.1687,
+      "step": 1706
+    },
+    {
+      "epoch": 0.2765716137394686,
+      "grad_norm": 1.1593130826950073,
+      "learning_rate": 4.949501298730021e-06,
+      "loss": 0.1611,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2767336357744653,
+      "grad_norm": 1.2367503643035889,
+      "learning_rate": 4.949413810339392e-06,
+      "loss": 0.1803,
+      "step": 1708
+    },
+    {
+      "epoch": 0.27689565780946207,
+      "grad_norm": 1.1531428098678589,
+      "learning_rate": 4.9493262470026286e-06,
+      "loss": 0.183,
+      "step": 1709
+    },
+    {
+      "epoch": 0.27705767984445884,
+      "grad_norm": 1.1906907558441162,
+      "learning_rate": 4.949238608722408e-06,
+      "loss": 0.1486,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2772197018794556,
+      "grad_norm": 1.1683154106140137,
+      "learning_rate": 4.949150895501414e-06,
+      "loss": 0.1657,
+      "step": 1711
+    },
+    {
+      "epoch": 0.27738172391445237,
+      "grad_norm": 1.2222694158554077,
+      "learning_rate": 4.949063107342329e-06,
+      "loss": 0.1983,
+      "step": 1712
+    },
+    {
+      "epoch": 0.27754374594944914,
+      "grad_norm": 1.0516256093978882,
+      "learning_rate": 4.948975244247839e-06,
+      "loss": 0.1497,
+      "step": 1713
+    },
+    {
+      "epoch": 0.2777057679844459,
+      "grad_norm": 1.630317211151123,
+      "learning_rate": 4.948887306220634e-06,
+      "loss": 0.2039,
+      "step": 1714
+    },
+    {
+      "epoch": 0.2778677900194426,
+      "grad_norm": 1.3137788772583008,
+      "learning_rate": 4.948799293263403e-06,
+      "loss": 0.1633,
+      "step": 1715
+    },
+    {
+      "epoch": 0.2780298120544394,
+      "grad_norm": 1.0964162349700928,
+      "learning_rate": 4.94871120537884e-06,
+      "loss": 0.1566,
+      "step": 1716
+    },
+    {
+      "epoch": 0.27819183408943615,
+      "grad_norm": 1.437134027481079,
+      "learning_rate": 4.948623042569639e-06,
+      "loss": 0.1775,
+      "step": 1717
+    },
+    {
+      "epoch": 0.2783538561244329,
+      "grad_norm": 1.0482769012451172,
+      "learning_rate": 4.9485348048385e-06,
+      "loss": 0.1494,
+      "step": 1718
+    },
+    {
+      "epoch": 0.2785158781594297,
+      "grad_norm": 1.0254579782485962,
+      "learning_rate": 4.94844649218812e-06,
+      "loss": 0.1427,
+      "step": 1719
+    },
+    {
+      "epoch": 0.27867790019442645,
+      "grad_norm": 1.1733806133270264,
+      "learning_rate": 4.9483581046212025e-06,
+      "loss": 0.1733,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2788399222294232,
+      "grad_norm": 1.1621792316436768,
+      "learning_rate": 4.948269642140453e-06,
+      "loss": 0.1532,
+      "step": 1721
+    },
+    {
+      "epoch": 0.27900194426442,
+      "grad_norm": 1.2027387619018555,
+      "learning_rate": 4.948181104748576e-06,
+      "loss": 0.1711,
+      "step": 1722
+    },
+    {
+      "epoch": 0.2791639662994167,
+      "grad_norm": 1.3323115110397339,
+      "learning_rate": 4.9480924924482824e-06,
+      "loss": 0.185,
+      "step": 1723
+    },
+    {
+      "epoch": 0.27932598833441347,
+      "grad_norm": 1.1466425657272339,
+      "learning_rate": 4.948003805242282e-06,
+      "loss": 0.1606,
+      "step": 1724
+    },
+    {
+      "epoch": 0.27948801036941023,
+      "grad_norm": 1.1077498197555542,
+      "learning_rate": 4.94791504313329e-06,
+      "loss": 0.1582,
+      "step": 1725
+    },
+    {
+      "epoch": 0.279650032404407,
+      "grad_norm": 1.1000559329986572,
+      "learning_rate": 4.9478262061240216e-06,
+      "loss": 0.1451,
+      "step": 1726
+    },
+    {
+      "epoch": 0.27981205443940377,
+      "grad_norm": 1.2915170192718506,
+      "learning_rate": 4.9477372942171945e-06,
+      "loss": 0.1759,
+      "step": 1727
+    },
+    {
+      "epoch": 0.27997407647440054,
+      "grad_norm": 1.1661245822906494,
+      "learning_rate": 4.947648307415529e-06,
+      "loss": 0.1615,
+      "step": 1728
+    },
+    {
+      "epoch": 0.2801360985093973,
+      "grad_norm": 1.1373041868209839,
+      "learning_rate": 4.947559245721749e-06,
+      "loss": 0.1458,
+      "step": 1729
+    },
+    {
+      "epoch": 0.280298120544394,
+      "grad_norm": 1.0064786672592163,
+      "learning_rate": 4.947470109138579e-06,
+      "loss": 0.1409,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2804601425793908,
+      "grad_norm": 0.9815242290496826,
+      "learning_rate": 4.947380897668747e-06,
+      "loss": 0.1361,
+      "step": 1731
+    },
+    {
+      "epoch": 0.28062216461438755,
+      "grad_norm": 0.9623909592628479,
+      "learning_rate": 4.947291611314981e-06,
+      "loss": 0.1412,
+      "step": 1732
+    },
+    {
+      "epoch": 0.2807841866493843,
+      "grad_norm": 1.241988182067871,
+      "learning_rate": 4.947202250080015e-06,
+      "loss": 0.1754,
+      "step": 1733
+    },
+    {
+      "epoch": 0.2809462086843811,
+      "grad_norm": 1.213842749595642,
+      "learning_rate": 4.9471128139665826e-06,
+      "loss": 0.1757,
+      "step": 1734
+    },
+    {
+      "epoch": 0.28110823071937785,
+      "grad_norm": 1.181797742843628,
+      "learning_rate": 4.9470233029774195e-06,
+      "loss": 0.1591,
+      "step": 1735
+    },
+    {
+      "epoch": 0.2812702527543746,
+      "grad_norm": 1.0789086818695068,
+      "learning_rate": 4.9469337171152645e-06,
+      "loss": 0.1524,
+      "step": 1736
+    },
+    {
+      "epoch": 0.28143227478937133,
+      "grad_norm": 1.2921925783157349,
+      "learning_rate": 4.94684405638286e-06,
+      "loss": 0.1836,
+      "step": 1737
+    },
+    {
+      "epoch": 0.2815942968243681,
+      "grad_norm": 1.1767724752426147,
+      "learning_rate": 4.946754320782948e-06,
+      "loss": 0.1598,
+      "step": 1738
+    },
+    {
+      "epoch": 0.28175631885936486,
+      "grad_norm": 1.1282929182052612,
+      "learning_rate": 4.946664510318275e-06,
+      "loss": 0.1573,
+      "step": 1739
+    },
+    {
+      "epoch": 0.28191834089436163,
+      "grad_norm": 1.316778302192688,
+      "learning_rate": 4.946574624991589e-06,
+      "loss": 0.1816,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2820803629293584,
+      "grad_norm": 1.234459638595581,
+      "learning_rate": 4.9464846648056396e-06,
+      "loss": 0.1854,
+      "step": 1741
+    },
+    {
+      "epoch": 0.28224238496435516,
+      "grad_norm": 1.1687872409820557,
+      "learning_rate": 4.946394629763181e-06,
+      "loss": 0.1638,
+      "step": 1742
+    },
+    {
+      "epoch": 0.28240440699935193,
+      "grad_norm": 1.0062087774276733,
+      "learning_rate": 4.946304519866966e-06,
+      "loss": 0.1454,
+      "step": 1743
+    },
+    {
+      "epoch": 0.2825664290343487,
+      "grad_norm": 1.048824429512024,
+      "learning_rate": 4.946214335119752e-06,
+      "loss": 0.1415,
+      "step": 1744
+    },
+    {
+      "epoch": 0.2827284510693454,
+      "grad_norm": 1.0653513669967651,
+      "learning_rate": 4.9461240755243e-06,
+      "loss": 0.1454,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2828904731043422,
+      "grad_norm": 1.1784061193466187,
+      "learning_rate": 4.94603374108337e-06,
+      "loss": 0.1578,
+      "step": 1746
+    },
+    {
+      "epoch": 0.28305249513933894,
+      "grad_norm": 1.1248995065689087,
+      "learning_rate": 4.945943331799728e-06,
+      "loss": 0.1534,
+      "step": 1747
+    },
+    {
+      "epoch": 0.2832145171743357,
+      "grad_norm": 1.266843557357788,
+      "learning_rate": 4.945852847676138e-06,
+      "loss": 0.1917,
+      "step": 1748
+    },
+    {
+      "epoch": 0.2833765392093325,
+      "grad_norm": 1.1719672679901123,
+      "learning_rate": 4.945762288715371e-06,
+      "loss": 0.1702,
+      "step": 1749
+    },
+    {
+      "epoch": 0.28353856124432925,
+      "grad_norm": 1.1670610904693604,
+      "learning_rate": 4.945671654920195e-06,
+      "loss": 0.1615,
+      "step": 1750
+    },
+    {
+      "epoch": 0.283700583279326,
+      "grad_norm": 1.1604700088500977,
+      "learning_rate": 4.945580946293386e-06,
+      "loss": 0.1599,
+      "step": 1751
+    },
+    {
+      "epoch": 0.2838626053143227,
+      "grad_norm": 1.085565447807312,
+      "learning_rate": 4.945490162837718e-06,
+      "loss": 0.1558,
+      "step": 1752
+    },
+    {
+      "epoch": 0.2840246273493195,
+      "grad_norm": 1.1299009323120117,
+      "learning_rate": 4.945399304555968e-06,
+      "loss": 0.1584,
+      "step": 1753
+    },
+    {
+      "epoch": 0.28418664938431626,
+      "grad_norm": 1.1173944473266602,
+      "learning_rate": 4.945308371450919e-06,
+      "loss": 0.1524,
+      "step": 1754
+    },
+    {
+      "epoch": 0.284348671419313,
+      "grad_norm": 0.8910510540008545,
+      "learning_rate": 4.945217363525349e-06,
+      "loss": 0.1184,
+      "step": 1755
+    },
+    {
+      "epoch": 0.2845106934543098,
+      "grad_norm": 1.2654694318771362,
+      "learning_rate": 4.945126280782047e-06,
+      "loss": 0.2026,
+      "step": 1756
+    },
+    {
+      "epoch": 0.28467271548930656,
+      "grad_norm": 1.1252206563949585,
+      "learning_rate": 4.945035123223797e-06,
+      "loss": 0.1871,
+      "step": 1757
+    },
+    {
+      "epoch": 0.2848347375243033,
+      "grad_norm": 1.245661973953247,
+      "learning_rate": 4.944943890853389e-06,
+      "loss": 0.1596,
+      "step": 1758
+    },
+    {
+      "epoch": 0.28499675955930004,
+      "grad_norm": 1.2262191772460938,
+      "learning_rate": 4.944852583673615e-06,
+      "loss": 0.1797,
+      "step": 1759
+    },
+    {
+      "epoch": 0.2851587815942968,
+      "grad_norm": 1.0503751039505005,
+      "learning_rate": 4.944761201687268e-06,
+      "loss": 0.1437,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2853208036292936,
+      "grad_norm": 1.460869550704956,
+      "learning_rate": 4.944669744897144e-06,
+      "loss": 0.1576,
+      "step": 1761
+    },
+    {
+      "epoch": 0.28548282566429034,
+      "grad_norm": 1.1839721202850342,
+      "learning_rate": 4.944578213306043e-06,
+      "loss": 0.1463,
+      "step": 1762
+    },
+    {
+      "epoch": 0.2856448476992871,
+      "grad_norm": 1.1517478227615356,
+      "learning_rate": 4.944486606916764e-06,
+      "loss": 0.1873,
+      "step": 1763
+    },
+    {
+      "epoch": 0.2858068697342839,
+      "grad_norm": 1.1901956796646118,
+      "learning_rate": 4.94439492573211e-06,
+      "loss": 0.159,
+      "step": 1764
+    },
+    {
+      "epoch": 0.28596889176928064,
+      "grad_norm": 1.0697051286697388,
+      "learning_rate": 4.944303169754887e-06,
+      "loss": 0.1668,
+      "step": 1765
+    },
+    {
+      "epoch": 0.2861309138042774,
+      "grad_norm": 1.2297983169555664,
+      "learning_rate": 4.944211338987901e-06,
+      "loss": 0.1695,
+      "step": 1766
+    },
+    {
+      "epoch": 0.2862929358392741,
+      "grad_norm": 1.160058617591858,
+      "learning_rate": 4.944119433433964e-06,
+      "loss": 0.1674,
+      "step": 1767
+    },
+    {
+      "epoch": 0.2864549578742709,
+      "grad_norm": 1.1287837028503418,
+      "learning_rate": 4.944027453095887e-06,
+      "loss": 0.1578,
+      "step": 1768
+    },
+    {
+      "epoch": 0.28661697990926766,
+      "grad_norm": 1.0510386228561401,
+      "learning_rate": 4.943935397976484e-06,
+      "loss": 0.154,
+      "step": 1769
+    },
+    {
+      "epoch": 0.2867790019442644,
+      "grad_norm": 1.118772268295288,
+      "learning_rate": 4.943843268078572e-06,
+      "loss": 0.151,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2869410239792612,
+      "grad_norm": 1.0504902601242065,
+      "learning_rate": 4.94375106340497e-06,
+      "loss": 0.1569,
+      "step": 1771
+    },
+    {
+      "epoch": 0.28710304601425796,
+      "grad_norm": 1.1878551244735718,
+      "learning_rate": 4.9436587839585e-06,
+      "loss": 0.1696,
+      "step": 1772
+    },
+    {
+      "epoch": 0.2872650680492547,
+      "grad_norm": 1.1703705787658691,
+      "learning_rate": 4.9435664297419836e-06,
+      "loss": 0.1505,
+      "step": 1773
+    },
+    {
+      "epoch": 0.28742709008425144,
+      "grad_norm": 1.1289130449295044,
+      "learning_rate": 4.9434740007582485e-06,
+      "loss": 0.1568,
+      "step": 1774
+    },
+    {
+      "epoch": 0.2875891121192482,
+      "grad_norm": 1.1618683338165283,
+      "learning_rate": 4.943381497010122e-06,
+      "loss": 0.1708,
+      "step": 1775
+    },
+    {
+      "epoch": 0.28775113415424497,
+      "grad_norm": 1.0598828792572021,
+      "learning_rate": 4.943288918500434e-06,
+      "loss": 0.1608,
+      "step": 1776
+    },
+    {
+      "epoch": 0.28791315618924174,
+      "grad_norm": 1.1021727323532104,
+      "learning_rate": 4.943196265232018e-06,
+      "loss": 0.1598,
+      "step": 1777
+    },
+    {
+      "epoch": 0.2880751782242385,
+      "grad_norm": 1.0794800519943237,
+      "learning_rate": 4.94310353720771e-06,
+      "loss": 0.1445,
+      "step": 1778
+    },
+    {
+      "epoch": 0.28823720025923527,
+      "grad_norm": 1.0673820972442627,
+      "learning_rate": 4.9430107344303445e-06,
+      "loss": 0.1494,
+      "step": 1779
+    },
+    {
+      "epoch": 0.28839922229423204,
+      "grad_norm": 1.0401991605758667,
+      "learning_rate": 4.942917856902763e-06,
+      "loss": 0.1386,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28856124432922875,
+      "grad_norm": 1.2189371585845947,
+      "learning_rate": 4.9428249046278065e-06,
+      "loss": 0.1655,
+      "step": 1781
+    },
+    {
+      "epoch": 0.2887232663642255,
+      "grad_norm": 1.2600358724594116,
+      "learning_rate": 4.942731877608319e-06,
+      "loss": 0.1771,
+      "step": 1782
+    },
+    {
+      "epoch": 0.2888852883992223,
+      "grad_norm": 1.2140158414840698,
+      "learning_rate": 4.942638775847149e-06,
+      "loss": 0.1741,
+      "step": 1783
+    },
+    {
+      "epoch": 0.28904731043421905,
+      "grad_norm": 1.0172653198242188,
+      "learning_rate": 4.942545599347142e-06,
+      "loss": 0.1417,
+      "step": 1784
+    },
+    {
+      "epoch": 0.2892093324692158,
+      "grad_norm": 1.2738704681396484,
+      "learning_rate": 4.942452348111151e-06,
+      "loss": 0.1731,
+      "step": 1785
+    },
+    {
+      "epoch": 0.2893713545042126,
+      "grad_norm": 1.1915194988250732,
+      "learning_rate": 4.942359022142028e-06,
+      "loss": 0.1663,
+      "step": 1786
+    },
+    {
+      "epoch": 0.28953337653920935,
+      "grad_norm": 1.04640531539917,
+      "learning_rate": 4.94226562144263e-06,
+      "loss": 0.1371,
+      "step": 1787
+    },
+    {
+      "epoch": 0.28969539857420606,
+      "grad_norm": 1.054979681968689,
+      "learning_rate": 4.942172146015814e-06,
+      "loss": 0.1486,
+      "step": 1788
+    },
+    {
+      "epoch": 0.28985742060920283,
+      "grad_norm": 1.1150860786437988,
+      "learning_rate": 4.942078595864441e-06,
+      "loss": 0.1521,
+      "step": 1789
+    },
+    {
+      "epoch": 0.2900194426441996,
+      "grad_norm": 1.2064958810806274,
+      "learning_rate": 4.941984970991372e-06,
+      "loss": 0.1682,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29018146467919637,
+      "grad_norm": 1.0891704559326172,
+      "learning_rate": 4.941891271399473e-06,
+      "loss": 0.1581,
+      "step": 1791
+    },
+    {
+      "epoch": 0.29034348671419313,
+      "grad_norm": 1.2679988145828247,
+      "learning_rate": 4.9417974970916096e-06,
+      "loss": 0.1671,
+      "step": 1792
+    },
+    {
+      "epoch": 0.2905055087491899,
+      "grad_norm": 1.0891644954681396,
+      "learning_rate": 4.941703648070653e-06,
+      "loss": 0.1621,
+      "step": 1793
+    },
+    {
+      "epoch": 0.29066753078418667,
+      "grad_norm": 0.9570609331130981,
+      "learning_rate": 4.9416097243394725e-06,
+      "loss": 0.1428,
+      "step": 1794
+    },
+    {
+      "epoch": 0.29082955281918343,
+      "grad_norm": 1.0233310461044312,
+      "learning_rate": 4.941515725900943e-06,
+      "loss": 0.158,
+      "step": 1795
+    },
+    {
+      "epoch": 0.29099157485418015,
+      "grad_norm": 1.1454052925109863,
+      "learning_rate": 4.94142165275794e-06,
+      "loss": 0.1722,
+      "step": 1796
+    },
+    {
+      "epoch": 0.2911535968891769,
+      "grad_norm": 1.0904475450515747,
+      "learning_rate": 4.941327504913344e-06,
+      "loss": 0.164,
+      "step": 1797
+    },
+    {
+      "epoch": 0.2913156189241737,
+      "grad_norm": 1.0754122734069824,
+      "learning_rate": 4.941233282370034e-06,
+      "loss": 0.1507,
+      "step": 1798
+    },
+    {
+      "epoch": 0.29147764095917045,
+      "grad_norm": 1.1219291687011719,
+      "learning_rate": 4.941138985130893e-06,
+      "loss": 0.1742,
+      "step": 1799
+    },
+    {
+      "epoch": 0.2916396629941672,
+      "grad_norm": 1.2050334215164185,
+      "learning_rate": 4.941044613198807e-06,
+      "loss": 0.1694,
+      "step": 1800
+    },
+    {
+      "epoch": 0.291801685029164,
+      "grad_norm": 1.0843576192855835,
+      "learning_rate": 4.940950166576661e-06,
+      "loss": 0.1558,
+      "step": 1801
+    },
+    {
+      "epoch": 0.29196370706416075,
+      "grad_norm": 1.1150481700897217,
+      "learning_rate": 4.940855645267349e-06,
+      "loss": 0.1559,
+      "step": 1802
+    },
+    {
+      "epoch": 0.29212572909915746,
+      "grad_norm": 1.0459260940551758,
+      "learning_rate": 4.94076104927376e-06,
+      "loss": 0.1608,
+      "step": 1803
+    },
+    {
+      "epoch": 0.29228775113415423,
+      "grad_norm": 1.014258623123169,
+      "learning_rate": 4.94066637859879e-06,
+      "loss": 0.133,
+      "step": 1804
+    },
+    {
+      "epoch": 0.292449773169151,
+      "grad_norm": 1.1400861740112305,
+      "learning_rate": 4.940571633245335e-06,
+      "loss": 0.1778,
+      "step": 1805
+    },
+    {
+      "epoch": 0.29261179520414776,
+      "grad_norm": 1.1275112628936768,
+      "learning_rate": 4.940476813216294e-06,
+      "loss": 0.1555,
+      "step": 1806
+    },
+    {
+      "epoch": 0.29277381723914453,
+      "grad_norm": 1.0518170595169067,
+      "learning_rate": 4.940381918514568e-06,
+      "loss": 0.159,
+      "step": 1807
+    },
+    {
+      "epoch": 0.2929358392741413,
+      "grad_norm": 1.0463672876358032,
+      "learning_rate": 4.940286949143061e-06,
+      "loss": 0.1614,
+      "step": 1808
+    },
+    {
+      "epoch": 0.29309786130913806,
+      "grad_norm": 1.200708270072937,
+      "learning_rate": 4.94019190510468e-06,
+      "loss": 0.1809,
+      "step": 1809
+    },
+    {
+      "epoch": 0.2932598833441348,
+      "grad_norm": 1.2362070083618164,
+      "learning_rate": 4.940096786402331e-06,
+      "loss": 0.1431,
+      "step": 1810
+    },
+    {
+      "epoch": 0.29342190537913154,
+      "grad_norm": 1.1330715417861938,
+      "learning_rate": 4.940001593038925e-06,
+      "loss": 0.1431,
+      "step": 1811
+    },
+    {
+      "epoch": 0.2935839274141283,
+      "grad_norm": 1.2720601558685303,
+      "learning_rate": 4.939906325017374e-06,
+      "loss": 0.1729,
+      "step": 1812
+    },
+    {
+      "epoch": 0.2937459494491251,
+      "grad_norm": 1.2702455520629883,
+      "learning_rate": 4.939810982340595e-06,
+      "loss": 0.1722,
+      "step": 1813
+    },
+    {
+      "epoch": 0.29390797148412184,
+      "grad_norm": 1.1870192289352417,
+      "learning_rate": 4.939715565011504e-06,
+      "loss": 0.1647,
+      "step": 1814
+    },
+    {
+      "epoch": 0.2940699935191186,
+      "grad_norm": 1.2095550298690796,
+      "learning_rate": 4.939620073033021e-06,
+      "loss": 0.1647,
+      "step": 1815
+    },
+    {
+      "epoch": 0.2942320155541154,
+      "grad_norm": 1.1870856285095215,
+      "learning_rate": 4.939524506408068e-06,
+      "loss": 0.1554,
+      "step": 1816
+    },
+    {
+      "epoch": 0.29439403758911215,
+      "grad_norm": 1.080191969871521,
+      "learning_rate": 4.939428865139568e-06,
+      "loss": 0.1572,
+      "step": 1817
+    },
+    {
+      "epoch": 0.29455605962410886,
+      "grad_norm": 1.1094964742660522,
+      "learning_rate": 4.939333149230447e-06,
+      "loss": 0.1623,
+      "step": 1818
+    },
+    {
+      "epoch": 0.2947180816591056,
+      "grad_norm": 1.0105398893356323,
+      "learning_rate": 4.939237358683636e-06,
+      "loss": 0.1591,
+      "step": 1819
+    },
+    {
+      "epoch": 0.2948801036941024,
+      "grad_norm": 1.1917641162872314,
+      "learning_rate": 4.9391414935020656e-06,
+      "loss": 0.1728,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29504212572909916,
+      "grad_norm": 1.0402076244354248,
+      "learning_rate": 4.939045553688666e-06,
+      "loss": 0.1525,
+      "step": 1821
+    },
+    {
+      "epoch": 0.2952041477640959,
+      "grad_norm": 1.235533356666565,
+      "learning_rate": 4.938949539246376e-06,
+      "loss": 0.1522,
+      "step": 1822
+    },
+    {
+      "epoch": 0.2953661697990927,
+      "grad_norm": 0.9965629577636719,
+      "learning_rate": 4.9388534501781325e-06,
+      "loss": 0.1515,
+      "step": 1823
+    },
+    {
+      "epoch": 0.29552819183408946,
+      "grad_norm": 1.0826505422592163,
+      "learning_rate": 4.938757286486874e-06,
+      "loss": 0.14,
+      "step": 1824
+    },
+    {
+      "epoch": 0.29569021386908617,
+      "grad_norm": 1.1033453941345215,
+      "learning_rate": 4.938661048175545e-06,
+      "loss": 0.1591,
+      "step": 1825
+    },
+    {
+      "epoch": 0.29585223590408294,
+      "grad_norm": 1.169182538986206,
+      "learning_rate": 4.93856473524709e-06,
+      "loss": 0.154,
+      "step": 1826
+    },
+    {
+      "epoch": 0.2960142579390797,
+      "grad_norm": 1.1595622301101685,
+      "learning_rate": 4.938468347704455e-06,
+      "loss": 0.1506,
+      "step": 1827
+    },
+    {
+      "epoch": 0.2961762799740765,
+      "grad_norm": 1.1768068075180054,
+      "learning_rate": 4.938371885550589e-06,
+      "loss": 0.1738,
+      "step": 1828
+    },
+    {
+      "epoch": 0.29633830200907324,
+      "grad_norm": 1.0577977895736694,
+      "learning_rate": 4.938275348788443e-06,
+      "loss": 0.1349,
+      "step": 1829
+    },
+    {
+      "epoch": 0.29650032404407,
+      "grad_norm": 1.1384469270706177,
+      "learning_rate": 4.938178737420974e-06,
+      "loss": 0.1506,
+      "step": 1830
+    },
+    {
+      "epoch": 0.2966623460790668,
+      "grad_norm": 1.1064151525497437,
+      "learning_rate": 4.938082051451135e-06,
+      "loss": 0.1583,
+      "step": 1831
+    },
+    {
+      "epoch": 0.2968243681140635,
+      "grad_norm": 1.0915504693984985,
+      "learning_rate": 4.937985290881886e-06,
+      "loss": 0.1642,
+      "step": 1832
+    },
+    {
+      "epoch": 0.29698639014906025,
+      "grad_norm": 0.9604697823524475,
+      "learning_rate": 4.937888455716186e-06,
+      "loss": 0.1342,
+      "step": 1833
+    },
+    {
+      "epoch": 0.297148412184057,
+      "grad_norm": 1.1359755992889404,
+      "learning_rate": 4.9377915459569995e-06,
+      "loss": 0.161,
+      "step": 1834
+    },
+    {
+      "epoch": 0.2973104342190538,
+      "grad_norm": 1.0202888250350952,
+      "learning_rate": 4.93769456160729e-06,
+      "loss": 0.144,
+      "step": 1835
+    },
+    {
+      "epoch": 0.29747245625405055,
+      "grad_norm": 1.0656861066818237,
+      "learning_rate": 4.937597502670027e-06,
+      "loss": 0.1468,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2976344782890473,
+      "grad_norm": 1.2268567085266113,
+      "learning_rate": 4.937500369148179e-06,
+      "loss": 0.1796,
+      "step": 1837
+    },
+    {
+      "epoch": 0.2977965003240441,
+      "grad_norm": 1.1543471813201904,
+      "learning_rate": 4.9374031610447185e-06,
+      "loss": 0.1666,
+      "step": 1838
+    },
+    {
+      "epoch": 0.29795852235904086,
+      "grad_norm": 1.1135607957839966,
+      "learning_rate": 4.9373058783626195e-06,
+      "loss": 0.1525,
+      "step": 1839
+    },
+    {
+      "epoch": 0.29812054439403757,
+      "grad_norm": 0.995367169380188,
+      "learning_rate": 4.937208521104858e-06,
+      "loss": 0.1383,
+      "step": 1840
+    },
+    {
+      "epoch": 0.29828256642903433,
+      "grad_norm": 1.0049978494644165,
+      "learning_rate": 4.9371110892744146e-06,
+      "loss": 0.1365,
+      "step": 1841
+    },
+    {
+      "epoch": 0.2984445884640311,
+      "grad_norm": 1.1172624826431274,
+      "learning_rate": 4.937013582874269e-06,
+      "loss": 0.1624,
+      "step": 1842
+    },
+    {
+      "epoch": 0.29860661049902787,
+      "grad_norm": 1.1392813920974731,
+      "learning_rate": 4.936916001907406e-06,
+      "loss": 0.1725,
+      "step": 1843
+    },
+    {
+      "epoch": 0.29876863253402464,
+      "grad_norm": 1.1414449214935303,
+      "learning_rate": 4.93681834637681e-06,
+      "loss": 0.1639,
+      "step": 1844
+    },
+    {
+      "epoch": 0.2989306545690214,
+      "grad_norm": 1.2181705236434937,
+      "learning_rate": 4.9367206162854695e-06,
+      "loss": 0.1662,
+      "step": 1845
+    },
+    {
+      "epoch": 0.29909267660401817,
+      "grad_norm": 1.1003068685531616,
+      "learning_rate": 4.936622811636376e-06,
+      "loss": 0.1654,
+      "step": 1846
+    },
+    {
+      "epoch": 0.2992546986390149,
+      "grad_norm": 1.1408618688583374,
+      "learning_rate": 4.93652493243252e-06,
+      "loss": 0.1622,
+      "step": 1847
+    },
+    {
+      "epoch": 0.29941672067401165,
+      "grad_norm": 1.5164649486541748,
+      "learning_rate": 4.936426978676897e-06,
+      "loss": 0.1466,
+      "step": 1848
+    },
+    {
+      "epoch": 0.2995787427090084,
+      "grad_norm": 1.0539342164993286,
+      "learning_rate": 4.9363289503725055e-06,
+      "loss": 0.1605,
+      "step": 1849
+    },
+    {
+      "epoch": 0.2997407647440052,
+      "grad_norm": 1.0239052772521973,
+      "learning_rate": 4.936230847522343e-06,
+      "loss": 0.1398,
+      "step": 1850
+    },
+    {
+      "epoch": 0.29990278677900195,
+      "grad_norm": 0.9674493670463562,
+      "learning_rate": 4.9361326701294124e-06,
+      "loss": 0.1389,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3000648088139987,
+      "grad_norm": 1.2970346212387085,
+      "learning_rate": 4.936034418196718e-06,
+      "loss": 0.1836,
+      "step": 1852
+    },
+    {
+      "epoch": 0.3002268308489955,
+      "grad_norm": 1.1987262964248657,
+      "learning_rate": 4.935936091727264e-06,
+      "loss": 0.1852,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3003888528839922,
+      "grad_norm": 1.051525354385376,
+      "learning_rate": 4.935837690724063e-06,
+      "loss": 0.1552,
+      "step": 1854
+    },
+    {
+      "epoch": 0.30055087491898896,
+      "grad_norm": 1.2423964738845825,
+      "learning_rate": 4.9357392151901204e-06,
+      "loss": 0.1768,
+      "step": 1855
+    },
+    {
+      "epoch": 0.30071289695398573,
+      "grad_norm": 1.0378652811050415,
+      "learning_rate": 4.935640665128454e-06,
+      "loss": 0.1486,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3008749189889825,
+      "grad_norm": 0.9993836879730225,
+      "learning_rate": 4.935542040542077e-06,
+      "loss": 0.1506,
+      "step": 1857
+    },
+    {
+      "epoch": 0.30103694102397927,
+      "grad_norm": 1.056410789489746,
+      "learning_rate": 4.935443341434008e-06,
+      "loss": 0.1442,
+      "step": 1858
+    },
+    {
+      "epoch": 0.30119896305897603,
+      "grad_norm": 1.1144357919692993,
+      "learning_rate": 4.935344567807265e-06,
+      "loss": 0.1679,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3013609850939728,
+      "grad_norm": 0.9949638247489929,
+      "learning_rate": 4.935245719664873e-06,
+      "loss": 0.1405,
+      "step": 1860
+    },
+    {
+      "epoch": 0.30152300712896957,
+      "grad_norm": 1.090957760810852,
+      "learning_rate": 4.935146797009854e-06,
+      "loss": 0.1663,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3016850291639663,
+      "grad_norm": 1.0422848463058472,
+      "learning_rate": 4.935047799845238e-06,
+      "loss": 0.1413,
+      "step": 1862
+    },
+    {
+      "epoch": 0.30184705119896305,
+      "grad_norm": 1.0720633268356323,
+      "learning_rate": 4.93494872817405e-06,
+      "loss": 0.1583,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3020090732339598,
+      "grad_norm": 1.1506516933441162,
+      "learning_rate": 4.9348495819993235e-06,
+      "loss": 0.1554,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3021710952689566,
+      "grad_norm": 1.3008663654327393,
+      "learning_rate": 4.934750361324092e-06,
+      "loss": 0.1636,
+      "step": 1865
+    },
+    {
+      "epoch": 0.30233311730395335,
+      "grad_norm": 1.1331379413604736,
+      "learning_rate": 4.9346510661513924e-06,
+      "loss": 0.149,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3024951393389501,
+      "grad_norm": 1.184205174446106,
+      "learning_rate": 4.934551696484262e-06,
+      "loss": 0.1718,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3026571613739469,
+      "grad_norm": 1.1299738883972168,
+      "learning_rate": 4.93445225232574e-06,
+      "loss": 0.1697,
+      "step": 1868
+    },
+    {
+      "epoch": 0.3028191834089436,
+      "grad_norm": 1.1674683094024658,
+      "learning_rate": 4.934352733678871e-06,
+      "loss": 0.1672,
+      "step": 1869
+    },
+    {
+      "epoch": 0.30298120544394036,
+      "grad_norm": 1.0782699584960938,
+      "learning_rate": 4.9342531405467e-06,
+      "loss": 0.165,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3031432274789371,
+      "grad_norm": 1.0888804197311401,
+      "learning_rate": 4.934153472932272e-06,
+      "loss": 0.1549,
+      "step": 1871
+    },
+    {
+      "epoch": 0.3033052495139339,
+      "grad_norm": 1.1337841749191284,
+      "learning_rate": 4.934053730838639e-06,
+      "loss": 0.1743,
+      "step": 1872
+    },
+    {
+      "epoch": 0.30346727154893066,
+      "grad_norm": 1.0406060218811035,
+      "learning_rate": 4.933953914268853e-06,
+      "loss": 0.1504,
+      "step": 1873
+    },
+    {
+      "epoch": 0.30362929358392743,
+      "grad_norm": 1.1255491971969604,
+      "learning_rate": 4.9338540232259664e-06,
+      "loss": 0.1566,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3037913156189242,
+      "grad_norm": 1.169628381729126,
+      "learning_rate": 4.933754057713037e-06,
+      "loss": 0.1566,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3039533376539209,
+      "grad_norm": 1.4170974493026733,
+      "learning_rate": 4.9336540177331225e-06,
+      "loss": 0.1623,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3041153596889177,
+      "grad_norm": 1.1771212816238403,
+      "learning_rate": 4.933553903289285e-06,
+      "loss": 0.1438,
+      "step": 1877
+    },
+    {
+      "epoch": 0.30427738172391444,
+      "grad_norm": 1.251401662826538,
+      "learning_rate": 4.9334537143845876e-06,
+      "loss": 0.1762,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3044394037589112,
+      "grad_norm": 1.1183122396469116,
+      "learning_rate": 4.933353451022094e-06,
+      "loss": 0.1617,
+      "step": 1879
+    },
+    {
+      "epoch": 0.304601425793908,
+      "grad_norm": 1.187208652496338,
+      "learning_rate": 4.933253113204874e-06,
+      "loss": 0.1815,
+      "step": 1880
+    },
+    {
+      "epoch": 0.30476344782890474,
+      "grad_norm": 1.1539027690887451,
+      "learning_rate": 4.933152700935997e-06,
+      "loss": 0.1718,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3049254698639015,
+      "grad_norm": 1.0791608095169067,
+      "learning_rate": 4.933052214218535e-06,
+      "loss": 0.1562,
+      "step": 1882
+    },
+    {
+      "epoch": 0.3050874918988983,
+      "grad_norm": 1.1433926820755005,
+      "learning_rate": 4.932951653055564e-06,
+      "loss": 0.1717,
+      "step": 1883
+    },
+    {
+      "epoch": 0.305249513933895,
+      "grad_norm": 1.2988801002502441,
+      "learning_rate": 4.93285101745016e-06,
+      "loss": 0.1815,
+      "step": 1884
+    },
+    {
+      "epoch": 0.30541153596889176,
+      "grad_norm": 1.1645463705062866,
+      "learning_rate": 4.932750307405402e-06,
+      "loss": 0.1674,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3055735580038885,
+      "grad_norm": 1.1217708587646484,
+      "learning_rate": 4.932649522924372e-06,
+      "loss": 0.1636,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3057355800388853,
+      "grad_norm": 1.087047815322876,
+      "learning_rate": 4.932548664010153e-06,
+      "loss": 0.1462,
+      "step": 1887
+    },
+    {
+      "epoch": 0.30589760207388206,
+      "grad_norm": 1.0934066772460938,
+      "learning_rate": 4.932447730665832e-06,
+      "loss": 0.176,
+      "step": 1888
+    },
+    {
+      "epoch": 0.3060596241088788,
+      "grad_norm": 1.0176986455917358,
+      "learning_rate": 4.9323467228944965e-06,
+      "loss": 0.1458,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3062216461438756,
+      "grad_norm": 0.9754590392112732,
+      "learning_rate": 4.932245640699238e-06,
+      "loss": 0.1338,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3063836681788723,
+      "grad_norm": 1.1034765243530273,
+      "learning_rate": 4.932144484083148e-06,
+      "loss": 0.16,
+      "step": 1891
+    },
+    {
+      "epoch": 0.30654569021386907,
+      "grad_norm": 1.0700440406799316,
+      "learning_rate": 4.932043253049323e-06,
+      "loss": 0.1531,
+      "step": 1892
+    },
+    {
+      "epoch": 0.30670771224886584,
+      "grad_norm": 1.2314177751541138,
+      "learning_rate": 4.93194194760086e-06,
+      "loss": 0.1607,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3068697342838626,
+      "grad_norm": 1.1649889945983887,
+      "learning_rate": 4.931840567740858e-06,
+      "loss": 0.1783,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3070317563188594,
+      "grad_norm": 1.099457025527954,
+      "learning_rate": 4.9317391134724195e-06,
+      "loss": 0.1763,
+      "step": 1895
+    },
+    {
+      "epoch": 0.30719377835385614,
+      "grad_norm": 1.0466156005859375,
+      "learning_rate": 4.93163758479865e-06,
+      "loss": 0.1406,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3073558003888529,
+      "grad_norm": 1.3971292972564697,
+      "learning_rate": 4.931535981722654e-06,
+      "loss": 0.1728,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3075178224238496,
+      "grad_norm": 1.2202054262161255,
+      "learning_rate": 4.931434304247541e-06,
+      "loss": 0.1857,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3076798444588464,
+      "grad_norm": 1.1348539590835571,
+      "learning_rate": 4.931332552376422e-06,
+      "loss": 0.1688,
+      "step": 1899
+    },
+    {
+      "epoch": 0.30784186649384315,
+      "grad_norm": 1.024888038635254,
+      "learning_rate": 4.931230726112412e-06,
+      "loss": 0.1495,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3080038885288399,
+      "grad_norm": 1.1341696977615356,
+      "learning_rate": 4.931128825458623e-06,
+      "loss": 0.1586,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3081659105638367,
+      "grad_norm": 1.094712495803833,
+      "learning_rate": 4.9310268504181764e-06,
+      "loss": 0.1744,
+      "step": 1902
+    },
+    {
+      "epoch": 0.30832793259883345,
+      "grad_norm": 1.1410554647445679,
+      "learning_rate": 4.930924800994192e-06,
+      "loss": 0.1684,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3084899546338302,
+      "grad_norm": 1.1105761528015137,
+      "learning_rate": 4.930822677189791e-06,
+      "loss": 0.1649,
+      "step": 1904
+    },
+    {
+      "epoch": 0.30865197666882693,
+      "grad_norm": 0.9734887480735779,
+      "learning_rate": 4.930720479008098e-06,
+      "loss": 0.1453,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3088139987038237,
+      "grad_norm": 1.0480237007141113,
+      "learning_rate": 4.93061820645224e-06,
+      "loss": 0.1511,
+      "step": 1906
+    },
+    {
+      "epoch": 0.30897602073882047,
+      "grad_norm": 1.168448567390442,
+      "learning_rate": 4.930515859525348e-06,
+      "loss": 0.151,
+      "step": 1907
+    },
+    {
+      "epoch": 0.30913804277381723,
+      "grad_norm": 1.0188759565353394,
+      "learning_rate": 4.930413438230552e-06,
+      "loss": 0.1355,
+      "step": 1908
+    },
+    {
+      "epoch": 0.309300064808814,
+      "grad_norm": 1.064790964126587,
+      "learning_rate": 4.930310942570987e-06,
+      "loss": 0.1482,
+      "step": 1909
+    },
+    {
+      "epoch": 0.30946208684381077,
+      "grad_norm": 1.1251689195632935,
+      "learning_rate": 4.930208372549787e-06,
+      "loss": 0.1616,
+      "step": 1910
+    },
+    {
+      "epoch": 0.30962410887880754,
+      "grad_norm": 1.050378441810608,
+      "learning_rate": 4.930105728170093e-06,
+      "loss": 0.1405,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3097861309138043,
+      "grad_norm": 1.0885461568832397,
+      "learning_rate": 4.930003009435043e-06,
+      "loss": 0.1513,
+      "step": 1912
+    },
+    {
+      "epoch": 0.309948152948801,
+      "grad_norm": 1.0849794149398804,
+      "learning_rate": 4.929900216347783e-06,
+      "loss": 0.1529,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3101101749837978,
+      "grad_norm": 1.1687252521514893,
+      "learning_rate": 4.9297973489114565e-06,
+      "loss": 0.1697,
+      "step": 1914
+    },
+    {
+      "epoch": 0.31027219701879455,
+      "grad_norm": 1.0808390378952026,
+      "learning_rate": 4.929694407129211e-06,
+      "loss": 0.1474,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3104342190537913,
+      "grad_norm": 1.0042861700057983,
+      "learning_rate": 4.929591391004196e-06,
+      "loss": 0.1445,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3105962410887881,
+      "grad_norm": 1.0821775197982788,
+      "learning_rate": 4.929488300539564e-06,
+      "loss": 0.1594,
+      "step": 1917
+    },
+    {
+      "epoch": 0.31075826312378485,
+      "grad_norm": 1.214570164680481,
+      "learning_rate": 4.929385135738469e-06,
+      "loss": 0.1792,
+      "step": 1918
+    },
+    {
+      "epoch": 0.3109202851587816,
+      "grad_norm": 1.0570710897445679,
+      "learning_rate": 4.929281896604068e-06,
+      "loss": 0.1638,
+      "step": 1919
+    },
+    {
+      "epoch": 0.31108230719377833,
+      "grad_norm": 0.8534372448921204,
+      "learning_rate": 4.92917858313952e-06,
+      "loss": 0.121,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3112443292287751,
+      "grad_norm": 1.1174410581588745,
+      "learning_rate": 4.9290751953479856e-06,
+      "loss": 0.1623,
+      "step": 1921
+    },
+    {
+      "epoch": 0.31140635126377186,
+      "grad_norm": 1.2499920129776,
+      "learning_rate": 4.928971733232628e-06,
+      "loss": 0.1853,
+      "step": 1922
+    },
+    {
+      "epoch": 0.31156837329876863,
+      "grad_norm": 0.961334764957428,
+      "learning_rate": 4.928868196796615e-06,
+      "loss": 0.1464,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3117303953337654,
+      "grad_norm": 1.0711408853530884,
+      "learning_rate": 4.928764586043111e-06,
+      "loss": 0.1597,
+      "step": 1924
+    },
+    {
+      "epoch": 0.31189241736876216,
+      "grad_norm": 1.0463709831237793,
+      "learning_rate": 4.928660900975289e-06,
+      "loss": 0.1594,
+      "step": 1925
+    },
+    {
+      "epoch": 0.31205443940375893,
+      "grad_norm": 1.1414794921875,
+      "learning_rate": 4.9285571415963205e-06,
+      "loss": 0.1703,
+      "step": 1926
+    },
+    {
+      "epoch": 0.31221646143875564,
+      "grad_norm": 1.1475234031677246,
+      "learning_rate": 4.928453307909381e-06,
+      "loss": 0.177,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3123784834737524,
+      "grad_norm": 0.9466526508331299,
+      "learning_rate": 4.928349399917646e-06,
+      "loss": 0.1405,
+      "step": 1928
+    },
+    {
+      "epoch": 0.3125405055087492,
+      "grad_norm": 1.0631221532821655,
+      "learning_rate": 4.928245417624297e-06,
+      "loss": 0.1643,
+      "step": 1929
+    },
+    {
+      "epoch": 0.31270252754374595,
+      "grad_norm": 1.1224422454833984,
+      "learning_rate": 4.928141361032513e-06,
+      "loss": 0.1689,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3128645495787427,
+      "grad_norm": 1.0592671632766724,
+      "learning_rate": 4.928037230145481e-06,
+      "loss": 0.1495,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3130265716137395,
+      "grad_norm": 1.071808099746704,
+      "learning_rate": 4.927933024966385e-06,
+      "loss": 0.1601,
+      "step": 1932
+    },
+    {
+      "epoch": 0.31318859364873625,
+      "grad_norm": 1.0835838317871094,
+      "learning_rate": 4.927828745498414e-06,
+      "loss": 0.1623,
+      "step": 1933
+    },
+    {
+      "epoch": 0.313350615683733,
+      "grad_norm": 1.1354283094406128,
+      "learning_rate": 4.927724391744758e-06,
+      "loss": 0.145,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3135126377187297,
+      "grad_norm": 1.1181296110153198,
+      "learning_rate": 4.9276199637086106e-06,
+      "loss": 0.1633,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3136746597537265,
+      "grad_norm": 1.1335471868515015,
+      "learning_rate": 4.927515461393167e-06,
+      "loss": 0.1455,
+      "step": 1936
+    },
+    {
+      "epoch": 0.31383668178872326,
+      "grad_norm": 1.091052770614624,
+      "learning_rate": 4.927410884801626e-06,
+      "loss": 0.1619,
+      "step": 1937
+    },
+    {
+      "epoch": 0.31399870382372,
+      "grad_norm": 1.1227446794509888,
+      "learning_rate": 4.927306233937185e-06,
+      "loss": 0.1757,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3141607258587168,
+      "grad_norm": 0.9883224368095398,
+      "learning_rate": 4.927201508803048e-06,
+      "loss": 0.1359,
+      "step": 1939
+    },
+    {
+      "epoch": 0.31432274789371356,
+      "grad_norm": 1.1337419748306274,
+      "learning_rate": 4.927096709402417e-06,
+      "loss": 0.1501,
+      "step": 1940
+    },
+    {
+      "epoch": 0.31448476992871033,
+      "grad_norm": 1.0160683393478394,
+      "learning_rate": 4.9269918357385015e-06,
+      "loss": 0.143,
+      "step": 1941
+    },
+    {
+      "epoch": 0.31464679196370704,
+      "grad_norm": 1.1797884702682495,
+      "learning_rate": 4.926886887814509e-06,
+      "loss": 0.1696,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3148088139987038,
+      "grad_norm": 1.0038928985595703,
+      "learning_rate": 4.92678186563365e-06,
+      "loss": 0.1523,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3149708360337006,
+      "grad_norm": 1.1223050355911255,
+      "learning_rate": 4.926676769199139e-06,
+      "loss": 0.1695,
+      "step": 1944
+    },
+    {
+      "epoch": 0.31513285806869734,
+      "grad_norm": 1.1693401336669922,
+      "learning_rate": 4.9265715985141914e-06,
+      "loss": 0.1676,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3152948801036941,
+      "grad_norm": 1.0387099981307983,
+      "learning_rate": 4.9264663535820256e-06,
+      "loss": 0.1502,
+      "step": 1946
+    },
+    {
+      "epoch": 0.3154569021386909,
+      "grad_norm": 1.1783015727996826,
+      "learning_rate": 4.926361034405861e-06,
+      "loss": 0.1697,
+      "step": 1947
+    },
+    {
+      "epoch": 0.31561892417368764,
+      "grad_norm": 1.137247920036316,
+      "learning_rate": 4.926255640988919e-06,
+      "loss": 0.1691,
+      "step": 1948
+    },
+    {
+      "epoch": 0.31578094620868435,
+      "grad_norm": 1.0531809329986572,
+      "learning_rate": 4.926150173334427e-06,
+      "loss": 0.1597,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3159429682436811,
+      "grad_norm": 1.0189039707183838,
+      "learning_rate": 4.926044631445611e-06,
+      "loss": 0.1597,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3161049902786779,
+      "grad_norm": 1.1217700242996216,
+      "learning_rate": 4.9259390153257006e-06,
+      "loss": 0.1609,
+      "step": 1951
+    },
+    {
+      "epoch": 0.31626701231367466,
+      "grad_norm": 1.0535531044006348,
+      "learning_rate": 4.925833324977926e-06,
+      "loss": 0.1584,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3164290343486714,
+      "grad_norm": 1.066357970237732,
+      "learning_rate": 4.925727560405522e-06,
+      "loss": 0.1622,
+      "step": 1953
+    },
+    {
+      "epoch": 0.3165910563836682,
+      "grad_norm": 1.065629005432129,
+      "learning_rate": 4.925621721611726e-06,
+      "loss": 0.1709,
+      "step": 1954
+    },
+    {
+      "epoch": 0.31675307841866496,
+      "grad_norm": 1.0723075866699219,
+      "learning_rate": 4.925515808599774e-06,
+      "loss": 0.165,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3169151004536617,
+      "grad_norm": 1.1291669607162476,
+      "learning_rate": 4.925409821372908e-06,
+      "loss": 0.1531,
+      "step": 1956
+    },
+    {
+      "epoch": 0.31707712248865844,
+      "grad_norm": 1.150839924812317,
+      "learning_rate": 4.925303759934372e-06,
+      "loss": 0.1676,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3172391445236552,
+      "grad_norm": 1.0867832899093628,
+      "learning_rate": 4.925197624287409e-06,
+      "loss": 0.1749,
+      "step": 1958
+    },
+    {
+      "epoch": 0.31740116655865197,
+      "grad_norm": 1.1858301162719727,
+      "learning_rate": 4.925091414435268e-06,
+      "loss": 0.1802,
+      "step": 1959
+    },
+    {
+      "epoch": 0.31756318859364874,
+      "grad_norm": 1.1005200147628784,
+      "learning_rate": 4.924985130381198e-06,
+      "loss": 0.1735,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3177252106286455,
+      "grad_norm": 1.015781044960022,
+      "learning_rate": 4.924878772128452e-06,
+      "loss": 0.1494,
+      "step": 1961
+    },
+    {
+      "epoch": 0.31788723266364227,
+      "grad_norm": 1.1282938718795776,
+      "learning_rate": 4.924772339680283e-06,
+      "loss": 0.1528,
+      "step": 1962
+    },
+    {
+      "epoch": 0.31804925469863904,
+      "grad_norm": 1.103499412536621,
+      "learning_rate": 4.9246658330399474e-06,
+      "loss": 0.1722,
+      "step": 1963
+    },
+    {
+      "epoch": 0.31821127673363575,
+      "grad_norm": 1.0378453731536865,
+      "learning_rate": 4.9245592522107065e-06,
+      "loss": 0.1648,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3183732987686325,
+      "grad_norm": 1.1685363054275513,
+      "learning_rate": 4.924452597195819e-06,
+      "loss": 0.1796,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3185353208036293,
+      "grad_norm": 1.0074273347854614,
+      "learning_rate": 4.92434586799855e-06,
+      "loss": 0.1525,
+      "step": 1966
+    },
+    {
+      "epoch": 0.31869734283862605,
+      "grad_norm": 0.9873900413513184,
+      "learning_rate": 4.924239064622163e-06,
+      "loss": 0.146,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3188593648736228,
+      "grad_norm": 1.1330924034118652,
+      "learning_rate": 4.924132187069928e-06,
+      "loss": 0.1609,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3190213869086196,
+      "grad_norm": 1.1219487190246582,
+      "learning_rate": 4.924025235345114e-06,
+      "loss": 0.1646,
+      "step": 1969
+    },
+    {
+      "epoch": 0.31918340894361635,
+      "grad_norm": 1.068410873413086,
+      "learning_rate": 4.923918209450994e-06,
+      "loss": 0.154,
+      "step": 1970
+    },
+    {
+      "epoch": 0.31934543097861307,
+      "grad_norm": 1.062517762184143,
+      "learning_rate": 4.923811109390843e-06,
+      "loss": 0.1454,
+      "step": 1971
+    },
+    {
+      "epoch": 0.31950745301360983,
+      "grad_norm": 1.0570393800735474,
+      "learning_rate": 4.9237039351679365e-06,
+      "loss": 0.1551,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3196694750486066,
+      "grad_norm": 1.120102882385254,
+      "learning_rate": 4.923596686785556e-06,
+      "loss": 0.1628,
+      "step": 1973
+    },
+    {
+      "epoch": 0.31983149708360337,
+      "grad_norm": 1.1168919801712036,
+      "learning_rate": 4.923489364246981e-06,
+      "loss": 0.1715,
+      "step": 1974
+    },
+    {
+      "epoch": 0.31999351911860013,
+      "grad_norm": 1.2127281427383423,
+      "learning_rate": 4.923381967555496e-06,
+      "loss": 0.1598,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3201555411535969,
+      "grad_norm": 1.1189041137695312,
+      "learning_rate": 4.923274496714387e-06,
+      "loss": 0.1604,
+      "step": 1976
+    },
+    {
+      "epoch": 0.32031756318859367,
+      "grad_norm": 1.0939730405807495,
+      "learning_rate": 4.923166951726945e-06,
+      "loss": 0.1557,
+      "step": 1977
+    },
+    {
+      "epoch": 0.32047958522359044,
+      "grad_norm": 1.0992319583892822,
+      "learning_rate": 4.923059332596456e-06,
+      "loss": 0.1615,
+      "step": 1978
+    },
+    {
+      "epoch": 0.32064160725858715,
+      "grad_norm": 1.0739459991455078,
+      "learning_rate": 4.922951639326215e-06,
+      "loss": 0.1537,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3208036292935839,
+      "grad_norm": 1.1589041948318481,
+      "learning_rate": 4.922843871919518e-06,
+      "loss": 0.1663,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3209656513285807,
+      "grad_norm": 1.0140794515609741,
+      "learning_rate": 4.922736030379662e-06,
+      "loss": 0.1435,
+      "step": 1981
+    },
+    {
+      "epoch": 0.32112767336357745,
+      "grad_norm": 1.1577235460281372,
+      "learning_rate": 4.922628114709945e-06,
+      "loss": 0.1815,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3212896953985742,
+      "grad_norm": 0.9740235805511475,
+      "learning_rate": 4.922520124913672e-06,
+      "loss": 0.1442,
+      "step": 1983
+    },
+    {
+      "epoch": 0.321451717433571,
+      "grad_norm": 1.3087810277938843,
+      "learning_rate": 4.922412060994145e-06,
+      "loss": 0.1556,
+      "step": 1984
+    },
+    {
+      "epoch": 0.32161373946856775,
+      "grad_norm": 1.1714460849761963,
+      "learning_rate": 4.922303922954671e-06,
+      "loss": 0.1828,
+      "step": 1985
+    },
+    {
+      "epoch": 0.32177576150356446,
+      "grad_norm": 0.9607410430908203,
+      "learning_rate": 4.922195710798559e-06,
+      "loss": 0.1292,
+      "step": 1986
+    },
+    {
+      "epoch": 0.32193778353856123,
+      "grad_norm": 1.0208156108856201,
+      "learning_rate": 4.9220874245291194e-06,
+      "loss": 0.143,
+      "step": 1987
+    },
+    {
+      "epoch": 0.322099805573558,
+      "grad_norm": 1.004679560661316,
+      "learning_rate": 4.9219790641496656e-06,
+      "loss": 0.1454,
+      "step": 1988
+    },
+    {
+      "epoch": 0.32226182760855476,
+      "grad_norm": 1.2176995277404785,
+      "learning_rate": 4.921870629663514e-06,
+      "loss": 0.1631,
+      "step": 1989
+    },
+    {
+      "epoch": 0.32242384964355153,
+      "grad_norm": 1.067450761795044,
+      "learning_rate": 4.9217621210739826e-06,
+      "loss": 0.1605,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3225858716785483,
+      "grad_norm": 1.1650382280349731,
+      "learning_rate": 4.92165353838439e-06,
+      "loss": 0.1758,
+      "step": 1991
+    },
+    {
+      "epoch": 0.32274789371354506,
+      "grad_norm": 1.0696114301681519,
+      "learning_rate": 4.921544881598059e-06,
+      "loss": 0.1483,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3229099157485418,
+      "grad_norm": 1.1148940324783325,
+      "learning_rate": 4.921436150718316e-06,
+      "loss": 0.1585,
+      "step": 1993
+    },
+    {
+      "epoch": 0.32307193778353854,
+      "grad_norm": 1.0661381483078003,
+      "learning_rate": 4.921327345748486e-06,
+      "loss": 0.1714,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3232339598185353,
+      "grad_norm": 1.0887075662612915,
+      "learning_rate": 4.921218466691898e-06,
+      "loss": 0.1494,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3233959818535321,
+      "grad_norm": 1.2316148281097412,
+      "learning_rate": 4.921109513551885e-06,
+      "loss": 0.1681,
+      "step": 1996
+    },
+    {
+      "epoch": 0.32355800388852884,
+      "grad_norm": 0.982879102230072,
+      "learning_rate": 4.92100048633178e-06,
+      "loss": 0.1363,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3237200259235256,
+      "grad_norm": 1.0092542171478271,
+      "learning_rate": 4.920891385034918e-06,
+      "loss": 0.1371,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3238820479585224,
+      "grad_norm": 1.0851917266845703,
+      "learning_rate": 4.9207822096646385e-06,
+      "loss": 0.144,
+      "step": 1999
+    },
+    {
+      "epoch": 0.32404406999351915,
+      "grad_norm": 0.9787352681159973,
+      "learning_rate": 4.920672960224282e-06,
+      "loss": 0.146,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32420609202851586,
+      "grad_norm": 1.1132274866104126,
+      "learning_rate": 4.92056363671719e-06,
+      "loss": 0.1693,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3243681140635126,
+      "grad_norm": 1.1033443212509155,
+      "learning_rate": 4.920454239146709e-06,
+      "loss": 0.1584,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3245301360985094,
+      "grad_norm": 1.1150498390197754,
+      "learning_rate": 4.920344767516186e-06,
+      "loss": 0.1472,
+      "step": 2003
+    },
+    {
+      "epoch": 0.32469215813350616,
+      "grad_norm": 1.13274347782135,
+      "learning_rate": 4.92023522182897e-06,
+      "loss": 0.1729,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3248541801685029,
+      "grad_norm": 0.9556041359901428,
+      "learning_rate": 4.920125602088412e-06,
+      "loss": 0.1365,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3250162022034997,
+      "grad_norm": 0.979815661907196,
+      "learning_rate": 4.9200159082978685e-06,
+      "loss": 0.1376,
+      "step": 2006
+    },
+    {
+      "epoch": 0.32517822423849646,
+      "grad_norm": 1.1807565689086914,
+      "learning_rate": 4.919906140460693e-06,
+      "loss": 0.1665,
+      "step": 2007
+    },
+    {
+      "epoch": 0.32534024627349317,
+      "grad_norm": 1.0704436302185059,
+      "learning_rate": 4.919796298580247e-06,
+      "loss": 0.158,
+      "step": 2008
+    },
+    {
+      "epoch": 0.32550226830848994,
+      "grad_norm": 1.0892672538757324,
+      "learning_rate": 4.919686382659889e-06,
+      "loss": 0.1461,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3256642903434867,
+      "grad_norm": 1.1033027172088623,
+      "learning_rate": 4.919576392702984e-06,
+      "loss": 0.1592,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3258263123784835,
+      "grad_norm": 1.2258738279342651,
+      "learning_rate": 4.919466328712897e-06,
+      "loss": 0.1911,
+      "step": 2011
+    },
+    {
+      "epoch": 0.32598833441348024,
+      "grad_norm": 0.9703921675682068,
+      "learning_rate": 4.9193561906929945e-06,
+      "loss": 0.1397,
+      "step": 2012
+    },
+    {
+      "epoch": 0.326150356448477,
+      "grad_norm": 1.1200995445251465,
+      "learning_rate": 4.919245978646648e-06,
+      "loss": 0.1705,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3263123784834738,
+      "grad_norm": 1.0445135831832886,
+      "learning_rate": 4.919135692577229e-06,
+      "loss": 0.1488,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3264744005184705,
+      "grad_norm": 0.9916481375694275,
+      "learning_rate": 4.919025332488111e-06,
+      "loss": 0.134,
+      "step": 2015
+    },
+    {
+      "epoch": 0.32663642255346725,
+      "grad_norm": 1.1209403276443481,
+      "learning_rate": 4.918914898382673e-06,
+      "loss": 0.1423,
+      "step": 2016
+    },
+    {
+      "epoch": 0.326798444588464,
+      "grad_norm": 0.954210638999939,
+      "learning_rate": 4.918804390264292e-06,
+      "loss": 0.1461,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3269604666234608,
+      "grad_norm": 1.1691701412200928,
+      "learning_rate": 4.91869380813635e-06,
+      "loss": 0.1693,
+      "step": 2018
+    },
+    {
+      "epoch": 0.32712248865845756,
+      "grad_norm": 1.2081804275512695,
+      "learning_rate": 4.918583152002231e-06,
+      "loss": 0.1663,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3272845106934543,
+      "grad_norm": 1.0673002004623413,
+      "learning_rate": 4.91847242186532e-06,
+      "loss": 0.1625,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3274465327284511,
+      "grad_norm": 1.0329972505569458,
+      "learning_rate": 4.918361617729006e-06,
+      "loss": 0.1458,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3276085547634478,
+      "grad_norm": 1.0796326398849487,
+      "learning_rate": 4.918250739596678e-06,
+      "loss": 0.1394,
+      "step": 2022
+    },
+    {
+      "epoch": 0.32777057679844457,
+      "grad_norm": 1.0734831094741821,
+      "learning_rate": 4.91813978747173e-06,
+      "loss": 0.1557,
+      "step": 2023
+    },
+    {
+      "epoch": 0.32793259883344134,
+      "grad_norm": 1.1349564790725708,
+      "learning_rate": 4.918028761357557e-06,
+      "loss": 0.1694,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3280946208684381,
+      "grad_norm": 1.027542233467102,
+      "learning_rate": 4.917917661257554e-06,
+      "loss": 0.1479,
+      "step": 2025
+    },
+    {
+      "epoch": 0.32825664290343487,
+      "grad_norm": 1.1970611810684204,
+      "learning_rate": 4.917806487175123e-06,
+      "loss": 0.1549,
+      "step": 2026
+    },
+    {
+      "epoch": 0.32841866493843164,
+      "grad_norm": 1.0326762199401855,
+      "learning_rate": 4.917695239113665e-06,
+      "loss": 0.1454,
+      "step": 2027
+    },
+    {
+      "epoch": 0.3285806869734284,
+      "grad_norm": 1.1361476182937622,
+      "learning_rate": 4.917583917076581e-06,
+      "loss": 0.1529,
+      "step": 2028
+    },
+    {
+      "epoch": 0.32874270900842517,
+      "grad_norm": 1.0551778078079224,
+      "learning_rate": 4.917472521067281e-06,
+      "loss": 0.155,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3289047310434219,
+      "grad_norm": 1.0310840606689453,
+      "learning_rate": 4.917361051089172e-06,
+      "loss": 0.1433,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32906675307841865,
+      "grad_norm": 1.0534615516662598,
+      "learning_rate": 4.917249507145665e-06,
+      "loss": 0.1675,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3292287751134154,
+      "grad_norm": 1.0969111919403076,
+      "learning_rate": 4.917137889240172e-06,
+      "loss": 0.1741,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3293907971484122,
+      "grad_norm": 1.0823769569396973,
+      "learning_rate": 4.91702619737611e-06,
+      "loss": 0.1621,
+      "step": 2033
+    },
+    {
+      "epoch": 0.32955281918340895,
+      "grad_norm": 1.058327317237854,
+      "learning_rate": 4.916914431556895e-06,
+      "loss": 0.1553,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3297148412184057,
+      "grad_norm": 1.0909065008163452,
+      "learning_rate": 4.9168025917859465e-06,
+      "loss": 0.1621,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3298768632534025,
+      "grad_norm": 1.1126458644866943,
+      "learning_rate": 4.916690678066688e-06,
+      "loss": 0.1628,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3300388852883992,
+      "grad_norm": 1.1230723857879639,
+      "learning_rate": 4.916578690402542e-06,
+      "loss": 0.1714,
+      "step": 2037
+    },
+    {
+      "epoch": 0.33020090732339596,
+      "grad_norm": 1.2173502445220947,
+      "learning_rate": 4.916466628796938e-06,
+      "loss": 0.1657,
+      "step": 2038
+    },
+    {
+      "epoch": 0.33036292935839273,
+      "grad_norm": 0.9629350900650024,
+      "learning_rate": 4.916354493253301e-06,
+      "loss": 0.1262,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3305249513933895,
+      "grad_norm": 1.159567952156067,
+      "learning_rate": 4.9162422837750654e-06,
+      "loss": 0.1867,
+      "step": 2040
+    },
+    {
+      "epoch": 0.33068697342838627,
+      "grad_norm": 1.039483904838562,
+      "learning_rate": 4.916130000365662e-06,
+      "loss": 0.1369,
+      "step": 2041
+    },
+    {
+      "epoch": 0.33084899546338303,
+      "grad_norm": 1.0981976985931396,
+      "learning_rate": 4.916017643028529e-06,
+      "loss": 0.1542,
+      "step": 2042
+    },
+    {
+      "epoch": 0.3310110174983798,
+      "grad_norm": 1.1774072647094727,
+      "learning_rate": 4.915905211767101e-06,
+      "loss": 0.1406,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3311730395333765,
+      "grad_norm": 1.048284649848938,
+      "learning_rate": 4.915792706584821e-06,
+      "loss": 0.1514,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3313350615683733,
+      "grad_norm": 1.050802230834961,
+      "learning_rate": 4.9156801274851295e-06,
+      "loss": 0.1515,
+      "step": 2045
+    },
+    {
+      "epoch": 0.33149708360337005,
+      "grad_norm": 1.0515334606170654,
+      "learning_rate": 4.9155674744714725e-06,
+      "loss": 0.1386,
+      "step": 2046
+    },
+    {
+      "epoch": 0.3316591056383668,
+      "grad_norm": 0.9953398704528809,
+      "learning_rate": 4.915454747547296e-06,
+      "loss": 0.1465,
+      "step": 2047
+    },
+    {
+      "epoch": 0.3318211276733636,
+      "grad_norm": 0.9482908844947815,
+      "learning_rate": 4.91534194671605e-06,
+      "loss": 0.14,
+      "step": 2048
+    },
+    {
+      "epoch": 0.33198314970836035,
+      "grad_norm": 1.0365067720413208,
+      "learning_rate": 4.915229071981186e-06,
+      "loss": 0.1446,
+      "step": 2049
+    },
+    {
+      "epoch": 0.3321451717433571,
+      "grad_norm": 1.1342229843139648,
+      "learning_rate": 4.915116123346155e-06,
+      "loss": 0.1796,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3323071937783539,
+      "grad_norm": 1.1378239393234253,
+      "learning_rate": 4.915003100814417e-06,
+      "loss": 0.1679,
+      "step": 2051
+    },
+    {
+      "epoch": 0.3324692158133506,
+      "grad_norm": 1.1816859245300293,
+      "learning_rate": 4.9148900043894275e-06,
+      "loss": 0.185,
+      "step": 2052
+    },
+    {
+      "epoch": 0.33263123784834736,
+      "grad_norm": 1.2354105710983276,
+      "learning_rate": 4.9147768340746486e-06,
+      "loss": 0.1913,
+      "step": 2053
+    },
+    {
+      "epoch": 0.33279325988334413,
+      "grad_norm": 1.0721561908721924,
+      "learning_rate": 4.914663589873541e-06,
+      "loss": 0.156,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3329552819183409,
+      "grad_norm": 1.0587486028671265,
+      "learning_rate": 4.914550271789572e-06,
+      "loss": 0.1607,
+      "step": 2055
+    },
+    {
+      "epoch": 0.33311730395333766,
+      "grad_norm": 1.0155506134033203,
+      "learning_rate": 4.914436879826207e-06,
+      "loss": 0.1528,
+      "step": 2056
+    },
+    {
+      "epoch": 0.33327932598833443,
+      "grad_norm": 1.0975433588027954,
+      "learning_rate": 4.914323413986917e-06,
+      "loss": 0.1606,
+      "step": 2057
+    },
+    {
+      "epoch": 0.3334413480233312,
+      "grad_norm": 1.12659752368927,
+      "learning_rate": 4.9142098742751726e-06,
+      "loss": 0.1677,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3336033700583279,
+      "grad_norm": 1.0046424865722656,
+      "learning_rate": 4.914096260694449e-06,
+      "loss": 0.163,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3337653920933247,
+      "grad_norm": 1.103347897529602,
+      "learning_rate": 4.9139825732482205e-06,
+      "loss": 0.1583,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33392741412832144,
+      "grad_norm": 1.109889268875122,
+      "learning_rate": 4.913868811939968e-06,
+      "loss": 0.162,
+      "step": 2061
+    },
+    {
+      "epoch": 0.3340894361633182,
+      "grad_norm": 1.036608338356018,
+      "learning_rate": 4.91375497677317e-06,
+      "loss": 0.1487,
+      "step": 2062
+    },
+    {
+      "epoch": 0.334251458198315,
+      "grad_norm": 1.199803113937378,
+      "learning_rate": 4.913641067751313e-06,
+      "loss": 0.1858,
+      "step": 2063
+    },
+    {
+      "epoch": 0.33441348023331174,
+      "grad_norm": 1.304969072341919,
+      "learning_rate": 4.913527084877879e-06,
+      "loss": 0.1704,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3345755022683085,
+      "grad_norm": 1.1703951358795166,
+      "learning_rate": 4.913413028156358e-06,
+      "loss": 0.1688,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3347375243033052,
+      "grad_norm": 1.0002027750015259,
+      "learning_rate": 4.913298897590237e-06,
+      "loss": 0.1413,
+      "step": 2066
+    },
+    {
+      "epoch": 0.334899546338302,
+      "grad_norm": 1.1760587692260742,
+      "learning_rate": 4.913184693183011e-06,
+      "loss": 0.177,
+      "step": 2067
+    },
+    {
+      "epoch": 0.33506156837329876,
+      "grad_norm": 1.080187439918518,
+      "learning_rate": 4.913070414938172e-06,
+      "loss": 0.1556,
+      "step": 2068
+    },
+    {
+      "epoch": 0.3352235904082955,
+      "grad_norm": 1.124131202697754,
+      "learning_rate": 4.912956062859219e-06,
+      "loss": 0.1701,
+      "step": 2069
+    },
+    {
+      "epoch": 0.3353856124432923,
+      "grad_norm": 1.0667952299118042,
+      "learning_rate": 4.912841636949649e-06,
+      "loss": 0.1503,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33554763447828906,
+      "grad_norm": 1.2242783308029175,
+      "learning_rate": 4.912727137212964e-06,
+      "loss": 0.1686,
+      "step": 2071
+    },
+    {
+      "epoch": 0.3357096565132858,
+      "grad_norm": 1.0249295234680176,
+      "learning_rate": 4.912612563652667e-06,
+      "loss": 0.1517,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3358716785482826,
+      "grad_norm": 1.5885037183761597,
+      "learning_rate": 4.912497916272264e-06,
+      "loss": 0.1591,
+      "step": 2073
+    },
+    {
+      "epoch": 0.3360337005832793,
+      "grad_norm": 1.040554404258728,
+      "learning_rate": 4.912383195075264e-06,
+      "loss": 0.1571,
+      "step": 2074
+    },
+    {
+      "epoch": 0.33619572261827607,
+      "grad_norm": 1.0694094896316528,
+      "learning_rate": 4.912268400065175e-06,
+      "loss": 0.1511,
+      "step": 2075
+    },
+    {
+      "epoch": 0.33635774465327284,
+      "grad_norm": 2.055079460144043,
+      "learning_rate": 4.912153531245511e-06,
+      "loss": 0.1697,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3365197666882696,
+      "grad_norm": 1.066028356552124,
+      "learning_rate": 4.912038588619786e-06,
+      "loss": 0.1572,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3366817887232664,
+      "grad_norm": 1.1714937686920166,
+      "learning_rate": 4.9119235721915174e-06,
+      "loss": 0.1764,
+      "step": 2078
+    },
+    {
+      "epoch": 0.33684381075826314,
+      "grad_norm": 1.0577641725540161,
+      "learning_rate": 4.911808481964224e-06,
+      "loss": 0.1411,
+      "step": 2079
+    },
+    {
+      "epoch": 0.3370058327932599,
+      "grad_norm": 1.0684007406234741,
+      "learning_rate": 4.911693317941428e-06,
+      "loss": 0.154,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3371678548282566,
+      "grad_norm": 1.107500672340393,
+      "learning_rate": 4.911578080126652e-06,
+      "loss": 0.1549,
+      "step": 2081
+    },
+    {
+      "epoch": 0.3373298768632534,
+      "grad_norm": 1.0562461614608765,
+      "learning_rate": 4.911462768523423e-06,
+      "loss": 0.1475,
+      "step": 2082
+    },
+    {
+      "epoch": 0.33749189889825015,
+      "grad_norm": 1.016905426979065,
+      "learning_rate": 4.911347383135269e-06,
+      "loss": 0.1391,
+      "step": 2083
+    },
+    {
+      "epoch": 0.3376539209332469,
+      "grad_norm": 1.1356723308563232,
+      "learning_rate": 4.9112319239657204e-06,
+      "loss": 0.1747,
+      "step": 2084
+    },
+    {
+      "epoch": 0.3378159429682437,
+      "grad_norm": 1.1550018787384033,
+      "learning_rate": 4.91111639101831e-06,
+      "loss": 0.1509,
+      "step": 2085
+    },
+    {
+      "epoch": 0.33797796500324045,
+      "grad_norm": 1.108013391494751,
+      "learning_rate": 4.911000784296572e-06,
+      "loss": 0.1712,
+      "step": 2086
+    },
+    {
+      "epoch": 0.3381399870382372,
+      "grad_norm": 1.057617425918579,
+      "learning_rate": 4.910885103804046e-06,
+      "loss": 0.1624,
+      "step": 2087
+    },
+    {
+      "epoch": 0.33830200907323393,
+      "grad_norm": 1.2254884243011475,
+      "learning_rate": 4.910769349544269e-06,
+      "loss": 0.1485,
+      "step": 2088
+    },
+    {
+      "epoch": 0.3384640311082307,
+      "grad_norm": 1.2375078201293945,
+      "learning_rate": 4.910653521520784e-06,
+      "loss": 0.1587,
+      "step": 2089
+    },
+    {
+      "epoch": 0.33862605314322747,
+      "grad_norm": 1.054060459136963,
+      "learning_rate": 4.9105376197371355e-06,
+      "loss": 0.1391,
+      "step": 2090
+    },
+    {
+      "epoch": 0.33878807517822424,
+      "grad_norm": 1.0424585342407227,
+      "learning_rate": 4.910421644196868e-06,
+      "loss": 0.1532,
+      "step": 2091
+    },
+    {
+      "epoch": 0.338950097213221,
+      "grad_norm": 1.091741681098938,
+      "learning_rate": 4.9103055949035326e-06,
+      "loss": 0.1619,
+      "step": 2092
+    },
+    {
+      "epoch": 0.33911211924821777,
+      "grad_norm": 1.0904366970062256,
+      "learning_rate": 4.910189471860678e-06,
+      "loss": 0.1675,
+      "step": 2093
+    },
+    {
+      "epoch": 0.33927414128321454,
+      "grad_norm": 1.1803327798843384,
+      "learning_rate": 4.910073275071858e-06,
+      "loss": 0.1856,
+      "step": 2094
+    },
+    {
+      "epoch": 0.3394361633182113,
+      "grad_norm": 1.1525360345840454,
+      "learning_rate": 4.909957004540629e-06,
+      "loss": 0.1703,
+      "step": 2095
+    },
+    {
+      "epoch": 0.339598185353208,
+      "grad_norm": 1.070910096168518,
+      "learning_rate": 4.909840660270547e-06,
+      "loss": 0.1768,
+      "step": 2096
+    },
+    {
+      "epoch": 0.3397602073882048,
+      "grad_norm": 0.9406764507293701,
+      "learning_rate": 4.909724242265172e-06,
+      "loss": 0.1367,
+      "step": 2097
+    },
+    {
+      "epoch": 0.33992222942320155,
+      "grad_norm": 1.0764198303222656,
+      "learning_rate": 4.909607750528068e-06,
+      "loss": 0.1583,
+      "step": 2098
+    },
+    {
+      "epoch": 0.3400842514581983,
+      "grad_norm": 0.9914731383323669,
+      "learning_rate": 4.909491185062797e-06,
+      "loss": 0.1486,
+      "step": 2099
+    },
+    {
+      "epoch": 0.3402462734931951,
+      "grad_norm": 1.056030511856079,
+      "learning_rate": 4.909374545872927e-06,
+      "loss": 0.1546,
+      "step": 2100
+    },
+    {
+      "epoch": 0.34040829552819185,
+      "grad_norm": 1.0676145553588867,
+      "learning_rate": 4.909257832962026e-06,
+      "loss": 0.1284,
+      "step": 2101
+    },
+    {
+      "epoch": 0.3405703175631886,
+      "grad_norm": 1.0468038320541382,
+      "learning_rate": 4.909141046333666e-06,
+      "loss": 0.1365,
+      "step": 2102
+    },
+    {
+      "epoch": 0.34073233959818533,
+      "grad_norm": 1.0132776498794556,
+      "learning_rate": 4.90902418599142e-06,
+      "loss": 0.1521,
+      "step": 2103
+    },
+    {
+      "epoch": 0.3408943616331821,
+      "grad_norm": 1.1214451789855957,
+      "learning_rate": 4.908907251938864e-06,
+      "loss": 0.1561,
+      "step": 2104
+    },
+    {
+      "epoch": 0.34105638366817886,
+      "grad_norm": 1.1212339401245117,
+      "learning_rate": 4.9087902441795745e-06,
+      "loss": 0.161,
+      "step": 2105
+    },
+    {
+      "epoch": 0.34121840570317563,
+      "grad_norm": 1.0603835582733154,
+      "learning_rate": 4.908673162717133e-06,
+      "loss": 0.1596,
+      "step": 2106
+    },
+    {
+      "epoch": 0.3413804277381724,
+      "grad_norm": 1.171034812927246,
+      "learning_rate": 4.908556007555122e-06,
+      "loss": 0.1534,
+      "step": 2107
+    },
+    {
+      "epoch": 0.34154244977316917,
+      "grad_norm": 1.055219292640686,
+      "learning_rate": 4.908438778697125e-06,
+      "loss": 0.1511,
+      "step": 2108
+    },
+    {
+      "epoch": 0.34170447180816593,
+      "grad_norm": 1.1049814224243164,
+      "learning_rate": 4.90832147614673e-06,
+      "loss": 0.1679,
+      "step": 2109
+    },
+    {
+      "epoch": 0.34186649384316264,
+      "grad_norm": 1.144884705543518,
+      "learning_rate": 4.908204099907527e-06,
+      "loss": 0.1635,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3420285158781594,
+      "grad_norm": 1.1654030084609985,
+      "learning_rate": 4.908086649983105e-06,
+      "loss": 0.1663,
+      "step": 2111
+    },
+    {
+      "epoch": 0.3421905379131562,
+      "grad_norm": 1.1463876962661743,
+      "learning_rate": 4.907969126377059e-06,
+      "loss": 0.1601,
+      "step": 2112
+    },
+    {
+      "epoch": 0.34235255994815295,
+      "grad_norm": 1.105735182762146,
+      "learning_rate": 4.9078515290929855e-06,
+      "loss": 0.1573,
+      "step": 2113
+    },
+    {
+      "epoch": 0.3425145819831497,
+      "grad_norm": 1.1214326620101929,
+      "learning_rate": 4.907733858134482e-06,
+      "loss": 0.1705,
+      "step": 2114
+    },
+    {
+      "epoch": 0.3426766040181465,
+      "grad_norm": 0.9656097292900085,
+      "learning_rate": 4.907616113505149e-06,
+      "loss": 0.1477,
+      "step": 2115
+    },
+    {
+      "epoch": 0.34283862605314325,
+      "grad_norm": 1.0422648191452026,
+      "learning_rate": 4.907498295208589e-06,
+      "loss": 0.1577,
+      "step": 2116
+    },
+    {
+      "epoch": 0.34300064808814,
+      "grad_norm": 1.0416967868804932,
+      "learning_rate": 4.907380403248408e-06,
+      "loss": 0.1674,
+      "step": 2117
+    },
+    {
+      "epoch": 0.3431626701231367,
+      "grad_norm": 1.0211541652679443,
+      "learning_rate": 4.907262437628211e-06,
+      "loss": 0.1615,
+      "step": 2118
+    },
+    {
+      "epoch": 0.3433246921581335,
+      "grad_norm": 1.0481899976730347,
+      "learning_rate": 4.90714439835161e-06,
+      "loss": 0.1624,
+      "step": 2119
+    },
+    {
+      "epoch": 0.34348671419313026,
+      "grad_norm": 0.9321041703224182,
+      "learning_rate": 4.907026285422215e-06,
+      "loss": 0.149,
+      "step": 2120
+    },
+    {
+      "epoch": 0.343648736228127,
+      "grad_norm": 1.062423825263977,
+      "learning_rate": 4.9069080988436405e-06,
+      "loss": 0.1582,
+      "step": 2121
+    },
+    {
+      "epoch": 0.3438107582631238,
+      "grad_norm": 1.213869571685791,
+      "learning_rate": 4.906789838619504e-06,
+      "loss": 0.1728,
+      "step": 2122
+    },
+    {
+      "epoch": 0.34397278029812056,
+      "grad_norm": 1.011985421180725,
+      "learning_rate": 4.9066715047534205e-06,
+      "loss": 0.1525,
+      "step": 2123
+    },
+    {
+      "epoch": 0.34413480233311733,
+      "grad_norm": 1.0821608304977417,
+      "learning_rate": 4.906553097249015e-06,
+      "loss": 0.1607,
+      "step": 2124
+    },
+    {
+      "epoch": 0.34429682436811404,
+      "grad_norm": 0.9949021339416504,
+      "learning_rate": 4.906434616109907e-06,
+      "loss": 0.1306,
+      "step": 2125
+    },
+    {
+      "epoch": 0.3444588464031108,
+      "grad_norm": 1.1297369003295898,
+      "learning_rate": 4.906316061339724e-06,
+      "loss": 0.1498,
+      "step": 2126
+    },
+    {
+      "epoch": 0.3446208684381076,
+      "grad_norm": 1.0807654857635498,
+      "learning_rate": 4.906197432942093e-06,
+      "loss": 0.1492,
+      "step": 2127
+    },
+    {
+      "epoch": 0.34478289047310434,
+      "grad_norm": 1.2625532150268555,
+      "learning_rate": 4.9060787309206436e-06,
+      "loss": 0.1784,
+      "step": 2128
+    },
+    {
+      "epoch": 0.3449449125081011,
+      "grad_norm": 1.1583614349365234,
+      "learning_rate": 4.905959955279007e-06,
+      "loss": 0.1652,
+      "step": 2129
+    },
+    {
+      "epoch": 0.3451069345430979,
+      "grad_norm": 1.1260557174682617,
+      "learning_rate": 4.905841106020818e-06,
+      "loss": 0.1649,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34526895657809464,
+      "grad_norm": 1.1886639595031738,
+      "learning_rate": 4.905722183149714e-06,
+      "loss": 0.1724,
+      "step": 2131
+    },
+    {
+      "epoch": 0.34543097861309136,
+      "grad_norm": 1.138461709022522,
+      "learning_rate": 4.905603186669332e-06,
+      "loss": 0.177,
+      "step": 2132
+    },
+    {
+      "epoch": 0.3455930006480881,
+      "grad_norm": 1.1662522554397583,
+      "learning_rate": 4.905484116583314e-06,
+      "loss": 0.1436,
+      "step": 2133
+    },
+    {
+      "epoch": 0.3457550226830849,
+      "grad_norm": 1.0118292570114136,
+      "learning_rate": 4.905364972895304e-06,
+      "loss": 0.1573,
+      "step": 2134
+    },
+    {
+      "epoch": 0.34591704471808166,
+      "grad_norm": 0.9873135685920715,
+      "learning_rate": 4.905245755608946e-06,
+      "loss": 0.1484,
+      "step": 2135
+    },
+    {
+      "epoch": 0.3460790667530784,
+      "grad_norm": 1.1140849590301514,
+      "learning_rate": 4.9051264647278886e-06,
+      "loss": 0.1773,
+      "step": 2136
+    },
+    {
+      "epoch": 0.3462410887880752,
+      "grad_norm": 1.053364872932434,
+      "learning_rate": 4.9050071002557815e-06,
+      "loss": 0.1462,
+      "step": 2137
+    },
+    {
+      "epoch": 0.34640311082307196,
+      "grad_norm": 1.0225468873977661,
+      "learning_rate": 4.904887662196277e-06,
+      "loss": 0.165,
+      "step": 2138
+    },
+    {
+      "epoch": 0.34656513285806867,
+      "grad_norm": 1.1107808351516724,
+      "learning_rate": 4.90476815055303e-06,
+      "loss": 0.1427,
+      "step": 2139
+    },
+    {
+      "epoch": 0.34672715489306544,
+      "grad_norm": 1.0493844747543335,
+      "learning_rate": 4.904648565329697e-06,
+      "loss": 0.1416,
+      "step": 2140
+    },
+    {
+      "epoch": 0.3468891769280622,
+      "grad_norm": 1.1383299827575684,
+      "learning_rate": 4.9045289065299375e-06,
+      "loss": 0.1915,
+      "step": 2141
+    },
+    {
+      "epoch": 0.34705119896305897,
+      "grad_norm": 1.1016839742660522,
+      "learning_rate": 4.904409174157412e-06,
+      "loss": 0.1629,
+      "step": 2142
+    },
+    {
+      "epoch": 0.34721322099805574,
+      "grad_norm": 1.0556694269180298,
+      "learning_rate": 4.904289368215784e-06,
+      "loss": 0.17,
+      "step": 2143
+    },
+    {
+      "epoch": 0.3473752430330525,
+      "grad_norm": 1.0352901220321655,
+      "learning_rate": 4.90416948870872e-06,
+      "loss": 0.149,
+      "step": 2144
+    },
+    {
+      "epoch": 0.3475372650680493,
+      "grad_norm": 1.0276658535003662,
+      "learning_rate": 4.9040495356398874e-06,
+      "loss": 0.1466,
+      "step": 2145
+    },
+    {
+      "epoch": 0.34769928710304604,
+      "grad_norm": 1.1721683740615845,
+      "learning_rate": 4.903929509012957e-06,
+      "loss": 0.1652,
+      "step": 2146
+    },
+    {
+      "epoch": 0.34786130913804275,
+      "grad_norm": 1.0136381387710571,
+      "learning_rate": 4.903809408831601e-06,
+      "loss": 0.1425,
+      "step": 2147
+    },
+    {
+      "epoch": 0.3480233311730395,
+      "grad_norm": 1.0642050504684448,
+      "learning_rate": 4.9036892350994935e-06,
+      "loss": 0.1454,
+      "step": 2148
+    },
+    {
+      "epoch": 0.3481853532080363,
+      "grad_norm": 1.0709917545318604,
+      "learning_rate": 4.903568987820313e-06,
+      "loss": 0.1478,
+      "step": 2149
+    },
+    {
+      "epoch": 0.34834737524303305,
+      "grad_norm": 1.0638947486877441,
+      "learning_rate": 4.9034486669977375e-06,
+      "loss": 0.1625,
+      "step": 2150
+    },
+    {
+      "epoch": 0.3485093972780298,
+      "grad_norm": 1.0096039772033691,
+      "learning_rate": 4.903328272635449e-06,
+      "loss": 0.1462,
+      "step": 2151
+    },
+    {
+      "epoch": 0.3486714193130266,
+      "grad_norm": 1.1130759716033936,
+      "learning_rate": 4.903207804737132e-06,
+      "loss": 0.1488,
+      "step": 2152
+    },
+    {
+      "epoch": 0.34883344134802335,
+      "grad_norm": 1.105833649635315,
+      "learning_rate": 4.9030872633064715e-06,
+      "loss": 0.1669,
+      "step": 2153
+    },
+    {
+      "epoch": 0.34899546338302007,
+      "grad_norm": 1.2638037204742432,
+      "learning_rate": 4.902966648347156e-06,
+      "loss": 0.1613,
+      "step": 2154
+    },
+    {
+      "epoch": 0.34915748541801683,
+      "grad_norm": 0.990151584148407,
+      "learning_rate": 4.902845959862876e-06,
+      "loss": 0.1456,
+      "step": 2155
+    },
+    {
+      "epoch": 0.3493195074530136,
+      "grad_norm": 1.108185887336731,
+      "learning_rate": 4.902725197857325e-06,
+      "loss": 0.1522,
+      "step": 2156
+    },
+    {
+      "epoch": 0.34948152948801037,
+      "grad_norm": 0.9440566897392273,
+      "learning_rate": 4.902604362334197e-06,
+      "loss": 0.1266,
+      "step": 2157
+    },
+    {
+      "epoch": 0.34964355152300713,
+      "grad_norm": 1.2154903411865234,
+      "learning_rate": 4.902483453297189e-06,
+      "loss": 0.1911,
+      "step": 2158
+    },
+    {
+      "epoch": 0.3498055735580039,
+      "grad_norm": 1.0809717178344727,
+      "learning_rate": 4.902362470750002e-06,
+      "loss": 0.1564,
+      "step": 2159
+    },
+    {
+      "epoch": 0.34996759559300067,
+      "grad_norm": 1.0994709730148315,
+      "learning_rate": 4.902241414696337e-06,
+      "loss": 0.1707,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3501296176279974,
+      "grad_norm": 1.0126726627349854,
+      "learning_rate": 4.902120285139898e-06,
+      "loss": 0.1451,
+      "step": 2161
+    },
+    {
+      "epoch": 0.35029163966299415,
+      "grad_norm": 1.068403720855713,
+      "learning_rate": 4.901999082084391e-06,
+      "loss": 0.1578,
+      "step": 2162
+    },
+    {
+      "epoch": 0.3504536616979909,
+      "grad_norm": 0.9774585962295532,
+      "learning_rate": 4.901877805533525e-06,
+      "loss": 0.1406,
+      "step": 2163
+    },
+    {
+      "epoch": 0.3506156837329877,
+      "grad_norm": 1.0109790563583374,
+      "learning_rate": 4.901756455491011e-06,
+      "loss": 0.1425,
+      "step": 2164
+    },
+    {
+      "epoch": 0.35077770576798445,
+      "grad_norm": 1.182918667793274,
+      "learning_rate": 4.901635031960561e-06,
+      "loss": 0.1952,
+      "step": 2165
+    },
+    {
+      "epoch": 0.3509397278029812,
+      "grad_norm": 0.9629144668579102,
+      "learning_rate": 4.901513534945891e-06,
+      "loss": 0.1487,
+      "step": 2166
+    },
+    {
+      "epoch": 0.351101749837978,
+      "grad_norm": 1.0952849388122559,
+      "learning_rate": 4.901391964450718e-06,
+      "loss": 0.171,
+      "step": 2167
+    },
+    {
+      "epoch": 0.35126377187297475,
+      "grad_norm": 1.0250297784805298,
+      "learning_rate": 4.901270320478763e-06,
+      "loss": 0.1532,
+      "step": 2168
+    },
+    {
+      "epoch": 0.35142579390797146,
+      "grad_norm": 1.1184073686599731,
+      "learning_rate": 4.901148603033747e-06,
+      "loss": 0.1641,
+      "step": 2169
+    },
+    {
+      "epoch": 0.35158781594296823,
+      "grad_norm": 1.0301399230957031,
+      "learning_rate": 4.901026812119394e-06,
+      "loss": 0.1515,
+      "step": 2170
+    },
+    {
+      "epoch": 0.351749837977965,
+      "grad_norm": 1.188010573387146,
+      "learning_rate": 4.900904947739431e-06,
+      "loss": 0.1565,
+      "step": 2171
+    },
+    {
+      "epoch": 0.35191186001296176,
+      "grad_norm": 1.0016369819641113,
+      "learning_rate": 4.9007830098975875e-06,
+      "loss": 0.1523,
+      "step": 2172
+    },
+    {
+      "epoch": 0.35207388204795853,
+      "grad_norm": 1.049594521522522,
+      "learning_rate": 4.9006609985975925e-06,
+      "loss": 0.1426,
+      "step": 2173
+    },
+    {
+      "epoch": 0.3522359040829553,
+      "grad_norm": 1.197577714920044,
+      "learning_rate": 4.900538913843181e-06,
+      "loss": 0.1812,
+      "step": 2174
+    },
+    {
+      "epoch": 0.35239792611795207,
+      "grad_norm": 1.06314218044281,
+      "learning_rate": 4.900416755638087e-06,
+      "loss": 0.1601,
+      "step": 2175
+    },
+    {
+      "epoch": 0.3525599481529488,
+      "grad_norm": 1.0126889944076538,
+      "learning_rate": 4.900294523986051e-06,
+      "loss": 0.1472,
+      "step": 2176
+    },
+    {
+      "epoch": 0.35272197018794554,
+      "grad_norm": 1.1102066040039062,
+      "learning_rate": 4.900172218890809e-06,
+      "loss": 0.1699,
+      "step": 2177
+    },
+    {
+      "epoch": 0.3528839922229423,
+      "grad_norm": 1.1450811624526978,
+      "learning_rate": 4.900049840356107e-06,
+      "loss": 0.1804,
+      "step": 2178
+    },
+    {
+      "epoch": 0.3530460142579391,
+      "grad_norm": 1.1297409534454346,
+      "learning_rate": 4.899927388385688e-06,
+      "loss": 0.1678,
+      "step": 2179
+    },
+    {
+      "epoch": 0.35320803629293585,
+      "grad_norm": 1.0835684537887573,
+      "learning_rate": 4.899804862983298e-06,
+      "loss": 0.173,
+      "step": 2180
+    },
+    {
+      "epoch": 0.3533700583279326,
+      "grad_norm": 1.0187311172485352,
+      "learning_rate": 4.899682264152686e-06,
+      "loss": 0.1353,
+      "step": 2181
+    },
+    {
+      "epoch": 0.3535320803629294,
+      "grad_norm": 0.9485295414924622,
+      "learning_rate": 4.899559591897604e-06,
+      "loss": 0.1516,
+      "step": 2182
+    },
+    {
+      "epoch": 0.3536941023979261,
+      "grad_norm": 0.9085932970046997,
+      "learning_rate": 4.899436846221807e-06,
+      "loss": 0.148,
+      "step": 2183
+    },
+    {
+      "epoch": 0.35385612443292286,
+      "grad_norm": 1.1273099184036255,
+      "learning_rate": 4.899314027129047e-06,
+      "loss": 0.1685,
+      "step": 2184
+    },
+    {
+      "epoch": 0.3540181464679196,
+      "grad_norm": 1.0366731882095337,
+      "learning_rate": 4.899191134623086e-06,
+      "loss": 0.1542,
+      "step": 2185
+    },
+    {
+      "epoch": 0.3541801685029164,
+      "grad_norm": 1.049047589302063,
+      "learning_rate": 4.89906816870768e-06,
+      "loss": 0.172,
+      "step": 2186
+    },
+    {
+      "epoch": 0.35434219053791316,
+      "grad_norm": 1.1898120641708374,
+      "learning_rate": 4.8989451293865955e-06,
+      "loss": 0.1729,
+      "step": 2187
+    },
+    {
+      "epoch": 0.3545042125729099,
+      "grad_norm": 1.0170189142227173,
+      "learning_rate": 4.898822016663595e-06,
+      "loss": 0.1539,
+      "step": 2188
+    },
+    {
+      "epoch": 0.3546662346079067,
+      "grad_norm": 0.9323112368583679,
+      "learning_rate": 4.8986988305424445e-06,
+      "loss": 0.1392,
+      "step": 2189
+    },
+    {
+      "epoch": 0.35482825664290346,
+      "grad_norm": 1.097751259803772,
+      "learning_rate": 4.898575571026916e-06,
+      "loss": 0.1697,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3549902786779002,
+      "grad_norm": 1.2580288648605347,
+      "learning_rate": 4.898452238120779e-06,
+      "loss": 0.1925,
+      "step": 2191
+    },
+    {
+      "epoch": 0.35515230071289694,
+      "grad_norm": 1.1584991216659546,
+      "learning_rate": 4.898328831827808e-06,
+      "loss": 0.1696,
+      "step": 2192
+    },
+    {
+      "epoch": 0.3553143227478937,
+      "grad_norm": 1.0582317113876343,
+      "learning_rate": 4.898205352151777e-06,
+      "loss": 0.1498,
+      "step": 2193
+    },
+    {
+      "epoch": 0.3554763447828905,
+      "grad_norm": 1.039017915725708,
+      "learning_rate": 4.898081799096467e-06,
+      "loss": 0.1458,
+      "step": 2194
+    },
+    {
+      "epoch": 0.35563836681788724,
+      "grad_norm": 1.237728476524353,
+      "learning_rate": 4.897958172665658e-06,
+      "loss": 0.1588,
+      "step": 2195
+    },
+    {
+      "epoch": 0.355800388852884,
+      "grad_norm": 0.9922459125518799,
+      "learning_rate": 4.897834472863131e-06,
+      "loss": 0.146,
+      "step": 2196
+    },
+    {
+      "epoch": 0.3559624108878808,
+      "grad_norm": 1.1066429615020752,
+      "learning_rate": 4.897710699692672e-06,
+      "loss": 0.1492,
+      "step": 2197
+    },
+    {
+      "epoch": 0.3561244329228775,
+      "grad_norm": 1.1246432065963745,
+      "learning_rate": 4.897586853158067e-06,
+      "loss": 0.1644,
+      "step": 2198
+    },
+    {
+      "epoch": 0.35628645495787425,
+      "grad_norm": 1.1196368932724,
+      "learning_rate": 4.897462933263107e-06,
+      "loss": 0.1517,
+      "step": 2199
+    },
+    {
+      "epoch": 0.356448476992871,
+      "grad_norm": 1.1763758659362793,
+      "learning_rate": 4.897338940011583e-06,
+      "loss": 0.169,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3566104990278678,
+      "grad_norm": 1.0583901405334473,
+      "learning_rate": 4.897214873407289e-06,
+      "loss": 0.1462,
+      "step": 2201
+    },
+    {
+      "epoch": 0.35677252106286456,
+      "grad_norm": 0.9856936931610107,
+      "learning_rate": 4.897090733454021e-06,
+      "loss": 0.1408,
+      "step": 2202
+    },
+    {
+      "epoch": 0.3569345430978613,
+      "grad_norm": 1.0380568504333496,
+      "learning_rate": 4.8969665201555775e-06,
+      "loss": 0.1692,
+      "step": 2203
+    },
+    {
+      "epoch": 0.3570965651328581,
+      "grad_norm": 1.0581004619598389,
+      "learning_rate": 4.896842233515759e-06,
+      "loss": 0.1714,
+      "step": 2204
+    },
+    {
+      "epoch": 0.3572585871678548,
+      "grad_norm": 1.2082271575927734,
+      "learning_rate": 4.896717873538368e-06,
+      "loss": 0.1791,
+      "step": 2205
+    },
+    {
+      "epoch": 0.35742060920285157,
+      "grad_norm": 1.0210479497909546,
+      "learning_rate": 4.89659344022721e-06,
+      "loss": 0.1652,
+      "step": 2206
+    },
+    {
+      "epoch": 0.35758263123784834,
+      "grad_norm": 0.9992541074752808,
+      "learning_rate": 4.896468933586094e-06,
+      "loss": 0.162,
+      "step": 2207
+    },
+    {
+      "epoch": 0.3577446532728451,
+      "grad_norm": 1.0144470930099487,
+      "learning_rate": 4.896344353618826e-06,
+      "loss": 0.1597,
+      "step": 2208
+    },
+    {
+      "epoch": 0.35790667530784187,
+      "grad_norm": 1.1150000095367432,
+      "learning_rate": 4.89621970032922e-06,
+      "loss": 0.167,
+      "step": 2209
+    },
+    {
+      "epoch": 0.35806869734283864,
+      "grad_norm": 1.0578536987304688,
+      "learning_rate": 4.896094973721091e-06,
+      "loss": 0.1402,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3582307193778354,
+      "grad_norm": 0.9122249484062195,
+      "learning_rate": 4.8959701737982535e-06,
+      "loss": 0.1362,
+      "step": 2211
+    },
+    {
+      "epoch": 0.35839274141283217,
+      "grad_norm": 1.0043046474456787,
+      "learning_rate": 4.8958453005645265e-06,
+      "loss": 0.1537,
+      "step": 2212
+    },
+    {
+      "epoch": 0.3585547634478289,
+      "grad_norm": 1.1452409029006958,
+      "learning_rate": 4.895720354023732e-06,
+      "loss": 0.1239,
+      "step": 2213
+    },
+    {
+      "epoch": 0.35871678548282565,
+      "grad_norm": 1.0220675468444824,
+      "learning_rate": 4.895595334179692e-06,
+      "loss": 0.1461,
+      "step": 2214
+    },
+    {
+      "epoch": 0.3588788075178224,
+      "grad_norm": 1.3352243900299072,
+      "learning_rate": 4.895470241036232e-06,
+      "loss": 0.2061,
+      "step": 2215
+    },
+    {
+      "epoch": 0.3590408295528192,
+      "grad_norm": 1.1318044662475586,
+      "learning_rate": 4.89534507459718e-06,
+      "loss": 0.1479,
+      "step": 2216
+    },
+    {
+      "epoch": 0.35920285158781595,
+      "grad_norm": 1.2025889158248901,
+      "learning_rate": 4.895219834866364e-06,
+      "loss": 0.1721,
+      "step": 2217
+    },
+    {
+      "epoch": 0.3593648736228127,
+      "grad_norm": 1.0799037218093872,
+      "learning_rate": 4.895094521847617e-06,
+      "loss": 0.1504,
+      "step": 2218
+    },
+    {
+      "epoch": 0.3595268956578095,
+      "grad_norm": 0.9984576106071472,
+      "learning_rate": 4.894969135544776e-06,
+      "loss": 0.1485,
+      "step": 2219
+    },
+    {
+      "epoch": 0.3596889176928062,
+      "grad_norm": 1.1945985555648804,
+      "learning_rate": 4.894843675961673e-06,
+      "loss": 0.1784,
+      "step": 2220
+    },
+    {
+      "epoch": 0.35985093972780297,
+      "grad_norm": 0.962485671043396,
+      "learning_rate": 4.894718143102151e-06,
+      "loss": 0.1345,
+      "step": 2221
+    },
+    {
+      "epoch": 0.36001296176279973,
+      "grad_norm": 0.9608977437019348,
+      "learning_rate": 4.894592536970047e-06,
+      "loss": 0.144,
+      "step": 2222
+    },
+    {
+      "epoch": 0.3601749837977965,
+      "grad_norm": 1.0343858003616333,
+      "learning_rate": 4.894466857569207e-06,
+      "loss": 0.1507,
+      "step": 2223
+    },
+    {
+      "epoch": 0.36033700583279327,
+      "grad_norm": 1.0216684341430664,
+      "learning_rate": 4.894341104903476e-06,
+      "loss": 0.1354,
+      "step": 2224
+    },
+    {
+      "epoch": 0.36049902786779003,
+      "grad_norm": 1.18091881275177,
+      "learning_rate": 4.8942152789767e-06,
+      "loss": 0.1631,
+      "step": 2225
+    },
+    {
+      "epoch": 0.3606610499027868,
+      "grad_norm": 1.0483477115631104,
+      "learning_rate": 4.894089379792731e-06,
+      "loss": 0.1607,
+      "step": 2226
+    },
+    {
+      "epoch": 0.3608230719377835,
+      "grad_norm": 1.1260876655578613,
+      "learning_rate": 4.893963407355422e-06,
+      "loss": 0.1611,
+      "step": 2227
+    },
+    {
+      "epoch": 0.3609850939727803,
+      "grad_norm": 0.9986515045166016,
+      "learning_rate": 4.893837361668624e-06,
+      "loss": 0.1565,
+      "step": 2228
+    },
+    {
+      "epoch": 0.36114711600777705,
+      "grad_norm": 1.1207728385925293,
+      "learning_rate": 4.893711242736197e-06,
+      "loss": 0.1611,
+      "step": 2229
+    },
+    {
+      "epoch": 0.3613091380427738,
+      "grad_norm": 1.1995974779129028,
+      "learning_rate": 4.8935850505619985e-06,
+      "loss": 0.1705,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3614711600777706,
+      "grad_norm": 1.1663684844970703,
+      "learning_rate": 4.893458785149889e-06,
+      "loss": 0.1841,
+      "step": 2231
+    },
+    {
+      "epoch": 0.36163318211276735,
+      "grad_norm": 0.9084141850471497,
+      "learning_rate": 4.8933324465037334e-06,
+      "loss": 0.1425,
+      "step": 2232
+    },
+    {
+      "epoch": 0.3617952041477641,
+      "grad_norm": 0.9766511917114258,
+      "learning_rate": 4.893206034627397e-06,
+      "loss": 0.1585,
+      "step": 2233
+    },
+    {
+      "epoch": 0.3619572261827609,
+      "grad_norm": 1.0167397260665894,
+      "learning_rate": 4.893079549524747e-06,
+      "loss": 0.167,
+      "step": 2234
+    },
+    {
+      "epoch": 0.3621192482177576,
+      "grad_norm": 1.0378706455230713,
+      "learning_rate": 4.892952991199654e-06,
+      "loss": 0.1502,
+      "step": 2235
+    },
+    {
+      "epoch": 0.36228127025275436,
+      "grad_norm": 0.9949929118156433,
+      "learning_rate": 4.89282635965599e-06,
+      "loss": 0.1482,
+      "step": 2236
+    },
+    {
+      "epoch": 0.36244329228775113,
+      "grad_norm": 1.006990909576416,
+      "learning_rate": 4.89269965489763e-06,
+      "loss": 0.1504,
+      "step": 2237
+    },
+    {
+      "epoch": 0.3626053143227479,
+      "grad_norm": 0.947759747505188,
+      "learning_rate": 4.8925728769284504e-06,
+      "loss": 0.1436,
+      "step": 2238
+    },
+    {
+      "epoch": 0.36276733635774466,
+      "grad_norm": 0.9912741184234619,
+      "learning_rate": 4.892446025752332e-06,
+      "loss": 0.1401,
+      "step": 2239
+    },
+    {
+      "epoch": 0.36292935839274143,
+      "grad_norm": 0.9576963782310486,
+      "learning_rate": 4.892319101373154e-06,
+      "loss": 0.1406,
+      "step": 2240
+    },
+    {
+      "epoch": 0.3630913804277382,
+      "grad_norm": 1.0038036108016968,
+      "learning_rate": 4.892192103794801e-06,
+      "loss": 0.1418,
+      "step": 2241
+    },
+    {
+      "epoch": 0.3632534024627349,
+      "grad_norm": 1.1145362854003906,
+      "learning_rate": 4.892065033021158e-06,
+      "loss": 0.1654,
+      "step": 2242
+    },
+    {
+      "epoch": 0.3634154244977317,
+      "grad_norm": 1.185004472732544,
+      "learning_rate": 4.8919378890561145e-06,
+      "loss": 0.1686,
+      "step": 2243
+    },
+    {
+      "epoch": 0.36357744653272844,
+      "grad_norm": 1.0970295667648315,
+      "learning_rate": 4.8918106719035594e-06,
+      "loss": 0.1504,
+      "step": 2244
+    },
+    {
+      "epoch": 0.3637394685677252,
+      "grad_norm": 1.0193977355957031,
+      "learning_rate": 4.891683381567386e-06,
+      "loss": 0.1405,
+      "step": 2245
+    },
+    {
+      "epoch": 0.363901490602722,
+      "grad_norm": 0.9931899309158325,
+      "learning_rate": 4.891556018051489e-06,
+      "loss": 0.1569,
+      "step": 2246
+    },
+    {
+      "epoch": 0.36406351263771874,
+      "grad_norm": 1.1715176105499268,
+      "learning_rate": 4.891428581359764e-06,
+      "loss": 0.1799,
+      "step": 2247
+    },
+    {
+      "epoch": 0.3642255346727155,
+      "grad_norm": 1.0806795358657837,
+      "learning_rate": 4.891301071496113e-06,
+      "loss": 0.1547,
+      "step": 2248
+    },
+    {
+      "epoch": 0.3643875567077122,
+      "grad_norm": 1.0172020196914673,
+      "learning_rate": 4.891173488464436e-06,
+      "loss": 0.1409,
+      "step": 2249
+    },
+    {
+      "epoch": 0.364549578742709,
+      "grad_norm": 0.9692556858062744,
+      "learning_rate": 4.891045832268637e-06,
+      "loss": 0.1519,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36471160077770576,
+      "grad_norm": 1.108995795249939,
+      "learning_rate": 4.890918102912621e-06,
+      "loss": 0.1695,
+      "step": 2251
+    },
+    {
+      "epoch": 0.3648736228127025,
+      "grad_norm": 1.081821322441101,
+      "learning_rate": 4.890790300400297e-06,
+      "loss": 0.1719,
+      "step": 2252
+    },
+    {
+      "epoch": 0.3650356448476993,
+      "grad_norm": 1.0040013790130615,
+      "learning_rate": 4.890662424735576e-06,
+      "loss": 0.1603,
+      "step": 2253
+    },
+    {
+      "epoch": 0.36519766688269606,
+      "grad_norm": 1.0947988033294678,
+      "learning_rate": 4.8905344759223696e-06,
+      "loss": 0.1791,
+      "step": 2254
+    },
+    {
+      "epoch": 0.3653596889176928,
+      "grad_norm": 1.1201766729354858,
+      "learning_rate": 4.890406453964594e-06,
+      "loss": 0.1646,
+      "step": 2255
+    },
+    {
+      "epoch": 0.36552171095268954,
+      "grad_norm": 1.0785022974014282,
+      "learning_rate": 4.890278358866165e-06,
+      "loss": 0.1617,
+      "step": 2256
+    },
+    {
+      "epoch": 0.3656837329876863,
+      "grad_norm": 1.1495450735092163,
+      "learning_rate": 4.8901501906310024e-06,
+      "loss": 0.1674,
+      "step": 2257
+    },
+    {
+      "epoch": 0.3658457550226831,
+      "grad_norm": 1.0032192468643188,
+      "learning_rate": 4.890021949263027e-06,
+      "loss": 0.1427,
+      "step": 2258
+    },
+    {
+      "epoch": 0.36600777705767984,
+      "grad_norm": 1.148429274559021,
+      "learning_rate": 4.889893634766165e-06,
+      "loss": 0.1654,
+      "step": 2259
+    },
+    {
+      "epoch": 0.3661697990926766,
+      "grad_norm": 1.0179355144500732,
+      "learning_rate": 4.889765247144341e-06,
+      "loss": 0.1496,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3663318211276734,
+      "grad_norm": 0.9481959939002991,
+      "learning_rate": 4.889636786401484e-06,
+      "loss": 0.1423,
+      "step": 2261
+    },
+    {
+      "epoch": 0.36649384316267014,
+      "grad_norm": 1.1533952951431274,
+      "learning_rate": 4.889508252541524e-06,
+      "loss": 0.1582,
+      "step": 2262
+    },
+    {
+      "epoch": 0.3666558651976669,
+      "grad_norm": 1.0846741199493408,
+      "learning_rate": 4.889379645568394e-06,
+      "loss": 0.1599,
+      "step": 2263
+    },
+    {
+      "epoch": 0.3668178872326636,
+      "grad_norm": 1.2179301977157593,
+      "learning_rate": 4.889250965486029e-06,
+      "loss": 0.1812,
+      "step": 2264
+    },
+    {
+      "epoch": 0.3669799092676604,
+      "grad_norm": 0.9992818832397461,
+      "learning_rate": 4.889122212298366e-06,
+      "loss": 0.138,
+      "step": 2265
+    },
+    {
+      "epoch": 0.36714193130265715,
+      "grad_norm": 1.1005460023880005,
+      "learning_rate": 4.888993386009345e-06,
+      "loss": 0.1519,
+      "step": 2266
+    },
+    {
+      "epoch": 0.3673039533376539,
+      "grad_norm": 1.0345803499221802,
+      "learning_rate": 4.888864486622907e-06,
+      "loss": 0.1539,
+      "step": 2267
+    },
+    {
+      "epoch": 0.3674659753726507,
+      "grad_norm": 1.0553728342056274,
+      "learning_rate": 4.888735514142998e-06,
+      "loss": 0.1459,
+      "step": 2268
+    },
+    {
+      "epoch": 0.36762799740764746,
+      "grad_norm": 1.0219756364822388,
+      "learning_rate": 4.888606468573562e-06,
+      "loss": 0.1459,
+      "step": 2269
+    },
+    {
+      "epoch": 0.3677900194426442,
+      "grad_norm": 1.0199123620986938,
+      "learning_rate": 4.8884773499185485e-06,
+      "loss": 0.1466,
+      "step": 2270
+    },
+    {
+      "epoch": 0.36795204147764093,
+      "grad_norm": 0.9618728160858154,
+      "learning_rate": 4.888348158181908e-06,
+      "loss": 0.1478,
+      "step": 2271
+    },
+    {
+      "epoch": 0.3681140635126377,
+      "grad_norm": 1.018974781036377,
+      "learning_rate": 4.8882188933675935e-06,
+      "loss": 0.1493,
+      "step": 2272
+    },
+    {
+      "epoch": 0.36827608554763447,
+      "grad_norm": 1.1692665815353394,
+      "learning_rate": 4.888089555479561e-06,
+      "loss": 0.1571,
+      "step": 2273
+    },
+    {
+      "epoch": 0.36843810758263124,
+      "grad_norm": 1.0094574689865112,
+      "learning_rate": 4.887960144521766e-06,
+      "loss": 0.1413,
+      "step": 2274
+    },
+    {
+      "epoch": 0.368600129617628,
+      "grad_norm": 1.0901294946670532,
+      "learning_rate": 4.8878306604981704e-06,
+      "loss": 0.1495,
+      "step": 2275
+    },
+    {
+      "epoch": 0.36876215165262477,
+      "grad_norm": 1.043537974357605,
+      "learning_rate": 4.887701103412734e-06,
+      "loss": 0.1641,
+      "step": 2276
+    },
+    {
+      "epoch": 0.36892417368762154,
+      "grad_norm": 1.0010807514190674,
+      "learning_rate": 4.887571473269422e-06,
+      "loss": 0.1562,
+      "step": 2277
+    },
+    {
+      "epoch": 0.36908619572261825,
+      "grad_norm": 0.9946657419204712,
+      "learning_rate": 4.8874417700722025e-06,
+      "loss": 0.1406,
+      "step": 2278
+    },
+    {
+      "epoch": 0.369248217757615,
+      "grad_norm": 1.007975697517395,
+      "learning_rate": 4.887311993825041e-06,
+      "loss": 0.1448,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3694102397926118,
+      "grad_norm": 1.072035312652588,
+      "learning_rate": 4.887182144531909e-06,
+      "loss": 0.1686,
+      "step": 2280
+    },
+    {
+      "epoch": 0.36957226182760855,
+      "grad_norm": 2.578970432281494,
+      "learning_rate": 4.887052222196782e-06,
+      "loss": 0.1594,
+      "step": 2281
+    },
+    {
+      "epoch": 0.3697342838626053,
+      "grad_norm": 0.9519162774085999,
+      "learning_rate": 4.886922226823632e-06,
+      "loss": 0.1369,
+      "step": 2282
+    },
+    {
+      "epoch": 0.3698963058976021,
+      "grad_norm": 1.0015640258789062,
+      "learning_rate": 4.886792158416439e-06,
+      "loss": 0.1469,
+      "step": 2283
+    },
+    {
+      "epoch": 0.37005832793259885,
+      "grad_norm": 0.9746085405349731,
+      "learning_rate": 4.8866620169791815e-06,
+      "loss": 0.1342,
+      "step": 2284
+    },
+    {
+      "epoch": 0.3702203499675956,
+      "grad_norm": 1.0695668458938599,
+      "learning_rate": 4.886531802515842e-06,
+      "loss": 0.1645,
+      "step": 2285
+    },
+    {
+      "epoch": 0.37038237200259233,
+      "grad_norm": 1.1742403507232666,
+      "learning_rate": 4.886401515030404e-06,
+      "loss": 0.1668,
+      "step": 2286
+    },
+    {
+      "epoch": 0.3705443940375891,
+      "grad_norm": 1.0946766138076782,
+      "learning_rate": 4.886271154526856e-06,
+      "loss": 0.1486,
+      "step": 2287
+    },
+    {
+      "epoch": 0.37070641607258586,
+      "grad_norm": 1.0523529052734375,
+      "learning_rate": 4.886140721009184e-06,
+      "loss": 0.1458,
+      "step": 2288
+    },
+    {
+      "epoch": 0.37086843810758263,
+      "grad_norm": 1.1310120820999146,
+      "learning_rate": 4.88601021448138e-06,
+      "loss": 0.1568,
+      "step": 2289
+    },
+    {
+      "epoch": 0.3710304601425794,
+      "grad_norm": 1.176133155822754,
+      "learning_rate": 4.885879634947439e-06,
+      "loss": 0.1723,
+      "step": 2290
+    },
+    {
+      "epoch": 0.37119248217757617,
+      "grad_norm": 1.064237117767334,
+      "learning_rate": 4.8857489824113544e-06,
+      "loss": 0.1507,
+      "step": 2291
+    },
+    {
+      "epoch": 0.37135450421257293,
+      "grad_norm": 1.0737059116363525,
+      "learning_rate": 4.885618256877123e-06,
+      "loss": 0.1391,
+      "step": 2292
+    },
+    {
+      "epoch": 0.37151652624756965,
+      "grad_norm": 1.0706037282943726,
+      "learning_rate": 4.885487458348748e-06,
+      "loss": 0.1442,
+      "step": 2293
+    },
+    {
+      "epoch": 0.3716785482825664,
+      "grad_norm": 1.3316296339035034,
+      "learning_rate": 4.885356586830229e-06,
+      "loss": 0.1965,
+      "step": 2294
+    },
+    {
+      "epoch": 0.3718405703175632,
+      "grad_norm": 0.9738705158233643,
+      "learning_rate": 4.8852256423255706e-06,
+      "loss": 0.1412,
+      "step": 2295
+    },
+    {
+      "epoch": 0.37200259235255995,
+      "grad_norm": 0.9631845355033875,
+      "learning_rate": 4.8850946248387795e-06,
+      "loss": 0.1446,
+      "step": 2296
+    },
+    {
+      "epoch": 0.3721646143875567,
+      "grad_norm": 1.002110481262207,
+      "learning_rate": 4.884963534373864e-06,
+      "loss": 0.1489,
+      "step": 2297
+    },
+    {
+      "epoch": 0.3723266364225535,
+      "grad_norm": 0.8810906410217285,
+      "learning_rate": 4.884832370934838e-06,
+      "loss": 0.1249,
+      "step": 2298
+    },
+    {
+      "epoch": 0.37248865845755025,
+      "grad_norm": 1.2521817684173584,
+      "learning_rate": 4.88470113452571e-06,
+      "loss": 0.155,
+      "step": 2299
+    },
+    {
+      "epoch": 0.37265068049254696,
+      "grad_norm": 1.0234705209732056,
+      "learning_rate": 4.8845698251505e-06,
+      "loss": 0.1481,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3728127025275437,
+      "grad_norm": 1.0289057493209839,
+      "learning_rate": 4.884438442813223e-06,
+      "loss": 0.1226,
+      "step": 2301
+    },
+    {
+      "epoch": 0.3729747245625405,
+      "grad_norm": 1.0109435319900513,
+      "learning_rate": 4.8843069875179005e-06,
+      "loss": 0.1527,
+      "step": 2302
+    },
+    {
+      "epoch": 0.37313674659753726,
+      "grad_norm": 1.0381628274917603,
+      "learning_rate": 4.884175459268554e-06,
+      "loss": 0.1689,
+      "step": 2303
+    },
+    {
+      "epoch": 0.37329876863253403,
+      "grad_norm": 1.0785562992095947,
+      "learning_rate": 4.884043858069208e-06,
+      "loss": 0.1507,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3734607906675308,
+      "grad_norm": 1.0197018384933472,
+      "learning_rate": 4.883912183923889e-06,
+      "loss": 0.1564,
+      "step": 2305
+    },
+    {
+      "epoch": 0.37362281270252756,
+      "grad_norm": 1.1440730094909668,
+      "learning_rate": 4.883780436836627e-06,
+      "loss": 0.1594,
+      "step": 2306
+    },
+    {
+      "epoch": 0.37378483473752433,
+      "grad_norm": 1.0093541145324707,
+      "learning_rate": 4.883648616811451e-06,
+      "loss": 0.1565,
+      "step": 2307
+    },
+    {
+      "epoch": 0.37394685677252104,
+      "grad_norm": 1.0944989919662476,
+      "learning_rate": 4.883516723852396e-06,
+      "loss": 0.1549,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3741088788075178,
+      "grad_norm": 1.0716201066970825,
+      "learning_rate": 4.883384757963498e-06,
+      "loss": 0.1666,
+      "step": 2309
+    },
+    {
+      "epoch": 0.3742709008425146,
+      "grad_norm": 0.946195125579834,
+      "learning_rate": 4.883252719148794e-06,
+      "loss": 0.1422,
+      "step": 2310
+    },
+    {
+      "epoch": 0.37443292287751134,
+      "grad_norm": 0.9485779404640198,
+      "learning_rate": 4.883120607412323e-06,
+      "loss": 0.1536,
+      "step": 2311
+    },
+    {
+      "epoch": 0.3745949449125081,
+      "grad_norm": 1.0378730297088623,
+      "learning_rate": 4.8829884227581294e-06,
+      "loss": 0.1448,
+      "step": 2312
+    },
+    {
+      "epoch": 0.3747569669475049,
+      "grad_norm": 1.0789496898651123,
+      "learning_rate": 4.882856165190256e-06,
+      "loss": 0.1549,
+      "step": 2313
+    },
+    {
+      "epoch": 0.37491898898250164,
+      "grad_norm": 1.0196281671524048,
+      "learning_rate": 4.88272383471275e-06,
+      "loss": 0.1511,
+      "step": 2314
+    },
+    {
+      "epoch": 0.37508101101749836,
+      "grad_norm": 0.9408444762229919,
+      "learning_rate": 4.882591431329662e-06,
+      "loss": 0.1301,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3752430330524951,
+      "grad_norm": 1.0712037086486816,
+      "learning_rate": 4.8824589550450415e-06,
+      "loss": 0.1601,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3754050550874919,
+      "grad_norm": 0.8943222165107727,
+      "learning_rate": 4.8823264058629426e-06,
+      "loss": 0.1318,
+      "step": 2317
+    },
+    {
+      "epoch": 0.37556707712248866,
+      "grad_norm": 0.9707005620002747,
+      "learning_rate": 4.882193783787421e-06,
+      "loss": 0.1476,
+      "step": 2318
+    },
+    {
+      "epoch": 0.3757290991574854,
+      "grad_norm": 1.0680534839630127,
+      "learning_rate": 4.882061088822534e-06,
+      "loss": 0.1753,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3758911211924822,
+      "grad_norm": 1.0867925882339478,
+      "learning_rate": 4.881928320972342e-06,
+      "loss": 0.1571,
+      "step": 2320
+    },
+    {
+      "epoch": 0.37605314322747896,
+      "grad_norm": 1.1130952835083008,
+      "learning_rate": 4.881795480240908e-06,
+      "loss": 0.169,
+      "step": 2321
+    },
+    {
+      "epoch": 0.37621516526247567,
+      "grad_norm": 1.034469485282898,
+      "learning_rate": 4.881662566632296e-06,
+      "loss": 0.1709,
+      "step": 2322
+    },
+    {
+      "epoch": 0.37637718729747244,
+      "grad_norm": 1.1244481801986694,
+      "learning_rate": 4.881529580150573e-06,
+      "loss": 0.1594,
+      "step": 2323
+    },
+    {
+      "epoch": 0.3765392093324692,
+      "grad_norm": 1.132101058959961,
+      "learning_rate": 4.881396520799808e-06,
+      "loss": 0.1757,
+      "step": 2324
+    },
+    {
+      "epoch": 0.37670123136746597,
+      "grad_norm": 1.098602294921875,
+      "learning_rate": 4.881263388584072e-06,
+      "loss": 0.154,
+      "step": 2325
+    },
+    {
+      "epoch": 0.37686325340246274,
+      "grad_norm": 0.9612120389938354,
+      "learning_rate": 4.8811301835074384e-06,
+      "loss": 0.1427,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3770252754374595,
+      "grad_norm": 1.0000402927398682,
+      "learning_rate": 4.880996905573985e-06,
+      "loss": 0.1531,
+      "step": 2327
+    },
+    {
+      "epoch": 0.3771872974724563,
+      "grad_norm": 1.076193928718567,
+      "learning_rate": 4.880863554787787e-06,
+      "loss": 0.1791,
+      "step": 2328
+    },
+    {
+      "epoch": 0.37734931950745304,
+      "grad_norm": 0.9946293234825134,
+      "learning_rate": 4.8807301311529266e-06,
+      "loss": 0.1507,
+      "step": 2329
+    },
+    {
+      "epoch": 0.37751134154244975,
+      "grad_norm": 0.9795472025871277,
+      "learning_rate": 4.880596634673484e-06,
+      "loss": 0.1498,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3776733635774465,
+      "grad_norm": 1.1108335256576538,
+      "learning_rate": 4.880463065353547e-06,
+      "loss": 0.1735,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3778353856124433,
+      "grad_norm": 1.0525665283203125,
+      "learning_rate": 4.8803294231972e-06,
+      "loss": 0.1655,
+      "step": 2332
+    },
+    {
+      "epoch": 0.37799740764744005,
+      "grad_norm": 1.1461604833602905,
+      "learning_rate": 4.880195708208533e-06,
+      "loss": 0.1622,
+      "step": 2333
+    },
+    {
+      "epoch": 0.3781594296824368,
+      "grad_norm": 1.1616642475128174,
+      "learning_rate": 4.8800619203916376e-06,
+      "loss": 0.1538,
+      "step": 2334
+    },
+    {
+      "epoch": 0.3783214517174336,
+      "grad_norm": 1.3006972074508667,
+      "learning_rate": 4.879928059750606e-06,
+      "loss": 0.1981,
+      "step": 2335
+    },
+    {
+      "epoch": 0.37848347375243035,
+      "grad_norm": 1.2075080871582031,
+      "learning_rate": 4.8797941262895365e-06,
+      "loss": 0.1655,
+      "step": 2336
+    },
+    {
+      "epoch": 0.37864549578742707,
+      "grad_norm": 1.0310864448547363,
+      "learning_rate": 4.8796601200125245e-06,
+      "loss": 0.1668,
+      "step": 2337
+    },
+    {
+      "epoch": 0.37880751782242383,
+      "grad_norm": 1.1267403364181519,
+      "learning_rate": 4.8795260409236725e-06,
+      "loss": 0.1574,
+      "step": 2338
+    },
+    {
+      "epoch": 0.3789695398574206,
+      "grad_norm": 1.0787062644958496,
+      "learning_rate": 4.879391889027081e-06,
+      "loss": 0.1417,
+      "step": 2339
+    },
+    {
+      "epoch": 0.37913156189241737,
+      "grad_norm": 0.9812328219413757,
+      "learning_rate": 4.879257664326856e-06,
+      "loss": 0.1436,
+      "step": 2340
+    },
+    {
+      "epoch": 0.37929358392741414,
+      "grad_norm": 1.0274299383163452,
+      "learning_rate": 4.8791233668271055e-06,
+      "loss": 0.1532,
+      "step": 2341
+    },
+    {
+      "epoch": 0.3794556059624109,
+      "grad_norm": 0.916479766368866,
+      "learning_rate": 4.8789889965319355e-06,
+      "loss": 0.1369,
+      "step": 2342
+    },
+    {
+      "epoch": 0.37961762799740767,
+      "grad_norm": 1.0698994398117065,
+      "learning_rate": 4.87885455344546e-06,
+      "loss": 0.1632,
+      "step": 2343
+    },
+    {
+      "epoch": 0.3797796500324044,
+      "grad_norm": 1.1658228635787964,
+      "learning_rate": 4.878720037571792e-06,
+      "loss": 0.1735,
+      "step": 2344
+    },
+    {
+      "epoch": 0.37994167206740115,
+      "grad_norm": 1.0319538116455078,
+      "learning_rate": 4.878585448915047e-06,
+      "loss": 0.1601,
+      "step": 2345
+    },
+    {
+      "epoch": 0.3801036941023979,
+      "grad_norm": 1.0444625616073608,
+      "learning_rate": 4.878450787479344e-06,
+      "loss": 0.1496,
+      "step": 2346
+    },
+    {
+      "epoch": 0.3802657161373947,
+      "grad_norm": 0.9932622909545898,
+      "learning_rate": 4.8783160532688026e-06,
+      "loss": 0.1479,
+      "step": 2347
+    },
+    {
+      "epoch": 0.38042773817239145,
+      "grad_norm": 1.0771230459213257,
+      "learning_rate": 4.878181246287544e-06,
+      "loss": 0.172,
+      "step": 2348
+    },
+    {
+      "epoch": 0.3805897602073882,
+      "grad_norm": 0.9821139574050903,
+      "learning_rate": 4.878046366539696e-06,
+      "loss": 0.1355,
+      "step": 2349
+    },
+    {
+      "epoch": 0.380751782242385,
+      "grad_norm": 1.1744256019592285,
+      "learning_rate": 4.877911414029382e-06,
+      "loss": 0.1728,
+      "step": 2350
+    },
+    {
+      "epoch": 0.38091380427738175,
+      "grad_norm": 1.0486180782318115,
+      "learning_rate": 4.877776388760735e-06,
+      "loss": 0.151,
+      "step": 2351
+    },
+    {
+      "epoch": 0.38107582631237846,
+      "grad_norm": 0.974438488483429,
+      "learning_rate": 4.8776412907378845e-06,
+      "loss": 0.1425,
+      "step": 2352
+    },
+    {
+      "epoch": 0.38123784834737523,
+      "grad_norm": 0.9871600270271301,
+      "learning_rate": 4.877506119964964e-06,
+      "loss": 0.1509,
+      "step": 2353
+    },
+    {
+      "epoch": 0.381399870382372,
+      "grad_norm": 1.0788719654083252,
+      "learning_rate": 4.877370876446109e-06,
+      "loss": 0.1502,
+      "step": 2354
+    },
+    {
+      "epoch": 0.38156189241736876,
+      "grad_norm": 0.8554217219352722,
+      "learning_rate": 4.877235560185459e-06,
+      "loss": 0.1317,
+      "step": 2355
+    },
+    {
+      "epoch": 0.38172391445236553,
+      "grad_norm": 1.0223811864852905,
+      "learning_rate": 4.877100171187154e-06,
+      "loss": 0.1618,
+      "step": 2356
+    },
+    {
+      "epoch": 0.3818859364873623,
+      "grad_norm": 1.042266607284546,
+      "learning_rate": 4.876964709455335e-06,
+      "loss": 0.1565,
+      "step": 2357
+    },
+    {
+      "epoch": 0.38204795852235907,
+      "grad_norm": 1.0862993001937866,
+      "learning_rate": 4.876829174994149e-06,
+      "loss": 0.1689,
+      "step": 2358
+    },
+    {
+      "epoch": 0.3822099805573558,
+      "grad_norm": 1.0016382932662964,
+      "learning_rate": 4.876693567807741e-06,
+      "loss": 0.1428,
+      "step": 2359
+    },
+    {
+      "epoch": 0.38237200259235254,
+      "grad_norm": 1.009976863861084,
+      "learning_rate": 4.8765578879002625e-06,
+      "loss": 0.1525,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3825340246273493,
+      "grad_norm": 0.9850345253944397,
+      "learning_rate": 4.876422135275863e-06,
+      "loss": 0.1373,
+      "step": 2361
+    },
+    {
+      "epoch": 0.3826960466623461,
+      "grad_norm": 1.0292307138442993,
+      "learning_rate": 4.8762863099386984e-06,
+      "loss": 0.1444,
+      "step": 2362
+    },
+    {
+      "epoch": 0.38285806869734285,
+      "grad_norm": 0.9896308779716492,
+      "learning_rate": 4.876150411892922e-06,
+      "loss": 0.1527,
+      "step": 2363
+    },
+    {
+      "epoch": 0.3830200907323396,
+      "grad_norm": 1.091084599494934,
+      "learning_rate": 4.876014441142693e-06,
+      "loss": 0.1541,
+      "step": 2364
+    },
+    {
+      "epoch": 0.3831821127673364,
+      "grad_norm": 1.06158447265625,
+      "learning_rate": 4.875878397692172e-06,
+      "loss": 0.1555,
+      "step": 2365
+    },
+    {
+      "epoch": 0.3833441348023331,
+      "grad_norm": 1.0953679084777832,
+      "learning_rate": 4.8757422815455215e-06,
+      "loss": 0.1643,
+      "step": 2366
+    },
+    {
+      "epoch": 0.38350615683732986,
+      "grad_norm": 1.0464844703674316,
+      "learning_rate": 4.8756060927069075e-06,
+      "loss": 0.1512,
+      "step": 2367
+    },
+    {
+      "epoch": 0.3836681788723266,
+      "grad_norm": 1.0464715957641602,
+      "learning_rate": 4.875469831180495e-06,
+      "loss": 0.1573,
+      "step": 2368
+    },
+    {
+      "epoch": 0.3838302009073234,
+      "grad_norm": 0.9663591384887695,
+      "learning_rate": 4.875333496970454e-06,
+      "loss": 0.1446,
+      "step": 2369
+    },
+    {
+      "epoch": 0.38399222294232016,
+      "grad_norm": 1.192948341369629,
+      "learning_rate": 4.875197090080957e-06,
+      "loss": 0.1528,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3841542449773169,
+      "grad_norm": 1.0319154262542725,
+      "learning_rate": 4.875060610516176e-06,
+      "loss": 0.1489,
+      "step": 2371
+    },
+    {
+      "epoch": 0.3843162670123137,
+      "grad_norm": 1.1265735626220703,
+      "learning_rate": 4.874924058280288e-06,
+      "loss": 0.1558,
+      "step": 2372
+    },
+    {
+      "epoch": 0.38447828904731046,
+      "grad_norm": 1.1732511520385742,
+      "learning_rate": 4.874787433377472e-06,
+      "loss": 0.1631,
+      "step": 2373
+    },
+    {
+      "epoch": 0.3846403110823072,
+      "grad_norm": 1.2255806922912598,
+      "learning_rate": 4.874650735811906e-06,
+      "loss": 0.1717,
+      "step": 2374
+    },
+    {
+      "epoch": 0.38480233311730394,
+      "grad_norm": 1.0163358449935913,
+      "learning_rate": 4.874513965587775e-06,
+      "loss": 0.1531,
+      "step": 2375
+    },
+    {
+      "epoch": 0.3849643551523007,
+      "grad_norm": 1.047625184059143,
+      "learning_rate": 4.874377122709263e-06,
+      "loss": 0.1702,
+      "step": 2376
+    },
+    {
+      "epoch": 0.3851263771872975,
+      "grad_norm": 1.0815023183822632,
+      "learning_rate": 4.874240207180556e-06,
+      "loss": 0.1649,
+      "step": 2377
+    },
+    {
+      "epoch": 0.38528839922229424,
+      "grad_norm": 0.9695912003517151,
+      "learning_rate": 4.874103219005845e-06,
+      "loss": 0.1392,
+      "step": 2378
+    },
+    {
+      "epoch": 0.385450421257291,
+      "grad_norm": 1.0727660655975342,
+      "learning_rate": 4.873966158189321e-06,
+      "loss": 0.1428,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3856124432922878,
+      "grad_norm": 1.184160590171814,
+      "learning_rate": 4.873829024735176e-06,
+      "loss": 0.1532,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3857744653272845,
+      "grad_norm": 0.9503794312477112,
+      "learning_rate": 4.873691818647609e-06,
+      "loss": 0.1313,
+      "step": 2381
+    },
+    {
+      "epoch": 0.38593648736228126,
+      "grad_norm": 1.0387951135635376,
+      "learning_rate": 4.873554539930815e-06,
+      "loss": 0.1699,
+      "step": 2382
+    },
+    {
+      "epoch": 0.386098509397278,
+      "grad_norm": 1.1206920146942139,
+      "learning_rate": 4.873417188588997e-06,
+      "loss": 0.1645,
+      "step": 2383
+    },
+    {
+      "epoch": 0.3862605314322748,
+      "grad_norm": 1.0586580038070679,
+      "learning_rate": 4.873279764626357e-06,
+      "loss": 0.1505,
+      "step": 2384
+    },
+    {
+      "epoch": 0.38642255346727156,
+      "grad_norm": 1.080195665359497,
+      "learning_rate": 4.873142268047099e-06,
+      "loss": 0.1518,
+      "step": 2385
+    },
+    {
+      "epoch": 0.3865845755022683,
+      "grad_norm": 1.0671923160552979,
+      "learning_rate": 4.87300469885543e-06,
+      "loss": 0.17,
+      "step": 2386
+    },
+    {
+      "epoch": 0.3867465975372651,
+      "grad_norm": 0.9395395517349243,
+      "learning_rate": 4.87286705705556e-06,
+      "loss": 0.1497,
+      "step": 2387
+    },
+    {
+      "epoch": 0.3869086195722618,
+      "grad_norm": 1.0905933380126953,
+      "learning_rate": 4.872729342651701e-06,
+      "loss": 0.1711,
+      "step": 2388
+    },
+    {
+      "epoch": 0.38707064160725857,
+      "grad_norm": 1.1106778383255005,
+      "learning_rate": 4.8725915556480655e-06,
+      "loss": 0.1736,
+      "step": 2389
+    },
+    {
+      "epoch": 0.38723266364225534,
+      "grad_norm": 1.1426775455474854,
+      "learning_rate": 4.87245369604887e-06,
+      "loss": 0.178,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3873946856772521,
+      "grad_norm": 1.0927367210388184,
+      "learning_rate": 4.872315763858332e-06,
+      "loss": 0.1624,
+      "step": 2391
+    },
+    {
+      "epoch": 0.38755670771224887,
+      "grad_norm": 1.1219305992126465,
+      "learning_rate": 4.872177759080673e-06,
+      "loss": 0.1561,
+      "step": 2392
+    },
+    {
+      "epoch": 0.38771872974724564,
+      "grad_norm": 1.0665841102600098,
+      "learning_rate": 4.872039681720116e-06,
+      "loss": 0.1651,
+      "step": 2393
+    },
+    {
+      "epoch": 0.3878807517822424,
+      "grad_norm": 1.0562465190887451,
+      "learning_rate": 4.8719015317808835e-06,
+      "loss": 0.1674,
+      "step": 2394
+    },
+    {
+      "epoch": 0.3880427738172391,
+      "grad_norm": 1.0711188316345215,
+      "learning_rate": 4.8717633092672045e-06,
+      "loss": 0.1534,
+      "step": 2395
+    },
+    {
+      "epoch": 0.3882047958522359,
+      "grad_norm": 1.0176469087600708,
+      "learning_rate": 4.871625014183308e-06,
+      "loss": 0.1558,
+      "step": 2396
+    },
+    {
+      "epoch": 0.38836681788723265,
+      "grad_norm": 1.0027252435684204,
+      "learning_rate": 4.871486646533425e-06,
+      "loss": 0.1381,
+      "step": 2397
+    },
+    {
+      "epoch": 0.3885288399222294,
+      "grad_norm": 1.0068401098251343,
+      "learning_rate": 4.8713482063217895e-06,
+      "loss": 0.1579,
+      "step": 2398
+    },
+    {
+      "epoch": 0.3886908619572262,
+      "grad_norm": 0.9579773545265198,
+      "learning_rate": 4.871209693552638e-06,
+      "loss": 0.1335,
+      "step": 2399
+    },
+    {
+      "epoch": 0.38885288399222295,
+      "grad_norm": 1.1602519750595093,
+      "learning_rate": 4.871071108230208e-06,
+      "loss": 0.1684,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3890149060272197,
+      "grad_norm": 1.4036099910736084,
+      "learning_rate": 4.87093245035874e-06,
+      "loss": 0.1646,
+      "step": 2401
+    },
+    {
+      "epoch": 0.3891769280622165,
+      "grad_norm": 1.1225903034210205,
+      "learning_rate": 4.8707937199424756e-06,
+      "loss": 0.1643,
+      "step": 2402
+    },
+    {
+      "epoch": 0.3893389500972132,
+      "grad_norm": 1.1143039464950562,
+      "learning_rate": 4.870654916985661e-06,
+      "loss": 0.1577,
+      "step": 2403
+    },
+    {
+      "epoch": 0.38950097213220997,
+      "grad_norm": 1.080872893333435,
+      "learning_rate": 4.870516041492543e-06,
+      "loss": 0.1504,
+      "step": 2404
+    },
+    {
+      "epoch": 0.38966299416720673,
+      "grad_norm": 1.2829567193984985,
+      "learning_rate": 4.870377093467371e-06,
+      "loss": 0.1817,
+      "step": 2405
+    },
+    {
+      "epoch": 0.3898250162022035,
+      "grad_norm": 1.2042837142944336,
+      "learning_rate": 4.870238072914396e-06,
+      "loss": 0.1548,
+      "step": 2406
+    },
+    {
+      "epoch": 0.38998703823720027,
+      "grad_norm": 0.9398330450057983,
+      "learning_rate": 4.870098979837871e-06,
+      "loss": 0.1389,
+      "step": 2407
+    },
+    {
+      "epoch": 0.39014906027219703,
+      "grad_norm": 1.0796722173690796,
+      "learning_rate": 4.869959814242054e-06,
+      "loss": 0.157,
+      "step": 2408
+    },
+    {
+      "epoch": 0.3903110823071938,
+      "grad_norm": 1.1259878873825073,
+      "learning_rate": 4.869820576131202e-06,
+      "loss": 0.153,
+      "step": 2409
+    },
+    {
+      "epoch": 0.3904731043421905,
+      "grad_norm": 1.019061803817749,
+      "learning_rate": 4.8696812655095744e-06,
+      "loss": 0.1538,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3906351263771873,
+      "grad_norm": 0.9436379075050354,
+      "learning_rate": 4.869541882381435e-06,
+      "loss": 0.1485,
+      "step": 2411
+    },
+    {
+      "epoch": 0.39079714841218405,
+      "grad_norm": 1.071523904800415,
+      "learning_rate": 4.869402426751048e-06,
+      "loss": 0.1594,
+      "step": 2412
+    },
+    {
+      "epoch": 0.3909591704471808,
+      "grad_norm": 1.1812318563461304,
+      "learning_rate": 4.86926289862268e-06,
+      "loss": 0.1701,
+      "step": 2413
+    },
+    {
+      "epoch": 0.3911211924821776,
+      "grad_norm": 1.1143845319747925,
+      "learning_rate": 4.8691232980006015e-06,
+      "loss": 0.1499,
+      "step": 2414
+    },
+    {
+      "epoch": 0.39128321451717435,
+      "grad_norm": 1.0867897272109985,
+      "learning_rate": 4.868983624889083e-06,
+      "loss": 0.1617,
+      "step": 2415
+    },
+    {
+      "epoch": 0.3914452365521711,
+      "grad_norm": 1.0865579843521118,
+      "learning_rate": 4.868843879292399e-06,
+      "loss": 0.1634,
+      "step": 2416
+    },
+    {
+      "epoch": 0.39160725858716783,
+      "grad_norm": 0.9552433490753174,
+      "learning_rate": 4.868704061214824e-06,
+      "loss": 0.1485,
+      "step": 2417
+    },
+    {
+      "epoch": 0.3917692806221646,
+      "grad_norm": 1.0676617622375488,
+      "learning_rate": 4.868564170660637e-06,
+      "loss": 0.1744,
+      "step": 2418
+    },
+    {
+      "epoch": 0.39193130265716136,
+      "grad_norm": 0.9854423403739929,
+      "learning_rate": 4.868424207634118e-06,
+      "loss": 0.1511,
+      "step": 2419
+    },
+    {
+      "epoch": 0.39209332469215813,
+      "grad_norm": 0.8792446851730347,
+      "learning_rate": 4.868284172139551e-06,
+      "loss": 0.1247,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3922553467271549,
+      "grad_norm": 1.0896074771881104,
+      "learning_rate": 4.868144064181218e-06,
+      "loss": 0.1551,
+      "step": 2421
+    },
+    {
+      "epoch": 0.39241736876215166,
+      "grad_norm": 1.0454200506210327,
+      "learning_rate": 4.868003883763408e-06,
+      "loss": 0.1417,
+      "step": 2422
+    },
+    {
+      "epoch": 0.39257939079714843,
+      "grad_norm": 0.9985229969024658,
+      "learning_rate": 4.8678636308904095e-06,
+      "loss": 0.1619,
+      "step": 2423
+    },
+    {
+      "epoch": 0.3927414128321452,
+      "grad_norm": 0.9544141888618469,
+      "learning_rate": 4.867723305566514e-06,
+      "loss": 0.1435,
+      "step": 2424
+    },
+    {
+      "epoch": 0.3929034348671419,
+      "grad_norm": 1.029191017150879,
+      "learning_rate": 4.867582907796016e-06,
+      "loss": 0.1544,
+      "step": 2425
+    },
+    {
+      "epoch": 0.3930654569021387,
+      "grad_norm": 1.372963786125183,
+      "learning_rate": 4.86744243758321e-06,
+      "loss": 0.1815,
+      "step": 2426
+    },
+    {
+      "epoch": 0.39322747893713544,
+      "grad_norm": 1.0754982233047485,
+      "learning_rate": 4.867301894932394e-06,
+      "loss": 0.1632,
+      "step": 2427
+    },
+    {
+      "epoch": 0.3933895009721322,
+      "grad_norm": 1.3161104917526245,
+      "learning_rate": 4.8671612798478685e-06,
+      "loss": 0.1773,
+      "step": 2428
+    },
+    {
+      "epoch": 0.393551523007129,
+      "grad_norm": 1.0129326581954956,
+      "learning_rate": 4.867020592333937e-06,
+      "loss": 0.1519,
+      "step": 2429
+    },
+    {
+      "epoch": 0.39371354504212575,
+      "grad_norm": 0.9305555820465088,
+      "learning_rate": 4.866879832394903e-06,
+      "loss": 0.1458,
+      "step": 2430
+    },
+    {
+      "epoch": 0.3938755670771225,
+      "grad_norm": 1.048586130142212,
+      "learning_rate": 4.866739000035074e-06,
+      "loss": 0.1473,
+      "step": 2431
+    },
+    {
+      "epoch": 0.3940375891121192,
+      "grad_norm": 1.050463318824768,
+      "learning_rate": 4.86659809525876e-06,
+      "loss": 0.1673,
+      "step": 2432
+    },
+    {
+      "epoch": 0.394199611147116,
+      "grad_norm": 0.9650267958641052,
+      "learning_rate": 4.86645711807027e-06,
+      "loss": 0.1479,
+      "step": 2433
+    },
+    {
+      "epoch": 0.39436163318211276,
+      "grad_norm": 1.1384695768356323,
+      "learning_rate": 4.866316068473919e-06,
+      "loss": 0.1779,
+      "step": 2434
+    },
+    {
+      "epoch": 0.3945236552171095,
+      "grad_norm": 1.1191463470458984,
+      "learning_rate": 4.866174946474023e-06,
+      "loss": 0.1442,
+      "step": 2435
+    },
+    {
+      "epoch": 0.3946856772521063,
+      "grad_norm": 1.047663688659668,
+      "learning_rate": 4.8660337520749e-06,
+      "loss": 0.1437,
+      "step": 2436
+    },
+    {
+      "epoch": 0.39484769928710306,
+      "grad_norm": 1.0196114778518677,
+      "learning_rate": 4.865892485280869e-06,
+      "loss": 0.1477,
+      "step": 2437
+    },
+    {
+      "epoch": 0.3950097213220998,
+      "grad_norm": 0.9830042719841003,
+      "learning_rate": 4.865751146096255e-06,
+      "loss": 0.1492,
+      "step": 2438
+    },
+    {
+      "epoch": 0.39517174335709654,
+      "grad_norm": 1.0708327293395996,
+      "learning_rate": 4.865609734525379e-06,
+      "loss": 0.1685,
+      "step": 2439
+    },
+    {
+      "epoch": 0.3953337653920933,
+      "grad_norm": 1.0699785947799683,
+      "learning_rate": 4.865468250572571e-06,
+      "loss": 0.173,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3954957874270901,
+      "grad_norm": 0.9510377049446106,
+      "learning_rate": 4.8653266942421585e-06,
+      "loss": 0.1457,
+      "step": 2441
+    },
+    {
+      "epoch": 0.39565780946208684,
+      "grad_norm": 0.93174147605896,
+      "learning_rate": 4.865185065538472e-06,
+      "loss": 0.1385,
+      "step": 2442
+    },
+    {
+      "epoch": 0.3958198314970836,
+      "grad_norm": 0.9724693894386292,
+      "learning_rate": 4.865043364465848e-06,
+      "loss": 0.1609,
+      "step": 2443
+    },
+    {
+      "epoch": 0.3959818535320804,
+      "grad_norm": 0.9042225480079651,
+      "learning_rate": 4.86490159102862e-06,
+      "loss": 0.1339,
+      "step": 2444
+    },
+    {
+      "epoch": 0.39614387556707714,
+      "grad_norm": 1.1967670917510986,
+      "learning_rate": 4.864759745231126e-06,
+      "loss": 0.1431,
+      "step": 2445
+    },
+    {
+      "epoch": 0.3963058976020739,
+      "grad_norm": 1.0818557739257812,
+      "learning_rate": 4.8646178270777055e-06,
+      "loss": 0.1786,
+      "step": 2446
+    },
+    {
+      "epoch": 0.3964679196370706,
+      "grad_norm": 1.0691944360733032,
+      "learning_rate": 4.864475836572703e-06,
+      "loss": 0.1758,
+      "step": 2447
+    },
+    {
+      "epoch": 0.3966299416720674,
+      "grad_norm": 0.9906453490257263,
+      "learning_rate": 4.864333773720461e-06,
+      "loss": 0.1508,
+      "step": 2448
+    },
+    {
+      "epoch": 0.39679196370706415,
+      "grad_norm": 0.993923008441925,
+      "learning_rate": 4.864191638525328e-06,
+      "loss": 0.1369,
+      "step": 2449
+    },
+    {
+      "epoch": 0.3969539857420609,
+      "grad_norm": 1.0036680698394775,
+      "learning_rate": 4.8640494309916506e-06,
+      "loss": 0.1497,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3971160077770577,
+      "grad_norm": 0.9663056135177612,
+      "learning_rate": 4.863907151123782e-06,
+      "loss": 0.1443,
+      "step": 2451
+    },
+    {
+      "epoch": 0.39727802981205446,
+      "grad_norm": 1.1457011699676514,
+      "learning_rate": 4.863764798926076e-06,
+      "loss": 0.1601,
+      "step": 2452
+    },
+    {
+      "epoch": 0.3974400518470512,
+      "grad_norm": 1.0735100507736206,
+      "learning_rate": 4.863622374402887e-06,
+      "loss": 0.1587,
+      "step": 2453
+    },
+    {
+      "epoch": 0.39760207388204793,
+      "grad_norm": 1.3038159608840942,
+      "learning_rate": 4.863479877558573e-06,
+      "loss": 0.2004,
+      "step": 2454
+    },
+    {
+      "epoch": 0.3977640959170447,
+      "grad_norm": 1.0412776470184326,
+      "learning_rate": 4.863337308397495e-06,
+      "loss": 0.1383,
+      "step": 2455
+    },
+    {
+      "epoch": 0.39792611795204147,
+      "grad_norm": 1.0914403200149536,
+      "learning_rate": 4.863194666924013e-06,
+      "loss": 0.1643,
+      "step": 2456
+    },
+    {
+      "epoch": 0.39808813998703824,
+      "grad_norm": 0.943895697593689,
+      "learning_rate": 4.863051953142494e-06,
+      "loss": 0.1351,
+      "step": 2457
+    },
+    {
+      "epoch": 0.398250162022035,
+      "grad_norm": 1.0537207126617432,
+      "learning_rate": 4.862909167057304e-06,
+      "loss": 0.1631,
+      "step": 2458
+    },
+    {
+      "epoch": 0.39841218405703177,
+      "grad_norm": 1.0041303634643555,
+      "learning_rate": 4.862766308672811e-06,
+      "loss": 0.1668,
+      "step": 2459
+    },
+    {
+      "epoch": 0.39857420609202854,
+      "grad_norm": 0.9804761409759521,
+      "learning_rate": 4.862623377993387e-06,
+      "loss": 0.1657,
+      "step": 2460
+    },
+    {
+      "epoch": 0.39873622812702525,
+      "grad_norm": 0.928268551826477,
+      "learning_rate": 4.862480375023405e-06,
+      "loss": 0.1355,
+      "step": 2461
+    },
+    {
+      "epoch": 0.398898250162022,
+      "grad_norm": 1.046536922454834,
+      "learning_rate": 4.862337299767241e-06,
+      "loss": 0.1444,
+      "step": 2462
+    },
+    {
+      "epoch": 0.3990602721970188,
+      "grad_norm": 1.052215814590454,
+      "learning_rate": 4.862194152229271e-06,
+      "loss": 0.1699,
+      "step": 2463
+    },
+    {
+      "epoch": 0.39922229423201555,
+      "grad_norm": 0.983517050743103,
+      "learning_rate": 4.862050932413878e-06,
+      "loss": 0.1326,
+      "step": 2464
+    },
+    {
+      "epoch": 0.3993843162670123,
+      "grad_norm": 1.0024183988571167,
+      "learning_rate": 4.861907640325442e-06,
+      "loss": 0.1512,
+      "step": 2465
+    },
+    {
+      "epoch": 0.3995463383020091,
+      "grad_norm": 1.0899640321731567,
+      "learning_rate": 4.8617642759683474e-06,
+      "loss": 0.1622,
+      "step": 2466
+    },
+    {
+      "epoch": 0.39970836033700585,
+      "grad_norm": 1.0080574750900269,
+      "learning_rate": 4.861620839346982e-06,
+      "loss": 0.147,
+      "step": 2467
+    },
+    {
+      "epoch": 0.3998703823720026,
+      "grad_norm": 1.098242163658142,
+      "learning_rate": 4.861477330465734e-06,
+      "loss": 0.1628,
+      "step": 2468
+    },
+    {
+      "epoch": 0.40003240440699933,
+      "grad_norm": 1.0178964138031006,
+      "learning_rate": 4.861333749328993e-06,
+      "loss": 0.1432,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4001944264419961,
+      "grad_norm": 1.0105563402175903,
+      "learning_rate": 4.861190095941155e-06,
+      "loss": 0.148,
+      "step": 2470
+    },
+    {
+      "epoch": 0.40035644847699287,
+      "grad_norm": 1.041428565979004,
+      "learning_rate": 4.861046370306613e-06,
+      "loss": 0.15,
+      "step": 2471
+    },
+    {
+      "epoch": 0.40051847051198963,
+      "grad_norm": 1.3379833698272705,
+      "learning_rate": 4.860902572429767e-06,
+      "loss": 0.17,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4006804925469864,
+      "grad_norm": 1.0125343799591064,
+      "learning_rate": 4.8607587023150145e-06,
+      "loss": 0.1607,
+      "step": 2473
+    },
+    {
+      "epoch": 0.40084251458198317,
+      "grad_norm": 0.9886438846588135,
+      "learning_rate": 4.86061475996676e-06,
+      "loss": 0.1524,
+      "step": 2474
+    },
+    {
+      "epoch": 0.40100453661697993,
+      "grad_norm": 1.094799280166626,
+      "learning_rate": 4.860470745389405e-06,
+      "loss": 0.1548,
+      "step": 2475
+    },
+    {
+      "epoch": 0.40116655865197665,
+      "grad_norm": 0.9762154221534729,
+      "learning_rate": 4.860326658587358e-06,
+      "loss": 0.1545,
+      "step": 2476
+    },
+    {
+      "epoch": 0.4013285806869734,
+      "grad_norm": 1.0075474977493286,
+      "learning_rate": 4.860182499565027e-06,
+      "loss": 0.1675,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4014906027219702,
+      "grad_norm": 0.9806322455406189,
+      "learning_rate": 4.860038268326823e-06,
+      "loss": 0.1658,
+      "step": 2478
+    },
+    {
+      "epoch": 0.40165262475696695,
+      "grad_norm": 0.9867870211601257,
+      "learning_rate": 4.859893964877159e-06,
+      "loss": 0.1424,
+      "step": 2479
+    },
+    {
+      "epoch": 0.4018146467919637,
+      "grad_norm": 1.0405807495117188,
+      "learning_rate": 4.85974958922045e-06,
+      "loss": 0.1644,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4019766688269605,
+      "grad_norm": 0.9794811606407166,
+      "learning_rate": 4.8596051413611155e-06,
+      "loss": 0.1547,
+      "step": 2481
+    },
+    {
+      "epoch": 0.40213869086195725,
+      "grad_norm": 1.097029209136963,
+      "learning_rate": 4.859460621303572e-06,
+      "loss": 0.1774,
+      "step": 2482
+    },
+    {
+      "epoch": 0.40230071289695396,
+      "grad_norm": 0.9748839139938354,
+      "learning_rate": 4.859316029052245e-06,
+      "loss": 0.1384,
+      "step": 2483
+    },
+    {
+      "epoch": 0.4024627349319507,
+      "grad_norm": 1.0535310506820679,
+      "learning_rate": 4.859171364611556e-06,
+      "loss": 0.1546,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4026247569669475,
+      "grad_norm": 0.8713586330413818,
+      "learning_rate": 4.859026627985933e-06,
+      "loss": 0.1416,
+      "step": 2485
+    },
+    {
+      "epoch": 0.40278677900194426,
+      "grad_norm": 1.1586978435516357,
+      "learning_rate": 4.8588818191798035e-06,
+      "loss": 0.1747,
+      "step": 2486
+    },
+    {
+      "epoch": 0.40294880103694103,
+      "grad_norm": 1.052422285079956,
+      "learning_rate": 4.858736938197599e-06,
+      "loss": 0.1516,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4031108230719378,
+      "grad_norm": 0.9205193519592285,
+      "learning_rate": 4.858591985043751e-06,
+      "loss": 0.1432,
+      "step": 2488
+    },
+    {
+      "epoch": 0.40327284510693456,
+      "grad_norm": 0.9910753965377808,
+      "learning_rate": 4.858446959722698e-06,
+      "loss": 0.1499,
+      "step": 2489
+    },
+    {
+      "epoch": 0.40343486714193133,
+      "grad_norm": 1.0436080694198608,
+      "learning_rate": 4.858301862238874e-06,
+      "loss": 0.1561,
+      "step": 2490
+    },
+    {
+      "epoch": 0.40359688917692804,
+      "grad_norm": 1.0225828886032104,
+      "learning_rate": 4.858156692596721e-06,
+      "loss": 0.151,
+      "step": 2491
+    },
+    {
+      "epoch": 0.4037589112119248,
+      "grad_norm": 1.093070387840271,
+      "learning_rate": 4.858011450800678e-06,
+      "loss": 0.1608,
+      "step": 2492
+    },
+    {
+      "epoch": 0.4039209332469216,
+      "grad_norm": 1.1206576824188232,
+      "learning_rate": 4.857866136855192e-06,
+      "loss": 0.1707,
+      "step": 2493
+    },
+    {
+      "epoch": 0.40408295528191834,
+      "grad_norm": 0.9448593854904175,
+      "learning_rate": 4.857720750764708e-06,
+      "loss": 0.1395,
+      "step": 2494
+    },
+    {
+      "epoch": 0.4042449773169151,
+      "grad_norm": 0.9939359426498413,
+      "learning_rate": 4.857575292533675e-06,
+      "loss": 0.1436,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4044069993519119,
+      "grad_norm": 1.074859857559204,
+      "learning_rate": 4.857429762166543e-06,
+      "loss": 0.1645,
+      "step": 2496
+    },
+    {
+      "epoch": 0.40456902138690864,
+      "grad_norm": 0.9814433455467224,
+      "learning_rate": 4.857284159667766e-06,
+      "loss": 0.1486,
+      "step": 2497
+    },
+    {
+      "epoch": 0.40473104342190536,
+      "grad_norm": 1.1380925178527832,
+      "learning_rate": 4.857138485041797e-06,
+      "loss": 0.1672,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4048930654569021,
+      "grad_norm": 1.0334879159927368,
+      "learning_rate": 4.8569927382930945e-06,
+      "loss": 0.147,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4050550874918989,
+      "grad_norm": 1.037170171737671,
+      "learning_rate": 4.856846919426118e-06,
+      "loss": 0.1521,
+      "step": 2500
+    },
+    {
+      "epoch": 0.40521710952689566,
+      "grad_norm": 0.9876829385757446,
+      "learning_rate": 4.85670102844533e-06,
+      "loss": 0.1543,
+      "step": 2501
+    },
+    {
+      "epoch": 0.4053791315618924,
+      "grad_norm": 1.055932879447937,
+      "learning_rate": 4.856555065355193e-06,
+      "loss": 0.1641,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4055411535968892,
+      "grad_norm": 0.9136426448822021,
+      "learning_rate": 4.856409030160174e-06,
+      "loss": 0.1424,
+      "step": 2503
+    },
+    {
+      "epoch": 0.40570317563188596,
+      "grad_norm": 1.0281522274017334,
+      "learning_rate": 4.856262922864741e-06,
+      "loss": 0.1646,
+      "step": 2504
+    },
+    {
+      "epoch": 0.40586519766688267,
+      "grad_norm": 0.9944785237312317,
+      "learning_rate": 4.8561167434733655e-06,
+      "loss": 0.144,
+      "step": 2505
+    },
+    {
+      "epoch": 0.40602721970187944,
+      "grad_norm": 1.1220792531967163,
+      "learning_rate": 4.855970491990518e-06,
+      "loss": 0.1582,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4061892417368762,
+      "grad_norm": 0.8896439671516418,
+      "learning_rate": 4.855824168420675e-06,
+      "loss": 0.1386,
+      "step": 2507
+    },
+    {
+      "epoch": 0.406351263771873,
+      "grad_norm": 1.0499117374420166,
+      "learning_rate": 4.855677772768315e-06,
+      "loss": 0.1566,
+      "step": 2508
+    },
+    {
+      "epoch": 0.40651328580686974,
+      "grad_norm": 0.9450101256370544,
+      "learning_rate": 4.855531305037914e-06,
+      "loss": 0.1459,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4066753078418665,
+      "grad_norm": 1.0647412538528442,
+      "learning_rate": 4.855384765233956e-06,
+      "loss": 0.1569,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4068373298768633,
+      "grad_norm": 0.9091509580612183,
+      "learning_rate": 4.855238153360924e-06,
+      "loss": 0.1313,
+      "step": 2511
+    },
+    {
+      "epoch": 0.40699935191186,
+      "grad_norm": 1.0979092121124268,
+      "learning_rate": 4.8550914694233045e-06,
+      "loss": 0.1662,
+      "step": 2512
+    },
+    {
+      "epoch": 0.40716137394685675,
+      "grad_norm": 1.062098503112793,
+      "learning_rate": 4.854944713425585e-06,
+      "loss": 0.1572,
+      "step": 2513
+    },
+    {
+      "epoch": 0.4073233959818535,
+      "grad_norm": 0.9755269289016724,
+      "learning_rate": 4.854797885372255e-06,
+      "loss": 0.146,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4074854180168503,
+      "grad_norm": 1.0798227787017822,
+      "learning_rate": 4.854650985267809e-06,
+      "loss": 0.1669,
+      "step": 2515
+    },
+    {
+      "epoch": 0.40764744005184705,
+      "grad_norm": 1.0882648229599,
+      "learning_rate": 4.854504013116741e-06,
+      "loss": 0.163,
+      "step": 2516
+    },
+    {
+      "epoch": 0.4078094620868438,
+      "grad_norm": 0.9996983408927917,
+      "learning_rate": 4.854356968923549e-06,
+      "loss": 0.1403,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4079714841218406,
+      "grad_norm": 1.076771855354309,
+      "learning_rate": 4.8542098526927304e-06,
+      "loss": 0.1401,
+      "step": 2518
+    },
+    {
+      "epoch": 0.40813350615683736,
+      "grad_norm": 1.0781205892562866,
+      "learning_rate": 4.854062664428787e-06,
+      "loss": 0.1554,
+      "step": 2519
+    },
+    {
+      "epoch": 0.40829552819183407,
+      "grad_norm": 0.9651884436607361,
+      "learning_rate": 4.853915404136223e-06,
+      "loss": 0.1401,
+      "step": 2520
+    },
+    {
+      "epoch": 0.40845755022683083,
+      "grad_norm": 1.0154225826263428,
+      "learning_rate": 4.853768071819544e-06,
+      "loss": 0.1669,
+      "step": 2521
+    },
+    {
+      "epoch": 0.4086195722618276,
+      "grad_norm": 0.9365798830986023,
+      "learning_rate": 4.853620667483259e-06,
+      "loss": 0.1321,
+      "step": 2522
+    },
+    {
+      "epoch": 0.40878159429682437,
+      "grad_norm": 0.9837552309036255,
+      "learning_rate": 4.8534731911318755e-06,
+      "loss": 0.1488,
+      "step": 2523
+    },
+    {
+      "epoch": 0.40894361633182114,
+      "grad_norm": 1.0226497650146484,
+      "learning_rate": 4.853325642769908e-06,
+      "loss": 0.1504,
+      "step": 2524
+    },
+    {
+      "epoch": 0.4091056383668179,
+      "grad_norm": 1.206057071685791,
+      "learning_rate": 4.853178022401872e-06,
+      "loss": 0.1996,
+      "step": 2525
+    },
+    {
+      "epoch": 0.40926766040181467,
+      "grad_norm": 1.0094561576843262,
+      "learning_rate": 4.853030330032283e-06,
+      "loss": 0.1633,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4094296824368114,
+      "grad_norm": 1.0124329328536987,
+      "learning_rate": 4.8528825656656585e-06,
+      "loss": 0.158,
+      "step": 2527
+    },
+    {
+      "epoch": 0.40959170447180815,
+      "grad_norm": 0.9299939274787903,
+      "learning_rate": 4.852734729306523e-06,
+      "loss": 0.1398,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4097537265068049,
+      "grad_norm": 1.040216088294983,
+      "learning_rate": 4.852586820959398e-06,
+      "loss": 0.1592,
+      "step": 2529
+    },
+    {
+      "epoch": 0.4099157485418017,
+      "grad_norm": 1.0537763833999634,
+      "learning_rate": 4.852438840628808e-06,
+      "loss": 0.1607,
+      "step": 2530
+    },
+    {
+      "epoch": 0.41007777057679845,
+      "grad_norm": 0.9010412693023682,
+      "learning_rate": 4.852290788319284e-06,
+      "loss": 0.1365,
+      "step": 2531
+    },
+    {
+      "epoch": 0.4102397926117952,
+      "grad_norm": 0.9371026754379272,
+      "learning_rate": 4.852142664035353e-06,
+      "loss": 0.1374,
+      "step": 2532
+    },
+    {
+      "epoch": 0.410401814646792,
+      "grad_norm": 0.8150733113288879,
+      "learning_rate": 4.8519944677815495e-06,
+      "loss": 0.1214,
+      "step": 2533
+    },
+    {
+      "epoch": 0.4105638366817887,
+      "grad_norm": 1.083972454071045,
+      "learning_rate": 4.8518461995624064e-06,
+      "loss": 0.1656,
+      "step": 2534
+    },
+    {
+      "epoch": 0.41072585871678546,
+      "grad_norm": 0.9189485907554626,
+      "learning_rate": 4.851697859382461e-06,
+      "loss": 0.1295,
+      "step": 2535
+    },
+    {
+      "epoch": 0.41088788075178223,
+      "grad_norm": 1.0893805027008057,
+      "learning_rate": 4.851549447246253e-06,
+      "loss": 0.1678,
+      "step": 2536
+    },
+    {
+      "epoch": 0.411049902786779,
+      "grad_norm": 1.0611486434936523,
+      "learning_rate": 4.8514009631583215e-06,
+      "loss": 0.1484,
+      "step": 2537
+    },
+    {
+      "epoch": 0.41121192482177576,
+      "grad_norm": 1.0211429595947266,
+      "learning_rate": 4.851252407123211e-06,
+      "loss": 0.1585,
+      "step": 2538
+    },
+    {
+      "epoch": 0.41137394685677253,
+      "grad_norm": 1.0858979225158691,
+      "learning_rate": 4.851103779145467e-06,
+      "loss": 0.1493,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4115359688917693,
+      "grad_norm": 1.152523398399353,
+      "learning_rate": 4.850955079229637e-06,
+      "loss": 0.1583,
+      "step": 2540
+    },
+    {
+      "epoch": 0.41169799092676607,
+      "grad_norm": 1.043539047241211,
+      "learning_rate": 4.8508063073802715e-06,
+      "loss": 0.1549,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4118600129617628,
+      "grad_norm": 0.9617857933044434,
+      "learning_rate": 4.850657463601921e-06,
+      "loss": 0.1388,
+      "step": 2542
+    },
+    {
+      "epoch": 0.41202203499675955,
+      "grad_norm": 0.9477269649505615,
+      "learning_rate": 4.85050854789914e-06,
+      "loss": 0.1371,
+      "step": 2543
+    },
+    {
+      "epoch": 0.4121840570317563,
+      "grad_norm": 1.0493178367614746,
+      "learning_rate": 4.850359560276486e-06,
+      "loss": 0.1549,
+      "step": 2544
+    },
+    {
+      "epoch": 0.4123460790667531,
+      "grad_norm": 0.9363582134246826,
+      "learning_rate": 4.850210500738518e-06,
+      "loss": 0.1428,
+      "step": 2545
+    },
+    {
+      "epoch": 0.41250810110174985,
+      "grad_norm": 0.9234378337860107,
+      "learning_rate": 4.850061369289795e-06,
+      "loss": 0.1389,
+      "step": 2546
+    },
+    {
+      "epoch": 0.4126701231367466,
+      "grad_norm": 1.1004607677459717,
+      "learning_rate": 4.849912165934882e-06,
+      "loss": 0.166,
+      "step": 2547
+    },
+    {
+      "epoch": 0.4128321451717434,
+      "grad_norm": 1.073050856590271,
+      "learning_rate": 4.8497628906783425e-06,
+      "loss": 0.1658,
+      "step": 2548
+    },
+    {
+      "epoch": 0.4129941672067401,
+      "grad_norm": 1.1251699924468994,
+      "learning_rate": 4.849613543524746e-06,
+      "loss": 0.1618,
+      "step": 2549
+    },
+    {
+      "epoch": 0.41315618924173686,
+      "grad_norm": 0.9501916766166687,
+      "learning_rate": 4.84946412447866e-06,
+      "loss": 0.1468,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4133182112767336,
+      "grad_norm": 1.087240219116211,
+      "learning_rate": 4.849314633544659e-06,
+      "loss": 0.1596,
+      "step": 2551
+    },
+    {
+      "epoch": 0.4134802333117304,
+      "grad_norm": 1.0362616777420044,
+      "learning_rate": 4.849165070727313e-06,
+      "loss": 0.1707,
+      "step": 2552
+    },
+    {
+      "epoch": 0.41364225534672716,
+      "grad_norm": 0.9123900532722473,
+      "learning_rate": 4.849015436031202e-06,
+      "loss": 0.1363,
+      "step": 2553
+    },
+    {
+      "epoch": 0.41380427738172393,
+      "grad_norm": 0.9124921560287476,
+      "learning_rate": 4.848865729460903e-06,
+      "loss": 0.1288,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4139662994167207,
+      "grad_norm": 0.9334890842437744,
+      "learning_rate": 4.848715951020997e-06,
+      "loss": 0.1289,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4141283214517174,
+      "grad_norm": 1.121118187904358,
+      "learning_rate": 4.848566100716066e-06,
+      "loss": 0.1475,
+      "step": 2556
+    },
+    {
+      "epoch": 0.4142903434867142,
+      "grad_norm": 1.0138769149780273,
+      "learning_rate": 4.848416178550697e-06,
+      "loss": 0.1556,
+      "step": 2557
+    },
+    {
+      "epoch": 0.41445236552171094,
+      "grad_norm": 1.0010759830474854,
+      "learning_rate": 4.848266184529475e-06,
+      "loss": 0.151,
+      "step": 2558
+    },
+    {
+      "epoch": 0.4146143875567077,
+      "grad_norm": 1.0928181409835815,
+      "learning_rate": 4.848116118656991e-06,
+      "loss": 0.1721,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4147764095917045,
+      "grad_norm": 1.068526029586792,
+      "learning_rate": 4.847965980937836e-06,
+      "loss": 0.1644,
+      "step": 2560
+    },
+    {
+      "epoch": 0.41493843162670124,
+      "grad_norm": 1.0553364753723145,
+      "learning_rate": 4.847815771376604e-06,
+      "loss": 0.1501,
+      "step": 2561
+    },
+    {
+      "epoch": 0.415100453661698,
+      "grad_norm": 0.9555181860923767,
+      "learning_rate": 4.847665489977891e-06,
+      "loss": 0.1336,
+      "step": 2562
+    },
+    {
+      "epoch": 0.4152624756966948,
+      "grad_norm": 0.970905601978302,
+      "learning_rate": 4.847515136746295e-06,
+      "loss": 0.1426,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4154244977316915,
+      "grad_norm": 1.1288583278656006,
+      "learning_rate": 4.847364711686417e-06,
+      "loss": 0.1637,
+      "step": 2564
+    },
+    {
+      "epoch": 0.41558651976668826,
+      "grad_norm": 1.0185610055923462,
+      "learning_rate": 4.8472142148028585e-06,
+      "loss": 0.1561,
+      "step": 2565
+    },
+    {
+      "epoch": 0.415748541801685,
+      "grad_norm": 1.0335227251052856,
+      "learning_rate": 4.847063646100226e-06,
+      "loss": 0.1576,
+      "step": 2566
+    },
+    {
+      "epoch": 0.4159105638366818,
+      "grad_norm": 0.9629384875297546,
+      "learning_rate": 4.846913005583125e-06,
+      "loss": 0.1378,
+      "step": 2567
+    },
+    {
+      "epoch": 0.41607258587167856,
+      "grad_norm": 1.0787895917892456,
+      "learning_rate": 4.846762293256167e-06,
+      "loss": 0.1794,
+      "step": 2568
+    },
+    {
+      "epoch": 0.4162346079066753,
+      "grad_norm": 1.1242948770523071,
+      "learning_rate": 4.84661150912396e-06,
+      "loss": 0.1591,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4163966299416721,
+      "grad_norm": 0.9214066863059998,
+      "learning_rate": 4.846460653191121e-06,
+      "loss": 0.1397,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4165586519766688,
+      "grad_norm": 1.106182336807251,
+      "learning_rate": 4.846309725462264e-06,
+      "loss": 0.1657,
+      "step": 2571
+    },
+    {
+      "epoch": 0.41672067401166557,
+      "grad_norm": 1.0790849924087524,
+      "learning_rate": 4.846158725942006e-06,
+      "loss": 0.1565,
+      "step": 2572
+    },
+    {
+      "epoch": 0.41688269604666234,
+      "grad_norm": 1.0249879360198975,
+      "learning_rate": 4.84600765463497e-06,
+      "loss": 0.1579,
+      "step": 2573
+    },
+    {
+      "epoch": 0.4170447180816591,
+      "grad_norm": 1.0132466554641724,
+      "learning_rate": 4.845856511545777e-06,
+      "loss": 0.1618,
+      "step": 2574
+    },
+    {
+      "epoch": 0.41720674011665587,
+      "grad_norm": 1.0668222904205322,
+      "learning_rate": 4.845705296679051e-06,
+      "loss": 0.1762,
+      "step": 2575
+    },
+    {
+      "epoch": 0.41736876215165264,
+      "grad_norm": 1.0214135646820068,
+      "learning_rate": 4.84555401003942e-06,
+      "loss": 0.1508,
+      "step": 2576
+    },
+    {
+      "epoch": 0.4175307841866494,
+      "grad_norm": 0.97760009765625,
+      "learning_rate": 4.845402651631512e-06,
+      "loss": 0.1552,
+      "step": 2577
+    },
+    {
+      "epoch": 0.4176928062216461,
+      "grad_norm": 0.9772970080375671,
+      "learning_rate": 4.845251221459958e-06,
+      "loss": 0.1478,
+      "step": 2578
+    },
+    {
+      "epoch": 0.4178548282566429,
+      "grad_norm": 0.999789297580719,
+      "learning_rate": 4.845099719529393e-06,
+      "loss": 0.1563,
+      "step": 2579
+    },
+    {
+      "epoch": 0.41801685029163965,
+      "grad_norm": 1.1174688339233398,
+      "learning_rate": 4.844948145844452e-06,
+      "loss": 0.1767,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4181788723266364,
+      "grad_norm": 0.9995019435882568,
+      "learning_rate": 4.844796500409771e-06,
+      "loss": 0.1353,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4183408943616332,
+      "grad_norm": 0.9696946144104004,
+      "learning_rate": 4.844644783229993e-06,
+      "loss": 0.1431,
+      "step": 2582
+    },
+    {
+      "epoch": 0.41850291639662995,
+      "grad_norm": 1.0892887115478516,
+      "learning_rate": 4.844492994309757e-06,
+      "loss": 0.1597,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4186649384316267,
+      "grad_norm": 1.0031565427780151,
+      "learning_rate": 4.844341133653709e-06,
+      "loss": 0.1602,
+      "step": 2584
+    },
+    {
+      "epoch": 0.4188269604666235,
+      "grad_norm": 1.021321177482605,
+      "learning_rate": 4.844189201266497e-06,
+      "loss": 0.162,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4189889825016202,
+      "grad_norm": 1.1473654508590698,
+      "learning_rate": 4.844037197152767e-06,
+      "loss": 0.1854,
+      "step": 2586
+    },
+    {
+      "epoch": 0.41915100453661697,
+      "grad_norm": 1.1398388147354126,
+      "learning_rate": 4.8438851213171715e-06,
+      "loss": 0.1643,
+      "step": 2587
+    },
+    {
+      "epoch": 0.41931302657161373,
+      "grad_norm": 0.9572474956512451,
+      "learning_rate": 4.843732973764363e-06,
+      "loss": 0.1537,
+      "step": 2588
+    },
+    {
+      "epoch": 0.4194750486066105,
+      "grad_norm": 0.9969721436500549,
+      "learning_rate": 4.843580754498999e-06,
+      "loss": 0.1557,
+      "step": 2589
+    },
+    {
+      "epoch": 0.41963707064160727,
+      "grad_norm": 0.9691717624664307,
+      "learning_rate": 4.8434284635257335e-06,
+      "loss": 0.156,
+      "step": 2590
+    },
+    {
+      "epoch": 0.41979909267660404,
+      "grad_norm": 1.0504859685897827,
+      "learning_rate": 4.8432761008492284e-06,
+      "loss": 0.1416,
+      "step": 2591
+    },
+    {
+      "epoch": 0.4199611147116008,
+      "grad_norm": 1.0523217916488647,
+      "learning_rate": 4.843123666474146e-06,
+      "loss": 0.1704,
+      "step": 2592
+    },
+    {
+      "epoch": 0.4201231367465975,
+      "grad_norm": 1.1369895935058594,
+      "learning_rate": 4.842971160405149e-06,
+      "loss": 0.1514,
+      "step": 2593
+    },
+    {
+      "epoch": 0.4202851587815943,
+      "grad_norm": 1.0184272527694702,
+      "learning_rate": 4.842818582646904e-06,
+      "loss": 0.1512,
+      "step": 2594
+    },
+    {
+      "epoch": 0.42044718081659105,
+      "grad_norm": 1.0141220092773438,
+      "learning_rate": 4.84266593320408e-06,
+      "loss": 0.1563,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4206092028515878,
+      "grad_norm": 1.072980284690857,
+      "learning_rate": 4.842513212081348e-06,
+      "loss": 0.1629,
+      "step": 2596
+    },
+    {
+      "epoch": 0.4207712248865846,
+      "grad_norm": 1.1694176197052002,
+      "learning_rate": 4.842360419283381e-06,
+      "loss": 0.169,
+      "step": 2597
+    },
+    {
+      "epoch": 0.42093324692158135,
+      "grad_norm": 1.0028104782104492,
+      "learning_rate": 4.8422075548148525e-06,
+      "loss": 0.1462,
+      "step": 2598
+    },
+    {
+      "epoch": 0.4210952689565781,
+      "grad_norm": 0.9839292764663696,
+      "learning_rate": 4.84205461868044e-06,
+      "loss": 0.1486,
+      "step": 2599
+    },
+    {
+      "epoch": 0.42125729099157483,
+      "grad_norm": 0.9316514730453491,
+      "learning_rate": 4.841901610884826e-06,
+      "loss": 0.14,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4214193130265716,
+      "grad_norm": 1.1105822324752808,
+      "learning_rate": 4.8417485314326895e-06,
+      "loss": 0.1726,
+      "step": 2601
+    },
+    {
+      "epoch": 0.42158133506156836,
+      "grad_norm": 0.9990409016609192,
+      "learning_rate": 4.841595380328714e-06,
+      "loss": 0.1669,
+      "step": 2602
+    },
+    {
+      "epoch": 0.42174335709656513,
+      "grad_norm": 0.978840172290802,
+      "learning_rate": 4.841442157577587e-06,
+      "loss": 0.1542,
+      "step": 2603
+    },
+    {
+      "epoch": 0.4219053791315619,
+      "grad_norm": 1.0527600049972534,
+      "learning_rate": 4.841288863183996e-06,
+      "loss": 0.1514,
+      "step": 2604
+    },
+    {
+      "epoch": 0.42206740116655866,
+      "grad_norm": 1.0424004793167114,
+      "learning_rate": 4.8411354971526316e-06,
+      "loss": 0.1565,
+      "step": 2605
+    },
+    {
+      "epoch": 0.42222942320155543,
+      "grad_norm": 0.9171922206878662,
+      "learning_rate": 4.840982059488186e-06,
+      "loss": 0.1424,
+      "step": 2606
+    },
+    {
+      "epoch": 0.4223914452365522,
+      "grad_norm": 0.9491299986839294,
+      "learning_rate": 4.840828550195355e-06,
+      "loss": 0.14,
+      "step": 2607
+    },
+    {
+      "epoch": 0.4225534672715489,
+      "grad_norm": 1.1257144212722778,
+      "learning_rate": 4.840674969278836e-06,
+      "loss": 0.1776,
+      "step": 2608
+    },
+    {
+      "epoch": 0.4227154893065457,
+      "grad_norm": 0.9474332928657532,
+      "learning_rate": 4.840521316743326e-06,
+      "loss": 0.1463,
+      "step": 2609
+    },
+    {
+      "epoch": 0.42287751134154244,
+      "grad_norm": 1.0280464887619019,
+      "learning_rate": 4.8403675925935275e-06,
+      "loss": 0.1528,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4230395333765392,
+      "grad_norm": 1.010370135307312,
+      "learning_rate": 4.840213796834145e-06,
+      "loss": 0.1519,
+      "step": 2611
+    },
+    {
+      "epoch": 0.423201555411536,
+      "grad_norm": 1.0161646604537964,
+      "learning_rate": 4.8400599294698825e-06,
+      "loss": 0.1563,
+      "step": 2612
+    },
+    {
+      "epoch": 0.42336357744653275,
+      "grad_norm": 1.0096008777618408,
+      "learning_rate": 4.83990599050545e-06,
+      "loss": 0.1559,
+      "step": 2613
+    },
+    {
+      "epoch": 0.4235255994815295,
+      "grad_norm": 0.9370619654655457,
+      "learning_rate": 4.839751979945556e-06,
+      "loss": 0.122,
+      "step": 2614
+    },
+    {
+      "epoch": 0.4236876215165262,
+      "grad_norm": 0.9001135230064392,
+      "learning_rate": 4.839597897794915e-06,
+      "loss": 0.133,
+      "step": 2615
+    },
+    {
+      "epoch": 0.423849643551523,
+      "grad_norm": 0.9671546816825867,
+      "learning_rate": 4.839443744058238e-06,
+      "loss": 0.1472,
+      "step": 2616
+    },
+    {
+      "epoch": 0.42401166558651976,
+      "grad_norm": 0.9636013507843018,
+      "learning_rate": 4.839289518740245e-06,
+      "loss": 0.1435,
+      "step": 2617
+    },
+    {
+      "epoch": 0.4241736876215165,
+      "grad_norm": 1.114350438117981,
+      "learning_rate": 4.839135221845654e-06,
+      "loss": 0.1679,
+      "step": 2618
+    },
+    {
+      "epoch": 0.4243357096565133,
+      "grad_norm": 1.0020900964736938,
+      "learning_rate": 4.838980853379184e-06,
+      "loss": 0.1374,
+      "step": 2619
+    },
+    {
+      "epoch": 0.42449773169151006,
+      "grad_norm": 1.0408121347427368,
+      "learning_rate": 4.838826413345561e-06,
+      "loss": 0.1569,
+      "step": 2620
+    },
+    {
+      "epoch": 0.42465975372650683,
+      "grad_norm": 1.0487539768218994,
+      "learning_rate": 4.83867190174951e-06,
+      "loss": 0.1568,
+      "step": 2621
+    },
+    {
+      "epoch": 0.42482177576150354,
+      "grad_norm": 1.0748343467712402,
+      "learning_rate": 4.838517318595758e-06,
+      "loss": 0.1628,
+      "step": 2622
+    },
+    {
+      "epoch": 0.4249837977965003,
+      "grad_norm": 1.2027356624603271,
+      "learning_rate": 4.8383626638890355e-06,
+      "loss": 0.1863,
+      "step": 2623
+    },
+    {
+      "epoch": 0.4251458198314971,
+      "grad_norm": 1.07537841796875,
+      "learning_rate": 4.838207937634074e-06,
+      "loss": 0.1492,
+      "step": 2624
+    },
+    {
+      "epoch": 0.42530784186649384,
+      "grad_norm": 1.2119613885879517,
+      "learning_rate": 4.838053139835608e-06,
+      "loss": 0.1757,
+      "step": 2625
+    },
+    {
+      "epoch": 0.4254698639014906,
+      "grad_norm": 0.9143966436386108,
+      "learning_rate": 4.837898270498374e-06,
+      "loss": 0.1334,
+      "step": 2626
+    },
+    {
+      "epoch": 0.4256318859364874,
+      "grad_norm": 1.0104649066925049,
+      "learning_rate": 4.83774332962711e-06,
+      "loss": 0.1594,
+      "step": 2627
+    },
+    {
+      "epoch": 0.42579390797148414,
+      "grad_norm": 0.9676437377929688,
+      "learning_rate": 4.837588317226558e-06,
+      "loss": 0.1587,
+      "step": 2628
+    },
+    {
+      "epoch": 0.42595593000648085,
+      "grad_norm": 0.922566831111908,
+      "learning_rate": 4.837433233301461e-06,
+      "loss": 0.1529,
+      "step": 2629
+    },
+    {
+      "epoch": 0.4261179520414776,
+      "grad_norm": 1.1346608400344849,
+      "learning_rate": 4.837278077856562e-06,
+      "loss": 0.1793,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4262799740764744,
+      "grad_norm": 0.9662337303161621,
+      "learning_rate": 4.837122850896611e-06,
+      "loss": 0.1447,
+      "step": 2631
+    },
+    {
+      "epoch": 0.42644199611147116,
+      "grad_norm": 1.0667489767074585,
+      "learning_rate": 4.836967552426355e-06,
+      "loss": 0.1784,
+      "step": 2632
+    },
+    {
+      "epoch": 0.4266040181464679,
+      "grad_norm": 1.030396580696106,
+      "learning_rate": 4.836812182450549e-06,
+      "loss": 0.1485,
+      "step": 2633
+    },
+    {
+      "epoch": 0.4267660401814647,
+      "grad_norm": 0.9403509497642517,
+      "learning_rate": 4.836656740973944e-06,
+      "loss": 0.1476,
+      "step": 2634
+    },
+    {
+      "epoch": 0.42692806221646146,
+      "grad_norm": 1.0584341287612915,
+      "learning_rate": 4.836501228001298e-06,
+      "loss": 0.1499,
+      "step": 2635
+    },
+    {
+      "epoch": 0.4270900842514582,
+      "grad_norm": 1.0108885765075684,
+      "learning_rate": 4.836345643537368e-06,
+      "loss": 0.1485,
+      "step": 2636
+    },
+    {
+      "epoch": 0.42725210628645494,
+      "grad_norm": 0.9687297940254211,
+      "learning_rate": 4.8361899875869165e-06,
+      "loss": 0.1516,
+      "step": 2637
+    },
+    {
+      "epoch": 0.4274141283214517,
+      "grad_norm": 0.9748333096504211,
+      "learning_rate": 4.836034260154704e-06,
+      "loss": 0.1523,
+      "step": 2638
+    },
+    {
+      "epoch": 0.42757615035644847,
+      "grad_norm": 1.1141382455825806,
+      "learning_rate": 4.835878461245496e-06,
+      "loss": 0.1485,
+      "step": 2639
+    },
+    {
+      "epoch": 0.42773817239144524,
+      "grad_norm": 0.8912030458450317,
+      "learning_rate": 4.83572259086406e-06,
+      "loss": 0.133,
+      "step": 2640
+    },
+    {
+      "epoch": 0.427900194426442,
+      "grad_norm": 1.0493932962417603,
+      "learning_rate": 4.835566649015165e-06,
+      "loss": 0.1706,
+      "step": 2641
+    },
+    {
+      "epoch": 0.42806221646143877,
+      "grad_norm": 1.024950385093689,
+      "learning_rate": 4.835410635703582e-06,
+      "loss": 0.1362,
+      "step": 2642
+    },
+    {
+      "epoch": 0.42822423849643554,
+      "grad_norm": 1.094906210899353,
+      "learning_rate": 4.8352545509340865e-06,
+      "loss": 0.1505,
+      "step": 2643
+    },
+    {
+      "epoch": 0.42838626053143225,
+      "grad_norm": 1.0383245944976807,
+      "learning_rate": 4.835098394711451e-06,
+      "loss": 0.1569,
+      "step": 2644
+    },
+    {
+      "epoch": 0.428548282566429,
+      "grad_norm": 0.9308538436889648,
+      "learning_rate": 4.834942167040457e-06,
+      "loss": 0.131,
+      "step": 2645
+    },
+    {
+      "epoch": 0.4287103046014258,
+      "grad_norm": 0.9128456711769104,
+      "learning_rate": 4.834785867925883e-06,
+      "loss": 0.1431,
+      "step": 2646
+    },
+    {
+      "epoch": 0.42887232663642255,
+      "grad_norm": 1.047548770904541,
+      "learning_rate": 4.8346294973725115e-06,
+      "loss": 0.1592,
+      "step": 2647
+    },
+    {
+      "epoch": 0.4290343486714193,
+      "grad_norm": 0.9378132820129395,
+      "learning_rate": 4.8344730553851275e-06,
+      "loss": 0.146,
+      "step": 2648
+    },
+    {
+      "epoch": 0.4291963707064161,
+      "grad_norm": 1.0160073041915894,
+      "learning_rate": 4.8343165419685155e-06,
+      "loss": 0.1598,
+      "step": 2649
+    },
+    {
+      "epoch": 0.42935839274141285,
+      "grad_norm": 1.0141408443450928,
+      "learning_rate": 4.834159957127468e-06,
+      "loss": 0.1496,
+      "step": 2650
+    },
+    {
+      "epoch": 0.42952041477640956,
+      "grad_norm": 1.0546174049377441,
+      "learning_rate": 4.834003300866773e-06,
+      "loss": 0.1479,
+      "step": 2651
+    },
+    {
+      "epoch": 0.42968243681140633,
+      "grad_norm": 1.0627154111862183,
+      "learning_rate": 4.833846573191227e-06,
+      "loss": 0.1585,
+      "step": 2652
+    },
+    {
+      "epoch": 0.4298444588464031,
+      "grad_norm": 1.0541164875030518,
+      "learning_rate": 4.833689774105622e-06,
+      "loss": 0.1551,
+      "step": 2653
+    },
+    {
+      "epoch": 0.43000648088139987,
+      "grad_norm": 0.9814350008964539,
+      "learning_rate": 4.833532903614758e-06,
+      "loss": 0.1397,
+      "step": 2654
+    },
+    {
+      "epoch": 0.43016850291639663,
+      "grad_norm": 1.0309319496154785,
+      "learning_rate": 4.8333759617234344e-06,
+      "loss": 0.1603,
+      "step": 2655
+    },
+    {
+      "epoch": 0.4303305249513934,
+      "grad_norm": 1.044171929359436,
+      "learning_rate": 4.833218948436453e-06,
+      "loss": 0.1588,
+      "step": 2656
+    },
+    {
+      "epoch": 0.43049254698639017,
+      "grad_norm": 1.0367193222045898,
+      "learning_rate": 4.833061863758618e-06,
+      "loss": 0.1734,
+      "step": 2657
+    },
+    {
+      "epoch": 0.43065456902138693,
+      "grad_norm": 0.9801927804946899,
+      "learning_rate": 4.832904707694736e-06,
+      "loss": 0.1459,
+      "step": 2658
+    },
+    {
+      "epoch": 0.43081659105638365,
+      "grad_norm": 0.9928318858146667,
+      "learning_rate": 4.8327474802496145e-06,
+      "loss": 0.1478,
+      "step": 2659
+    },
+    {
+      "epoch": 0.4309786130913804,
+      "grad_norm": 0.9557510614395142,
+      "learning_rate": 4.832590181428066e-06,
+      "loss": 0.1511,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4311406351263772,
+      "grad_norm": 1.0317496061325073,
+      "learning_rate": 4.832432811234902e-06,
+      "loss": 0.1646,
+      "step": 2661
+    },
+    {
+      "epoch": 0.43130265716137395,
+      "grad_norm": 1.0222221612930298,
+      "learning_rate": 4.832275369674939e-06,
+      "loss": 0.1483,
+      "step": 2662
+    },
+    {
+      "epoch": 0.4314646791963707,
+      "grad_norm": 0.9168278574943542,
+      "learning_rate": 4.832117856752994e-06,
+      "loss": 0.1329,
+      "step": 2663
+    },
+    {
+      "epoch": 0.4316267012313675,
+      "grad_norm": 0.9790104031562805,
+      "learning_rate": 4.831960272473886e-06,
+      "loss": 0.1395,
+      "step": 2664
+    },
+    {
+      "epoch": 0.43178872326636425,
+      "grad_norm": 0.9434686303138733,
+      "learning_rate": 4.831802616842436e-06,
+      "loss": 0.138,
+      "step": 2665
+    },
+    {
+      "epoch": 0.43195074530136096,
+      "grad_norm": 1.0316851139068604,
+      "learning_rate": 4.831644889863471e-06,
+      "loss": 0.1684,
+      "step": 2666
+    },
+    {
+      "epoch": 0.43211276733635773,
+      "grad_norm": 1.1002922058105469,
+      "learning_rate": 4.831487091541812e-06,
+      "loss": 0.151,
+      "step": 2667
+    },
+    {
+      "epoch": 0.4322747893713545,
+      "grad_norm": 1.2835469245910645,
+      "learning_rate": 4.831329221882291e-06,
+      "loss": 0.1845,
+      "step": 2668
+    },
+    {
+      "epoch": 0.43243681140635126,
+      "grad_norm": 1.0473268032073975,
+      "learning_rate": 4.831171280889739e-06,
+      "loss": 0.1447,
+      "step": 2669
+    },
+    {
+      "epoch": 0.43259883344134803,
+      "grad_norm": 1.0189570188522339,
+      "learning_rate": 4.831013268568986e-06,
+      "loss": 0.1545,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4327608554763448,
+      "grad_norm": 1.0731713771820068,
+      "learning_rate": 4.830855184924868e-06,
+      "loss": 0.1656,
+      "step": 2671
+    },
+    {
+      "epoch": 0.43292287751134156,
+      "grad_norm": 0.955833911895752,
+      "learning_rate": 4.830697029962222e-06,
+      "loss": 0.1447,
+      "step": 2672
+    },
+    {
+      "epoch": 0.4330848995463383,
+      "grad_norm": 1.0004793405532837,
+      "learning_rate": 4.830538803685887e-06,
+      "loss": 0.1471,
+      "step": 2673
+    },
+    {
+      "epoch": 0.43324692158133504,
+      "grad_norm": 1.0488044023513794,
+      "learning_rate": 4.830380506100704e-06,
+      "loss": 0.1557,
+      "step": 2674
+    },
+    {
+      "epoch": 0.4334089436163318,
+      "grad_norm": 1.1116876602172852,
+      "learning_rate": 4.830222137211518e-06,
+      "loss": 0.1584,
+      "step": 2675
+    },
+    {
+      "epoch": 0.4335709656513286,
+      "grad_norm": 0.9907661080360413,
+      "learning_rate": 4.830063697023173e-06,
+      "loss": 0.1414,
+      "step": 2676
+    },
+    {
+      "epoch": 0.43373298768632534,
+      "grad_norm": 1.0337917804718018,
+      "learning_rate": 4.829905185540517e-06,
+      "loss": 0.1692,
+      "step": 2677
+    },
+    {
+      "epoch": 0.4338950097213221,
+      "grad_norm": 0.9875041246414185,
+      "learning_rate": 4.829746602768401e-06,
+      "loss": 0.1462,
+      "step": 2678
+    },
+    {
+      "epoch": 0.4340570317563189,
+      "grad_norm": 1.0087636709213257,
+      "learning_rate": 4.829587948711677e-06,
+      "loss": 0.1484,
+      "step": 2679
+    },
+    {
+      "epoch": 0.43421905379131565,
+      "grad_norm": 1.0348845720291138,
+      "learning_rate": 4.8294292233752e-06,
+      "loss": 0.1457,
+      "step": 2680
+    },
+    {
+      "epoch": 0.43438107582631236,
+      "grad_norm": 0.8487588167190552,
+      "learning_rate": 4.829270426763824e-06,
+      "loss": 0.1168,
+      "step": 2681
+    },
+    {
+      "epoch": 0.4345430978613091,
+      "grad_norm": 1.086271047592163,
+      "learning_rate": 4.829111558882411e-06,
+      "loss": 0.1501,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4347051198963059,
+      "grad_norm": 0.9329110980033875,
+      "learning_rate": 4.828952619735821e-06,
+      "loss": 0.1482,
+      "step": 2683
+    },
+    {
+      "epoch": 0.43486714193130266,
+      "grad_norm": 1.0108580589294434,
+      "learning_rate": 4.828793609328916e-06,
+      "loss": 0.1457,
+      "step": 2684
+    },
+    {
+      "epoch": 0.4350291639662994,
+      "grad_norm": 1.005904197692871,
+      "learning_rate": 4.828634527666562e-06,
+      "loss": 0.1391,
+      "step": 2685
+    },
+    {
+      "epoch": 0.4351911860012962,
+      "grad_norm": 0.9748127460479736,
+      "learning_rate": 4.828475374753627e-06,
+      "loss": 0.1528,
+      "step": 2686
+    },
+    {
+      "epoch": 0.43535320803629296,
+      "grad_norm": 1.1028996706008911,
+      "learning_rate": 4.82831615059498e-06,
+      "loss": 0.1504,
+      "step": 2687
+    },
+    {
+      "epoch": 0.43551523007128967,
+      "grad_norm": 1.1191354990005493,
+      "learning_rate": 4.828156855195493e-06,
+      "loss": 0.1613,
+      "step": 2688
+    },
+    {
+      "epoch": 0.43567725210628644,
+      "grad_norm": 0.9752424955368042,
+      "learning_rate": 4.827997488560041e-06,
+      "loss": 0.1479,
+      "step": 2689
+    },
+    {
+      "epoch": 0.4358392741412832,
+      "grad_norm": 1.1131165027618408,
+      "learning_rate": 4.827838050693499e-06,
+      "loss": 0.1846,
+      "step": 2690
+    },
+    {
+      "epoch": 0.43600129617628,
+      "grad_norm": 0.9194788336753845,
+      "learning_rate": 4.827678541600747e-06,
+      "loss": 0.1299,
+      "step": 2691
+    },
+    {
+      "epoch": 0.43616331821127674,
+      "grad_norm": 1.0693033933639526,
+      "learning_rate": 4.827518961286663e-06,
+      "loss": 0.1686,
+      "step": 2692
+    },
+    {
+      "epoch": 0.4363253402462735,
+      "grad_norm": 1.0003294944763184,
+      "learning_rate": 4.827359309756132e-06,
+      "loss": 0.1645,
+      "step": 2693
+    },
+    {
+      "epoch": 0.4364873622812703,
+      "grad_norm": 1.1136189699172974,
+      "learning_rate": 4.827199587014038e-06,
+      "loss": 0.1632,
+      "step": 2694
+    },
+    {
+      "epoch": 0.436649384316267,
+      "grad_norm": 0.9603433012962341,
+      "learning_rate": 4.8270397930652685e-06,
+      "loss": 0.1411,
+      "step": 2695
+    },
+    {
+      "epoch": 0.43681140635126375,
+      "grad_norm": 1.076905369758606,
+      "learning_rate": 4.826879927914713e-06,
+      "loss": 0.1593,
+      "step": 2696
+    },
+    {
+      "epoch": 0.4369734283862605,
+      "grad_norm": 0.9769685864448547,
+      "learning_rate": 4.826719991567262e-06,
+      "loss": 0.1612,
+      "step": 2697
+    },
+    {
+      "epoch": 0.4371354504212573,
+      "grad_norm": 0.9933087825775146,
+      "learning_rate": 4.82655998402781e-06,
+      "loss": 0.1777,
+      "step": 2698
+    },
+    {
+      "epoch": 0.43729747245625405,
+      "grad_norm": 0.9159439206123352,
+      "learning_rate": 4.826399905301252e-06,
+      "loss": 0.1424,
+      "step": 2699
+    },
+    {
+      "epoch": 0.4374594944912508,
+      "grad_norm": 0.9449082016944885,
+      "learning_rate": 4.826239755392488e-06,
+      "loss": 0.1445,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4376215165262476,
+      "grad_norm": 1.004356861114502,
+      "learning_rate": 4.826079534306417e-06,
+      "loss": 0.1589,
+      "step": 2701
+    },
+    {
+      "epoch": 0.43778353856124436,
+      "grad_norm": 1.0486918687820435,
+      "learning_rate": 4.8259192420479395e-06,
+      "loss": 0.1625,
+      "step": 2702
+    },
+    {
+      "epoch": 0.43794556059624107,
+      "grad_norm": 1.0001510381698608,
+      "learning_rate": 4.825758878621963e-06,
+      "loss": 0.1594,
+      "step": 2703
+    },
+    {
+      "epoch": 0.43810758263123784,
+      "grad_norm": 1.1367384195327759,
+      "learning_rate": 4.825598444033393e-06,
+      "loss": 0.1772,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4382696046662346,
+      "grad_norm": 1.0560551881790161,
+      "learning_rate": 4.825437938287139e-06,
+      "loss": 0.164,
+      "step": 2705
+    },
+    {
+      "epoch": 0.43843162670123137,
+      "grad_norm": 0.9951310157775879,
+      "learning_rate": 4.82527736138811e-06,
+      "loss": 0.1679,
+      "step": 2706
+    },
+    {
+      "epoch": 0.43859364873622814,
+      "grad_norm": 1.0460946559906006,
+      "learning_rate": 4.825116713341223e-06,
+      "loss": 0.1679,
+      "step": 2707
+    },
+    {
+      "epoch": 0.4387556707712249,
+      "grad_norm": 1.0097354650497437,
+      "learning_rate": 4.824955994151389e-06,
+      "loss": 0.1549,
+      "step": 2708
+    },
+    {
+      "epoch": 0.43891769280622167,
+      "grad_norm": 0.9873436093330383,
+      "learning_rate": 4.824795203823529e-06,
+      "loss": 0.1522,
+      "step": 2709
+    },
+    {
+      "epoch": 0.4390797148412184,
+      "grad_norm": 0.9897306561470032,
+      "learning_rate": 4.824634342362561e-06,
+      "loss": 0.1603,
+      "step": 2710
+    },
+    {
+      "epoch": 0.43924173687621515,
+      "grad_norm": 1.0298486948013306,
+      "learning_rate": 4.824473409773408e-06,
+      "loss": 0.1711,
+      "step": 2711
+    },
+    {
+      "epoch": 0.4394037589112119,
+      "grad_norm": 1.0957099199295044,
+      "learning_rate": 4.824312406060995e-06,
+      "loss": 0.1806,
+      "step": 2712
+    },
+    {
+      "epoch": 0.4395657809462087,
+      "grad_norm": 0.9915971755981445,
+      "learning_rate": 4.824151331230245e-06,
+      "loss": 0.1612,
+      "step": 2713
+    },
+    {
+      "epoch": 0.43972780298120545,
+      "grad_norm": 0.9229410886764526,
+      "learning_rate": 4.82399018528609e-06,
+      "loss": 0.1444,
+      "step": 2714
+    },
+    {
+      "epoch": 0.4398898250162022,
+      "grad_norm": 1.0651377439498901,
+      "learning_rate": 4.823828968233459e-06,
+      "loss": 0.1627,
+      "step": 2715
+    },
+    {
+      "epoch": 0.440051847051199,
+      "grad_norm": 0.9680843353271484,
+      "learning_rate": 4.823667680077285e-06,
+      "loss": 0.1354,
+      "step": 2716
+    },
+    {
+      "epoch": 0.4402138690861957,
+      "grad_norm": 1.1376385688781738,
+      "learning_rate": 4.823506320822503e-06,
+      "loss": 0.1696,
+      "step": 2717
+    },
+    {
+      "epoch": 0.44037589112119246,
+      "grad_norm": 0.9446072578430176,
+      "learning_rate": 4.8233448904740505e-06,
+      "loss": 0.1395,
+      "step": 2718
+    },
+    {
+      "epoch": 0.44053791315618923,
+      "grad_norm": 1.1791229248046875,
+      "learning_rate": 4.823183389036867e-06,
+      "loss": 0.1819,
+      "step": 2719
+    },
+    {
+      "epoch": 0.440699935191186,
+      "grad_norm": 1.0275377035140991,
+      "learning_rate": 4.823021816515893e-06,
+      "loss": 0.1589,
+      "step": 2720
+    },
+    {
+      "epoch": 0.44086195722618277,
+      "grad_norm": 1.0309733152389526,
+      "learning_rate": 4.822860172916074e-06,
+      "loss": 0.1415,
+      "step": 2721
+    },
+    {
+      "epoch": 0.44102397926117953,
+      "grad_norm": 1.0780889987945557,
+      "learning_rate": 4.8226984582423545e-06,
+      "loss": 0.1693,
+      "step": 2722
+    },
+    {
+      "epoch": 0.4411860012961763,
+      "grad_norm": 1.0635453462600708,
+      "learning_rate": 4.8225366724996826e-06,
+      "loss": 0.1393,
+      "step": 2723
+    },
+    {
+      "epoch": 0.44134802333117307,
+      "grad_norm": 0.916522204875946,
+      "learning_rate": 4.82237481569301e-06,
+      "loss": 0.1435,
+      "step": 2724
+    },
+    {
+      "epoch": 0.4415100453661698,
+      "grad_norm": 0.949371337890625,
+      "learning_rate": 4.822212887827287e-06,
+      "loss": 0.1441,
+      "step": 2725
+    },
+    {
+      "epoch": 0.44167206740116655,
+      "grad_norm": 0.9767130613327026,
+      "learning_rate": 4.822050888907469e-06,
+      "loss": 0.1539,
+      "step": 2726
+    },
+    {
+      "epoch": 0.4418340894361633,
+      "grad_norm": 0.9982846975326538,
+      "learning_rate": 4.8218888189385145e-06,
+      "loss": 0.1326,
+      "step": 2727
+    },
+    {
+      "epoch": 0.4419961114711601,
+      "grad_norm": 0.99046790599823,
+      "learning_rate": 4.82172667792538e-06,
+      "loss": 0.1545,
+      "step": 2728
+    },
+    {
+      "epoch": 0.44215813350615685,
+      "grad_norm": 0.9805136919021606,
+      "learning_rate": 4.821564465873027e-06,
+      "loss": 0.1505,
+      "step": 2729
+    },
+    {
+      "epoch": 0.4423201555411536,
+      "grad_norm": 0.9637601375579834,
+      "learning_rate": 4.821402182786421e-06,
+      "loss": 0.1468,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4424821775761504,
+      "grad_norm": 1.0324764251708984,
+      "learning_rate": 4.821239828670525e-06,
+      "loss": 0.1655,
+      "step": 2731
+    },
+    {
+      "epoch": 0.4426441996111471,
+      "grad_norm": 1.0406447649002075,
+      "learning_rate": 4.8210774035303085e-06,
+      "loss": 0.1528,
+      "step": 2732
+    },
+    {
+      "epoch": 0.44280622164614386,
+      "grad_norm": 1.000382661819458,
+      "learning_rate": 4.820914907370739e-06,
+      "loss": 0.1424,
+      "step": 2733
+    },
+    {
+      "epoch": 0.4429682436811406,
+      "grad_norm": 1.0306026935577393,
+      "learning_rate": 4.82075234019679e-06,
+      "loss": 0.1612,
+      "step": 2734
+    },
+    {
+      "epoch": 0.4431302657161374,
+      "grad_norm": 0.9841228723526001,
+      "learning_rate": 4.820589702013436e-06,
+      "loss": 0.1515,
+      "step": 2735
+    },
+    {
+      "epoch": 0.44329228775113416,
+      "grad_norm": 0.9914823770523071,
+      "learning_rate": 4.820426992825653e-06,
+      "loss": 0.1586,
+      "step": 2736
+    },
+    {
+      "epoch": 0.44345430978613093,
+      "grad_norm": 1.0662131309509277,
+      "learning_rate": 4.820264212638419e-06,
+      "loss": 0.1617,
+      "step": 2737
+    },
+    {
+      "epoch": 0.4436163318211277,
+      "grad_norm": 1.1415921449661255,
+      "learning_rate": 4.820101361456715e-06,
+      "loss": 0.1786,
+      "step": 2738
+    },
+    {
+      "epoch": 0.4437783538561244,
+      "grad_norm": 1.0659246444702148,
+      "learning_rate": 4.819938439285524e-06,
+      "loss": 0.1655,
+      "step": 2739
+    },
+    {
+      "epoch": 0.4439403758911212,
+      "grad_norm": 1.2058014869689941,
+      "learning_rate": 4.819775446129832e-06,
+      "loss": 0.1672,
+      "step": 2740
+    },
+    {
+      "epoch": 0.44410239792611794,
+      "grad_norm": 1.0413726568222046,
+      "learning_rate": 4.819612381994624e-06,
+      "loss": 0.1361,
+      "step": 2741
+    },
+    {
+      "epoch": 0.4442644199611147,
+      "grad_norm": 0.9703432321548462,
+      "learning_rate": 4.8194492468848895e-06,
+      "loss": 0.156,
+      "step": 2742
+    },
+    {
+      "epoch": 0.4444264419961115,
+      "grad_norm": 1.1094372272491455,
+      "learning_rate": 4.819286040805622e-06,
+      "loss": 0.1517,
+      "step": 2743
+    },
+    {
+      "epoch": 0.44458846403110824,
+      "grad_norm": 1.0455727577209473,
+      "learning_rate": 4.8191227637618145e-06,
+      "loss": 0.1484,
+      "step": 2744
+    },
+    {
+      "epoch": 0.444750486066105,
+      "grad_norm": 0.9904317855834961,
+      "learning_rate": 4.818959415758463e-06,
+      "loss": 0.1528,
+      "step": 2745
+    },
+    {
+      "epoch": 0.4449125081011017,
+      "grad_norm": 0.9994639158248901,
+      "learning_rate": 4.818795996800564e-06,
+      "loss": 0.1466,
+      "step": 2746
+    },
+    {
+      "epoch": 0.4450745301360985,
+      "grad_norm": 1.0804579257965088,
+      "learning_rate": 4.818632506893119e-06,
+      "loss": 0.1715,
+      "step": 2747
+    },
+    {
+      "epoch": 0.44523655217109526,
+      "grad_norm": 1.024933099746704,
+      "learning_rate": 4.8184689460411306e-06,
+      "loss": 0.1411,
+      "step": 2748
+    },
+    {
+      "epoch": 0.445398574206092,
+      "grad_norm": 1.01506507396698,
+      "learning_rate": 4.8183053142496025e-06,
+      "loss": 0.1706,
+      "step": 2749
+    },
+    {
+      "epoch": 0.4455605962410888,
+      "grad_norm": 0.9807841777801514,
+      "learning_rate": 4.818141611523543e-06,
+      "loss": 0.1517,
+      "step": 2750
+    },
+    {
+      "epoch": 0.44572261827608556,
+      "grad_norm": 0.8835573196411133,
+      "learning_rate": 4.81797783786796e-06,
+      "loss": 0.1406,
+      "step": 2751
+    },
+    {
+      "epoch": 0.4458846403110823,
+      "grad_norm": 0.9974625110626221,
+      "learning_rate": 4.817813993287863e-06,
+      "loss": 0.1729,
+      "step": 2752
+    },
+    {
+      "epoch": 0.4460466623460791,
+      "grad_norm": 1.12760329246521,
+      "learning_rate": 4.817650077788268e-06,
+      "loss": 0.1616,
+      "step": 2753
+    },
+    {
+      "epoch": 0.4462086843810758,
+      "grad_norm": 1.0579090118408203,
+      "learning_rate": 4.817486091374189e-06,
+      "loss": 0.138,
+      "step": 2754
+    },
+    {
+      "epoch": 0.44637070641607257,
+      "grad_norm": 0.9531564116477966,
+      "learning_rate": 4.817322034050645e-06,
+      "loss": 0.1276,
+      "step": 2755
+    },
+    {
+      "epoch": 0.44653272845106934,
+      "grad_norm": 0.9502880573272705,
+      "learning_rate": 4.817157905822652e-06,
+      "loss": 0.1437,
+      "step": 2756
+    },
+    {
+      "epoch": 0.4466947504860661,
+      "grad_norm": 1.0466476678848267,
+      "learning_rate": 4.816993706695237e-06,
+      "loss": 0.1648,
+      "step": 2757
+    },
+    {
+      "epoch": 0.4468567725210629,
+      "grad_norm": 1.0479638576507568,
+      "learning_rate": 4.816829436673421e-06,
+      "loss": 0.1539,
+      "step": 2758
+    },
+    {
+      "epoch": 0.44701879455605964,
+      "grad_norm": 1.0833531618118286,
+      "learning_rate": 4.81666509576223e-06,
+      "loss": 0.1675,
+      "step": 2759
+    },
+    {
+      "epoch": 0.4471808165910564,
+      "grad_norm": 0.9449465274810791,
+      "learning_rate": 4.816500683966694e-06,
+      "loss": 0.1481,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4473428386260531,
+      "grad_norm": 1.0143202543258667,
+      "learning_rate": 4.816336201291842e-06,
+      "loss": 0.152,
+      "step": 2761
+    },
+    {
+      "epoch": 0.4475048606610499,
+      "grad_norm": 1.0159945487976074,
+      "learning_rate": 4.816171647742708e-06,
+      "loss": 0.1393,
+      "step": 2762
+    },
+    {
+      "epoch": 0.44766688269604665,
+      "grad_norm": 1.0358235836029053,
+      "learning_rate": 4.816007023324327e-06,
+      "loss": 0.1563,
+      "step": 2763
+    },
+    {
+      "epoch": 0.4478289047310434,
+      "grad_norm": 0.9171221852302551,
+      "learning_rate": 4.815842328041736e-06,
+      "loss": 0.1449,
+      "step": 2764
+    },
+    {
+      "epoch": 0.4479909267660402,
+      "grad_norm": 1.0671533346176147,
+      "learning_rate": 4.815677561899973e-06,
+      "loss": 0.1561,
+      "step": 2765
+    },
+    {
+      "epoch": 0.44815294880103695,
+      "grad_norm": 1.0144374370574951,
+      "learning_rate": 4.815512724904081e-06,
+      "loss": 0.1648,
+      "step": 2766
+    },
+    {
+      "epoch": 0.4483149708360337,
+      "grad_norm": 0.9934388399124146,
+      "learning_rate": 4.815347817059103e-06,
+      "loss": 0.143,
+      "step": 2767
+    },
+    {
+      "epoch": 0.44847699287103043,
+      "grad_norm": 0.9061841368675232,
+      "learning_rate": 4.815182838370085e-06,
+      "loss": 0.139,
+      "step": 2768
+    },
+    {
+      "epoch": 0.4486390149060272,
+      "grad_norm": 1.0116674900054932,
+      "learning_rate": 4.815017788842075e-06,
+      "loss": 0.1504,
+      "step": 2769
+    },
+    {
+      "epoch": 0.44880103694102397,
+      "grad_norm": 1.012621283531189,
+      "learning_rate": 4.814852668480122e-06,
+      "loss": 0.164,
+      "step": 2770
+    },
+    {
+      "epoch": 0.44896305897602073,
+      "grad_norm": 0.9846920967102051,
+      "learning_rate": 4.81468747728928e-06,
+      "loss": 0.1494,
+      "step": 2771
+    },
+    {
+      "epoch": 0.4491250810110175,
+      "grad_norm": 0.9715894460678101,
+      "learning_rate": 4.814522215274603e-06,
+      "loss": 0.1446,
+      "step": 2772
+    },
+    {
+      "epoch": 0.44928710304601427,
+      "grad_norm": 1.0353641510009766,
+      "learning_rate": 4.814356882441147e-06,
+      "loss": 0.1578,
+      "step": 2773
+    },
+    {
+      "epoch": 0.44944912508101104,
+      "grad_norm": 0.9785324931144714,
+      "learning_rate": 4.81419147879397e-06,
+      "loss": 0.1495,
+      "step": 2774
+    },
+    {
+      "epoch": 0.4496111471160078,
+      "grad_norm": 1.190711259841919,
+      "learning_rate": 4.814026004338135e-06,
+      "loss": 0.1844,
+      "step": 2775
+    },
+    {
+      "epoch": 0.4497731691510045,
+      "grad_norm": 1.0003420114517212,
+      "learning_rate": 4.813860459078703e-06,
+      "loss": 0.1485,
+      "step": 2776
+    },
+    {
+      "epoch": 0.4499351911860013,
+      "grad_norm": 0.9324052929878235,
+      "learning_rate": 4.8136948430207415e-06,
+      "loss": 0.1421,
+      "step": 2777
+    },
+    {
+      "epoch": 0.45009721322099805,
+      "grad_norm": 1.0191559791564941,
+      "learning_rate": 4.813529156169317e-06,
+      "loss": 0.1593,
+      "step": 2778
+    },
+    {
+      "epoch": 0.4502592352559948,
+      "grad_norm": 1.0456205606460571,
+      "learning_rate": 4.813363398529498e-06,
+      "loss": 0.1552,
+      "step": 2779
+    },
+    {
+      "epoch": 0.4504212572909916,
+      "grad_norm": 0.9817255735397339,
+      "learning_rate": 4.813197570106357e-06,
+      "loss": 0.1554,
+      "step": 2780
+    },
+    {
+      "epoch": 0.45058327932598835,
+      "grad_norm": 1.0291125774383545,
+      "learning_rate": 4.813031670904969e-06,
+      "loss": 0.1549,
+      "step": 2781
+    },
+    {
+      "epoch": 0.4507453013609851,
+      "grad_norm": 0.9215115904808044,
+      "learning_rate": 4.8128657009304096e-06,
+      "loss": 0.1342,
+      "step": 2782
+    },
+    {
+      "epoch": 0.45090732339598183,
+      "grad_norm": 0.9075199961662292,
+      "learning_rate": 4.8126996601877564e-06,
+      "loss": 0.1471,
+      "step": 2783
+    },
+    {
+      "epoch": 0.4510693454309786,
+      "grad_norm": 0.9484758377075195,
+      "learning_rate": 4.8125335486820905e-06,
+      "loss": 0.1392,
+      "step": 2784
+    },
+    {
+      "epoch": 0.45123136746597536,
+      "grad_norm": 1.10005521774292,
+      "learning_rate": 4.812367366418493e-06,
+      "loss": 0.162,
+      "step": 2785
+    },
+    {
+      "epoch": 0.45139338950097213,
+      "grad_norm": 0.9598970413208008,
+      "learning_rate": 4.8122011134020505e-06,
+      "loss": 0.1256,
+      "step": 2786
+    },
+    {
+      "epoch": 0.4515554115359689,
+      "grad_norm": 1.010698914527893,
+      "learning_rate": 4.81203478963785e-06,
+      "loss": 0.1521,
+      "step": 2787
+    },
+    {
+      "epoch": 0.45171743357096567,
+      "grad_norm": 0.8810874819755554,
+      "learning_rate": 4.8118683951309795e-06,
+      "loss": 0.1325,
+      "step": 2788
+    },
+    {
+      "epoch": 0.45187945560596243,
+      "grad_norm": 1.020269751548767,
+      "learning_rate": 4.811701929886531e-06,
+      "loss": 0.1457,
+      "step": 2789
+    },
+    {
+      "epoch": 0.45204147764095914,
+      "grad_norm": 0.9705826640129089,
+      "learning_rate": 4.811535393909598e-06,
+      "loss": 0.1452,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4522034996759559,
+      "grad_norm": 1.204943060874939,
+      "learning_rate": 4.811368787205275e-06,
+      "loss": 0.1884,
+      "step": 2791
+    },
+    {
+      "epoch": 0.4523655217109527,
+      "grad_norm": 1.057442545890808,
+      "learning_rate": 4.811202109778661e-06,
+      "loss": 0.1642,
+      "step": 2792
+    },
+    {
+      "epoch": 0.45252754374594945,
+      "grad_norm": 1.0493812561035156,
+      "learning_rate": 4.811035361634855e-06,
+      "loss": 0.1466,
+      "step": 2793
+    },
+    {
+      "epoch": 0.4526895657809462,
+      "grad_norm": 1.0692343711853027,
+      "learning_rate": 4.810868542778959e-06,
+      "loss": 0.1522,
+      "step": 2794
+    },
+    {
+      "epoch": 0.452851587815943,
+      "grad_norm": 0.98219895362854,
+      "learning_rate": 4.8107016532160784e-06,
+      "loss": 0.1417,
+      "step": 2795
+    },
+    {
+      "epoch": 0.45301360985093975,
+      "grad_norm": 1.0239708423614502,
+      "learning_rate": 4.81053469295132e-06,
+      "loss": 0.1593,
+      "step": 2796
+    },
+    {
+      "epoch": 0.4531756318859365,
+      "grad_norm": 0.94671630859375,
+      "learning_rate": 4.81036766198979e-06,
+      "loss": 0.1411,
+      "step": 2797
+    },
+    {
+      "epoch": 0.4533376539209332,
+      "grad_norm": 1.0224705934524536,
+      "learning_rate": 4.810200560336601e-06,
+      "loss": 0.1615,
+      "step": 2798
+    },
+    {
+      "epoch": 0.45349967595593,
+      "grad_norm": 0.970841109752655,
+      "learning_rate": 4.810033387996865e-06,
+      "loss": 0.1588,
+      "step": 2799
+    },
+    {
+      "epoch": 0.45366169799092676,
+      "grad_norm": 1.008506417274475,
+      "learning_rate": 4.809866144975699e-06,
+      "loss": 0.1653,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4538237200259235,
+      "grad_norm": 0.9327491521835327,
+      "learning_rate": 4.809698831278217e-06,
+      "loss": 0.1419,
+      "step": 2801
+    },
+    {
+      "epoch": 0.4539857420609203,
+      "grad_norm": 0.9707223176956177,
+      "learning_rate": 4.809531446909541e-06,
+      "loss": 0.161,
+      "step": 2802
+    },
+    {
+      "epoch": 0.45414776409591706,
+      "grad_norm": 1.0070552825927734,
+      "learning_rate": 4.8093639918747915e-06,
+      "loss": 0.1703,
+      "step": 2803
+    },
+    {
+      "epoch": 0.45430978613091383,
+      "grad_norm": 1.0490121841430664,
+      "learning_rate": 4.8091964661790926e-06,
+      "loss": 0.1556,
+      "step": 2804
+    },
+    {
+      "epoch": 0.45447180816591054,
+      "grad_norm": 1.058758020401001,
+      "learning_rate": 4.80902886982757e-06,
+      "loss": 0.1536,
+      "step": 2805
+    },
+    {
+      "epoch": 0.4546338302009073,
+      "grad_norm": 0.9229541420936584,
+      "learning_rate": 4.808861202825351e-06,
+      "loss": 0.1402,
+      "step": 2806
+    },
+    {
+      "epoch": 0.4547958522359041,
+      "grad_norm": 1.017575979232788,
+      "learning_rate": 4.8086934651775675e-06,
+      "loss": 0.1572,
+      "step": 2807
+    },
+    {
+      "epoch": 0.45495787427090084,
+      "grad_norm": 0.8822991847991943,
+      "learning_rate": 4.80852565688935e-06,
+      "loss": 0.1379,
+      "step": 2808
+    },
+    {
+      "epoch": 0.4551198963058976,
+      "grad_norm": 0.9270743131637573,
+      "learning_rate": 4.8083577779658344e-06,
+      "loss": 0.1278,
+      "step": 2809
+    },
+    {
+      "epoch": 0.4552819183408944,
+      "grad_norm": 0.9304889440536499,
+      "learning_rate": 4.808189828412157e-06,
+      "loss": 0.1374,
+      "step": 2810
+    },
+    {
+      "epoch": 0.45544394037589114,
+      "grad_norm": 0.9572131633758545,
+      "learning_rate": 4.8080218082334566e-06,
+      "loss": 0.1386,
+      "step": 2811
+    },
+    {
+      "epoch": 0.45560596241088785,
+      "grad_norm": 0.9808574318885803,
+      "learning_rate": 4.807853717434874e-06,
+      "loss": 0.1678,
+      "step": 2812
+    },
+    {
+      "epoch": 0.4557679844458846,
+      "grad_norm": 1.0128886699676514,
+      "learning_rate": 4.807685556021552e-06,
+      "loss": 0.1619,
+      "step": 2813
+    },
+    {
+      "epoch": 0.4559300064808814,
+      "grad_norm": 1.1356350183486938,
+      "learning_rate": 4.807517323998637e-06,
+      "loss": 0.1639,
+      "step": 2814
+    },
+    {
+      "epoch": 0.45609202851587816,
+      "grad_norm": 0.9979428052902222,
+      "learning_rate": 4.807349021371276e-06,
+      "loss": 0.1587,
+      "step": 2815
+    },
+    {
+      "epoch": 0.4562540505508749,
+      "grad_norm": 0.9667853116989136,
+      "learning_rate": 4.8071806481446194e-06,
+      "loss": 0.1401,
+      "step": 2816
+    },
+    {
+      "epoch": 0.4564160725858717,
+      "grad_norm": 0.9587422013282776,
+      "learning_rate": 4.807012204323817e-06,
+      "loss": 0.1515,
+      "step": 2817
+    },
+    {
+      "epoch": 0.45657809462086846,
+      "grad_norm": 0.9749378561973572,
+      "learning_rate": 4.806843689914025e-06,
+      "loss": 0.1488,
+      "step": 2818
+    },
+    {
+      "epoch": 0.4567401166558652,
+      "grad_norm": 1.0103932619094849,
+      "learning_rate": 4.806675104920397e-06,
+      "loss": 0.1579,
+      "step": 2819
+    },
+    {
+      "epoch": 0.45690213869086194,
+      "grad_norm": 1.1067167520523071,
+      "learning_rate": 4.806506449348094e-06,
+      "loss": 0.1602,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4570641607258587,
+      "grad_norm": 1.0808192491531372,
+      "learning_rate": 4.8063377232022755e-06,
+      "loss": 0.1781,
+      "step": 2821
+    },
+    {
+      "epoch": 0.45722618276085547,
+      "grad_norm": 0.9388707280158997,
+      "learning_rate": 4.8061689264881036e-06,
+      "loss": 0.1362,
+      "step": 2822
+    },
+    {
+      "epoch": 0.45738820479585224,
+      "grad_norm": 0.9397252202033997,
+      "learning_rate": 4.806000059210744e-06,
+      "loss": 0.1385,
+      "step": 2823
+    },
+    {
+      "epoch": 0.457550226830849,
+      "grad_norm": 1.0616799592971802,
+      "learning_rate": 4.805831121375361e-06,
+      "loss": 0.1753,
+      "step": 2824
+    },
+    {
+      "epoch": 0.45771224886584577,
+      "grad_norm": 0.973735511302948,
+      "learning_rate": 4.805662112987127e-06,
+      "loss": 0.152,
+      "step": 2825
+    },
+    {
+      "epoch": 0.45787427090084254,
+      "grad_norm": 0.9995198845863342,
+      "learning_rate": 4.805493034051212e-06,
+      "loss": 0.1607,
+      "step": 2826
+    },
+    {
+      "epoch": 0.45803629293583925,
+      "grad_norm": 0.9708908200263977,
+      "learning_rate": 4.80532388457279e-06,
+      "loss": 0.141,
+      "step": 2827
+    },
+    {
+      "epoch": 0.458198314970836,
+      "grad_norm": 1.112001895904541,
+      "learning_rate": 4.805154664557034e-06,
+      "loss": 0.1623,
+      "step": 2828
+    },
+    {
+      "epoch": 0.4583603370058328,
+      "grad_norm": 0.9360017776489258,
+      "learning_rate": 4.804985374009125e-06,
+      "loss": 0.1294,
+      "step": 2829
+    },
+    {
+      "epoch": 0.45852235904082955,
+      "grad_norm": 0.8773815631866455,
+      "learning_rate": 4.804816012934242e-06,
+      "loss": 0.1369,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4586843810758263,
+      "grad_norm": 0.9873615503311157,
+      "learning_rate": 4.8046465813375655e-06,
+      "loss": 0.1471,
+      "step": 2831
+    },
+    {
+      "epoch": 0.4588464031108231,
+      "grad_norm": 1.0159286260604858,
+      "learning_rate": 4.8044770792242815e-06,
+      "loss": 0.1654,
+      "step": 2832
+    },
+    {
+      "epoch": 0.45900842514581985,
+      "grad_norm": 0.9192469120025635,
+      "learning_rate": 4.8043075065995755e-06,
+      "loss": 0.1376,
+      "step": 2833
+    },
+    {
+      "epoch": 0.45917044718081657,
+      "grad_norm": 0.8338161110877991,
+      "learning_rate": 4.8041378634686355e-06,
+      "loss": 0.13,
+      "step": 2834
+    },
+    {
+      "epoch": 0.45933246921581333,
+      "grad_norm": 0.9531093835830688,
+      "learning_rate": 4.803968149836653e-06,
+      "loss": 0.1326,
+      "step": 2835
+    },
+    {
+      "epoch": 0.4594944912508101,
+      "grad_norm": 1.2418529987335205,
+      "learning_rate": 4.803798365708821e-06,
+      "loss": 0.1606,
+      "step": 2836
+    },
+    {
+      "epoch": 0.45965651328580687,
+      "grad_norm": 0.956653892993927,
+      "learning_rate": 4.803628511090333e-06,
+      "loss": 0.1383,
+      "step": 2837
+    },
+    {
+      "epoch": 0.45981853532080363,
+      "grad_norm": 0.9836128950119019,
+      "learning_rate": 4.803458585986389e-06,
+      "loss": 0.1557,
+      "step": 2838
+    },
+    {
+      "epoch": 0.4599805573558004,
+      "grad_norm": 1.1878564357757568,
+      "learning_rate": 4.803288590402185e-06,
+      "loss": 0.1793,
+      "step": 2839
+    },
+    {
+      "epoch": 0.46014257939079717,
+      "grad_norm": 1.0045480728149414,
+      "learning_rate": 4.803118524342925e-06,
+      "loss": 0.1658,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46030460142579394,
+      "grad_norm": 1.1326467990875244,
+      "learning_rate": 4.802948387813812e-06,
+      "loss": 0.1524,
+      "step": 2841
+    },
+    {
+      "epoch": 0.46046662346079065,
+      "grad_norm": 0.9452898502349854,
+      "learning_rate": 4.80277818082005e-06,
+      "loss": 0.149,
+      "step": 2842
+    },
+    {
+      "epoch": 0.4606286454957874,
+      "grad_norm": 1.0143297910690308,
+      "learning_rate": 4.802607903366849e-06,
+      "loss": 0.1439,
+      "step": 2843
+    },
+    {
+      "epoch": 0.4607906675307842,
+      "grad_norm": 0.9265870451927185,
+      "learning_rate": 4.802437555459418e-06,
+      "loss": 0.1461,
+      "step": 2844
+    },
+    {
+      "epoch": 0.46095268956578095,
+      "grad_norm": 0.9968574643135071,
+      "learning_rate": 4.80226713710297e-06,
+      "loss": 0.1549,
+      "step": 2845
+    },
+    {
+      "epoch": 0.4611147116007777,
+      "grad_norm": 0.9609559774398804,
+      "learning_rate": 4.802096648302718e-06,
+      "loss": 0.1511,
+      "step": 2846
+    },
+    {
+      "epoch": 0.4612767336357745,
+      "grad_norm": 1.0482757091522217,
+      "learning_rate": 4.8019260890638805e-06,
+      "loss": 0.1849,
+      "step": 2847
+    },
+    {
+      "epoch": 0.46143875567077125,
+      "grad_norm": 1.261745572090149,
+      "learning_rate": 4.801755459391675e-06,
+      "loss": 0.2021,
+      "step": 2848
+    },
+    {
+      "epoch": 0.46160077770576796,
+      "grad_norm": 1.0810171365737915,
+      "learning_rate": 4.801584759291323e-06,
+      "loss": 0.1696,
+      "step": 2849
+    },
+    {
+      "epoch": 0.46176279974076473,
+      "grad_norm": 1.0707732439041138,
+      "learning_rate": 4.801413988768047e-06,
+      "loss": 0.1727,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4619248217757615,
+      "grad_norm": 1.021913766860962,
+      "learning_rate": 4.8012431478270716e-06,
+      "loss": 0.1585,
+      "step": 2851
+    },
+    {
+      "epoch": 0.46208684381075826,
+      "grad_norm": 0.9242175817489624,
+      "learning_rate": 4.801072236473625e-06,
+      "loss": 0.1406,
+      "step": 2852
+    },
+    {
+      "epoch": 0.46224886584575503,
+      "grad_norm": 0.953057587146759,
+      "learning_rate": 4.800901254712936e-06,
+      "loss": 0.1605,
+      "step": 2853
+    },
+    {
+      "epoch": 0.4624108878807518,
+      "grad_norm": 0.9746292233467102,
+      "learning_rate": 4.800730202550237e-06,
+      "loss": 0.141,
+      "step": 2854
+    },
+    {
+      "epoch": 0.46257290991574856,
+      "grad_norm": 0.9918842315673828,
+      "learning_rate": 4.800559079990762e-06,
+      "loss": 0.1639,
+      "step": 2855
+    },
+    {
+      "epoch": 0.4627349319507453,
+      "grad_norm": 0.9181191325187683,
+      "learning_rate": 4.800387887039747e-06,
+      "loss": 0.1396,
+      "step": 2856
+    },
+    {
+      "epoch": 0.46289695398574204,
+      "grad_norm": 1.0667685270309448,
+      "learning_rate": 4.800216623702428e-06,
+      "loss": 0.1612,
+      "step": 2857
+    },
+    {
+      "epoch": 0.4630589760207388,
+      "grad_norm": 0.8903344869613647,
+      "learning_rate": 4.800045289984047e-06,
+      "loss": 0.1219,
+      "step": 2858
+    },
+    {
+      "epoch": 0.4632209980557356,
+      "grad_norm": 1.1288483142852783,
+      "learning_rate": 4.7998738858898475e-06,
+      "loss": 0.1552,
+      "step": 2859
+    },
+    {
+      "epoch": 0.46338302009073234,
+      "grad_norm": 0.9569337964057922,
+      "learning_rate": 4.799702411425071e-06,
+      "loss": 0.1304,
+      "step": 2860
+    },
+    {
+      "epoch": 0.4635450421257291,
+      "grad_norm": 0.9571936130523682,
+      "learning_rate": 4.799530866594967e-06,
+      "loss": 0.1502,
+      "step": 2861
+    },
+    {
+      "epoch": 0.4637070641607259,
+      "grad_norm": 0.955163836479187,
+      "learning_rate": 4.7993592514047825e-06,
+      "loss": 0.1475,
+      "step": 2862
+    },
+    {
+      "epoch": 0.4638690861957226,
+      "grad_norm": 0.8739729523658752,
+      "learning_rate": 4.79918756585977e-06,
+      "loss": 0.1334,
+      "step": 2863
+    },
+    {
+      "epoch": 0.46403110823071936,
+      "grad_norm": 1.0791646242141724,
+      "learning_rate": 4.7990158099651815e-06,
+      "loss": 0.1593,
+      "step": 2864
+    },
+    {
+      "epoch": 0.4641931302657161,
+      "grad_norm": 1.0804657936096191,
+      "learning_rate": 4.798843983726272e-06,
+      "loss": 0.148,
+      "step": 2865
+    },
+    {
+      "epoch": 0.4643551523007129,
+      "grad_norm": 0.9415281414985657,
+      "learning_rate": 4.798672087148301e-06,
+      "loss": 0.1481,
+      "step": 2866
+    },
+    {
+      "epoch": 0.46451717433570966,
+      "grad_norm": 0.9962906837463379,
+      "learning_rate": 4.798500120236526e-06,
+      "loss": 0.154,
+      "step": 2867
+    },
+    {
+      "epoch": 0.4646791963707064,
+      "grad_norm": 1.016023874282837,
+      "learning_rate": 4.79832808299621e-06,
+      "loss": 0.1592,
+      "step": 2868
+    },
+    {
+      "epoch": 0.4648412184057032,
+      "grad_norm": 0.8804002404212952,
+      "learning_rate": 4.7981559754326154e-06,
+      "loss": 0.1345,
+      "step": 2869
+    },
+    {
+      "epoch": 0.46500324044069996,
+      "grad_norm": 0.9187561869621277,
+      "learning_rate": 4.797983797551011e-06,
+      "loss": 0.1399,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4651652624756967,
+      "grad_norm": 0.9533474445343018,
+      "learning_rate": 4.797811549356662e-06,
+      "loss": 0.1513,
+      "step": 2871
+    },
+    {
+      "epoch": 0.46532728451069344,
+      "grad_norm": 1.0123627185821533,
+      "learning_rate": 4.7976392308548416e-06,
+      "loss": 0.1563,
+      "step": 2872
+    },
+    {
+      "epoch": 0.4654893065456902,
+      "grad_norm": 1.0109021663665771,
+      "learning_rate": 4.7974668420508195e-06,
+      "loss": 0.1527,
+      "step": 2873
+    },
+    {
+      "epoch": 0.465651328580687,
+      "grad_norm": 1.1355345249176025,
+      "learning_rate": 4.797294382949873e-06,
+      "loss": 0.1783,
+      "step": 2874
+    },
+    {
+      "epoch": 0.46581335061568374,
+      "grad_norm": 1.0185506343841553,
+      "learning_rate": 4.797121853557277e-06,
+      "loss": 0.1591,
+      "step": 2875
+    },
+    {
+      "epoch": 0.4659753726506805,
+      "grad_norm": 0.9530416131019592,
+      "learning_rate": 4.796949253878311e-06,
+      "loss": 0.1403,
+      "step": 2876
+    },
+    {
+      "epoch": 0.4661373946856773,
+      "grad_norm": 0.9338066577911377,
+      "learning_rate": 4.796776583918256e-06,
+      "loss": 0.1505,
+      "step": 2877
+    },
+    {
+      "epoch": 0.466299416720674,
+      "grad_norm": 0.8232749700546265,
+      "learning_rate": 4.796603843682397e-06,
+      "loss": 0.1251,
+      "step": 2878
+    },
+    {
+      "epoch": 0.46646143875567075,
+      "grad_norm": 1.013901710510254,
+      "learning_rate": 4.7964310331760174e-06,
+      "loss": 0.1506,
+      "step": 2879
+    },
+    {
+      "epoch": 0.4666234607906675,
+      "grad_norm": 1.042008399963379,
+      "learning_rate": 4.796258152404406e-06,
+      "loss": 0.1578,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4667854828256643,
+      "grad_norm": 1.0777783393859863,
+      "learning_rate": 4.796085201372852e-06,
+      "loss": 0.1805,
+      "step": 2881
+    },
+    {
+      "epoch": 0.46694750486066106,
+      "grad_norm": 0.9231262803077698,
+      "learning_rate": 4.795912180086646e-06,
+      "loss": 0.1404,
+      "step": 2882
+    },
+    {
+      "epoch": 0.4671095268956578,
+      "grad_norm": 0.972801685333252,
+      "learning_rate": 4.795739088551084e-06,
+      "loss": 0.1478,
+      "step": 2883
+    },
+    {
+      "epoch": 0.4672715489306546,
+      "grad_norm": 0.9988305568695068,
+      "learning_rate": 4.795565926771461e-06,
+      "loss": 0.1507,
+      "step": 2884
+    },
+    {
+      "epoch": 0.4674335709656513,
+      "grad_norm": 0.9387997388839722,
+      "learning_rate": 4.795392694753077e-06,
+      "loss": 0.1514,
+      "step": 2885
+    },
+    {
+      "epoch": 0.46759559300064807,
+      "grad_norm": 0.9761515855789185,
+      "learning_rate": 4.79521939250123e-06,
+      "loss": 0.1604,
+      "step": 2886
+    },
+    {
+      "epoch": 0.46775761503564484,
+      "grad_norm": 0.9907017946243286,
+      "learning_rate": 4.7950460200212244e-06,
+      "loss": 0.1448,
+      "step": 2887
+    },
+    {
+      "epoch": 0.4679196370706416,
+      "grad_norm": 1.0086826086044312,
+      "learning_rate": 4.7948725773183645e-06,
+      "loss": 0.1539,
+      "step": 2888
+    },
+    {
+      "epoch": 0.46808165910563837,
+      "grad_norm": 0.9991224408149719,
+      "learning_rate": 4.794699064397957e-06,
+      "loss": 0.156,
+      "step": 2889
+    },
+    {
+      "epoch": 0.46824368114063514,
+      "grad_norm": 1.0062789916992188,
+      "learning_rate": 4.794525481265312e-06,
+      "loss": 0.1594,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4684057031756319,
+      "grad_norm": 1.0728275775909424,
+      "learning_rate": 4.794351827925739e-06,
+      "loss": 0.156,
+      "step": 2891
+    },
+    {
+      "epoch": 0.46856772521062867,
+      "grad_norm": 0.9946532249450684,
+      "learning_rate": 4.794178104384554e-06,
+      "loss": 0.1754,
+      "step": 2892
+    },
+    {
+      "epoch": 0.4687297472456254,
+      "grad_norm": 1.0101428031921387,
+      "learning_rate": 4.794004310647069e-06,
+      "loss": 0.1571,
+      "step": 2893
+    },
+    {
+      "epoch": 0.46889176928062215,
+      "grad_norm": 0.8463723063468933,
+      "learning_rate": 4.7938304467186036e-06,
+      "loss": 0.1326,
+      "step": 2894
+    },
+    {
+      "epoch": 0.4690537913156189,
+      "grad_norm": 0.8931572437286377,
+      "learning_rate": 4.793656512604478e-06,
+      "loss": 0.1415,
+      "step": 2895
+    },
+    {
+      "epoch": 0.4692158133506157,
+      "grad_norm": 1.0302973985671997,
+      "learning_rate": 4.793482508310014e-06,
+      "loss": 0.1637,
+      "step": 2896
+    },
+    {
+      "epoch": 0.46937783538561245,
+      "grad_norm": 1.800075888633728,
+      "learning_rate": 4.793308433840534e-06,
+      "loss": 0.1476,
+      "step": 2897
+    },
+    {
+      "epoch": 0.4695398574206092,
+      "grad_norm": 0.8768447041511536,
+      "learning_rate": 4.793134289201367e-06,
+      "loss": 0.1269,
+      "step": 2898
+    },
+    {
+      "epoch": 0.469701879455606,
+      "grad_norm": 0.9154813289642334,
+      "learning_rate": 4.792960074397839e-06,
+      "loss": 0.1353,
+      "step": 2899
+    },
+    {
+      "epoch": 0.4698639014906027,
+      "grad_norm": 1.0366954803466797,
+      "learning_rate": 4.792785789435283e-06,
+      "loss": 0.1506,
+      "step": 2900
+    },
+    {
+      "epoch": 0.47002592352559946,
+      "grad_norm": 0.8922173380851746,
+      "learning_rate": 4.792611434319029e-06,
+      "loss": 0.1285,
+      "step": 2901
+    },
+    {
+      "epoch": 0.47018794556059623,
+      "grad_norm": 0.9748638272285461,
+      "learning_rate": 4.792437009054413e-06,
+      "loss": 0.1438,
+      "step": 2902
+    },
+    {
+      "epoch": 0.470349967595593,
+      "grad_norm": 1.032471776008606,
+      "learning_rate": 4.792262513646773e-06,
+      "loss": 0.1685,
+      "step": 2903
+    },
+    {
+      "epoch": 0.47051198963058977,
+      "grad_norm": 1.1114881038665771,
+      "learning_rate": 4.792087948101447e-06,
+      "loss": 0.1312,
+      "step": 2904
+    },
+    {
+      "epoch": 0.47067401166558653,
+      "grad_norm": 2.0958445072174072,
+      "learning_rate": 4.791913312423776e-06,
+      "loss": 0.1668,
+      "step": 2905
+    },
+    {
+      "epoch": 0.4708360337005833,
+      "grad_norm": 1.0381637811660767,
+      "learning_rate": 4.791738606619105e-06,
+      "loss": 0.1517,
+      "step": 2906
+    },
+    {
+      "epoch": 0.47099805573558,
+      "grad_norm": 0.963405191898346,
+      "learning_rate": 4.7915638306927775e-06,
+      "loss": 0.1395,
+      "step": 2907
+    },
+    {
+      "epoch": 0.4711600777705768,
+      "grad_norm": 1.0255495309829712,
+      "learning_rate": 4.791388984650143e-06,
+      "loss": 0.1537,
+      "step": 2908
+    },
+    {
+      "epoch": 0.47132209980557355,
+      "grad_norm": 0.9075080156326294,
+      "learning_rate": 4.79121406849655e-06,
+      "loss": 0.142,
+      "step": 2909
+    },
+    {
+      "epoch": 0.4714841218405703,
+      "grad_norm": 0.981643557548523,
+      "learning_rate": 4.791039082237352e-06,
+      "loss": 0.1381,
+      "step": 2910
+    },
+    {
+      "epoch": 0.4716461438755671,
+      "grad_norm": 1.0010422468185425,
+      "learning_rate": 4.790864025877902e-06,
+      "loss": 0.1668,
+      "step": 2911
+    },
+    {
+      "epoch": 0.47180816591056385,
+      "grad_norm": 1.1024540662765503,
+      "learning_rate": 4.790688899423556e-06,
+      "loss": 0.1753,
+      "step": 2912
+    },
+    {
+      "epoch": 0.4719701879455606,
+      "grad_norm": 0.8568621277809143,
+      "learning_rate": 4.790513702879673e-06,
+      "loss": 0.1248,
+      "step": 2913
+    },
+    {
+      "epoch": 0.4721322099805574,
+      "grad_norm": 0.8953239917755127,
+      "learning_rate": 4.7903384362516135e-06,
+      "loss": 0.1436,
+      "step": 2914
+    },
+    {
+      "epoch": 0.4722942320155541,
+      "grad_norm": 1.0069905519485474,
+      "learning_rate": 4.790163099544741e-06,
+      "loss": 0.1673,
+      "step": 2915
+    },
+    {
+      "epoch": 0.47245625405055086,
+      "grad_norm": 0.9999399185180664,
+      "learning_rate": 4.78998769276442e-06,
+      "loss": 0.1499,
+      "step": 2916
+    },
+    {
+      "epoch": 0.47261827608554763,
+      "grad_norm": 1.05617356300354,
+      "learning_rate": 4.789812215916016e-06,
+      "loss": 0.1648,
+      "step": 2917
+    },
+    {
+      "epoch": 0.4727802981205444,
+      "grad_norm": 0.9993578195571899,
+      "learning_rate": 4.7896366690049016e-06,
+      "loss": 0.1546,
+      "step": 2918
+    },
+    {
+      "epoch": 0.47294232015554116,
+      "grad_norm": 0.8876897692680359,
+      "learning_rate": 4.789461052036444e-06,
+      "loss": 0.1327,
+      "step": 2919
+    },
+    {
+      "epoch": 0.47310434219053793,
+      "grad_norm": 0.9634594321250916,
+      "learning_rate": 4.789285365016019e-06,
+      "loss": 0.1576,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4732663642255347,
+      "grad_norm": 0.8948929905891418,
+      "learning_rate": 4.7891096079490016e-06,
+      "loss": 0.137,
+      "step": 2921
+    },
+    {
+      "epoch": 0.4734283862605314,
+      "grad_norm": 1.0438789129257202,
+      "learning_rate": 4.788933780840771e-06,
+      "loss": 0.1549,
+      "step": 2922
+    },
+    {
+      "epoch": 0.4735904082955282,
+      "grad_norm": 1.0046937465667725,
+      "learning_rate": 4.7887578836967045e-06,
+      "loss": 0.1338,
+      "step": 2923
+    },
+    {
+      "epoch": 0.47375243033052494,
+      "grad_norm": 0.9940471649169922,
+      "learning_rate": 4.788581916522186e-06,
+      "loss": 0.1524,
+      "step": 2924
+    },
+    {
+      "epoch": 0.4739144523655217,
+      "grad_norm": 0.9437041878700256,
+      "learning_rate": 4.788405879322599e-06,
+      "loss": 0.1619,
+      "step": 2925
+    },
+    {
+      "epoch": 0.4740764744005185,
+      "grad_norm": 0.9506624937057495,
+      "learning_rate": 4.78822977210333e-06,
+      "loss": 0.1405,
+      "step": 2926
+    },
+    {
+      "epoch": 0.47423849643551524,
+      "grad_norm": 1.014735460281372,
+      "learning_rate": 4.788053594869767e-06,
+      "loss": 0.132,
+      "step": 2927
+    },
+    {
+      "epoch": 0.474400518470512,
+      "grad_norm": 1.1211403608322144,
+      "learning_rate": 4.787877347627302e-06,
+      "loss": 0.1818,
+      "step": 2928
+    },
+    {
+      "epoch": 0.4745625405055087,
+      "grad_norm": 1.0037801265716553,
+      "learning_rate": 4.787701030381326e-06,
+      "loss": 0.1575,
+      "step": 2929
+    },
+    {
+      "epoch": 0.4747245625405055,
+      "grad_norm": 0.911635160446167,
+      "learning_rate": 4.787524643137235e-06,
+      "loss": 0.1325,
+      "step": 2930
+    },
+    {
+      "epoch": 0.47488658457550226,
+      "grad_norm": 0.9881319999694824,
+      "learning_rate": 4.7873481859004245e-06,
+      "loss": 0.1486,
+      "step": 2931
+    },
+    {
+      "epoch": 0.475048606610499,
+      "grad_norm": 1.0285166501998901,
+      "learning_rate": 4.7871716586762965e-06,
+      "loss": 0.1545,
+      "step": 2932
+    },
+    {
+      "epoch": 0.4752106286454958,
+      "grad_norm": 1.2436772584915161,
+      "learning_rate": 4.786995061470249e-06,
+      "loss": 0.1594,
+      "step": 2933
+    },
+    {
+      "epoch": 0.47537265068049256,
+      "grad_norm": 1.0951595306396484,
+      "learning_rate": 4.786818394287688e-06,
+      "loss": 0.1712,
+      "step": 2934
+    },
+    {
+      "epoch": 0.4755346727154893,
+      "grad_norm": 1.0446057319641113,
+      "learning_rate": 4.786641657134017e-06,
+      "loss": 0.1583,
+      "step": 2935
+    },
+    {
+      "epoch": 0.4756966947504861,
+      "grad_norm": 0.9601659774780273,
+      "learning_rate": 4.786464850014646e-06,
+      "loss": 0.1574,
+      "step": 2936
+    },
+    {
+      "epoch": 0.4758587167854828,
+      "grad_norm": 0.9458787441253662,
+      "learning_rate": 4.786287972934984e-06,
+      "loss": 0.1306,
+      "step": 2937
+    },
+    {
+      "epoch": 0.47602073882047957,
+      "grad_norm": 1.0759549140930176,
+      "learning_rate": 4.786111025900442e-06,
+      "loss": 0.1646,
+      "step": 2938
+    },
+    {
+      "epoch": 0.47618276085547634,
+      "grad_norm": 0.9733848571777344,
+      "learning_rate": 4.785934008916435e-06,
+      "loss": 0.1616,
+      "step": 2939
+    },
+    {
+      "epoch": 0.4763447828904731,
+      "grad_norm": 0.9360319375991821,
+      "learning_rate": 4.785756921988379e-06,
+      "loss": 0.1504,
+      "step": 2940
+    },
+    {
+      "epoch": 0.4765068049254699,
+      "grad_norm": 0.9589405059814453,
+      "learning_rate": 4.785579765121693e-06,
+      "loss": 0.1566,
+      "step": 2941
+    },
+    {
+      "epoch": 0.47666882696046664,
+      "grad_norm": 1.0184801816940308,
+      "learning_rate": 4.785402538321798e-06,
+      "loss": 0.1502,
+      "step": 2942
+    },
+    {
+      "epoch": 0.4768308489954634,
+      "grad_norm": 1.011683702468872,
+      "learning_rate": 4.785225241594114e-06,
+      "loss": 0.165,
+      "step": 2943
+    },
+    {
+      "epoch": 0.4769928710304601,
+      "grad_norm": 0.9309144020080566,
+      "learning_rate": 4.785047874944069e-06,
+      "loss": 0.1353,
+      "step": 2944
+    },
+    {
+      "epoch": 0.4771548930654569,
+      "grad_norm": 1.050207495689392,
+      "learning_rate": 4.7848704383770875e-06,
+      "loss": 0.1605,
+      "step": 2945
+    },
+    {
+      "epoch": 0.47731691510045365,
+      "grad_norm": 0.9079956412315369,
+      "learning_rate": 4.784692931898601e-06,
+      "loss": 0.1404,
+      "step": 2946
+    },
+    {
+      "epoch": 0.4774789371354504,
+      "grad_norm": 0.9349779486656189,
+      "learning_rate": 4.784515355514039e-06,
+      "loss": 0.1332,
+      "step": 2947
+    },
+    {
+      "epoch": 0.4776409591704472,
+      "grad_norm": 1.089605689048767,
+      "learning_rate": 4.7843377092288365e-06,
+      "loss": 0.1575,
+      "step": 2948
+    },
+    {
+      "epoch": 0.47780298120544396,
+      "grad_norm": 1.0942819118499756,
+      "learning_rate": 4.784159993048427e-06,
+      "loss": 0.1675,
+      "step": 2949
+    },
+    {
+      "epoch": 0.4779650032404407,
+      "grad_norm": 0.9137088656425476,
+      "learning_rate": 4.7839822069782505e-06,
+      "loss": 0.1388,
+      "step": 2950
+    },
+    {
+      "epoch": 0.47812702527543743,
+      "grad_norm": 0.9597694277763367,
+      "learning_rate": 4.783804351023745e-06,
+      "loss": 0.155,
+      "step": 2951
+    },
+    {
+      "epoch": 0.4782890473104342,
+      "grad_norm": 0.9747596383094788,
+      "learning_rate": 4.783626425190353e-06,
+      "loss": 0.1504,
+      "step": 2952
+    },
+    {
+      "epoch": 0.47845106934543097,
+      "grad_norm": 0.9618208408355713,
+      "learning_rate": 4.783448429483518e-06,
+      "loss": 0.1479,
+      "step": 2953
+    },
+    {
+      "epoch": 0.47861309138042774,
+      "grad_norm": 0.9732487201690674,
+      "learning_rate": 4.783270363908687e-06,
+      "loss": 0.1515,
+      "step": 2954
+    },
+    {
+      "epoch": 0.4787751134154245,
+      "grad_norm": 1.075685977935791,
+      "learning_rate": 4.78309222847131e-06,
+      "loss": 0.1613,
+      "step": 2955
+    },
+    {
+      "epoch": 0.47893713545042127,
+      "grad_norm": 0.9794430136680603,
+      "learning_rate": 4.782914023176834e-06,
+      "loss": 0.1425,
+      "step": 2956
+    },
+    {
+      "epoch": 0.47909915748541804,
+      "grad_norm": 1.0761573314666748,
+      "learning_rate": 4.782735748030714e-06,
+      "loss": 0.1697,
+      "step": 2957
+    },
+    {
+      "epoch": 0.4792611795204148,
+      "grad_norm": 1.1438524723052979,
+      "learning_rate": 4.782557403038404e-06,
+      "loss": 0.1605,
+      "step": 2958
+    },
+    {
+      "epoch": 0.4794232015554115,
+      "grad_norm": 0.9872993230819702,
+      "learning_rate": 4.782378988205362e-06,
+      "loss": 0.1534,
+      "step": 2959
+    },
+    {
+      "epoch": 0.4795852235904083,
+      "grad_norm": 0.964551568031311,
+      "learning_rate": 4.7822005035370455e-06,
+      "loss": 0.1568,
+      "step": 2960
+    },
+    {
+      "epoch": 0.47974724562540505,
+      "grad_norm": 0.9149985313415527,
+      "learning_rate": 4.782021949038916e-06,
+      "loss": 0.138,
+      "step": 2961
+    },
+    {
+      "epoch": 0.4799092676604018,
+      "grad_norm": 0.9830363392829895,
+      "learning_rate": 4.781843324716437e-06,
+      "loss": 0.1573,
+      "step": 2962
+    },
+    {
+      "epoch": 0.4800712896953986,
+      "grad_norm": 0.9936883449554443,
+      "learning_rate": 4.781664630575076e-06,
+      "loss": 0.1575,
+      "step": 2963
+    },
+    {
+      "epoch": 0.48023331173039535,
+      "grad_norm": 1.0400924682617188,
+      "learning_rate": 4.7814858666202975e-06,
+      "loss": 0.1731,
+      "step": 2964
+    },
+    {
+      "epoch": 0.4803953337653921,
+      "grad_norm": 0.9391010999679565,
+      "learning_rate": 4.781307032857573e-06,
+      "loss": 0.1357,
+      "step": 2965
+    },
+    {
+      "epoch": 0.48055735580038883,
+      "grad_norm": 1.0018044710159302,
+      "learning_rate": 4.781128129292374e-06,
+      "loss": 0.1614,
+      "step": 2966
+    },
+    {
+      "epoch": 0.4807193778353856,
+      "grad_norm": 0.8832952976226807,
+      "learning_rate": 4.780949155930174e-06,
+      "loss": 0.142,
+      "step": 2967
+    },
+    {
+      "epoch": 0.48088139987038236,
+      "grad_norm": 1.0228934288024902,
+      "learning_rate": 4.7807701127764506e-06,
+      "loss": 0.166,
+      "step": 2968
+    },
+    {
+      "epoch": 0.48104342190537913,
+      "grad_norm": 0.8861875534057617,
+      "learning_rate": 4.78059099983668e-06,
+      "loss": 0.1308,
+      "step": 2969
+    },
+    {
+      "epoch": 0.4812054439403759,
+      "grad_norm": 1.0241353511810303,
+      "learning_rate": 4.780411817116344e-06,
+      "loss": 0.167,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48136746597537267,
+      "grad_norm": 1.0048408508300781,
+      "learning_rate": 4.7802325646209255e-06,
+      "loss": 0.1637,
+      "step": 2971
+    },
+    {
+      "epoch": 0.48152948801036943,
+      "grad_norm": 0.9971985816955566,
+      "learning_rate": 4.780053242355908e-06,
+      "loss": 0.1671,
+      "step": 2972
+    },
+    {
+      "epoch": 0.48169151004536614,
+      "grad_norm": 1.110739827156067,
+      "learning_rate": 4.779873850326778e-06,
+      "loss": 0.1776,
+      "step": 2973
+    },
+    {
+      "epoch": 0.4818535320803629,
+      "grad_norm": 0.9932035207748413,
+      "learning_rate": 4.779694388539027e-06,
+      "loss": 0.1502,
+      "step": 2974
+    },
+    {
+      "epoch": 0.4820155541153597,
+      "grad_norm": 0.9594526290893555,
+      "learning_rate": 4.779514856998144e-06,
+      "loss": 0.1553,
+      "step": 2975
+    },
+    {
+      "epoch": 0.48217757615035645,
+      "grad_norm": 0.9843863844871521,
+      "learning_rate": 4.779335255709623e-06,
+      "loss": 0.1436,
+      "step": 2976
+    },
+    {
+      "epoch": 0.4823395981853532,
+      "grad_norm": 1.0879184007644653,
+      "learning_rate": 4.7791555846789585e-06,
+      "loss": 0.1591,
+      "step": 2977
+    },
+    {
+      "epoch": 0.48250162022035,
+      "grad_norm": 1.059070348739624,
+      "learning_rate": 4.778975843911649e-06,
+      "loss": 0.1468,
+      "step": 2978
+    },
+    {
+      "epoch": 0.48266364225534675,
+      "grad_norm": 0.9296880960464478,
+      "learning_rate": 4.778796033413193e-06,
+      "loss": 0.1387,
+      "step": 2979
+    },
+    {
+      "epoch": 0.48282566429034346,
+      "grad_norm": 1.030938982963562,
+      "learning_rate": 4.778616153189093e-06,
+      "loss": 0.1576,
+      "step": 2980
+    },
+    {
+      "epoch": 0.4829876863253402,
+      "grad_norm": 0.967088520526886,
+      "learning_rate": 4.778436203244853e-06,
+      "loss": 0.1516,
+      "step": 2981
+    },
+    {
+      "epoch": 0.483149708360337,
+      "grad_norm": 0.9913309216499329,
+      "learning_rate": 4.7782561835859795e-06,
+      "loss": 0.1429,
+      "step": 2982
+    },
+    {
+      "epoch": 0.48331173039533376,
+      "grad_norm": 1.0217386484146118,
+      "learning_rate": 4.77807609421798e-06,
+      "loss": 0.1632,
+      "step": 2983
+    },
+    {
+      "epoch": 0.48347375243033053,
+      "grad_norm": 0.9644971489906311,
+      "learning_rate": 4.777895935146364e-06,
+      "loss": 0.1548,
+      "step": 2984
+    },
+    {
+      "epoch": 0.4836357744653273,
+      "grad_norm": 0.9488604068756104,
+      "learning_rate": 4.777715706376645e-06,
+      "loss": 0.1591,
+      "step": 2985
+    },
+    {
+      "epoch": 0.48379779650032406,
+      "grad_norm": 1.0006232261657715,
+      "learning_rate": 4.777535407914338e-06,
+      "loss": 0.1522,
+      "step": 2986
+    },
+    {
+      "epoch": 0.48395981853532083,
+      "grad_norm": 1.0294928550720215,
+      "learning_rate": 4.777355039764958e-06,
+      "loss": 0.1633,
+      "step": 2987
+    },
+    {
+      "epoch": 0.48412184057031754,
+      "grad_norm": 0.8555730581283569,
+      "learning_rate": 4.777174601934026e-06,
+      "loss": 0.1297,
+      "step": 2988
+    },
+    {
+      "epoch": 0.4842838626053143,
+      "grad_norm": 0.8943042755126953,
+      "learning_rate": 4.776994094427061e-06,
+      "loss": 0.1294,
+      "step": 2989
+    },
+    {
+      "epoch": 0.4844458846403111,
+      "grad_norm": 0.976116418838501,
+      "learning_rate": 4.776813517249588e-06,
+      "loss": 0.1681,
+      "step": 2990
+    },
+    {
+      "epoch": 0.48460790667530784,
+      "grad_norm": 0.9380887746810913,
+      "learning_rate": 4.77663287040713e-06,
+      "loss": 0.156,
+      "step": 2991
+    },
+    {
+      "epoch": 0.4847699287103046,
+      "grad_norm": 0.9756038188934326,
+      "learning_rate": 4.776452153905216e-06,
+      "loss": 0.1531,
+      "step": 2992
+    },
+    {
+      "epoch": 0.4849319507453014,
+      "grad_norm": 0.9774467945098877,
+      "learning_rate": 4.776271367749375e-06,
+      "loss": 0.1688,
+      "step": 2993
+    },
+    {
+      "epoch": 0.48509397278029814,
+      "grad_norm": 1.0772470235824585,
+      "learning_rate": 4.776090511945139e-06,
+      "loss": 0.174,
+      "step": 2994
+    },
+    {
+      "epoch": 0.48525599481529486,
+      "grad_norm": 0.983753502368927,
+      "learning_rate": 4.77590958649804e-06,
+      "loss": 0.1422,
+      "step": 2995
+    },
+    {
+      "epoch": 0.4854180168502916,
+      "grad_norm": 1.1349477767944336,
+      "learning_rate": 4.775728591413616e-06,
+      "loss": 0.1654,
+      "step": 2996
+    },
+    {
+      "epoch": 0.4855800388852884,
+      "grad_norm": 1.0424829721450806,
+      "learning_rate": 4.775547526697405e-06,
+      "loss": 0.1515,
+      "step": 2997
+    },
+    {
+      "epoch": 0.48574206092028516,
+      "grad_norm": 1.055623173713684,
+      "learning_rate": 4.775366392354946e-06,
+      "loss": 0.1482,
+      "step": 2998
+    },
+    {
+      "epoch": 0.4859040829552819,
+      "grad_norm": 0.9909132122993469,
+      "learning_rate": 4.775185188391781e-06,
+      "loss": 0.1529,
+      "step": 2999
+    },
+    {
+      "epoch": 0.4860661049902787,
+      "grad_norm": 0.981861412525177,
+      "learning_rate": 4.775003914813456e-06,
+      "loss": 0.1505,
+      "step": 3000
+    },
+    {
+      "epoch": 0.48622812702527546,
+      "grad_norm": 0.9199234247207642,
+      "learning_rate": 4.774822571625516e-06,
+      "loss": 0.1376,
+      "step": 3001
+    },
+    {
+      "epoch": 0.48639014906027217,
+      "grad_norm": 1.0599560737609863,
+      "learning_rate": 4.7746411588335105e-06,
+      "loss": 0.1656,
+      "step": 3002
+    },
+    {
+      "epoch": 0.48655217109526894,
+      "grad_norm": 0.9341553449630737,
+      "learning_rate": 4.774459676442991e-06,
+      "loss": 0.1573,
+      "step": 3003
+    },
+    {
+      "epoch": 0.4867141931302657,
+      "grad_norm": 1.0569862127304077,
+      "learning_rate": 4.774278124459509e-06,
+      "loss": 0.1664,
+      "step": 3004
+    },
+    {
+      "epoch": 0.48687621516526247,
+      "grad_norm": 0.970799446105957,
+      "learning_rate": 4.774096502888619e-06,
+      "loss": 0.1452,
+      "step": 3005
+    },
+    {
+      "epoch": 0.48703823720025924,
+      "grad_norm": 1.03872811794281,
+      "learning_rate": 4.773914811735879e-06,
+      "loss": 0.173,
+      "step": 3006
+    },
+    {
+      "epoch": 0.487200259235256,
+      "grad_norm": 0.8187316656112671,
+      "learning_rate": 4.773733051006849e-06,
+      "loss": 0.1248,
+      "step": 3007
+    },
+    {
+      "epoch": 0.4873622812702528,
+      "grad_norm": 0.9771019816398621,
+      "learning_rate": 4.773551220707091e-06,
+      "loss": 0.1625,
+      "step": 3008
+    },
+    {
+      "epoch": 0.48752430330524954,
+      "grad_norm": 0.9272666573524475,
+      "learning_rate": 4.773369320842167e-06,
+      "loss": 0.1544,
+      "step": 3009
+    },
+    {
+      "epoch": 0.48768632534024625,
+      "grad_norm": 0.886873185634613,
+      "learning_rate": 4.773187351417643e-06,
+      "loss": 0.1422,
+      "step": 3010
+    },
+    {
+      "epoch": 0.487848347375243,
+      "grad_norm": 1.0471525192260742,
+      "learning_rate": 4.773005312439087e-06,
+      "loss": 0.1653,
+      "step": 3011
+    },
+    {
+      "epoch": 0.4880103694102398,
+      "grad_norm": 1.0167343616485596,
+      "learning_rate": 4.772823203912069e-06,
+      "loss": 0.1555,
+      "step": 3012
+    },
+    {
+      "epoch": 0.48817239144523655,
+      "grad_norm": 0.9511725902557373,
+      "learning_rate": 4.7726410258421616e-06,
+      "loss": 0.1502,
+      "step": 3013
+    },
+    {
+      "epoch": 0.4883344134802333,
+      "grad_norm": 1.0081369876861572,
+      "learning_rate": 4.772458778234938e-06,
+      "loss": 0.1521,
+      "step": 3014
+    },
+    {
+      "epoch": 0.4884964355152301,
+      "grad_norm": 1.0319949388504028,
+      "learning_rate": 4.772276461095975e-06,
+      "loss": 0.1658,
+      "step": 3015
+    },
+    {
+      "epoch": 0.48865845755022685,
+      "grad_norm": 1.0191717147827148,
+      "learning_rate": 4.772094074430852e-06,
+      "loss": 0.1671,
+      "step": 3016
+    },
+    {
+      "epoch": 0.48882047958522357,
+      "grad_norm": 0.9789111614227295,
+      "learning_rate": 4.771911618245148e-06,
+      "loss": 0.1515,
+      "step": 3017
+    },
+    {
+      "epoch": 0.48898250162022033,
+      "grad_norm": 1.0732017755508423,
+      "learning_rate": 4.771729092544446e-06,
+      "loss": 0.1736,
+      "step": 3018
+    },
+    {
+      "epoch": 0.4891445236552171,
+      "grad_norm": 0.9397895932197571,
+      "learning_rate": 4.771546497334331e-06,
+      "loss": 0.1399,
+      "step": 3019
+    },
+    {
+      "epoch": 0.48930654569021387,
+      "grad_norm": 0.9154207706451416,
+      "learning_rate": 4.771363832620391e-06,
+      "loss": 0.146,
+      "step": 3020
+    },
+    {
+      "epoch": 0.48946856772521063,
+      "grad_norm": 0.9853832125663757,
+      "learning_rate": 4.771181098408214e-06,
+      "loss": 0.1648,
+      "step": 3021
+    },
+    {
+      "epoch": 0.4896305897602074,
+      "grad_norm": 1.054186463356018,
+      "learning_rate": 4.770998294703392e-06,
+      "loss": 0.1758,
+      "step": 3022
+    },
+    {
+      "epoch": 0.48979261179520417,
+      "grad_norm": 1.047492504119873,
+      "learning_rate": 4.770815421511517e-06,
+      "loss": 0.1563,
+      "step": 3023
+    },
+    {
+      "epoch": 0.4899546338302009,
+      "grad_norm": 0.9914636015892029,
+      "learning_rate": 4.7706324788381865e-06,
+      "loss": 0.1574,
+      "step": 3024
+    },
+    {
+      "epoch": 0.49011665586519765,
+      "grad_norm": 1.0021421909332275,
+      "learning_rate": 4.770449466688996e-06,
+      "loss": 0.1657,
+      "step": 3025
+    },
+    {
+      "epoch": 0.4902786779001944,
+      "grad_norm": 0.9965320229530334,
+      "learning_rate": 4.770266385069547e-06,
+      "loss": 0.1498,
+      "step": 3026
+    },
+    {
+      "epoch": 0.4904406999351912,
+      "grad_norm": 1.0368542671203613,
+      "learning_rate": 4.77008323398544e-06,
+      "loss": 0.1701,
+      "step": 3027
+    },
+    {
+      "epoch": 0.49060272197018795,
+      "grad_norm": 0.9540989398956299,
+      "learning_rate": 4.769900013442279e-06,
+      "loss": 0.1525,
+      "step": 3028
+    },
+    {
+      "epoch": 0.4907647440051847,
+      "grad_norm": 0.9963111877441406,
+      "learning_rate": 4.769716723445672e-06,
+      "loss": 0.1578,
+      "step": 3029
+    },
+    {
+      "epoch": 0.4909267660401815,
+      "grad_norm": 0.9703772664070129,
+      "learning_rate": 4.769533364001225e-06,
+      "loss": 0.1462,
+      "step": 3030
+    },
+    {
+      "epoch": 0.49108878807517825,
+      "grad_norm": 1.1421406269073486,
+      "learning_rate": 4.769349935114549e-06,
+      "loss": 0.1572,
+      "step": 3031
+    },
+    {
+      "epoch": 0.49125081011017496,
+      "grad_norm": 0.9838953018188477,
+      "learning_rate": 4.769166436791257e-06,
+      "loss": 0.1629,
+      "step": 3032
+    },
+    {
+      "epoch": 0.49141283214517173,
+      "grad_norm": 0.8788158893585205,
+      "learning_rate": 4.768982869036964e-06,
+      "loss": 0.1342,
+      "step": 3033
+    },
+    {
+      "epoch": 0.4915748541801685,
+      "grad_norm": 0.9803071618080139,
+      "learning_rate": 4.768799231857285e-06,
+      "loss": 0.1427,
+      "step": 3034
+    },
+    {
+      "epoch": 0.49173687621516526,
+      "grad_norm": 0.9650258421897888,
+      "learning_rate": 4.76861552525784e-06,
+      "loss": 0.1321,
+      "step": 3035
+    },
+    {
+      "epoch": 0.49189889825016203,
+      "grad_norm": 0.854787290096283,
+      "learning_rate": 4.768431749244251e-06,
+      "loss": 0.1314,
+      "step": 3036
+    },
+    {
+      "epoch": 0.4920609202851588,
+      "grad_norm": 1.0625677108764648,
+      "learning_rate": 4.768247903822139e-06,
+      "loss": 0.1617,
+      "step": 3037
+    },
+    {
+      "epoch": 0.49222294232015557,
+      "grad_norm": 1.0747294425964355,
+      "learning_rate": 4.76806398899713e-06,
+      "loss": 0.1586,
+      "step": 3038
+    },
+    {
+      "epoch": 0.4923849643551523,
+      "grad_norm": 1.0550459623336792,
+      "learning_rate": 4.767880004774853e-06,
+      "loss": 0.1567,
+      "step": 3039
+    },
+    {
+      "epoch": 0.49254698639014904,
+      "grad_norm": 1.0270591974258423,
+      "learning_rate": 4.767695951160934e-06,
+      "loss": 0.1546,
+      "step": 3040
+    },
+    {
+      "epoch": 0.4927090084251458,
+      "grad_norm": 0.9791101813316345,
+      "learning_rate": 4.767511828161008e-06,
+      "loss": 0.1442,
+      "step": 3041
+    },
+    {
+      "epoch": 0.4928710304601426,
+      "grad_norm": 1.110756754875183,
+      "learning_rate": 4.767327635780707e-06,
+      "loss": 0.1766,
+      "step": 3042
+    },
+    {
+      "epoch": 0.49303305249513935,
+      "grad_norm": 1.0451438426971436,
+      "learning_rate": 4.7671433740256664e-06,
+      "loss": 0.1476,
+      "step": 3043
+    },
+    {
+      "epoch": 0.4931950745301361,
+      "grad_norm": 1.0751953125,
+      "learning_rate": 4.7669590429015265e-06,
+      "loss": 0.1748,
+      "step": 3044
+    },
+    {
+      "epoch": 0.4933570965651329,
+      "grad_norm": 0.9066848158836365,
+      "learning_rate": 4.766774642413925e-06,
+      "loss": 0.1459,
+      "step": 3045
+    },
+    {
+      "epoch": 0.4935191186001296,
+      "grad_norm": 0.9705883860588074,
+      "learning_rate": 4.7665901725685045e-06,
+      "loss": 0.1495,
+      "step": 3046
+    },
+    {
+      "epoch": 0.49368114063512636,
+      "grad_norm": 1.0257165431976318,
+      "learning_rate": 4.76640563337091e-06,
+      "loss": 0.1661,
+      "step": 3047
+    },
+    {
+      "epoch": 0.4938431626701231,
+      "grad_norm": 0.933527410030365,
+      "learning_rate": 4.766221024826788e-06,
+      "loss": 0.1444,
+      "step": 3048
+    },
+    {
+      "epoch": 0.4940051847051199,
+      "grad_norm": 0.9257835745811462,
+      "learning_rate": 4.766036346941787e-06,
+      "loss": 0.1586,
+      "step": 3049
+    },
+    {
+      "epoch": 0.49416720674011666,
+      "grad_norm": 1.0416429042816162,
+      "learning_rate": 4.765851599721557e-06,
+      "loss": 0.1379,
+      "step": 3050
+    },
+    {
+      "epoch": 0.4943292287751134,
+      "grad_norm": 1.0839523077011108,
+      "learning_rate": 4.7656667831717514e-06,
+      "loss": 0.1593,
+      "step": 3051
+    },
+    {
+      "epoch": 0.4944912508101102,
+      "grad_norm": 1.0328129529953003,
+      "learning_rate": 4.765481897298025e-06,
+      "loss": 0.1518,
+      "step": 3052
+    },
+    {
+      "epoch": 0.49465327284510696,
+      "grad_norm": 1.078291654586792,
+      "learning_rate": 4.765296942106035e-06,
+      "loss": 0.1678,
+      "step": 3053
+    },
+    {
+      "epoch": 0.4948152948801037,
+      "grad_norm": 0.9669474363327026,
+      "learning_rate": 4.76511191760144e-06,
+      "loss": 0.1472,
+      "step": 3054
+    },
+    {
+      "epoch": 0.49497731691510044,
+      "grad_norm": 0.9748349189758301,
+      "learning_rate": 4.764926823789903e-06,
+      "loss": 0.1488,
+      "step": 3055
+    },
+    {
+      "epoch": 0.4951393389500972,
+      "grad_norm": 0.9624823927879333,
+      "learning_rate": 4.764741660677085e-06,
+      "loss": 0.1376,
+      "step": 3056
+    },
+    {
+      "epoch": 0.495301360985094,
+      "grad_norm": 1.2067596912384033,
+      "learning_rate": 4.7645564282686534e-06,
+      "loss": 0.189,
+      "step": 3057
+    },
+    {
+      "epoch": 0.49546338302009074,
+      "grad_norm": 1.1277415752410889,
+      "learning_rate": 4.764371126570275e-06,
+      "loss": 0.1515,
+      "step": 3058
+    },
+    {
+      "epoch": 0.4956254050550875,
+      "grad_norm": 1.0058308839797974,
+      "learning_rate": 4.76418575558762e-06,
+      "loss": 0.1585,
+      "step": 3059
+    },
+    {
+      "epoch": 0.4957874270900843,
+      "grad_norm": 0.9348163604736328,
+      "learning_rate": 4.76400031532636e-06,
+      "loss": 0.1388,
+      "step": 3060
+    },
+    {
+      "epoch": 0.495949449125081,
+      "grad_norm": 0.9315893054008484,
+      "learning_rate": 4.763814805792169e-06,
+      "loss": 0.1327,
+      "step": 3061
+    },
+    {
+      "epoch": 0.49611147116007775,
+      "grad_norm": 0.9435446262359619,
+      "learning_rate": 4.763629226990724e-06,
+      "loss": 0.1536,
+      "step": 3062
+    },
+    {
+      "epoch": 0.4962734931950745,
+      "grad_norm": 1.0565837621688843,
+      "learning_rate": 4.763443578927701e-06,
+      "loss": 0.1814,
+      "step": 3063
+    },
+    {
+      "epoch": 0.4964355152300713,
+      "grad_norm": 0.9601361155509949,
+      "learning_rate": 4.763257861608783e-06,
+      "loss": 0.1412,
+      "step": 3064
+    },
+    {
+      "epoch": 0.49659753726506806,
+      "grad_norm": 0.9786501526832581,
+      "learning_rate": 4.763072075039651e-06,
+      "loss": 0.1485,
+      "step": 3065
+    },
+    {
+      "epoch": 0.4967595593000648,
+      "grad_norm": 0.9912653565406799,
+      "learning_rate": 4.762886219225991e-06,
+      "loss": 0.1415,
+      "step": 3066
+    },
+    {
+      "epoch": 0.4969215813350616,
+      "grad_norm": 0.8689770102500916,
+      "learning_rate": 4.762700294173487e-06,
+      "loss": 0.1392,
+      "step": 3067
+    },
+    {
+      "epoch": 0.4970836033700583,
+      "grad_norm": 1.0026378631591797,
+      "learning_rate": 4.762514299887831e-06,
+      "loss": 0.1389,
+      "step": 3068
+    },
+    {
+      "epoch": 0.49724562540505507,
+      "grad_norm": 0.907612144947052,
+      "learning_rate": 4.762328236374713e-06,
+      "loss": 0.1334,
+      "step": 3069
+    },
+    {
+      "epoch": 0.49740764744005184,
+      "grad_norm": 1.0089789628982544,
+      "learning_rate": 4.762142103639824e-06,
+      "loss": 0.1653,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4975696694750486,
+      "grad_norm": 1.0177544355392456,
+      "learning_rate": 4.761955901688862e-06,
+      "loss": 0.1704,
+      "step": 3071
+    },
+    {
+      "epoch": 0.49773169151004537,
+      "grad_norm": 1.010158658027649,
+      "learning_rate": 4.761769630527523e-06,
+      "loss": 0.1641,
+      "step": 3072
+    },
+    {
+      "epoch": 0.49789371354504214,
+      "grad_norm": 0.8676300644874573,
+      "learning_rate": 4.761583290161507e-06,
+      "loss": 0.1422,
+      "step": 3073
+    },
+    {
+      "epoch": 0.4980557355800389,
+      "grad_norm": 0.9401457905769348,
+      "learning_rate": 4.761396880596515e-06,
+      "loss": 0.1441,
+      "step": 3074
+    },
+    {
+      "epoch": 0.4982177576150357,
+      "grad_norm": 1.0235344171524048,
+      "learning_rate": 4.761210401838251e-06,
+      "loss": 0.1493,
+      "step": 3075
+    },
+    {
+      "epoch": 0.4983797796500324,
+      "grad_norm": 0.9820840358734131,
+      "learning_rate": 4.76102385389242e-06,
+      "loss": 0.135,
+      "step": 3076
+    },
+    {
+      "epoch": 0.49854180168502915,
+      "grad_norm": 0.96462482213974,
+      "learning_rate": 4.760837236764731e-06,
+      "loss": 0.1605,
+      "step": 3077
+    },
+    {
+      "epoch": 0.4987038237200259,
+      "grad_norm": 0.929373025894165,
+      "learning_rate": 4.760650550460895e-06,
+      "loss": 0.1183,
+      "step": 3078
+    },
+    {
+      "epoch": 0.4988658457550227,
+      "grad_norm": 0.9171031713485718,
+      "learning_rate": 4.760463794986622e-06,
+      "loss": 0.1457,
+      "step": 3079
+    },
+    {
+      "epoch": 0.49902786779001945,
+      "grad_norm": 0.9517503976821899,
+      "learning_rate": 4.760276970347627e-06,
+      "loss": 0.136,
+      "step": 3080
+    },
+    {
+      "epoch": 0.4991898898250162,
+      "grad_norm": 0.9725329875946045,
+      "learning_rate": 4.760090076549626e-06,
+      "loss": 0.152,
+      "step": 3081
+    },
+    {
+      "epoch": 0.499351911860013,
+      "grad_norm": 0.9958840012550354,
+      "learning_rate": 4.759903113598338e-06,
+      "loss": 0.1427,
+      "step": 3082
+    },
+    {
+      "epoch": 0.4995139338950097,
+      "grad_norm": 1.0073212385177612,
+      "learning_rate": 4.759716081499484e-06,
+      "loss": 0.154,
+      "step": 3083
+    },
+    {
+      "epoch": 0.49967595593000647,
+      "grad_norm": 0.9372782707214355,
+      "learning_rate": 4.759528980258786e-06,
+      "loss": 0.1594,
+      "step": 3084
+    },
+    {
+      "epoch": 0.49983797796500323,
+      "grad_norm": 0.9391396641731262,
+      "learning_rate": 4.7593418098819695e-06,
+      "loss": 0.1285,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9143754839897156,
+      "learning_rate": 4.759154570374761e-06,
+      "loss": 0.1312,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5001620220349967,
+      "grad_norm": 1.103759765625,
+      "learning_rate": 4.75896726174289e-06,
+      "loss": 0.1854,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5003240440699935,
+      "grad_norm": 0.9957083463668823,
+      "learning_rate": 4.758779883992087e-06,
+      "loss": 0.1459,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5004860661049902,
+      "grad_norm": 1.0584758520126343,
+      "learning_rate": 4.758592437128086e-06,
+      "loss": 0.1514,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5006480881399871,
+      "grad_norm": 0.9559106230735779,
+      "learning_rate": 4.758404921156622e-06,
+      "loss": 0.1483,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5008101101749838,
+      "grad_norm": 1.1724073886871338,
+      "learning_rate": 4.7582173360834326e-06,
+      "loss": 0.1854,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5009721322099806,
+      "grad_norm": 1.016783595085144,
+      "learning_rate": 4.7580296819142565e-06,
+      "loss": 0.1646,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5011341542449773,
+      "grad_norm": 0.9434977769851685,
+      "learning_rate": 4.757841958654838e-06,
+      "loss": 0.1467,
+      "step": 3093
+    },
+    {
+      "epoch": 0.501296176279974,
+      "grad_norm": 0.8541172742843628,
+      "learning_rate": 4.757654166310919e-06,
+      "loss": 0.1325,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5014581983149708,
+      "grad_norm": 0.8900263905525208,
+      "learning_rate": 4.757466304888245e-06,
+      "loss": 0.1388,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5016202203499676,
+      "grad_norm": 0.9139759540557861,
+      "learning_rate": 4.757278374392567e-06,
+      "loss": 0.1381,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5017822423849644,
+      "grad_norm": 0.8878260254859924,
+      "learning_rate": 4.757090374829631e-06,
+      "loss": 0.1404,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5019442644199611,
+      "grad_norm": 1.07386314868927,
+      "learning_rate": 4.7569023062051936e-06,
+      "loss": 0.1747,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5021062864549579,
+      "grad_norm": 0.9467965960502625,
+      "learning_rate": 4.756714168525006e-06,
+      "loss": 0.1505,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5022683084899546,
+      "grad_norm": 0.9320406913757324,
+      "learning_rate": 4.756525961794826e-06,
+      "loss": 0.1414,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5024303305249513,
+      "grad_norm": 0.9513562321662903,
+      "learning_rate": 4.756337686020413e-06,
+      "loss": 0.1516,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5025923525599482,
+      "grad_norm": 0.9535584449768066,
+      "learning_rate": 4.756149341207526e-06,
+      "loss": 0.1461,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5027543745949449,
+      "grad_norm": 0.9998999834060669,
+      "learning_rate": 4.75596092736193e-06,
+      "loss": 0.1684,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5029163966299417,
+      "grad_norm": 1.0532805919647217,
+      "learning_rate": 4.755772444489388e-06,
+      "loss": 0.1307,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5030784186649384,
+      "grad_norm": 0.9271810054779053,
+      "learning_rate": 4.7555838925956686e-06,
+      "loss": 0.1446,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5032404406999352,
+      "grad_norm": 1.0862815380096436,
+      "learning_rate": 4.75539527168654e-06,
+      "loss": 0.1484,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5034024627349319,
+      "grad_norm": 1.0652508735656738,
+      "learning_rate": 4.755206581767775e-06,
+      "loss": 0.1749,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5035644847699287,
+      "grad_norm": 0.9044068455696106,
+      "learning_rate": 4.755017822845145e-06,
+      "loss": 0.1525,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5037265068049255,
+      "grad_norm": 0.9966428279876709,
+      "learning_rate": 4.754828994924428e-06,
+      "loss": 0.1536,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5038885288399222,
+      "grad_norm": 1.0621579885482788,
+      "learning_rate": 4.754640098011399e-06,
+      "loss": 0.1773,
+      "step": 3110
+    },
+    {
+      "epoch": 0.504050550874919,
+      "grad_norm": 1.0683941841125488,
+      "learning_rate": 4.754451132111839e-06,
+      "loss": 0.1586,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5042125729099157,
+      "grad_norm": 0.7734267115592957,
+      "learning_rate": 4.754262097231531e-06,
+      "loss": 0.1281,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5043745949449125,
+      "grad_norm": 1.0103263854980469,
+      "learning_rate": 4.754072993376258e-06,
+      "loss": 0.1583,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5045366169799093,
+      "grad_norm": 0.9807709455490112,
+      "learning_rate": 4.753883820551806e-06,
+      "loss": 0.148,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5046986390149061,
+      "grad_norm": 1.0581525564193726,
+      "learning_rate": 4.753694578763963e-06,
+      "loss": 0.1637,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5048606610499028,
+      "grad_norm": 1.050081491470337,
+      "learning_rate": 4.75350526801852e-06,
+      "loss": 0.1468,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5050226830848995,
+      "grad_norm": 0.991671621799469,
+      "learning_rate": 4.753315888321269e-06,
+      "loss": 0.1562,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5051847051198963,
+      "grad_norm": 0.9450017809867859,
+      "learning_rate": 4.753126439678005e-06,
+      "loss": 0.1554,
+      "step": 3118
+    },
+    {
+      "epoch": 0.505346727154893,
+      "grad_norm": 0.8860989809036255,
+      "learning_rate": 4.752936922094524e-06,
+      "loss": 0.1439,
+      "step": 3119
+    },
+    {
+      "epoch": 0.5055087491898899,
+      "grad_norm": 1.0846068859100342,
+      "learning_rate": 4.752747335576626e-06,
+      "loss": 0.1612,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5056707712248866,
+      "grad_norm": 1.1103551387786865,
+      "learning_rate": 4.75255768013011e-06,
+      "loss": 0.1659,
+      "step": 3121
+    },
+    {
+      "epoch": 0.5058327932598834,
+      "grad_norm": 0.9545550346374512,
+      "learning_rate": 4.752367955760781e-06,
+      "loss": 0.1395,
+      "step": 3122
+    },
+    {
+      "epoch": 0.5059948152948801,
+      "grad_norm": 1.050429344177246,
+      "learning_rate": 4.752178162474443e-06,
+      "loss": 0.1587,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5061568373298768,
+      "grad_norm": 0.9017760157585144,
+      "learning_rate": 4.751988300276903e-06,
+      "loss": 0.1398,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5063188593648736,
+      "grad_norm": 1.020270586013794,
+      "learning_rate": 4.751798369173971e-06,
+      "loss": 0.1452,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5064808813998704,
+      "grad_norm": 1.0533658266067505,
+      "learning_rate": 4.751608369171458e-06,
+      "loss": 0.1577,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5066429034348672,
+      "grad_norm": 0.9330847263336182,
+      "learning_rate": 4.751418300275178e-06,
+      "loss": 0.142,
+      "step": 3127
+    },
+    {
+      "epoch": 0.5068049254698639,
+      "grad_norm": 0.9158099293708801,
+      "learning_rate": 4.751228162490946e-06,
+      "loss": 0.1455,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5069669475048607,
+      "grad_norm": 0.9073064923286438,
+      "learning_rate": 4.75103795582458e-06,
+      "loss": 0.136,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5071289695398574,
+      "grad_norm": 1.0237650871276855,
+      "learning_rate": 4.750847680281901e-06,
+      "loss": 0.1599,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5072909915748541,
+      "grad_norm": 0.9799794554710388,
+      "learning_rate": 4.750657335868728e-06,
+      "loss": 0.1526,
+      "step": 3131
+    },
+    {
+      "epoch": 0.507453013609851,
+      "grad_norm": 0.9370033740997314,
+      "learning_rate": 4.750466922590888e-06,
+      "loss": 0.1508,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5076150356448477,
+      "grad_norm": 0.8960480093955994,
+      "learning_rate": 4.750276440454207e-06,
+      "loss": 0.1249,
+      "step": 3133
+    },
+    {
+      "epoch": 0.5077770576798445,
+      "grad_norm": 0.9389232993125916,
+      "learning_rate": 4.750085889464512e-06,
+      "loss": 0.1578,
+      "step": 3134
+    },
+    {
+      "epoch": 0.5079390797148412,
+      "grad_norm": 0.9818000197410583,
+      "learning_rate": 4.749895269627633e-06,
+      "loss": 0.1501,
+      "step": 3135
+    },
+    {
+      "epoch": 0.508101101749838,
+      "grad_norm": 0.9019086956977844,
+      "learning_rate": 4.749704580949404e-06,
+      "loss": 0.1541,
+      "step": 3136
+    },
+    {
+      "epoch": 0.5082631237848347,
+      "grad_norm": 1.0174736976623535,
+      "learning_rate": 4.749513823435659e-06,
+      "loss": 0.1599,
+      "step": 3137
+    },
+    {
+      "epoch": 0.5084251458198314,
+      "grad_norm": 1.0276374816894531,
+      "learning_rate": 4.749322997092235e-06,
+      "loss": 0.1574,
+      "step": 3138
+    },
+    {
+      "epoch": 0.5085871678548283,
+      "grad_norm": 0.8995428681373596,
+      "learning_rate": 4.74913210192497e-06,
+      "loss": 0.1356,
+      "step": 3139
+    },
+    {
+      "epoch": 0.508749189889825,
+      "grad_norm": 1.0277477502822876,
+      "learning_rate": 4.748941137939706e-06,
+      "loss": 0.1604,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5089112119248218,
+      "grad_norm": 1.0544497966766357,
+      "learning_rate": 4.748750105142285e-06,
+      "loss": 0.1749,
+      "step": 3141
+    },
+    {
+      "epoch": 0.5090732339598185,
+      "grad_norm": 0.8580374717712402,
+      "learning_rate": 4.748559003538553e-06,
+      "loss": 0.1397,
+      "step": 3142
+    },
+    {
+      "epoch": 0.5092352559948153,
+      "grad_norm": 1.0036252737045288,
+      "learning_rate": 4.748367833134357e-06,
+      "loss": 0.1463,
+      "step": 3143
+    },
+    {
+      "epoch": 0.509397278029812,
+      "grad_norm": 0.8418014049530029,
+      "learning_rate": 4.748176593935546e-06,
+      "loss": 0.1334,
+      "step": 3144
+    },
+    {
+      "epoch": 0.5095593000648088,
+      "grad_norm": 1.012681245803833,
+      "learning_rate": 4.747985285947972e-06,
+      "loss": 0.1758,
+      "step": 3145
+    },
+    {
+      "epoch": 0.5097213220998056,
+      "grad_norm": 0.9466359615325928,
+      "learning_rate": 4.7477939091774885e-06,
+      "loss": 0.1552,
+      "step": 3146
+    },
+    {
+      "epoch": 0.5098833441348023,
+      "grad_norm": 1.010433316230774,
+      "learning_rate": 4.74760246362995e-06,
+      "loss": 0.158,
+      "step": 3147
+    },
+    {
+      "epoch": 0.5100453661697991,
+      "grad_norm": 0.9007865786552429,
+      "learning_rate": 4.7474109493112154e-06,
+      "loss": 0.1287,
+      "step": 3148
+    },
+    {
+      "epoch": 0.5102073882047958,
+      "grad_norm": 0.9816288352012634,
+      "learning_rate": 4.747219366227145e-06,
+      "loss": 0.147,
+      "step": 3149
+    },
+    {
+      "epoch": 0.5103694102397927,
+      "grad_norm": 1.0334168672561646,
+      "learning_rate": 4.7470277143836e-06,
+      "loss": 0.1545,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5105314322747894,
+      "grad_norm": 0.978326678276062,
+      "learning_rate": 4.746835993786445e-06,
+      "loss": 0.1715,
+      "step": 3151
+    },
+    {
+      "epoch": 0.5106934543097861,
+      "grad_norm": 0.9503743052482605,
+      "learning_rate": 4.746644204441545e-06,
+      "loss": 0.147,
+      "step": 3152
+    },
+    {
+      "epoch": 0.5108554763447829,
+      "grad_norm": 0.8829200863838196,
+      "learning_rate": 4.7464523463547695e-06,
+      "loss": 0.1584,
+      "step": 3153
+    },
+    {
+      "epoch": 0.5110174983797796,
+      "grad_norm": 1.0434906482696533,
+      "learning_rate": 4.746260419531989e-06,
+      "loss": 0.1657,
+      "step": 3154
+    },
+    {
+      "epoch": 0.5111795204147764,
+      "grad_norm": 0.9752653241157532,
+      "learning_rate": 4.746068423979074e-06,
+      "loss": 0.1518,
+      "step": 3155
+    },
+    {
+      "epoch": 0.5113415424497731,
+      "grad_norm": 1.0544825792312622,
+      "learning_rate": 4.745876359701902e-06,
+      "loss": 0.1599,
+      "step": 3156
+    },
+    {
+      "epoch": 0.51150356448477,
+      "grad_norm": 1.0769002437591553,
+      "learning_rate": 4.745684226706348e-06,
+      "loss": 0.154,
+      "step": 3157
+    },
+    {
+      "epoch": 0.5116655865197667,
+      "grad_norm": 0.8938611745834351,
+      "learning_rate": 4.745492024998291e-06,
+      "loss": 0.1473,
+      "step": 3158
+    },
+    {
+      "epoch": 0.5118276085547635,
+      "grad_norm": 0.9698961973190308,
+      "learning_rate": 4.745299754583612e-06,
+      "loss": 0.149,
+      "step": 3159
+    },
+    {
+      "epoch": 0.5119896305897602,
+      "grad_norm": 0.9677153825759888,
+      "learning_rate": 4.745107415468194e-06,
+      "loss": 0.1546,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5121516526247569,
+      "grad_norm": 1.0147168636322021,
+      "learning_rate": 4.744915007657922e-06,
+      "loss": 0.1603,
+      "step": 3161
+    },
+    {
+      "epoch": 0.5123136746597537,
+      "grad_norm": 1.016696572303772,
+      "learning_rate": 4.744722531158683e-06,
+      "loss": 0.1499,
+      "step": 3162
+    },
+    {
+      "epoch": 0.5124756966947505,
+      "grad_norm": 0.9147832989692688,
+      "learning_rate": 4.744529985976368e-06,
+      "loss": 0.1331,
+      "step": 3163
+    },
+    {
+      "epoch": 0.5126377187297473,
+      "grad_norm": 1.0391688346862793,
+      "learning_rate": 4.744337372116866e-06,
+      "loss": 0.1582,
+      "step": 3164
+    },
+    {
+      "epoch": 0.512799740764744,
+      "grad_norm": 0.8867128491401672,
+      "learning_rate": 4.744144689586072e-06,
+      "loss": 0.135,
+      "step": 3165
+    },
+    {
+      "epoch": 0.5129617627997408,
+      "grad_norm": 0.9849241971969604,
+      "learning_rate": 4.743951938389881e-06,
+      "loss": 0.1613,
+      "step": 3166
+    },
+    {
+      "epoch": 0.5131237848347375,
+      "grad_norm": 1.0392513275146484,
+      "learning_rate": 4.743759118534191e-06,
+      "loss": 0.1329,
+      "step": 3167
+    },
+    {
+      "epoch": 0.5132858068697342,
+      "grad_norm": 0.9350460767745972,
+      "learning_rate": 4.743566230024902e-06,
+      "loss": 0.146,
+      "step": 3168
+    },
+    {
+      "epoch": 0.5134478289047311,
+      "grad_norm": 1.038048505783081,
+      "learning_rate": 4.743373272867916e-06,
+      "loss": 0.1474,
+      "step": 3169
+    },
+    {
+      "epoch": 0.5136098509397278,
+      "grad_norm": 0.8922699689865112,
+      "learning_rate": 4.7431802470691355e-06,
+      "loss": 0.1342,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5137718729747246,
+      "grad_norm": 1.0452051162719727,
+      "learning_rate": 4.742987152634469e-06,
+      "loss": 0.1422,
+      "step": 3171
+    },
+    {
+      "epoch": 0.5139338950097213,
+      "grad_norm": 0.9297380447387695,
+      "learning_rate": 4.7427939895698235e-06,
+      "loss": 0.1562,
+      "step": 3172
+    },
+    {
+      "epoch": 0.5140959170447181,
+      "grad_norm": 1.0185930728912354,
+      "learning_rate": 4.7426007578811085e-06,
+      "loss": 0.1658,
+      "step": 3173
+    },
+    {
+      "epoch": 0.5142579390797148,
+      "grad_norm": 0.8730944991111755,
+      "learning_rate": 4.742407457574238e-06,
+      "loss": 0.1304,
+      "step": 3174
+    },
+    {
+      "epoch": 0.5144199611147116,
+      "grad_norm": 1.0207183361053467,
+      "learning_rate": 4.742214088655126e-06,
+      "loss": 0.1488,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5145819831497084,
+      "grad_norm": 0.978741466999054,
+      "learning_rate": 4.7420206511296885e-06,
+      "loss": 0.1657,
+      "step": 3176
+    },
+    {
+      "epoch": 0.5147440051847051,
+      "grad_norm": 0.8174691200256348,
+      "learning_rate": 4.7418271450038444e-06,
+      "loss": 0.1233,
+      "step": 3177
+    },
+    {
+      "epoch": 0.5149060272197019,
+      "grad_norm": 0.9762808084487915,
+      "learning_rate": 4.7416335702835155e-06,
+      "loss": 0.1507,
+      "step": 3178
+    },
+    {
+      "epoch": 0.5150680492546986,
+      "grad_norm": 0.8684691786766052,
+      "learning_rate": 4.7414399269746235e-06,
+      "loss": 0.1492,
+      "step": 3179
+    },
+    {
+      "epoch": 0.5152300712896954,
+      "grad_norm": 0.9834610223770142,
+      "learning_rate": 4.741246215083094e-06,
+      "loss": 0.1429,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5153920933246922,
+      "grad_norm": 0.9805695414543152,
+      "learning_rate": 4.741052434614854e-06,
+      "loss": 0.1474,
+      "step": 3181
+    },
+    {
+      "epoch": 0.5155541153596889,
+      "grad_norm": 1.0447092056274414,
+      "learning_rate": 4.740858585575832e-06,
+      "loss": 0.1786,
+      "step": 3182
+    },
+    {
+      "epoch": 0.5157161373946857,
+      "grad_norm": 1.1218931674957275,
+      "learning_rate": 4.740664667971962e-06,
+      "loss": 0.1708,
+      "step": 3183
+    },
+    {
+      "epoch": 0.5158781594296824,
+      "grad_norm": 0.8798757195472717,
+      "learning_rate": 4.7404706818091736e-06,
+      "loss": 0.1393,
+      "step": 3184
+    },
+    {
+      "epoch": 0.5160401814646792,
+      "grad_norm": 0.8483942747116089,
+      "learning_rate": 4.740276627093405e-06,
+      "loss": 0.1447,
+      "step": 3185
+    },
+    {
+      "epoch": 0.5162022034996759,
+      "grad_norm": 1.1056780815124512,
+      "learning_rate": 4.740082503830593e-06,
+      "loss": 0.1656,
+      "step": 3186
+    },
+    {
+      "epoch": 0.5163642255346728,
+      "grad_norm": 0.8028669953346252,
+      "learning_rate": 4.739888312026677e-06,
+      "loss": 0.1112,
+      "step": 3187
+    },
+    {
+      "epoch": 0.5165262475696695,
+      "grad_norm": 1.00840163230896,
+      "learning_rate": 4.7396940516875996e-06,
+      "loss": 0.1385,
+      "step": 3188
+    },
+    {
+      "epoch": 0.5166882696046662,
+      "grad_norm": 1.072086215019226,
+      "learning_rate": 4.739499722819304e-06,
+      "loss": 0.1697,
+      "step": 3189
+    },
+    {
+      "epoch": 0.516850291639663,
+      "grad_norm": 0.9224776029586792,
+      "learning_rate": 4.739305325427736e-06,
+      "loss": 0.1426,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5170123136746597,
+      "grad_norm": 0.8584015369415283,
+      "learning_rate": 4.739110859518844e-06,
+      "loss": 0.1288,
+      "step": 3191
+    },
+    {
+      "epoch": 0.5171743357096565,
+      "grad_norm": 1.027614712715149,
+      "learning_rate": 4.738916325098579e-06,
+      "loss": 0.1508,
+      "step": 3192
+    },
+    {
+      "epoch": 0.5173363577446533,
+      "grad_norm": 1.064180612564087,
+      "learning_rate": 4.738721722172891e-06,
+      "loss": 0.1641,
+      "step": 3193
+    },
+    {
+      "epoch": 0.5174983797796501,
+      "grad_norm": 0.8860505223274231,
+      "learning_rate": 4.738527050747738e-06,
+      "loss": 0.1344,
+      "step": 3194
+    },
+    {
+      "epoch": 0.5176604018146468,
+      "grad_norm": 0.8795318603515625,
+      "learning_rate": 4.738332310829073e-06,
+      "loss": 0.1374,
+      "step": 3195
+    },
+    {
+      "epoch": 0.5178224238496435,
+      "grad_norm": 1.0304006338119507,
+      "learning_rate": 4.738137502422856e-06,
+      "loss": 0.1534,
+      "step": 3196
+    },
+    {
+      "epoch": 0.5179844458846403,
+      "grad_norm": 1.116926670074463,
+      "learning_rate": 4.737942625535048e-06,
+      "loss": 0.1539,
+      "step": 3197
+    },
+    {
+      "epoch": 0.518146467919637,
+      "grad_norm": 0.8668916821479797,
+      "learning_rate": 4.737747680171611e-06,
+      "loss": 0.1338,
+      "step": 3198
+    },
+    {
+      "epoch": 0.5183084899546339,
+      "grad_norm": 1.1048635244369507,
+      "learning_rate": 4.737552666338511e-06,
+      "loss": 0.1753,
+      "step": 3199
+    },
+    {
+      "epoch": 0.5184705119896306,
+      "grad_norm": 0.8767505884170532,
+      "learning_rate": 4.737357584041713e-06,
+      "loss": 0.1328,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5186325340246274,
+      "grad_norm": 0.9243037700653076,
+      "learning_rate": 4.737162433287188e-06,
+      "loss": 0.1562,
+      "step": 3201
+    },
+    {
+      "epoch": 0.5187945560596241,
+      "grad_norm": 1.0530647039413452,
+      "learning_rate": 4.7369672140809065e-06,
+      "loss": 0.1638,
+      "step": 3202
+    },
+    {
+      "epoch": 0.5189565780946209,
+      "grad_norm": 0.9306735396385193,
+      "learning_rate": 4.736771926428841e-06,
+      "loss": 0.1411,
+      "step": 3203
+    },
+    {
+      "epoch": 0.5191186001296176,
+      "grad_norm": 1.0215457677841187,
+      "learning_rate": 4.736576570336968e-06,
+      "loss": 0.1399,
+      "step": 3204
+    },
+    {
+      "epoch": 0.5192806221646143,
+      "grad_norm": 0.9567215442657471,
+      "learning_rate": 4.736381145811264e-06,
+      "loss": 0.1628,
+      "step": 3205
+    },
+    {
+      "epoch": 0.5194426441996112,
+      "grad_norm": 0.9945108294487,
+      "learning_rate": 4.736185652857709e-06,
+      "loss": 0.1668,
+      "step": 3206
+    },
+    {
+      "epoch": 0.5196046662346079,
+      "grad_norm": 0.9826516509056091,
+      "learning_rate": 4.735990091482284e-06,
+      "loss": 0.1528,
+      "step": 3207
+    },
+    {
+      "epoch": 0.5197666882696047,
+      "grad_norm": 0.8901511430740356,
+      "learning_rate": 4.7357944616909745e-06,
+      "loss": 0.14,
+      "step": 3208
+    },
+    {
+      "epoch": 0.5199287103046014,
+      "grad_norm": 0.9824097156524658,
+      "learning_rate": 4.735598763489764e-06,
+      "loss": 0.1584,
+      "step": 3209
+    },
+    {
+      "epoch": 0.5200907323395982,
+      "grad_norm": 0.8953663110733032,
+      "learning_rate": 4.735402996884642e-06,
+      "loss": 0.1347,
+      "step": 3210
+    },
+    {
+      "epoch": 0.520252754374595,
+      "grad_norm": 0.8961206674575806,
+      "learning_rate": 4.735207161881596e-06,
+      "loss": 0.1368,
+      "step": 3211
+    },
+    {
+      "epoch": 0.5204147764095917,
+      "grad_norm": 0.9631252884864807,
+      "learning_rate": 4.7350112584866225e-06,
+      "loss": 0.1638,
+      "step": 3212
+    },
+    {
+      "epoch": 0.5205767984445885,
+      "grad_norm": 0.7925556898117065,
+      "learning_rate": 4.734815286705712e-06,
+      "loss": 0.13,
+      "step": 3213
+    },
+    {
+      "epoch": 0.5207388204795852,
+      "grad_norm": 1.1014782190322876,
+      "learning_rate": 4.734619246544862e-06,
+      "loss": 0.1628,
+      "step": 3214
+    },
+    {
+      "epoch": 0.520900842514582,
+      "grad_norm": 0.9577755928039551,
+      "learning_rate": 4.73442313801007e-06,
+      "loss": 0.1562,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5210628645495787,
+      "grad_norm": 0.9577489495277405,
+      "learning_rate": 4.734226961107338e-06,
+      "loss": 0.1627,
+      "step": 3216
+    },
+    {
+      "epoch": 0.5212248865845756,
+      "grad_norm": 0.9195597171783447,
+      "learning_rate": 4.734030715842667e-06,
+      "loss": 0.1485,
+      "step": 3217
+    },
+    {
+      "epoch": 0.5213869086195723,
+      "grad_norm": 1.0537441968917847,
+      "learning_rate": 4.733834402222064e-06,
+      "loss": 0.1548,
+      "step": 3218
+    },
+    {
+      "epoch": 0.521548930654569,
+      "grad_norm": 1.0351758003234863,
+      "learning_rate": 4.733638020251532e-06,
+      "loss": 0.1574,
+      "step": 3219
+    },
+    {
+      "epoch": 0.5217109526895658,
+      "grad_norm": 0.934990406036377,
+      "learning_rate": 4.7334415699370825e-06,
+      "loss": 0.1425,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5218729747245625,
+      "grad_norm": 1.002612590789795,
+      "learning_rate": 4.733245051284727e-06,
+      "loss": 0.1538,
+      "step": 3221
+    },
+    {
+      "epoch": 0.5220349967595593,
+      "grad_norm": 1.078823208808899,
+      "learning_rate": 4.733048464300476e-06,
+      "loss": 0.1703,
+      "step": 3222
+    },
+    {
+      "epoch": 0.522197018794556,
+      "grad_norm": 0.9427096843719482,
+      "learning_rate": 4.732851808990346e-06,
+      "loss": 0.1432,
+      "step": 3223
+    },
+    {
+      "epoch": 0.5223590408295529,
+      "grad_norm": 0.9111664891242981,
+      "learning_rate": 4.732655085360355e-06,
+      "loss": 0.1408,
+      "step": 3224
+    },
+    {
+      "epoch": 0.5225210628645496,
+      "grad_norm": 1.06932532787323,
+      "learning_rate": 4.732458293416519e-06,
+      "loss": 0.1691,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5226830848995463,
+      "grad_norm": 1.0022755861282349,
+      "learning_rate": 4.7322614331648645e-06,
+      "loss": 0.161,
+      "step": 3226
+    },
+    {
+      "epoch": 0.5228451069345431,
+      "grad_norm": 0.9708278775215149,
+      "learning_rate": 4.73206450461141e-06,
+      "loss": 0.1496,
+      "step": 3227
+    },
+    {
+      "epoch": 0.5230071289695398,
+      "grad_norm": 1.0193688869476318,
+      "learning_rate": 4.731867507762184e-06,
+      "loss": 0.1484,
+      "step": 3228
+    },
+    {
+      "epoch": 0.5231691510045366,
+      "grad_norm": 0.90104079246521,
+      "learning_rate": 4.731670442623214e-06,
+      "loss": 0.1321,
+      "step": 3229
+    },
+    {
+      "epoch": 0.5233311730395334,
+      "grad_norm": 1.0102612972259521,
+      "learning_rate": 4.731473309200528e-06,
+      "loss": 0.1552,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5234931950745302,
+      "grad_norm": 0.8935184478759766,
+      "learning_rate": 4.731276107500159e-06,
+      "loss": 0.1449,
+      "step": 3231
+    },
+    {
+      "epoch": 0.5236552171095269,
+      "grad_norm": 0.8871548771858215,
+      "learning_rate": 4.731078837528141e-06,
+      "loss": 0.1353,
+      "step": 3232
+    },
+    {
+      "epoch": 0.5238172391445236,
+      "grad_norm": 1.071672797203064,
+      "learning_rate": 4.730881499290509e-06,
+      "loss": 0.1683,
+      "step": 3233
+    },
+    {
+      "epoch": 0.5239792611795204,
+      "grad_norm": 0.872079074382782,
+      "learning_rate": 4.730684092793302e-06,
+      "loss": 0.1411,
+      "step": 3234
+    },
+    {
+      "epoch": 0.5241412832145171,
+      "grad_norm": 0.931596040725708,
+      "learning_rate": 4.73048661804256e-06,
+      "loss": 0.1633,
+      "step": 3235
+    },
+    {
+      "epoch": 0.524303305249514,
+      "grad_norm": 0.8818705081939697,
+      "learning_rate": 4.730289075044326e-06,
+      "loss": 0.1281,
+      "step": 3236
+    },
+    {
+      "epoch": 0.5244653272845107,
+      "grad_norm": 0.9318626523017883,
+      "learning_rate": 4.730091463804642e-06,
+      "loss": 0.1438,
+      "step": 3237
+    },
+    {
+      "epoch": 0.5246273493195075,
+      "grad_norm": 1.0269670486450195,
+      "learning_rate": 4.729893784329557e-06,
+      "loss": 0.1696,
+      "step": 3238
+    },
+    {
+      "epoch": 0.5247893713545042,
+      "grad_norm": 0.8802597522735596,
+      "learning_rate": 4.729696036625119e-06,
+      "loss": 0.1319,
+      "step": 3239
+    },
+    {
+      "epoch": 0.5249513933895009,
+      "grad_norm": 0.9144219756126404,
+      "learning_rate": 4.729498220697377e-06,
+      "loss": 0.1412,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5251134154244977,
+      "grad_norm": 0.9358303546905518,
+      "learning_rate": 4.729300336552385e-06,
+      "loss": 0.1604,
+      "step": 3241
+    },
+    {
+      "epoch": 0.5252754374594945,
+      "grad_norm": 0.8559423089027405,
+      "learning_rate": 4.729102384196197e-06,
+      "loss": 0.1352,
+      "step": 3242
+    },
+    {
+      "epoch": 0.5254374594944913,
+      "grad_norm": 0.9450469017028809,
+      "learning_rate": 4.728904363634871e-06,
+      "loss": 0.1367,
+      "step": 3243
+    },
+    {
+      "epoch": 0.525599481529488,
+      "grad_norm": 0.9907295107841492,
+      "learning_rate": 4.728706274874465e-06,
+      "loss": 0.1418,
+      "step": 3244
+    },
+    {
+      "epoch": 0.5257615035644848,
+      "grad_norm": 0.9787341356277466,
+      "learning_rate": 4.72850811792104e-06,
+      "loss": 0.1519,
+      "step": 3245
+    },
+    {
+      "epoch": 0.5259235255994815,
+      "grad_norm": 0.9200270771980286,
+      "learning_rate": 4.72830989278066e-06,
+      "loss": 0.1517,
+      "step": 3246
+    },
+    {
+      "epoch": 0.5260855476344782,
+      "grad_norm": 0.9773831367492676,
+      "learning_rate": 4.72811159945939e-06,
+      "loss": 0.1637,
+      "step": 3247
+    },
+    {
+      "epoch": 0.5262475696694751,
+      "grad_norm": 1.007265329360962,
+      "learning_rate": 4.727913237963296e-06,
+      "loss": 0.1624,
+      "step": 3248
+    },
+    {
+      "epoch": 0.5264095917044718,
+      "grad_norm": 0.9802579283714294,
+      "learning_rate": 4.7277148082984495e-06,
+      "loss": 0.1675,
+      "step": 3249
+    },
+    {
+      "epoch": 0.5265716137394686,
+      "grad_norm": 0.9074774980545044,
+      "learning_rate": 4.72751631047092e-06,
+      "loss": 0.1461,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5267336357744653,
+      "grad_norm": 1.1058282852172852,
+      "learning_rate": 4.727317744486783e-06,
+      "loss": 0.1912,
+      "step": 3251
+    },
+    {
+      "epoch": 0.5268956578094621,
+      "grad_norm": 0.9917949438095093,
+      "learning_rate": 4.727119110352112e-06,
+      "loss": 0.1563,
+      "step": 3252
+    },
+    {
+      "epoch": 0.5270576798444588,
+      "grad_norm": 0.9093613624572754,
+      "learning_rate": 4.726920408072985e-06,
+      "loss": 0.1402,
+      "step": 3253
+    },
+    {
+      "epoch": 0.5272197018794557,
+      "grad_norm": 0.9544733166694641,
+      "learning_rate": 4.726721637655484e-06,
+      "loss": 0.1594,
+      "step": 3254
+    },
+    {
+      "epoch": 0.5273817239144524,
+      "grad_norm": 0.920921802520752,
+      "learning_rate": 4.726522799105689e-06,
+      "loss": 0.152,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5275437459494491,
+      "grad_norm": 1.037856101989746,
+      "learning_rate": 4.7263238924296835e-06,
+      "loss": 0.1575,
+      "step": 3256
+    },
+    {
+      "epoch": 0.5277057679844459,
+      "grad_norm": 0.9055295586585999,
+      "learning_rate": 4.726124917633556e-06,
+      "loss": 0.1405,
+      "step": 3257
+    },
+    {
+      "epoch": 0.5278677900194426,
+      "grad_norm": 0.952745258808136,
+      "learning_rate": 4.725925874723393e-06,
+      "loss": 0.1487,
+      "step": 3258
+    },
+    {
+      "epoch": 0.5280298120544394,
+      "grad_norm": 1.02610445022583,
+      "learning_rate": 4.725726763705284e-06,
+      "loss": 0.1586,
+      "step": 3259
+    },
+    {
+      "epoch": 0.5281918340894362,
+      "grad_norm": 0.9188981652259827,
+      "learning_rate": 4.725527584585322e-06,
+      "loss": 0.1264,
+      "step": 3260
+    },
+    {
+      "epoch": 0.528353856124433,
+      "grad_norm": 1.0480659008026123,
+      "learning_rate": 4.725328337369602e-06,
+      "loss": 0.149,
+      "step": 3261
+    },
+    {
+      "epoch": 0.5285158781594297,
+      "grad_norm": 0.9283803105354309,
+      "learning_rate": 4.725129022064221e-06,
+      "loss": 0.1442,
+      "step": 3262
+    },
+    {
+      "epoch": 0.5286779001944264,
+      "grad_norm": 0.8994842171669006,
+      "learning_rate": 4.7249296386752754e-06,
+      "loss": 0.1296,
+      "step": 3263
+    },
+    {
+      "epoch": 0.5288399222294232,
+      "grad_norm": 0.9448267817497253,
+      "learning_rate": 4.724730187208868e-06,
+      "loss": 0.1365,
+      "step": 3264
+    },
+    {
+      "epoch": 0.5290019442644199,
+      "grad_norm": 1.0473707914352417,
+      "learning_rate": 4.7245306676711e-06,
+      "loss": 0.1619,
+      "step": 3265
+    },
+    {
+      "epoch": 0.5291639662994168,
+      "grad_norm": 0.9334203004837036,
+      "learning_rate": 4.724331080068077e-06,
+      "loss": 0.1456,
+      "step": 3266
+    },
+    {
+      "epoch": 0.5293259883344135,
+      "grad_norm": 1.053633451461792,
+      "learning_rate": 4.724131424405906e-06,
+      "loss": 0.1759,
+      "step": 3267
+    },
+    {
+      "epoch": 0.5294880103694103,
+      "grad_norm": 0.9441131949424744,
+      "learning_rate": 4.723931700690695e-06,
+      "loss": 0.1525,
+      "step": 3268
+    },
+    {
+      "epoch": 0.529650032404407,
+      "grad_norm": 0.875206470489502,
+      "learning_rate": 4.723731908928556e-06,
+      "loss": 0.1461,
+      "step": 3269
+    },
+    {
+      "epoch": 0.5298120544394037,
+      "grad_norm": 0.9882826805114746,
+      "learning_rate": 4.7235320491256026e-06,
+      "loss": 0.1414,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5299740764744005,
+      "grad_norm": 0.8946253061294556,
+      "learning_rate": 4.723332121287949e-06,
+      "loss": 0.1441,
+      "step": 3271
+    },
+    {
+      "epoch": 0.5301360985093972,
+      "grad_norm": 0.8672228455543518,
+      "learning_rate": 4.723132125421712e-06,
+      "loss": 0.1381,
+      "step": 3272
+    },
+    {
+      "epoch": 0.5302981205443941,
+      "grad_norm": 0.9539285898208618,
+      "learning_rate": 4.7229320615330136e-06,
+      "loss": 0.1544,
+      "step": 3273
+    },
+    {
+      "epoch": 0.5304601425793908,
+      "grad_norm": 1.0140817165374756,
+      "learning_rate": 4.722731929627971e-06,
+      "loss": 0.1621,
+      "step": 3274
+    },
+    {
+      "epoch": 0.5306221646143876,
+      "grad_norm": 1.0421953201293945,
+      "learning_rate": 4.7225317297127125e-06,
+      "loss": 0.1633,
+      "step": 3275
+    },
+    {
+      "epoch": 0.5307841866493843,
+      "grad_norm": 1.0012775659561157,
+      "learning_rate": 4.722331461793361e-06,
+      "loss": 0.1464,
+      "step": 3276
+    },
+    {
+      "epoch": 0.530946208684381,
+      "grad_norm": 0.9863828420639038,
+      "learning_rate": 4.722131125876044e-06,
+      "loss": 0.1476,
+      "step": 3277
+    },
+    {
+      "epoch": 0.5311082307193778,
+      "grad_norm": 0.9448386430740356,
+      "learning_rate": 4.721930721966893e-06,
+      "loss": 0.1432,
+      "step": 3278
+    },
+    {
+      "epoch": 0.5312702527543746,
+      "grad_norm": 0.898563802242279,
+      "learning_rate": 4.721730250072038e-06,
+      "loss": 0.1365,
+      "step": 3279
+    },
+    {
+      "epoch": 0.5314322747893714,
+      "grad_norm": 0.8489564061164856,
+      "learning_rate": 4.7215297101976145e-06,
+      "loss": 0.1388,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5315942968243681,
+      "grad_norm": 0.9576735496520996,
+      "learning_rate": 4.721329102349757e-06,
+      "loss": 0.1527,
+      "step": 3281
+    },
+    {
+      "epoch": 0.5317563188593649,
+      "grad_norm": 0.9543820023536682,
+      "learning_rate": 4.721128426534605e-06,
+      "loss": 0.1396,
+      "step": 3282
+    },
+    {
+      "epoch": 0.5319183408943616,
+      "grad_norm": 1.031876564025879,
+      "learning_rate": 4.720927682758298e-06,
+      "loss": 0.1608,
+      "step": 3283
+    },
+    {
+      "epoch": 0.5320803629293583,
+      "grad_norm": 0.9371171593666077,
+      "learning_rate": 4.720726871026978e-06,
+      "loss": 0.1366,
+      "step": 3284
+    },
+    {
+      "epoch": 0.5322423849643552,
+      "grad_norm": 0.9507832527160645,
+      "learning_rate": 4.720525991346791e-06,
+      "loss": 0.1423,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5324044069993519,
+      "grad_norm": 1.2056124210357666,
+      "learning_rate": 4.720325043723881e-06,
+      "loss": 0.1797,
+      "step": 3286
+    },
+    {
+      "epoch": 0.5325664290343487,
+      "grad_norm": 1.0787984132766724,
+      "learning_rate": 4.720124028164399e-06,
+      "loss": 0.1548,
+      "step": 3287
+    },
+    {
+      "epoch": 0.5327284510693454,
+      "grad_norm": 1.0400218963623047,
+      "learning_rate": 4.719922944674494e-06,
+      "loss": 0.1581,
+      "step": 3288
+    },
+    {
+      "epoch": 0.5328904731043422,
+      "grad_norm": 0.9209017753601074,
+      "learning_rate": 4.719721793260318e-06,
+      "loss": 0.1439,
+      "step": 3289
+    },
+    {
+      "epoch": 0.5330524951393389,
+      "grad_norm": 0.8602655529975891,
+      "learning_rate": 4.719520573928028e-06,
+      "loss": 0.1304,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5332145171743357,
+      "grad_norm": 0.9528794884681702,
+      "learning_rate": 4.719319286683779e-06,
+      "loss": 0.1422,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5333765392093325,
+      "grad_norm": 1.055066466331482,
+      "learning_rate": 4.71911793153373e-06,
+      "loss": 0.1698,
+      "step": 3292
+    },
+    {
+      "epoch": 0.5335385612443292,
+      "grad_norm": 0.9883852601051331,
+      "learning_rate": 4.718916508484043e-06,
+      "loss": 0.1581,
+      "step": 3293
+    },
+    {
+      "epoch": 0.533700583279326,
+      "grad_norm": 0.9446104764938354,
+      "learning_rate": 4.7187150175408805e-06,
+      "loss": 0.1551,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5338626053143227,
+      "grad_norm": 1.1086978912353516,
+      "learning_rate": 4.7185134587104075e-06,
+      "loss": 0.1783,
+      "step": 3295
+    },
+    {
+      "epoch": 0.5340246273493195,
+      "grad_norm": 0.9049498438835144,
+      "learning_rate": 4.718311831998792e-06,
+      "loss": 0.146,
+      "step": 3296
+    },
+    {
+      "epoch": 0.5341866493843163,
+      "grad_norm": 1.0368142127990723,
+      "learning_rate": 4.718110137412201e-06,
+      "loss": 0.1789,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5343486714193131,
+      "grad_norm": 0.9038468599319458,
+      "learning_rate": 4.71790837495681e-06,
+      "loss": 0.1441,
+      "step": 3298
+    },
+    {
+      "epoch": 0.5345106934543098,
+      "grad_norm": 0.9993532299995422,
+      "learning_rate": 4.717706544638788e-06,
+      "loss": 0.1552,
+      "step": 3299
+    },
+    {
+      "epoch": 0.5346727154893065,
+      "grad_norm": 0.8008403778076172,
+      "learning_rate": 4.717504646464314e-06,
+      "loss": 0.1378,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5348347375243033,
+      "grad_norm": 0.9350391030311584,
+      "learning_rate": 4.717302680439563e-06,
+      "loss": 0.1507,
+      "step": 3301
+    },
+    {
+      "epoch": 0.5349967595593,
+      "grad_norm": 0.8926771879196167,
+      "learning_rate": 4.717100646570716e-06,
+      "loss": 0.1515,
+      "step": 3302
+    },
+    {
+      "epoch": 0.5351587815942969,
+      "grad_norm": 0.8540582060813904,
+      "learning_rate": 4.716898544863954e-06,
+      "loss": 0.1348,
+      "step": 3303
+    },
+    {
+      "epoch": 0.5353208036292936,
+      "grad_norm": 0.9826602935791016,
+      "learning_rate": 4.7166963753254616e-06,
+      "loss": 0.1562,
+      "step": 3304
+    },
+    {
+      "epoch": 0.5354828256642904,
+      "grad_norm": 0.9145976901054382,
+      "learning_rate": 4.716494137961425e-06,
+      "loss": 0.1367,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5356448476992871,
+      "grad_norm": 0.8795981407165527,
+      "learning_rate": 4.716291832778031e-06,
+      "loss": 0.1399,
+      "step": 3306
+    },
+    {
+      "epoch": 0.5358068697342838,
+      "grad_norm": 0.932141125202179,
+      "learning_rate": 4.71608945978147e-06,
+      "loss": 0.1474,
+      "step": 3307
+    },
+    {
+      "epoch": 0.5359688917692806,
+      "grad_norm": 0.9054761528968811,
+      "learning_rate": 4.715887018977935e-06,
+      "loss": 0.1486,
+      "step": 3308
+    },
+    {
+      "epoch": 0.5361309138042774,
+      "grad_norm": 0.9789231419563293,
+      "learning_rate": 4.715684510373619e-06,
+      "loss": 0.1493,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5362929358392742,
+      "grad_norm": 0.9172061681747437,
+      "learning_rate": 4.715481933974719e-06,
+      "loss": 0.1352,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5364549578742709,
+      "grad_norm": 0.9078491926193237,
+      "learning_rate": 4.715279289787434e-06,
+      "loss": 0.141,
+      "step": 3311
+    },
+    {
+      "epoch": 0.5366169799092677,
+      "grad_norm": 1.0831925868988037,
+      "learning_rate": 4.715076577817963e-06,
+      "loss": 0.153,
+      "step": 3312
+    },
+    {
+      "epoch": 0.5367790019442644,
+      "grad_norm": 1.0742647647857666,
+      "learning_rate": 4.714873798072509e-06,
+      "loss": 0.165,
+      "step": 3313
+    },
+    {
+      "epoch": 0.5369410239792611,
+      "grad_norm": 1.015164852142334,
+      "learning_rate": 4.714670950557276e-06,
+      "loss": 0.1408,
+      "step": 3314
+    },
+    {
+      "epoch": 0.537103046014258,
+      "grad_norm": 1.0785647630691528,
+      "learning_rate": 4.714468035278473e-06,
+      "loss": 0.1709,
+      "step": 3315
+    },
+    {
+      "epoch": 0.5372650680492547,
+      "grad_norm": 1.0501564741134644,
+      "learning_rate": 4.714265052242306e-06,
+      "loss": 0.1692,
+      "step": 3316
+    },
+    {
+      "epoch": 0.5374270900842515,
+      "grad_norm": 0.8828537464141846,
+      "learning_rate": 4.714062001454986e-06,
+      "loss": 0.1389,
+      "step": 3317
+    },
+    {
+      "epoch": 0.5375891121192482,
+      "grad_norm": 1.0610477924346924,
+      "learning_rate": 4.7138588829227285e-06,
+      "loss": 0.1675,
+      "step": 3318
+    },
+    {
+      "epoch": 0.537751134154245,
+      "grad_norm": 0.8431143760681152,
+      "learning_rate": 4.713655696651746e-06,
+      "loss": 0.1328,
+      "step": 3319
+    },
+    {
+      "epoch": 0.5379131561892417,
+      "grad_norm": 1.091422438621521,
+      "learning_rate": 4.713452442648255e-06,
+      "loss": 0.1691,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5380751782242384,
+      "grad_norm": 0.9422255158424377,
+      "learning_rate": 4.713249120918476e-06,
+      "loss": 0.1448,
+      "step": 3321
+    },
+    {
+      "epoch": 0.5382372002592353,
+      "grad_norm": 0.9391464591026306,
+      "learning_rate": 4.7130457314686316e-06,
+      "loss": 0.1406,
+      "step": 3322
+    },
+    {
+      "epoch": 0.538399222294232,
+      "grad_norm": 0.9336527585983276,
+      "learning_rate": 4.712842274304942e-06,
+      "loss": 0.1559,
+      "step": 3323
+    },
+    {
+      "epoch": 0.5385612443292288,
+      "grad_norm": 1.0600395202636719,
+      "learning_rate": 4.712638749433634e-06,
+      "loss": 0.1621,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5387232663642255,
+      "grad_norm": 0.8601197004318237,
+      "learning_rate": 4.712435156860934e-06,
+      "loss": 0.1338,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5388852883992223,
+      "grad_norm": 0.9096711277961731,
+      "learning_rate": 4.7122314965930724e-06,
+      "loss": 0.1335,
+      "step": 3326
+    },
+    {
+      "epoch": 0.539047310434219,
+      "grad_norm": 0.9492406845092773,
+      "learning_rate": 4.712027768636282e-06,
+      "loss": 0.1475,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5392093324692158,
+      "grad_norm": 1.070809245109558,
+      "learning_rate": 4.711823972996793e-06,
+      "loss": 0.1636,
+      "step": 3328
+    },
+    {
+      "epoch": 0.5393713545042126,
+      "grad_norm": 1.0436028242111206,
+      "learning_rate": 4.711620109680843e-06,
+      "loss": 0.1788,
+      "step": 3329
+    },
+    {
+      "epoch": 0.5395333765392093,
+      "grad_norm": 0.9128246307373047,
+      "learning_rate": 4.711416178694671e-06,
+      "loss": 0.1435,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5396953985742061,
+      "grad_norm": 0.9323081970214844,
+      "learning_rate": 4.7112121800445146e-06,
+      "loss": 0.1534,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5398574206092028,
+      "grad_norm": 1.0300610065460205,
+      "learning_rate": 4.711008113736617e-06,
+      "loss": 0.1807,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5400194426441997,
+      "grad_norm": 0.9709299206733704,
+      "learning_rate": 4.710803979777221e-06,
+      "loss": 0.1371,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5401814646791964,
+      "grad_norm": 0.9480953812599182,
+      "learning_rate": 4.710599778172575e-06,
+      "loss": 0.1549,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5403434867141931,
+      "grad_norm": 0.871699869632721,
+      "learning_rate": 4.710395508928923e-06,
+      "loss": 0.1341,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5405055087491899,
+      "grad_norm": 1.0588428974151611,
+      "learning_rate": 4.7101911720525186e-06,
+      "loss": 0.1603,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5406675307841866,
+      "grad_norm": 0.9104177951812744,
+      "learning_rate": 4.709986767549612e-06,
+      "loss": 0.1527,
+      "step": 3337
+    },
+    {
+      "epoch": 0.5408295528191834,
+      "grad_norm": 0.8339903950691223,
+      "learning_rate": 4.70978229542646e-06,
+      "loss": 0.1166,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5409915748541801,
+      "grad_norm": 1.0834025144577026,
+      "learning_rate": 4.709577755689316e-06,
+      "loss": 0.1623,
+      "step": 3339
+    },
+    {
+      "epoch": 0.541153596889177,
+      "grad_norm": 0.9392449259757996,
+      "learning_rate": 4.709373148344441e-06,
+      "loss": 0.1518,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5413156189241737,
+      "grad_norm": 1.014581561088562,
+      "learning_rate": 4.709168473398094e-06,
+      "loss": 0.1585,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5414776409591704,
+      "grad_norm": 0.9345294833183289,
+      "learning_rate": 4.708963730856536e-06,
+      "loss": 0.1385,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5416396629941672,
+      "grad_norm": 0.975787341594696,
+      "learning_rate": 4.708758920726036e-06,
+      "loss": 0.1515,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5418016850291639,
+      "grad_norm": 0.9591774940490723,
+      "learning_rate": 4.708554043012857e-06,
+      "loss": 0.1466,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5419637070641607,
+      "grad_norm": 0.9647694230079651,
+      "learning_rate": 4.708349097723268e-06,
+      "loss": 0.15,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5421257290991575,
+      "grad_norm": 1.0015144348144531,
+      "learning_rate": 4.708144084863541e-06,
+      "loss": 0.1497,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5422877511341543,
+      "grad_norm": 0.9959015846252441,
+      "learning_rate": 4.70793900443995e-06,
+      "loss": 0.1338,
+      "step": 3347
+    },
+    {
+      "epoch": 0.542449773169151,
+      "grad_norm": 1.017600655555725,
+      "learning_rate": 4.707733856458767e-06,
+      "loss": 0.1643,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5426117952041478,
+      "grad_norm": 1.165099859237671,
+      "learning_rate": 4.707528640926271e-06,
+      "loss": 0.1774,
+      "step": 3349
+    },
+    {
+      "epoch": 0.5427738172391445,
+      "grad_norm": 1.0708073377609253,
+      "learning_rate": 4.707323357848741e-06,
+      "loss": 0.174,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5429358392741412,
+      "grad_norm": 0.931759238243103,
+      "learning_rate": 4.707118007232457e-06,
+      "loss": 0.142,
+      "step": 3351
+    },
+    {
+      "epoch": 0.5430978613091381,
+      "grad_norm": 1.1331273317337036,
+      "learning_rate": 4.706912589083704e-06,
+      "loss": 0.171,
+      "step": 3352
+    },
+    {
+      "epoch": 0.5432598833441348,
+      "grad_norm": 0.9179681539535522,
+      "learning_rate": 4.706707103408767e-06,
+      "loss": 0.1463,
+      "step": 3353
+    },
+    {
+      "epoch": 0.5434219053791316,
+      "grad_norm": 0.9513891935348511,
+      "learning_rate": 4.706501550213932e-06,
+      "loss": 0.1706,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5435839274141283,
+      "grad_norm": 0.9615473747253418,
+      "learning_rate": 4.706295929505489e-06,
+      "loss": 0.1522,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5437459494491251,
+      "grad_norm": 0.9769582152366638,
+      "learning_rate": 4.70609024128973e-06,
+      "loss": 0.1464,
+      "step": 3356
+    },
+    {
+      "epoch": 0.5439079714841218,
+      "grad_norm": 0.9356233477592468,
+      "learning_rate": 4.705884485572948e-06,
+      "loss": 0.1442,
+      "step": 3357
+    },
+    {
+      "epoch": 0.5440699935191186,
+      "grad_norm": 1.0967140197753906,
+      "learning_rate": 4.7056786623614395e-06,
+      "loss": 0.1843,
+      "step": 3358
+    },
+    {
+      "epoch": 0.5442320155541154,
+      "grad_norm": 1.0115594863891602,
+      "learning_rate": 4.705472771661501e-06,
+      "loss": 0.1405,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5443940375891121,
+      "grad_norm": 1.0418258905410767,
+      "learning_rate": 4.705266813479434e-06,
+      "loss": 0.1626,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5445560596241089,
+      "grad_norm": 0.8964266180992126,
+      "learning_rate": 4.7050607878215375e-06,
+      "loss": 0.1398,
+      "step": 3361
+    },
+    {
+      "epoch": 0.5447180816591056,
+      "grad_norm": 0.9070065021514893,
+      "learning_rate": 4.704854694694117e-06,
+      "loss": 0.137,
+      "step": 3362
+    },
+    {
+      "epoch": 0.5448801036941024,
+      "grad_norm": 0.9841972589492798,
+      "learning_rate": 4.704648534103479e-06,
+      "loss": 0.1477,
+      "step": 3363
+    },
+    {
+      "epoch": 0.5450421257290992,
+      "grad_norm": 0.9104343056678772,
+      "learning_rate": 4.704442306055932e-06,
+      "loss": 0.1468,
+      "step": 3364
+    },
+    {
+      "epoch": 0.5452041477640959,
+      "grad_norm": 0.8985049724578857,
+      "learning_rate": 4.704236010557784e-06,
+      "loss": 0.1294,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5453661697990927,
+      "grad_norm": 0.9941574335098267,
+      "learning_rate": 4.704029647615348e-06,
+      "loss": 0.1539,
+      "step": 3366
+    },
+    {
+      "epoch": 0.5455281918340894,
+      "grad_norm": 1.0449243783950806,
+      "learning_rate": 4.7038232172349394e-06,
+      "loss": 0.1488,
+      "step": 3367
+    },
+    {
+      "epoch": 0.5456902138690862,
+      "grad_norm": 1.0465726852416992,
+      "learning_rate": 4.703616719422873e-06,
+      "loss": 0.1555,
+      "step": 3368
+    },
+    {
+      "epoch": 0.5458522359040829,
+      "grad_norm": 1.0622889995574951,
+      "learning_rate": 4.703410154185467e-06,
+      "loss": 0.1572,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5460142579390798,
+      "grad_norm": 0.9786834716796875,
+      "learning_rate": 4.703203521529044e-06,
+      "loss": 0.1558,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5461762799740765,
+      "grad_norm": 1.072187066078186,
+      "learning_rate": 4.702996821459923e-06,
+      "loss": 0.1553,
+      "step": 3371
+    },
+    {
+      "epoch": 0.5463383020090732,
+      "grad_norm": 0.9539135098457336,
+      "learning_rate": 4.702790053984432e-06,
+      "loss": 0.1469,
+      "step": 3372
+    },
+    {
+      "epoch": 0.54650032404407,
+      "grad_norm": 0.9775049686431885,
+      "learning_rate": 4.702583219108895e-06,
+      "loss": 0.154,
+      "step": 3373
+    },
+    {
+      "epoch": 0.5466623460790667,
+      "grad_norm": 0.8917540907859802,
+      "learning_rate": 4.702376316839642e-06,
+      "loss": 0.1335,
+      "step": 3374
+    },
+    {
+      "epoch": 0.5468243681140635,
+      "grad_norm": 0.9353879690170288,
+      "learning_rate": 4.7021693471830035e-06,
+      "loss": 0.1562,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5469863901490603,
+      "grad_norm": 0.9763373732566833,
+      "learning_rate": 4.701962310145312e-06,
+      "loss": 0.1546,
+      "step": 3376
+    },
+    {
+      "epoch": 0.5471484121840571,
+      "grad_norm": 0.9837629795074463,
+      "learning_rate": 4.701755205732902e-06,
+      "loss": 0.1645,
+      "step": 3377
+    },
+    {
+      "epoch": 0.5473104342190538,
+      "grad_norm": 0.8612603545188904,
+      "learning_rate": 4.7015480339521115e-06,
+      "loss": 0.1393,
+      "step": 3378
+    },
+    {
+      "epoch": 0.5474724562540505,
+      "grad_norm": 0.8714431524276733,
+      "learning_rate": 4.701340794809278e-06,
+      "loss": 0.1376,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5476344782890473,
+      "grad_norm": 0.9637095332145691,
+      "learning_rate": 4.701133488310744e-06,
+      "loss": 0.1527,
+      "step": 3380
+    },
+    {
+      "epoch": 0.547796500324044,
+      "grad_norm": 1.0376760959625244,
+      "learning_rate": 4.700926114462852e-06,
+      "loss": 0.1709,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5479585223590409,
+      "grad_norm": 0.9160656332969666,
+      "learning_rate": 4.700718673271947e-06,
+      "loss": 0.1439,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5481205443940376,
+      "grad_norm": 0.8436545729637146,
+      "learning_rate": 4.700511164744376e-06,
+      "loss": 0.1203,
+      "step": 3383
+    },
+    {
+      "epoch": 0.5482825664290344,
+      "grad_norm": 0.9517103433609009,
+      "learning_rate": 4.700303588886489e-06,
+      "loss": 0.1516,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5484445884640311,
+      "grad_norm": 0.9557342529296875,
+      "learning_rate": 4.700095945704636e-06,
+      "loss": 0.1502,
+      "step": 3385
+    },
+    {
+      "epoch": 0.5486066104990278,
+      "grad_norm": 0.9375651478767395,
+      "learning_rate": 4.699888235205172e-06,
+      "loss": 0.1455,
+      "step": 3386
+    },
+    {
+      "epoch": 0.5487686325340246,
+      "grad_norm": 0.9998489022254944,
+      "learning_rate": 4.699680457394451e-06,
+      "loss": 0.1429,
+      "step": 3387
+    },
+    {
+      "epoch": 0.5489306545690213,
+      "grad_norm": 0.9671725630760193,
+      "learning_rate": 4.699472612278831e-06,
+      "loss": 0.1548,
+      "step": 3388
+    },
+    {
+      "epoch": 0.5490926766040182,
+      "grad_norm": 1.063137412071228,
+      "learning_rate": 4.699264699864672e-06,
+      "loss": 0.1602,
+      "step": 3389
+    },
+    {
+      "epoch": 0.5492546986390149,
+      "grad_norm": 1.0313401222229004,
+      "learning_rate": 4.699056720158336e-06,
+      "loss": 0.1663,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5494167206740117,
+      "grad_norm": 1.0677765607833862,
+      "learning_rate": 4.698848673166185e-06,
+      "loss": 0.1436,
+      "step": 3391
+    },
+    {
+      "epoch": 0.5495787427090084,
+      "grad_norm": 1.0899100303649902,
+      "learning_rate": 4.698640558894586e-06,
+      "loss": 0.1608,
+      "step": 3392
+    },
+    {
+      "epoch": 0.5497407647440052,
+      "grad_norm": 0.9777064323425293,
+      "learning_rate": 4.6984323773499066e-06,
+      "loss": 0.1537,
+      "step": 3393
+    },
+    {
+      "epoch": 0.549902786779002,
+      "grad_norm": 1.0204713344573975,
+      "learning_rate": 4.698224128538517e-06,
+      "loss": 0.1535,
+      "step": 3394
+    },
+    {
+      "epoch": 0.5500648088139987,
+      "grad_norm": 0.9061859846115112,
+      "learning_rate": 4.698015812466787e-06,
+      "loss": 0.1429,
+      "step": 3395
+    },
+    {
+      "epoch": 0.5502268308489955,
+      "grad_norm": 0.9810669422149658,
+      "learning_rate": 4.6978074291410936e-06,
+      "loss": 0.1544,
+      "step": 3396
+    },
+    {
+      "epoch": 0.5503888528839922,
+      "grad_norm": 0.9356314539909363,
+      "learning_rate": 4.697598978567811e-06,
+      "loss": 0.1427,
+      "step": 3397
+    },
+    {
+      "epoch": 0.550550874918989,
+      "grad_norm": 1.0116984844207764,
+      "learning_rate": 4.697390460753318e-06,
+      "loss": 0.148,
+      "step": 3398
+    },
+    {
+      "epoch": 0.5507128969539857,
+      "grad_norm": 0.9845331907272339,
+      "learning_rate": 4.697181875703995e-06,
+      "loss": 0.1722,
+      "step": 3399
+    },
+    {
+      "epoch": 0.5508749189889826,
+      "grad_norm": 1.221174716949463,
+      "learning_rate": 4.696973223426224e-06,
+      "loss": 0.1573,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5510369410239793,
+      "grad_norm": 0.8805899620056152,
+      "learning_rate": 4.696764503926387e-06,
+      "loss": 0.137,
+      "step": 3401
+    },
+    {
+      "epoch": 0.551198963058976,
+      "grad_norm": 0.8739351630210876,
+      "learning_rate": 4.696555717210873e-06,
+      "loss": 0.125,
+      "step": 3402
+    },
+    {
+      "epoch": 0.5513609850939728,
+      "grad_norm": 0.8733137249946594,
+      "learning_rate": 4.696346863286071e-06,
+      "loss": 0.1306,
+      "step": 3403
+    },
+    {
+      "epoch": 0.5515230071289695,
+      "grad_norm": 0.967108428478241,
+      "learning_rate": 4.6961379421583685e-06,
+      "loss": 0.1513,
+      "step": 3404
+    },
+    {
+      "epoch": 0.5516850291639663,
+      "grad_norm": 0.9212470054626465,
+      "learning_rate": 4.69592895383416e-06,
+      "loss": 0.1431,
+      "step": 3405
+    },
+    {
+      "epoch": 0.551847051198963,
+      "grad_norm": 0.9542239904403687,
+      "learning_rate": 4.695719898319839e-06,
+      "loss": 0.149,
+      "step": 3406
+    },
+    {
+      "epoch": 0.5520090732339599,
+      "grad_norm": 0.8492709994316101,
+      "learning_rate": 4.6955107756218035e-06,
+      "loss": 0.1331,
+      "step": 3407
+    },
+    {
+      "epoch": 0.5521710952689566,
+      "grad_norm": 0.9666154384613037,
+      "learning_rate": 4.695301585746451e-06,
+      "loss": 0.1457,
+      "step": 3408
+    },
+    {
+      "epoch": 0.5523331173039533,
+      "grad_norm": 0.9427182674407959,
+      "learning_rate": 4.695092328700182e-06,
+      "loss": 0.1505,
+      "step": 3409
+    },
+    {
+      "epoch": 0.5524951393389501,
+      "grad_norm": 0.9889812469482422,
+      "learning_rate": 4.6948830044894016e-06,
+      "loss": 0.1627,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5526571613739468,
+      "grad_norm": 1.1829040050506592,
+      "learning_rate": 4.694673613120511e-06,
+      "loss": 0.1849,
+      "step": 3411
+    },
+    {
+      "epoch": 0.5528191834089436,
+      "grad_norm": 0.8475874662399292,
+      "learning_rate": 4.6944641545999194e-06,
+      "loss": 0.133,
+      "step": 3412
+    },
+    {
+      "epoch": 0.5529812054439404,
+      "grad_norm": 0.9569193720817566,
+      "learning_rate": 4.694254628934035e-06,
+      "loss": 0.1498,
+      "step": 3413
+    },
+    {
+      "epoch": 0.5531432274789372,
+      "grad_norm": 1.0036123991012573,
+      "learning_rate": 4.694045036129269e-06,
+      "loss": 0.1409,
+      "step": 3414
+    },
+    {
+      "epoch": 0.5533052495139339,
+      "grad_norm": 0.9770514369010925,
+      "learning_rate": 4.6938353761920345e-06,
+      "loss": 0.1484,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5534672715489306,
+      "grad_norm": 0.9479352235794067,
+      "learning_rate": 4.693625649128746e-06,
+      "loss": 0.1481,
+      "step": 3416
+    },
+    {
+      "epoch": 0.5536292935839274,
+      "grad_norm": 1.000531792640686,
+      "learning_rate": 4.693415854945822e-06,
+      "loss": 0.145,
+      "step": 3417
+    },
+    {
+      "epoch": 0.5537913156189241,
+      "grad_norm": 1.1228997707366943,
+      "learning_rate": 4.69320599364968e-06,
+      "loss": 0.1561,
+      "step": 3418
+    },
+    {
+      "epoch": 0.553953337653921,
+      "grad_norm": 0.9270069003105164,
+      "learning_rate": 4.692996065246742e-06,
+      "loss": 0.1384,
+      "step": 3419
+    },
+    {
+      "epoch": 0.5541153596889177,
+      "grad_norm": 1.013025164604187,
+      "learning_rate": 4.692786069743432e-06,
+      "loss": 0.1832,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5542773817239145,
+      "grad_norm": 0.9863829612731934,
+      "learning_rate": 4.692576007146175e-06,
+      "loss": 0.149,
+      "step": 3421
+    },
+    {
+      "epoch": 0.5544394037589112,
+      "grad_norm": 0.9370028376579285,
+      "learning_rate": 4.692365877461397e-06,
+      "loss": 0.1392,
+      "step": 3422
+    },
+    {
+      "epoch": 0.5546014257939079,
+      "grad_norm": 1.1558516025543213,
+      "learning_rate": 4.692155680695529e-06,
+      "loss": 0.1732,
+      "step": 3423
+    },
+    {
+      "epoch": 0.5547634478289047,
+      "grad_norm": 0.8533956408500671,
+      "learning_rate": 4.691945416855002e-06,
+      "loss": 0.1409,
+      "step": 3424
+    },
+    {
+      "epoch": 0.5549254698639015,
+      "grad_norm": 1.1222232580184937,
+      "learning_rate": 4.69173508594625e-06,
+      "loss": 0.1489,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5550874918988983,
+      "grad_norm": 1.032179594039917,
+      "learning_rate": 4.6915246879757084e-06,
+      "loss": 0.162,
+      "step": 3426
+    },
+    {
+      "epoch": 0.555249513933895,
+      "grad_norm": 1.0500115156173706,
+      "learning_rate": 4.691314222949814e-06,
+      "loss": 0.1491,
+      "step": 3427
+    },
+    {
+      "epoch": 0.5554115359688918,
+      "grad_norm": 0.944101095199585,
+      "learning_rate": 4.691103690875007e-06,
+      "loss": 0.1565,
+      "step": 3428
+    },
+    {
+      "epoch": 0.5555735580038885,
+      "grad_norm": 0.9684193134307861,
+      "learning_rate": 4.690893091757731e-06,
+      "loss": 0.1523,
+      "step": 3429
+    },
+    {
+      "epoch": 0.5557355800388852,
+      "grad_norm": 0.9088751077651978,
+      "learning_rate": 4.690682425604427e-06,
+      "loss": 0.1326,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5558976020738821,
+      "grad_norm": 0.8529621958732605,
+      "learning_rate": 4.6904716924215425e-06,
+      "loss": 0.1359,
+      "step": 3431
+    },
+    {
+      "epoch": 0.5560596241088788,
+      "grad_norm": 1.1265101432800293,
+      "learning_rate": 4.690260892215525e-06,
+      "loss": 0.1779,
+      "step": 3432
+    },
+    {
+      "epoch": 0.5562216461438756,
+      "grad_norm": 0.9452961087226868,
+      "learning_rate": 4.690050024992825e-06,
+      "loss": 0.1409,
+      "step": 3433
+    },
+    {
+      "epoch": 0.5563836681788723,
+      "grad_norm": 0.9862555265426636,
+      "learning_rate": 4.689839090759893e-06,
+      "loss": 0.1692,
+      "step": 3434
+    },
+    {
+      "epoch": 0.5565456902138691,
+      "grad_norm": 0.9753914475440979,
+      "learning_rate": 4.689628089523185e-06,
+      "loss": 0.1589,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5567077122488658,
+      "grad_norm": 0.8871638178825378,
+      "learning_rate": 4.689417021289157e-06,
+      "loss": 0.1438,
+      "step": 3436
+    },
+    {
+      "epoch": 0.5568697342838627,
+      "grad_norm": 0.890839159488678,
+      "learning_rate": 4.689205886064265e-06,
+      "loss": 0.1378,
+      "step": 3437
+    },
+    {
+      "epoch": 0.5570317563188594,
+      "grad_norm": 1.1297328472137451,
+      "learning_rate": 4.68899468385497e-06,
+      "loss": 0.1829,
+      "step": 3438
+    },
+    {
+      "epoch": 0.5571937783538561,
+      "grad_norm": 1.0578821897506714,
+      "learning_rate": 4.6887834146677365e-06,
+      "loss": 0.1538,
+      "step": 3439
+    },
+    {
+      "epoch": 0.5573558003888529,
+      "grad_norm": 0.9971956014633179,
+      "learning_rate": 4.688572078509027e-06,
+      "loss": 0.1608,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5575178224238496,
+      "grad_norm": 1.0291696786880493,
+      "learning_rate": 4.688360675385308e-06,
+      "loss": 0.1569,
+      "step": 3441
+    },
+    {
+      "epoch": 0.5576798444588464,
+      "grad_norm": 0.9045324921607971,
+      "learning_rate": 4.688149205303048e-06,
+      "loss": 0.1374,
+      "step": 3442
+    },
+    {
+      "epoch": 0.5578418664938432,
+      "grad_norm": 1.0575282573699951,
+      "learning_rate": 4.687937668268718e-06,
+      "loss": 0.1511,
+      "step": 3443
+    },
+    {
+      "epoch": 0.55800388852884,
+      "grad_norm": 0.9568415880203247,
+      "learning_rate": 4.687726064288789e-06,
+      "loss": 0.1441,
+      "step": 3444
+    },
+    {
+      "epoch": 0.5581659105638367,
+      "grad_norm": 0.9325804710388184,
+      "learning_rate": 4.687514393369738e-06,
+      "loss": 0.1417,
+      "step": 3445
+    },
+    {
+      "epoch": 0.5583279325988334,
+      "grad_norm": 1.066925048828125,
+      "learning_rate": 4.6873026555180386e-06,
+      "loss": 0.1646,
+      "step": 3446
+    },
+    {
+      "epoch": 0.5584899546338302,
+      "grad_norm": 0.9176453948020935,
+      "learning_rate": 4.687090850740172e-06,
+      "loss": 0.1265,
+      "step": 3447
+    },
+    {
+      "epoch": 0.5586519766688269,
+      "grad_norm": 0.9668508768081665,
+      "learning_rate": 4.6868789790426185e-06,
+      "loss": 0.1433,
+      "step": 3448
+    },
+    {
+      "epoch": 0.5588139987038238,
+      "grad_norm": 0.8948216438293457,
+      "learning_rate": 4.68666704043186e-06,
+      "loss": 0.1344,
+      "step": 3449
+    },
+    {
+      "epoch": 0.5589760207388205,
+      "grad_norm": 0.9884295463562012,
+      "learning_rate": 4.6864550349143815e-06,
+      "loss": 0.1584,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5591380427738173,
+      "grad_norm": 0.8515805006027222,
+      "learning_rate": 4.6862429624966695e-06,
+      "loss": 0.144,
+      "step": 3451
+    },
+    {
+      "epoch": 0.559300064808814,
+      "grad_norm": 0.9054228663444519,
+      "learning_rate": 4.686030823185215e-06,
+      "loss": 0.1445,
+      "step": 3452
+    },
+    {
+      "epoch": 0.5594620868438107,
+      "grad_norm": 1.0499194860458374,
+      "learning_rate": 4.685818616986506e-06,
+      "loss": 0.1536,
+      "step": 3453
+    },
+    {
+      "epoch": 0.5596241088788075,
+      "grad_norm": 0.8390993475914001,
+      "learning_rate": 4.685606343907038e-06,
+      "loss": 0.1301,
+      "step": 3454
+    },
+    {
+      "epoch": 0.5597861309138042,
+      "grad_norm": 0.996111273765564,
+      "learning_rate": 4.685394003953304e-06,
+      "loss": 0.1515,
+      "step": 3455
+    },
+    {
+      "epoch": 0.5599481529488011,
+      "grad_norm": 1.1226462125778198,
+      "learning_rate": 4.685181597131802e-06,
+      "loss": 0.1795,
+      "step": 3456
+    },
+    {
+      "epoch": 0.5601101749837978,
+      "grad_norm": 1.1062334775924683,
+      "learning_rate": 4.684969123449032e-06,
+      "loss": 0.1587,
+      "step": 3457
+    },
+    {
+      "epoch": 0.5602721970187946,
+      "grad_norm": 0.9502385258674622,
+      "learning_rate": 4.684756582911494e-06,
+      "loss": 0.1506,
+      "step": 3458
+    },
+    {
+      "epoch": 0.5604342190537913,
+      "grad_norm": 0.9273852109909058,
+      "learning_rate": 4.684543975525691e-06,
+      "loss": 0.1421,
+      "step": 3459
+    },
+    {
+      "epoch": 0.560596241088788,
+      "grad_norm": 1.0246306657791138,
+      "learning_rate": 4.6843313012981295e-06,
+      "loss": 0.1696,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5607582631237849,
+      "grad_norm": 0.9162903428077698,
+      "learning_rate": 4.684118560235315e-06,
+      "loss": 0.1468,
+      "step": 3461
+    },
+    {
+      "epoch": 0.5609202851587816,
+      "grad_norm": 1.003145456314087,
+      "learning_rate": 4.6839057523437606e-06,
+      "loss": 0.1484,
+      "step": 3462
+    },
+    {
+      "epoch": 0.5610823071937784,
+      "grad_norm": 0.9194111824035645,
+      "learning_rate": 4.683692877629973e-06,
+      "loss": 0.1415,
+      "step": 3463
+    },
+    {
+      "epoch": 0.5612443292287751,
+      "grad_norm": 1.1066536903381348,
+      "learning_rate": 4.683479936100468e-06,
+      "loss": 0.1734,
+      "step": 3464
+    },
+    {
+      "epoch": 0.5614063512637719,
+      "grad_norm": 0.9368934631347656,
+      "learning_rate": 4.683266927761762e-06,
+      "loss": 0.1652,
+      "step": 3465
+    },
+    {
+      "epoch": 0.5615683732987686,
+      "grad_norm": 0.9979063868522644,
+      "learning_rate": 4.68305385262037e-06,
+      "loss": 0.1536,
+      "step": 3466
+    },
+    {
+      "epoch": 0.5617303953337653,
+      "grad_norm": 0.9077597856521606,
+      "learning_rate": 4.6828407106828135e-06,
+      "loss": 0.1353,
+      "step": 3467
+    },
+    {
+      "epoch": 0.5618924173687622,
+      "grad_norm": 1.0348340272903442,
+      "learning_rate": 4.682627501955614e-06,
+      "loss": 0.1543,
+      "step": 3468
+    },
+    {
+      "epoch": 0.5620544394037589,
+      "grad_norm": 0.8664389252662659,
+      "learning_rate": 4.6824142264452945e-06,
+      "loss": 0.1277,
+      "step": 3469
+    },
+    {
+      "epoch": 0.5622164614387557,
+      "grad_norm": 0.9749402403831482,
+      "learning_rate": 4.682200884158381e-06,
+      "loss": 0.1528,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5623784834737524,
+      "grad_norm": 0.9896659255027771,
+      "learning_rate": 4.6819874751014015e-06,
+      "loss": 0.1463,
+      "step": 3471
+    },
+    {
+      "epoch": 0.5625405055087492,
+      "grad_norm": 0.9059292674064636,
+      "learning_rate": 4.6817739992808855e-06,
+      "loss": 0.1469,
+      "step": 3472
+    },
+    {
+      "epoch": 0.562702527543746,
+      "grad_norm": 0.8704636693000793,
+      "learning_rate": 4.681560456703364e-06,
+      "loss": 0.1475,
+      "step": 3473
+    },
+    {
+      "epoch": 0.5628645495787427,
+      "grad_norm": 0.9798029065132141,
+      "learning_rate": 4.681346847375373e-06,
+      "loss": 0.1693,
+      "step": 3474
+    },
+    {
+      "epoch": 0.5630265716137395,
+      "grad_norm": 0.971039891242981,
+      "learning_rate": 4.681133171303447e-06,
+      "loss": 0.1549,
+      "step": 3475
+    },
+    {
+      "epoch": 0.5631885936487362,
+      "grad_norm": 0.9427266120910645,
+      "learning_rate": 4.6809194284941236e-06,
+      "loss": 0.157,
+      "step": 3476
+    },
+    {
+      "epoch": 0.563350615683733,
+      "grad_norm": 1.0199015140533447,
+      "learning_rate": 4.680705618953944e-06,
+      "loss": 0.1631,
+      "step": 3477
+    },
+    {
+      "epoch": 0.5635126377187297,
+      "grad_norm": 0.9573736786842346,
+      "learning_rate": 4.6804917426894495e-06,
+      "loss": 0.1644,
+      "step": 3478
+    },
+    {
+      "epoch": 0.5636746597537265,
+      "grad_norm": 0.992607831954956,
+      "learning_rate": 4.680277799707185e-06,
+      "loss": 0.1475,
+      "step": 3479
+    },
+    {
+      "epoch": 0.5638366817887233,
+      "grad_norm": 0.8607563376426697,
+      "learning_rate": 4.6800637900136944e-06,
+      "loss": 0.1461,
+      "step": 3480
+    },
+    {
+      "epoch": 0.56399870382372,
+      "grad_norm": 0.8701785206794739,
+      "learning_rate": 4.6798497136155286e-06,
+      "loss": 0.1307,
+      "step": 3481
+    },
+    {
+      "epoch": 0.5641607258587168,
+      "grad_norm": 0.8611844778060913,
+      "learning_rate": 4.679635570519236e-06,
+      "loss": 0.1326,
+      "step": 3482
+    },
+    {
+      "epoch": 0.5643227478937135,
+      "grad_norm": 1.2296451330184937,
+      "learning_rate": 4.679421360731371e-06,
+      "loss": 0.1533,
+      "step": 3483
+    },
+    {
+      "epoch": 0.5644847699287103,
+      "grad_norm": 0.9519492983818054,
+      "learning_rate": 4.6792070842584855e-06,
+      "loss": 0.1453,
+      "step": 3484
+    },
+    {
+      "epoch": 0.564646791963707,
+      "grad_norm": 0.9066160321235657,
+      "learning_rate": 4.678992741107136e-06,
+      "loss": 0.1469,
+      "step": 3485
+    },
+    {
+      "epoch": 0.5648088139987039,
+      "grad_norm": 1.1821013689041138,
+      "learning_rate": 4.678778331283883e-06,
+      "loss": 0.1911,
+      "step": 3486
+    },
+    {
+      "epoch": 0.5649708360337006,
+      "grad_norm": 0.9712387919425964,
+      "learning_rate": 4.678563854795285e-06,
+      "loss": 0.166,
+      "step": 3487
+    },
+    {
+      "epoch": 0.5651328580686974,
+      "grad_norm": 0.964177668094635,
+      "learning_rate": 4.678349311647905e-06,
+      "loss": 0.1541,
+      "step": 3488
+    },
+    {
+      "epoch": 0.5652948801036941,
+      "grad_norm": 0.864924967288971,
+      "learning_rate": 4.678134701848308e-06,
+      "loss": 0.1395,
+      "step": 3489
+    },
+    {
+      "epoch": 0.5654569021386908,
+      "grad_norm": 0.9088456630706787,
+      "learning_rate": 4.67792002540306e-06,
+      "loss": 0.1444,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5656189241736876,
+      "grad_norm": 0.9394357204437256,
+      "learning_rate": 4.67770528231873e-06,
+      "loss": 0.1482,
+      "step": 3491
+    },
+    {
+      "epoch": 0.5657809462086844,
+      "grad_norm": 0.8865934014320374,
+      "learning_rate": 4.677490472601888e-06,
+      "loss": 0.1456,
+      "step": 3492
+    },
+    {
+      "epoch": 0.5659429682436812,
+      "grad_norm": 1.0510127544403076,
+      "learning_rate": 4.677275596259107e-06,
+      "loss": 0.1593,
+      "step": 3493
+    },
+    {
+      "epoch": 0.5661049902786779,
+      "grad_norm": 0.8334269523620605,
+      "learning_rate": 4.677060653296961e-06,
+      "loss": 0.123,
+      "step": 3494
+    },
+    {
+      "epoch": 0.5662670123136747,
+      "grad_norm": 1.0426408052444458,
+      "learning_rate": 4.676845643722028e-06,
+      "loss": 0.1603,
+      "step": 3495
+    },
+    {
+      "epoch": 0.5664290343486714,
+      "grad_norm": 0.9913350343704224,
+      "learning_rate": 4.676630567540886e-06,
+      "loss": 0.1548,
+      "step": 3496
+    },
+    {
+      "epoch": 0.5665910563836681,
+      "grad_norm": 0.9299079775810242,
+      "learning_rate": 4.676415424760115e-06,
+      "loss": 0.1387,
+      "step": 3497
+    },
+    {
+      "epoch": 0.566753078418665,
+      "grad_norm": 0.8800061345100403,
+      "learning_rate": 4.6762002153863e-06,
+      "loss": 0.1399,
+      "step": 3498
+    },
+    {
+      "epoch": 0.5669151004536617,
+      "grad_norm": 0.827441930770874,
+      "learning_rate": 4.675984939426026e-06,
+      "loss": 0.1308,
+      "step": 3499
+    },
+    {
+      "epoch": 0.5670771224886585,
+      "grad_norm": 0.914867639541626,
+      "learning_rate": 4.675769596885877e-06,
+      "loss": 0.1541,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5672391445236552,
+      "grad_norm": 0.7989079356193542,
+      "learning_rate": 4.675554187772444e-06,
+      "loss": 0.1272,
+      "step": 3501
+    },
+    {
+      "epoch": 0.567401166558652,
+      "grad_norm": 1.044179916381836,
+      "learning_rate": 4.675338712092316e-06,
+      "loss": 0.1792,
+      "step": 3502
+    },
+    {
+      "epoch": 0.5675631885936487,
+      "grad_norm": 1.00692617893219,
+      "learning_rate": 4.67512316985209e-06,
+      "loss": 0.1667,
+      "step": 3503
+    },
+    {
+      "epoch": 0.5677252106286454,
+      "grad_norm": 0.9069098830223083,
+      "learning_rate": 4.674907561058358e-06,
+      "loss": 0.1426,
+      "step": 3504
+    },
+    {
+      "epoch": 0.5678872326636423,
+      "grad_norm": 0.9729463458061218,
+      "learning_rate": 4.674691885717717e-06,
+      "loss": 0.1551,
+      "step": 3505
+    },
+    {
+      "epoch": 0.568049254698639,
+      "grad_norm": 0.8880230188369751,
+      "learning_rate": 4.674476143836768e-06,
+      "loss": 0.1346,
+      "step": 3506
+    },
+    {
+      "epoch": 0.5682112767336358,
+      "grad_norm": 0.9867011904716492,
+      "learning_rate": 4.6742603354221105e-06,
+      "loss": 0.1409,
+      "step": 3507
+    },
+    {
+      "epoch": 0.5683732987686325,
+      "grad_norm": 0.9959049820899963,
+      "learning_rate": 4.674044460480348e-06,
+      "loss": 0.1525,
+      "step": 3508
+    },
+    {
+      "epoch": 0.5685353208036293,
+      "grad_norm": 0.9901296496391296,
+      "learning_rate": 4.6738285190180865e-06,
+      "loss": 0.145,
+      "step": 3509
+    },
+    {
+      "epoch": 0.568697342838626,
+      "grad_norm": 0.9741671085357666,
+      "learning_rate": 4.673612511041933e-06,
+      "loss": 0.157,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5688593648736228,
+      "grad_norm": 0.9612037539482117,
+      "learning_rate": 4.673396436558497e-06,
+      "loss": 0.1535,
+      "step": 3511
+    },
+    {
+      "epoch": 0.5690213869086196,
+      "grad_norm": 1.0780646800994873,
+      "learning_rate": 4.673180295574389e-06,
+      "loss": 0.1713,
+      "step": 3512
+    },
+    {
+      "epoch": 0.5691834089436163,
+      "grad_norm": 1.152421236038208,
+      "learning_rate": 4.672964088096223e-06,
+      "loss": 0.1358,
+      "step": 3513
+    },
+    {
+      "epoch": 0.5693454309786131,
+      "grad_norm": 0.9687821269035339,
+      "learning_rate": 4.672747814130615e-06,
+      "loss": 0.157,
+      "step": 3514
+    },
+    {
+      "epoch": 0.5695074530136098,
+      "grad_norm": 1.0094846487045288,
+      "learning_rate": 4.6725314736841806e-06,
+      "loss": 0.1582,
+      "step": 3515
+    },
+    {
+      "epoch": 0.5696694750486067,
+      "grad_norm": 1.0530569553375244,
+      "learning_rate": 4.672315066763542e-06,
+      "loss": 0.1764,
+      "step": 3516
+    },
+    {
+      "epoch": 0.5698314970836034,
+      "grad_norm": 0.8865777254104614,
+      "learning_rate": 4.672098593375319e-06,
+      "loss": 0.1384,
+      "step": 3517
+    },
+    {
+      "epoch": 0.5699935191186001,
+      "grad_norm": 0.9916114211082458,
+      "learning_rate": 4.671882053526135e-06,
+      "loss": 0.1531,
+      "step": 3518
+    },
+    {
+      "epoch": 0.5701555411535969,
+      "grad_norm": 0.9261521100997925,
+      "learning_rate": 4.671665447222615e-06,
+      "loss": 0.1444,
+      "step": 3519
+    },
+    {
+      "epoch": 0.5703175631885936,
+      "grad_norm": 0.8022376894950867,
+      "learning_rate": 4.671448774471389e-06,
+      "loss": 0.1337,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5704795852235904,
+      "grad_norm": 0.8965422511100769,
+      "learning_rate": 4.671232035279085e-06,
+      "loss": 0.136,
+      "step": 3521
+    },
+    {
+      "epoch": 0.5706416072585871,
+      "grad_norm": 0.9427691102027893,
+      "learning_rate": 4.671015229652335e-06,
+      "loss": 0.1415,
+      "step": 3522
+    },
+    {
+      "epoch": 0.570803629293584,
+      "grad_norm": 0.9803239107131958,
+      "learning_rate": 4.6707983575977724e-06,
+      "loss": 0.1568,
+      "step": 3523
+    },
+    {
+      "epoch": 0.5709656513285807,
+      "grad_norm": 1.0181081295013428,
+      "learning_rate": 4.670581419122034e-06,
+      "loss": 0.1559,
+      "step": 3524
+    },
+    {
+      "epoch": 0.5711276733635774,
+      "grad_norm": 0.9415149688720703,
+      "learning_rate": 4.670364414231756e-06,
+      "loss": 0.1375,
+      "step": 3525
+    },
+    {
+      "epoch": 0.5712896953985742,
+      "grad_norm": 0.9571231007575989,
+      "learning_rate": 4.67014734293358e-06,
+      "loss": 0.1489,
+      "step": 3526
+    },
+    {
+      "epoch": 0.5714517174335709,
+      "grad_norm": 0.9342589378356934,
+      "learning_rate": 4.669930205234146e-06,
+      "loss": 0.1547,
+      "step": 3527
+    },
+    {
+      "epoch": 0.5716137394685677,
+      "grad_norm": 1.1925525665283203,
+      "learning_rate": 4.6697130011401e-06,
+      "loss": 0.1662,
+      "step": 3528
+    },
+    {
+      "epoch": 0.5717757615035645,
+      "grad_norm": 0.8905578851699829,
+      "learning_rate": 4.669495730658086e-06,
+      "loss": 0.1552,
+      "step": 3529
+    },
+    {
+      "epoch": 0.5719377835385613,
+      "grad_norm": 0.839648425579071,
+      "learning_rate": 4.669278393794753e-06,
+      "loss": 0.1288,
+      "step": 3530
+    },
+    {
+      "epoch": 0.572099805573558,
+      "grad_norm": 0.937215268611908,
+      "learning_rate": 4.669060990556751e-06,
+      "loss": 0.1486,
+      "step": 3531
+    },
+    {
+      "epoch": 0.5722618276085548,
+      "grad_norm": 1.0918093919754028,
+      "learning_rate": 4.6688435209507305e-06,
+      "loss": 0.1545,
+      "step": 3532
+    },
+    {
+      "epoch": 0.5724238496435515,
+      "grad_norm": 1.0010573863983154,
+      "learning_rate": 4.668625984983347e-06,
+      "loss": 0.1635,
+      "step": 3533
+    },
+    {
+      "epoch": 0.5725858716785482,
+      "grad_norm": 0.9082236289978027,
+      "learning_rate": 4.668408382661257e-06,
+      "loss": 0.136,
+      "step": 3534
+    },
+    {
+      "epoch": 0.5727478937135451,
+      "grad_norm": 0.9376429319381714,
+      "learning_rate": 4.6681907139911185e-06,
+      "loss": 0.1462,
+      "step": 3535
+    },
+    {
+      "epoch": 0.5729099157485418,
+      "grad_norm": 0.9339603185653687,
+      "learning_rate": 4.66797297897959e-06,
+      "loss": 0.1397,
+      "step": 3536
+    },
+    {
+      "epoch": 0.5730719377835386,
+      "grad_norm": 0.9145956635475159,
+      "learning_rate": 4.667755177633335e-06,
+      "loss": 0.1423,
+      "step": 3537
+    },
+    {
+      "epoch": 0.5732339598185353,
+      "grad_norm": 0.9105459451675415,
+      "learning_rate": 4.667537309959018e-06,
+      "loss": 0.1487,
+      "step": 3538
+    },
+    {
+      "epoch": 0.5733959818535321,
+      "grad_norm": 0.9851331114768982,
+      "learning_rate": 4.667319375963304e-06,
+      "loss": 0.15,
+      "step": 3539
+    },
+    {
+      "epoch": 0.5735580038885288,
+      "grad_norm": 0.9845294952392578,
+      "learning_rate": 4.667101375652862e-06,
+      "loss": 0.1442,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5737200259235256,
+      "grad_norm": 0.8981902599334717,
+      "learning_rate": 4.666883309034362e-06,
+      "loss": 0.143,
+      "step": 3541
+    },
+    {
+      "epoch": 0.5738820479585224,
+      "grad_norm": 1.1543735265731812,
+      "learning_rate": 4.666665176114477e-06,
+      "loss": 0.1609,
+      "step": 3542
+    },
+    {
+      "epoch": 0.5740440699935191,
+      "grad_norm": 0.9345577359199524,
+      "learning_rate": 4.666446976899881e-06,
+      "loss": 0.1438,
+      "step": 3543
+    },
+    {
+      "epoch": 0.5742060920285159,
+      "grad_norm": 0.9611945152282715,
+      "learning_rate": 4.666228711397249e-06,
+      "loss": 0.1577,
+      "step": 3544
+    },
+    {
+      "epoch": 0.5743681140635126,
+      "grad_norm": 0.9625419974327087,
+      "learning_rate": 4.6660103796132615e-06,
+      "loss": 0.1464,
+      "step": 3545
+    },
+    {
+      "epoch": 0.5745301360985094,
+      "grad_norm": 0.9665799140930176,
+      "learning_rate": 4.665791981554598e-06,
+      "loss": 0.1501,
+      "step": 3546
+    },
+    {
+      "epoch": 0.5746921581335062,
+      "grad_norm": 1.2185211181640625,
+      "learning_rate": 4.665573517227942e-06,
+      "loss": 0.1475,
+      "step": 3547
+    },
+    {
+      "epoch": 0.5748541801685029,
+      "grad_norm": 0.938981294631958,
+      "learning_rate": 4.665354986639975e-06,
+      "loss": 0.1545,
+      "step": 3548
+    },
+    {
+      "epoch": 0.5750162022034997,
+      "grad_norm": 0.9329091310501099,
+      "learning_rate": 4.665136389797387e-06,
+      "loss": 0.1437,
+      "step": 3549
+    },
+    {
+      "epoch": 0.5751782242384964,
+      "grad_norm": 1.1116046905517578,
+      "learning_rate": 4.664917726706864e-06,
+      "loss": 0.1516,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5753402462734932,
+      "grad_norm": 0.8297967910766602,
+      "learning_rate": 4.664698997375098e-06,
+      "loss": 0.1187,
+      "step": 3551
+    },
+    {
+      "epoch": 0.5755022683084899,
+      "grad_norm": 0.9134968519210815,
+      "learning_rate": 4.6644802018087806e-06,
+      "loss": 0.1374,
+      "step": 3552
+    },
+    {
+      "epoch": 0.5756642903434868,
+      "grad_norm": 1.091581106185913,
+      "learning_rate": 4.664261340014608e-06,
+      "loss": 0.1626,
+      "step": 3553
+    },
+    {
+      "epoch": 0.5758263123784835,
+      "grad_norm": 0.9347933530807495,
+      "learning_rate": 4.664042411999276e-06,
+      "loss": 0.139,
+      "step": 3554
+    },
+    {
+      "epoch": 0.5759883344134802,
+      "grad_norm": 1.0377494096755981,
+      "learning_rate": 4.663823417769482e-06,
+      "loss": 0.1703,
+      "step": 3555
+    },
+    {
+      "epoch": 0.576150356448477,
+      "grad_norm": 0.9047291874885559,
+      "learning_rate": 4.663604357331928e-06,
+      "loss": 0.14,
+      "step": 3556
+    },
+    {
+      "epoch": 0.5763123784834737,
+      "grad_norm": 0.9950608611106873,
+      "learning_rate": 4.663385230693316e-06,
+      "loss": 0.1631,
+      "step": 3557
+    },
+    {
+      "epoch": 0.5764744005184705,
+      "grad_norm": 0.8619599938392639,
+      "learning_rate": 4.6631660378603526e-06,
+      "loss": 0.1388,
+      "step": 3558
+    },
+    {
+      "epoch": 0.5766364225534673,
+      "grad_norm": 0.9685336947441101,
+      "learning_rate": 4.662946778839742e-06,
+      "loss": 0.1649,
+      "step": 3559
+    },
+    {
+      "epoch": 0.5767984445884641,
+      "grad_norm": 0.9717000722885132,
+      "learning_rate": 4.662727453638195e-06,
+      "loss": 0.1489,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5769604666234608,
+      "grad_norm": 0.8928928375244141,
+      "learning_rate": 4.662508062262421e-06,
+      "loss": 0.1475,
+      "step": 3561
+    },
+    {
+      "epoch": 0.5771224886584575,
+      "grad_norm": 0.8069142699241638,
+      "learning_rate": 4.662288604719134e-06,
+      "loss": 0.125,
+      "step": 3562
+    },
+    {
+      "epoch": 0.5772845106934543,
+      "grad_norm": 0.8762704133987427,
+      "learning_rate": 4.662069081015047e-06,
+      "loss": 0.1369,
+      "step": 3563
+    },
+    {
+      "epoch": 0.577446532728451,
+      "grad_norm": 0.9283084869384766,
+      "learning_rate": 4.66184949115688e-06,
+      "loss": 0.1376,
+      "step": 3564
+    },
+    {
+      "epoch": 0.5776085547634479,
+      "grad_norm": 1.0266906023025513,
+      "learning_rate": 4.66162983515135e-06,
+      "loss": 0.1532,
+      "step": 3565
+    },
+    {
+      "epoch": 0.5777705767984446,
+      "grad_norm": 1.1265392303466797,
+      "learning_rate": 4.661410113005177e-06,
+      "loss": 0.1586,
+      "step": 3566
+    },
+    {
+      "epoch": 0.5779325988334414,
+      "grad_norm": 0.9027408361434937,
+      "learning_rate": 4.661190324725085e-06,
+      "loss": 0.1427,
+      "step": 3567
+    },
+    {
+      "epoch": 0.5780946208684381,
+      "grad_norm": 0.9591796398162842,
+      "learning_rate": 4.6609704703178e-06,
+      "loss": 0.1442,
+      "step": 3568
+    },
+    {
+      "epoch": 0.5782566429034348,
+      "grad_norm": 1.0705912113189697,
+      "learning_rate": 4.6607505497900475e-06,
+      "loss": 0.156,
+      "step": 3569
+    },
+    {
+      "epoch": 0.5784186649384316,
+      "grad_norm": 1.016605257987976,
+      "learning_rate": 4.660530563148557e-06,
+      "loss": 0.1494,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5785806869734283,
+      "grad_norm": 0.9418259263038635,
+      "learning_rate": 4.66031051040006e-06,
+      "loss": 0.144,
+      "step": 3571
+    },
+    {
+      "epoch": 0.5787427090084252,
+      "grad_norm": 1.0305404663085938,
+      "learning_rate": 4.66009039155129e-06,
+      "loss": 0.1371,
+      "step": 3572
+    },
+    {
+      "epoch": 0.5789047310434219,
+      "grad_norm": 0.9588339328765869,
+      "learning_rate": 4.65987020660898e-06,
+      "loss": 0.1428,
+      "step": 3573
+    },
+    {
+      "epoch": 0.5790667530784187,
+      "grad_norm": 0.9257709980010986,
+      "learning_rate": 4.659649955579869e-06,
+      "loss": 0.1238,
+      "step": 3574
+    },
+    {
+      "epoch": 0.5792287751134154,
+      "grad_norm": 0.9726823568344116,
+      "learning_rate": 4.659429638470695e-06,
+      "loss": 0.1537,
+      "step": 3575
+    },
+    {
+      "epoch": 0.5793907971484121,
+      "grad_norm": 0.880847156047821,
+      "learning_rate": 4.659209255288201e-06,
+      "loss": 0.1404,
+      "step": 3576
+    },
+    {
+      "epoch": 0.579552819183409,
+      "grad_norm": 0.9704412817955017,
+      "learning_rate": 4.658988806039129e-06,
+      "loss": 0.1457,
+      "step": 3577
+    },
+    {
+      "epoch": 0.5797148412184057,
+      "grad_norm": 0.9277657866477966,
+      "learning_rate": 4.658768290730222e-06,
+      "loss": 0.1439,
+      "step": 3578
+    },
+    {
+      "epoch": 0.5798768632534025,
+      "grad_norm": 1.1359479427337646,
+      "learning_rate": 4.658547709368232e-06,
+      "loss": 0.181,
+      "step": 3579
+    },
+    {
+      "epoch": 0.5800388852883992,
+      "grad_norm": 1.0015472173690796,
+      "learning_rate": 4.658327061959904e-06,
+      "loss": 0.148,
+      "step": 3580
+    },
+    {
+      "epoch": 0.580200907323396,
+      "grad_norm": 0.9326437711715698,
+      "learning_rate": 4.658106348511992e-06,
+      "loss": 0.1476,
+      "step": 3581
+    },
+    {
+      "epoch": 0.5803629293583927,
+      "grad_norm": 1.0903030633926392,
+      "learning_rate": 4.6578855690312474e-06,
+      "loss": 0.1691,
+      "step": 3582
+    },
+    {
+      "epoch": 0.5805249513933896,
+      "grad_norm": 0.8469444513320923,
+      "learning_rate": 4.657664723524426e-06,
+      "loss": 0.1308,
+      "step": 3583
+    },
+    {
+      "epoch": 0.5806869734283863,
+      "grad_norm": 0.9137740135192871,
+      "learning_rate": 4.657443811998287e-06,
+      "loss": 0.1478,
+      "step": 3584
+    },
+    {
+      "epoch": 0.580848995463383,
+      "grad_norm": 0.9702494740486145,
+      "learning_rate": 4.657222834459588e-06,
+      "loss": 0.1455,
+      "step": 3585
+    },
+    {
+      "epoch": 0.5810110174983798,
+      "grad_norm": 0.9724593162536621,
+      "learning_rate": 4.65700179091509e-06,
+      "loss": 0.1515,
+      "step": 3586
+    },
+    {
+      "epoch": 0.5811730395333765,
+      "grad_norm": 0.8013084530830383,
+      "learning_rate": 4.6567806813715575e-06,
+      "loss": 0.1202,
+      "step": 3587
+    },
+    {
+      "epoch": 0.5813350615683733,
+      "grad_norm": 0.9419142007827759,
+      "learning_rate": 4.656559505835755e-06,
+      "loss": 0.1389,
+      "step": 3588
+    },
+    {
+      "epoch": 0.58149708360337,
+      "grad_norm": 0.9750432372093201,
+      "learning_rate": 4.6563382643144505e-06,
+      "loss": 0.1489,
+      "step": 3589
+    },
+    {
+      "epoch": 0.5816591056383669,
+      "grad_norm": 0.8804764151573181,
+      "learning_rate": 4.656116956814414e-06,
+      "loss": 0.1381,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5818211276733636,
+      "grad_norm": 0.9245219230651855,
+      "learning_rate": 4.655895583342415e-06,
+      "loss": 0.1461,
+      "step": 3591
+    },
+    {
+      "epoch": 0.5819831497083603,
+      "grad_norm": 0.8934546113014221,
+      "learning_rate": 4.655674143905229e-06,
+      "loss": 0.141,
+      "step": 3592
+    },
+    {
+      "epoch": 0.5821451717433571,
+      "grad_norm": 0.992603600025177,
+      "learning_rate": 4.655452638509631e-06,
+      "loss": 0.1639,
+      "step": 3593
+    },
+    {
+      "epoch": 0.5823071937783538,
+      "grad_norm": 1.0013014078140259,
+      "learning_rate": 4.655231067162398e-06,
+      "loss": 0.1502,
+      "step": 3594
+    },
+    {
+      "epoch": 0.5824692158133506,
+      "grad_norm": 0.9464520812034607,
+      "learning_rate": 4.655009429870311e-06,
+      "loss": 0.1467,
+      "step": 3595
+    },
+    {
+      "epoch": 0.5826312378483474,
+      "grad_norm": 0.8180822730064392,
+      "learning_rate": 4.65478772664015e-06,
+      "loss": 0.1202,
+      "step": 3596
+    },
+    {
+      "epoch": 0.5827932598833442,
+      "grad_norm": 0.7912185788154602,
+      "learning_rate": 4.6545659574786985e-06,
+      "loss": 0.1202,
+      "step": 3597
+    },
+    {
+      "epoch": 0.5829552819183409,
+      "grad_norm": 0.8654948472976685,
+      "learning_rate": 4.654344122392742e-06,
+      "loss": 0.1296,
+      "step": 3598
+    },
+    {
+      "epoch": 0.5831173039533376,
+      "grad_norm": 1.000624418258667,
+      "learning_rate": 4.65412222138907e-06,
+      "loss": 0.1421,
+      "step": 3599
+    },
+    {
+      "epoch": 0.5832793259883344,
+      "grad_norm": 0.9236934185028076,
+      "learning_rate": 4.6539002544744705e-06,
+      "loss": 0.139,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5834413480233311,
+      "grad_norm": 0.8793966174125671,
+      "learning_rate": 4.653678221655735e-06,
+      "loss": 0.1356,
+      "step": 3601
+    },
+    {
+      "epoch": 0.583603370058328,
+      "grad_norm": 0.9714186787605286,
+      "learning_rate": 4.653456122939659e-06,
+      "loss": 0.132,
+      "step": 3602
+    },
+    {
+      "epoch": 0.5837653920933247,
+      "grad_norm": 1.067306399345398,
+      "learning_rate": 4.653233958333036e-06,
+      "loss": 0.159,
+      "step": 3603
+    },
+    {
+      "epoch": 0.5839274141283215,
+      "grad_norm": 0.8334851264953613,
+      "learning_rate": 4.653011727842665e-06,
+      "loss": 0.1267,
+      "step": 3604
+    },
+    {
+      "epoch": 0.5840894361633182,
+      "grad_norm": 0.8716873526573181,
+      "learning_rate": 4.652789431475346e-06,
+      "loss": 0.1423,
+      "step": 3605
+    },
+    {
+      "epoch": 0.5842514581983149,
+      "grad_norm": 0.8517516851425171,
+      "learning_rate": 4.652567069237877e-06,
+      "loss": 0.1191,
+      "step": 3606
+    },
+    {
+      "epoch": 0.5844134802333117,
+      "grad_norm": 0.9621455669403076,
+      "learning_rate": 4.652344641137068e-06,
+      "loss": 0.1457,
+      "step": 3607
+    },
+    {
+      "epoch": 0.5845755022683085,
+      "grad_norm": 1.0080509185791016,
+      "learning_rate": 4.652122147179721e-06,
+      "loss": 0.1642,
+      "step": 3608
+    },
+    {
+      "epoch": 0.5847375243033053,
+      "grad_norm": 0.8625085353851318,
+      "learning_rate": 4.6518995873726434e-06,
+      "loss": 0.1353,
+      "step": 3609
+    },
+    {
+      "epoch": 0.584899546338302,
+      "grad_norm": 0.8965188264846802,
+      "learning_rate": 4.651676961722647e-06,
+      "loss": 0.1345,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5850615683732988,
+      "grad_norm": 1.0228919982910156,
+      "learning_rate": 4.651454270236541e-06,
+      "loss": 0.1283,
+      "step": 3611
+    },
+    {
+      "epoch": 0.5852235904082955,
+      "grad_norm": 1.0728081464767456,
+      "learning_rate": 4.651231512921142e-06,
+      "loss": 0.1795,
+      "step": 3612
+    },
+    {
+      "epoch": 0.5853856124432922,
+      "grad_norm": 1.0229250192642212,
+      "learning_rate": 4.651008689783264e-06,
+      "loss": 0.1516,
+      "step": 3613
+    },
+    {
+      "epoch": 0.5855476344782891,
+      "grad_norm": 0.8845334649085999,
+      "learning_rate": 4.650785800829726e-06,
+      "loss": 0.1361,
+      "step": 3614
+    },
+    {
+      "epoch": 0.5857096565132858,
+      "grad_norm": 0.9456861615180969,
+      "learning_rate": 4.650562846067347e-06,
+      "loss": 0.159,
+      "step": 3615
+    },
+    {
+      "epoch": 0.5858716785482826,
+      "grad_norm": 1.1445139646530151,
+      "learning_rate": 4.650339825502949e-06,
+      "loss": 0.163,
+      "step": 3616
+    },
+    {
+      "epoch": 0.5860337005832793,
+      "grad_norm": 0.9295792579650879,
+      "learning_rate": 4.650116739143356e-06,
+      "loss": 0.1346,
+      "step": 3617
+    },
+    {
+      "epoch": 0.5861957226182761,
+      "grad_norm": 0.9515857696533203,
+      "learning_rate": 4.6498935869953945e-06,
+      "loss": 0.1431,
+      "step": 3618
+    },
+    {
+      "epoch": 0.5863577446532728,
+      "grad_norm": 0.8937877416610718,
+      "learning_rate": 4.649670369065891e-06,
+      "loss": 0.1438,
+      "step": 3619
+    },
+    {
+      "epoch": 0.5865197666882696,
+      "grad_norm": 0.9099476337432861,
+      "learning_rate": 4.649447085361677e-06,
+      "loss": 0.1447,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5866817887232664,
+      "grad_norm": 0.973203182220459,
+      "learning_rate": 4.649223735889583e-06,
+      "loss": 0.1532,
+      "step": 3621
+    },
+    {
+      "epoch": 0.5868438107582631,
+      "grad_norm": 0.8838689923286438,
+      "learning_rate": 4.649000320656445e-06,
+      "loss": 0.1507,
+      "step": 3622
+    },
+    {
+      "epoch": 0.5870058327932599,
+      "grad_norm": 0.8679401874542236,
+      "learning_rate": 4.6487768396690965e-06,
+      "loss": 0.1281,
+      "step": 3623
+    },
+    {
+      "epoch": 0.5871678548282566,
+      "grad_norm": 0.9477259516716003,
+      "learning_rate": 4.648553292934377e-06,
+      "loss": 0.1491,
+      "step": 3624
+    },
+    {
+      "epoch": 0.5873298768632534,
+      "grad_norm": 0.923620879650116,
+      "learning_rate": 4.648329680459127e-06,
+      "loss": 0.1535,
+      "step": 3625
+    },
+    {
+      "epoch": 0.5874918988982502,
+      "grad_norm": 0.9671549797058105,
+      "learning_rate": 4.648106002250186e-06,
+      "loss": 0.1598,
+      "step": 3626
+    },
+    {
+      "epoch": 0.587653920933247,
+      "grad_norm": 0.845306396484375,
+      "learning_rate": 4.6478822583144015e-06,
+      "loss": 0.1349,
+      "step": 3627
+    },
+    {
+      "epoch": 0.5878159429682437,
+      "grad_norm": 1.0081162452697754,
+      "learning_rate": 4.647658448658616e-06,
+      "loss": 0.1665,
+      "step": 3628
+    },
+    {
+      "epoch": 0.5879779650032404,
+      "grad_norm": 0.9138731360435486,
+      "learning_rate": 4.64743457328968e-06,
+      "loss": 0.1496,
+      "step": 3629
+    },
+    {
+      "epoch": 0.5881399870382372,
+      "grad_norm": 0.9138461351394653,
+      "learning_rate": 4.647210632214443e-06,
+      "loss": 0.1446,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5883020090732339,
+      "grad_norm": 0.8777061104774475,
+      "learning_rate": 4.6469866254397564e-06,
+      "loss": 0.1461,
+      "step": 3631
+    },
+    {
+      "epoch": 0.5884640311082308,
+      "grad_norm": 0.984879732131958,
+      "learning_rate": 4.646762552972475e-06,
+      "loss": 0.1488,
+      "step": 3632
+    },
+    {
+      "epoch": 0.5886260531432275,
+      "grad_norm": 0.9393163919448853,
+      "learning_rate": 4.646538414819454e-06,
+      "loss": 0.1398,
+      "step": 3633
+    },
+    {
+      "epoch": 0.5887880751782243,
+      "grad_norm": 0.8989644646644592,
+      "learning_rate": 4.646314210987552e-06,
+      "loss": 0.1317,
+      "step": 3634
+    },
+    {
+      "epoch": 0.588950097213221,
+      "grad_norm": 1.0422638654708862,
+      "learning_rate": 4.646089941483629e-06,
+      "loss": 0.1567,
+      "step": 3635
+    },
+    {
+      "epoch": 0.5891121192482177,
+      "grad_norm": 0.9822480082511902,
+      "learning_rate": 4.645865606314548e-06,
+      "loss": 0.1557,
+      "step": 3636
+    },
+    {
+      "epoch": 0.5892741412832145,
+      "grad_norm": 1.0314334630966187,
+      "learning_rate": 4.645641205487172e-06,
+      "loss": 0.1502,
+      "step": 3637
+    },
+    {
+      "epoch": 0.5894361633182112,
+      "grad_norm": 1.0244547128677368,
+      "learning_rate": 4.645416739008367e-06,
+      "loss": 0.159,
+      "step": 3638
+    },
+    {
+      "epoch": 0.5895981853532081,
+      "grad_norm": 1.0546983480453491,
+      "learning_rate": 4.645192206885003e-06,
+      "loss": 0.1466,
+      "step": 3639
+    },
+    {
+      "epoch": 0.5897602073882048,
+      "grad_norm": 0.9143233895301819,
+      "learning_rate": 4.644967609123947e-06,
+      "loss": 0.1341,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5899222294232016,
+      "grad_norm": 0.9980403184890747,
+      "learning_rate": 4.644742945732074e-06,
+      "loss": 0.1516,
+      "step": 3641
+    },
+    {
+      "epoch": 0.5900842514581983,
+      "grad_norm": 0.9705789685249329,
+      "learning_rate": 4.644518216716256e-06,
+      "loss": 0.1653,
+      "step": 3642
+    },
+    {
+      "epoch": 0.590246273493195,
+      "grad_norm": 0.951261579990387,
+      "learning_rate": 4.6442934220833716e-06,
+      "loss": 0.1499,
+      "step": 3643
+    },
+    {
+      "epoch": 0.5904082955281919,
+      "grad_norm": 0.9710079431533813,
+      "learning_rate": 4.644068561840297e-06,
+      "loss": 0.1623,
+      "step": 3644
+    },
+    {
+      "epoch": 0.5905703175631886,
+      "grad_norm": 0.9166867733001709,
+      "learning_rate": 4.643843635993913e-06,
+      "loss": 0.1437,
+      "step": 3645
+    },
+    {
+      "epoch": 0.5907323395981854,
+      "grad_norm": 0.9674573540687561,
+      "learning_rate": 4.643618644551101e-06,
+      "loss": 0.1575,
+      "step": 3646
+    },
+    {
+      "epoch": 0.5908943616331821,
+      "grad_norm": 0.9722270369529724,
+      "learning_rate": 4.643393587518747e-06,
+      "loss": 0.1585,
+      "step": 3647
+    },
+    {
+      "epoch": 0.5910563836681789,
+      "grad_norm": 0.8995879888534546,
+      "learning_rate": 4.643168464903736e-06,
+      "loss": 0.1396,
+      "step": 3648
+    },
+    {
+      "epoch": 0.5912184057031756,
+      "grad_norm": 0.9309381246566772,
+      "learning_rate": 4.642943276712956e-06,
+      "loss": 0.1465,
+      "step": 3649
+    },
+    {
+      "epoch": 0.5913804277381723,
+      "grad_norm": 0.9563421010971069,
+      "learning_rate": 4.642718022953297e-06,
+      "loss": 0.1389,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5915424497731692,
+      "grad_norm": 0.9059107303619385,
+      "learning_rate": 4.642492703631652e-06,
+      "loss": 0.1399,
+      "step": 3651
+    },
+    {
+      "epoch": 0.5917044718081659,
+      "grad_norm": 1.094211220741272,
+      "learning_rate": 4.642267318754915e-06,
+      "loss": 0.1802,
+      "step": 3652
+    },
+    {
+      "epoch": 0.5918664938431627,
+      "grad_norm": 0.952056348323822,
+      "learning_rate": 4.6420418683299825e-06,
+      "loss": 0.1658,
+      "step": 3653
+    },
+    {
+      "epoch": 0.5920285158781594,
+      "grad_norm": 0.9111037254333496,
+      "learning_rate": 4.641816352363753e-06,
+      "loss": 0.1377,
+      "step": 3654
+    },
+    {
+      "epoch": 0.5921905379131562,
+      "grad_norm": 0.8917416334152222,
+      "learning_rate": 4.641590770863126e-06,
+      "loss": 0.1374,
+      "step": 3655
+    },
+    {
+      "epoch": 0.592352559948153,
+      "grad_norm": 0.9995194673538208,
+      "learning_rate": 4.641365123835004e-06,
+      "loss": 0.1524,
+      "step": 3656
+    },
+    {
+      "epoch": 0.5925145819831497,
+      "grad_norm": 0.972366213798523,
+      "learning_rate": 4.641139411286291e-06,
+      "loss": 0.1565,
+      "step": 3657
+    },
+    {
+      "epoch": 0.5926766040181465,
+      "grad_norm": 1.033241868019104,
+      "learning_rate": 4.640913633223893e-06,
+      "loss": 0.1777,
+      "step": 3658
+    },
+    {
+      "epoch": 0.5928386260531432,
+      "grad_norm": 0.9050754904747009,
+      "learning_rate": 4.640687789654719e-06,
+      "loss": 0.1473,
+      "step": 3659
+    },
+    {
+      "epoch": 0.59300064808814,
+      "grad_norm": 0.962814211845398,
+      "learning_rate": 4.64046188058568e-06,
+      "loss": 0.1508,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5931626701231367,
+      "grad_norm": 1.0121870040893555,
+      "learning_rate": 4.640235906023686e-06,
+      "loss": 0.1686,
+      "step": 3661
+    },
+    {
+      "epoch": 0.5933246921581335,
+      "grad_norm": 0.9426042437553406,
+      "learning_rate": 4.6400098659756525e-06,
+      "loss": 0.1478,
+      "step": 3662
+    },
+    {
+      "epoch": 0.5934867141931303,
+      "grad_norm": 1.017554759979248,
+      "learning_rate": 4.639783760448497e-06,
+      "loss": 0.1571,
+      "step": 3663
+    },
+    {
+      "epoch": 0.593648736228127,
+      "grad_norm": 0.9146802425384521,
+      "learning_rate": 4.639557589449135e-06,
+      "loss": 0.1372,
+      "step": 3664
+    },
+    {
+      "epoch": 0.5938107582631238,
+      "grad_norm": 0.938179612159729,
+      "learning_rate": 4.6393313529844895e-06,
+      "loss": 0.1441,
+      "step": 3665
+    },
+    {
+      "epoch": 0.5939727802981205,
+      "grad_norm": 0.8877557516098022,
+      "learning_rate": 4.639105051061481e-06,
+      "loss": 0.1409,
+      "step": 3666
+    },
+    {
+      "epoch": 0.5941348023331173,
+      "grad_norm": 0.9249974489212036,
+      "learning_rate": 4.638878683687036e-06,
+      "loss": 0.1561,
+      "step": 3667
+    },
+    {
+      "epoch": 0.594296824368114,
+      "grad_norm": 0.906252384185791,
+      "learning_rate": 4.638652250868078e-06,
+      "loss": 0.1418,
+      "step": 3668
+    },
+    {
+      "epoch": 0.5944588464031109,
+      "grad_norm": 0.9417980909347534,
+      "learning_rate": 4.638425752611536e-06,
+      "loss": 0.1536,
+      "step": 3669
+    },
+    {
+      "epoch": 0.5946208684381076,
+      "grad_norm": 0.8455900549888611,
+      "learning_rate": 4.6381991889243416e-06,
+      "loss": 0.1347,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5947828904731044,
+      "grad_norm": 1.0007245540618896,
+      "learning_rate": 4.6379725598134265e-06,
+      "loss": 0.1662,
+      "step": 3671
+    },
+    {
+      "epoch": 0.5949449125081011,
+      "grad_norm": 0.8600866198539734,
+      "learning_rate": 4.637745865285725e-06,
+      "loss": 0.1389,
+      "step": 3672
+    },
+    {
+      "epoch": 0.5951069345430978,
+      "grad_norm": 0.9751458764076233,
+      "learning_rate": 4.637519105348173e-06,
+      "loss": 0.154,
+      "step": 3673
+    },
+    {
+      "epoch": 0.5952689565780946,
+      "grad_norm": 0.9357625842094421,
+      "learning_rate": 4.637292280007709e-06,
+      "loss": 0.1592,
+      "step": 3674
+    },
+    {
+      "epoch": 0.5954309786130914,
+      "grad_norm": 0.8890631794929504,
+      "learning_rate": 4.637065389271274e-06,
+      "loss": 0.137,
+      "step": 3675
+    },
+    {
+      "epoch": 0.5955930006480882,
+      "grad_norm": 0.8981730341911316,
+      "learning_rate": 4.6368384331458085e-06,
+      "loss": 0.1411,
+      "step": 3676
+    },
+    {
+      "epoch": 0.5957550226830849,
+      "grad_norm": 0.987389862537384,
+      "learning_rate": 4.636611411638259e-06,
+      "loss": 0.1455,
+      "step": 3677
+    },
+    {
+      "epoch": 0.5959170447180817,
+      "grad_norm": 1.0186084508895874,
+      "learning_rate": 4.63638432475557e-06,
+      "loss": 0.1597,
+      "step": 3678
+    },
+    {
+      "epoch": 0.5960790667530784,
+      "grad_norm": 1.0408918857574463,
+      "learning_rate": 4.636157172504692e-06,
+      "loss": 0.1647,
+      "step": 3679
+    },
+    {
+      "epoch": 0.5962410887880751,
+      "grad_norm": 0.9388018250465393,
+      "learning_rate": 4.635929954892572e-06,
+      "loss": 0.1402,
+      "step": 3680
+    },
+    {
+      "epoch": 0.596403110823072,
+      "grad_norm": 0.8321965932846069,
+      "learning_rate": 4.635702671926166e-06,
+      "loss": 0.1253,
+      "step": 3681
+    },
+    {
+      "epoch": 0.5965651328580687,
+      "grad_norm": 0.8523208498954773,
+      "learning_rate": 4.6354753236124254e-06,
+      "loss": 0.1384,
+      "step": 3682
+    },
+    {
+      "epoch": 0.5967271548930655,
+      "grad_norm": 1.0064256191253662,
+      "learning_rate": 4.635247909958308e-06,
+      "loss": 0.1559,
+      "step": 3683
+    },
+    {
+      "epoch": 0.5968891769280622,
+      "grad_norm": 0.8501699566841125,
+      "learning_rate": 4.635020430970771e-06,
+      "loss": 0.1278,
+      "step": 3684
+    },
+    {
+      "epoch": 0.597051198963059,
+      "grad_norm": 0.8985577821731567,
+      "learning_rate": 4.634792886656777e-06,
+      "loss": 0.1393,
+      "step": 3685
+    },
+    {
+      "epoch": 0.5972132209980557,
+      "grad_norm": 1.0347974300384521,
+      "learning_rate": 4.6345652770232856e-06,
+      "loss": 0.1612,
+      "step": 3686
+    },
+    {
+      "epoch": 0.5973752430330524,
+      "grad_norm": 0.8934628367424011,
+      "learning_rate": 4.634337602077263e-06,
+      "loss": 0.1367,
+      "step": 3687
+    },
+    {
+      "epoch": 0.5975372650680493,
+      "grad_norm": 1.1091103553771973,
+      "learning_rate": 4.6341098618256745e-06,
+      "loss": 0.16,
+      "step": 3688
+    },
+    {
+      "epoch": 0.597699287103046,
+      "grad_norm": 1.0096241235733032,
+      "learning_rate": 4.633882056275488e-06,
+      "loss": 0.1496,
+      "step": 3689
+    },
+    {
+      "epoch": 0.5978613091380428,
+      "grad_norm": 0.793545126914978,
+      "learning_rate": 4.633654185433676e-06,
+      "loss": 0.1135,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5980233311730395,
+      "grad_norm": 0.9871333241462708,
+      "learning_rate": 4.633426249307208e-06,
+      "loss": 0.1621,
+      "step": 3691
+    },
+    {
+      "epoch": 0.5981853532080363,
+      "grad_norm": 0.9502182602882385,
+      "learning_rate": 4.63319824790306e-06,
+      "loss": 0.1588,
+      "step": 3692
+    },
+    {
+      "epoch": 0.598347375243033,
+      "grad_norm": 0.9651210904121399,
+      "learning_rate": 4.632970181228208e-06,
+      "loss": 0.1581,
+      "step": 3693
+    },
+    {
+      "epoch": 0.5985093972780298,
+      "grad_norm": 0.91889488697052,
+      "learning_rate": 4.6327420492896295e-06,
+      "loss": 0.1405,
+      "step": 3694
+    },
+    {
+      "epoch": 0.5986714193130266,
+      "grad_norm": 0.9224652051925659,
+      "learning_rate": 4.632513852094306e-06,
+      "loss": 0.1439,
+      "step": 3695
+    },
+    {
+      "epoch": 0.5988334413480233,
+      "grad_norm": 1.0088536739349365,
+      "learning_rate": 4.632285589649219e-06,
+      "loss": 0.1434,
+      "step": 3696
+    },
+    {
+      "epoch": 0.5989954633830201,
+      "grad_norm": 0.9861193299293518,
+      "learning_rate": 4.632057261961353e-06,
+      "loss": 0.1488,
+      "step": 3697
+    },
+    {
+      "epoch": 0.5991574854180168,
+      "grad_norm": 0.9628931879997253,
+      "learning_rate": 4.631828869037694e-06,
+      "loss": 0.1572,
+      "step": 3698
+    },
+    {
+      "epoch": 0.5993195074530137,
+      "grad_norm": 0.9910879135131836,
+      "learning_rate": 4.631600410885231e-06,
+      "loss": 0.1529,
+      "step": 3699
+    },
+    {
+      "epoch": 0.5994815294880104,
+      "grad_norm": 0.876059889793396,
+      "learning_rate": 4.631371887510954e-06,
+      "loss": 0.1401,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5996435515230071,
+      "grad_norm": 0.9525460004806519,
+      "learning_rate": 4.6311432989218545e-06,
+      "loss": 0.1607,
+      "step": 3701
+    },
+    {
+      "epoch": 0.5998055735580039,
+      "grad_norm": 0.8962662816047668,
+      "learning_rate": 4.630914645124928e-06,
+      "loss": 0.155,
+      "step": 3702
+    },
+    {
+      "epoch": 0.5999675955930006,
+      "grad_norm": 0.852394163608551,
+      "learning_rate": 4.630685926127169e-06,
+      "loss": 0.1339,
+      "step": 3703
+    },
+    {
+      "epoch": 0.6001296176279974,
+      "grad_norm": 1.0351753234863281,
+      "learning_rate": 4.630457141935577e-06,
+      "loss": 0.1648,
+      "step": 3704
+    },
+    {
+      "epoch": 0.6002916396629941,
+      "grad_norm": 0.8926696181297302,
+      "learning_rate": 4.630228292557153e-06,
+      "loss": 0.1472,
+      "step": 3705
+    },
+    {
+      "epoch": 0.600453661697991,
+      "grad_norm": 0.8851702213287354,
+      "learning_rate": 4.629999377998898e-06,
+      "loss": 0.1355,
+      "step": 3706
+    },
+    {
+      "epoch": 0.6006156837329877,
+      "grad_norm": 0.9555678367614746,
+      "learning_rate": 4.629770398267815e-06,
+      "loss": 0.1617,
+      "step": 3707
+    },
+    {
+      "epoch": 0.6007777057679844,
+      "grad_norm": 0.979743242263794,
+      "learning_rate": 4.629541353370914e-06,
+      "loss": 0.1756,
+      "step": 3708
+    },
+    {
+      "epoch": 0.6009397278029812,
+      "grad_norm": 0.8310829997062683,
+      "learning_rate": 4.6293122433152e-06,
+      "loss": 0.127,
+      "step": 3709
+    },
+    {
+      "epoch": 0.6011017498379779,
+      "grad_norm": 0.9136250615119934,
+      "learning_rate": 4.629083068107684e-06,
+      "loss": 0.1452,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6012637718729748,
+      "grad_norm": 0.915210485458374,
+      "learning_rate": 4.628853827755378e-06,
+      "loss": 0.1557,
+      "step": 3711
+    },
+    {
+      "epoch": 0.6014257939079715,
+      "grad_norm": 0.902146577835083,
+      "learning_rate": 4.628624522265298e-06,
+      "loss": 0.1351,
+      "step": 3712
+    },
+    {
+      "epoch": 0.6015878159429683,
+      "grad_norm": 0.8265898823738098,
+      "learning_rate": 4.628395151644458e-06,
+      "loss": 0.1417,
+      "step": 3713
+    },
+    {
+      "epoch": 0.601749837977965,
+      "grad_norm": 1.038116693496704,
+      "learning_rate": 4.628165715899877e-06,
+      "loss": 0.1567,
+      "step": 3714
+    },
+    {
+      "epoch": 0.6019118600129617,
+      "grad_norm": 1.0465978384017944,
+      "learning_rate": 4.6279362150385755e-06,
+      "loss": 0.1566,
+      "step": 3715
+    },
+    {
+      "epoch": 0.6020738820479585,
+      "grad_norm": 0.9295293688774109,
+      "learning_rate": 4.627706649067575e-06,
+      "loss": 0.1515,
+      "step": 3716
+    },
+    {
+      "epoch": 0.6022359040829552,
+      "grad_norm": 0.8935041427612305,
+      "learning_rate": 4.6274770179939e-06,
+      "loss": 0.1467,
+      "step": 3717
+    },
+    {
+      "epoch": 0.6023979261179521,
+      "grad_norm": 1.1107021570205688,
+      "learning_rate": 4.627247321824576e-06,
+      "loss": 0.1565,
+      "step": 3718
+    },
+    {
+      "epoch": 0.6025599481529488,
+      "grad_norm": 0.9548209309577942,
+      "learning_rate": 4.627017560566633e-06,
+      "loss": 0.1464,
+      "step": 3719
+    },
+    {
+      "epoch": 0.6027219701879456,
+      "grad_norm": 0.8951996564865112,
+      "learning_rate": 4.6267877342271e-06,
+      "loss": 0.1415,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6028839922229423,
+      "grad_norm": 0.8305820226669312,
+      "learning_rate": 4.626557842813008e-06,
+      "loss": 0.1319,
+      "step": 3721
+    },
+    {
+      "epoch": 0.6030460142579391,
+      "grad_norm": 0.8860799074172974,
+      "learning_rate": 4.626327886331392e-06,
+      "loss": 0.1335,
+      "step": 3722
+    },
+    {
+      "epoch": 0.6032080362929358,
+      "grad_norm": 0.9841209053993225,
+      "learning_rate": 4.626097864789289e-06,
+      "loss": 0.1727,
+      "step": 3723
+    },
+    {
+      "epoch": 0.6033700583279326,
+      "grad_norm": 0.8832147717475891,
+      "learning_rate": 4.625867778193737e-06,
+      "loss": 0.1418,
+      "step": 3724
+    },
+    {
+      "epoch": 0.6035320803629294,
+      "grad_norm": 0.9186353087425232,
+      "learning_rate": 4.625637626551774e-06,
+      "loss": 0.1462,
+      "step": 3725
+    },
+    {
+      "epoch": 0.6036941023979261,
+      "grad_norm": 0.8324099779129028,
+      "learning_rate": 4.625407409870444e-06,
+      "loss": 0.1441,
+      "step": 3726
+    },
+    {
+      "epoch": 0.6038561244329229,
+      "grad_norm": 0.7835507988929749,
+      "learning_rate": 4.625177128156791e-06,
+      "loss": 0.1322,
+      "step": 3727
+    },
+    {
+      "epoch": 0.6040181464679196,
+      "grad_norm": 0.9128600358963013,
+      "learning_rate": 4.624946781417861e-06,
+      "loss": 0.1451,
+      "step": 3728
+    },
+    {
+      "epoch": 0.6041801685029164,
+      "grad_norm": 0.7894352078437805,
+      "learning_rate": 4.624716369660701e-06,
+      "loss": 0.1217,
+      "step": 3729
+    },
+    {
+      "epoch": 0.6043421905379132,
+      "grad_norm": 0.8810085654258728,
+      "learning_rate": 4.624485892892363e-06,
+      "loss": 0.1463,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6045042125729099,
+      "grad_norm": 1.0495206117630005,
+      "learning_rate": 4.624255351119897e-06,
+      "loss": 0.1634,
+      "step": 3731
+    },
+    {
+      "epoch": 0.6046662346079067,
+      "grad_norm": 1.0064868927001953,
+      "learning_rate": 4.624024744350358e-06,
+      "loss": 0.1705,
+      "step": 3732
+    },
+    {
+      "epoch": 0.6048282566429034,
+      "grad_norm": 0.985504150390625,
+      "learning_rate": 4.6237940725908014e-06,
+      "loss": 0.1366,
+      "step": 3733
+    },
+    {
+      "epoch": 0.6049902786779002,
+      "grad_norm": 0.9360939264297485,
+      "learning_rate": 4.623563335848286e-06,
+      "loss": 0.1567,
+      "step": 3734
+    },
+    {
+      "epoch": 0.6051523007128969,
+      "grad_norm": 0.9163661003112793,
+      "learning_rate": 4.623332534129872e-06,
+      "loss": 0.1453,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6053143227478938,
+      "grad_norm": 1.0676296949386597,
+      "learning_rate": 4.62310166744262e-06,
+      "loss": 0.1633,
+      "step": 3736
+    },
+    {
+      "epoch": 0.6054763447828905,
+      "grad_norm": 0.9336465001106262,
+      "learning_rate": 4.622870735793595e-06,
+      "loss": 0.1483,
+      "step": 3737
+    },
+    {
+      "epoch": 0.6056383668178872,
+      "grad_norm": 0.8538773655891418,
+      "learning_rate": 4.622639739189863e-06,
+      "loss": 0.128,
+      "step": 3738
+    },
+    {
+      "epoch": 0.605800388852884,
+      "grad_norm": 0.9462392926216125,
+      "learning_rate": 4.622408677638491e-06,
+      "loss": 0.1465,
+      "step": 3739
+    },
+    {
+      "epoch": 0.6059624108878807,
+      "grad_norm": 0.965332567691803,
+      "learning_rate": 4.62217755114655e-06,
+      "loss": 0.1309,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6061244329228775,
+      "grad_norm": 0.9652532339096069,
+      "learning_rate": 4.621946359721112e-06,
+      "loss": 0.1488,
+      "step": 3741
+    },
+    {
+      "epoch": 0.6062864549578743,
+      "grad_norm": 0.9327961802482605,
+      "learning_rate": 4.62171510336925e-06,
+      "loss": 0.1618,
+      "step": 3742
+    },
+    {
+      "epoch": 0.6064484769928711,
+      "grad_norm": 0.9430671334266663,
+      "learning_rate": 4.621483782098041e-06,
+      "loss": 0.1497,
+      "step": 3743
+    },
+    {
+      "epoch": 0.6066104990278678,
+      "grad_norm": 1.0052106380462646,
+      "learning_rate": 4.621252395914561e-06,
+      "loss": 0.1474,
+      "step": 3744
+    },
+    {
+      "epoch": 0.6067725210628645,
+      "grad_norm": 0.8685867786407471,
+      "learning_rate": 4.621020944825891e-06,
+      "loss": 0.1379,
+      "step": 3745
+    },
+    {
+      "epoch": 0.6069345430978613,
+      "grad_norm": 0.9735004305839539,
+      "learning_rate": 4.620789428839114e-06,
+      "loss": 0.1687,
+      "step": 3746
+    },
+    {
+      "epoch": 0.607096565132858,
+      "grad_norm": 0.8425636291503906,
+      "learning_rate": 4.6205578479613125e-06,
+      "loss": 0.1331,
+      "step": 3747
+    },
+    {
+      "epoch": 0.6072585871678549,
+      "grad_norm": 0.9645400047302246,
+      "learning_rate": 4.620326202199572e-06,
+      "loss": 0.1448,
+      "step": 3748
+    },
+    {
+      "epoch": 0.6074206092028516,
+      "grad_norm": 0.8548397421836853,
+      "learning_rate": 4.62009449156098e-06,
+      "loss": 0.139,
+      "step": 3749
+    },
+    {
+      "epoch": 0.6075826312378484,
+      "grad_norm": 0.92023104429245,
+      "learning_rate": 4.619862716052629e-06,
+      "loss": 0.1418,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6077446532728451,
+      "grad_norm": 1.001118540763855,
+      "learning_rate": 4.6196308756816075e-06,
+      "loss": 0.1444,
+      "step": 3751
+    },
+    {
+      "epoch": 0.6079066753078418,
+      "grad_norm": 1.030665397644043,
+      "learning_rate": 4.6193989704550105e-06,
+      "loss": 0.1619,
+      "step": 3752
+    },
+    {
+      "epoch": 0.6080686973428386,
+      "grad_norm": 0.9593086242675781,
+      "learning_rate": 4.619167000379934e-06,
+      "loss": 0.1527,
+      "step": 3753
+    },
+    {
+      "epoch": 0.6082307193778353,
+      "grad_norm": 0.9325686097145081,
+      "learning_rate": 4.6189349654634766e-06,
+      "loss": 0.1493,
+      "step": 3754
+    },
+    {
+      "epoch": 0.6083927414128322,
+      "grad_norm": 1.058605670928955,
+      "learning_rate": 4.618702865712736e-06,
+      "loss": 0.1687,
+      "step": 3755
+    },
+    {
+      "epoch": 0.6085547634478289,
+      "grad_norm": 0.9116065502166748,
+      "learning_rate": 4.618470701134815e-06,
+      "loss": 0.1231,
+      "step": 3756
+    },
+    {
+      "epoch": 0.6087167854828257,
+      "grad_norm": 1.0380024909973145,
+      "learning_rate": 4.6182384717368174e-06,
+      "loss": 0.1602,
+      "step": 3757
+    },
+    {
+      "epoch": 0.6088788075178224,
+      "grad_norm": 1.04635488986969,
+      "learning_rate": 4.618006177525849e-06,
+      "loss": 0.1619,
+      "step": 3758
+    },
+    {
+      "epoch": 0.6090408295528191,
+      "grad_norm": 0.9052918553352356,
+      "learning_rate": 4.617773818509016e-06,
+      "loss": 0.1408,
+      "step": 3759
+    },
+    {
+      "epoch": 0.609202851587816,
+      "grad_norm": 0.8943411111831665,
+      "learning_rate": 4.61754139469343e-06,
+      "loss": 0.1294,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6093648736228127,
+      "grad_norm": 0.9428607225418091,
+      "learning_rate": 4.617308906086201e-06,
+      "loss": 0.1561,
+      "step": 3761
+    },
+    {
+      "epoch": 0.6095268956578095,
+      "grad_norm": 0.9149422645568848,
+      "learning_rate": 4.6170763526944425e-06,
+      "loss": 0.1324,
+      "step": 3762
+    },
+    {
+      "epoch": 0.6096889176928062,
+      "grad_norm": 0.9159535765647888,
+      "learning_rate": 4.616843734525272e-06,
+      "loss": 0.1459,
+      "step": 3763
+    },
+    {
+      "epoch": 0.609850939727803,
+      "grad_norm": 0.915641188621521,
+      "learning_rate": 4.616611051585806e-06,
+      "loss": 0.1408,
+      "step": 3764
+    },
+    {
+      "epoch": 0.6100129617627997,
+      "grad_norm": 0.9071922898292542,
+      "learning_rate": 4.616378303883163e-06,
+      "loss": 0.145,
+      "step": 3765
+    },
+    {
+      "epoch": 0.6101749837977966,
+      "grad_norm": 0.8846275806427002,
+      "learning_rate": 4.6161454914244665e-06,
+      "loss": 0.1196,
+      "step": 3766
+    },
+    {
+      "epoch": 0.6103370058327933,
+      "grad_norm": 0.8655915856361389,
+      "learning_rate": 4.615912614216838e-06,
+      "loss": 0.1349,
+      "step": 3767
+    },
+    {
+      "epoch": 0.61049902786779,
+      "grad_norm": 0.8834684491157532,
+      "learning_rate": 4.615679672267405e-06,
+      "loss": 0.144,
+      "step": 3768
+    },
+    {
+      "epoch": 0.6106610499027868,
+      "grad_norm": 0.9629760384559631,
+      "learning_rate": 4.615446665583293e-06,
+      "loss": 0.149,
+      "step": 3769
+    },
+    {
+      "epoch": 0.6108230719377835,
+      "grad_norm": 0.8233225345611572,
+      "learning_rate": 4.615213594171633e-06,
+      "loss": 0.1256,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6109850939727803,
+      "grad_norm": 0.8711490631103516,
+      "learning_rate": 4.6149804580395555e-06,
+      "loss": 0.1413,
+      "step": 3771
+    },
+    {
+      "epoch": 0.611147116007777,
+      "grad_norm": 0.8631579875946045,
+      "learning_rate": 4.614747257194194e-06,
+      "loss": 0.1279,
+      "step": 3772
+    },
+    {
+      "epoch": 0.6113091380427739,
+      "grad_norm": 1.0351872444152832,
+      "learning_rate": 4.614513991642684e-06,
+      "loss": 0.1701,
+      "step": 3773
+    },
+    {
+      "epoch": 0.6114711600777706,
+      "grad_norm": 0.8989577293395996,
+      "learning_rate": 4.614280661392163e-06,
+      "loss": 0.1516,
+      "step": 3774
+    },
+    {
+      "epoch": 0.6116331821127673,
+      "grad_norm": 0.9219076037406921,
+      "learning_rate": 4.61404726644977e-06,
+      "loss": 0.1482,
+      "step": 3775
+    },
+    {
+      "epoch": 0.6117952041477641,
+      "grad_norm": 1.016165852546692,
+      "learning_rate": 4.613813806822647e-06,
+      "loss": 0.1632,
+      "step": 3776
+    },
+    {
+      "epoch": 0.6119572261827608,
+      "grad_norm": 0.8687697649002075,
+      "learning_rate": 4.613580282517936e-06,
+      "loss": 0.1438,
+      "step": 3777
+    },
+    {
+      "epoch": 0.6121192482177576,
+      "grad_norm": 0.941464900970459,
+      "learning_rate": 4.613346693542784e-06,
+      "loss": 0.1477,
+      "step": 3778
+    },
+    {
+      "epoch": 0.6122812702527544,
+      "grad_norm": 0.8832119703292847,
+      "learning_rate": 4.613113039904337e-06,
+      "loss": 0.1303,
+      "step": 3779
+    },
+    {
+      "epoch": 0.6124432922877512,
+      "grad_norm": 0.837990403175354,
+      "learning_rate": 4.6128793216097445e-06,
+      "loss": 0.1366,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6126053143227479,
+      "grad_norm": 0.8604587912559509,
+      "learning_rate": 4.612645538666157e-06,
+      "loss": 0.1297,
+      "step": 3781
+    },
+    {
+      "epoch": 0.6127673363577446,
+      "grad_norm": 0.9468358755111694,
+      "learning_rate": 4.61241169108073e-06,
+      "loss": 0.1561,
+      "step": 3782
+    },
+    {
+      "epoch": 0.6129293583927414,
+      "grad_norm": 0.8678791522979736,
+      "learning_rate": 4.612177778860617e-06,
+      "loss": 0.1352,
+      "step": 3783
+    },
+    {
+      "epoch": 0.6130913804277381,
+      "grad_norm": 0.8907611966133118,
+      "learning_rate": 4.611943802012975e-06,
+      "loss": 0.1373,
+      "step": 3784
+    },
+    {
+      "epoch": 0.613253402462735,
+      "grad_norm": 0.8536321520805359,
+      "learning_rate": 4.611709760544963e-06,
+      "loss": 0.1302,
+      "step": 3785
+    },
+    {
+      "epoch": 0.6134154244977317,
+      "grad_norm": 0.8262412548065186,
+      "learning_rate": 4.611475654463743e-06,
+      "loss": 0.1426,
+      "step": 3786
+    },
+    {
+      "epoch": 0.6135774465327285,
+      "grad_norm": 0.9103348255157471,
+      "learning_rate": 4.611241483776478e-06,
+      "loss": 0.1445,
+      "step": 3787
+    },
+    {
+      "epoch": 0.6137394685677252,
+      "grad_norm": 0.9351794123649597,
+      "learning_rate": 4.6110072484903326e-06,
+      "loss": 0.1353,
+      "step": 3788
+    },
+    {
+      "epoch": 0.6139014906027219,
+      "grad_norm": 1.0249840021133423,
+      "learning_rate": 4.610772948612473e-06,
+      "loss": 0.1455,
+      "step": 3789
+    },
+    {
+      "epoch": 0.6140635126377187,
+      "grad_norm": 0.9844290018081665,
+      "learning_rate": 4.610538584150071e-06,
+      "loss": 0.1643,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6142255346727155,
+      "grad_norm": 0.9268943667411804,
+      "learning_rate": 4.6103041551102935e-06,
+      "loss": 0.1388,
+      "step": 3791
+    },
+    {
+      "epoch": 0.6143875567077123,
+      "grad_norm": 0.9610852599143982,
+      "learning_rate": 4.610069661500317e-06,
+      "loss": 0.1476,
+      "step": 3792
+    },
+    {
+      "epoch": 0.614549578742709,
+      "grad_norm": 0.9968616366386414,
+      "learning_rate": 4.609835103327315e-06,
+      "loss": 0.1548,
+      "step": 3793
+    },
+    {
+      "epoch": 0.6147116007777058,
+      "grad_norm": 0.9769883155822754,
+      "learning_rate": 4.609600480598464e-06,
+      "loss": 0.1619,
+      "step": 3794
+    },
+    {
+      "epoch": 0.6148736228127025,
+      "grad_norm": 0.9648511409759521,
+      "learning_rate": 4.609365793320944e-06,
+      "loss": 0.1464,
+      "step": 3795
+    },
+    {
+      "epoch": 0.6150356448476992,
+      "grad_norm": 0.9245080351829529,
+      "learning_rate": 4.6091310415019355e-06,
+      "loss": 0.1612,
+      "step": 3796
+    },
+    {
+      "epoch": 0.6151976668826961,
+      "grad_norm": 0.8954278230667114,
+      "learning_rate": 4.608896225148621e-06,
+      "loss": 0.1421,
+      "step": 3797
+    },
+    {
+      "epoch": 0.6153596889176928,
+      "grad_norm": 1.078444004058838,
+      "learning_rate": 4.608661344268185e-06,
+      "loss": 0.1863,
+      "step": 3798
+    },
+    {
+      "epoch": 0.6155217109526896,
+      "grad_norm": 0.7994515299797058,
+      "learning_rate": 4.608426398867815e-06,
+      "loss": 0.1162,
+      "step": 3799
+    },
+    {
+      "epoch": 0.6156837329876863,
+      "grad_norm": 0.8549524545669556,
+      "learning_rate": 4.608191388954699e-06,
+      "loss": 0.1416,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6158457550226831,
+      "grad_norm": 0.8880654573440552,
+      "learning_rate": 4.607956314536029e-06,
+      "loss": 0.1308,
+      "step": 3801
+    },
+    {
+      "epoch": 0.6160077770576798,
+      "grad_norm": 0.9206951856613159,
+      "learning_rate": 4.607721175618997e-06,
+      "loss": 0.1543,
+      "step": 3802
+    },
+    {
+      "epoch": 0.6161697990926766,
+      "grad_norm": 1.0166141986846924,
+      "learning_rate": 4.607485972210797e-06,
+      "loss": 0.1533,
+      "step": 3803
+    },
+    {
+      "epoch": 0.6163318211276734,
+      "grad_norm": 0.9442442655563354,
+      "learning_rate": 4.6072507043186265e-06,
+      "loss": 0.1442,
+      "step": 3804
+    },
+    {
+      "epoch": 0.6164938431626701,
+      "grad_norm": 1.1184486150741577,
+      "learning_rate": 4.607015371949683e-06,
+      "loss": 0.158,
+      "step": 3805
+    },
+    {
+      "epoch": 0.6166558651976669,
+      "grad_norm": 0.9163842797279358,
+      "learning_rate": 4.60677997511117e-06,
+      "loss": 0.142,
+      "step": 3806
+    },
+    {
+      "epoch": 0.6168178872326636,
+      "grad_norm": 0.8781442642211914,
+      "learning_rate": 4.606544513810287e-06,
+      "loss": 0.1387,
+      "step": 3807
+    },
+    {
+      "epoch": 0.6169799092676604,
+      "grad_norm": 0.9204709529876709,
+      "learning_rate": 4.606308988054239e-06,
+      "loss": 0.1549,
+      "step": 3808
+    },
+    {
+      "epoch": 0.6171419313026572,
+      "grad_norm": 0.9622353315353394,
+      "learning_rate": 4.606073397850234e-06,
+      "loss": 0.1534,
+      "step": 3809
+    },
+    {
+      "epoch": 0.6173039533376539,
+      "grad_norm": 1.0953677892684937,
+      "learning_rate": 4.605837743205479e-06,
+      "loss": 0.144,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6174659753726507,
+      "grad_norm": 0.8764045834541321,
+      "learning_rate": 4.6056020241271855e-06,
+      "loss": 0.1359,
+      "step": 3811
+    },
+    {
+      "epoch": 0.6176279974076474,
+      "grad_norm": 0.9213500022888184,
+      "learning_rate": 4.605366240622565e-06,
+      "loss": 0.1521,
+      "step": 3812
+    },
+    {
+      "epoch": 0.6177900194426442,
+      "grad_norm": 0.9212419986724854,
+      "learning_rate": 4.605130392698833e-06,
+      "loss": 0.1425,
+      "step": 3813
+    },
+    {
+      "epoch": 0.6179520414776409,
+      "grad_norm": 0.9497610330581665,
+      "learning_rate": 4.604894480363205e-06,
+      "loss": 0.1536,
+      "step": 3814
+    },
+    {
+      "epoch": 0.6181140635126378,
+      "grad_norm": 0.9440318942070007,
+      "learning_rate": 4.6046585036229005e-06,
+      "loss": 0.1481,
+      "step": 3815
+    },
+    {
+      "epoch": 0.6182760855476345,
+      "grad_norm": 0.8853862881660461,
+      "learning_rate": 4.604422462485138e-06,
+      "loss": 0.141,
+      "step": 3816
+    },
+    {
+      "epoch": 0.6184381075826313,
+      "grad_norm": 0.9499521851539612,
+      "learning_rate": 4.604186356957141e-06,
+      "loss": 0.147,
+      "step": 3817
+    },
+    {
+      "epoch": 0.618600129617628,
+      "grad_norm": 0.8915985822677612,
+      "learning_rate": 4.603950187046134e-06,
+      "loss": 0.1515,
+      "step": 3818
+    },
+    {
+      "epoch": 0.6187621516526247,
+      "grad_norm": 0.9126549959182739,
+      "learning_rate": 4.6037139527593424e-06,
+      "loss": 0.1505,
+      "step": 3819
+    },
+    {
+      "epoch": 0.6189241736876215,
+      "grad_norm": 1.0714317560195923,
+      "learning_rate": 4.603477654103994e-06,
+      "loss": 0.171,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6190861957226182,
+      "grad_norm": 1.0061242580413818,
+      "learning_rate": 4.60324129108732e-06,
+      "loss": 0.1606,
+      "step": 3821
+    },
+    {
+      "epoch": 0.6192482177576151,
+      "grad_norm": 1.0254570245742798,
+      "learning_rate": 4.603004863716553e-06,
+      "loss": 0.177,
+      "step": 3822
+    },
+    {
+      "epoch": 0.6194102397926118,
+      "grad_norm": 0.7951987981796265,
+      "learning_rate": 4.602768371998925e-06,
+      "loss": 0.1198,
+      "step": 3823
+    },
+    {
+      "epoch": 0.6195722618276086,
+      "grad_norm": 0.9415546655654907,
+      "learning_rate": 4.602531815941676e-06,
+      "loss": 0.1581,
+      "step": 3824
+    },
+    {
+      "epoch": 0.6197342838626053,
+      "grad_norm": 0.9311978220939636,
+      "learning_rate": 4.602295195552039e-06,
+      "loss": 0.1474,
+      "step": 3825
+    },
+    {
+      "epoch": 0.619896305897602,
+      "grad_norm": 0.8883768320083618,
+      "learning_rate": 4.602058510837257e-06,
+      "loss": 0.1383,
+      "step": 3826
+    },
+    {
+      "epoch": 0.6200583279325989,
+      "grad_norm": 1.01768958568573,
+      "learning_rate": 4.601821761804572e-06,
+      "loss": 0.1721,
+      "step": 3827
+    },
+    {
+      "epoch": 0.6202203499675956,
+      "grad_norm": 0.9163593053817749,
+      "learning_rate": 4.6015849484612265e-06,
+      "loss": 0.1258,
+      "step": 3828
+    },
+    {
+      "epoch": 0.6203823720025924,
+      "grad_norm": 0.9427400231361389,
+      "learning_rate": 4.601348070814468e-06,
+      "loss": 0.1308,
+      "step": 3829
+    },
+    {
+      "epoch": 0.6205443940375891,
+      "grad_norm": 1.179629921913147,
+      "learning_rate": 4.601111128871544e-06,
+      "loss": 0.1838,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6207064160725859,
+      "grad_norm": 1.0116398334503174,
+      "learning_rate": 4.600874122639703e-06,
+      "loss": 0.147,
+      "step": 3831
+    },
+    {
+      "epoch": 0.6208684381075826,
+      "grad_norm": 0.8127241134643555,
+      "learning_rate": 4.600637052126199e-06,
+      "loss": 0.1333,
+      "step": 3832
+    },
+    {
+      "epoch": 0.6210304601425793,
+      "grad_norm": 0.9073969125747681,
+      "learning_rate": 4.600399917338284e-06,
+      "loss": 0.1369,
+      "step": 3833
+    },
+    {
+      "epoch": 0.6211924821775762,
+      "grad_norm": 0.8607815504074097,
+      "learning_rate": 4.600162718283215e-06,
+      "loss": 0.1357,
+      "step": 3834
+    },
+    {
+      "epoch": 0.6213545042125729,
+      "grad_norm": 0.9723119735717773,
+      "learning_rate": 4.5999254549682484e-06,
+      "loss": 0.1456,
+      "step": 3835
+    },
+    {
+      "epoch": 0.6215165262475697,
+      "grad_norm": 0.9572594165802002,
+      "learning_rate": 4.599688127400645e-06,
+      "loss": 0.1548,
+      "step": 3836
+    },
+    {
+      "epoch": 0.6216785482825664,
+      "grad_norm": 0.9637323021888733,
+      "learning_rate": 4.599450735587666e-06,
+      "loss": 0.1533,
+      "step": 3837
+    },
+    {
+      "epoch": 0.6218405703175632,
+      "grad_norm": 0.9875199794769287,
+      "learning_rate": 4.599213279536575e-06,
+      "loss": 0.159,
+      "step": 3838
+    },
+    {
+      "epoch": 0.62200259235256,
+      "grad_norm": 0.9891331195831299,
+      "learning_rate": 4.598975759254638e-06,
+      "loss": 0.1594,
+      "step": 3839
+    },
+    {
+      "epoch": 0.6221646143875567,
+      "grad_norm": 0.941134512424469,
+      "learning_rate": 4.598738174749121e-06,
+      "loss": 0.1466,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6223266364225535,
+      "grad_norm": 0.8841784596443176,
+      "learning_rate": 4.598500526027296e-06,
+      "loss": 0.1571,
+      "step": 3841
+    },
+    {
+      "epoch": 0.6224886584575502,
+      "grad_norm": 0.9083060622215271,
+      "learning_rate": 4.598262813096432e-06,
+      "loss": 0.1543,
+      "step": 3842
+    },
+    {
+      "epoch": 0.622650680492547,
+      "grad_norm": 0.8198730945587158,
+      "learning_rate": 4.598025035963805e-06,
+      "loss": 0.1286,
+      "step": 3843
+    },
+    {
+      "epoch": 0.6228127025275437,
+      "grad_norm": 0.9744411706924438,
+      "learning_rate": 4.597787194636688e-06,
+      "loss": 0.1523,
+      "step": 3844
+    },
+    {
+      "epoch": 0.6229747245625405,
+      "grad_norm": 0.8268557190895081,
+      "learning_rate": 4.597549289122361e-06,
+      "loss": 0.1294,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6231367465975373,
+      "grad_norm": 1.0291190147399902,
+      "learning_rate": 4.597311319428099e-06,
+      "loss": 0.1556,
+      "step": 3846
+    },
+    {
+      "epoch": 0.623298768632534,
+      "grad_norm": 0.8949023485183716,
+      "learning_rate": 4.597073285561188e-06,
+      "loss": 0.1354,
+      "step": 3847
+    },
+    {
+      "epoch": 0.6234607906675308,
+      "grad_norm": 0.9648371338844299,
+      "learning_rate": 4.596835187528908e-06,
+      "loss": 0.1517,
+      "step": 3848
+    },
+    {
+      "epoch": 0.6236228127025275,
+      "grad_norm": 1.00344717502594,
+      "learning_rate": 4.596597025338547e-06,
+      "loss": 0.1659,
+      "step": 3849
+    },
+    {
+      "epoch": 0.6237848347375243,
+      "grad_norm": 0.8660650849342346,
+      "learning_rate": 4.59635879899739e-06,
+      "loss": 0.1356,
+      "step": 3850
+    },
+    {
+      "epoch": 0.623946856772521,
+      "grad_norm": 1.0011287927627563,
+      "learning_rate": 4.596120508512727e-06,
+      "loss": 0.1452,
+      "step": 3851
+    },
+    {
+      "epoch": 0.6241088788075179,
+      "grad_norm": 0.8602931499481201,
+      "learning_rate": 4.595882153891849e-06,
+      "loss": 0.1315,
+      "step": 3852
+    },
+    {
+      "epoch": 0.6242709008425146,
+      "grad_norm": 0.8699659705162048,
+      "learning_rate": 4.595643735142049e-06,
+      "loss": 0.1395,
+      "step": 3853
+    },
+    {
+      "epoch": 0.6244329228775113,
+      "grad_norm": 0.782909631729126,
+      "learning_rate": 4.595405252270622e-06,
+      "loss": 0.1228,
+      "step": 3854
+    },
+    {
+      "epoch": 0.6245949449125081,
+      "grad_norm": 0.9927259683609009,
+      "learning_rate": 4.595166705284864e-06,
+      "loss": 0.1533,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6247569669475048,
+      "grad_norm": 0.9355255365371704,
+      "learning_rate": 4.594928094192076e-06,
+      "loss": 0.1478,
+      "step": 3856
+    },
+    {
+      "epoch": 0.6249189889825016,
+      "grad_norm": 0.9448959231376648,
+      "learning_rate": 4.594689418999558e-06,
+      "loss": 0.1456,
+      "step": 3857
+    },
+    {
+      "epoch": 0.6250810110174984,
+      "grad_norm": 0.8999738097190857,
+      "learning_rate": 4.594450679714613e-06,
+      "loss": 0.144,
+      "step": 3858
+    },
+    {
+      "epoch": 0.6252430330524952,
+      "grad_norm": 0.852476954460144,
+      "learning_rate": 4.594211876344545e-06,
+      "loss": 0.1304,
+      "step": 3859
+    },
+    {
+      "epoch": 0.6254050550874919,
+      "grad_norm": 0.8585977554321289,
+      "learning_rate": 4.593973008896662e-06,
+      "loss": 0.14,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6255670771224887,
+      "grad_norm": 0.8411082625389099,
+      "learning_rate": 4.593734077378273e-06,
+      "loss": 0.1301,
+      "step": 3861
+    },
+    {
+      "epoch": 0.6257290991574854,
+      "grad_norm": 0.9660947322845459,
+      "learning_rate": 4.593495081796686e-06,
+      "loss": 0.1478,
+      "step": 3862
+    },
+    {
+      "epoch": 0.6258911211924821,
+      "grad_norm": 0.9528719186782837,
+      "learning_rate": 4.593256022159217e-06,
+      "loss": 0.141,
+      "step": 3863
+    },
+    {
+      "epoch": 0.626053143227479,
+      "grad_norm": 0.8987707495689392,
+      "learning_rate": 4.59301689847318e-06,
+      "loss": 0.1377,
+      "step": 3864
+    },
+    {
+      "epoch": 0.6262151652624757,
+      "grad_norm": 0.8041302561759949,
+      "learning_rate": 4.592777710745889e-06,
+      "loss": 0.1266,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6263771872974725,
+      "grad_norm": 0.9248653650283813,
+      "learning_rate": 4.592538458984666e-06,
+      "loss": 0.1373,
+      "step": 3866
+    },
+    {
+      "epoch": 0.6265392093324692,
+      "grad_norm": 0.9307569265365601,
+      "learning_rate": 4.592299143196829e-06,
+      "loss": 0.1531,
+      "step": 3867
+    },
+    {
+      "epoch": 0.626701231367466,
+      "grad_norm": 1.1055402755737305,
+      "learning_rate": 4.5920597633897015e-06,
+      "loss": 0.1396,
+      "step": 3868
+    },
+    {
+      "epoch": 0.6268632534024627,
+      "grad_norm": 0.9206196069717407,
+      "learning_rate": 4.591820319570609e-06,
+      "loss": 0.1401,
+      "step": 3869
+    },
+    {
+      "epoch": 0.6270252754374595,
+      "grad_norm": 0.8944006562232971,
+      "learning_rate": 4.5915808117468766e-06,
+      "loss": 0.1455,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6271872974724563,
+      "grad_norm": 0.87538743019104,
+      "learning_rate": 4.591341239925831e-06,
+      "loss": 0.1393,
+      "step": 3871
+    },
+    {
+      "epoch": 0.627349319507453,
+      "grad_norm": 1.0033069849014282,
+      "learning_rate": 4.591101604114807e-06,
+      "loss": 0.1531,
+      "step": 3872
+    },
+    {
+      "epoch": 0.6275113415424498,
+      "grad_norm": 1.0902032852172852,
+      "learning_rate": 4.590861904321133e-06,
+      "loss": 0.1743,
+      "step": 3873
+    },
+    {
+      "epoch": 0.6276733635774465,
+      "grad_norm": 1.0226129293441772,
+      "learning_rate": 4.590622140552144e-06,
+      "loss": 0.1443,
+      "step": 3874
+    },
+    {
+      "epoch": 0.6278353856124433,
+      "grad_norm": 0.9415045380592346,
+      "learning_rate": 4.590382312815178e-06,
+      "loss": 0.1412,
+      "step": 3875
+    },
+    {
+      "epoch": 0.62799740764744,
+      "grad_norm": 1.0171313285827637,
+      "learning_rate": 4.5901424211175715e-06,
+      "loss": 0.1658,
+      "step": 3876
+    },
+    {
+      "epoch": 0.6281594296824368,
+      "grad_norm": 0.9125556945800781,
+      "learning_rate": 4.589902465466665e-06,
+      "loss": 0.1417,
+      "step": 3877
+    },
+    {
+      "epoch": 0.6283214517174336,
+      "grad_norm": 0.9008194804191589,
+      "learning_rate": 4.5896624458698e-06,
+      "loss": 0.1405,
+      "step": 3878
+    },
+    {
+      "epoch": 0.6284834737524303,
+      "grad_norm": 0.9593300819396973,
+      "learning_rate": 4.589422362334321e-06,
+      "loss": 0.1403,
+      "step": 3879
+    },
+    {
+      "epoch": 0.6286454957874271,
+      "grad_norm": 0.9999030232429504,
+      "learning_rate": 4.5891822148675745e-06,
+      "loss": 0.1457,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6288075178224238,
+      "grad_norm": 0.9375879764556885,
+      "learning_rate": 4.588942003476907e-06,
+      "loss": 0.1425,
+      "step": 3881
+    },
+    {
+      "epoch": 0.6289695398574207,
+      "grad_norm": 0.8549643754959106,
+      "learning_rate": 4.588701728169671e-06,
+      "loss": 0.1354,
+      "step": 3882
+    },
+    {
+      "epoch": 0.6291315618924174,
+      "grad_norm": 0.9586682915687561,
+      "learning_rate": 4.588461388953216e-06,
+      "loss": 0.1707,
+      "step": 3883
+    },
+    {
+      "epoch": 0.6292935839274141,
+      "grad_norm": 0.8284324407577515,
+      "learning_rate": 4.5882209858348956e-06,
+      "loss": 0.1326,
+      "step": 3884
+    },
+    {
+      "epoch": 0.6294556059624109,
+      "grad_norm": 0.9033730030059814,
+      "learning_rate": 4.587980518822067e-06,
+      "loss": 0.1452,
+      "step": 3885
+    },
+    {
+      "epoch": 0.6296176279974076,
+      "grad_norm": 0.9468452334403992,
+      "learning_rate": 4.587739987922087e-06,
+      "loss": 0.1466,
+      "step": 3886
+    },
+    {
+      "epoch": 0.6297796500324044,
+      "grad_norm": 0.97198885679245,
+      "learning_rate": 4.587499393142316e-06,
+      "loss": 0.1459,
+      "step": 3887
+    },
+    {
+      "epoch": 0.6299416720674011,
+      "grad_norm": 1.0194365978240967,
+      "learning_rate": 4.587258734490115e-06,
+      "loss": 0.1597,
+      "step": 3888
+    },
+    {
+      "epoch": 0.630103694102398,
+      "grad_norm": 0.9163743257522583,
+      "learning_rate": 4.587018011972848e-06,
+      "loss": 0.1414,
+      "step": 3889
+    },
+    {
+      "epoch": 0.6302657161373947,
+      "grad_norm": 0.7911040186882019,
+      "learning_rate": 4.586777225597881e-06,
+      "loss": 0.1259,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6304277381723914,
+      "grad_norm": 0.9265354871749878,
+      "learning_rate": 4.58653637537258e-06,
+      "loss": 0.1403,
+      "step": 3891
+    },
+    {
+      "epoch": 0.6305897602073882,
+      "grad_norm": 0.9925926923751831,
+      "learning_rate": 4.586295461304315e-06,
+      "loss": 0.1521,
+      "step": 3892
+    },
+    {
+      "epoch": 0.6307517822423849,
+      "grad_norm": 1.0709103345870972,
+      "learning_rate": 4.586054483400459e-06,
+      "loss": 0.1772,
+      "step": 3893
+    },
+    {
+      "epoch": 0.6309138042773818,
+      "grad_norm": 0.9420156478881836,
+      "learning_rate": 4.585813441668383e-06,
+      "loss": 0.1495,
+      "step": 3894
+    },
+    {
+      "epoch": 0.6310758263123785,
+      "grad_norm": 0.840775728225708,
+      "learning_rate": 4.585572336115463e-06,
+      "loss": 0.1301,
+      "step": 3895
+    },
+    {
+      "epoch": 0.6312378483473753,
+      "grad_norm": 0.8908707499504089,
+      "learning_rate": 4.585331166749077e-06,
+      "loss": 0.1455,
+      "step": 3896
+    },
+    {
+      "epoch": 0.631399870382372,
+      "grad_norm": 0.9626536965370178,
+      "learning_rate": 4.5850899335766034e-06,
+      "loss": 0.1543,
+      "step": 3897
+    },
+    {
+      "epoch": 0.6315618924173687,
+      "grad_norm": 0.8213000297546387,
+      "learning_rate": 4.584848636605423e-06,
+      "loss": 0.133,
+      "step": 3898
+    },
+    {
+      "epoch": 0.6317239144523655,
+      "grad_norm": 0.9156107306480408,
+      "learning_rate": 4.584607275842921e-06,
+      "loss": 0.1383,
+      "step": 3899
+    },
+    {
+      "epoch": 0.6318859364873622,
+      "grad_norm": 0.9265947341918945,
+      "learning_rate": 4.58436585129648e-06,
+      "loss": 0.1336,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6320479585223591,
+      "grad_norm": 0.9135920405387878,
+      "learning_rate": 4.584124362973488e-06,
+      "loss": 0.1347,
+      "step": 3901
+    },
+    {
+      "epoch": 0.6322099805573558,
+      "grad_norm": 1.0750758647918701,
+      "learning_rate": 4.583882810881334e-06,
+      "loss": 0.1476,
+      "step": 3902
+    },
+    {
+      "epoch": 0.6323720025923526,
+      "grad_norm": 1.0379818677902222,
+      "learning_rate": 4.583641195027409e-06,
+      "loss": 0.1563,
+      "step": 3903
+    },
+    {
+      "epoch": 0.6325340246273493,
+      "grad_norm": 0.9261550903320312,
+      "learning_rate": 4.583399515419106e-06,
+      "loss": 0.139,
+      "step": 3904
+    },
+    {
+      "epoch": 0.6326960466623461,
+      "grad_norm": 0.9920796155929565,
+      "learning_rate": 4.58315777206382e-06,
+      "loss": 0.1358,
+      "step": 3905
+    },
+    {
+      "epoch": 0.6328580686973428,
+      "grad_norm": 0.9471383690834045,
+      "learning_rate": 4.582915964968946e-06,
+      "loss": 0.1418,
+      "step": 3906
+    },
+    {
+      "epoch": 0.6330200907323396,
+      "grad_norm": 1.0869243144989014,
+      "learning_rate": 4.582674094141885e-06,
+      "loss": 0.166,
+      "step": 3907
+    },
+    {
+      "epoch": 0.6331821127673364,
+      "grad_norm": 0.9842541813850403,
+      "learning_rate": 4.582432159590037e-06,
+      "loss": 0.1565,
+      "step": 3908
+    },
+    {
+      "epoch": 0.6333441348023331,
+      "grad_norm": 1.0878725051879883,
+      "learning_rate": 4.582190161320803e-06,
+      "loss": 0.173,
+      "step": 3909
+    },
+    {
+      "epoch": 0.6335061568373299,
+      "grad_norm": 0.9759945869445801,
+      "learning_rate": 4.58194809934159e-06,
+      "loss": 0.1576,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6336681788723266,
+      "grad_norm": 0.9469559192657471,
+      "learning_rate": 4.581705973659803e-06,
+      "loss": 0.1445,
+      "step": 3911
+    },
+    {
+      "epoch": 0.6338302009073234,
+      "grad_norm": 0.9183310866355896,
+      "learning_rate": 4.5814637842828506e-06,
+      "loss": 0.1411,
+      "step": 3912
+    },
+    {
+      "epoch": 0.6339922229423202,
+      "grad_norm": 1.0365331172943115,
+      "learning_rate": 4.581221531218144e-06,
+      "loss": 0.154,
+      "step": 3913
+    },
+    {
+      "epoch": 0.6341542449773169,
+      "grad_norm": 0.9276525974273682,
+      "learning_rate": 4.580979214473095e-06,
+      "loss": 0.1452,
+      "step": 3914
+    },
+    {
+      "epoch": 0.6343162670123137,
+      "grad_norm": 0.9412997961044312,
+      "learning_rate": 4.580736834055117e-06,
+      "loss": 0.1418,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6344782890473104,
+      "grad_norm": 0.906779408454895,
+      "learning_rate": 4.580494389971628e-06,
+      "loss": 0.152,
+      "step": 3916
+    },
+    {
+      "epoch": 0.6346403110823072,
+      "grad_norm": 0.8911001682281494,
+      "learning_rate": 4.580251882230045e-06,
+      "loss": 0.1523,
+      "step": 3917
+    },
+    {
+      "epoch": 0.6348023331173039,
+      "grad_norm": 0.9510197639465332,
+      "learning_rate": 4.580009310837789e-06,
+      "loss": 0.1451,
+      "step": 3918
+    },
+    {
+      "epoch": 0.6349643551523008,
+      "grad_norm": 0.9477640986442566,
+      "learning_rate": 4.579766675802281e-06,
+      "loss": 0.1485,
+      "step": 3919
+    },
+    {
+      "epoch": 0.6351263771872975,
+      "grad_norm": 0.9456864595413208,
+      "learning_rate": 4.579523977130946e-06,
+      "loss": 0.1489,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6352883992222942,
+      "grad_norm": 0.8579930663108826,
+      "learning_rate": 4.57928121483121e-06,
+      "loss": 0.1462,
+      "step": 3921
+    },
+    {
+      "epoch": 0.635450421257291,
+      "grad_norm": 0.8796806931495667,
+      "learning_rate": 4.579038388910499e-06,
+      "loss": 0.1349,
+      "step": 3922
+    },
+    {
+      "epoch": 0.6356124432922877,
+      "grad_norm": 0.9751520156860352,
+      "learning_rate": 4.578795499376246e-06,
+      "loss": 0.1583,
+      "step": 3923
+    },
+    {
+      "epoch": 0.6357744653272845,
+      "grad_norm": 0.9027172327041626,
+      "learning_rate": 4.578552546235882e-06,
+      "loss": 0.1483,
+      "step": 3924
+    },
+    {
+      "epoch": 0.6359364873622813,
+      "grad_norm": 0.9351487159729004,
+      "learning_rate": 4.578309529496839e-06,
+      "loss": 0.1398,
+      "step": 3925
+    },
+    {
+      "epoch": 0.6360985093972781,
+      "grad_norm": 0.99167400598526,
+      "learning_rate": 4.578066449166554e-06,
+      "loss": 0.172,
+      "step": 3926
+    },
+    {
+      "epoch": 0.6362605314322748,
+      "grad_norm": 1.0527704954147339,
+      "learning_rate": 4.577823305252464e-06,
+      "loss": 0.1422,
+      "step": 3927
+    },
+    {
+      "epoch": 0.6364225534672715,
+      "grad_norm": 0.8460220694541931,
+      "learning_rate": 4.57758009776201e-06,
+      "loss": 0.1374,
+      "step": 3928
+    },
+    {
+      "epoch": 0.6365845755022683,
+      "grad_norm": 0.8465033769607544,
+      "learning_rate": 4.577336826702631e-06,
+      "loss": 0.1408,
+      "step": 3929
+    },
+    {
+      "epoch": 0.636746597537265,
+      "grad_norm": 0.9382408261299133,
+      "learning_rate": 4.577093492081774e-06,
+      "loss": 0.1412,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6369086195722619,
+      "grad_norm": 0.9168792366981506,
+      "learning_rate": 4.576850093906881e-06,
+      "loss": 0.143,
+      "step": 3931
+    },
+    {
+      "epoch": 0.6370706416072586,
+      "grad_norm": 0.8806384205818176,
+      "learning_rate": 4.576606632185403e-06,
+      "loss": 0.1321,
+      "step": 3932
+    },
+    {
+      "epoch": 0.6372326636422554,
+      "grad_norm": 0.8602378964424133,
+      "learning_rate": 4.576363106924785e-06,
+      "loss": 0.122,
+      "step": 3933
+    },
+    {
+      "epoch": 0.6373946856772521,
+      "grad_norm": 0.9462723135948181,
+      "learning_rate": 4.576119518132483e-06,
+      "loss": 0.1421,
+      "step": 3934
+    },
+    {
+      "epoch": 0.6375567077122488,
+      "grad_norm": 0.9258326292037964,
+      "learning_rate": 4.5758758658159465e-06,
+      "loss": 0.1483,
+      "step": 3935
+    },
+    {
+      "epoch": 0.6377187297472456,
+      "grad_norm": 0.9369986653327942,
+      "learning_rate": 4.575632149982631e-06,
+      "loss": 0.1457,
+      "step": 3936
+    },
+    {
+      "epoch": 0.6378807517822424,
+      "grad_norm": 0.8889022469520569,
+      "learning_rate": 4.575388370639997e-06,
+      "loss": 0.1267,
+      "step": 3937
+    },
+    {
+      "epoch": 0.6380427738172392,
+      "grad_norm": 0.9667481780052185,
+      "learning_rate": 4.5751445277955e-06,
+      "loss": 0.1417,
+      "step": 3938
+    },
+    {
+      "epoch": 0.6382047958522359,
+      "grad_norm": 0.9078060388565063,
+      "learning_rate": 4.574900621456602e-06,
+      "loss": 0.1371,
+      "step": 3939
+    },
+    {
+      "epoch": 0.6383668178872327,
+      "grad_norm": 0.9612594246864319,
+      "learning_rate": 4.574656651630767e-06,
+      "loss": 0.1446,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6385288399222294,
+      "grad_norm": 0.9944157004356384,
+      "learning_rate": 4.574412618325458e-06,
+      "loss": 0.146,
+      "step": 3941
+    },
+    {
+      "epoch": 0.6386908619572261,
+      "grad_norm": 0.9619837999343872,
+      "learning_rate": 4.574168521548144e-06,
+      "loss": 0.1485,
+      "step": 3942
+    },
+    {
+      "epoch": 0.638852883992223,
+      "grad_norm": 1.0251742601394653,
+      "learning_rate": 4.5739243613062915e-06,
+      "loss": 0.1626,
+      "step": 3943
+    },
+    {
+      "epoch": 0.6390149060272197,
+      "grad_norm": 1.0261211395263672,
+      "learning_rate": 4.573680137607373e-06,
+      "loss": 0.1482,
+      "step": 3944
+    },
+    {
+      "epoch": 0.6391769280622165,
+      "grad_norm": 0.9001795053482056,
+      "learning_rate": 4.57343585045886e-06,
+      "loss": 0.1326,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6393389500972132,
+      "grad_norm": 1.0013598203659058,
+      "learning_rate": 4.573191499868228e-06,
+      "loss": 0.1509,
+      "step": 3946
+    },
+    {
+      "epoch": 0.63950097213221,
+      "grad_norm": 1.0281734466552734,
+      "learning_rate": 4.572947085842952e-06,
+      "loss": 0.1414,
+      "step": 3947
+    },
+    {
+      "epoch": 0.6396629941672067,
+      "grad_norm": 0.9392046928405762,
+      "learning_rate": 4.572702608390513e-06,
+      "loss": 0.1421,
+      "step": 3948
+    },
+    {
+      "epoch": 0.6398250162022034,
+      "grad_norm": 0.9891103506088257,
+      "learning_rate": 4.57245806751839e-06,
+      "loss": 0.1592,
+      "step": 3949
+    },
+    {
+      "epoch": 0.6399870382372003,
+      "grad_norm": 0.8271520137786865,
+      "learning_rate": 4.572213463234065e-06,
+      "loss": 0.1301,
+      "step": 3950
+    },
+    {
+      "epoch": 0.640149060272197,
+      "grad_norm": 0.8745349645614624,
+      "learning_rate": 4.571968795545023e-06,
+      "loss": 0.1386,
+      "step": 3951
+    },
+    {
+      "epoch": 0.6403110823071938,
+      "grad_norm": 0.9697176218032837,
+      "learning_rate": 4.5717240644587495e-06,
+      "loss": 0.1628,
+      "step": 3952
+    },
+    {
+      "epoch": 0.6404731043421905,
+      "grad_norm": 1.1860628128051758,
+      "learning_rate": 4.571479269982734e-06,
+      "loss": 0.1444,
+      "step": 3953
+    },
+    {
+      "epoch": 0.6406351263771873,
+      "grad_norm": 0.9648879766464233,
+      "learning_rate": 4.571234412124464e-06,
+      "loss": 0.1604,
+      "step": 3954
+    },
+    {
+      "epoch": 0.640797148412184,
+      "grad_norm": 0.9528911709785461,
+      "learning_rate": 4.570989490891434e-06,
+      "loss": 0.1569,
+      "step": 3955
+    },
+    {
+      "epoch": 0.6409591704471809,
+      "grad_norm": 0.8974754214286804,
+      "learning_rate": 4.570744506291138e-06,
+      "loss": 0.1485,
+      "step": 3956
+    },
+    {
+      "epoch": 0.6411211924821776,
+      "grad_norm": 0.8654069304466248,
+      "learning_rate": 4.570499458331071e-06,
+      "loss": 0.14,
+      "step": 3957
+    },
+    {
+      "epoch": 0.6412832145171743,
+      "grad_norm": 0.906300961971283,
+      "learning_rate": 4.570254347018731e-06,
+      "loss": 0.1508,
+      "step": 3958
+    },
+    {
+      "epoch": 0.6414452365521711,
+      "grad_norm": 1.011763572692871,
+      "learning_rate": 4.570009172361617e-06,
+      "loss": 0.1649,
+      "step": 3959
+    },
+    {
+      "epoch": 0.6416072585871678,
+      "grad_norm": 0.956451952457428,
+      "learning_rate": 4.5697639343672325e-06,
+      "loss": 0.1585,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6417692806221647,
+      "grad_norm": 0.8895651698112488,
+      "learning_rate": 4.569518633043081e-06,
+      "loss": 0.1425,
+      "step": 3961
+    },
+    {
+      "epoch": 0.6419313026571614,
+      "grad_norm": 0.922642171382904,
+      "learning_rate": 4.569273268396667e-06,
+      "loss": 0.1554,
+      "step": 3962
+    },
+    {
+      "epoch": 0.6420933246921582,
+      "grad_norm": 0.8361677527427673,
+      "learning_rate": 4.569027840435498e-06,
+      "loss": 0.1271,
+      "step": 3963
+    },
+    {
+      "epoch": 0.6422553467271549,
+      "grad_norm": 0.762519121170044,
+      "learning_rate": 4.568782349167084e-06,
+      "loss": 0.1148,
+      "step": 3964
+    },
+    {
+      "epoch": 0.6424173687621516,
+      "grad_norm": 1.0175034999847412,
+      "learning_rate": 4.568536794598937e-06,
+      "loss": 0.1408,
+      "step": 3965
+    },
+    {
+      "epoch": 0.6425793907971484,
+      "grad_norm": 1.0413769483566284,
+      "learning_rate": 4.56829117673857e-06,
+      "loss": 0.17,
+      "step": 3966
+    },
+    {
+      "epoch": 0.6427414128321451,
+      "grad_norm": 0.8488859534263611,
+      "learning_rate": 4.568045495593497e-06,
+      "loss": 0.1435,
+      "step": 3967
+    },
+    {
+      "epoch": 0.642903434867142,
+      "grad_norm": 0.8258789777755737,
+      "learning_rate": 4.567799751171237e-06,
+      "loss": 0.1261,
+      "step": 3968
+    },
+    {
+      "epoch": 0.6430654569021387,
+      "grad_norm": 0.8980733752250671,
+      "learning_rate": 4.567553943479309e-06,
+      "loss": 0.1372,
+      "step": 3969
+    },
+    {
+      "epoch": 0.6432274789371355,
+      "grad_norm": 0.8406203389167786,
+      "learning_rate": 4.567308072525233e-06,
+      "loss": 0.122,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6433895009721322,
+      "grad_norm": 0.9138239622116089,
+      "learning_rate": 4.567062138316534e-06,
+      "loss": 0.1483,
+      "step": 3971
+    },
+    {
+      "epoch": 0.6435515230071289,
+      "grad_norm": 0.9522116780281067,
+      "learning_rate": 4.566816140860735e-06,
+      "loss": 0.1437,
+      "step": 3972
+    },
+    {
+      "epoch": 0.6437135450421257,
+      "grad_norm": 0.8937114477157593,
+      "learning_rate": 4.566570080165363e-06,
+      "loss": 0.1242,
+      "step": 3973
+    },
+    {
+      "epoch": 0.6438755670771225,
+      "grad_norm": 0.9830919504165649,
+      "learning_rate": 4.566323956237948e-06,
+      "loss": 0.1407,
+      "step": 3974
+    },
+    {
+      "epoch": 0.6440375891121193,
+      "grad_norm": 1.0365149974822998,
+      "learning_rate": 4.566077769086022e-06,
+      "loss": 0.1556,
+      "step": 3975
+    },
+    {
+      "epoch": 0.644199611147116,
+      "grad_norm": 0.9378160238265991,
+      "learning_rate": 4.565831518717114e-06,
+      "loss": 0.1469,
+      "step": 3976
+    },
+    {
+      "epoch": 0.6443616331821128,
+      "grad_norm": 0.9110143184661865,
+      "learning_rate": 4.565585205138761e-06,
+      "loss": 0.1368,
+      "step": 3977
+    },
+    {
+      "epoch": 0.6445236552171095,
+      "grad_norm": 0.936230480670929,
+      "learning_rate": 4.5653388283585e-06,
+      "loss": 0.1328,
+      "step": 3978
+    },
+    {
+      "epoch": 0.6446856772521062,
+      "grad_norm": 1.0089045763015747,
+      "learning_rate": 4.565092388383869e-06,
+      "loss": 0.1427,
+      "step": 3979
+    },
+    {
+      "epoch": 0.6448476992871031,
+      "grad_norm": 0.9380981922149658,
+      "learning_rate": 4.564845885222407e-06,
+      "loss": 0.1466,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6450097213220998,
+      "grad_norm": 0.9660612344741821,
+      "learning_rate": 4.564599318881659e-06,
+      "loss": 0.1387,
+      "step": 3981
+    },
+    {
+      "epoch": 0.6451717433570966,
+      "grad_norm": 0.8815035223960876,
+      "learning_rate": 4.564352689369168e-06,
+      "loss": 0.1407,
+      "step": 3982
+    },
+    {
+      "epoch": 0.6453337653920933,
+      "grad_norm": 0.8399519324302673,
+      "learning_rate": 4.56410599669248e-06,
+      "loss": 0.1371,
+      "step": 3983
+    },
+    {
+      "epoch": 0.6454957874270901,
+      "grad_norm": 0.965580403804779,
+      "learning_rate": 4.563859240859144e-06,
+      "loss": 0.1474,
+      "step": 3984
+    },
+    {
+      "epoch": 0.6456578094620868,
+      "grad_norm": 0.9897104501724243,
+      "learning_rate": 4.5636124218767095e-06,
+      "loss": 0.1633,
+      "step": 3985
+    },
+    {
+      "epoch": 0.6458198314970836,
+      "grad_norm": 1.1038873195648193,
+      "learning_rate": 4.563365539752728e-06,
+      "loss": 0.1622,
+      "step": 3986
+    },
+    {
+      "epoch": 0.6459818535320804,
+      "grad_norm": 0.9341737031936646,
+      "learning_rate": 4.563118594494755e-06,
+      "loss": 0.1524,
+      "step": 3987
+    },
+    {
+      "epoch": 0.6461438755670771,
+      "grad_norm": 1.0048877000808716,
+      "learning_rate": 4.5628715861103455e-06,
+      "loss": 0.1549,
+      "step": 3988
+    },
+    {
+      "epoch": 0.6463058976020739,
+      "grad_norm": 0.8541507124900818,
+      "learning_rate": 4.562624514607058e-06,
+      "loss": 0.1366,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6464679196370706,
+      "grad_norm": 0.8617687821388245,
+      "learning_rate": 4.562377379992451e-06,
+      "loss": 0.1352,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6466299416720674,
+      "grad_norm": 0.8900352120399475,
+      "learning_rate": 4.5621301822740875e-06,
+      "loss": 0.1451,
+      "step": 3991
+    },
+    {
+      "epoch": 0.6467919637070642,
+      "grad_norm": 0.917232871055603,
+      "learning_rate": 4.56188292145953e-06,
+      "loss": 0.1397,
+      "step": 3992
+    },
+    {
+      "epoch": 0.6469539857420609,
+      "grad_norm": 0.8828876614570618,
+      "learning_rate": 4.5616355975563456e-06,
+      "loss": 0.1346,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6471160077770577,
+      "grad_norm": 1.0017690658569336,
+      "learning_rate": 4.561388210572101e-06,
+      "loss": 0.1667,
+      "step": 3994
+    },
+    {
+      "epoch": 0.6472780298120544,
+      "grad_norm": 0.9873506426811218,
+      "learning_rate": 4.561140760514365e-06,
+      "loss": 0.1538,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6474400518470512,
+      "grad_norm": 0.9482616782188416,
+      "learning_rate": 4.56089324739071e-06,
+      "loss": 0.1576,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6476020738820479,
+      "grad_norm": 0.8878632187843323,
+      "learning_rate": 4.560645671208709e-06,
+      "loss": 0.1395,
+      "step": 3997
+    },
+    {
+      "epoch": 0.6477640959170448,
+      "grad_norm": 1.057747483253479,
+      "learning_rate": 4.560398031975937e-06,
+      "loss": 0.1718,
+      "step": 3998
+    },
+    {
+      "epoch": 0.6479261179520415,
+      "grad_norm": 0.8953585028648376,
+      "learning_rate": 4.560150329699971e-06,
+      "loss": 0.1451,
+      "step": 3999
+    },
+    {
+      "epoch": 0.6480881399870383,
+      "grad_norm": 0.900346577167511,
+      "learning_rate": 4.55990256438839e-06,
+      "loss": 0.1447,
+      "step": 4000
+    },
+    {
+      "epoch": 0.648250162022035,
+      "grad_norm": 0.8216709494590759,
+      "learning_rate": 4.559654736048776e-06,
+      "loss": 0.1327,
+      "step": 4001
+    },
+    {
+      "epoch": 0.6484121840570317,
+      "grad_norm": 0.850792646408081,
+      "learning_rate": 4.559406844688711e-06,
+      "loss": 0.1341,
+      "step": 4002
+    },
+    {
+      "epoch": 0.6485742060920285,
+      "grad_norm": 0.9442715644836426,
+      "learning_rate": 4.5591588903157816e-06,
+      "loss": 0.1442,
+      "step": 4003
+    },
+    {
+      "epoch": 0.6487362281270252,
+      "grad_norm": 0.8908452391624451,
+      "learning_rate": 4.558910872937572e-06,
+      "loss": 0.1427,
+      "step": 4004
+    },
+    {
+      "epoch": 0.6488982501620221,
+      "grad_norm": 0.9406809210777283,
+      "learning_rate": 4.558662792561672e-06,
+      "loss": 0.1616,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6490602721970188,
+      "grad_norm": 0.8570629954338074,
+      "learning_rate": 4.558414649195673e-06,
+      "loss": 0.1443,
+      "step": 4006
+    },
+    {
+      "epoch": 0.6492222942320156,
+      "grad_norm": 0.8232815861701965,
+      "learning_rate": 4.558166442847166e-06,
+      "loss": 0.1351,
+      "step": 4007
+    },
+    {
+      "epoch": 0.6493843162670123,
+      "grad_norm": 0.9287505745887756,
+      "learning_rate": 4.557918173523747e-06,
+      "loss": 0.1492,
+      "step": 4008
+    },
+    {
+      "epoch": 0.649546338302009,
+      "grad_norm": 0.9767974615097046,
+      "learning_rate": 4.557669841233013e-06,
+      "loss": 0.1583,
+      "step": 4009
+    },
+    {
+      "epoch": 0.6497083603370059,
+      "grad_norm": 0.8061318397521973,
+      "learning_rate": 4.55742144598256e-06,
+      "loss": 0.1257,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6498703823720026,
+      "grad_norm": 0.8351037502288818,
+      "learning_rate": 4.557172987779991e-06,
+      "loss": 0.143,
+      "step": 4011
+    },
+    {
+      "epoch": 0.6500324044069994,
+      "grad_norm": 0.8559525609016418,
+      "learning_rate": 4.5569244666329055e-06,
+      "loss": 0.1358,
+      "step": 4012
+    },
+    {
+      "epoch": 0.6501944264419961,
+      "grad_norm": 0.9949284195899963,
+      "learning_rate": 4.556675882548909e-06,
+      "loss": 0.1463,
+      "step": 4013
+    },
+    {
+      "epoch": 0.6503564484769929,
+      "grad_norm": 0.8767020106315613,
+      "learning_rate": 4.5564272355356085e-06,
+      "loss": 0.1366,
+      "step": 4014
+    },
+    {
+      "epoch": 0.6505184705119896,
+      "grad_norm": 0.9822929501533508,
+      "learning_rate": 4.556178525600611e-06,
+      "loss": 0.1454,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6506804925469863,
+      "grad_norm": 0.8953209519386292,
+      "learning_rate": 4.555929752751526e-06,
+      "loss": 0.1393,
+      "step": 4016
+    },
+    {
+      "epoch": 0.6508425145819832,
+      "grad_norm": 0.9238772988319397,
+      "learning_rate": 4.555680916995965e-06,
+      "loss": 0.15,
+      "step": 4017
+    },
+    {
+      "epoch": 0.6510045366169799,
+      "grad_norm": 0.8017407655715942,
+      "learning_rate": 4.5554320183415435e-06,
+      "loss": 0.1256,
+      "step": 4018
+    },
+    {
+      "epoch": 0.6511665586519767,
+      "grad_norm": 0.9731351137161255,
+      "learning_rate": 4.555183056795877e-06,
+      "loss": 0.1371,
+      "step": 4019
+    },
+    {
+      "epoch": 0.6513285806869734,
+      "grad_norm": 0.8913125991821289,
+      "learning_rate": 4.5549340323665815e-06,
+      "loss": 0.1392,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6514906027219702,
+      "grad_norm": 0.9241263270378113,
+      "learning_rate": 4.5546849450612774e-06,
+      "loss": 0.1483,
+      "step": 4021
+    },
+    {
+      "epoch": 0.651652624756967,
+      "grad_norm": 0.925363302230835,
+      "learning_rate": 4.554435794887586e-06,
+      "loss": 0.1376,
+      "step": 4022
+    },
+    {
+      "epoch": 0.6518146467919637,
+      "grad_norm": 1.0443159341812134,
+      "learning_rate": 4.5541865818531315e-06,
+      "loss": 0.1505,
+      "step": 4023
+    },
+    {
+      "epoch": 0.6519766688269605,
+      "grad_norm": 0.9155001044273376,
+      "learning_rate": 4.553937305965539e-06,
+      "loss": 0.1499,
+      "step": 4024
+    },
+    {
+      "epoch": 0.6521386908619572,
+      "grad_norm": 0.920136034488678,
+      "learning_rate": 4.5536879672324345e-06,
+      "loss": 0.1312,
+      "step": 4025
+    },
+    {
+      "epoch": 0.652300712896954,
+      "grad_norm": 0.8369788527488708,
+      "learning_rate": 4.553438565661448e-06,
+      "loss": 0.1225,
+      "step": 4026
+    },
+    {
+      "epoch": 0.6524627349319507,
+      "grad_norm": 0.9391274452209473,
+      "learning_rate": 4.553189101260211e-06,
+      "loss": 0.1424,
+      "step": 4027
+    },
+    {
+      "epoch": 0.6526247569669476,
+      "grad_norm": 0.9724919199943542,
+      "learning_rate": 4.552939574036356e-06,
+      "loss": 0.1591,
+      "step": 4028
+    },
+    {
+      "epoch": 0.6527867790019443,
+      "grad_norm": 0.8575350642204285,
+      "learning_rate": 4.552689983997519e-06,
+      "loss": 0.1452,
+      "step": 4029
+    },
+    {
+      "epoch": 0.652948801036941,
+      "grad_norm": 0.9005478620529175,
+      "learning_rate": 4.552440331151334e-06,
+      "loss": 0.152,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6531108230719378,
+      "grad_norm": 0.9679160118103027,
+      "learning_rate": 4.552190615505444e-06,
+      "loss": 0.1539,
+      "step": 4031
+    },
+    {
+      "epoch": 0.6532728451069345,
+      "grad_norm": 0.8400557041168213,
+      "learning_rate": 4.551940837067486e-06,
+      "loss": 0.1406,
+      "step": 4032
+    },
+    {
+      "epoch": 0.6534348671419313,
+      "grad_norm": 0.9820280075073242,
+      "learning_rate": 4.551690995845104e-06,
+      "loss": 0.1618,
+      "step": 4033
+    },
+    {
+      "epoch": 0.653596889176928,
+      "grad_norm": 0.862453818321228,
+      "learning_rate": 4.551441091845942e-06,
+      "loss": 0.1438,
+      "step": 4034
+    },
+    {
+      "epoch": 0.6537589112119249,
+      "grad_norm": 0.8491960167884827,
+      "learning_rate": 4.551191125077647e-06,
+      "loss": 0.1303,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6539209332469216,
+      "grad_norm": 0.8754890561103821,
+      "learning_rate": 4.550941095547869e-06,
+      "loss": 0.1446,
+      "step": 4036
+    },
+    {
+      "epoch": 0.6540829552819183,
+      "grad_norm": 0.9268251657485962,
+      "learning_rate": 4.550691003264256e-06,
+      "loss": 0.1556,
+      "step": 4037
+    },
+    {
+      "epoch": 0.6542449773169151,
+      "grad_norm": 0.8505957126617432,
+      "learning_rate": 4.55044084823446e-06,
+      "loss": 0.1343,
+      "step": 4038
+    },
+    {
+      "epoch": 0.6544069993519118,
+      "grad_norm": 0.8797240853309631,
+      "learning_rate": 4.550190630466137e-06,
+      "loss": 0.1445,
+      "step": 4039
+    },
+    {
+      "epoch": 0.6545690213869086,
+      "grad_norm": 0.8217586278915405,
+      "learning_rate": 4.5499403499669415e-06,
+      "loss": 0.1309,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6547310434219054,
+      "grad_norm": 1.0774375200271606,
+      "learning_rate": 4.549690006744531e-06,
+      "loss": 0.1524,
+      "step": 4041
+    },
+    {
+      "epoch": 0.6548930654569022,
+      "grad_norm": 0.9880624413490295,
+      "learning_rate": 4.549439600806568e-06,
+      "loss": 0.1523,
+      "step": 4042
+    },
+    {
+      "epoch": 0.6550550874918989,
+      "grad_norm": 0.9190874695777893,
+      "learning_rate": 4.549189132160713e-06,
+      "loss": 0.133,
+      "step": 4043
+    },
+    {
+      "epoch": 0.6552171095268956,
+      "grad_norm": 0.9123296141624451,
+      "learning_rate": 4.548938600814629e-06,
+      "loss": 0.1377,
+      "step": 4044
+    },
+    {
+      "epoch": 0.6553791315618924,
+      "grad_norm": 1.0104409456253052,
+      "learning_rate": 4.548688006775981e-06,
+      "loss": 0.1488,
+      "step": 4045
+    },
+    {
+      "epoch": 0.6555411535968891,
+      "grad_norm": 0.955862283706665,
+      "learning_rate": 4.5484373500524395e-06,
+      "loss": 0.1588,
+      "step": 4046
+    },
+    {
+      "epoch": 0.655703175631886,
+      "grad_norm": 0.7813193202018738,
+      "learning_rate": 4.548186630651671e-06,
+      "loss": 0.1188,
+      "step": 4047
+    },
+    {
+      "epoch": 0.6558651976668827,
+      "grad_norm": 1.0729082822799683,
+      "learning_rate": 4.547935848581349e-06,
+      "loss": 0.178,
+      "step": 4048
+    },
+    {
+      "epoch": 0.6560272197018795,
+      "grad_norm": 0.8658989071846008,
+      "learning_rate": 4.547685003849145e-06,
+      "loss": 0.1391,
+      "step": 4049
+    },
+    {
+      "epoch": 0.6561892417368762,
+      "grad_norm": 1.0283029079437256,
+      "learning_rate": 4.5474340964627365e-06,
+      "loss": 0.1565,
+      "step": 4050
+    },
+    {
+      "epoch": 0.656351263771873,
+      "grad_norm": 0.9804076552391052,
+      "learning_rate": 4.547183126429798e-06,
+      "loss": 0.1667,
+      "step": 4051
+    },
+    {
+      "epoch": 0.6565132858068697,
+      "grad_norm": 0.7968271374702454,
+      "learning_rate": 4.5469320937580105e-06,
+      "loss": 0.1288,
+      "step": 4052
+    },
+    {
+      "epoch": 0.6566753078418665,
+      "grad_norm": 0.9019933938980103,
+      "learning_rate": 4.546680998455054e-06,
+      "loss": 0.1439,
+      "step": 4053
+    },
+    {
+      "epoch": 0.6568373298768633,
+      "grad_norm": 0.8464137315750122,
+      "learning_rate": 4.546429840528612e-06,
+      "loss": 0.1405,
+      "step": 4054
+    },
+    {
+      "epoch": 0.65699935191186,
+      "grad_norm": 0.8122446537017822,
+      "learning_rate": 4.54617861998637e-06,
+      "loss": 0.1338,
+      "step": 4055
+    },
+    {
+      "epoch": 0.6571613739468568,
+      "grad_norm": 0.9290385246276855,
+      "learning_rate": 4.545927336836013e-06,
+      "loss": 0.1323,
+      "step": 4056
+    },
+    {
+      "epoch": 0.6573233959818535,
+      "grad_norm": 0.9269915223121643,
+      "learning_rate": 4.545675991085231e-06,
+      "loss": 0.1515,
+      "step": 4057
+    },
+    {
+      "epoch": 0.6574854180168503,
+      "grad_norm": 1.0693535804748535,
+      "learning_rate": 4.545424582741714e-06,
+      "loss": 0.1695,
+      "step": 4058
+    },
+    {
+      "epoch": 0.657647440051847,
+      "grad_norm": 0.9048995971679688,
+      "learning_rate": 4.545173111813154e-06,
+      "loss": 0.1444,
+      "step": 4059
+    },
+    {
+      "epoch": 0.6578094620868438,
+      "grad_norm": 0.8983332514762878,
+      "learning_rate": 4.544921578307246e-06,
+      "loss": 0.1323,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6579714841218406,
+      "grad_norm": 1.1153713464736938,
+      "learning_rate": 4.544669982231688e-06,
+      "loss": 0.1742,
+      "step": 4061
+    },
+    {
+      "epoch": 0.6581335061568373,
+      "grad_norm": 0.9735933542251587,
+      "learning_rate": 4.544418323594175e-06,
+      "loss": 0.1485,
+      "step": 4062
+    },
+    {
+      "epoch": 0.6582955281918341,
+      "grad_norm": 0.9374943375587463,
+      "learning_rate": 4.544166602402409e-06,
+      "loss": 0.144,
+      "step": 4063
+    },
+    {
+      "epoch": 0.6584575502268308,
+      "grad_norm": 0.9201976656913757,
+      "learning_rate": 4.543914818664092e-06,
+      "loss": 0.1497,
+      "step": 4064
+    },
+    {
+      "epoch": 0.6586195722618277,
+      "grad_norm": 1.0155818462371826,
+      "learning_rate": 4.543662972386927e-06,
+      "loss": 0.1638,
+      "step": 4065
+    },
+    {
+      "epoch": 0.6587815942968244,
+      "grad_norm": 0.913718044757843,
+      "learning_rate": 4.543411063578621e-06,
+      "loss": 0.1466,
+      "step": 4066
+    },
+    {
+      "epoch": 0.6589436163318211,
+      "grad_norm": 0.813508927822113,
+      "learning_rate": 4.5431590922468815e-06,
+      "loss": 0.1324,
+      "step": 4067
+    },
+    {
+      "epoch": 0.6591056383668179,
+      "grad_norm": 0.7492986917495728,
+      "learning_rate": 4.5429070583994185e-06,
+      "loss": 0.1263,
+      "step": 4068
+    },
+    {
+      "epoch": 0.6592676604018146,
+      "grad_norm": 0.8122268915176392,
+      "learning_rate": 4.542654962043943e-06,
+      "loss": 0.1274,
+      "step": 4069
+    },
+    {
+      "epoch": 0.6594296824368114,
+      "grad_norm": 0.8873049020767212,
+      "learning_rate": 4.542402803188168e-06,
+      "loss": 0.1413,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6595917044718081,
+      "grad_norm": 0.9349288940429688,
+      "learning_rate": 4.542150581839811e-06,
+      "loss": 0.1534,
+      "step": 4071
+    },
+    {
+      "epoch": 0.659753726506805,
+      "grad_norm": 0.8754650354385376,
+      "learning_rate": 4.5418982980065874e-06,
+      "loss": 0.1503,
+      "step": 4072
+    },
+    {
+      "epoch": 0.6599157485418017,
+      "grad_norm": 0.8489086627960205,
+      "learning_rate": 4.541645951696217e-06,
+      "loss": 0.1353,
+      "step": 4073
+    },
+    {
+      "epoch": 0.6600777705767984,
+      "grad_norm": 0.8746541142463684,
+      "learning_rate": 4.541393542916423e-06,
+      "loss": 0.1541,
+      "step": 4074
+    },
+    {
+      "epoch": 0.6602397926117952,
+      "grad_norm": 0.9938592910766602,
+      "learning_rate": 4.541141071674924e-06,
+      "loss": 0.1663,
+      "step": 4075
+    },
+    {
+      "epoch": 0.6604018146467919,
+      "grad_norm": 0.9091808199882507,
+      "learning_rate": 4.540888537979449e-06,
+      "loss": 0.1333,
+      "step": 4076
+    },
+    {
+      "epoch": 0.6605638366817888,
+      "grad_norm": 0.8329112529754639,
+      "learning_rate": 4.540635941837723e-06,
+      "loss": 0.1314,
+      "step": 4077
+    },
+    {
+      "epoch": 0.6607258587167855,
+      "grad_norm": 0.8982328176498413,
+      "learning_rate": 4.540383283257477e-06,
+      "loss": 0.1536,
+      "step": 4078
+    },
+    {
+      "epoch": 0.6608878807517823,
+      "grad_norm": 0.9931498169898987,
+      "learning_rate": 4.540130562246439e-06,
+      "loss": 0.1613,
+      "step": 4079
+    },
+    {
+      "epoch": 0.661049902786779,
+      "grad_norm": 0.9610978364944458,
+      "learning_rate": 4.539877778812342e-06,
+      "loss": 0.1464,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6612119248217757,
+      "grad_norm": 0.8849239349365234,
+      "learning_rate": 4.539624932962923e-06,
+      "loss": 0.1327,
+      "step": 4081
+    },
+    {
+      "epoch": 0.6613739468567725,
+      "grad_norm": 0.9079961776733398,
+      "learning_rate": 4.539372024705916e-06,
+      "loss": 0.1298,
+      "step": 4082
+    },
+    {
+      "epoch": 0.6615359688917692,
+      "grad_norm": 0.8636612296104431,
+      "learning_rate": 4.5391190540490595e-06,
+      "loss": 0.1354,
+      "step": 4083
+    },
+    {
+      "epoch": 0.6616979909267661,
+      "grad_norm": 1.0267540216445923,
+      "learning_rate": 4.538866021000096e-06,
+      "loss": 0.1476,
+      "step": 4084
+    },
+    {
+      "epoch": 0.6618600129617628,
+      "grad_norm": 0.9484116435050964,
+      "learning_rate": 4.538612925566765e-06,
+      "loss": 0.1392,
+      "step": 4085
+    },
+    {
+      "epoch": 0.6620220349967596,
+      "grad_norm": 0.9751802086830139,
+      "learning_rate": 4.538359767756813e-06,
+      "loss": 0.1462,
+      "step": 4086
+    },
+    {
+      "epoch": 0.6621840570317563,
+      "grad_norm": 0.9586076140403748,
+      "learning_rate": 4.538106547577984e-06,
+      "loss": 0.1576,
+      "step": 4087
+    },
+    {
+      "epoch": 0.662346079066753,
+      "grad_norm": 0.9770367741584778,
+      "learning_rate": 4.537853265038027e-06,
+      "loss": 0.1509,
+      "step": 4088
+    },
+    {
+      "epoch": 0.6625081011017498,
+      "grad_norm": 0.9630382657051086,
+      "learning_rate": 4.537599920144692e-06,
+      "loss": 0.1464,
+      "step": 4089
+    },
+    {
+      "epoch": 0.6626701231367466,
+      "grad_norm": 0.9468212127685547,
+      "learning_rate": 4.537346512905729e-06,
+      "loss": 0.1563,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6628321451717434,
+      "grad_norm": 1.0076836347579956,
+      "learning_rate": 4.537093043328894e-06,
+      "loss": 0.1756,
+      "step": 4091
+    },
+    {
+      "epoch": 0.6629941672067401,
+      "grad_norm": 0.879906415939331,
+      "learning_rate": 4.536839511421941e-06,
+      "loss": 0.1328,
+      "step": 4092
+    },
+    {
+      "epoch": 0.6631561892417369,
+      "grad_norm": 0.8465386629104614,
+      "learning_rate": 4.536585917192629e-06,
+      "loss": 0.1299,
+      "step": 4093
+    },
+    {
+      "epoch": 0.6633182112767336,
+      "grad_norm": 0.960082471370697,
+      "learning_rate": 4.536332260648716e-06,
+      "loss": 0.1401,
+      "step": 4094
+    },
+    {
+      "epoch": 0.6634802333117304,
+      "grad_norm": 0.8401903510093689,
+      "learning_rate": 4.536078541797964e-06,
+      "loss": 0.1229,
+      "step": 4095
+    },
+    {
+      "epoch": 0.6636422553467272,
+      "grad_norm": 1.0268646478652954,
+      "learning_rate": 4.535824760648135e-06,
+      "loss": 0.1713,
+      "step": 4096
+    },
+    {
+      "epoch": 0.6638042773817239,
+      "grad_norm": 0.8968318104743958,
+      "learning_rate": 4.535570917206995e-06,
+      "loss": 0.1454,
+      "step": 4097
+    },
+    {
+      "epoch": 0.6639662994167207,
+      "grad_norm": 0.9677649736404419,
+      "learning_rate": 4.535317011482311e-06,
+      "loss": 0.1556,
+      "step": 4098
+    },
+    {
+      "epoch": 0.6641283214517174,
+      "grad_norm": 0.9355975389480591,
+      "learning_rate": 4.535063043481852e-06,
+      "loss": 0.1549,
+      "step": 4099
+    },
+    {
+      "epoch": 0.6642903434867142,
+      "grad_norm": 0.8558212518692017,
+      "learning_rate": 4.534809013213389e-06,
+      "loss": 0.1448,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6644523655217109,
+      "grad_norm": 0.8515587449073792,
+      "learning_rate": 4.534554920684694e-06,
+      "loss": 0.1419,
+      "step": 4101
+    },
+    {
+      "epoch": 0.6646143875567078,
+      "grad_norm": 0.807841956615448,
+      "learning_rate": 4.534300765903542e-06,
+      "loss": 0.1292,
+      "step": 4102
+    },
+    {
+      "epoch": 0.6647764095917045,
+      "grad_norm": 0.8981359601020813,
+      "learning_rate": 4.534046548877709e-06,
+      "loss": 0.1352,
+      "step": 4103
+    },
+    {
+      "epoch": 0.6649384316267012,
+      "grad_norm": 0.893144965171814,
+      "learning_rate": 4.533792269614974e-06,
+      "loss": 0.1303,
+      "step": 4104
+    },
+    {
+      "epoch": 0.665100453661698,
+      "grad_norm": 0.9923272728919983,
+      "learning_rate": 4.533537928123118e-06,
+      "loss": 0.1647,
+      "step": 4105
+    },
+    {
+      "epoch": 0.6652624756966947,
+      "grad_norm": 0.8646336793899536,
+      "learning_rate": 4.533283524409922e-06,
+      "loss": 0.1513,
+      "step": 4106
+    },
+    {
+      "epoch": 0.6654244977316915,
+      "grad_norm": 0.952130138874054,
+      "learning_rate": 4.53302905848317e-06,
+      "loss": 0.1633,
+      "step": 4107
+    },
+    {
+      "epoch": 0.6655865197666883,
+      "grad_norm": 0.8783090710639954,
+      "learning_rate": 4.53277453035065e-06,
+      "loss": 0.1316,
+      "step": 4108
+    },
+    {
+      "epoch": 0.6657485418016851,
+      "grad_norm": 0.8041051626205444,
+      "learning_rate": 4.532519940020148e-06,
+      "loss": 0.1197,
+      "step": 4109
+    },
+    {
+      "epoch": 0.6659105638366818,
+      "grad_norm": 0.9836557507514954,
+      "learning_rate": 4.532265287499454e-06,
+      "loss": 0.1687,
+      "step": 4110
+    },
+    {
+      "epoch": 0.6660725858716785,
+      "grad_norm": 0.8173419833183289,
+      "learning_rate": 4.532010572796361e-06,
+      "loss": 0.1313,
+      "step": 4111
+    },
+    {
+      "epoch": 0.6662346079066753,
+      "grad_norm": 0.8361591696739197,
+      "learning_rate": 4.531755795918661e-06,
+      "loss": 0.128,
+      "step": 4112
+    },
+    {
+      "epoch": 0.666396629941672,
+      "grad_norm": 0.9987451434135437,
+      "learning_rate": 4.531500956874151e-06,
+      "loss": 0.1556,
+      "step": 4113
+    },
+    {
+      "epoch": 0.6665586519766689,
+      "grad_norm": 0.9050056338310242,
+      "learning_rate": 4.531246055670627e-06,
+      "loss": 0.1359,
+      "step": 4114
+    },
+    {
+      "epoch": 0.6667206740116656,
+      "grad_norm": 0.8897657990455627,
+      "learning_rate": 4.53099109231589e-06,
+      "loss": 0.1415,
+      "step": 4115
+    },
+    {
+      "epoch": 0.6668826960466624,
+      "grad_norm": 1.0086582899093628,
+      "learning_rate": 4.53073606681774e-06,
+      "loss": 0.1757,
+      "step": 4116
+    },
+    {
+      "epoch": 0.6670447180816591,
+      "grad_norm": 0.9577541351318359,
+      "learning_rate": 4.530480979183981e-06,
+      "loss": 0.1415,
+      "step": 4117
+    },
+    {
+      "epoch": 0.6672067401166558,
+      "grad_norm": 0.8575126528739929,
+      "learning_rate": 4.530225829422418e-06,
+      "loss": 0.1258,
+      "step": 4118
+    },
+    {
+      "epoch": 0.6673687621516526,
+      "grad_norm": 0.9692230224609375,
+      "learning_rate": 4.529970617540857e-06,
+      "loss": 0.1456,
+      "step": 4119
+    },
+    {
+      "epoch": 0.6675307841866494,
+      "grad_norm": 0.8811309337615967,
+      "learning_rate": 4.529715343547107e-06,
+      "loss": 0.1295,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6676928062216462,
+      "grad_norm": 0.9429895281791687,
+      "learning_rate": 4.529460007448981e-06,
+      "loss": 0.1667,
+      "step": 4121
+    },
+    {
+      "epoch": 0.6678548282566429,
+      "grad_norm": 0.9843415021896362,
+      "learning_rate": 4.5292046092542885e-06,
+      "loss": 0.1663,
+      "step": 4122
+    },
+    {
+      "epoch": 0.6680168502916397,
+      "grad_norm": 0.8283978700637817,
+      "learning_rate": 4.528949148970846e-06,
+      "loss": 0.1327,
+      "step": 4123
+    },
+    {
+      "epoch": 0.6681788723266364,
+      "grad_norm": 0.9020947217941284,
+      "learning_rate": 4.52869362660647e-06,
+      "loss": 0.1461,
+      "step": 4124
+    },
+    {
+      "epoch": 0.6683408943616331,
+      "grad_norm": 0.8452586531639099,
+      "learning_rate": 4.528438042168978e-06,
+      "loss": 0.145,
+      "step": 4125
+    },
+    {
+      "epoch": 0.66850291639663,
+      "grad_norm": 0.914142370223999,
+      "learning_rate": 4.5281823956661905e-06,
+      "loss": 0.1416,
+      "step": 4126
+    },
+    {
+      "epoch": 0.6686649384316267,
+      "grad_norm": 0.9229021072387695,
+      "learning_rate": 4.52792668710593e-06,
+      "loss": 0.1534,
+      "step": 4127
+    },
+    {
+      "epoch": 0.6688269604666235,
+      "grad_norm": 0.8286415338516235,
+      "learning_rate": 4.527670916496021e-06,
+      "loss": 0.1303,
+      "step": 4128
+    },
+    {
+      "epoch": 0.6689889825016202,
+      "grad_norm": 1.0366954803466797,
+      "learning_rate": 4.5274150838442875e-06,
+      "loss": 0.1616,
+      "step": 4129
+    },
+    {
+      "epoch": 0.669151004536617,
+      "grad_norm": 0.886921226978302,
+      "learning_rate": 4.52715918915856e-06,
+      "loss": 0.1416,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6693130265716137,
+      "grad_norm": 0.9015594124794006,
+      "learning_rate": 4.5269032324466656e-06,
+      "loss": 0.1492,
+      "step": 4131
+    },
+    {
+      "epoch": 0.6694750486066104,
+      "grad_norm": 0.8952576518058777,
+      "learning_rate": 4.526647213716438e-06,
+      "loss": 0.1392,
+      "step": 4132
+    },
+    {
+      "epoch": 0.6696370706416073,
+      "grad_norm": 0.8199195861816406,
+      "learning_rate": 4.526391132975711e-06,
+      "loss": 0.1342,
+      "step": 4133
+    },
+    {
+      "epoch": 0.669799092676604,
+      "grad_norm": 0.822807252407074,
+      "learning_rate": 4.526134990232317e-06,
+      "loss": 0.1377,
+      "step": 4134
+    },
+    {
+      "epoch": 0.6699611147116008,
+      "grad_norm": 0.8194425702095032,
+      "learning_rate": 4.525878785494097e-06,
+      "loss": 0.1341,
+      "step": 4135
+    },
+    {
+      "epoch": 0.6701231367465975,
+      "grad_norm": 1.0225645303726196,
+      "learning_rate": 4.525622518768888e-06,
+      "loss": 0.1631,
+      "step": 4136
+    },
+    {
+      "epoch": 0.6702851587815943,
+      "grad_norm": 0.8572471737861633,
+      "learning_rate": 4.5253661900645315e-06,
+      "loss": 0.1328,
+      "step": 4137
+    },
+    {
+      "epoch": 0.670447180816591,
+      "grad_norm": 0.9486353397369385,
+      "learning_rate": 4.5251097993888726e-06,
+      "loss": 0.1543,
+      "step": 4138
+    },
+    {
+      "epoch": 0.6706092028515879,
+      "grad_norm": 0.9721882343292236,
+      "learning_rate": 4.524853346749753e-06,
+      "loss": 0.1558,
+      "step": 4139
+    },
+    {
+      "epoch": 0.6707712248865846,
+      "grad_norm": 0.8730854988098145,
+      "learning_rate": 4.524596832155022e-06,
+      "loss": 0.1354,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6709332469215813,
+      "grad_norm": 0.8489865660667419,
+      "learning_rate": 4.524340255612526e-06,
+      "loss": 0.1314,
+      "step": 4141
+    },
+    {
+      "epoch": 0.6710952689565781,
+      "grad_norm": 0.8910239934921265,
+      "learning_rate": 4.524083617130118e-06,
+      "loss": 0.1545,
+      "step": 4142
+    },
+    {
+      "epoch": 0.6712572909915748,
+      "grad_norm": 0.7663037776947021,
+      "learning_rate": 4.523826916715649e-06,
+      "loss": 0.1255,
+      "step": 4143
+    },
+    {
+      "epoch": 0.6714193130265717,
+      "grad_norm": 0.9194484353065491,
+      "learning_rate": 4.523570154376975e-06,
+      "loss": 0.1457,
+      "step": 4144
+    },
+    {
+      "epoch": 0.6715813350615684,
+      "grad_norm": 0.9784338474273682,
+      "learning_rate": 4.52331333012195e-06,
+      "loss": 0.1645,
+      "step": 4145
+    },
+    {
+      "epoch": 0.6717433570965652,
+      "grad_norm": 0.8883211612701416,
+      "learning_rate": 4.5230564439584335e-06,
+      "loss": 0.1359,
+      "step": 4146
+    },
+    {
+      "epoch": 0.6719053791315619,
+      "grad_norm": 0.9348551034927368,
+      "learning_rate": 4.522799495894286e-06,
+      "loss": 0.1549,
+      "step": 4147
+    },
+    {
+      "epoch": 0.6720674011665586,
+      "grad_norm": 0.76348876953125,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.116,
+      "step": 4148
+    },
+    {
+      "epoch": 0.6722294232015554,
+      "grad_norm": 0.8616440296173096,
+      "learning_rate": 4.522285414095547e-06,
+      "loss": 0.1296,
+      "step": 4149
+    },
+    {
+      "epoch": 0.6723914452365521,
+      "grad_norm": 0.901292622089386,
+      "learning_rate": 4.522028280376683e-06,
+      "loss": 0.1495,
+      "step": 4150
+    },
+    {
+      "epoch": 0.672553467271549,
+      "grad_norm": 0.8072662353515625,
+      "learning_rate": 4.521771084788649e-06,
+      "loss": 0.1316,
+      "step": 4151
+    },
+    {
+      "epoch": 0.6727154893065457,
+      "grad_norm": 0.9773443937301636,
+      "learning_rate": 4.521513827339311e-06,
+      "loss": 0.164,
+      "step": 4152
+    },
+    {
+      "epoch": 0.6728775113415425,
+      "grad_norm": 0.8923728466033936,
+      "learning_rate": 4.521256508036543e-06,
+      "loss": 0.1437,
+      "step": 4153
+    },
+    {
+      "epoch": 0.6730395333765392,
+      "grad_norm": 1.0206551551818848,
+      "learning_rate": 4.5209991268882165e-06,
+      "loss": 0.1608,
+      "step": 4154
+    },
+    {
+      "epoch": 0.6732015554115359,
+      "grad_norm": 0.9942020177841187,
+      "learning_rate": 4.520741683902208e-06,
+      "loss": 0.1553,
+      "step": 4155
+    },
+    {
+      "epoch": 0.6733635774465327,
+      "grad_norm": 0.9606051445007324,
+      "learning_rate": 4.520484179086394e-06,
+      "loss": 0.1577,
+      "step": 4156
+    },
+    {
+      "epoch": 0.6735255994815295,
+      "grad_norm": 0.9637233018875122,
+      "learning_rate": 4.520226612448653e-06,
+      "loss": 0.1646,
+      "step": 4157
+    },
+    {
+      "epoch": 0.6736876215165263,
+      "grad_norm": 0.8612246513366699,
+      "learning_rate": 4.519968983996867e-06,
+      "loss": 0.1358,
+      "step": 4158
+    },
+    {
+      "epoch": 0.673849643551523,
+      "grad_norm": 0.9165799617767334,
+      "learning_rate": 4.519711293738918e-06,
+      "loss": 0.154,
+      "step": 4159
+    },
+    {
+      "epoch": 0.6740116655865198,
+      "grad_norm": 0.853792130947113,
+      "learning_rate": 4.519453541682691e-06,
+      "loss": 0.1378,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6741736876215165,
+      "grad_norm": 0.9091629385948181,
+      "learning_rate": 4.519195727836073e-06,
+      "loss": 0.1311,
+      "step": 4161
+    },
+    {
+      "epoch": 0.6743357096565132,
+      "grad_norm": 1.033930778503418,
+      "learning_rate": 4.518937852206952e-06,
+      "loss": 0.1692,
+      "step": 4162
+    },
+    {
+      "epoch": 0.6744977316915101,
+      "grad_norm": 0.9184512495994568,
+      "learning_rate": 4.518679914803218e-06,
+      "loss": 0.1417,
+      "step": 4163
+    },
+    {
+      "epoch": 0.6746597537265068,
+      "grad_norm": 0.8692227602005005,
+      "learning_rate": 4.518421915632764e-06,
+      "loss": 0.1491,
+      "step": 4164
+    },
+    {
+      "epoch": 0.6748217757615036,
+      "grad_norm": 0.817550778388977,
+      "learning_rate": 4.518163854703484e-06,
+      "loss": 0.1407,
+      "step": 4165
+    },
+    {
+      "epoch": 0.6749837977965003,
+      "grad_norm": 0.8003614544868469,
+      "learning_rate": 4.5179057320232735e-06,
+      "loss": 0.126,
+      "step": 4166
+    },
+    {
+      "epoch": 0.6751458198314971,
+      "grad_norm": 0.847512423992157,
+      "learning_rate": 4.517647547600032e-06,
+      "loss": 0.1203,
+      "step": 4167
+    },
+    {
+      "epoch": 0.6753078418664938,
+      "grad_norm": 0.9135223031044006,
+      "learning_rate": 4.517389301441657e-06,
+      "loss": 0.1311,
+      "step": 4168
+    },
+    {
+      "epoch": 0.6754698639014906,
+      "grad_norm": 0.9252517819404602,
+      "learning_rate": 4.517130993556051e-06,
+      "loss": 0.1576,
+      "step": 4169
+    },
+    {
+      "epoch": 0.6756318859364874,
+      "grad_norm": 0.8751834034919739,
+      "learning_rate": 4.51687262395112e-06,
+      "loss": 0.1332,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6757939079714841,
+      "grad_norm": 0.8638715147972107,
+      "learning_rate": 4.516614192634765e-06,
+      "loss": 0.1313,
+      "step": 4171
+    },
+    {
+      "epoch": 0.6759559300064809,
+      "grad_norm": 0.910304844379425,
+      "learning_rate": 4.516355699614897e-06,
+      "loss": 0.1441,
+      "step": 4172
+    },
+    {
+      "epoch": 0.6761179520414776,
+      "grad_norm": 0.8590049743652344,
+      "learning_rate": 4.516097144899424e-06,
+      "loss": 0.1324,
+      "step": 4173
+    },
+    {
+      "epoch": 0.6762799740764744,
+      "grad_norm": 0.9081874489784241,
+      "learning_rate": 4.515838528496257e-06,
+      "loss": 0.1493,
+      "step": 4174
+    },
+    {
+      "epoch": 0.6764419961114712,
+      "grad_norm": 0.9282753467559814,
+      "learning_rate": 4.51557985041331e-06,
+      "loss": 0.1555,
+      "step": 4175
+    },
+    {
+      "epoch": 0.6766040181464679,
+      "grad_norm": 0.9835655093193054,
+      "learning_rate": 4.5153211106584965e-06,
+      "loss": 0.1559,
+      "step": 4176
+    },
+    {
+      "epoch": 0.6767660401814647,
+      "grad_norm": 1.0135793685913086,
+      "learning_rate": 4.515062309239734e-06,
+      "loss": 0.1568,
+      "step": 4177
+    },
+    {
+      "epoch": 0.6769280622164614,
+      "grad_norm": 0.8695417046546936,
+      "learning_rate": 4.514803446164941e-06,
+      "loss": 0.1355,
+      "step": 4178
+    },
+    {
+      "epoch": 0.6770900842514582,
+      "grad_norm": 0.7575681805610657,
+      "learning_rate": 4.514544521442039e-06,
+      "loss": 0.1289,
+      "step": 4179
+    },
+    {
+      "epoch": 0.6772521062864549,
+      "grad_norm": 0.9172993302345276,
+      "learning_rate": 4.514285535078949e-06,
+      "loss": 0.1565,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6774141283214518,
+      "grad_norm": 0.8855761885643005,
+      "learning_rate": 4.5140264870835974e-06,
+      "loss": 0.1476,
+      "step": 4181
+    },
+    {
+      "epoch": 0.6775761503564485,
+      "grad_norm": 0.9483667016029358,
+      "learning_rate": 4.513767377463908e-06,
+      "loss": 0.1516,
+      "step": 4182
+    },
+    {
+      "epoch": 0.6777381723914452,
+      "grad_norm": 0.8638037443161011,
+      "learning_rate": 4.51350820622781e-06,
+      "loss": 0.1359,
+      "step": 4183
+    },
+    {
+      "epoch": 0.677900194426442,
+      "grad_norm": 0.8898397088050842,
+      "learning_rate": 4.513248973383234e-06,
+      "loss": 0.1323,
+      "step": 4184
+    },
+    {
+      "epoch": 0.6780622164614387,
+      "grad_norm": 0.9830394983291626,
+      "learning_rate": 4.512989678938111e-06,
+      "loss": 0.1544,
+      "step": 4185
+    },
+    {
+      "epoch": 0.6782242384964355,
+      "grad_norm": 0.8503865599632263,
+      "learning_rate": 4.512730322900375e-06,
+      "loss": 0.1273,
+      "step": 4186
+    },
+    {
+      "epoch": 0.6783862605314323,
+      "grad_norm": 1.0824391841888428,
+      "learning_rate": 4.5124709052779626e-06,
+      "loss": 0.1775,
+      "step": 4187
+    },
+    {
+      "epoch": 0.6785482825664291,
+      "grad_norm": 0.9633022546768188,
+      "learning_rate": 4.51221142607881e-06,
+      "loss": 0.1557,
+      "step": 4188
+    },
+    {
+      "epoch": 0.6787103046014258,
+      "grad_norm": 0.9733066558837891,
+      "learning_rate": 4.511951885310858e-06,
+      "loss": 0.1404,
+      "step": 4189
+    },
+    {
+      "epoch": 0.6788723266364226,
+      "grad_norm": 1.0073535442352295,
+      "learning_rate": 4.511692282982047e-06,
+      "loss": 0.1678,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6790343486714193,
+      "grad_norm": 0.8856287002563477,
+      "learning_rate": 4.511432619100319e-06,
+      "loss": 0.131,
+      "step": 4191
+    },
+    {
+      "epoch": 0.679196370706416,
+      "grad_norm": 0.896113395690918,
+      "learning_rate": 4.511172893673621e-06,
+      "loss": 0.1373,
+      "step": 4192
+    },
+    {
+      "epoch": 0.6793583927414129,
+      "grad_norm": 0.9079231023788452,
+      "learning_rate": 4.5109131067099e-06,
+      "loss": 0.1411,
+      "step": 4193
+    },
+    {
+      "epoch": 0.6795204147764096,
+      "grad_norm": 0.7973667979240417,
+      "learning_rate": 4.510653258217103e-06,
+      "loss": 0.1338,
+      "step": 4194
+    },
+    {
+      "epoch": 0.6796824368114064,
+      "grad_norm": 0.9222817420959473,
+      "learning_rate": 4.510393348203184e-06,
+      "loss": 0.1556,
+      "step": 4195
+    },
+    {
+      "epoch": 0.6798444588464031,
+      "grad_norm": 1.0345180034637451,
+      "learning_rate": 4.5101333766760926e-06,
+      "loss": 0.1483,
+      "step": 4196
+    },
+    {
+      "epoch": 0.6800064808813999,
+      "grad_norm": 0.8539285659790039,
+      "learning_rate": 4.509873343643783e-06,
+      "loss": 0.1385,
+      "step": 4197
+    },
+    {
+      "epoch": 0.6801685029163966,
+      "grad_norm": 0.8832521438598633,
+      "learning_rate": 4.509613249114215e-06,
+      "loss": 0.1512,
+      "step": 4198
+    },
+    {
+      "epoch": 0.6803305249513933,
+      "grad_norm": 1.0024913549423218,
+      "learning_rate": 4.509353093095344e-06,
+      "loss": 0.1768,
+      "step": 4199
+    },
+    {
+      "epoch": 0.6804925469863902,
+      "grad_norm": 0.8852626085281372,
+      "learning_rate": 4.509092875595131e-06,
+      "loss": 0.1321,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6806545690213869,
+      "grad_norm": 0.8801068663597107,
+      "learning_rate": 4.5088325966215375e-06,
+      "loss": 0.147,
+      "step": 4201
+    },
+    {
+      "epoch": 0.6808165910563837,
+      "grad_norm": 0.850282609462738,
+      "learning_rate": 4.508572256182528e-06,
+      "loss": 0.1439,
+      "step": 4202
+    },
+    {
+      "epoch": 0.6809786130913804,
+      "grad_norm": 0.880113959312439,
+      "learning_rate": 4.508311854286068e-06,
+      "loss": 0.1409,
+      "step": 4203
+    },
+    {
+      "epoch": 0.6811406351263772,
+      "grad_norm": 0.895704984664917,
+      "learning_rate": 4.508051390940125e-06,
+      "loss": 0.1374,
+      "step": 4204
+    },
+    {
+      "epoch": 0.681302657161374,
+      "grad_norm": 0.9211687445640564,
+      "learning_rate": 4.507790866152669e-06,
+      "loss": 0.1412,
+      "step": 4205
+    },
+    {
+      "epoch": 0.6814646791963707,
+      "grad_norm": 0.7838913798332214,
+      "learning_rate": 4.507530279931673e-06,
+      "loss": 0.1248,
+      "step": 4206
+    },
+    {
+      "epoch": 0.6816267012313675,
+      "grad_norm": 0.923410952091217,
+      "learning_rate": 4.507269632285106e-06,
+      "loss": 0.1537,
+      "step": 4207
+    },
+    {
+      "epoch": 0.6817887232663642,
+      "grad_norm": 0.8629544377326965,
+      "learning_rate": 4.5070089232209465e-06,
+      "loss": 0.1259,
+      "step": 4208
+    },
+    {
+      "epoch": 0.681950745301361,
+      "grad_norm": 1.0048764944076538,
+      "learning_rate": 4.506748152747171e-06,
+      "loss": 0.1619,
+      "step": 4209
+    },
+    {
+      "epoch": 0.6821127673363577,
+      "grad_norm": 1.133623719215393,
+      "learning_rate": 4.506487320871758e-06,
+      "loss": 0.1596,
+      "step": 4210
+    },
+    {
+      "epoch": 0.6822747893713546,
+      "grad_norm": 0.979425311088562,
+      "learning_rate": 4.50622642760269e-06,
+      "loss": 0.1672,
+      "step": 4211
+    },
+    {
+      "epoch": 0.6824368114063513,
+      "grad_norm": 0.9712530374526978,
+      "learning_rate": 4.5059654729479474e-06,
+      "loss": 0.1245,
+      "step": 4212
+    },
+    {
+      "epoch": 0.682598833441348,
+      "grad_norm": 0.7488314509391785,
+      "learning_rate": 4.505704456915515e-06,
+      "loss": 0.1177,
+      "step": 4213
+    },
+    {
+      "epoch": 0.6827608554763448,
+      "grad_norm": 0.8927283883094788,
+      "learning_rate": 4.505443379513381e-06,
+      "loss": 0.1482,
+      "step": 4214
+    },
+    {
+      "epoch": 0.6829228775113415,
+      "grad_norm": 0.9022080898284912,
+      "learning_rate": 4.5051822407495314e-06,
+      "loss": 0.1382,
+      "step": 4215
+    },
+    {
+      "epoch": 0.6830848995463383,
+      "grad_norm": 1.0110646486282349,
+      "learning_rate": 4.5049210406319585e-06,
+      "loss": 0.1619,
+      "step": 4216
+    },
+    {
+      "epoch": 0.683246921581335,
+      "grad_norm": 0.8575459122657776,
+      "learning_rate": 4.504659779168654e-06,
+      "loss": 0.147,
+      "step": 4217
+    },
+    {
+      "epoch": 0.6834089436163319,
+      "grad_norm": 0.9155558943748474,
+      "learning_rate": 4.5043984563676105e-06,
+      "loss": 0.1438,
+      "step": 4218
+    },
+    {
+      "epoch": 0.6835709656513286,
+      "grad_norm": 0.9820864796638489,
+      "learning_rate": 4.504137072236825e-06,
+      "loss": 0.1536,
+      "step": 4219
+    },
+    {
+      "epoch": 0.6837329876863253,
+      "grad_norm": 0.8118696808815002,
+      "learning_rate": 4.503875626784295e-06,
+      "loss": 0.1275,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6838950097213221,
+      "grad_norm": 0.9118869304656982,
+      "learning_rate": 4.5036141200180206e-06,
+      "loss": 0.1514,
+      "step": 4221
+    },
+    {
+      "epoch": 0.6840570317563188,
+      "grad_norm": 0.895452082157135,
+      "learning_rate": 4.503352551946003e-06,
+      "loss": 0.1496,
+      "step": 4222
+    },
+    {
+      "epoch": 0.6842190537913156,
+      "grad_norm": 0.8009963035583496,
+      "learning_rate": 4.503090922576244e-06,
+      "loss": 0.1173,
+      "step": 4223
+    },
+    {
+      "epoch": 0.6843810758263124,
+      "grad_norm": 0.8450411558151245,
+      "learning_rate": 4.5028292319167515e-06,
+      "loss": 0.1401,
+      "step": 4224
+    },
+    {
+      "epoch": 0.6845430978613092,
+      "grad_norm": 0.884223997592926,
+      "learning_rate": 4.5025674799755306e-06,
+      "loss": 0.1356,
+      "step": 4225
+    },
+    {
+      "epoch": 0.6847051198963059,
+      "grad_norm": 1.1534299850463867,
+      "learning_rate": 4.502305666760592e-06,
+      "loss": 0.1775,
+      "step": 4226
+    },
+    {
+      "epoch": 0.6848671419313026,
+      "grad_norm": 0.9515455365180969,
+      "learning_rate": 4.502043792279943e-06,
+      "loss": 0.1428,
+      "step": 4227
+    },
+    {
+      "epoch": 0.6850291639662994,
+      "grad_norm": 0.9897893071174622,
+      "learning_rate": 4.501781856541601e-06,
+      "loss": 0.1498,
+      "step": 4228
+    },
+    {
+      "epoch": 0.6851911860012961,
+      "grad_norm": 0.8131996393203735,
+      "learning_rate": 4.501519859553578e-06,
+      "loss": 0.1341,
+      "step": 4229
+    },
+    {
+      "epoch": 0.685353208036293,
+      "grad_norm": 0.8663079142570496,
+      "learning_rate": 4.50125780132389e-06,
+      "loss": 0.1286,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6855152300712897,
+      "grad_norm": 0.8376840949058533,
+      "learning_rate": 4.500995681860557e-06,
+      "loss": 0.1301,
+      "step": 4231
+    },
+    {
+      "epoch": 0.6856772521062865,
+      "grad_norm": 0.8951122164726257,
+      "learning_rate": 4.500733501171599e-06,
+      "loss": 0.1411,
+      "step": 4232
+    },
+    {
+      "epoch": 0.6858392741412832,
+      "grad_norm": 1.0260791778564453,
+      "learning_rate": 4.500471259265037e-06,
+      "loss": 0.1705,
+      "step": 4233
+    },
+    {
+      "epoch": 0.68600129617628,
+      "grad_norm": 1.0505539178848267,
+      "learning_rate": 4.500208956148895e-06,
+      "loss": 0.1712,
+      "step": 4234
+    },
+    {
+      "epoch": 0.6861633182112767,
+      "grad_norm": 1.0583857297897339,
+      "learning_rate": 4.4999465918312e-06,
+      "loss": 0.1724,
+      "step": 4235
+    },
+    {
+      "epoch": 0.6863253402462735,
+      "grad_norm": 0.8932731747627258,
+      "learning_rate": 4.499684166319978e-06,
+      "loss": 0.1523,
+      "step": 4236
+    },
+    {
+      "epoch": 0.6864873622812703,
+      "grad_norm": 0.9509609937667847,
+      "learning_rate": 4.499421679623261e-06,
+      "loss": 0.1455,
+      "step": 4237
+    },
+    {
+      "epoch": 0.686649384316267,
+      "grad_norm": 0.9110451340675354,
+      "learning_rate": 4.499159131749079e-06,
+      "loss": 0.1367,
+      "step": 4238
+    },
+    {
+      "epoch": 0.6868114063512638,
+      "grad_norm": 0.9273454546928406,
+      "learning_rate": 4.498896522705465e-06,
+      "loss": 0.1397,
+      "step": 4239
+    },
+    {
+      "epoch": 0.6869734283862605,
+      "grad_norm": 0.8918704390525818,
+      "learning_rate": 4.498633852500455e-06,
+      "loss": 0.1376,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6871354504212573,
+      "grad_norm": 0.9055836200714111,
+      "learning_rate": 4.4983711211420844e-06,
+      "loss": 0.1428,
+      "step": 4241
+    },
+    {
+      "epoch": 0.687297472456254,
+      "grad_norm": 0.9966475963592529,
+      "learning_rate": 4.498108328638395e-06,
+      "loss": 0.1578,
+      "step": 4242
+    },
+    {
+      "epoch": 0.6874594944912508,
+      "grad_norm": 0.954372227191925,
+      "learning_rate": 4.497845474997425e-06,
+      "loss": 0.1375,
+      "step": 4243
+    },
+    {
+      "epoch": 0.6876215165262476,
+      "grad_norm": 1.3921228647232056,
+      "learning_rate": 4.4975825602272185e-06,
+      "loss": 0.1365,
+      "step": 4244
+    },
+    {
+      "epoch": 0.6877835385612443,
+      "grad_norm": 0.9259489178657532,
+      "learning_rate": 4.497319584335821e-06,
+      "loss": 0.1462,
+      "step": 4245
+    },
+    {
+      "epoch": 0.6879455605962411,
+      "grad_norm": 0.9492351412773132,
+      "learning_rate": 4.497056547331276e-06,
+      "loss": 0.149,
+      "step": 4246
+    },
+    {
+      "epoch": 0.6881075826312378,
+      "grad_norm": 0.9279522895812988,
+      "learning_rate": 4.496793449221634e-06,
+      "loss": 0.1434,
+      "step": 4247
+    },
+    {
+      "epoch": 0.6882696046662347,
+      "grad_norm": 1.0066229104995728,
+      "learning_rate": 4.496530290014945e-06,
+      "loss": 0.1634,
+      "step": 4248
+    },
+    {
+      "epoch": 0.6884316267012314,
+      "grad_norm": 0.853121817111969,
+      "learning_rate": 4.496267069719259e-06,
+      "loss": 0.1469,
+      "step": 4249
+    },
+    {
+      "epoch": 0.6885936487362281,
+      "grad_norm": 1.041290521621704,
+      "learning_rate": 4.496003788342633e-06,
+      "loss": 0.1604,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6887556707712249,
+      "grad_norm": 1.0122541189193726,
+      "learning_rate": 4.495740445893121e-06,
+      "loss": 0.1334,
+      "step": 4251
+    },
+    {
+      "epoch": 0.6889176928062216,
+      "grad_norm": 1.0344403982162476,
+      "learning_rate": 4.495477042378781e-06,
+      "loss": 0.1643,
+      "step": 4252
+    },
+    {
+      "epoch": 0.6890797148412184,
+      "grad_norm": 0.811650276184082,
+      "learning_rate": 4.495213577807672e-06,
+      "loss": 0.1343,
+      "step": 4253
+    },
+    {
+      "epoch": 0.6892417368762151,
+      "grad_norm": 0.842697024345398,
+      "learning_rate": 4.494950052187857e-06,
+      "loss": 0.1383,
+      "step": 4254
+    },
+    {
+      "epoch": 0.689403758911212,
+      "grad_norm": 0.8029190897941589,
+      "learning_rate": 4.494686465527397e-06,
+      "loss": 0.1306,
+      "step": 4255
+    },
+    {
+      "epoch": 0.6895657809462087,
+      "grad_norm": 0.8258494734764099,
+      "learning_rate": 4.494422817834359e-06,
+      "loss": 0.1334,
+      "step": 4256
+    },
+    {
+      "epoch": 0.6897278029812054,
+      "grad_norm": 0.8976909518241882,
+      "learning_rate": 4.494159109116809e-06,
+      "loss": 0.1548,
+      "step": 4257
+    },
+    {
+      "epoch": 0.6898898250162022,
+      "grad_norm": 0.8805419206619263,
+      "learning_rate": 4.493895339382815e-06,
+      "loss": 0.1286,
+      "step": 4258
+    },
+    {
+      "epoch": 0.6900518470511989,
+      "grad_norm": 0.7982555627822876,
+      "learning_rate": 4.49363150864045e-06,
+      "loss": 0.1213,
+      "step": 4259
+    },
+    {
+      "epoch": 0.6902138690861958,
+      "grad_norm": 1.0019187927246094,
+      "learning_rate": 4.493367616897785e-06,
+      "loss": 0.1542,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6903758911211925,
+      "grad_norm": 0.832612931728363,
+      "learning_rate": 4.4931036641628946e-06,
+      "loss": 0.1404,
+      "step": 4261
+    },
+    {
+      "epoch": 0.6905379131561893,
+      "grad_norm": 0.9622425436973572,
+      "learning_rate": 4.4928396504438555e-06,
+      "loss": 0.1418,
+      "step": 4262
+    },
+    {
+      "epoch": 0.690699935191186,
+      "grad_norm": 0.8642094731330872,
+      "learning_rate": 4.492575575748746e-06,
+      "loss": 0.1437,
+      "step": 4263
+    },
+    {
+      "epoch": 0.6908619572261827,
+      "grad_norm": 1.0327246189117432,
+      "learning_rate": 4.4923114400856445e-06,
+      "loss": 0.1633,
+      "step": 4264
+    },
+    {
+      "epoch": 0.6910239792611795,
+      "grad_norm": 1.0118496417999268,
+      "learning_rate": 4.492047243462636e-06,
+      "loss": 0.1497,
+      "step": 4265
+    },
+    {
+      "epoch": 0.6911860012961762,
+      "grad_norm": 0.860984742641449,
+      "learning_rate": 4.491782985887802e-06,
+      "loss": 0.1416,
+      "step": 4266
+    },
+    {
+      "epoch": 0.6913480233311731,
+      "grad_norm": 0.9283857941627502,
+      "learning_rate": 4.491518667369228e-06,
+      "loss": 0.1471,
+      "step": 4267
+    },
+    {
+      "epoch": 0.6915100453661698,
+      "grad_norm": 0.9157022833824158,
+      "learning_rate": 4.491254287915003e-06,
+      "loss": 0.1482,
+      "step": 4268
+    },
+    {
+      "epoch": 0.6916720674011666,
+      "grad_norm": 0.9066836833953857,
+      "learning_rate": 4.490989847533214e-06,
+      "loss": 0.1549,
+      "step": 4269
+    },
+    {
+      "epoch": 0.6918340894361633,
+      "grad_norm": 0.7749550938606262,
+      "learning_rate": 4.490725346231954e-06,
+      "loss": 0.1114,
+      "step": 4270
+    },
+    {
+      "epoch": 0.69199611147116,
+      "grad_norm": 0.7799788117408752,
+      "learning_rate": 4.490460784019317e-06,
+      "loss": 0.1278,
+      "step": 4271
+    },
+    {
+      "epoch": 0.6921581335061568,
+      "grad_norm": 1.0473695993423462,
+      "learning_rate": 4.4901961609033965e-06,
+      "loss": 0.19,
+      "step": 4272
+    },
+    {
+      "epoch": 0.6923201555411536,
+      "grad_norm": 0.9344850778579712,
+      "learning_rate": 4.489931476892289e-06,
+      "loss": 0.1531,
+      "step": 4273
+    },
+    {
+      "epoch": 0.6924821775761504,
+      "grad_norm": 0.990973711013794,
+      "learning_rate": 4.489666731994095e-06,
+      "loss": 0.1538,
+      "step": 4274
+    },
+    {
+      "epoch": 0.6926441996111471,
+      "grad_norm": 0.9593062400817871,
+      "learning_rate": 4.4894019262169134e-06,
+      "loss": 0.1559,
+      "step": 4275
+    },
+    {
+      "epoch": 0.6928062216461439,
+      "grad_norm": 0.8276656866073608,
+      "learning_rate": 4.489137059568847e-06,
+      "loss": 0.1293,
+      "step": 4276
+    },
+    {
+      "epoch": 0.6929682436811406,
+      "grad_norm": 0.8525620698928833,
+      "learning_rate": 4.488872132058001e-06,
+      "loss": 0.1359,
+      "step": 4277
+    },
+    {
+      "epoch": 0.6931302657161373,
+      "grad_norm": 0.933274507522583,
+      "learning_rate": 4.48860714369248e-06,
+      "loss": 0.1463,
+      "step": 4278
+    },
+    {
+      "epoch": 0.6932922877511342,
+      "grad_norm": 1.014924168586731,
+      "learning_rate": 4.4883420944803925e-06,
+      "loss": 0.1627,
+      "step": 4279
+    },
+    {
+      "epoch": 0.6934543097861309,
+      "grad_norm": 0.9470938444137573,
+      "learning_rate": 4.488076984429849e-06,
+      "loss": 0.1612,
+      "step": 4280
+    },
+    {
+      "epoch": 0.6936163318211277,
+      "grad_norm": 0.861634373664856,
+      "learning_rate": 4.48781181354896e-06,
+      "loss": 0.1445,
+      "step": 4281
+    },
+    {
+      "epoch": 0.6937783538561244,
+      "grad_norm": 0.8474523425102234,
+      "learning_rate": 4.4875465818458404e-06,
+      "loss": 0.1379,
+      "step": 4282
+    },
+    {
+      "epoch": 0.6939403758911212,
+      "grad_norm": 0.8774752616882324,
+      "learning_rate": 4.487281289328605e-06,
+      "loss": 0.1374,
+      "step": 4283
+    },
+    {
+      "epoch": 0.6941023979261179,
+      "grad_norm": 0.8172398805618286,
+      "learning_rate": 4.4870159360053725e-06,
+      "loss": 0.1332,
+      "step": 4284
+    },
+    {
+      "epoch": 0.6942644199611148,
+      "grad_norm": 0.8533669114112854,
+      "learning_rate": 4.4867505218842596e-06,
+      "loss": 0.1421,
+      "step": 4285
+    },
+    {
+      "epoch": 0.6944264419961115,
+      "grad_norm": 0.909589946269989,
+      "learning_rate": 4.4864850469733886e-06,
+      "loss": 0.1532,
+      "step": 4286
+    },
+    {
+      "epoch": 0.6945884640311082,
+      "grad_norm": 0.7759212255477905,
+      "learning_rate": 4.486219511280883e-06,
+      "loss": 0.1142,
+      "step": 4287
+    },
+    {
+      "epoch": 0.694750486066105,
+      "grad_norm": 0.9085809588432312,
+      "learning_rate": 4.485953914814867e-06,
+      "loss": 0.1354,
+      "step": 4288
+    },
+    {
+      "epoch": 0.6949125081011017,
+      "grad_norm": 0.857865035533905,
+      "learning_rate": 4.485688257583467e-06,
+      "loss": 0.1398,
+      "step": 4289
+    },
+    {
+      "epoch": 0.6950745301360985,
+      "grad_norm": 0.8624712824821472,
+      "learning_rate": 4.485422539594811e-06,
+      "loss": 0.1342,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6952365521710953,
+      "grad_norm": 0.8267756104469299,
+      "learning_rate": 4.48515676085703e-06,
+      "loss": 0.1381,
+      "step": 4291
+    },
+    {
+      "epoch": 0.6953985742060921,
+      "grad_norm": 0.9393563270568848,
+      "learning_rate": 4.4848909213782566e-06,
+      "loss": 0.1485,
+      "step": 4292
+    },
+    {
+      "epoch": 0.6955605962410888,
+      "grad_norm": 0.9284929037094116,
+      "learning_rate": 4.484625021166624e-06,
+      "loss": 0.1549,
+      "step": 4293
+    },
+    {
+      "epoch": 0.6957226182760855,
+      "grad_norm": 0.9308271408081055,
+      "learning_rate": 4.484359060230269e-06,
+      "loss": 0.1443,
+      "step": 4294
+    },
+    {
+      "epoch": 0.6958846403110823,
+      "grad_norm": 0.9445673227310181,
+      "learning_rate": 4.484093038577329e-06,
+      "loss": 0.127,
+      "step": 4295
+    },
+    {
+      "epoch": 0.696046662346079,
+      "grad_norm": 0.8204240798950195,
+      "learning_rate": 4.483826956215942e-06,
+      "loss": 0.1279,
+      "step": 4296
+    },
+    {
+      "epoch": 0.6962086843810759,
+      "grad_norm": 0.9389863610267639,
+      "learning_rate": 4.483560813154252e-06,
+      "loss": 0.1493,
+      "step": 4297
+    },
+    {
+      "epoch": 0.6963707064160726,
+      "grad_norm": 0.8122605085372925,
+      "learning_rate": 4.4832946094004e-06,
+      "loss": 0.1314,
+      "step": 4298
+    },
+    {
+      "epoch": 0.6965327284510694,
+      "grad_norm": 0.9366967082023621,
+      "learning_rate": 4.483028344962534e-06,
+      "loss": 0.1396,
+      "step": 4299
+    },
+    {
+      "epoch": 0.6966947504860661,
+      "grad_norm": 0.8493170738220215,
+      "learning_rate": 4.482762019848799e-06,
+      "loss": 0.1372,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6968567725210628,
+      "grad_norm": 0.9560868740081787,
+      "learning_rate": 4.482495634067344e-06,
+      "loss": 0.1642,
+      "step": 4301
+    },
+    {
+      "epoch": 0.6970187945560596,
+      "grad_norm": 0.8868750333786011,
+      "learning_rate": 4.48222918762632e-06,
+      "loss": 0.1407,
+      "step": 4302
+    },
+    {
+      "epoch": 0.6971808165910564,
+      "grad_norm": 0.8307478427886963,
+      "learning_rate": 4.48196268053388e-06,
+      "loss": 0.136,
+      "step": 4303
+    },
+    {
+      "epoch": 0.6973428386260532,
+      "grad_norm": 0.923088014125824,
+      "learning_rate": 4.481696112798179e-06,
+      "loss": 0.1425,
+      "step": 4304
+    },
+    {
+      "epoch": 0.6975048606610499,
+      "grad_norm": 0.8834422826766968,
+      "learning_rate": 4.481429484427372e-06,
+      "loss": 0.1395,
+      "step": 4305
+    },
+    {
+      "epoch": 0.6976668826960467,
+      "grad_norm": 0.9474985599517822,
+      "learning_rate": 4.481162795429618e-06,
+      "loss": 0.153,
+      "step": 4306
+    },
+    {
+      "epoch": 0.6978289047310434,
+      "grad_norm": 0.9343580603599548,
+      "learning_rate": 4.480896045813076e-06,
+      "loss": 0.145,
+      "step": 4307
+    },
+    {
+      "epoch": 0.6979909267660401,
+      "grad_norm": 0.746100127696991,
+      "learning_rate": 4.480629235585909e-06,
+      "loss": 0.1194,
+      "step": 4308
+    },
+    {
+      "epoch": 0.698152948801037,
+      "grad_norm": 0.9636530876159668,
+      "learning_rate": 4.480362364756281e-06,
+      "loss": 0.1526,
+      "step": 4309
+    },
+    {
+      "epoch": 0.6983149708360337,
+      "grad_norm": 0.7834259271621704,
+      "learning_rate": 4.480095433332357e-06,
+      "loss": 0.1296,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6984769928710305,
+      "grad_norm": 0.9413818717002869,
+      "learning_rate": 4.479828441322304e-06,
+      "loss": 0.1542,
+      "step": 4311
+    },
+    {
+      "epoch": 0.6986390149060272,
+      "grad_norm": 0.9346475005149841,
+      "learning_rate": 4.4795613887342916e-06,
+      "loss": 0.1429,
+      "step": 4312
+    },
+    {
+      "epoch": 0.698801036941024,
+      "grad_norm": 0.917719841003418,
+      "learning_rate": 4.479294275576492e-06,
+      "loss": 0.1366,
+      "step": 4313
+    },
+    {
+      "epoch": 0.6989630589760207,
+      "grad_norm": 0.8321374654769897,
+      "learning_rate": 4.479027101857076e-06,
+      "loss": 0.1395,
+      "step": 4314
+    },
+    {
+      "epoch": 0.6991250810110174,
+      "grad_norm": 0.9253593683242798,
+      "learning_rate": 4.478759867584221e-06,
+      "loss": 0.1371,
+      "step": 4315
+    },
+    {
+      "epoch": 0.6992871030460143,
+      "grad_norm": 0.8277562856674194,
+      "learning_rate": 4.4784925727661025e-06,
+      "loss": 0.1368,
+      "step": 4316
+    },
+    {
+      "epoch": 0.699449125081011,
+      "grad_norm": 1.0059376955032349,
+      "learning_rate": 4.478225217410898e-06,
+      "loss": 0.1577,
+      "step": 4317
+    },
+    {
+      "epoch": 0.6996111471160078,
+      "grad_norm": 0.9551058411598206,
+      "learning_rate": 4.47795780152679e-06,
+      "loss": 0.1414,
+      "step": 4318
+    },
+    {
+      "epoch": 0.6997731691510045,
+      "grad_norm": 1.059830665588379,
+      "learning_rate": 4.47769032512196e-06,
+      "loss": 0.1566,
+      "step": 4319
+    },
+    {
+      "epoch": 0.6999351911860013,
+      "grad_norm": 0.8938872814178467,
+      "learning_rate": 4.477422788204592e-06,
+      "loss": 0.1362,
+      "step": 4320
+    },
+    {
+      "epoch": 0.700097213220998,
+      "grad_norm": 0.9738820195198059,
+      "learning_rate": 4.4771551907828714e-06,
+      "loss": 0.1627,
+      "step": 4321
+    },
+    {
+      "epoch": 0.7002592352559948,
+      "grad_norm": 0.812980592250824,
+      "learning_rate": 4.476887532864986e-06,
+      "loss": 0.1344,
+      "step": 4322
+    },
+    {
+      "epoch": 0.7004212572909916,
+      "grad_norm": 0.9212116599082947,
+      "learning_rate": 4.476619814459128e-06,
+      "loss": 0.1346,
+      "step": 4323
+    },
+    {
+      "epoch": 0.7005832793259883,
+      "grad_norm": 0.7404506802558899,
+      "learning_rate": 4.476352035573486e-06,
+      "loss": 0.1101,
+      "step": 4324
+    },
+    {
+      "epoch": 0.7007453013609851,
+      "grad_norm": 0.8756342530250549,
+      "learning_rate": 4.4760841962162535e-06,
+      "loss": 0.1492,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7009073233959818,
+      "grad_norm": 0.9470415711402893,
+      "learning_rate": 4.475816296395627e-06,
+      "loss": 0.1547,
+      "step": 4326
+    },
+    {
+      "epoch": 0.7010693454309787,
+      "grad_norm": 0.7365562915802002,
+      "learning_rate": 4.475548336119804e-06,
+      "loss": 0.1169,
+      "step": 4327
+    },
+    {
+      "epoch": 0.7012313674659754,
+      "grad_norm": 0.9501516819000244,
+      "learning_rate": 4.475280315396982e-06,
+      "loss": 0.1503,
+      "step": 4328
+    },
+    {
+      "epoch": 0.7013933895009722,
+      "grad_norm": 0.8905579447746277,
+      "learning_rate": 4.475012234235363e-06,
+      "loss": 0.1577,
+      "step": 4329
+    },
+    {
+      "epoch": 0.7015554115359689,
+      "grad_norm": 0.9142157435417175,
+      "learning_rate": 4.474744092643149e-06,
+      "loss": 0.1435,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7017174335709656,
+      "grad_norm": 0.8763073086738586,
+      "learning_rate": 4.474475890628545e-06,
+      "loss": 0.1384,
+      "step": 4331
+    },
+    {
+      "epoch": 0.7018794556059624,
+      "grad_norm": 0.8825966715812683,
+      "learning_rate": 4.474207628199756e-06,
+      "loss": 0.1515,
+      "step": 4332
+    },
+    {
+      "epoch": 0.7020414776409591,
+      "grad_norm": 0.876340389251709,
+      "learning_rate": 4.473939305364991e-06,
+      "loss": 0.1513,
+      "step": 4333
+    },
+    {
+      "epoch": 0.702203499675956,
+      "grad_norm": 0.8109468221664429,
+      "learning_rate": 4.47367092213246e-06,
+      "loss": 0.1298,
+      "step": 4334
+    },
+    {
+      "epoch": 0.7023655217109527,
+      "grad_norm": 0.9166135787963867,
+      "learning_rate": 4.473402478510376e-06,
+      "loss": 0.1428,
+      "step": 4335
+    },
+    {
+      "epoch": 0.7025275437459495,
+      "grad_norm": 0.838951051235199,
+      "learning_rate": 4.473133974506951e-06,
+      "loss": 0.1315,
+      "step": 4336
+    },
+    {
+      "epoch": 0.7026895657809462,
+      "grad_norm": 0.866460382938385,
+      "learning_rate": 4.472865410130401e-06,
+      "loss": 0.1348,
+      "step": 4337
+    },
+    {
+      "epoch": 0.7028515878159429,
+      "grad_norm": 0.8910289406776428,
+      "learning_rate": 4.472596785388944e-06,
+      "loss": 0.1366,
+      "step": 4338
+    },
+    {
+      "epoch": 0.7030136098509397,
+      "grad_norm": 1.0762733221054077,
+      "learning_rate": 4.472328100290799e-06,
+      "loss": 0.173,
+      "step": 4339
+    },
+    {
+      "epoch": 0.7031756318859365,
+      "grad_norm": 0.9918913841247559,
+      "learning_rate": 4.472059354844187e-06,
+      "loss": 0.1718,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7033376539209333,
+      "grad_norm": 0.9252434968948364,
+      "learning_rate": 4.471790549057332e-06,
+      "loss": 0.147,
+      "step": 4341
+    },
+    {
+      "epoch": 0.70349967595593,
+      "grad_norm": 0.9707101583480835,
+      "learning_rate": 4.4715216829384566e-06,
+      "loss": 0.1484,
+      "step": 4342
+    },
+    {
+      "epoch": 0.7036616979909268,
+      "grad_norm": 0.9649327397346497,
+      "learning_rate": 4.471252756495789e-06,
+      "loss": 0.1663,
+      "step": 4343
+    },
+    {
+      "epoch": 0.7038237200259235,
+      "grad_norm": 0.8951583504676819,
+      "learning_rate": 4.470983769737557e-06,
+      "loss": 0.1424,
+      "step": 4344
+    },
+    {
+      "epoch": 0.7039857420609202,
+      "grad_norm": 0.885223388671875,
+      "learning_rate": 4.470714722671992e-06,
+      "loss": 0.1515,
+      "step": 4345
+    },
+    {
+      "epoch": 0.7041477640959171,
+      "grad_norm": 1.0771913528442383,
+      "learning_rate": 4.470445615307325e-06,
+      "loss": 0.179,
+      "step": 4346
+    },
+    {
+      "epoch": 0.7043097861309138,
+      "grad_norm": 1.0075687170028687,
+      "learning_rate": 4.470176447651791e-06,
+      "loss": 0.1614,
+      "step": 4347
+    },
+    {
+      "epoch": 0.7044718081659106,
+      "grad_norm": 0.9713208675384521,
+      "learning_rate": 4.4699072197136255e-06,
+      "loss": 0.1478,
+      "step": 4348
+    },
+    {
+      "epoch": 0.7046338302009073,
+      "grad_norm": 0.9104232788085938,
+      "learning_rate": 4.469637931501066e-06,
+      "loss": 0.1519,
+      "step": 4349
+    },
+    {
+      "epoch": 0.7047958522359041,
+      "grad_norm": 0.9640597105026245,
+      "learning_rate": 4.469368583022352e-06,
+      "loss": 0.172,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7049578742709008,
+      "grad_norm": 0.966541588306427,
+      "learning_rate": 4.469099174285725e-06,
+      "loss": 0.1659,
+      "step": 4351
+    },
+    {
+      "epoch": 0.7051198963058976,
+      "grad_norm": 0.9105446338653564,
+      "learning_rate": 4.468829705299429e-06,
+      "loss": 0.1513,
+      "step": 4352
+    },
+    {
+      "epoch": 0.7052819183408944,
+      "grad_norm": 0.916776180267334,
+      "learning_rate": 4.4685601760717075e-06,
+      "loss": 0.1533,
+      "step": 4353
+    },
+    {
+      "epoch": 0.7054439403758911,
+      "grad_norm": 0.8211317658424377,
+      "learning_rate": 4.4682905866108094e-06,
+      "loss": 0.1411,
+      "step": 4354
+    },
+    {
+      "epoch": 0.7056059624108879,
+      "grad_norm": 0.908372163772583,
+      "learning_rate": 4.468020936924983e-06,
+      "loss": 0.1392,
+      "step": 4355
+    },
+    {
+      "epoch": 0.7057679844458846,
+      "grad_norm": 0.9611108303070068,
+      "learning_rate": 4.467751227022478e-06,
+      "loss": 0.1592,
+      "step": 4356
+    },
+    {
+      "epoch": 0.7059300064808814,
+      "grad_norm": 0.8858509063720703,
+      "learning_rate": 4.467481456911547e-06,
+      "loss": 0.1414,
+      "step": 4357
+    },
+    {
+      "epoch": 0.7060920285158782,
+      "grad_norm": 0.8831855654716492,
+      "learning_rate": 4.467211626600444e-06,
+      "loss": 0.1331,
+      "step": 4358
+    },
+    {
+      "epoch": 0.7062540505508749,
+      "grad_norm": 0.8268503546714783,
+      "learning_rate": 4.466941736097427e-06,
+      "loss": 0.1466,
+      "step": 4359
+    },
+    {
+      "epoch": 0.7064160725858717,
+      "grad_norm": 0.952812135219574,
+      "learning_rate": 4.466671785410752e-06,
+      "loss": 0.1511,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7065780946208684,
+      "grad_norm": 0.9500433802604675,
+      "learning_rate": 4.46640177454868e-06,
+      "loss": 0.1439,
+      "step": 4361
+    },
+    {
+      "epoch": 0.7067401166558652,
+      "grad_norm": 0.8875350952148438,
+      "learning_rate": 4.4661317035194716e-06,
+      "loss": 0.145,
+      "step": 4362
+    },
+    {
+      "epoch": 0.7069021386908619,
+      "grad_norm": 0.8836730718612671,
+      "learning_rate": 4.465861572331392e-06,
+      "loss": 0.1386,
+      "step": 4363
+    },
+    {
+      "epoch": 0.7070641607258588,
+      "grad_norm": 0.9382752180099487,
+      "learning_rate": 4.4655913809927045e-06,
+      "loss": 0.1529,
+      "step": 4364
+    },
+    {
+      "epoch": 0.7072261827608555,
+      "grad_norm": 0.968842625617981,
+      "learning_rate": 4.465321129511678e-06,
+      "loss": 0.1463,
+      "step": 4365
+    },
+    {
+      "epoch": 0.7073882047958522,
+      "grad_norm": 0.9342696666717529,
+      "learning_rate": 4.4650508178965814e-06,
+      "loss": 0.1376,
+      "step": 4366
+    },
+    {
+      "epoch": 0.707550226830849,
+      "grad_norm": 1.0476247072219849,
+      "learning_rate": 4.464780446155684e-06,
+      "loss": 0.1389,
+      "step": 4367
+    },
+    {
+      "epoch": 0.7077122488658457,
+      "grad_norm": 0.9861329197883606,
+      "learning_rate": 4.464510014297261e-06,
+      "loss": 0.1561,
+      "step": 4368
+    },
+    {
+      "epoch": 0.7078742709008425,
+      "grad_norm": 0.8797351717948914,
+      "learning_rate": 4.464239522329585e-06,
+      "loss": 0.1307,
+      "step": 4369
+    },
+    {
+      "epoch": 0.7080362929358393,
+      "grad_norm": 0.9944042563438416,
+      "learning_rate": 4.4639689702609326e-06,
+      "loss": 0.1431,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7081983149708361,
+      "grad_norm": 0.8967122435569763,
+      "learning_rate": 4.463698358099583e-06,
+      "loss": 0.1483,
+      "step": 4371
+    },
+    {
+      "epoch": 0.7083603370058328,
+      "grad_norm": 0.9998703598976135,
+      "learning_rate": 4.463427685853815e-06,
+      "loss": 0.1347,
+      "step": 4372
+    },
+    {
+      "epoch": 0.7085223590408296,
+      "grad_norm": 0.9268200397491455,
+      "learning_rate": 4.463156953531912e-06,
+      "loss": 0.1591,
+      "step": 4373
+    },
+    {
+      "epoch": 0.7086843810758263,
+      "grad_norm": 0.8729379177093506,
+      "learning_rate": 4.462886161142157e-06,
+      "loss": 0.1271,
+      "step": 4374
+    },
+    {
+      "epoch": 0.708846403110823,
+      "grad_norm": 0.8896871209144592,
+      "learning_rate": 4.462615308692835e-06,
+      "loss": 0.1595,
+      "step": 4375
+    },
+    {
+      "epoch": 0.7090084251458199,
+      "grad_norm": 0.8534110188484192,
+      "learning_rate": 4.4623443961922334e-06,
+      "loss": 0.1321,
+      "step": 4376
+    },
+    {
+      "epoch": 0.7091704471808166,
+      "grad_norm": 0.9220858216285706,
+      "learning_rate": 4.462073423648643e-06,
+      "loss": 0.1584,
+      "step": 4377
+    },
+    {
+      "epoch": 0.7093324692158134,
+      "grad_norm": 0.8980048298835754,
+      "learning_rate": 4.461802391070354e-06,
+      "loss": 0.1432,
+      "step": 4378
+    },
+    {
+      "epoch": 0.7094944912508101,
+      "grad_norm": 0.8806954622268677,
+      "learning_rate": 4.461531298465659e-06,
+      "loss": 0.1342,
+      "step": 4379
+    },
+    {
+      "epoch": 0.7096565132858069,
+      "grad_norm": 0.9965754747390747,
+      "learning_rate": 4.4612601458428525e-06,
+      "loss": 0.1537,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7098185353208036,
+      "grad_norm": 0.9695927500724792,
+      "learning_rate": 4.460988933210233e-06,
+      "loss": 0.1357,
+      "step": 4381
+    },
+    {
+      "epoch": 0.7099805573558003,
+      "grad_norm": 0.9649356007575989,
+      "learning_rate": 4.460717660576097e-06,
+      "loss": 0.1534,
+      "step": 4382
+    },
+    {
+      "epoch": 0.7101425793907972,
+      "grad_norm": 0.9325234293937683,
+      "learning_rate": 4.460446327948745e-06,
+      "loss": 0.1443,
+      "step": 4383
+    },
+    {
+      "epoch": 0.7103046014257939,
+      "grad_norm": 1.038757562637329,
+      "learning_rate": 4.46017493533648e-06,
+      "loss": 0.1623,
+      "step": 4384
+    },
+    {
+      "epoch": 0.7104666234607907,
+      "grad_norm": 0.959611177444458,
+      "learning_rate": 4.459903482747605e-06,
+      "loss": 0.1428,
+      "step": 4385
+    },
+    {
+      "epoch": 0.7106286454957874,
+      "grad_norm": 0.9437382817268372,
+      "learning_rate": 4.459631970190428e-06,
+      "loss": 0.1429,
+      "step": 4386
+    },
+    {
+      "epoch": 0.7107906675307842,
+      "grad_norm": 0.9380490779876709,
+      "learning_rate": 4.459360397673253e-06,
+      "loss": 0.1459,
+      "step": 4387
+    },
+    {
+      "epoch": 0.710952689565781,
+      "grad_norm": 0.8897026777267456,
+      "learning_rate": 4.4590887652043925e-06,
+      "loss": 0.1397,
+      "step": 4388
+    },
+    {
+      "epoch": 0.7111147116007777,
+      "grad_norm": 0.9225705862045288,
+      "learning_rate": 4.458817072792155e-06,
+      "loss": 0.1516,
+      "step": 4389
+    },
+    {
+      "epoch": 0.7112767336357745,
+      "grad_norm": 0.8232748508453369,
+      "learning_rate": 4.458545320444857e-06,
+      "loss": 0.1326,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7114387556707712,
+      "grad_norm": 0.9818866848945618,
+      "learning_rate": 4.458273508170812e-06,
+      "loss": 0.1557,
+      "step": 4391
+    },
+    {
+      "epoch": 0.711600777705768,
+      "grad_norm": 1.086909294128418,
+      "learning_rate": 4.458001635978335e-06,
+      "loss": 0.1492,
+      "step": 4392
+    },
+    {
+      "epoch": 0.7117627997407647,
+      "grad_norm": 0.8664153814315796,
+      "learning_rate": 4.457729703875749e-06,
+      "loss": 0.135,
+      "step": 4393
+    },
+    {
+      "epoch": 0.7119248217757616,
+      "grad_norm": 0.8988549113273621,
+      "learning_rate": 4.457457711871369e-06,
+      "loss": 0.1473,
+      "step": 4394
+    },
+    {
+      "epoch": 0.7120868438107583,
+      "grad_norm": 0.9167562127113342,
+      "learning_rate": 4.45718565997352e-06,
+      "loss": 0.1553,
+      "step": 4395
+    },
+    {
+      "epoch": 0.712248865845755,
+      "grad_norm": 0.8664883971214294,
+      "learning_rate": 4.4569135481905274e-06,
+      "loss": 0.1404,
+      "step": 4396
+    },
+    {
+      "epoch": 0.7124108878807518,
+      "grad_norm": 0.7986405491828918,
+      "learning_rate": 4.456641376530715e-06,
+      "loss": 0.1239,
+      "step": 4397
+    },
+    {
+      "epoch": 0.7125729099157485,
+      "grad_norm": 0.9306362867355347,
+      "learning_rate": 4.456369145002412e-06,
+      "loss": 0.1489,
+      "step": 4398
+    },
+    {
+      "epoch": 0.7127349319507453,
+      "grad_norm": 0.9109500050544739,
+      "learning_rate": 4.456096853613948e-06,
+      "loss": 0.1439,
+      "step": 4399
+    },
+    {
+      "epoch": 0.712896953985742,
+      "grad_norm": 0.8157075643539429,
+      "learning_rate": 4.455824502373653e-06,
+      "loss": 0.1308,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7130589760207389,
+      "grad_norm": 0.8615176677703857,
+      "learning_rate": 4.4555520912898616e-06,
+      "loss": 0.1299,
+      "step": 4401
+    },
+    {
+      "epoch": 0.7132209980557356,
+      "grad_norm": 0.9204617142677307,
+      "learning_rate": 4.455279620370908e-06,
+      "loss": 0.1405,
+      "step": 4402
+    },
+    {
+      "epoch": 0.7133830200907323,
+      "grad_norm": 0.8997926115989685,
+      "learning_rate": 4.455007089625131e-06,
+      "loss": 0.1319,
+      "step": 4403
+    },
+    {
+      "epoch": 0.7135450421257291,
+      "grad_norm": 0.861057698726654,
+      "learning_rate": 4.454734499060867e-06,
+      "loss": 0.1556,
+      "step": 4404
+    },
+    {
+      "epoch": 0.7137070641607258,
+      "grad_norm": 0.9755337238311768,
+      "learning_rate": 4.4544618486864575e-06,
+      "loss": 0.1606,
+      "step": 4405
+    },
+    {
+      "epoch": 0.7138690861957226,
+      "grad_norm": 0.9111999869346619,
+      "learning_rate": 4.454189138510246e-06,
+      "loss": 0.1339,
+      "step": 4406
+    },
+    {
+      "epoch": 0.7140311082307194,
+      "grad_norm": 0.8962233662605286,
+      "learning_rate": 4.4539163685405755e-06,
+      "loss": 0.1345,
+      "step": 4407
+    },
+    {
+      "epoch": 0.7141931302657162,
+      "grad_norm": 0.9078991413116455,
+      "learning_rate": 4.453643538785793e-06,
+      "loss": 0.1504,
+      "step": 4408
+    },
+    {
+      "epoch": 0.7143551523007129,
+      "grad_norm": 0.9566717743873596,
+      "learning_rate": 4.453370649254245e-06,
+      "loss": 0.1546,
+      "step": 4409
+    },
+    {
+      "epoch": 0.7145171743357096,
+      "grad_norm": 0.8746940493583679,
+      "learning_rate": 4.453097699954282e-06,
+      "loss": 0.147,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7146791963707064,
+      "grad_norm": 0.9663369059562683,
+      "learning_rate": 4.452824690894257e-06,
+      "loss": 0.1594,
+      "step": 4411
+    },
+    {
+      "epoch": 0.7148412184057031,
+      "grad_norm": 0.9418163895606995,
+      "learning_rate": 4.452551622082522e-06,
+      "loss": 0.1533,
+      "step": 4412
+    },
+    {
+      "epoch": 0.7150032404407,
+      "grad_norm": 0.8881242871284485,
+      "learning_rate": 4.452278493527431e-06,
+      "loss": 0.1353,
+      "step": 4413
+    },
+    {
+      "epoch": 0.7151652624756967,
+      "grad_norm": 0.9275038838386536,
+      "learning_rate": 4.452005305237344e-06,
+      "loss": 0.1411,
+      "step": 4414
+    },
+    {
+      "epoch": 0.7153272845106935,
+      "grad_norm": 1.107954740524292,
+      "learning_rate": 4.451732057220618e-06,
+      "loss": 0.1572,
+      "step": 4415
+    },
+    {
+      "epoch": 0.7154893065456902,
+      "grad_norm": 0.9651256799697876,
+      "learning_rate": 4.451458749485614e-06,
+      "loss": 0.153,
+      "step": 4416
+    },
+    {
+      "epoch": 0.7156513285806869,
+      "grad_norm": 0.8925125598907471,
+      "learning_rate": 4.451185382040695e-06,
+      "loss": 0.1423,
+      "step": 4417
+    },
+    {
+      "epoch": 0.7158133506156837,
+      "grad_norm": 0.8692678213119507,
+      "learning_rate": 4.4509119548942245e-06,
+      "loss": 0.1448,
+      "step": 4418
+    },
+    {
+      "epoch": 0.7159753726506805,
+      "grad_norm": 0.9533895254135132,
+      "learning_rate": 4.45063846805457e-06,
+      "loss": 0.1495,
+      "step": 4419
+    },
+    {
+      "epoch": 0.7161373946856773,
+      "grad_norm": 0.8917126059532166,
+      "learning_rate": 4.450364921530099e-06,
+      "loss": 0.1437,
+      "step": 4420
+    },
+    {
+      "epoch": 0.716299416720674,
+      "grad_norm": 0.7850384712219238,
+      "learning_rate": 4.450091315329181e-06,
+      "loss": 0.1339,
+      "step": 4421
+    },
+    {
+      "epoch": 0.7164614387556708,
+      "grad_norm": 1.0266042947769165,
+      "learning_rate": 4.449817649460187e-06,
+      "loss": 0.1571,
+      "step": 4422
+    },
+    {
+      "epoch": 0.7166234607906675,
+      "grad_norm": 1.084693193435669,
+      "learning_rate": 4.449543923931493e-06,
+      "loss": 0.1458,
+      "step": 4423
+    },
+    {
+      "epoch": 0.7167854828256643,
+      "grad_norm": 0.9740076065063477,
+      "learning_rate": 4.449270138751471e-06,
+      "loss": 0.1395,
+      "step": 4424
+    },
+    {
+      "epoch": 0.7169475048606611,
+      "grad_norm": 0.8249508738517761,
+      "learning_rate": 4.4489962939285015e-06,
+      "loss": 0.133,
+      "step": 4425
+    },
+    {
+      "epoch": 0.7171095268956578,
+      "grad_norm": 0.9032569527626038,
+      "learning_rate": 4.4487223894709606e-06,
+      "loss": 0.1395,
+      "step": 4426
+    },
+    {
+      "epoch": 0.7172715489306546,
+      "grad_norm": 0.7856298089027405,
+      "learning_rate": 4.448448425387231e-06,
+      "loss": 0.1226,
+      "step": 4427
+    },
+    {
+      "epoch": 0.7174335709656513,
+      "grad_norm": 0.8503175377845764,
+      "learning_rate": 4.448174401685694e-06,
+      "loss": 0.1409,
+      "step": 4428
+    },
+    {
+      "epoch": 0.7175955930006481,
+      "grad_norm": 0.9140015244483948,
+      "learning_rate": 4.447900318374736e-06,
+      "loss": 0.1488,
+      "step": 4429
+    },
+    {
+      "epoch": 0.7177576150356448,
+      "grad_norm": 0.8657589554786682,
+      "learning_rate": 4.447626175462741e-06,
+      "loss": 0.1395,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7179196370706417,
+      "grad_norm": 0.8597119450569153,
+      "learning_rate": 4.447351972958099e-06,
+      "loss": 0.1285,
+      "step": 4431
+    },
+    {
+      "epoch": 0.7180816591056384,
+      "grad_norm": 1.0180467367172241,
+      "learning_rate": 4.447077710869199e-06,
+      "loss": 0.1668,
+      "step": 4432
+    },
+    {
+      "epoch": 0.7182436811406351,
+      "grad_norm": 0.924322783946991,
+      "learning_rate": 4.446803389204433e-06,
+      "loss": 0.1472,
+      "step": 4433
+    },
+    {
+      "epoch": 0.7184057031756319,
+      "grad_norm": 0.9639684557914734,
+      "learning_rate": 4.4465290079721935e-06,
+      "loss": 0.1419,
+      "step": 4434
+    },
+    {
+      "epoch": 0.7185677252106286,
+      "grad_norm": 0.9265577793121338,
+      "learning_rate": 4.446254567180877e-06,
+      "loss": 0.1461,
+      "step": 4435
+    },
+    {
+      "epoch": 0.7187297472456254,
+      "grad_norm": 0.9008076190948486,
+      "learning_rate": 4.445980066838882e-06,
+      "loss": 0.1408,
+      "step": 4436
+    },
+    {
+      "epoch": 0.7188917692806222,
+      "grad_norm": 0.8096259236335754,
+      "learning_rate": 4.445705506954605e-06,
+      "loss": 0.1268,
+      "step": 4437
+    },
+    {
+      "epoch": 0.719053791315619,
+      "grad_norm": 0.9201697111129761,
+      "learning_rate": 4.4454308875364486e-06,
+      "loss": 0.1484,
+      "step": 4438
+    },
+    {
+      "epoch": 0.7192158133506157,
+      "grad_norm": 1.0293805599212646,
+      "learning_rate": 4.445156208592814e-06,
+      "loss": 0.1722,
+      "step": 4439
+    },
+    {
+      "epoch": 0.7193778353856124,
+      "grad_norm": 0.9077914357185364,
+      "learning_rate": 4.444881470132108e-06,
+      "loss": 0.1476,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7195398574206092,
+      "grad_norm": 0.9666748046875,
+      "learning_rate": 4.444606672162735e-06,
+      "loss": 0.1537,
+      "step": 4441
+    },
+    {
+      "epoch": 0.7197018794556059,
+      "grad_norm": 0.942487895488739,
+      "learning_rate": 4.444331814693103e-06,
+      "loss": 0.1257,
+      "step": 4442
+    },
+    {
+      "epoch": 0.7198639014906028,
+      "grad_norm": 0.9764305353164673,
+      "learning_rate": 4.444056897731622e-06,
+      "loss": 0.1494,
+      "step": 4443
+    },
+    {
+      "epoch": 0.7200259235255995,
+      "grad_norm": 0.8661373257637024,
+      "learning_rate": 4.443781921286706e-06,
+      "loss": 0.1331,
+      "step": 4444
+    },
+    {
+      "epoch": 0.7201879455605963,
+      "grad_norm": 0.8567464351654053,
+      "learning_rate": 4.443506885366767e-06,
+      "loss": 0.143,
+      "step": 4445
+    },
+    {
+      "epoch": 0.720349967595593,
+      "grad_norm": 0.8924309015274048,
+      "learning_rate": 4.4432317899802205e-06,
+      "loss": 0.1297,
+      "step": 4446
+    },
+    {
+      "epoch": 0.7205119896305897,
+      "grad_norm": 0.874181866645813,
+      "learning_rate": 4.442956635135482e-06,
+      "loss": 0.1448,
+      "step": 4447
+    },
+    {
+      "epoch": 0.7206740116655865,
+      "grad_norm": 0.9796168804168701,
+      "learning_rate": 4.442681420840974e-06,
+      "loss": 0.1361,
+      "step": 4448
+    },
+    {
+      "epoch": 0.7208360337005832,
+      "grad_norm": 0.9041408896446228,
+      "learning_rate": 4.442406147105116e-06,
+      "loss": 0.1405,
+      "step": 4449
+    },
+    {
+      "epoch": 0.7209980557355801,
+      "grad_norm": 0.9344844222068787,
+      "learning_rate": 4.44213081393633e-06,
+      "loss": 0.1571,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7211600777705768,
+      "grad_norm": 0.927335798740387,
+      "learning_rate": 4.4418554213430405e-06,
+      "loss": 0.1417,
+      "step": 4451
+    },
+    {
+      "epoch": 0.7213220998055736,
+      "grad_norm": 0.7550799250602722,
+      "learning_rate": 4.441579969333675e-06,
+      "loss": 0.1111,
+      "step": 4452
+    },
+    {
+      "epoch": 0.7214841218405703,
+      "grad_norm": 0.8267915844917297,
+      "learning_rate": 4.44130445791666e-06,
+      "loss": 0.1308,
+      "step": 4453
+    },
+    {
+      "epoch": 0.721646143875567,
+      "grad_norm": 1.0004557371139526,
+      "learning_rate": 4.441028887100427e-06,
+      "loss": 0.1523,
+      "step": 4454
+    },
+    {
+      "epoch": 0.7218081659105638,
+      "grad_norm": 0.9250260591506958,
+      "learning_rate": 4.440753256893408e-06,
+      "loss": 0.1553,
+      "step": 4455
+    },
+    {
+      "epoch": 0.7219701879455606,
+      "grad_norm": 0.8540483713150024,
+      "learning_rate": 4.4404775673040346e-06,
+      "loss": 0.156,
+      "step": 4456
+    },
+    {
+      "epoch": 0.7221322099805574,
+      "grad_norm": 0.8714447617530823,
+      "learning_rate": 4.4402018183407435e-06,
+      "loss": 0.1415,
+      "step": 4457
+    },
+    {
+      "epoch": 0.7222942320155541,
+      "grad_norm": 0.9081652164459229,
+      "learning_rate": 4.4399260100119726e-06,
+      "loss": 0.148,
+      "step": 4458
+    },
+    {
+      "epoch": 0.7224562540505509,
+      "grad_norm": 0.9180371165275574,
+      "learning_rate": 4.439650142326161e-06,
+      "loss": 0.1537,
+      "step": 4459
+    },
+    {
+      "epoch": 0.7226182760855476,
+      "grad_norm": 0.9707509279251099,
+      "learning_rate": 4.439374215291748e-06,
+      "loss": 0.1333,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7227802981205443,
+      "grad_norm": 0.9013661742210388,
+      "learning_rate": 4.439098228917177e-06,
+      "loss": 0.1352,
+      "step": 4461
+    },
+    {
+      "epoch": 0.7229423201555412,
+      "grad_norm": 0.8960440158843994,
+      "learning_rate": 4.438822183210894e-06,
+      "loss": 0.1514,
+      "step": 4462
+    },
+    {
+      "epoch": 0.7231043421905379,
+      "grad_norm": 0.8287968635559082,
+      "learning_rate": 4.4385460781813426e-06,
+      "loss": 0.1281,
+      "step": 4463
+    },
+    {
+      "epoch": 0.7232663642255347,
+      "grad_norm": 0.8026749491691589,
+      "learning_rate": 4.438269913836972e-06,
+      "loss": 0.1293,
+      "step": 4464
+    },
+    {
+      "epoch": 0.7234283862605314,
+      "grad_norm": 0.9157952070236206,
+      "learning_rate": 4.437993690186234e-06,
+      "loss": 0.1577,
+      "step": 4465
+    },
+    {
+      "epoch": 0.7235904082955282,
+      "grad_norm": 0.8798143863677979,
+      "learning_rate": 4.437717407237578e-06,
+      "loss": 0.1296,
+      "step": 4466
+    },
+    {
+      "epoch": 0.7237524303305249,
+      "grad_norm": 0.8337207436561584,
+      "learning_rate": 4.437441064999459e-06,
+      "loss": 0.1351,
+      "step": 4467
+    },
+    {
+      "epoch": 0.7239144523655218,
+      "grad_norm": 0.9726783037185669,
+      "learning_rate": 4.437164663480332e-06,
+      "loss": 0.1518,
+      "step": 4468
+    },
+    {
+      "epoch": 0.7240764744005185,
+      "grad_norm": 0.9185786843299866,
+      "learning_rate": 4.436888202688654e-06,
+      "loss": 0.1466,
+      "step": 4469
+    },
+    {
+      "epoch": 0.7242384964355152,
+      "grad_norm": 1.0107709169387817,
+      "learning_rate": 4.436611682632884e-06,
+      "loss": 0.1403,
+      "step": 4470
+    },
+    {
+      "epoch": 0.724400518470512,
+      "grad_norm": 0.886813759803772,
+      "learning_rate": 4.436335103321484e-06,
+      "loss": 0.1544,
+      "step": 4471
+    },
+    {
+      "epoch": 0.7245625405055087,
+      "grad_norm": 0.8717353343963623,
+      "learning_rate": 4.436058464762915e-06,
+      "loss": 0.1332,
+      "step": 4472
+    },
+    {
+      "epoch": 0.7247245625405055,
+      "grad_norm": 0.9517882466316223,
+      "learning_rate": 4.435781766965641e-06,
+      "loss": 0.1582,
+      "step": 4473
+    },
+    {
+      "epoch": 0.7248865845755023,
+      "grad_norm": 0.946277379989624,
+      "learning_rate": 4.435505009938131e-06,
+      "loss": 0.1437,
+      "step": 4474
+    },
+    {
+      "epoch": 0.7250486066104991,
+      "grad_norm": 0.970646858215332,
+      "learning_rate": 4.435228193688851e-06,
+      "loss": 0.1502,
+      "step": 4475
+    },
+    {
+      "epoch": 0.7252106286454958,
+      "grad_norm": 0.8949056267738342,
+      "learning_rate": 4.434951318226272e-06,
+      "loss": 0.1389,
+      "step": 4476
+    },
+    {
+      "epoch": 0.7253726506804925,
+      "grad_norm": 0.8123460412025452,
+      "learning_rate": 4.434674383558865e-06,
+      "loss": 0.1244,
+      "step": 4477
+    },
+    {
+      "epoch": 0.7255346727154893,
+      "grad_norm": 0.8786303400993347,
+      "learning_rate": 4.434397389695102e-06,
+      "loss": 0.1267,
+      "step": 4478
+    },
+    {
+      "epoch": 0.725696694750486,
+      "grad_norm": 0.9474332332611084,
+      "learning_rate": 4.434120336643462e-06,
+      "loss": 0.1479,
+      "step": 4479
+    },
+    {
+      "epoch": 0.7258587167854829,
+      "grad_norm": 0.917175829410553,
+      "learning_rate": 4.433843224412419e-06,
+      "loss": 0.1538,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7260207388204796,
+      "grad_norm": 0.8738764524459839,
+      "learning_rate": 4.433566053010454e-06,
+      "loss": 0.1279,
+      "step": 4481
+    },
+    {
+      "epoch": 0.7261827608554764,
+      "grad_norm": 0.9041066765785217,
+      "learning_rate": 4.4332888224460466e-06,
+      "loss": 0.1384,
+      "step": 4482
+    },
+    {
+      "epoch": 0.7263447828904731,
+      "grad_norm": 0.8361867070198059,
+      "learning_rate": 4.433011532727679e-06,
+      "loss": 0.1305,
+      "step": 4483
+    },
+    {
+      "epoch": 0.7265068049254698,
+      "grad_norm": 0.9306924343109131,
+      "learning_rate": 4.432734183863837e-06,
+      "loss": 0.1426,
+      "step": 4484
+    },
+    {
+      "epoch": 0.7266688269604666,
+      "grad_norm": 0.9259305596351624,
+      "learning_rate": 4.432456775863006e-06,
+      "loss": 0.1332,
+      "step": 4485
+    },
+    {
+      "epoch": 0.7268308489954634,
+      "grad_norm": 0.8822481036186218,
+      "learning_rate": 4.432179308733674e-06,
+      "loss": 0.1458,
+      "step": 4486
+    },
+    {
+      "epoch": 0.7269928710304602,
+      "grad_norm": 0.922392725944519,
+      "learning_rate": 4.4319017824843315e-06,
+      "loss": 0.1534,
+      "step": 4487
+    },
+    {
+      "epoch": 0.7271548930654569,
+      "grad_norm": 0.8622284531593323,
+      "learning_rate": 4.43162419712347e-06,
+      "loss": 0.1243,
+      "step": 4488
+    },
+    {
+      "epoch": 0.7273169151004537,
+      "grad_norm": 0.9705631136894226,
+      "learning_rate": 4.431346552659581e-06,
+      "loss": 0.1426,
+      "step": 4489
+    },
+    {
+      "epoch": 0.7274789371354504,
+      "grad_norm": 0.9161760807037354,
+      "learning_rate": 4.431068849101162e-06,
+      "loss": 0.1513,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7276409591704471,
+      "grad_norm": 0.9033561944961548,
+      "learning_rate": 4.430791086456709e-06,
+      "loss": 0.133,
+      "step": 4491
+    },
+    {
+      "epoch": 0.727802981205444,
+      "grad_norm": 0.9011300206184387,
+      "learning_rate": 4.4305132647347215e-06,
+      "loss": 0.1466,
+      "step": 4492
+    },
+    {
+      "epoch": 0.7279650032404407,
+      "grad_norm": 1.0322880744934082,
+      "learning_rate": 4.4302353839437e-06,
+      "loss": 0.1472,
+      "step": 4493
+    },
+    {
+      "epoch": 0.7281270252754375,
+      "grad_norm": 0.9922388195991516,
+      "learning_rate": 4.429957444092146e-06,
+      "loss": 0.1596,
+      "step": 4494
+    },
+    {
+      "epoch": 0.7282890473104342,
+      "grad_norm": 1.0610861778259277,
+      "learning_rate": 4.4296794451885665e-06,
+      "loss": 0.178,
+      "step": 4495
+    },
+    {
+      "epoch": 0.728451069345431,
+      "grad_norm": 0.9445720314979553,
+      "learning_rate": 4.429401387241464e-06,
+      "loss": 0.1644,
+      "step": 4496
+    },
+    {
+      "epoch": 0.7286130913804277,
+      "grad_norm": 1.009913444519043,
+      "learning_rate": 4.429123270259348e-06,
+      "loss": 0.1726,
+      "step": 4497
+    },
+    {
+      "epoch": 0.7287751134154244,
+      "grad_norm": 0.9682244658470154,
+      "learning_rate": 4.428845094250729e-06,
+      "loss": 0.1474,
+      "step": 4498
+    },
+    {
+      "epoch": 0.7289371354504213,
+      "grad_norm": 0.9171390533447266,
+      "learning_rate": 4.4285668592241186e-06,
+      "loss": 0.1586,
+      "step": 4499
+    },
+    {
+      "epoch": 0.729099157485418,
+      "grad_norm": 0.9223143458366394,
+      "learning_rate": 4.428288565188028e-06,
+      "loss": 0.1357,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7292611795204148,
+      "grad_norm": 0.9437743425369263,
+      "learning_rate": 4.4280102121509734e-06,
+      "loss": 0.1514,
+      "step": 4501
+    },
+    {
+      "epoch": 0.7294232015554115,
+      "grad_norm": 0.8216745853424072,
+      "learning_rate": 4.427731800121473e-06,
+      "loss": 0.1336,
+      "step": 4502
+    },
+    {
+      "epoch": 0.7295852235904083,
+      "grad_norm": 0.8366663455963135,
+      "learning_rate": 4.427453329108045e-06,
+      "loss": 0.1307,
+      "step": 4503
+    },
+    {
+      "epoch": 0.729747245625405,
+      "grad_norm": 0.8991605043411255,
+      "learning_rate": 4.427174799119208e-06,
+      "loss": 0.151,
+      "step": 4504
+    },
+    {
+      "epoch": 0.7299092676604018,
+      "grad_norm": 0.7747397422790527,
+      "learning_rate": 4.426896210163487e-06,
+      "loss": 0.1273,
+      "step": 4505
+    },
+    {
+      "epoch": 0.7300712896953986,
+      "grad_norm": 1.0024263858795166,
+      "learning_rate": 4.426617562249405e-06,
+      "loss": 0.1545,
+      "step": 4506
+    },
+    {
+      "epoch": 0.7302333117303953,
+      "grad_norm": 0.8834472298622131,
+      "learning_rate": 4.426338855385487e-06,
+      "loss": 0.1525,
+      "step": 4507
+    },
+    {
+      "epoch": 0.7303953337653921,
+      "grad_norm": 0.8129113912582397,
+      "learning_rate": 4.426060089580262e-06,
+      "loss": 0.1393,
+      "step": 4508
+    },
+    {
+      "epoch": 0.7305573558003888,
+      "grad_norm": 1.0072026252746582,
+      "learning_rate": 4.42578126484226e-06,
+      "loss": 0.1591,
+      "step": 4509
+    },
+    {
+      "epoch": 0.7307193778353857,
+      "grad_norm": 0.9830167889595032,
+      "learning_rate": 4.42550238118001e-06,
+      "loss": 0.1498,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7308813998703824,
+      "grad_norm": 0.8229345083236694,
+      "learning_rate": 4.425223438602047e-06,
+      "loss": 0.133,
+      "step": 4511
+    },
+    {
+      "epoch": 0.7310434219053791,
+      "grad_norm": 0.8896669745445251,
+      "learning_rate": 4.424944437116907e-06,
+      "loss": 0.1397,
+      "step": 4512
+    },
+    {
+      "epoch": 0.7312054439403759,
+      "grad_norm": 0.9277743101119995,
+      "learning_rate": 4.424665376733125e-06,
+      "loss": 0.1421,
+      "step": 4513
+    },
+    {
+      "epoch": 0.7313674659753726,
+      "grad_norm": 0.9299547076225281,
+      "learning_rate": 4.424386257459241e-06,
+      "loss": 0.1507,
+      "step": 4514
+    },
+    {
+      "epoch": 0.7315294880103694,
+      "grad_norm": 0.9168359637260437,
+      "learning_rate": 4.424107079303793e-06,
+      "loss": 0.149,
+      "step": 4515
+    },
+    {
+      "epoch": 0.7316915100453661,
+      "grad_norm": 0.994986891746521,
+      "learning_rate": 4.423827842275325e-06,
+      "loss": 0.1692,
+      "step": 4516
+    },
+    {
+      "epoch": 0.731853532080363,
+      "grad_norm": 1.0216580629348755,
+      "learning_rate": 4.42354854638238e-06,
+      "loss": 0.1601,
+      "step": 4517
+    },
+    {
+      "epoch": 0.7320155541153597,
+      "grad_norm": 0.8891658782958984,
+      "learning_rate": 4.4232691916335055e-06,
+      "loss": 0.1446,
+      "step": 4518
+    },
+    {
+      "epoch": 0.7321775761503565,
+      "grad_norm": 0.8627836108207703,
+      "learning_rate": 4.422989778037248e-06,
+      "loss": 0.1371,
+      "step": 4519
+    },
+    {
+      "epoch": 0.7323395981853532,
+      "grad_norm": 0.8696764707565308,
+      "learning_rate": 4.422710305602156e-06,
+      "loss": 0.1213,
+      "step": 4520
+    },
+    {
+      "epoch": 0.7325016202203499,
+      "grad_norm": 0.8861494064331055,
+      "learning_rate": 4.422430774336782e-06,
+      "loss": 0.1454,
+      "step": 4521
+    },
+    {
+      "epoch": 0.7326636422553467,
+      "grad_norm": 0.9271621108055115,
+      "learning_rate": 4.422151184249679e-06,
+      "loss": 0.1513,
+      "step": 4522
+    },
+    {
+      "epoch": 0.7328256642903435,
+      "grad_norm": 0.8663860559463501,
+      "learning_rate": 4.4218715353494e-06,
+      "loss": 0.1484,
+      "step": 4523
+    },
+    {
+      "epoch": 0.7329876863253403,
+      "grad_norm": 0.8715428113937378,
+      "learning_rate": 4.421591827644503e-06,
+      "loss": 0.1486,
+      "step": 4524
+    },
+    {
+      "epoch": 0.733149708360337,
+      "grad_norm": 0.8773646950721741,
+      "learning_rate": 4.4213120611435475e-06,
+      "loss": 0.1462,
+      "step": 4525
+    },
+    {
+      "epoch": 0.7333117303953338,
+      "grad_norm": 0.775646984577179,
+      "learning_rate": 4.4210322358550915e-06,
+      "loss": 0.1371,
+      "step": 4526
+    },
+    {
+      "epoch": 0.7334737524303305,
+      "grad_norm": 0.9878357648849487,
+      "learning_rate": 4.420752351787698e-06,
+      "loss": 0.1479,
+      "step": 4527
+    },
+    {
+      "epoch": 0.7336357744653272,
+      "grad_norm": 1.0782580375671387,
+      "learning_rate": 4.420472408949931e-06,
+      "loss": 0.1573,
+      "step": 4528
+    },
+    {
+      "epoch": 0.7337977965003241,
+      "grad_norm": 0.8356068134307861,
+      "learning_rate": 4.420192407350355e-06,
+      "loss": 0.1349,
+      "step": 4529
+    },
+    {
+      "epoch": 0.7339598185353208,
+      "grad_norm": 0.8317657113075256,
+      "learning_rate": 4.419912346997539e-06,
+      "loss": 0.122,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7341218405703176,
+      "grad_norm": 0.8698350787162781,
+      "learning_rate": 4.4196322279000506e-06,
+      "loss": 0.1496,
+      "step": 4531
+    },
+    {
+      "epoch": 0.7342838626053143,
+      "grad_norm": 0.7942752242088318,
+      "learning_rate": 4.419352050066462e-06,
+      "loss": 0.1128,
+      "step": 4532
+    },
+    {
+      "epoch": 0.7344458846403111,
+      "grad_norm": 1.0515409708023071,
+      "learning_rate": 4.419071813505345e-06,
+      "loss": 0.165,
+      "step": 4533
+    },
+    {
+      "epoch": 0.7346079066753078,
+      "grad_norm": 0.9110780358314514,
+      "learning_rate": 4.418791518225275e-06,
+      "loss": 0.1305,
+      "step": 4534
+    },
+    {
+      "epoch": 0.7347699287103046,
+      "grad_norm": 0.7942949533462524,
+      "learning_rate": 4.4185111642348276e-06,
+      "loss": 0.1108,
+      "step": 4535
+    },
+    {
+      "epoch": 0.7349319507453014,
+      "grad_norm": 0.920792818069458,
+      "learning_rate": 4.418230751542581e-06,
+      "loss": 0.1577,
+      "step": 4536
+    },
+    {
+      "epoch": 0.7350939727802981,
+      "grad_norm": 1.1174674034118652,
+      "learning_rate": 4.417950280157115e-06,
+      "loss": 0.1558,
+      "step": 4537
+    },
+    {
+      "epoch": 0.7352559948152949,
+      "grad_norm": 0.9814813137054443,
+      "learning_rate": 4.417669750087014e-06,
+      "loss": 0.1552,
+      "step": 4538
+    },
+    {
+      "epoch": 0.7354180168502916,
+      "grad_norm": 0.7974295020103455,
+      "learning_rate": 4.417389161340857e-06,
+      "loss": 0.1274,
+      "step": 4539
+    },
+    {
+      "epoch": 0.7355800388852884,
+      "grad_norm": 0.9223327040672302,
+      "learning_rate": 4.417108513927233e-06,
+      "loss": 0.1425,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7357420609202852,
+      "grad_norm": 0.8545518517494202,
+      "learning_rate": 4.416827807854727e-06,
+      "loss": 0.1264,
+      "step": 4541
+    },
+    {
+      "epoch": 0.7359040829552819,
+      "grad_norm": 0.8425375819206238,
+      "learning_rate": 4.416547043131929e-06,
+      "loss": 0.1276,
+      "step": 4542
+    },
+    {
+      "epoch": 0.7360661049902787,
+      "grad_norm": 0.8444299697875977,
+      "learning_rate": 4.416266219767429e-06,
+      "loss": 0.1357,
+      "step": 4543
+    },
+    {
+      "epoch": 0.7362281270252754,
+      "grad_norm": 0.8412787914276123,
+      "learning_rate": 4.41598533776982e-06,
+      "loss": 0.135,
+      "step": 4544
+    },
+    {
+      "epoch": 0.7363901490602722,
+      "grad_norm": 0.9908415079116821,
+      "learning_rate": 4.415704397147698e-06,
+      "loss": 0.1655,
+      "step": 4545
+    },
+    {
+      "epoch": 0.7365521710952689,
+      "grad_norm": 0.8918123245239258,
+      "learning_rate": 4.415423397909655e-06,
+      "loss": 0.1396,
+      "step": 4546
+    },
+    {
+      "epoch": 0.7367141931302658,
+      "grad_norm": 0.8907253742218018,
+      "learning_rate": 4.4151423400642925e-06,
+      "loss": 0.1421,
+      "step": 4547
+    },
+    {
+      "epoch": 0.7368762151652625,
+      "grad_norm": 0.9191277027130127,
+      "learning_rate": 4.414861223620209e-06,
+      "loss": 0.1483,
+      "step": 4548
+    },
+    {
+      "epoch": 0.7370382372002592,
+      "grad_norm": 0.7574411034584045,
+      "learning_rate": 4.414580048586005e-06,
+      "loss": 0.1211,
+      "step": 4549
+    },
+    {
+      "epoch": 0.737200259235256,
+      "grad_norm": 0.7837011218070984,
+      "learning_rate": 4.414298814970286e-06,
+      "loss": 0.1293,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7373622812702527,
+      "grad_norm": 0.8395361304283142,
+      "learning_rate": 4.414017522781655e-06,
+      "loss": 0.1331,
+      "step": 4551
+    },
+    {
+      "epoch": 0.7375243033052495,
+      "grad_norm": 0.9138143658638,
+      "learning_rate": 4.41373617202872e-06,
+      "loss": 0.1533,
+      "step": 4552
+    },
+    {
+      "epoch": 0.7376863253402463,
+      "grad_norm": 0.8615904450416565,
+      "learning_rate": 4.413454762720088e-06,
+      "loss": 0.1347,
+      "step": 4553
+    },
+    {
+      "epoch": 0.7378483473752431,
+      "grad_norm": 0.8940994739532471,
+      "learning_rate": 4.413173294864373e-06,
+      "loss": 0.134,
+      "step": 4554
+    },
+    {
+      "epoch": 0.7380103694102398,
+      "grad_norm": 0.9082682132720947,
+      "learning_rate": 4.412891768470183e-06,
+      "loss": 0.1427,
+      "step": 4555
+    },
+    {
+      "epoch": 0.7381723914452365,
+      "grad_norm": 0.978569507598877,
+      "learning_rate": 4.412610183546135e-06,
+      "loss": 0.149,
+      "step": 4556
+    },
+    {
+      "epoch": 0.7383344134802333,
+      "grad_norm": 0.9564186930656433,
+      "learning_rate": 4.412328540100843e-06,
+      "loss": 0.1526,
+      "step": 4557
+    },
+    {
+      "epoch": 0.73849643551523,
+      "grad_norm": 0.756712794303894,
+      "learning_rate": 4.412046838142927e-06,
+      "loss": 0.1162,
+      "step": 4558
+    },
+    {
+      "epoch": 0.7386584575502269,
+      "grad_norm": 0.9216713309288025,
+      "learning_rate": 4.411765077681003e-06,
+      "loss": 0.1544,
+      "step": 4559
+    },
+    {
+      "epoch": 0.7388204795852236,
+      "grad_norm": 0.7456929087638855,
+      "learning_rate": 4.411483258723695e-06,
+      "loss": 0.1176,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7389825016202204,
+      "grad_norm": 0.8160392045974731,
+      "learning_rate": 4.411201381279625e-06,
+      "loss": 0.1369,
+      "step": 4561
+    },
+    {
+      "epoch": 0.7391445236552171,
+      "grad_norm": 0.8832926750183105,
+      "learning_rate": 4.410919445357418e-06,
+      "loss": 0.1453,
+      "step": 4562
+    },
+    {
+      "epoch": 0.7393065456902139,
+      "grad_norm": 0.9234940409660339,
+      "learning_rate": 4.410637450965699e-06,
+      "loss": 0.1416,
+      "step": 4563
+    },
+    {
+      "epoch": 0.7394685677252106,
+      "grad_norm": 0.9423860311508179,
+      "learning_rate": 4.410355398113099e-06,
+      "loss": 0.1502,
+      "step": 4564
+    },
+    {
+      "epoch": 0.7396305897602073,
+      "grad_norm": 0.9716334939002991,
+      "learning_rate": 4.410073286808247e-06,
+      "loss": 0.1462,
+      "step": 4565
+    },
+    {
+      "epoch": 0.7397926117952042,
+      "grad_norm": 0.8593721985816956,
+      "learning_rate": 4.409791117059773e-06,
+      "loss": 0.1287,
+      "step": 4566
+    },
+    {
+      "epoch": 0.7399546338302009,
+      "grad_norm": 0.9580432176589966,
+      "learning_rate": 4.409508888876313e-06,
+      "loss": 0.1527,
+      "step": 4567
+    },
+    {
+      "epoch": 0.7401166558651977,
+      "grad_norm": 0.8143693804740906,
+      "learning_rate": 4.409226602266503e-06,
+      "loss": 0.1272,
+      "step": 4568
+    },
+    {
+      "epoch": 0.7402786779001944,
+      "grad_norm": 0.8784132599830627,
+      "learning_rate": 4.408944257238979e-06,
+      "loss": 0.1321,
+      "step": 4569
+    },
+    {
+      "epoch": 0.7404406999351912,
+      "grad_norm": 0.9013129472732544,
+      "learning_rate": 4.408661853802379e-06,
+      "loss": 0.1524,
+      "step": 4570
+    },
+    {
+      "epoch": 0.740602721970188,
+      "grad_norm": 0.8545945286750793,
+      "learning_rate": 4.408379391965346e-06,
+      "loss": 0.1461,
+      "step": 4571
+    },
+    {
+      "epoch": 0.7407647440051847,
+      "grad_norm": 0.9265003800392151,
+      "learning_rate": 4.408096871736522e-06,
+      "loss": 0.1592,
+      "step": 4572
+    },
+    {
+      "epoch": 0.7409267660401815,
+      "grad_norm": 0.901270866394043,
+      "learning_rate": 4.407814293124551e-06,
+      "loss": 0.1615,
+      "step": 4573
+    },
+    {
+      "epoch": 0.7410887880751782,
+      "grad_norm": 0.9868318438529968,
+      "learning_rate": 4.407531656138079e-06,
+      "loss": 0.1498,
+      "step": 4574
+    },
+    {
+      "epoch": 0.741250810110175,
+      "grad_norm": 0.847172737121582,
+      "learning_rate": 4.407248960785756e-06,
+      "loss": 0.1408,
+      "step": 4575
+    },
+    {
+      "epoch": 0.7414128321451717,
+      "grad_norm": 0.941987156867981,
+      "learning_rate": 4.406966207076229e-06,
+      "loss": 0.1502,
+      "step": 4576
+    },
+    {
+      "epoch": 0.7415748541801686,
+      "grad_norm": 0.8444817662239075,
+      "learning_rate": 4.406683395018151e-06,
+      "loss": 0.1493,
+      "step": 4577
+    },
+    {
+      "epoch": 0.7417368762151653,
+      "grad_norm": 0.7697441577911377,
+      "learning_rate": 4.406400524620174e-06,
+      "loss": 0.1207,
+      "step": 4578
+    },
+    {
+      "epoch": 0.741898898250162,
+      "grad_norm": 0.8291786909103394,
+      "learning_rate": 4.406117595890956e-06,
+      "loss": 0.1369,
+      "step": 4579
+    },
+    {
+      "epoch": 0.7420609202851588,
+      "grad_norm": 0.8335320353507996,
+      "learning_rate": 4.405834608839152e-06,
+      "loss": 0.1313,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7422229423201555,
+      "grad_norm": 0.8066971302032471,
+      "learning_rate": 4.405551563473421e-06,
+      "loss": 0.1419,
+      "step": 4581
+    },
+    {
+      "epoch": 0.7423849643551523,
+      "grad_norm": 0.9154806137084961,
+      "learning_rate": 4.405268459802423e-06,
+      "loss": 0.1421,
+      "step": 4582
+    },
+    {
+      "epoch": 0.742546986390149,
+      "grad_norm": 0.8322879672050476,
+      "learning_rate": 4.404985297834821e-06,
+      "loss": 0.1335,
+      "step": 4583
+    },
+    {
+      "epoch": 0.7427090084251459,
+      "grad_norm": 0.9182136058807373,
+      "learning_rate": 4.404702077579279e-06,
+      "loss": 0.1531,
+      "step": 4584
+    },
+    {
+      "epoch": 0.7428710304601426,
+      "grad_norm": 0.9670420289039612,
+      "learning_rate": 4.404418799044463e-06,
+      "loss": 0.1728,
+      "step": 4585
+    },
+    {
+      "epoch": 0.7430330524951393,
+      "grad_norm": 1.0733063220977783,
+      "learning_rate": 4.4041354622390395e-06,
+      "loss": 0.1469,
+      "step": 4586
+    },
+    {
+      "epoch": 0.7431950745301361,
+      "grad_norm": 0.7664371132850647,
+      "learning_rate": 4.40385206717168e-06,
+      "loss": 0.1228,
+      "step": 4587
+    },
+    {
+      "epoch": 0.7433570965651328,
+      "grad_norm": 0.8765845894813538,
+      "learning_rate": 4.403568613851054e-06,
+      "loss": 0.1436,
+      "step": 4588
+    },
+    {
+      "epoch": 0.7435191186001296,
+      "grad_norm": 0.9102659225463867,
+      "learning_rate": 4.403285102285835e-06,
+      "loss": 0.148,
+      "step": 4589
+    },
+    {
+      "epoch": 0.7436811406351264,
+      "grad_norm": 1.0098685026168823,
+      "learning_rate": 4.403001532484697e-06,
+      "loss": 0.1589,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7438431626701232,
+      "grad_norm": 0.912421703338623,
+      "learning_rate": 4.402717904456318e-06,
+      "loss": 0.1474,
+      "step": 4591
+    },
+    {
+      "epoch": 0.7440051847051199,
+      "grad_norm": 0.960193932056427,
+      "learning_rate": 4.4024342182093745e-06,
+      "loss": 0.1535,
+      "step": 4592
+    },
+    {
+      "epoch": 0.7441672067401166,
+      "grad_norm": 0.8692451119422913,
+      "learning_rate": 4.402150473752549e-06,
+      "loss": 0.1506,
+      "step": 4593
+    },
+    {
+      "epoch": 0.7443292287751134,
+      "grad_norm": 0.9402160048484802,
+      "learning_rate": 4.401866671094522e-06,
+      "loss": 0.156,
+      "step": 4594
+    },
+    {
+      "epoch": 0.7444912508101101,
+      "grad_norm": 0.8891216516494751,
+      "learning_rate": 4.401582810243977e-06,
+      "loss": 0.1464,
+      "step": 4595
+    },
+    {
+      "epoch": 0.744653272845107,
+      "grad_norm": 0.9088370203971863,
+      "learning_rate": 4.4012988912096e-06,
+      "loss": 0.137,
+      "step": 4596
+    },
+    {
+      "epoch": 0.7448152948801037,
+      "grad_norm": 0.8495668172836304,
+      "learning_rate": 4.401014914000078e-06,
+      "loss": 0.1337,
+      "step": 4597
+    },
+    {
+      "epoch": 0.7449773169151005,
+      "grad_norm": 0.8261443376541138,
+      "learning_rate": 4.4007308786241e-06,
+      "loss": 0.1217,
+      "step": 4598
+    },
+    {
+      "epoch": 0.7451393389500972,
+      "grad_norm": 0.8053951263427734,
+      "learning_rate": 4.400446785090356e-06,
+      "loss": 0.1249,
+      "step": 4599
+    },
+    {
+      "epoch": 0.7453013609850939,
+      "grad_norm": 0.8850007653236389,
+      "learning_rate": 4.40016263340754e-06,
+      "loss": 0.1391,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7454633830200907,
+      "grad_norm": 0.8763007521629333,
+      "learning_rate": 4.399878423584345e-06,
+      "loss": 0.1301,
+      "step": 4601
+    },
+    {
+      "epoch": 0.7456254050550875,
+      "grad_norm": 0.9268038272857666,
+      "learning_rate": 4.399594155629469e-06,
+      "loss": 0.135,
+      "step": 4602
+    },
+    {
+      "epoch": 0.7457874270900843,
+      "grad_norm": 0.8517520427703857,
+      "learning_rate": 4.3993098295516085e-06,
+      "loss": 0.1419,
+      "step": 4603
+    },
+    {
+      "epoch": 0.745949449125081,
+      "grad_norm": 0.8787991404533386,
+      "learning_rate": 4.3990254453594634e-06,
+      "loss": 0.1233,
+      "step": 4604
+    },
+    {
+      "epoch": 0.7461114711600778,
+      "grad_norm": 0.9639225602149963,
+      "learning_rate": 4.398741003061735e-06,
+      "loss": 0.1407,
+      "step": 4605
+    },
+    {
+      "epoch": 0.7462734931950745,
+      "grad_norm": 0.8817579746246338,
+      "learning_rate": 4.398456502667127e-06,
+      "loss": 0.134,
+      "step": 4606
+    },
+    {
+      "epoch": 0.7464355152300713,
+      "grad_norm": 0.9382551908493042,
+      "learning_rate": 4.398171944184344e-06,
+      "loss": 0.1614,
+      "step": 4607
+    },
+    {
+      "epoch": 0.7465975372650681,
+      "grad_norm": 0.917212963104248,
+      "learning_rate": 4.397887327622093e-06,
+      "loss": 0.1304,
+      "step": 4608
+    },
+    {
+      "epoch": 0.7467595593000648,
+      "grad_norm": 1.028524398803711,
+      "learning_rate": 4.397602652989083e-06,
+      "loss": 0.1568,
+      "step": 4609
+    },
+    {
+      "epoch": 0.7469215813350616,
+      "grad_norm": 0.9571022391319275,
+      "learning_rate": 4.397317920294023e-06,
+      "loss": 0.1503,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7470836033700583,
+      "grad_norm": 0.9256038069725037,
+      "learning_rate": 4.397033129545627e-06,
+      "loss": 0.1341,
+      "step": 4611
+    },
+    {
+      "epoch": 0.7472456254050551,
+      "grad_norm": 0.8719132542610168,
+      "learning_rate": 4.396748280752608e-06,
+      "loss": 0.1495,
+      "step": 4612
+    },
+    {
+      "epoch": 0.7474076474400518,
+      "grad_norm": 0.8998635411262512,
+      "learning_rate": 4.39646337392368e-06,
+      "loss": 0.1421,
+      "step": 4613
+    },
+    {
+      "epoch": 0.7475696694750487,
+      "grad_norm": 0.9811137914657593,
+      "learning_rate": 4.396178409067564e-06,
+      "loss": 0.1531,
+      "step": 4614
+    },
+    {
+      "epoch": 0.7477316915100454,
+      "grad_norm": 0.962475061416626,
+      "learning_rate": 4.395893386192976e-06,
+      "loss": 0.1606,
+      "step": 4615
+    },
+    {
+      "epoch": 0.7478937135450421,
+      "grad_norm": 0.8668140172958374,
+      "learning_rate": 4.395608305308639e-06,
+      "loss": 0.1432,
+      "step": 4616
+    },
+    {
+      "epoch": 0.7480557355800389,
+      "grad_norm": 1.009864330291748,
+      "learning_rate": 4.3953231664232755e-06,
+      "loss": 0.1518,
+      "step": 4617
+    },
+    {
+      "epoch": 0.7482177576150356,
+      "grad_norm": 1.0338926315307617,
+      "learning_rate": 4.395037969545609e-06,
+      "loss": 0.1602,
+      "step": 4618
+    },
+    {
+      "epoch": 0.7483797796500324,
+      "grad_norm": 0.937615692615509,
+      "learning_rate": 4.394752714684367e-06,
+      "loss": 0.1508,
+      "step": 4619
+    },
+    {
+      "epoch": 0.7485418016850292,
+      "grad_norm": 1.010367512702942,
+      "learning_rate": 4.394467401848277e-06,
+      "loss": 0.1512,
+      "step": 4620
+    },
+    {
+      "epoch": 0.748703823720026,
+      "grad_norm": 0.8209222555160522,
+      "learning_rate": 4.394182031046069e-06,
+      "loss": 0.1412,
+      "step": 4621
+    },
+    {
+      "epoch": 0.7488658457550227,
+      "grad_norm": 0.8631600141525269,
+      "learning_rate": 4.393896602286475e-06,
+      "loss": 0.1371,
+      "step": 4622
+    },
+    {
+      "epoch": 0.7490278677900194,
+      "grad_norm": 0.7968021631240845,
+      "learning_rate": 4.393611115578228e-06,
+      "loss": 0.1263,
+      "step": 4623
+    },
+    {
+      "epoch": 0.7491898898250162,
+      "grad_norm": 0.9319000244140625,
+      "learning_rate": 4.3933255709300635e-06,
+      "loss": 0.158,
+      "step": 4624
+    },
+    {
+      "epoch": 0.7493519118600129,
+      "grad_norm": 0.8134043216705322,
+      "learning_rate": 4.393039968350718e-06,
+      "loss": 0.1389,
+      "step": 4625
+    },
+    {
+      "epoch": 0.7495139338950098,
+      "grad_norm": 0.968559205532074,
+      "learning_rate": 4.3927543078489295e-06,
+      "loss": 0.1761,
+      "step": 4626
+    },
+    {
+      "epoch": 0.7496759559300065,
+      "grad_norm": 0.8765543103218079,
+      "learning_rate": 4.392468589433441e-06,
+      "loss": 0.1496,
+      "step": 4627
+    },
+    {
+      "epoch": 0.7498379779650033,
+      "grad_norm": 0.9812090992927551,
+      "learning_rate": 4.392182813112993e-06,
+      "loss": 0.1566,
+      "step": 4628
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9258951544761658,
+      "learning_rate": 4.3918969788963295e-06,
+      "loss": 0.1479,
+      "step": 4629
+    },
+    {
+      "epoch": 0.7501620220349967,
+      "grad_norm": 0.8851484656333923,
+      "learning_rate": 4.391611086792198e-06,
+      "loss": 0.1463,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7503240440699935,
+      "grad_norm": 1.0019028186798096,
+      "learning_rate": 4.391325136809344e-06,
+      "loss": 0.1457,
+      "step": 4631
+    },
+    {
+      "epoch": 0.7504860661049902,
+      "grad_norm": 0.8665192723274231,
+      "learning_rate": 4.391039128956517e-06,
+      "loss": 0.1279,
+      "step": 4632
+    },
+    {
+      "epoch": 0.7506480881399871,
+      "grad_norm": 0.8466840982437134,
+      "learning_rate": 4.39075306324247e-06,
+      "loss": 0.1273,
+      "step": 4633
+    },
+    {
+      "epoch": 0.7508101101749838,
+      "grad_norm": 0.9541529417037964,
+      "learning_rate": 4.390466939675954e-06,
+      "loss": 0.1667,
+      "step": 4634
+    },
+    {
+      "epoch": 0.7509721322099806,
+      "grad_norm": 1.0194523334503174,
+      "learning_rate": 4.390180758265725e-06,
+      "loss": 0.1509,
+      "step": 4635
+    },
+    {
+      "epoch": 0.7511341542449773,
+      "grad_norm": 0.9222609996795654,
+      "learning_rate": 4.389894519020539e-06,
+      "loss": 0.1455,
+      "step": 4636
+    },
+    {
+      "epoch": 0.751296176279974,
+      "grad_norm": 0.9126465320587158,
+      "learning_rate": 4.389608221949153e-06,
+      "loss": 0.1303,
+      "step": 4637
+    },
+    {
+      "epoch": 0.7514581983149708,
+      "grad_norm": 0.8073568344116211,
+      "learning_rate": 4.38932186706033e-06,
+      "loss": 0.1254,
+      "step": 4638
+    },
+    {
+      "epoch": 0.7516202203499676,
+      "grad_norm": 0.9920159578323364,
+      "learning_rate": 4.389035454362829e-06,
+      "loss": 0.1476,
+      "step": 4639
+    },
+    {
+      "epoch": 0.7517822423849644,
+      "grad_norm": 0.8352819085121155,
+      "learning_rate": 4.388748983865414e-06,
+      "loss": 0.1289,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7519442644199611,
+      "grad_norm": 0.9489824175834656,
+      "learning_rate": 4.388462455576852e-06,
+      "loss": 0.1533,
+      "step": 4641
+    },
+    {
+      "epoch": 0.7521062864549579,
+      "grad_norm": 0.9182109832763672,
+      "learning_rate": 4.388175869505908e-06,
+      "loss": 0.1494,
+      "step": 4642
+    },
+    {
+      "epoch": 0.7522683084899546,
+      "grad_norm": 0.9489102959632874,
+      "learning_rate": 4.387889225661352e-06,
+      "loss": 0.1317,
+      "step": 4643
+    },
+    {
+      "epoch": 0.7524303305249513,
+      "grad_norm": 0.887995183467865,
+      "learning_rate": 4.387602524051954e-06,
+      "loss": 0.1433,
+      "step": 4644
+    },
+    {
+      "epoch": 0.7525923525599482,
+      "grad_norm": 0.934482753276825,
+      "learning_rate": 4.387315764686487e-06,
+      "loss": 0.1458,
+      "step": 4645
+    },
+    {
+      "epoch": 0.7527543745949449,
+      "grad_norm": 0.757559061050415,
+      "learning_rate": 4.387028947573724e-06,
+      "loss": 0.1199,
+      "step": 4646
+    },
+    {
+      "epoch": 0.7529163966299417,
+      "grad_norm": 0.8067179322242737,
+      "learning_rate": 4.386742072722443e-06,
+      "loss": 0.1321,
+      "step": 4647
+    },
+    {
+      "epoch": 0.7530784186649384,
+      "grad_norm": 0.9727403521537781,
+      "learning_rate": 4.3864551401414195e-06,
+      "loss": 0.1172,
+      "step": 4648
+    },
+    {
+      "epoch": 0.7532404406999352,
+      "grad_norm": 0.9205012917518616,
+      "learning_rate": 4.386168149839434e-06,
+      "loss": 0.1555,
+      "step": 4649
+    },
+    {
+      "epoch": 0.7534024627349319,
+      "grad_norm": 0.8873549103736877,
+      "learning_rate": 4.385881101825268e-06,
+      "loss": 0.1476,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7535644847699287,
+      "grad_norm": 0.9281837940216064,
+      "learning_rate": 4.3855939961077034e-06,
+      "loss": 0.1385,
+      "step": 4651
+    },
+    {
+      "epoch": 0.7537265068049255,
+      "grad_norm": 1.0171685218811035,
+      "learning_rate": 4.385306832695526e-06,
+      "loss": 0.1635,
+      "step": 4652
+    },
+    {
+      "epoch": 0.7538885288399222,
+      "grad_norm": 0.8333146572113037,
+      "learning_rate": 4.385019611597522e-06,
+      "loss": 0.1398,
+      "step": 4653
+    },
+    {
+      "epoch": 0.754050550874919,
+      "grad_norm": 0.9079596400260925,
+      "learning_rate": 4.384732332822479e-06,
+      "loss": 0.1396,
+      "step": 4654
+    },
+    {
+      "epoch": 0.7542125729099157,
+      "grad_norm": 0.8788375854492188,
+      "learning_rate": 4.384444996379188e-06,
+      "loss": 0.1384,
+      "step": 4655
+    },
+    {
+      "epoch": 0.7543745949449125,
+      "grad_norm": 0.8239230513572693,
+      "learning_rate": 4.38415760227644e-06,
+      "loss": 0.1278,
+      "step": 4656
+    },
+    {
+      "epoch": 0.7545366169799093,
+      "grad_norm": 0.8959946036338806,
+      "learning_rate": 4.383870150523029e-06,
+      "loss": 0.1499,
+      "step": 4657
+    },
+    {
+      "epoch": 0.7546986390149061,
+      "grad_norm": 0.8564679026603699,
+      "learning_rate": 4.38358264112775e-06,
+      "loss": 0.1381,
+      "step": 4658
+    },
+    {
+      "epoch": 0.7548606610499028,
+      "grad_norm": 0.8459985256195068,
+      "learning_rate": 4.383295074099402e-06,
+      "loss": 0.1358,
+      "step": 4659
+    },
+    {
+      "epoch": 0.7550226830848995,
+      "grad_norm": 0.9368695020675659,
+      "learning_rate": 4.3830074494467815e-06,
+      "loss": 0.1419,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7551847051198963,
+      "grad_norm": 0.8994658589363098,
+      "learning_rate": 4.382719767178689e-06,
+      "loss": 0.1558,
+      "step": 4661
+    },
+    {
+      "epoch": 0.755346727154893,
+      "grad_norm": 0.9667396545410156,
+      "learning_rate": 4.382432027303928e-06,
+      "loss": 0.1627,
+      "step": 4662
+    },
+    {
+      "epoch": 0.7555087491898899,
+      "grad_norm": 0.8609565496444702,
+      "learning_rate": 4.382144229831302e-06,
+      "loss": 0.1351,
+      "step": 4663
+    },
+    {
+      "epoch": 0.7556707712248866,
+      "grad_norm": 0.845554769039154,
+      "learning_rate": 4.381856374769617e-06,
+      "loss": 0.1273,
+      "step": 4664
+    },
+    {
+      "epoch": 0.7558327932598834,
+      "grad_norm": 0.9373748302459717,
+      "learning_rate": 4.3815684621276824e-06,
+      "loss": 0.1663,
+      "step": 4665
+    },
+    {
+      "epoch": 0.7559948152948801,
+      "grad_norm": 0.8447197079658508,
+      "learning_rate": 4.3812804919143055e-06,
+      "loss": 0.1337,
+      "step": 4666
+    },
+    {
+      "epoch": 0.7561568373298768,
+      "grad_norm": 0.8827223777770996,
+      "learning_rate": 4.380992464138298e-06,
+      "loss": 0.1419,
+      "step": 4667
+    },
+    {
+      "epoch": 0.7563188593648736,
+      "grad_norm": 0.8952583074569702,
+      "learning_rate": 4.380704378808473e-06,
+      "loss": 0.1249,
+      "step": 4668
+    },
+    {
+      "epoch": 0.7564808813998704,
+      "grad_norm": 0.8964098691940308,
+      "learning_rate": 4.380416235933646e-06,
+      "loss": 0.1544,
+      "step": 4669
+    },
+    {
+      "epoch": 0.7566429034348672,
+      "grad_norm": 1.0078829526901245,
+      "learning_rate": 4.380128035522632e-06,
+      "loss": 0.1629,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7568049254698639,
+      "grad_norm": 0.9272462725639343,
+      "learning_rate": 4.379839777584249e-06,
+      "loss": 0.158,
+      "step": 4671
+    },
+    {
+      "epoch": 0.7569669475048607,
+      "grad_norm": 0.9298140406608582,
+      "learning_rate": 4.379551462127319e-06,
+      "loss": 0.1442,
+      "step": 4672
+    },
+    {
+      "epoch": 0.7571289695398574,
+      "grad_norm": 1.0363284349441528,
+      "learning_rate": 4.3792630891606635e-06,
+      "loss": 0.1739,
+      "step": 4673
+    },
+    {
+      "epoch": 0.7572909915748541,
+      "grad_norm": 0.9505801796913147,
+      "learning_rate": 4.3789746586931034e-06,
+      "loss": 0.1657,
+      "step": 4674
+    },
+    {
+      "epoch": 0.757453013609851,
+      "grad_norm": 1.0468703508377075,
+      "learning_rate": 4.3786861707334676e-06,
+      "loss": 0.1587,
+      "step": 4675
+    },
+    {
+      "epoch": 0.7576150356448477,
+      "grad_norm": 0.8523300886154175,
+      "learning_rate": 4.37839762529058e-06,
+      "loss": 0.1465,
+      "step": 4676
+    },
+    {
+      "epoch": 0.7577770576798445,
+      "grad_norm": 0.9114488363265991,
+      "learning_rate": 4.378109022373272e-06,
+      "loss": 0.127,
+      "step": 4677
+    },
+    {
+      "epoch": 0.7579390797148412,
+      "grad_norm": 1.036751627922058,
+      "learning_rate": 4.3778203619903716e-06,
+      "loss": 0.1341,
+      "step": 4678
+    },
+    {
+      "epoch": 0.758101101749838,
+      "grad_norm": 1.0055261850357056,
+      "learning_rate": 4.377531644150712e-06,
+      "loss": 0.1549,
+      "step": 4679
+    },
+    {
+      "epoch": 0.7582631237848347,
+      "grad_norm": 1.0141489505767822,
+      "learning_rate": 4.3772428688631285e-06,
+      "loss": 0.1552,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7584251458198314,
+      "grad_norm": 0.9310604929924011,
+      "learning_rate": 4.376954036136456e-06,
+      "loss": 0.1502,
+      "step": 4681
+    },
+    {
+      "epoch": 0.7585871678548283,
+      "grad_norm": 0.7873426675796509,
+      "learning_rate": 4.376665145979532e-06,
+      "loss": 0.1249,
+      "step": 4682
+    },
+    {
+      "epoch": 0.758749189889825,
+      "grad_norm": 0.8380732536315918,
+      "learning_rate": 4.376376198401195e-06,
+      "loss": 0.154,
+      "step": 4683
+    },
+    {
+      "epoch": 0.7589112119248218,
+      "grad_norm": 0.762469470500946,
+      "learning_rate": 4.376087193410289e-06,
+      "loss": 0.1214,
+      "step": 4684
+    },
+    {
+      "epoch": 0.7590732339598185,
+      "grad_norm": 0.868675947189331,
+      "learning_rate": 4.375798131015654e-06,
+      "loss": 0.1446,
+      "step": 4685
+    },
+    {
+      "epoch": 0.7592352559948153,
+      "grad_norm": 0.998455286026001,
+      "learning_rate": 4.375509011226135e-06,
+      "loss": 0.1611,
+      "step": 4686
+    },
+    {
+      "epoch": 0.759397278029812,
+      "grad_norm": 1.0500266551971436,
+      "learning_rate": 4.3752198340505795e-06,
+      "loss": 0.1398,
+      "step": 4687
+    },
+    {
+      "epoch": 0.7595593000648088,
+      "grad_norm": 0.8594135642051697,
+      "learning_rate": 4.374930599497835e-06,
+      "loss": 0.1354,
+      "step": 4688
+    },
+    {
+      "epoch": 0.7597213220998056,
+      "grad_norm": 0.8801107406616211,
+      "learning_rate": 4.374641307576751e-06,
+      "loss": 0.1432,
+      "step": 4689
+    },
+    {
+      "epoch": 0.7598833441348023,
+      "grad_norm": 0.8573765158653259,
+      "learning_rate": 4.37435195829618e-06,
+      "loss": 0.1404,
+      "step": 4690
+    },
+    {
+      "epoch": 0.7600453661697991,
+      "grad_norm": 0.8854019045829773,
+      "learning_rate": 4.3740625516649755e-06,
+      "loss": 0.1406,
+      "step": 4691
+    },
+    {
+      "epoch": 0.7602073882047958,
+      "grad_norm": 0.899628758430481,
+      "learning_rate": 4.373773087691992e-06,
+      "loss": 0.1331,
+      "step": 4692
+    },
+    {
+      "epoch": 0.7603694102397927,
+      "grad_norm": 0.9409993886947632,
+      "learning_rate": 4.373483566386086e-06,
+      "loss": 0.1582,
+      "step": 4693
+    },
+    {
+      "epoch": 0.7605314322747894,
+      "grad_norm": 0.9930229783058167,
+      "learning_rate": 4.373193987756116e-06,
+      "loss": 0.1317,
+      "step": 4694
+    },
+    {
+      "epoch": 0.7606934543097861,
+      "grad_norm": 0.8028172254562378,
+      "learning_rate": 4.372904351810943e-06,
+      "loss": 0.1267,
+      "step": 4695
+    },
+    {
+      "epoch": 0.7608554763447829,
+      "grad_norm": 0.8876671195030212,
+      "learning_rate": 4.3726146585594296e-06,
+      "loss": 0.1482,
+      "step": 4696
+    },
+    {
+      "epoch": 0.7610174983797796,
+      "grad_norm": 0.8735641837120056,
+      "learning_rate": 4.3723249080104395e-06,
+      "loss": 0.1362,
+      "step": 4697
+    },
+    {
+      "epoch": 0.7611795204147764,
+      "grad_norm": 1.0168266296386719,
+      "learning_rate": 4.372035100172838e-06,
+      "loss": 0.1535,
+      "step": 4698
+    },
+    {
+      "epoch": 0.7613415424497731,
+      "grad_norm": 0.9453420042991638,
+      "learning_rate": 4.371745235055492e-06,
+      "loss": 0.1625,
+      "step": 4699
+    },
+    {
+      "epoch": 0.76150356448477,
+      "grad_norm": 0.8696652054786682,
+      "learning_rate": 4.371455312667272e-06,
+      "loss": 0.1469,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7616655865197667,
+      "grad_norm": 0.8512628078460693,
+      "learning_rate": 4.371165333017049e-06,
+      "loss": 0.1464,
+      "step": 4701
+    },
+    {
+      "epoch": 0.7618276085547635,
+      "grad_norm": 0.9412540793418884,
+      "learning_rate": 4.370875296113694e-06,
+      "loss": 0.1495,
+      "step": 4702
+    },
+    {
+      "epoch": 0.7619896305897602,
+      "grad_norm": 0.9683601260185242,
+      "learning_rate": 4.370585201966082e-06,
+      "loss": 0.1696,
+      "step": 4703
+    },
+    {
+      "epoch": 0.7621516526247569,
+      "grad_norm": 0.9484912753105164,
+      "learning_rate": 4.370295050583091e-06,
+      "loss": 0.162,
+      "step": 4704
+    },
+    {
+      "epoch": 0.7623136746597537,
+      "grad_norm": 0.8469820022583008,
+      "learning_rate": 4.370004841973596e-06,
+      "loss": 0.1332,
+      "step": 4705
+    },
+    {
+      "epoch": 0.7624756966947505,
+      "grad_norm": 0.9001607894897461,
+      "learning_rate": 4.3697145761464785e-06,
+      "loss": 0.1468,
+      "step": 4706
+    },
+    {
+      "epoch": 0.7626377187297473,
+      "grad_norm": 0.8998278975486755,
+      "learning_rate": 4.36942425311062e-06,
+      "loss": 0.1493,
+      "step": 4707
+    },
+    {
+      "epoch": 0.762799740764744,
+      "grad_norm": 0.9656628966331482,
+      "learning_rate": 4.369133872874903e-06,
+      "loss": 0.1528,
+      "step": 4708
+    },
+    {
+      "epoch": 0.7629617627997408,
+      "grad_norm": 1.009851098060608,
+      "learning_rate": 4.368843435448213e-06,
+      "loss": 0.1392,
+      "step": 4709
+    },
+    {
+      "epoch": 0.7631237848347375,
+      "grad_norm": 0.9109621644020081,
+      "learning_rate": 4.368552940839436e-06,
+      "loss": 0.1571,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7632858068697342,
+      "grad_norm": 0.9232974052429199,
+      "learning_rate": 4.368262389057462e-06,
+      "loss": 0.1551,
+      "step": 4711
+    },
+    {
+      "epoch": 0.7634478289047311,
+      "grad_norm": 0.8760313987731934,
+      "learning_rate": 4.367971780111179e-06,
+      "loss": 0.1587,
+      "step": 4712
+    },
+    {
+      "epoch": 0.7636098509397278,
+      "grad_norm": 0.945943295955658,
+      "learning_rate": 4.36768111400948e-06,
+      "loss": 0.1431,
+      "step": 4713
+    },
+    {
+      "epoch": 0.7637718729747246,
+      "grad_norm": 0.8429069519042969,
+      "learning_rate": 4.367390390761258e-06,
+      "loss": 0.1377,
+      "step": 4714
+    },
+    {
+      "epoch": 0.7639338950097213,
+      "grad_norm": 0.8647720813751221,
+      "learning_rate": 4.367099610375409e-06,
+      "loss": 0.1262,
+      "step": 4715
+    },
+    {
+      "epoch": 0.7640959170447181,
+      "grad_norm": 0.8221338987350464,
+      "learning_rate": 4.3668087728608314e-06,
+      "loss": 0.1269,
+      "step": 4716
+    },
+    {
+      "epoch": 0.7642579390797148,
+      "grad_norm": 0.8592271208763123,
+      "learning_rate": 4.366517878226423e-06,
+      "loss": 0.1344,
+      "step": 4717
+    },
+    {
+      "epoch": 0.7644199611147116,
+      "grad_norm": 0.8569756150245667,
+      "learning_rate": 4.366226926481083e-06,
+      "loss": 0.1417,
+      "step": 4718
+    },
+    {
+      "epoch": 0.7645819831497084,
+      "grad_norm": 0.8143072128295898,
+      "learning_rate": 4.365935917633716e-06,
+      "loss": 0.1295,
+      "step": 4719
+    },
+    {
+      "epoch": 0.7647440051847051,
+      "grad_norm": 0.9709054827690125,
+      "learning_rate": 4.365644851693226e-06,
+      "loss": 0.1597,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7649060272197019,
+      "grad_norm": 0.8787278532981873,
+      "learning_rate": 4.365353728668518e-06,
+      "loss": 0.14,
+      "step": 4721
+    },
+    {
+      "epoch": 0.7650680492546986,
+      "grad_norm": 0.945436418056488,
+      "learning_rate": 4.3650625485685e-06,
+      "loss": 0.153,
+      "step": 4722
+    },
+    {
+      "epoch": 0.7652300712896954,
+      "grad_norm": 0.9423735737800598,
+      "learning_rate": 4.3647713114020805e-06,
+      "loss": 0.1324,
+      "step": 4723
+    },
+    {
+      "epoch": 0.7653920933246922,
+      "grad_norm": 0.9254855513572693,
+      "learning_rate": 4.364480017178172e-06,
+      "loss": 0.149,
+      "step": 4724
+    },
+    {
+      "epoch": 0.7655541153596889,
+      "grad_norm": 0.8730553984642029,
+      "learning_rate": 4.364188665905687e-06,
+      "loss": 0.1386,
+      "step": 4725
+    },
+    {
+      "epoch": 0.7657161373946857,
+      "grad_norm": 0.8605020046234131,
+      "learning_rate": 4.36389725759354e-06,
+      "loss": 0.1381,
+      "step": 4726
+    },
+    {
+      "epoch": 0.7658781594296824,
+      "grad_norm": 1.0751097202301025,
+      "learning_rate": 4.363605792250648e-06,
+      "loss": 0.1849,
+      "step": 4727
+    },
+    {
+      "epoch": 0.7660401814646792,
+      "grad_norm": 0.9074150919914246,
+      "learning_rate": 4.363314269885928e-06,
+      "loss": 0.1377,
+      "step": 4728
+    },
+    {
+      "epoch": 0.7662022034996759,
+      "grad_norm": 0.8164258599281311,
+      "learning_rate": 4.363022690508301e-06,
+      "loss": 0.1357,
+      "step": 4729
+    },
+    {
+      "epoch": 0.7663642255346728,
+      "grad_norm": 0.9147708415985107,
+      "learning_rate": 4.362731054126687e-06,
+      "loss": 0.1494,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7665262475696695,
+      "grad_norm": 0.9929783344268799,
+      "learning_rate": 4.362439360750012e-06,
+      "loss": 0.162,
+      "step": 4731
+    },
+    {
+      "epoch": 0.7666882696046662,
+      "grad_norm": 0.977254331111908,
+      "learning_rate": 4.362147610387198e-06,
+      "loss": 0.1586,
+      "step": 4732
+    },
+    {
+      "epoch": 0.766850291639663,
+      "grad_norm": 0.9575954079627991,
+      "learning_rate": 4.361855803047175e-06,
+      "loss": 0.1561,
+      "step": 4733
+    },
+    {
+      "epoch": 0.7670123136746597,
+      "grad_norm": 0.8501836657524109,
+      "learning_rate": 4.361563938738869e-06,
+      "loss": 0.138,
+      "step": 4734
+    },
+    {
+      "epoch": 0.7671743357096565,
+      "grad_norm": 0.8874675035476685,
+      "learning_rate": 4.361272017471212e-06,
+      "loss": 0.1503,
+      "step": 4735
+    },
+    {
+      "epoch": 0.7673363577446533,
+      "grad_norm": 0.830808699131012,
+      "learning_rate": 4.3609800392531345e-06,
+      "loss": 0.1304,
+      "step": 4736
+    },
+    {
+      "epoch": 0.7674983797796501,
+      "grad_norm": 0.8549823760986328,
+      "learning_rate": 4.3606880040935714e-06,
+      "loss": 0.1475,
+      "step": 4737
+    },
+    {
+      "epoch": 0.7676604018146468,
+      "grad_norm": 0.9000952243804932,
+      "learning_rate": 4.36039591200146e-06,
+      "loss": 0.1549,
+      "step": 4738
+    },
+    {
+      "epoch": 0.7678224238496435,
+      "grad_norm": 0.9087912440299988,
+      "learning_rate": 4.360103762985734e-06,
+      "loss": 0.1516,
+      "step": 4739
+    },
+    {
+      "epoch": 0.7679844458846403,
+      "grad_norm": 0.9787373542785645,
+      "learning_rate": 4.359811557055335e-06,
+      "loss": 0.172,
+      "step": 4740
+    },
+    {
+      "epoch": 0.768146467919637,
+      "grad_norm": 0.8458462953567505,
+      "learning_rate": 4.359519294219201e-06,
+      "loss": 0.1462,
+      "step": 4741
+    },
+    {
+      "epoch": 0.7683084899546339,
+      "grad_norm": 0.8422055840492249,
+      "learning_rate": 4.3592269744862794e-06,
+      "loss": 0.1269,
+      "step": 4742
+    },
+    {
+      "epoch": 0.7684705119896306,
+      "grad_norm": 0.8210686445236206,
+      "learning_rate": 4.35893459786551e-06,
+      "loss": 0.1425,
+      "step": 4743
+    },
+    {
+      "epoch": 0.7686325340246274,
+      "grad_norm": 0.9287550449371338,
+      "learning_rate": 4.3586421643658404e-06,
+      "loss": 0.1574,
+      "step": 4744
+    },
+    {
+      "epoch": 0.7687945560596241,
+      "grad_norm": 0.9414339065551758,
+      "learning_rate": 4.3583496739962195e-06,
+      "loss": 0.1515,
+      "step": 4745
+    },
+    {
+      "epoch": 0.7689565780946209,
+      "grad_norm": 0.8130404353141785,
+      "learning_rate": 4.3580571267655945e-06,
+      "loss": 0.1309,
+      "step": 4746
+    },
+    {
+      "epoch": 0.7691186001296176,
+      "grad_norm": 0.7749109864234924,
+      "learning_rate": 4.357764522682919e-06,
+      "loss": 0.1363,
+      "step": 4747
+    },
+    {
+      "epoch": 0.7692806221646143,
+      "grad_norm": 0.874220609664917,
+      "learning_rate": 4.357471861757144e-06,
+      "loss": 0.1415,
+      "step": 4748
+    },
+    {
+      "epoch": 0.7694426441996112,
+      "grad_norm": 0.9805276393890381,
+      "learning_rate": 4.357179143997225e-06,
+      "loss": 0.1563,
+      "step": 4749
+    },
+    {
+      "epoch": 0.7696046662346079,
+      "grad_norm": 0.9434357285499573,
+      "learning_rate": 4.3568863694121185e-06,
+      "loss": 0.1492,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7697666882696047,
+      "grad_norm": 0.968651294708252,
+      "learning_rate": 4.356593538010783e-06,
+      "loss": 0.1611,
+      "step": 4751
+    },
+    {
+      "epoch": 0.7699287103046014,
+      "grad_norm": 0.7928199172019958,
+      "learning_rate": 4.356300649802178e-06,
+      "loss": 0.1264,
+      "step": 4752
+    },
+    {
+      "epoch": 0.7700907323395982,
+      "grad_norm": 0.8838950991630554,
+      "learning_rate": 4.356007704795265e-06,
+      "loss": 0.142,
+      "step": 4753
+    },
+    {
+      "epoch": 0.770252754374595,
+      "grad_norm": 0.8881464004516602,
+      "learning_rate": 4.355714702999008e-06,
+      "loss": 0.1438,
+      "step": 4754
+    },
+    {
+      "epoch": 0.7704147764095917,
+      "grad_norm": 0.938313364982605,
+      "learning_rate": 4.355421644422372e-06,
+      "loss": 0.1423,
+      "step": 4755
+    },
+    {
+      "epoch": 0.7705767984445885,
+      "grad_norm": 0.892318069934845,
+      "learning_rate": 4.355128529074323e-06,
+      "loss": 0.1451,
+      "step": 4756
+    },
+    {
+      "epoch": 0.7707388204795852,
+      "grad_norm": 0.8982798457145691,
+      "learning_rate": 4.354835356963831e-06,
+      "loss": 0.1375,
+      "step": 4757
+    },
+    {
+      "epoch": 0.770900842514582,
+      "grad_norm": 0.8019870519638062,
+      "learning_rate": 4.354542128099866e-06,
+      "loss": 0.1271,
+      "step": 4758
+    },
+    {
+      "epoch": 0.7710628645495787,
+      "grad_norm": 0.8931732773780823,
+      "learning_rate": 4.354248842491399e-06,
+      "loss": 0.1481,
+      "step": 4759
+    },
+    {
+      "epoch": 0.7712248865845756,
+      "grad_norm": 0.9127657413482666,
+      "learning_rate": 4.353955500147405e-06,
+      "loss": 0.1466,
+      "step": 4760
+    },
+    {
+      "epoch": 0.7713869086195723,
+      "grad_norm": 0.9457529783248901,
+      "learning_rate": 4.353662101076859e-06,
+      "loss": 0.1488,
+      "step": 4761
+    },
+    {
+      "epoch": 0.771548930654569,
+      "grad_norm": 0.9101020097732544,
+      "learning_rate": 4.353368645288738e-06,
+      "loss": 0.1464,
+      "step": 4762
+    },
+    {
+      "epoch": 0.7717109526895658,
+      "grad_norm": 0.9181843400001526,
+      "learning_rate": 4.353075132792023e-06,
+      "loss": 0.1379,
+      "step": 4763
+    },
+    {
+      "epoch": 0.7718729747245625,
+      "grad_norm": 0.8580861687660217,
+      "learning_rate": 4.352781563595691e-06,
+      "loss": 0.145,
+      "step": 4764
+    },
+    {
+      "epoch": 0.7720349967595593,
+      "grad_norm": 0.8708130717277527,
+      "learning_rate": 4.352487937708729e-06,
+      "loss": 0.1536,
+      "step": 4765
+    },
+    {
+      "epoch": 0.772197018794556,
+      "grad_norm": 0.8263076543807983,
+      "learning_rate": 4.352194255140118e-06,
+      "loss": 0.1327,
+      "step": 4766
+    },
+    {
+      "epoch": 0.7723590408295529,
+      "grad_norm": 0.9026300311088562,
+      "learning_rate": 4.351900515898846e-06,
+      "loss": 0.1477,
+      "step": 4767
+    },
+    {
+      "epoch": 0.7725210628645496,
+      "grad_norm": 0.9587908387184143,
+      "learning_rate": 4.351606719993899e-06,
+      "loss": 0.1537,
+      "step": 4768
+    },
+    {
+      "epoch": 0.7726830848995463,
+      "grad_norm": 0.78399258852005,
+      "learning_rate": 4.3513128674342665e-06,
+      "loss": 0.1308,
+      "step": 4769
+    },
+    {
+      "epoch": 0.7728451069345431,
+      "grad_norm": 0.8863502144813538,
+      "learning_rate": 4.351018958228941e-06,
+      "loss": 0.1467,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7730071289695398,
+      "grad_norm": 0.9153334498405457,
+      "learning_rate": 4.350724992386915e-06,
+      "loss": 0.1646,
+      "step": 4771
+    },
+    {
+      "epoch": 0.7731691510045366,
+      "grad_norm": 0.8299160003662109,
+      "learning_rate": 4.350430969917182e-06,
+      "loss": 0.1437,
+      "step": 4772
+    },
+    {
+      "epoch": 0.7733311730395334,
+      "grad_norm": 0.9863147139549255,
+      "learning_rate": 4.35013689082874e-06,
+      "loss": 0.1435,
+      "step": 4773
+    },
+    {
+      "epoch": 0.7734931950745302,
+      "grad_norm": 0.9008839130401611,
+      "learning_rate": 4.349842755130587e-06,
+      "loss": 0.1472,
+      "step": 4774
+    },
+    {
+      "epoch": 0.7736552171095269,
+      "grad_norm": 0.9455453157424927,
+      "learning_rate": 4.349548562831721e-06,
+      "loss": 0.1493,
+      "step": 4775
+    },
+    {
+      "epoch": 0.7738172391445236,
+      "grad_norm": 0.9574447870254517,
+      "learning_rate": 4.349254313941146e-06,
+      "loss": 0.1572,
+      "step": 4776
+    },
+    {
+      "epoch": 0.7739792611795204,
+      "grad_norm": 0.9903965592384338,
+      "learning_rate": 4.348960008467863e-06,
+      "loss": 0.1688,
+      "step": 4777
+    },
+    {
+      "epoch": 0.7741412832145171,
+      "grad_norm": 0.8893287181854248,
+      "learning_rate": 4.3486656464208785e-06,
+      "loss": 0.142,
+      "step": 4778
+    },
+    {
+      "epoch": 0.774303305249514,
+      "grad_norm": 0.8900960087776184,
+      "learning_rate": 4.348371227809199e-06,
+      "loss": 0.1408,
+      "step": 4779
+    },
+    {
+      "epoch": 0.7744653272845107,
+      "grad_norm": 0.9652172923088074,
+      "learning_rate": 4.348076752641834e-06,
+      "loss": 0.1675,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7746273493195075,
+      "grad_norm": 0.8140687346458435,
+      "learning_rate": 4.34778222092779e-06,
+      "loss": 0.1229,
+      "step": 4781
+    },
+    {
+      "epoch": 0.7747893713545042,
+      "grad_norm": 0.7977285385131836,
+      "learning_rate": 4.347487632676084e-06,
+      "loss": 0.122,
+      "step": 4782
+    },
+    {
+      "epoch": 0.7749513933895009,
+      "grad_norm": 0.8936629891395569,
+      "learning_rate": 4.347192987895726e-06,
+      "loss": 0.147,
+      "step": 4783
+    },
+    {
+      "epoch": 0.7751134154244977,
+      "grad_norm": 0.811756432056427,
+      "learning_rate": 4.346898286595733e-06,
+      "loss": 0.126,
+      "step": 4784
+    },
+    {
+      "epoch": 0.7752754374594945,
+      "grad_norm": 0.8740372657775879,
+      "learning_rate": 4.346603528785122e-06,
+      "loss": 0.1515,
+      "step": 4785
+    },
+    {
+      "epoch": 0.7754374594944913,
+      "grad_norm": 0.8331345319747925,
+      "learning_rate": 4.3463087144729115e-06,
+      "loss": 0.1341,
+      "step": 4786
+    },
+    {
+      "epoch": 0.775599481529488,
+      "grad_norm": 0.8488640785217285,
+      "learning_rate": 4.346013843668122e-06,
+      "loss": 0.1361,
+      "step": 4787
+    },
+    {
+      "epoch": 0.7757615035644848,
+      "grad_norm": 0.9174832701683044,
+      "learning_rate": 4.3457189163797776e-06,
+      "loss": 0.1343,
+      "step": 4788
+    },
+    {
+      "epoch": 0.7759235255994815,
+      "grad_norm": 0.8470262289047241,
+      "learning_rate": 4.345423932616899e-06,
+      "loss": 0.1319,
+      "step": 4789
+    },
+    {
+      "epoch": 0.7760855476344782,
+      "grad_norm": 0.8204308152198792,
+      "learning_rate": 4.345128892388515e-06,
+      "loss": 0.1331,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7762475696694751,
+      "grad_norm": 0.8398123383522034,
+      "learning_rate": 4.344833795703652e-06,
+      "loss": 0.1382,
+      "step": 4791
+    },
+    {
+      "epoch": 0.7764095917044718,
+      "grad_norm": 0.9840171933174133,
+      "learning_rate": 4.344538642571339e-06,
+      "loss": 0.1512,
+      "step": 4792
+    },
+    {
+      "epoch": 0.7765716137394686,
+      "grad_norm": 0.8746889233589172,
+      "learning_rate": 4.3442434330006075e-06,
+      "loss": 0.1425,
+      "step": 4793
+    },
+    {
+      "epoch": 0.7767336357744653,
+      "grad_norm": 0.8782622218132019,
+      "learning_rate": 4.3439481670004895e-06,
+      "loss": 0.1317,
+      "step": 4794
+    },
+    {
+      "epoch": 0.7768956578094621,
+      "grad_norm": 0.8894055485725403,
+      "learning_rate": 4.34365284458002e-06,
+      "loss": 0.1424,
+      "step": 4795
+    },
+    {
+      "epoch": 0.7770576798444588,
+      "grad_norm": 0.8703826069831848,
+      "learning_rate": 4.343357465748235e-06,
+      "loss": 0.1309,
+      "step": 4796
+    },
+    {
+      "epoch": 0.7772197018794557,
+      "grad_norm": 0.8910863399505615,
+      "learning_rate": 4.343062030514172e-06,
+      "loss": 0.1431,
+      "step": 4797
+    },
+    {
+      "epoch": 0.7773817239144524,
+      "grad_norm": 0.9824634194374084,
+      "learning_rate": 4.342766538886872e-06,
+      "loss": 0.1456,
+      "step": 4798
+    },
+    {
+      "epoch": 0.7775437459494491,
+      "grad_norm": 0.862075686454773,
+      "learning_rate": 4.342470990875375e-06,
+      "loss": 0.1338,
+      "step": 4799
+    },
+    {
+      "epoch": 0.7777057679844459,
+      "grad_norm": 0.7366591095924377,
+      "learning_rate": 4.342175386488724e-06,
+      "loss": 0.1238,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7778677900194426,
+      "grad_norm": 0.8682642579078674,
+      "learning_rate": 4.341879725735965e-06,
+      "loss": 0.1458,
+      "step": 4801
+    },
+    {
+      "epoch": 0.7780298120544394,
+      "grad_norm": 1.0655760765075684,
+      "learning_rate": 4.341584008626143e-06,
+      "loss": 0.1728,
+      "step": 4802
+    },
+    {
+      "epoch": 0.7781918340894362,
+      "grad_norm": 0.9638577103614807,
+      "learning_rate": 4.341288235168306e-06,
+      "loss": 0.1636,
+      "step": 4803
+    },
+    {
+      "epoch": 0.778353856124433,
+      "grad_norm": 0.9434710741043091,
+      "learning_rate": 4.340992405371506e-06,
+      "loss": 0.1469,
+      "step": 4804
+    },
+    {
+      "epoch": 0.7785158781594297,
+      "grad_norm": 0.9497054219245911,
+      "learning_rate": 4.340696519244794e-06,
+      "loss": 0.1343,
+      "step": 4805
+    },
+    {
+      "epoch": 0.7786779001944264,
+      "grad_norm": 0.8136816620826721,
+      "learning_rate": 4.340400576797221e-06,
+      "loss": 0.1239,
+      "step": 4806
+    },
+    {
+      "epoch": 0.7788399222294232,
+      "grad_norm": 0.81424480676651,
+      "learning_rate": 4.340104578037846e-06,
+      "loss": 0.1231,
+      "step": 4807
+    },
+    {
+      "epoch": 0.7790019442644199,
+      "grad_norm": 0.8290922045707703,
+      "learning_rate": 4.339808522975722e-06,
+      "loss": 0.1283,
+      "step": 4808
+    },
+    {
+      "epoch": 0.7791639662994168,
+      "grad_norm": 0.9101153016090393,
+      "learning_rate": 4.339512411619912e-06,
+      "loss": 0.1419,
+      "step": 4809
+    },
+    {
+      "epoch": 0.7793259883344135,
+      "grad_norm": 0.8872000575065613,
+      "learning_rate": 4.339216243979471e-06,
+      "loss": 0.1378,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7794880103694103,
+      "grad_norm": 0.9886912703514099,
+      "learning_rate": 4.338920020063465e-06,
+      "loss": 0.1599,
+      "step": 4811
+    },
+    {
+      "epoch": 0.779650032404407,
+      "grad_norm": 0.7226459383964539,
+      "learning_rate": 4.3386237398809576e-06,
+      "loss": 0.1087,
+      "step": 4812
+    },
+    {
+      "epoch": 0.7798120544394037,
+      "grad_norm": 0.9056746959686279,
+      "learning_rate": 4.338327403441012e-06,
+      "loss": 0.1485,
+      "step": 4813
+    },
+    {
+      "epoch": 0.7799740764744005,
+      "grad_norm": 0.831541121006012,
+      "learning_rate": 4.338031010752696e-06,
+      "loss": 0.1161,
+      "step": 4814
+    },
+    {
+      "epoch": 0.7801360985093972,
+      "grad_norm": 0.9287052750587463,
+      "learning_rate": 4.337734561825079e-06,
+      "loss": 0.1564,
+      "step": 4815
+    },
+    {
+      "epoch": 0.7802981205443941,
+      "grad_norm": 0.9070348739624023,
+      "learning_rate": 4.337438056667233e-06,
+      "loss": 0.1533,
+      "step": 4816
+    },
+    {
+      "epoch": 0.7804601425793908,
+      "grad_norm": 0.8945810794830322,
+      "learning_rate": 4.337141495288228e-06,
+      "loss": 0.141,
+      "step": 4817
+    },
+    {
+      "epoch": 0.7806221646143876,
+      "grad_norm": 0.9323962926864624,
+      "learning_rate": 4.336844877697139e-06,
+      "loss": 0.1525,
+      "step": 4818
+    },
+    {
+      "epoch": 0.7807841866493843,
+      "grad_norm": 0.8296406269073486,
+      "learning_rate": 4.336548203903042e-06,
+      "loss": 0.1358,
+      "step": 4819
+    },
+    {
+      "epoch": 0.780946208684381,
+      "grad_norm": 0.8179200887680054,
+      "learning_rate": 4.336251473915015e-06,
+      "loss": 0.138,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7811082307193778,
+      "grad_norm": 0.8966086506843567,
+      "learning_rate": 4.335954687742136e-06,
+      "loss": 0.1401,
+      "step": 4821
+    },
+    {
+      "epoch": 0.7812702527543746,
+      "grad_norm": 1.0740383863449097,
+      "learning_rate": 4.335657845393486e-06,
+      "loss": 0.1471,
+      "step": 4822
+    },
+    {
+      "epoch": 0.7814322747893714,
+      "grad_norm": 0.9319535493850708,
+      "learning_rate": 4.335360946878148e-06,
+      "loss": 0.1565,
+      "step": 4823
+    },
+    {
+      "epoch": 0.7815942968243681,
+      "grad_norm": 0.9597566723823547,
+      "learning_rate": 4.335063992205207e-06,
+      "loss": 0.1614,
+      "step": 4824
+    },
+    {
+      "epoch": 0.7817563188593649,
+      "grad_norm": 0.8953394293785095,
+      "learning_rate": 4.334766981383749e-06,
+      "loss": 0.1495,
+      "step": 4825
+    },
+    {
+      "epoch": 0.7819183408943616,
+      "grad_norm": 0.8695473074913025,
+      "learning_rate": 4.3344699144228605e-06,
+      "loss": 0.1353,
+      "step": 4826
+    },
+    {
+      "epoch": 0.7820803629293583,
+      "grad_norm": 0.8189257383346558,
+      "learning_rate": 4.334172791331633e-06,
+      "loss": 0.1269,
+      "step": 4827
+    },
+    {
+      "epoch": 0.7822423849643552,
+      "grad_norm": 0.8564877510070801,
+      "learning_rate": 4.333875612119156e-06,
+      "loss": 0.1302,
+      "step": 4828
+    },
+    {
+      "epoch": 0.7824044069993519,
+      "grad_norm": 0.8532876968383789,
+      "learning_rate": 4.3335783767945235e-06,
+      "loss": 0.1405,
+      "step": 4829
+    },
+    {
+      "epoch": 0.7825664290343487,
+      "grad_norm": 0.9211379289627075,
+      "learning_rate": 4.333281085366829e-06,
+      "loss": 0.1515,
+      "step": 4830
+    },
+    {
+      "epoch": 0.7827284510693454,
+      "grad_norm": 0.8540376424789429,
+      "learning_rate": 4.332983737845171e-06,
+      "loss": 0.1295,
+      "step": 4831
+    },
+    {
+      "epoch": 0.7828904731043422,
+      "grad_norm": 0.886938750743866,
+      "learning_rate": 4.332686334238646e-06,
+      "loss": 0.1318,
+      "step": 4832
+    },
+    {
+      "epoch": 0.7830524951393389,
+      "grad_norm": 0.8990640640258789,
+      "learning_rate": 4.3323888745563544e-06,
+      "loss": 0.1463,
+      "step": 4833
+    },
+    {
+      "epoch": 0.7832145171743357,
+      "grad_norm": 0.8602086305618286,
+      "learning_rate": 4.332091358807397e-06,
+      "loss": 0.1309,
+      "step": 4834
+    },
+    {
+      "epoch": 0.7833765392093325,
+      "grad_norm": 1.16105318069458,
+      "learning_rate": 4.331793787000878e-06,
+      "loss": 0.1761,
+      "step": 4835
+    },
+    {
+      "epoch": 0.7835385612443292,
+      "grad_norm": 0.9003329277038574,
+      "learning_rate": 4.3314961591459015e-06,
+      "loss": 0.1451,
+      "step": 4836
+    },
+    {
+      "epoch": 0.783700583279326,
+      "grad_norm": 0.7860771417617798,
+      "learning_rate": 4.3311984752515745e-06,
+      "loss": 0.1227,
+      "step": 4837
+    },
+    {
+      "epoch": 0.7838626053143227,
+      "grad_norm": 0.7542071342468262,
+      "learning_rate": 4.330900735327006e-06,
+      "loss": 0.1211,
+      "step": 4838
+    },
+    {
+      "epoch": 0.7840246273493195,
+      "grad_norm": 0.9031782150268555,
+      "learning_rate": 4.330602939381306e-06,
+      "loss": 0.1391,
+      "step": 4839
+    },
+    {
+      "epoch": 0.7841866493843163,
+      "grad_norm": 0.8484506607055664,
+      "learning_rate": 4.330305087423585e-06,
+      "loss": 0.1333,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7843486714193131,
+      "grad_norm": 1.0116180181503296,
+      "learning_rate": 4.3300071794629585e-06,
+      "loss": 0.1477,
+      "step": 4841
+    },
+    {
+      "epoch": 0.7845106934543098,
+      "grad_norm": 0.8738240003585815,
+      "learning_rate": 4.329709215508541e-06,
+      "loss": 0.1412,
+      "step": 4842
+    },
+    {
+      "epoch": 0.7846727154893065,
+      "grad_norm": 0.8802550435066223,
+      "learning_rate": 4.329411195569448e-06,
+      "loss": 0.1366,
+      "step": 4843
+    },
+    {
+      "epoch": 0.7848347375243033,
+      "grad_norm": 1.0167311429977417,
+      "learning_rate": 4.329113119654801e-06,
+      "loss": 0.1548,
+      "step": 4844
+    },
+    {
+      "epoch": 0.7849967595593,
+      "grad_norm": 0.9607383608818054,
+      "learning_rate": 4.328814987773718e-06,
+      "loss": 0.1532,
+      "step": 4845
+    },
+    {
+      "epoch": 0.7851587815942969,
+      "grad_norm": 0.8469568490982056,
+      "learning_rate": 4.328516799935323e-06,
+      "loss": 0.1294,
+      "step": 4846
+    },
+    {
+      "epoch": 0.7853208036292936,
+      "grad_norm": 0.8837965130805969,
+      "learning_rate": 4.328218556148738e-06,
+      "loss": 0.1422,
+      "step": 4847
+    },
+    {
+      "epoch": 0.7854828256642904,
+      "grad_norm": 1.0237815380096436,
+      "learning_rate": 4.327920256423089e-06,
+      "loss": 0.165,
+      "step": 4848
+    },
+    {
+      "epoch": 0.7856448476992871,
+      "grad_norm": 0.8523423671722412,
+      "learning_rate": 4.327621900767504e-06,
+      "loss": 0.1523,
+      "step": 4849
+    },
+    {
+      "epoch": 0.7858068697342838,
+      "grad_norm": 0.831685483455658,
+      "learning_rate": 4.3273234891911135e-06,
+      "loss": 0.137,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7859688917692806,
+      "grad_norm": 1.195661187171936,
+      "learning_rate": 4.327025021703044e-06,
+      "loss": 0.146,
+      "step": 4851
+    },
+    {
+      "epoch": 0.7861309138042774,
+      "grad_norm": 0.8539892435073853,
+      "learning_rate": 4.3267264983124304e-06,
+      "loss": 0.1462,
+      "step": 4852
+    },
+    {
+      "epoch": 0.7862929358392742,
+      "grad_norm": 0.8452509045600891,
+      "learning_rate": 4.326427919028407e-06,
+      "loss": 0.1356,
+      "step": 4853
+    },
+    {
+      "epoch": 0.7864549578742709,
+      "grad_norm": 0.9559705853462219,
+      "learning_rate": 4.326129283860109e-06,
+      "loss": 0.1544,
+      "step": 4854
+    },
+    {
+      "epoch": 0.7866169799092677,
+      "grad_norm": 0.9361833333969116,
+      "learning_rate": 4.325830592816675e-06,
+      "loss": 0.1396,
+      "step": 4855
+    },
+    {
+      "epoch": 0.7867790019442644,
+      "grad_norm": 0.9473642110824585,
+      "learning_rate": 4.3255318459072415e-06,
+      "loss": 0.1549,
+      "step": 4856
+    },
+    {
+      "epoch": 0.7869410239792611,
+      "grad_norm": 0.8166648149490356,
+      "learning_rate": 4.325233043140952e-06,
+      "loss": 0.1309,
+      "step": 4857
+    },
+    {
+      "epoch": 0.787103046014258,
+      "grad_norm": 0.8858399987220764,
+      "learning_rate": 4.324934184526949e-06,
+      "loss": 0.1317,
+      "step": 4858
+    },
+    {
+      "epoch": 0.7872650680492547,
+      "grad_norm": 0.9922860264778137,
+      "learning_rate": 4.324635270074375e-06,
+      "loss": 0.1545,
+      "step": 4859
+    },
+    {
+      "epoch": 0.7874270900842515,
+      "grad_norm": 0.9596473574638367,
+      "learning_rate": 4.324336299792378e-06,
+      "loss": 0.163,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7875891121192482,
+      "grad_norm": 0.8286867737770081,
+      "learning_rate": 4.3240372736901044e-06,
+      "loss": 0.1262,
+      "step": 4861
+    },
+    {
+      "epoch": 0.787751134154245,
+      "grad_norm": 0.8577548861503601,
+      "learning_rate": 4.3237381917767054e-06,
+      "loss": 0.1487,
+      "step": 4862
+    },
+    {
+      "epoch": 0.7879131561892417,
+      "grad_norm": 0.9122594594955444,
+      "learning_rate": 4.323439054061331e-06,
+      "loss": 0.1592,
+      "step": 4863
+    },
+    {
+      "epoch": 0.7880751782242384,
+      "grad_norm": 0.7763839364051819,
+      "learning_rate": 4.323139860553133e-06,
+      "loss": 0.1135,
+      "step": 4864
+    },
+    {
+      "epoch": 0.7882372002592353,
+      "grad_norm": 0.8384743332862854,
+      "learning_rate": 4.3228406112612686e-06,
+      "loss": 0.1249,
+      "step": 4865
+    },
+    {
+      "epoch": 0.788399222294232,
+      "grad_norm": 0.9874293804168701,
+      "learning_rate": 4.3225413061948915e-06,
+      "loss": 0.1608,
+      "step": 4866
+    },
+    {
+      "epoch": 0.7885612443292288,
+      "grad_norm": 0.8717676997184753,
+      "learning_rate": 4.322241945363161e-06,
+      "loss": 0.1304,
+      "step": 4867
+    },
+    {
+      "epoch": 0.7887232663642255,
+      "grad_norm": 0.8286629319190979,
+      "learning_rate": 4.321942528775238e-06,
+      "loss": 0.1303,
+      "step": 4868
+    },
+    {
+      "epoch": 0.7888852883992223,
+      "grad_norm": 0.7938746809959412,
+      "learning_rate": 4.3216430564402815e-06,
+      "loss": 0.1231,
+      "step": 4869
+    },
+    {
+      "epoch": 0.789047310434219,
+      "grad_norm": 1.0110461711883545,
+      "learning_rate": 4.3213435283674556e-06,
+      "loss": 0.149,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7892093324692158,
+      "grad_norm": 0.8735683560371399,
+      "learning_rate": 4.3210439445659255e-06,
+      "loss": 0.1276,
+      "step": 4871
+    },
+    {
+      "epoch": 0.7893713545042126,
+      "grad_norm": 1.0407592058181763,
+      "learning_rate": 4.320744305044858e-06,
+      "loss": 0.1411,
+      "step": 4872
+    },
+    {
+      "epoch": 0.7895333765392093,
+      "grad_norm": 0.9165899753570557,
+      "learning_rate": 4.3204446098134215e-06,
+      "loss": 0.1503,
+      "step": 4873
+    },
+    {
+      "epoch": 0.7896953985742061,
+      "grad_norm": 0.9459313154220581,
+      "learning_rate": 4.320144858880784e-06,
+      "loss": 0.1372,
+      "step": 4874
+    },
+    {
+      "epoch": 0.7898574206092028,
+      "grad_norm": 0.8336299061775208,
+      "learning_rate": 4.319845052256119e-06,
+      "loss": 0.1305,
+      "step": 4875
+    },
+    {
+      "epoch": 0.7900194426441997,
+      "grad_norm": 0.928596019744873,
+      "learning_rate": 4.319545189948599e-06,
+      "loss": 0.1473,
+      "step": 4876
+    },
+    {
+      "epoch": 0.7901814646791964,
+      "grad_norm": 0.8122962713241577,
+      "learning_rate": 4.3192452719674e-06,
+      "loss": 0.1276,
+      "step": 4877
+    },
+    {
+      "epoch": 0.7903434867141931,
+      "grad_norm": 0.7669326066970825,
+      "learning_rate": 4.318945298321698e-06,
+      "loss": 0.1314,
+      "step": 4878
+    },
+    {
+      "epoch": 0.7905055087491899,
+      "grad_norm": 0.8381757140159607,
+      "learning_rate": 4.318645269020671e-06,
+      "loss": 0.1312,
+      "step": 4879
+    },
+    {
+      "epoch": 0.7906675307841866,
+      "grad_norm": 0.8036909103393555,
+      "learning_rate": 4.3183451840735e-06,
+      "loss": 0.1092,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7908295528191834,
+      "grad_norm": 0.8607558608055115,
+      "learning_rate": 4.318045043489367e-06,
+      "loss": 0.137,
+      "step": 4881
+    },
+    {
+      "epoch": 0.7909915748541801,
+      "grad_norm": 0.8910214900970459,
+      "learning_rate": 4.3177448472774566e-06,
+      "loss": 0.153,
+      "step": 4882
+    },
+    {
+      "epoch": 0.791153596889177,
+      "grad_norm": 0.8832724094390869,
+      "learning_rate": 4.317444595446951e-06,
+      "loss": 0.148,
+      "step": 4883
+    },
+    {
+      "epoch": 0.7913156189241737,
+      "grad_norm": 0.9177297353744507,
+      "learning_rate": 4.317144288007039e-06,
+      "loss": 0.1537,
+      "step": 4884
+    },
+    {
+      "epoch": 0.7914776409591704,
+      "grad_norm": 0.9235965609550476,
+      "learning_rate": 4.316843924966909e-06,
+      "loss": 0.1404,
+      "step": 4885
+    },
+    {
+      "epoch": 0.7916396629941672,
+      "grad_norm": 0.8111236691474915,
+      "learning_rate": 4.316543506335752e-06,
+      "loss": 0.1276,
+      "step": 4886
+    },
+    {
+      "epoch": 0.7918016850291639,
+      "grad_norm": 0.8697571158409119,
+      "learning_rate": 4.31624303212276e-06,
+      "loss": 0.1325,
+      "step": 4887
+    },
+    {
+      "epoch": 0.7919637070641607,
+      "grad_norm": 0.8985753059387207,
+      "learning_rate": 4.315942502337126e-06,
+      "loss": 0.1418,
+      "step": 4888
+    },
+    {
+      "epoch": 0.7921257290991575,
+      "grad_norm": 0.8289467096328735,
+      "learning_rate": 4.315641916988046e-06,
+      "loss": 0.1169,
+      "step": 4889
+    },
+    {
+      "epoch": 0.7922877511341543,
+      "grad_norm": 1.004921555519104,
+      "learning_rate": 4.315341276084717e-06,
+      "loss": 0.1646,
+      "step": 4890
+    },
+    {
+      "epoch": 0.792449773169151,
+      "grad_norm": 0.9182339906692505,
+      "learning_rate": 4.315040579636339e-06,
+      "loss": 0.1565,
+      "step": 4891
+    },
+    {
+      "epoch": 0.7926117952041478,
+      "grad_norm": 0.8737165927886963,
+      "learning_rate": 4.3147398276521105e-06,
+      "loss": 0.1278,
+      "step": 4892
+    },
+    {
+      "epoch": 0.7927738172391445,
+      "grad_norm": 0.8931154608726501,
+      "learning_rate": 4.314439020141235e-06,
+      "loss": 0.145,
+      "step": 4893
+    },
+    {
+      "epoch": 0.7929358392741412,
+      "grad_norm": 0.9454561471939087,
+      "learning_rate": 4.314138157112916e-06,
+      "loss": 0.1496,
+      "step": 4894
+    },
+    {
+      "epoch": 0.7930978613091381,
+      "grad_norm": 0.8674540519714355,
+      "learning_rate": 4.313837238576361e-06,
+      "loss": 0.1277,
+      "step": 4895
+    },
+    {
+      "epoch": 0.7932598833441348,
+      "grad_norm": 0.8071511387825012,
+      "learning_rate": 4.313536264540774e-06,
+      "loss": 0.1258,
+      "step": 4896
+    },
+    {
+      "epoch": 0.7934219053791316,
+      "grad_norm": 0.9354859590530396,
+      "learning_rate": 4.313235235015367e-06,
+      "loss": 0.1537,
+      "step": 4897
+    },
+    {
+      "epoch": 0.7935839274141283,
+      "grad_norm": 0.9247655272483826,
+      "learning_rate": 4.312934150009351e-06,
+      "loss": 0.1535,
+      "step": 4898
+    },
+    {
+      "epoch": 0.7937459494491251,
+      "grad_norm": 0.8361133933067322,
+      "learning_rate": 4.312633009531935e-06,
+      "loss": 0.1397,
+      "step": 4899
+    },
+    {
+      "epoch": 0.7939079714841218,
+      "grad_norm": 1.0772885084152222,
+      "learning_rate": 4.3123318135923355e-06,
+      "loss": 0.1683,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7940699935191186,
+      "grad_norm": 0.7447131872177124,
+      "learning_rate": 4.312030562199769e-06,
+      "loss": 0.1155,
+      "step": 4901
+    },
+    {
+      "epoch": 0.7942320155541154,
+      "grad_norm": 0.7897053360939026,
+      "learning_rate": 4.311729255363453e-06,
+      "loss": 0.1229,
+      "step": 4902
+    },
+    {
+      "epoch": 0.7943940375891121,
+      "grad_norm": 0.8198691606521606,
+      "learning_rate": 4.3114278930926055e-06,
+      "loss": 0.1323,
+      "step": 4903
+    },
+    {
+      "epoch": 0.7945560596241089,
+      "grad_norm": 0.8728782534599304,
+      "learning_rate": 4.3111264753964475e-06,
+      "loss": 0.1368,
+      "step": 4904
+    },
+    {
+      "epoch": 0.7947180816591056,
+      "grad_norm": 0.8745116591453552,
+      "learning_rate": 4.3108250022842026e-06,
+      "loss": 0.1475,
+      "step": 4905
+    },
+    {
+      "epoch": 0.7948801036941024,
+      "grad_norm": 0.8484328389167786,
+      "learning_rate": 4.310523473765095e-06,
+      "loss": 0.1349,
+      "step": 4906
+    },
+    {
+      "epoch": 0.7950421257290992,
+      "grad_norm": 0.8352507948875427,
+      "learning_rate": 4.31022188984835e-06,
+      "loss": 0.1349,
+      "step": 4907
+    },
+    {
+      "epoch": 0.7952041477640959,
+      "grad_norm": 0.9030001759529114,
+      "learning_rate": 4.309920250543196e-06,
+      "loss": 0.1343,
+      "step": 4908
+    },
+    {
+      "epoch": 0.7953661697990927,
+      "grad_norm": 0.9628517627716064,
+      "learning_rate": 4.3096185558588625e-06,
+      "loss": 0.158,
+      "step": 4909
+    },
+    {
+      "epoch": 0.7955281918340894,
+      "grad_norm": 0.84720778465271,
+      "learning_rate": 4.30931680580458e-06,
+      "loss": 0.1336,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7956902138690862,
+      "grad_norm": 0.9344910383224487,
+      "learning_rate": 4.309015000389583e-06,
+      "loss": 0.1469,
+      "step": 4911
+    },
+    {
+      "epoch": 0.7958522359040829,
+      "grad_norm": 0.8999947309494019,
+      "learning_rate": 4.308713139623103e-06,
+      "loss": 0.1324,
+      "step": 4912
+    },
+    {
+      "epoch": 0.7960142579390798,
+      "grad_norm": 0.9617306590080261,
+      "learning_rate": 4.308411223514378e-06,
+      "loss": 0.1633,
+      "step": 4913
+    },
+    {
+      "epoch": 0.7961762799740765,
+      "grad_norm": 0.8142529726028442,
+      "learning_rate": 4.308109252072647e-06,
+      "loss": 0.1394,
+      "step": 4914
+    },
+    {
+      "epoch": 0.7963383020090732,
+      "grad_norm": 0.837073028087616,
+      "learning_rate": 4.307807225307148e-06,
+      "loss": 0.1277,
+      "step": 4915
+    },
+    {
+      "epoch": 0.79650032404407,
+      "grad_norm": 0.9646647572517395,
+      "learning_rate": 4.307505143227122e-06,
+      "loss": 0.1675,
+      "step": 4916
+    },
+    {
+      "epoch": 0.7966623460790667,
+      "grad_norm": 0.893146812915802,
+      "learning_rate": 4.307203005841813e-06,
+      "loss": 0.146,
+      "step": 4917
+    },
+    {
+      "epoch": 0.7968243681140635,
+      "grad_norm": 0.9233022928237915,
+      "learning_rate": 4.306900813160466e-06,
+      "loss": 0.1535,
+      "step": 4918
+    },
+    {
+      "epoch": 0.7969863901490603,
+      "grad_norm": 0.8315757513046265,
+      "learning_rate": 4.306598565192327e-06,
+      "loss": 0.133,
+      "step": 4919
+    },
+    {
+      "epoch": 0.7971484121840571,
+      "grad_norm": 0.8391256332397461,
+      "learning_rate": 4.306296261946643e-06,
+      "loss": 0.1378,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7973104342190538,
+      "grad_norm": 0.9489231109619141,
+      "learning_rate": 4.305993903432664e-06,
+      "loss": 0.1461,
+      "step": 4921
+    },
+    {
+      "epoch": 0.7974724562540505,
+      "grad_norm": 0.8281417489051819,
+      "learning_rate": 4.305691489659643e-06,
+      "loss": 0.1266,
+      "step": 4922
+    },
+    {
+      "epoch": 0.7976344782890473,
+      "grad_norm": 0.8485118746757507,
+      "learning_rate": 4.305389020636832e-06,
+      "loss": 0.1373,
+      "step": 4923
+    },
+    {
+      "epoch": 0.797796500324044,
+      "grad_norm": 0.8453590869903564,
+      "learning_rate": 4.3050864963734854e-06,
+      "loss": 0.1371,
+      "step": 4924
+    },
+    {
+      "epoch": 0.7979585223590409,
+      "grad_norm": 0.8031056523323059,
+      "learning_rate": 4.304783916878861e-06,
+      "loss": 0.1262,
+      "step": 4925
+    },
+    {
+      "epoch": 0.7981205443940376,
+      "grad_norm": 0.9986342191696167,
+      "learning_rate": 4.304481282162215e-06,
+      "loss": 0.1689,
+      "step": 4926
+    },
+    {
+      "epoch": 0.7982825664290344,
+      "grad_norm": 0.8515965938568115,
+      "learning_rate": 4.304178592232809e-06,
+      "loss": 0.1373,
+      "step": 4927
+    },
+    {
+      "epoch": 0.7984445884640311,
+      "grad_norm": 0.9090455174446106,
+      "learning_rate": 4.3038758470999056e-06,
+      "loss": 0.1508,
+      "step": 4928
+    },
+    {
+      "epoch": 0.7986066104990278,
+      "grad_norm": 0.8485004305839539,
+      "learning_rate": 4.303573046772765e-06,
+      "loss": 0.1513,
+      "step": 4929
+    },
+    {
+      "epoch": 0.7987686325340246,
+      "grad_norm": 1.051450490951538,
+      "learning_rate": 4.303270191260654e-06,
+      "loss": 0.1781,
+      "step": 4930
+    },
+    {
+      "epoch": 0.7989306545690213,
+      "grad_norm": 0.8531157374382019,
+      "learning_rate": 4.302967280572839e-06,
+      "loss": 0.1348,
+      "step": 4931
+    },
+    {
+      "epoch": 0.7990926766040182,
+      "grad_norm": 0.7763565182685852,
+      "learning_rate": 4.302664314718588e-06,
+      "loss": 0.1182,
+      "step": 4932
+    },
+    {
+      "epoch": 0.7992546986390149,
+      "grad_norm": 0.9212713837623596,
+      "learning_rate": 4.302361293707172e-06,
+      "loss": 0.1209,
+      "step": 4933
+    },
+    {
+      "epoch": 0.7994167206740117,
+      "grad_norm": 0.8668341040611267,
+      "learning_rate": 4.302058217547862e-06,
+      "loss": 0.1316,
+      "step": 4934
+    },
+    {
+      "epoch": 0.7995787427090084,
+      "grad_norm": 1.0723251104354858,
+      "learning_rate": 4.3017550862499314e-06,
+      "loss": 0.1604,
+      "step": 4935
+    },
+    {
+      "epoch": 0.7997407647440052,
+      "grad_norm": 0.9705539345741272,
+      "learning_rate": 4.301451899822655e-06,
+      "loss": 0.1541,
+      "step": 4936
+    },
+    {
+      "epoch": 0.799902786779002,
+      "grad_norm": 0.8101187348365784,
+      "learning_rate": 4.30114865827531e-06,
+      "loss": 0.1226,
+      "step": 4937
+    },
+    {
+      "epoch": 0.8000648088139987,
+      "grad_norm": 0.9303157925605774,
+      "learning_rate": 4.3008453616171746e-06,
+      "loss": 0.1661,
+      "step": 4938
+    },
+    {
+      "epoch": 0.8002268308489955,
+      "grad_norm": 0.8247880339622498,
+      "learning_rate": 4.300542009857529e-06,
+      "loss": 0.1287,
+      "step": 4939
+    },
+    {
+      "epoch": 0.8003888528839922,
+      "grad_norm": 0.8863083124160767,
+      "learning_rate": 4.300238603005656e-06,
+      "loss": 0.1557,
+      "step": 4940
+    },
+    {
+      "epoch": 0.800550874918989,
+      "grad_norm": 0.9141151905059814,
+      "learning_rate": 4.299935141070837e-06,
+      "loss": 0.1515,
+      "step": 4941
+    },
+    {
+      "epoch": 0.8007128969539857,
+      "grad_norm": 0.925873875617981,
+      "learning_rate": 4.299631624062359e-06,
+      "loss": 0.1467,
+      "step": 4942
+    },
+    {
+      "epoch": 0.8008749189889826,
+      "grad_norm": 0.9137710332870483,
+      "learning_rate": 4.299328051989509e-06,
+      "loss": 0.1325,
+      "step": 4943
+    },
+    {
+      "epoch": 0.8010369410239793,
+      "grad_norm": 1.1716556549072266,
+      "learning_rate": 4.299024424861574e-06,
+      "loss": 0.1653,
+      "step": 4944
+    },
+    {
+      "epoch": 0.801198963058976,
+      "grad_norm": 1.0240533351898193,
+      "learning_rate": 4.298720742687846e-06,
+      "loss": 0.1516,
+      "step": 4945
+    },
+    {
+      "epoch": 0.8013609850939728,
+      "grad_norm": 0.8595703840255737,
+      "learning_rate": 4.298417005477616e-06,
+      "loss": 0.1388,
+      "step": 4946
+    },
+    {
+      "epoch": 0.8015230071289695,
+      "grad_norm": 0.817602813243866,
+      "learning_rate": 4.298113213240176e-06,
+      "loss": 0.1203,
+      "step": 4947
+    },
+    {
+      "epoch": 0.8016850291639663,
+      "grad_norm": 0.9021492004394531,
+      "learning_rate": 4.2978093659848255e-06,
+      "loss": 0.1397,
+      "step": 4948
+    },
+    {
+      "epoch": 0.801847051198963,
+      "grad_norm": 0.756421685218811,
+      "learning_rate": 4.297505463720857e-06,
+      "loss": 0.1294,
+      "step": 4949
+    },
+    {
+      "epoch": 0.8020090732339599,
+      "grad_norm": 0.8492708206176758,
+      "learning_rate": 4.2972015064575726e-06,
+      "loss": 0.146,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8021710952689566,
+      "grad_norm": 0.8680381774902344,
+      "learning_rate": 4.29689749420427e-06,
+      "loss": 0.1556,
+      "step": 4951
+    },
+    {
+      "epoch": 0.8023331173039533,
+      "grad_norm": 0.7963783144950867,
+      "learning_rate": 4.2965934269702535e-06,
+      "loss": 0.1291,
+      "step": 4952
+    },
+    {
+      "epoch": 0.8024951393389501,
+      "grad_norm": 0.8368898630142212,
+      "learning_rate": 4.296289304764825e-06,
+      "loss": 0.1396,
+      "step": 4953
+    },
+    {
+      "epoch": 0.8026571613739468,
+      "grad_norm": 0.951833188533783,
+      "learning_rate": 4.295985127597291e-06,
+      "loss": 0.1592,
+      "step": 4954
+    },
+    {
+      "epoch": 0.8028191834089436,
+      "grad_norm": 0.8976386785507202,
+      "learning_rate": 4.295680895476959e-06,
+      "loss": 0.1454,
+      "step": 4955
+    },
+    {
+      "epoch": 0.8029812054439404,
+      "grad_norm": 0.8689819574356079,
+      "learning_rate": 4.295376608413136e-06,
+      "loss": 0.1393,
+      "step": 4956
+    },
+    {
+      "epoch": 0.8031432274789372,
+      "grad_norm": 0.9349297285079956,
+      "learning_rate": 4.295072266415135e-06,
+      "loss": 0.1617,
+      "step": 4957
+    },
+    {
+      "epoch": 0.8033052495139339,
+      "grad_norm": 0.9541299939155579,
+      "learning_rate": 4.294767869492265e-06,
+      "loss": 0.1545,
+      "step": 4958
+    },
+    {
+      "epoch": 0.8034672715489306,
+      "grad_norm": 0.9361594915390015,
+      "learning_rate": 4.294463417653842e-06,
+      "loss": 0.156,
+      "step": 4959
+    },
+    {
+      "epoch": 0.8036292935839274,
+      "grad_norm": 0.8772572875022888,
+      "learning_rate": 4.294158910909181e-06,
+      "loss": 0.14,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8037913156189241,
+      "grad_norm": 0.84719318151474,
+      "learning_rate": 4.2938543492676e-06,
+      "loss": 0.1324,
+      "step": 4961
+    },
+    {
+      "epoch": 0.803953337653921,
+      "grad_norm": 0.7897992730140686,
+      "learning_rate": 4.293549732738415e-06,
+      "loss": 0.1266,
+      "step": 4962
+    },
+    {
+      "epoch": 0.8041153596889177,
+      "grad_norm": 0.9460259079933167,
+      "learning_rate": 4.29324506133095e-06,
+      "loss": 0.1521,
+      "step": 4963
+    },
+    {
+      "epoch": 0.8042773817239145,
+      "grad_norm": 0.8699532747268677,
+      "learning_rate": 4.2929403350545255e-06,
+      "loss": 0.1435,
+      "step": 4964
+    },
+    {
+      "epoch": 0.8044394037589112,
+      "grad_norm": 0.9408538341522217,
+      "learning_rate": 4.292635553918466e-06,
+      "loss": 0.1431,
+      "step": 4965
+    },
+    {
+      "epoch": 0.8046014257939079,
+      "grad_norm": 0.885382354259491,
+      "learning_rate": 4.292330717932095e-06,
+      "loss": 0.1411,
+      "step": 4966
+    },
+    {
+      "epoch": 0.8047634478289047,
+      "grad_norm": 0.9126858115196228,
+      "learning_rate": 4.292025827104744e-06,
+      "loss": 0.1511,
+      "step": 4967
+    },
+    {
+      "epoch": 0.8049254698639015,
+      "grad_norm": 0.8485498428344727,
+      "learning_rate": 4.2917208814457364e-06,
+      "loss": 0.1325,
+      "step": 4968
+    },
+    {
+      "epoch": 0.8050874918988983,
+      "grad_norm": 0.9165229797363281,
+      "learning_rate": 4.291415880964407e-06,
+      "loss": 0.1521,
+      "step": 4969
+    },
+    {
+      "epoch": 0.805249513933895,
+      "grad_norm": 0.811354398727417,
+      "learning_rate": 4.291110825670087e-06,
+      "loss": 0.1297,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8054115359688918,
+      "grad_norm": 0.8046636581420898,
+      "learning_rate": 4.29080571557211e-06,
+      "loss": 0.1192,
+      "step": 4971
+    },
+    {
+      "epoch": 0.8055735580038885,
+      "grad_norm": 0.8413184881210327,
+      "learning_rate": 4.290500550679811e-06,
+      "loss": 0.1316,
+      "step": 4972
+    },
+    {
+      "epoch": 0.8057355800388852,
+      "grad_norm": 0.9264662861824036,
+      "learning_rate": 4.290195331002529e-06,
+      "loss": 0.15,
+      "step": 4973
+    },
+    {
+      "epoch": 0.8058976020738821,
+      "grad_norm": 0.917666494846344,
+      "learning_rate": 4.289890056549603e-06,
+      "loss": 0.1395,
+      "step": 4974
+    },
+    {
+      "epoch": 0.8060596241088788,
+      "grad_norm": 0.9180327653884888,
+      "learning_rate": 4.2895847273303705e-06,
+      "loss": 0.1427,
+      "step": 4975
+    },
+    {
+      "epoch": 0.8062216461438756,
+      "grad_norm": 0.902094841003418,
+      "learning_rate": 4.289279343354178e-06,
+      "loss": 0.1421,
+      "step": 4976
+    },
+    {
+      "epoch": 0.8063836681788723,
+      "grad_norm": 0.8890548944473267,
+      "learning_rate": 4.288973904630366e-06,
+      "loss": 0.134,
+      "step": 4977
+    },
+    {
+      "epoch": 0.8065456902138691,
+      "grad_norm": 0.8452367186546326,
+      "learning_rate": 4.288668411168283e-06,
+      "loss": 0.1293,
+      "step": 4978
+    },
+    {
+      "epoch": 0.8067077122488658,
+      "grad_norm": 0.9811566472053528,
+      "learning_rate": 4.288362862977274e-06,
+      "loss": 0.1597,
+      "step": 4979
+    },
+    {
+      "epoch": 0.8068697342838627,
+      "grad_norm": 0.9972265958786011,
+      "learning_rate": 4.28805726006669e-06,
+      "loss": 0.1607,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8070317563188594,
+      "grad_norm": 0.9556589722633362,
+      "learning_rate": 4.287751602445881e-06,
+      "loss": 0.1415,
+      "step": 4981
+    },
+    {
+      "epoch": 0.8071937783538561,
+      "grad_norm": 0.9611114859580994,
+      "learning_rate": 4.287445890124198e-06,
+      "loss": 0.1601,
+      "step": 4982
+    },
+    {
+      "epoch": 0.8073558003888529,
+      "grad_norm": 0.8882191181182861,
+      "learning_rate": 4.287140123110998e-06,
+      "loss": 0.1359,
+      "step": 4983
+    },
+    {
+      "epoch": 0.8075178224238496,
+      "grad_norm": 0.9331012964248657,
+      "learning_rate": 4.286834301415634e-06,
+      "loss": 0.1498,
+      "step": 4984
+    },
+    {
+      "epoch": 0.8076798444588464,
+      "grad_norm": 0.8011325001716614,
+      "learning_rate": 4.286528425047464e-06,
+      "loss": 0.1285,
+      "step": 4985
+    },
+    {
+      "epoch": 0.8078418664938432,
+      "grad_norm": 0.9134606719017029,
+      "learning_rate": 4.286222494015848e-06,
+      "loss": 0.155,
+      "step": 4986
+    },
+    {
+      "epoch": 0.80800388852884,
+      "grad_norm": 0.8747284412384033,
+      "learning_rate": 4.285916508330146e-06,
+      "loss": 0.143,
+      "step": 4987
+    },
+    {
+      "epoch": 0.8081659105638367,
+      "grad_norm": 0.8503855466842651,
+      "learning_rate": 4.285610467999722e-06,
+      "loss": 0.1233,
+      "step": 4988
+    },
+    {
+      "epoch": 0.8083279325988334,
+      "grad_norm": 0.8365163207054138,
+      "learning_rate": 4.285304373033938e-06,
+      "loss": 0.1299,
+      "step": 4989
+    },
+    {
+      "epoch": 0.8084899546338302,
+      "grad_norm": 0.7736684083938599,
+      "learning_rate": 4.28499822344216e-06,
+      "loss": 0.1336,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8086519766688269,
+      "grad_norm": 0.7120151519775391,
+      "learning_rate": 4.284692019233756e-06,
+      "loss": 0.1262,
+      "step": 4991
+    },
+    {
+      "epoch": 0.8088139987038238,
+      "grad_norm": 0.987061083316803,
+      "learning_rate": 4.2843857604180955e-06,
+      "loss": 0.1582,
+      "step": 4992
+    },
+    {
+      "epoch": 0.8089760207388205,
+      "grad_norm": 0.804357647895813,
+      "learning_rate": 4.2840794470045484e-06,
+      "loss": 0.1307,
+      "step": 4993
+    },
+    {
+      "epoch": 0.8091380427738173,
+      "grad_norm": 0.9282926917076111,
+      "learning_rate": 4.283773079002488e-06,
+      "loss": 0.1554,
+      "step": 4994
+    },
+    {
+      "epoch": 0.809300064808814,
+      "grad_norm": 0.8444812297821045,
+      "learning_rate": 4.283466656421289e-06,
+      "loss": 0.1413,
+      "step": 4995
+    },
+    {
+      "epoch": 0.8094620868438107,
+      "grad_norm": 0.9229618310928345,
+      "learning_rate": 4.283160179270325e-06,
+      "loss": 0.1442,
+      "step": 4996
+    },
+    {
+      "epoch": 0.8096241088788075,
+      "grad_norm": 0.8210486173629761,
+      "learning_rate": 4.282853647558974e-06,
+      "loss": 0.1277,
+      "step": 4997
+    },
+    {
+      "epoch": 0.8097861309138042,
+      "grad_norm": 0.8737907409667969,
+      "learning_rate": 4.282547061296618e-06,
+      "loss": 0.1377,
+      "step": 4998
+    },
+    {
+      "epoch": 0.8099481529488011,
+      "grad_norm": 0.8875065445899963,
+      "learning_rate": 4.2822404204926334e-06,
+      "loss": 0.1343,
+      "step": 4999
+    },
+    {
+      "epoch": 0.8101101749837978,
+      "grad_norm": 0.8774368762969971,
+      "learning_rate": 4.281933725156406e-06,
+      "loss": 0.1423,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8102721970187946,
+      "grad_norm": 0.8778204917907715,
+      "learning_rate": 4.281626975297319e-06,
+      "loss": 0.1279,
+      "step": 5001
+    },
+    {
+      "epoch": 0.8104342190537913,
+      "grad_norm": 0.8505150079727173,
+      "learning_rate": 4.281320170924758e-06,
+      "loss": 0.1398,
+      "step": 5002
+    },
+    {
+      "epoch": 0.810596241088788,
+      "grad_norm": 0.7801287770271301,
+      "learning_rate": 4.281013312048109e-06,
+      "loss": 0.1218,
+      "step": 5003
+    },
+    {
+      "epoch": 0.8107582631237849,
+      "grad_norm": 0.9131519794464111,
+      "learning_rate": 4.280706398676764e-06,
+      "loss": 0.1413,
+      "step": 5004
+    },
+    {
+      "epoch": 0.8109202851587816,
+      "grad_norm": 0.8857017159461975,
+      "learning_rate": 4.280399430820112e-06,
+      "loss": 0.1381,
+      "step": 5005
+    },
+    {
+      "epoch": 0.8110823071937784,
+      "grad_norm": 0.8741483092308044,
+      "learning_rate": 4.2800924084875465e-06,
+      "loss": 0.1353,
+      "step": 5006
+    },
+    {
+      "epoch": 0.8112443292287751,
+      "grad_norm": 0.816307783126831,
+      "learning_rate": 4.27978533168846e-06,
+      "loss": 0.1354,
+      "step": 5007
+    },
+    {
+      "epoch": 0.8114063512637719,
+      "grad_norm": 0.8606013059616089,
+      "learning_rate": 4.27947820043225e-06,
+      "loss": 0.1452,
+      "step": 5008
+    },
+    {
+      "epoch": 0.8115683732987686,
+      "grad_norm": 0.999919593334198,
+      "learning_rate": 4.279171014728314e-06,
+      "loss": 0.1466,
+      "step": 5009
+    },
+    {
+      "epoch": 0.8117303953337653,
+      "grad_norm": 0.8366535305976868,
+      "learning_rate": 4.278863774586049e-06,
+      "loss": 0.1267,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8118924173687622,
+      "grad_norm": 0.9078406691551208,
+      "learning_rate": 4.278556480014858e-06,
+      "loss": 0.1422,
+      "step": 5011
+    },
+    {
+      "epoch": 0.8120544394037589,
+      "grad_norm": 0.9016123414039612,
+      "learning_rate": 4.2782491310241426e-06,
+      "loss": 0.1574,
+      "step": 5012
+    },
+    {
+      "epoch": 0.8122164614387557,
+      "grad_norm": 0.7560967206954956,
+      "learning_rate": 4.277941727623307e-06,
+      "loss": 0.1211,
+      "step": 5013
+    },
+    {
+      "epoch": 0.8123784834737524,
+      "grad_norm": 0.8945764303207397,
+      "learning_rate": 4.2776342698217575e-06,
+      "loss": 0.148,
+      "step": 5014
+    },
+    {
+      "epoch": 0.8125405055087492,
+      "grad_norm": 0.8035917282104492,
+      "learning_rate": 4.277326757628901e-06,
+      "loss": 0.1312,
+      "step": 5015
+    },
+    {
+      "epoch": 0.812702527543746,
+      "grad_norm": 0.925362229347229,
+      "learning_rate": 4.277019191054146e-06,
+      "loss": 0.1421,
+      "step": 5016
+    },
+    {
+      "epoch": 0.8128645495787427,
+      "grad_norm": 0.8655949831008911,
+      "learning_rate": 4.276711570106905e-06,
+      "loss": 0.1454,
+      "step": 5017
+    },
+    {
+      "epoch": 0.8130265716137395,
+      "grad_norm": 0.9567270874977112,
+      "learning_rate": 4.276403894796589e-06,
+      "loss": 0.1528,
+      "step": 5018
+    },
+    {
+      "epoch": 0.8131885936487362,
+      "grad_norm": 0.8218672275543213,
+      "learning_rate": 4.276096165132613e-06,
+      "loss": 0.1435,
+      "step": 5019
+    },
+    {
+      "epoch": 0.813350615683733,
+      "grad_norm": 0.9071885943412781,
+      "learning_rate": 4.275788381124393e-06,
+      "loss": 0.1343,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8135126377187297,
+      "grad_norm": 0.801608681678772,
+      "learning_rate": 4.2754805427813455e-06,
+      "loss": 0.138,
+      "step": 5021
+    },
+    {
+      "epoch": 0.8136746597537265,
+      "grad_norm": 0.960382878780365,
+      "learning_rate": 4.275172650112889e-06,
+      "loss": 0.159,
+      "step": 5022
+    },
+    {
+      "epoch": 0.8138366817887233,
+      "grad_norm": 0.8618503212928772,
+      "learning_rate": 4.274864703128446e-06,
+      "loss": 0.1346,
+      "step": 5023
+    },
+    {
+      "epoch": 0.81399870382372,
+      "grad_norm": 0.8866019248962402,
+      "learning_rate": 4.274556701837438e-06,
+      "loss": 0.1402,
+      "step": 5024
+    },
+    {
+      "epoch": 0.8141607258587168,
+      "grad_norm": 0.8554643988609314,
+      "learning_rate": 4.27424864624929e-06,
+      "loss": 0.1376,
+      "step": 5025
+    },
+    {
+      "epoch": 0.8143227478937135,
+      "grad_norm": 0.8216258883476257,
+      "learning_rate": 4.273940536373426e-06,
+      "loss": 0.1221,
+      "step": 5026
+    },
+    {
+      "epoch": 0.8144847699287103,
+      "grad_norm": 1.0153284072875977,
+      "learning_rate": 4.273632372219274e-06,
+      "loss": 0.1503,
+      "step": 5027
+    },
+    {
+      "epoch": 0.814646791963707,
+      "grad_norm": 0.8892536759376526,
+      "learning_rate": 4.273324153796264e-06,
+      "loss": 0.1476,
+      "step": 5028
+    },
+    {
+      "epoch": 0.8148088139987039,
+      "grad_norm": 0.9032856225967407,
+      "learning_rate": 4.273015881113827e-06,
+      "loss": 0.1476,
+      "step": 5029
+    },
+    {
+      "epoch": 0.8149708360337006,
+      "grad_norm": 0.8178720474243164,
+      "learning_rate": 4.2727075541813945e-06,
+      "loss": 0.136,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8151328580686974,
+      "grad_norm": 0.8843294382095337,
+      "learning_rate": 4.2723991730084e-06,
+      "loss": 0.152,
+      "step": 5031
+    },
+    {
+      "epoch": 0.8152948801036941,
+      "grad_norm": 0.8823347091674805,
+      "learning_rate": 4.27209073760428e-06,
+      "loss": 0.1439,
+      "step": 5032
+    },
+    {
+      "epoch": 0.8154569021386908,
+      "grad_norm": 0.886072039604187,
+      "learning_rate": 4.271782247978473e-06,
+      "loss": 0.1416,
+      "step": 5033
+    },
+    {
+      "epoch": 0.8156189241736876,
+      "grad_norm": 0.8647177815437317,
+      "learning_rate": 4.271473704140415e-06,
+      "loss": 0.1335,
+      "step": 5034
+    },
+    {
+      "epoch": 0.8157809462086844,
+      "grad_norm": 0.8254427313804626,
+      "learning_rate": 4.271165106099549e-06,
+      "loss": 0.1284,
+      "step": 5035
+    },
+    {
+      "epoch": 0.8159429682436812,
+      "grad_norm": 0.8761928081512451,
+      "learning_rate": 4.270856453865318e-06,
+      "loss": 0.1494,
+      "step": 5036
+    },
+    {
+      "epoch": 0.8161049902786779,
+      "grad_norm": 0.8046637773513794,
+      "learning_rate": 4.2705477474471645e-06,
+      "loss": 0.1394,
+      "step": 5037
+    },
+    {
+      "epoch": 0.8162670123136747,
+      "grad_norm": 0.8774124383926392,
+      "learning_rate": 4.270238986854534e-06,
+      "loss": 0.1426,
+      "step": 5038
+    },
+    {
+      "epoch": 0.8164290343486714,
+      "grad_norm": 0.7928304672241211,
+      "learning_rate": 4.269930172096875e-06,
+      "loss": 0.1345,
+      "step": 5039
+    },
+    {
+      "epoch": 0.8165910563836681,
+      "grad_norm": 0.7432117462158203,
+      "learning_rate": 4.2696213031836355e-06,
+      "loss": 0.1129,
+      "step": 5040
+    },
+    {
+      "epoch": 0.816753078418665,
+      "grad_norm": 0.860908031463623,
+      "learning_rate": 4.269312380124268e-06,
+      "loss": 0.1405,
+      "step": 5041
+    },
+    {
+      "epoch": 0.8169151004536617,
+      "grad_norm": 0.9895366430282593,
+      "learning_rate": 4.2690034029282214e-06,
+      "loss": 0.1602,
+      "step": 5042
+    },
+    {
+      "epoch": 0.8170771224886585,
+      "grad_norm": 0.9045162200927734,
+      "learning_rate": 4.268694371604952e-06,
+      "loss": 0.1551,
+      "step": 5043
+    },
+    {
+      "epoch": 0.8172391445236552,
+      "grad_norm": 0.8091841340065002,
+      "learning_rate": 4.268385286163915e-06,
+      "loss": 0.1301,
+      "step": 5044
+    },
+    {
+      "epoch": 0.817401166558652,
+      "grad_norm": 0.8731530904769897,
+      "learning_rate": 4.2680761466145685e-06,
+      "loss": 0.1509,
+      "step": 5045
+    },
+    {
+      "epoch": 0.8175631885936487,
+      "grad_norm": 1.0045936107635498,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.1587,
+      "step": 5046
+    },
+    {
+      "epoch": 0.8177252106286454,
+      "grad_norm": 0.7668003439903259,
+      "learning_rate": 4.267457705228781e-06,
+      "loss": 0.1308,
+      "step": 5047
+    },
+    {
+      "epoch": 0.8178872326636423,
+      "grad_norm": 0.9017691612243652,
+      "learning_rate": 4.267148403411261e-06,
+      "loss": 0.1451,
+      "step": 5048
+    },
+    {
+      "epoch": 0.818049254698639,
+      "grad_norm": 0.8749794363975525,
+      "learning_rate": 4.266839047523279e-06,
+      "loss": 0.1391,
+      "step": 5049
+    },
+    {
+      "epoch": 0.8182112767336358,
+      "grad_norm": 0.7766733765602112,
+      "learning_rate": 4.266529637574297e-06,
+      "loss": 0.1278,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8183732987686325,
+      "grad_norm": 0.8146058320999146,
+      "learning_rate": 4.266220173573783e-06,
+      "loss": 0.1348,
+      "step": 5051
+    },
+    {
+      "epoch": 0.8185353208036293,
+      "grad_norm": 0.8790404796600342,
+      "learning_rate": 4.265910655531206e-06,
+      "loss": 0.1364,
+      "step": 5052
+    },
+    {
+      "epoch": 0.818697342838626,
+      "grad_norm": 0.8823887705802917,
+      "learning_rate": 4.265601083456036e-06,
+      "loss": 0.1339,
+      "step": 5053
+    },
+    {
+      "epoch": 0.8188593648736228,
+      "grad_norm": 0.8041099309921265,
+      "learning_rate": 4.265291457357746e-06,
+      "loss": 0.1253,
+      "step": 5054
+    },
+    {
+      "epoch": 0.8190213869086196,
+      "grad_norm": 0.9392186403274536,
+      "learning_rate": 4.264981777245809e-06,
+      "loss": 0.1374,
+      "step": 5055
+    },
+    {
+      "epoch": 0.8191834089436163,
+      "grad_norm": 0.8520307540893555,
+      "learning_rate": 4.2646720431297006e-06,
+      "loss": 0.1322,
+      "step": 5056
+    },
+    {
+      "epoch": 0.8193454309786131,
+      "grad_norm": 0.8722667694091797,
+      "learning_rate": 4.264362255018898e-06,
+      "loss": 0.1467,
+      "step": 5057
+    },
+    {
+      "epoch": 0.8195074530136098,
+      "grad_norm": 0.8683182001113892,
+      "learning_rate": 4.2640524129228815e-06,
+      "loss": 0.1249,
+      "step": 5058
+    },
+    {
+      "epoch": 0.8196694750486067,
+      "grad_norm": 0.8108190298080444,
+      "learning_rate": 4.263742516851128e-06,
+      "loss": 0.125,
+      "step": 5059
+    },
+    {
+      "epoch": 0.8198314970836034,
+      "grad_norm": 1.2350842952728271,
+      "learning_rate": 4.263432566813123e-06,
+      "loss": 0.159,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8199935191186001,
+      "grad_norm": 1.0421861410140991,
+      "learning_rate": 4.263122562818349e-06,
+      "loss": 0.1681,
+      "step": 5061
+    },
+    {
+      "epoch": 0.8201555411535969,
+      "grad_norm": 0.8750341534614563,
+      "learning_rate": 4.262812504876291e-06,
+      "loss": 0.1288,
+      "step": 5062
+    },
+    {
+      "epoch": 0.8203175631885936,
+      "grad_norm": 0.8388615846633911,
+      "learning_rate": 4.262502392996436e-06,
+      "loss": 0.1289,
+      "step": 5063
+    },
+    {
+      "epoch": 0.8204795852235904,
+      "grad_norm": 0.865910530090332,
+      "learning_rate": 4.262192227188273e-06,
+      "loss": 0.1329,
+      "step": 5064
+    },
+    {
+      "epoch": 0.8206416072585871,
+      "grad_norm": 0.9594810009002686,
+      "learning_rate": 4.261882007461292e-06,
+      "loss": 0.153,
+      "step": 5065
+    },
+    {
+      "epoch": 0.820803629293584,
+      "grad_norm": 0.8188930153846741,
+      "learning_rate": 4.261571733824986e-06,
+      "loss": 0.1377,
+      "step": 5066
+    },
+    {
+      "epoch": 0.8209656513285807,
+      "grad_norm": 0.9599230289459229,
+      "learning_rate": 4.261261406288847e-06,
+      "loss": 0.1379,
+      "step": 5067
+    },
+    {
+      "epoch": 0.8211276733635774,
+      "grad_norm": 0.9220377802848816,
+      "learning_rate": 4.260951024862372e-06,
+      "loss": 0.1523,
+      "step": 5068
+    },
+    {
+      "epoch": 0.8212896953985742,
+      "grad_norm": 0.8499447703361511,
+      "learning_rate": 4.2606405895550565e-06,
+      "loss": 0.1336,
+      "step": 5069
+    },
+    {
+      "epoch": 0.8214517174335709,
+      "grad_norm": 0.9115306735038757,
+      "learning_rate": 4.2603301003763994e-06,
+      "loss": 0.1438,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8216137394685677,
+      "grad_norm": 0.9516029357910156,
+      "learning_rate": 4.260019557335902e-06,
+      "loss": 0.153,
+      "step": 5071
+    },
+    {
+      "epoch": 0.8217757615035645,
+      "grad_norm": 0.7999269962310791,
+      "learning_rate": 4.259708960443065e-06,
+      "loss": 0.1234,
+      "step": 5072
+    },
+    {
+      "epoch": 0.8219377835385613,
+      "grad_norm": 0.8605833053588867,
+      "learning_rate": 4.259398309707392e-06,
+      "loss": 0.1404,
+      "step": 5073
+    },
+    {
+      "epoch": 0.822099805573558,
+      "grad_norm": 0.8901498317718506,
+      "learning_rate": 4.259087605138388e-06,
+      "loss": 0.1481,
+      "step": 5074
+    },
+    {
+      "epoch": 0.8222618276085548,
+      "grad_norm": 0.8830893039703369,
+      "learning_rate": 4.258776846745561e-06,
+      "loss": 0.1479,
+      "step": 5075
+    },
+    {
+      "epoch": 0.8224238496435515,
+      "grad_norm": 0.9161304831504822,
+      "learning_rate": 4.2584660345384176e-06,
+      "loss": 0.1487,
+      "step": 5076
+    },
+    {
+      "epoch": 0.8225858716785482,
+      "grad_norm": 0.8231655955314636,
+      "learning_rate": 4.25815516852647e-06,
+      "loss": 0.1412,
+      "step": 5077
+    },
+    {
+      "epoch": 0.8227478937135451,
+      "grad_norm": 0.7701215744018555,
+      "learning_rate": 4.257844248719229e-06,
+      "loss": 0.1308,
+      "step": 5078
+    },
+    {
+      "epoch": 0.8229099157485418,
+      "grad_norm": 0.9185945987701416,
+      "learning_rate": 4.257533275126206e-06,
+      "loss": 0.1517,
+      "step": 5079
+    },
+    {
+      "epoch": 0.8230719377835386,
+      "grad_norm": 0.9465458989143372,
+      "learning_rate": 4.25722224775692e-06,
+      "loss": 0.1686,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8232339598185353,
+      "grad_norm": 0.8534404635429382,
+      "learning_rate": 4.256911166620885e-06,
+      "loss": 0.1425,
+      "step": 5081
+    },
+    {
+      "epoch": 0.8233959818535321,
+      "grad_norm": 0.8880043625831604,
+      "learning_rate": 4.25660003172762e-06,
+      "loss": 0.1425,
+      "step": 5082
+    },
+    {
+      "epoch": 0.8235580038885288,
+      "grad_norm": 1.02664053440094,
+      "learning_rate": 4.256288843086645e-06,
+      "loss": 0.1425,
+      "step": 5083
+    },
+    {
+      "epoch": 0.8237200259235256,
+      "grad_norm": 0.8933080434799194,
+      "learning_rate": 4.255977600707481e-06,
+      "loss": 0.1529,
+      "step": 5084
+    },
+    {
+      "epoch": 0.8238820479585224,
+      "grad_norm": 0.9547656774520874,
+      "learning_rate": 4.255666304599653e-06,
+      "loss": 0.157,
+      "step": 5085
+    },
+    {
+      "epoch": 0.8240440699935191,
+      "grad_norm": 0.8259825706481934,
+      "learning_rate": 4.255354954772684e-06,
+      "loss": 0.1333,
+      "step": 5086
+    },
+    {
+      "epoch": 0.8242060920285159,
+      "grad_norm": 0.7905333638191223,
+      "learning_rate": 4.255043551236101e-06,
+      "loss": 0.1245,
+      "step": 5087
+    },
+    {
+      "epoch": 0.8243681140635126,
+      "grad_norm": 0.8866136074066162,
+      "learning_rate": 4.2547320939994315e-06,
+      "loss": 0.147,
+      "step": 5088
+    },
+    {
+      "epoch": 0.8245301360985094,
+      "grad_norm": 1.0445470809936523,
+      "learning_rate": 4.254420583072209e-06,
+      "loss": 0.1631,
+      "step": 5089
+    },
+    {
+      "epoch": 0.8246921581335062,
+      "grad_norm": 0.865193247795105,
+      "learning_rate": 4.25410901846396e-06,
+      "loss": 0.1331,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8248541801685029,
+      "grad_norm": 0.8681652545928955,
+      "learning_rate": 4.25379740018422e-06,
+      "loss": 0.1356,
+      "step": 5091
+    },
+    {
+      "epoch": 0.8250162022034997,
+      "grad_norm": 0.803639829158783,
+      "learning_rate": 4.253485728242525e-06,
+      "loss": 0.1306,
+      "step": 5092
+    },
+    {
+      "epoch": 0.8251782242384964,
+      "grad_norm": 0.8816442489624023,
+      "learning_rate": 4.253174002648409e-06,
+      "loss": 0.1355,
+      "step": 5093
+    },
+    {
+      "epoch": 0.8253402462734932,
+      "grad_norm": 0.8702470660209656,
+      "learning_rate": 4.252862223411412e-06,
+      "loss": 0.1454,
+      "step": 5094
+    },
+    {
+      "epoch": 0.8255022683084899,
+      "grad_norm": 0.7982497811317444,
+      "learning_rate": 4.2525503905410715e-06,
+      "loss": 0.1281,
+      "step": 5095
+    },
+    {
+      "epoch": 0.8256642903434868,
+      "grad_norm": 0.8184255957603455,
+      "learning_rate": 4.252238504046931e-06,
+      "loss": 0.1397,
+      "step": 5096
+    },
+    {
+      "epoch": 0.8258263123784835,
+      "grad_norm": 0.8369611501693726,
+      "learning_rate": 4.251926563938533e-06,
+      "loss": 0.1266,
+      "step": 5097
+    },
+    {
+      "epoch": 0.8259883344134802,
+      "grad_norm": 0.7753140926361084,
+      "learning_rate": 4.251614570225421e-06,
+      "loss": 0.1157,
+      "step": 5098
+    },
+    {
+      "epoch": 0.826150356448477,
+      "grad_norm": 1.0960832834243774,
+      "learning_rate": 4.251302522917142e-06,
+      "loss": 0.1757,
+      "step": 5099
+    },
+    {
+      "epoch": 0.8263123784834737,
+      "grad_norm": 0.8341733813285828,
+      "learning_rate": 4.250990422023243e-06,
+      "loss": 0.1455,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8264744005184705,
+      "grad_norm": 0.8899952173233032,
+      "learning_rate": 4.250678267553277e-06,
+      "loss": 0.1514,
+      "step": 5101
+    },
+    {
+      "epoch": 0.8266364225534673,
+      "grad_norm": 0.850158154964447,
+      "learning_rate": 4.250366059516791e-06,
+      "loss": 0.1365,
+      "step": 5102
+    },
+    {
+      "epoch": 0.8267984445884641,
+      "grad_norm": 0.8187429904937744,
+      "learning_rate": 4.250053797923339e-06,
+      "loss": 0.1324,
+      "step": 5103
+    },
+    {
+      "epoch": 0.8269604666234608,
+      "grad_norm": 0.9219319224357605,
+      "learning_rate": 4.249741482782476e-06,
+      "loss": 0.1514,
+      "step": 5104
+    },
+    {
+      "epoch": 0.8271224886584575,
+      "grad_norm": 0.9627091884613037,
+      "learning_rate": 4.249429114103758e-06,
+      "loss": 0.1689,
+      "step": 5105
+    },
+    {
+      "epoch": 0.8272845106934543,
+      "grad_norm": 0.7787367105484009,
+      "learning_rate": 4.249116691896743e-06,
+      "loss": 0.1235,
+      "step": 5106
+    },
+    {
+      "epoch": 0.827446532728451,
+      "grad_norm": 0.8211469054222107,
+      "learning_rate": 4.24880421617099e-06,
+      "loss": 0.1301,
+      "step": 5107
+    },
+    {
+      "epoch": 0.8276085547634479,
+      "grad_norm": 0.8499755859375,
+      "learning_rate": 4.248491686936059e-06,
+      "loss": 0.138,
+      "step": 5108
+    },
+    {
+      "epoch": 0.8277705767984446,
+      "grad_norm": 0.8932889699935913,
+      "learning_rate": 4.248179104201515e-06,
+      "loss": 0.1618,
+      "step": 5109
+    },
+    {
+      "epoch": 0.8279325988334414,
+      "grad_norm": 0.8149641156196594,
+      "learning_rate": 4.2478664679769196e-06,
+      "loss": 0.1358,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8280946208684381,
+      "grad_norm": 0.8986614942550659,
+      "learning_rate": 4.247553778271841e-06,
+      "loss": 0.1365,
+      "step": 5111
+    },
+    {
+      "epoch": 0.8282566429034348,
+      "grad_norm": 0.9837633371353149,
+      "learning_rate": 4.247241035095846e-06,
+      "loss": 0.1583,
+      "step": 5112
+    },
+    {
+      "epoch": 0.8284186649384316,
+      "grad_norm": 0.7755539417266846,
+      "learning_rate": 4.246928238458503e-06,
+      "loss": 0.1245,
+      "step": 5113
+    },
+    {
+      "epoch": 0.8285806869734283,
+      "grad_norm": 0.8996148109436035,
+      "learning_rate": 4.246615388369384e-06,
+      "loss": 0.1397,
+      "step": 5114
+    },
+    {
+      "epoch": 0.8287427090084252,
+      "grad_norm": 1.0767894983291626,
+      "learning_rate": 4.24630248483806e-06,
+      "loss": 0.1456,
+      "step": 5115
+    },
+    {
+      "epoch": 0.8289047310434219,
+      "grad_norm": 0.9750000238418579,
+      "learning_rate": 4.245989527874107e-06,
+      "loss": 0.1515,
+      "step": 5116
+    },
+    {
+      "epoch": 0.8290667530784187,
+      "grad_norm": 0.9112747311592102,
+      "learning_rate": 4.245676517487098e-06,
+      "loss": 0.1469,
+      "step": 5117
+    },
+    {
+      "epoch": 0.8292287751134154,
+      "grad_norm": 0.7812881469726562,
+      "learning_rate": 4.245363453686614e-06,
+      "loss": 0.1129,
+      "step": 5118
+    },
+    {
+      "epoch": 0.8293907971484121,
+      "grad_norm": 0.838598370552063,
+      "learning_rate": 4.245050336482231e-06,
+      "loss": 0.1267,
+      "step": 5119
+    },
+    {
+      "epoch": 0.829552819183409,
+      "grad_norm": 0.9569753408432007,
+      "learning_rate": 4.24473716588353e-06,
+      "loss": 0.1627,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8297148412184057,
+      "grad_norm": 0.9220561385154724,
+      "learning_rate": 4.244423941900095e-06,
+      "loss": 0.1537,
+      "step": 5121
+    },
+    {
+      "epoch": 0.8298768632534025,
+      "grad_norm": 0.9660263061523438,
+      "learning_rate": 4.2441106645415085e-06,
+      "loss": 0.1571,
+      "step": 5122
+    },
+    {
+      "epoch": 0.8300388852883992,
+      "grad_norm": 0.9250351190567017,
+      "learning_rate": 4.243797333817356e-06,
+      "loss": 0.1553,
+      "step": 5123
+    },
+    {
+      "epoch": 0.830200907323396,
+      "grad_norm": 0.8232597708702087,
+      "learning_rate": 4.243483949737225e-06,
+      "loss": 0.1277,
+      "step": 5124
+    },
+    {
+      "epoch": 0.8303629293583927,
+      "grad_norm": 0.9091588854789734,
+      "learning_rate": 4.2431705123107045e-06,
+      "loss": 0.1436,
+      "step": 5125
+    },
+    {
+      "epoch": 0.8305249513933896,
+      "grad_norm": 0.9420716762542725,
+      "learning_rate": 4.242857021547385e-06,
+      "loss": 0.1722,
+      "step": 5126
+    },
+    {
+      "epoch": 0.8306869734283863,
+      "grad_norm": 0.8245367407798767,
+      "learning_rate": 4.242543477456858e-06,
+      "loss": 0.1339,
+      "step": 5127
+    },
+    {
+      "epoch": 0.830848995463383,
+      "grad_norm": 0.8320567011833191,
+      "learning_rate": 4.242229880048718e-06,
+      "loss": 0.1509,
+      "step": 5128
+    },
+    {
+      "epoch": 0.8310110174983798,
+      "grad_norm": 0.8776043057441711,
+      "learning_rate": 4.241916229332559e-06,
+      "loss": 0.148,
+      "step": 5129
+    },
+    {
+      "epoch": 0.8311730395333765,
+      "grad_norm": 0.8687680959701538,
+      "learning_rate": 4.241602525317979e-06,
+      "loss": 0.1468,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8313350615683733,
+      "grad_norm": 1.0185246467590332,
+      "learning_rate": 4.241288768014576e-06,
+      "loss": 0.1771,
+      "step": 5131
+    },
+    {
+      "epoch": 0.83149708360337,
+      "grad_norm": 0.9902436137199402,
+      "learning_rate": 4.240974957431951e-06,
+      "loss": 0.1443,
+      "step": 5132
+    },
+    {
+      "epoch": 0.8316591056383669,
+      "grad_norm": 0.865923285484314,
+      "learning_rate": 4.240661093579705e-06,
+      "loss": 0.1428,
+      "step": 5133
+    },
+    {
+      "epoch": 0.8318211276733636,
+      "grad_norm": 0.8009111881256104,
+      "learning_rate": 4.240347176467442e-06,
+      "loss": 0.1209,
+      "step": 5134
+    },
+    {
+      "epoch": 0.8319831497083603,
+      "grad_norm": 0.8269287943840027,
+      "learning_rate": 4.240033206104767e-06,
+      "loss": 0.1354,
+      "step": 5135
+    },
+    {
+      "epoch": 0.8321451717433571,
+      "grad_norm": 0.8295719623565674,
+      "learning_rate": 4.2397191825012865e-06,
+      "loss": 0.1359,
+      "step": 5136
+    },
+    {
+      "epoch": 0.8323071937783538,
+      "grad_norm": 0.9293026328086853,
+      "learning_rate": 4.23940510566661e-06,
+      "loss": 0.1586,
+      "step": 5137
+    },
+    {
+      "epoch": 0.8324692158133506,
+      "grad_norm": 0.8136693239212036,
+      "learning_rate": 4.239090975610346e-06,
+      "loss": 0.1386,
+      "step": 5138
+    },
+    {
+      "epoch": 0.8326312378483474,
+      "grad_norm": 0.8754627108573914,
+      "learning_rate": 4.238776792342106e-06,
+      "loss": 0.1393,
+      "step": 5139
+    },
+    {
+      "epoch": 0.8327932598833442,
+      "grad_norm": 0.8677736520767212,
+      "learning_rate": 4.2384625558715045e-06,
+      "loss": 0.1352,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8329552819183409,
+      "grad_norm": 0.8522474765777588,
+      "learning_rate": 4.2381482662081555e-06,
+      "loss": 0.1534,
+      "step": 5141
+    },
+    {
+      "epoch": 0.8331173039533376,
+      "grad_norm": 0.7467418313026428,
+      "learning_rate": 4.237833923361676e-06,
+      "loss": 0.1175,
+      "step": 5142
+    },
+    {
+      "epoch": 0.8332793259883344,
+      "grad_norm": 0.905116856098175,
+      "learning_rate": 4.237519527341684e-06,
+      "loss": 0.1425,
+      "step": 5143
+    },
+    {
+      "epoch": 0.8334413480233311,
+      "grad_norm": 1.009063720703125,
+      "learning_rate": 4.237205078157799e-06,
+      "loss": 0.1333,
+      "step": 5144
+    },
+    {
+      "epoch": 0.833603370058328,
+      "grad_norm": 0.7382147908210754,
+      "learning_rate": 4.2368905758196436e-06,
+      "loss": 0.1214,
+      "step": 5145
+    },
+    {
+      "epoch": 0.8337653920933247,
+      "grad_norm": 1.0095723867416382,
+      "learning_rate": 4.236576020336838e-06,
+      "loss": 0.1717,
+      "step": 5146
+    },
+    {
+      "epoch": 0.8339274141283215,
+      "grad_norm": 0.9047743082046509,
+      "learning_rate": 4.236261411719009e-06,
+      "loss": 0.1274,
+      "step": 5147
+    },
+    {
+      "epoch": 0.8340894361633182,
+      "grad_norm": 0.8646296262741089,
+      "learning_rate": 4.235946749975783e-06,
+      "loss": 0.152,
+      "step": 5148
+    },
+    {
+      "epoch": 0.8342514581983149,
+      "grad_norm": 0.8817318677902222,
+      "learning_rate": 4.235632035116788e-06,
+      "loss": 0.1393,
+      "step": 5149
+    },
+    {
+      "epoch": 0.8344134802333117,
+      "grad_norm": 0.9330871105194092,
+      "learning_rate": 4.235317267151652e-06,
+      "loss": 0.1393,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8345755022683085,
+      "grad_norm": 0.9412763714790344,
+      "learning_rate": 4.235002446090007e-06,
+      "loss": 0.167,
+      "step": 5151
+    },
+    {
+      "epoch": 0.8347375243033053,
+      "grad_norm": 0.8925046920776367,
+      "learning_rate": 4.234687571941486e-06,
+      "loss": 0.1515,
+      "step": 5152
+    },
+    {
+      "epoch": 0.834899546338302,
+      "grad_norm": 0.7668562531471252,
+      "learning_rate": 4.234372644715723e-06,
+      "loss": 0.1195,
+      "step": 5153
+    },
+    {
+      "epoch": 0.8350615683732988,
+      "grad_norm": 0.8491714596748352,
+      "learning_rate": 4.234057664422354e-06,
+      "loss": 0.1408,
+      "step": 5154
+    },
+    {
+      "epoch": 0.8352235904082955,
+      "grad_norm": 0.8507294654846191,
+      "learning_rate": 4.233742631071017e-06,
+      "loss": 0.1333,
+      "step": 5155
+    },
+    {
+      "epoch": 0.8353856124432922,
+      "grad_norm": 0.9354631900787354,
+      "learning_rate": 4.2334275446713515e-06,
+      "loss": 0.1507,
+      "step": 5156
+    },
+    {
+      "epoch": 0.8355476344782891,
+      "grad_norm": 0.8924568295478821,
+      "learning_rate": 4.233112405232998e-06,
+      "loss": 0.1431,
+      "step": 5157
+    },
+    {
+      "epoch": 0.8357096565132858,
+      "grad_norm": 0.8328578472137451,
+      "learning_rate": 4.232797212765598e-06,
+      "loss": 0.1397,
+      "step": 5158
+    },
+    {
+      "epoch": 0.8358716785482826,
+      "grad_norm": 0.8072636127471924,
+      "learning_rate": 4.2324819672787976e-06,
+      "loss": 0.1458,
+      "step": 5159
+    },
+    {
+      "epoch": 0.8360337005832793,
+      "grad_norm": 0.813198447227478,
+      "learning_rate": 4.2321666687822405e-06,
+      "loss": 0.1361,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8361957226182761,
+      "grad_norm": 0.8552424311637878,
+      "learning_rate": 4.231851317285576e-06,
+      "loss": 0.1433,
+      "step": 5161
+    },
+    {
+      "epoch": 0.8363577446532728,
+      "grad_norm": 0.8645519018173218,
+      "learning_rate": 4.231535912798452e-06,
+      "loss": 0.1511,
+      "step": 5162
+    },
+    {
+      "epoch": 0.8365197666882696,
+      "grad_norm": 0.8857008814811707,
+      "learning_rate": 4.2312204553305195e-06,
+      "loss": 0.1396,
+      "step": 5163
+    },
+    {
+      "epoch": 0.8366817887232664,
+      "grad_norm": 0.9378755688667297,
+      "learning_rate": 4.23090494489143e-06,
+      "loss": 0.1589,
+      "step": 5164
+    },
+    {
+      "epoch": 0.8368438107582631,
+      "grad_norm": 0.8300303220748901,
+      "learning_rate": 4.230589381490837e-06,
+      "loss": 0.1385,
+      "step": 5165
+    },
+    {
+      "epoch": 0.8370058327932599,
+      "grad_norm": 0.8071770668029785,
+      "learning_rate": 4.230273765138399e-06,
+      "loss": 0.1184,
+      "step": 5166
+    },
+    {
+      "epoch": 0.8371678548282566,
+      "grad_norm": 0.7939083576202393,
+      "learning_rate": 4.229958095843769e-06,
+      "loss": 0.1276,
+      "step": 5167
+    },
+    {
+      "epoch": 0.8373298768632534,
+      "grad_norm": 0.8759216070175171,
+      "learning_rate": 4.229642373616609e-06,
+      "loss": 0.1323,
+      "step": 5168
+    },
+    {
+      "epoch": 0.8374918988982502,
+      "grad_norm": 0.8885746598243713,
+      "learning_rate": 4.2293265984665775e-06,
+      "loss": 0.159,
+      "step": 5169
+    },
+    {
+      "epoch": 0.837653920933247,
+      "grad_norm": 0.9272924065589905,
+      "learning_rate": 4.229010770403337e-06,
+      "loss": 0.1432,
+      "step": 5170
+    },
+    {
+      "epoch": 0.8378159429682437,
+      "grad_norm": 0.7635154724121094,
+      "learning_rate": 4.228694889436552e-06,
+      "loss": 0.1129,
+      "step": 5171
+    },
+    {
+      "epoch": 0.8379779650032404,
+      "grad_norm": 0.9018572568893433,
+      "learning_rate": 4.228378955575885e-06,
+      "loss": 0.1413,
+      "step": 5172
+    },
+    {
+      "epoch": 0.8381399870382372,
+      "grad_norm": 0.8228177428245544,
+      "learning_rate": 4.228062968831006e-06,
+      "loss": 0.1292,
+      "step": 5173
+    },
+    {
+      "epoch": 0.8383020090732339,
+      "grad_norm": 0.7802498936653137,
+      "learning_rate": 4.227746929211582e-06,
+      "loss": 0.1169,
+      "step": 5174
+    },
+    {
+      "epoch": 0.8384640311082308,
+      "grad_norm": 0.8804471492767334,
+      "learning_rate": 4.227430836727282e-06,
+      "loss": 0.1482,
+      "step": 5175
+    },
+    {
+      "epoch": 0.8386260531432275,
+      "grad_norm": 0.8976423740386963,
+      "learning_rate": 4.227114691387779e-06,
+      "loss": 0.1446,
+      "step": 5176
+    },
+    {
+      "epoch": 0.8387880751782243,
+      "grad_norm": 0.8988388180732727,
+      "learning_rate": 4.226798493202746e-06,
+      "loss": 0.1433,
+      "step": 5177
+    },
+    {
+      "epoch": 0.838950097213221,
+      "grad_norm": 0.8048079013824463,
+      "learning_rate": 4.226482242181859e-06,
+      "loss": 0.1248,
+      "step": 5178
+    },
+    {
+      "epoch": 0.8391121192482177,
+      "grad_norm": 0.9093589782714844,
+      "learning_rate": 4.226165938334792e-06,
+      "loss": 0.1449,
+      "step": 5179
+    },
+    {
+      "epoch": 0.8392741412832145,
+      "grad_norm": 0.9186120629310608,
+      "learning_rate": 4.225849581671225e-06,
+      "loss": 0.1496,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8394361633182112,
+      "grad_norm": 0.9053645730018616,
+      "learning_rate": 4.225533172200837e-06,
+      "loss": 0.1418,
+      "step": 5181
+    },
+    {
+      "epoch": 0.8395981853532081,
+      "grad_norm": 0.8142003417015076,
+      "learning_rate": 4.225216709933309e-06,
+      "loss": 0.132,
+      "step": 5182
+    },
+    {
+      "epoch": 0.8397602073882048,
+      "grad_norm": 0.8198449611663818,
+      "learning_rate": 4.224900194878326e-06,
+      "loss": 0.1387,
+      "step": 5183
+    },
+    {
+      "epoch": 0.8399222294232016,
+      "grad_norm": 0.8121005892753601,
+      "learning_rate": 4.2245836270455706e-06,
+      "loss": 0.1237,
+      "step": 5184
+    },
+    {
+      "epoch": 0.8400842514581983,
+      "grad_norm": 0.8428120017051697,
+      "learning_rate": 4.22426700644473e-06,
+      "loss": 0.1441,
+      "step": 5185
+    },
+    {
+      "epoch": 0.840246273493195,
+      "grad_norm": 0.9493716955184937,
+      "learning_rate": 4.223950333085492e-06,
+      "loss": 0.1443,
+      "step": 5186
+    },
+    {
+      "epoch": 0.8404082955281919,
+      "grad_norm": 0.794461190700531,
+      "learning_rate": 4.2236336069775445e-06,
+      "loss": 0.1181,
+      "step": 5187
+    },
+    {
+      "epoch": 0.8405703175631886,
+      "grad_norm": 0.9470990300178528,
+      "learning_rate": 4.223316828130581e-06,
+      "loss": 0.1715,
+      "step": 5188
+    },
+    {
+      "epoch": 0.8407323395981854,
+      "grad_norm": 0.7863299250602722,
+      "learning_rate": 4.222999996554291e-06,
+      "loss": 0.1198,
+      "step": 5189
+    },
+    {
+      "epoch": 0.8408943616331821,
+      "grad_norm": 0.916057288646698,
+      "learning_rate": 4.222683112258372e-06,
+      "loss": 0.148,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8410563836681789,
+      "grad_norm": 0.8652752041816711,
+      "learning_rate": 4.222366175252519e-06,
+      "loss": 0.1299,
+      "step": 5191
+    },
+    {
+      "epoch": 0.8412184057031756,
+      "grad_norm": 0.8845123648643494,
+      "learning_rate": 4.222049185546428e-06,
+      "loss": 0.1356,
+      "step": 5192
+    },
+    {
+      "epoch": 0.8413804277381723,
+      "grad_norm": 0.9572479724884033,
+      "learning_rate": 4.2217321431498e-06,
+      "loss": 0.1432,
+      "step": 5193
+    },
+    {
+      "epoch": 0.8415424497731692,
+      "grad_norm": 0.885009765625,
+      "learning_rate": 4.221415048072335e-06,
+      "loss": 0.1171,
+      "step": 5194
+    },
+    {
+      "epoch": 0.8417044718081659,
+      "grad_norm": 1.277252435684204,
+      "learning_rate": 4.221097900323735e-06,
+      "loss": 0.1719,
+      "step": 5195
+    },
+    {
+      "epoch": 0.8418664938431627,
+      "grad_norm": 1.0181325674057007,
+      "learning_rate": 4.220780699913704e-06,
+      "loss": 0.18,
+      "step": 5196
+    },
+    {
+      "epoch": 0.8420285158781594,
+      "grad_norm": 0.9635038375854492,
+      "learning_rate": 4.220463446851948e-06,
+      "loss": 0.1485,
+      "step": 5197
+    },
+    {
+      "epoch": 0.8421905379131562,
+      "grad_norm": 0.8910326361656189,
+      "learning_rate": 4.220146141148174e-06,
+      "loss": 0.1497,
+      "step": 5198
+    },
+    {
+      "epoch": 0.842352559948153,
+      "grad_norm": 0.8969202041625977,
+      "learning_rate": 4.2198287828120905e-06,
+      "loss": 0.1315,
+      "step": 5199
+    },
+    {
+      "epoch": 0.8425145819831497,
+      "grad_norm": 0.8396177291870117,
+      "learning_rate": 4.219511371853408e-06,
+      "loss": 0.1371,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8426766040181465,
+      "grad_norm": 0.8654994964599609,
+      "learning_rate": 4.21919390828184e-06,
+      "loss": 0.1358,
+      "step": 5201
+    },
+    {
+      "epoch": 0.8428386260531432,
+      "grad_norm": 0.877016544342041,
+      "learning_rate": 4.2188763921070974e-06,
+      "loss": 0.1539,
+      "step": 5202
+    },
+    {
+      "epoch": 0.84300064808814,
+      "grad_norm": 0.7908318042755127,
+      "learning_rate": 4.218558823338898e-06,
+      "loss": 0.1257,
+      "step": 5203
+    },
+    {
+      "epoch": 0.8431626701231367,
+      "grad_norm": 0.9762617349624634,
+      "learning_rate": 4.2182412019869556e-06,
+      "loss": 0.1717,
+      "step": 5204
+    },
+    {
+      "epoch": 0.8433246921581335,
+      "grad_norm": 0.8188162446022034,
+      "learning_rate": 4.217923528060992e-06,
+      "loss": 0.1281,
+      "step": 5205
+    },
+    {
+      "epoch": 0.8434867141931303,
+      "grad_norm": 0.8744212985038757,
+      "learning_rate": 4.217605801570725e-06,
+      "loss": 0.1317,
+      "step": 5206
+    },
+    {
+      "epoch": 0.843648736228127,
+      "grad_norm": 1.0196107625961304,
+      "learning_rate": 4.217288022525877e-06,
+      "loss": 0.1875,
+      "step": 5207
+    },
+    {
+      "epoch": 0.8438107582631238,
+      "grad_norm": 0.7780096530914307,
+      "learning_rate": 4.216970190936171e-06,
+      "loss": 0.1387,
+      "step": 5208
+    },
+    {
+      "epoch": 0.8439727802981205,
+      "grad_norm": 0.9348713755607605,
+      "learning_rate": 4.216652306811333e-06,
+      "loss": 0.1568,
+      "step": 5209
+    },
+    {
+      "epoch": 0.8441348023331173,
+      "grad_norm": 0.8681674599647522,
+      "learning_rate": 4.2163343701610884e-06,
+      "loss": 0.1436,
+      "step": 5210
+    },
+    {
+      "epoch": 0.844296824368114,
+      "grad_norm": 0.8520421981811523,
+      "learning_rate": 4.216016380995166e-06,
+      "loss": 0.1438,
+      "step": 5211
+    },
+    {
+      "epoch": 0.8444588464031109,
+      "grad_norm": 0.8671303391456604,
+      "learning_rate": 4.215698339323294e-06,
+      "loss": 0.1472,
+      "step": 5212
+    },
+    {
+      "epoch": 0.8446208684381076,
+      "grad_norm": 0.9922376275062561,
+      "learning_rate": 4.215380245155205e-06,
+      "loss": 0.1798,
+      "step": 5213
+    },
+    {
+      "epoch": 0.8447828904731044,
+      "grad_norm": 0.8721550703048706,
+      "learning_rate": 4.215062098500632e-06,
+      "loss": 0.1332,
+      "step": 5214
+    },
+    {
+      "epoch": 0.8449449125081011,
+      "grad_norm": 0.8614236116409302,
+      "learning_rate": 4.214743899369309e-06,
+      "loss": 0.1504,
+      "step": 5215
+    },
+    {
+      "epoch": 0.8451069345430978,
+      "grad_norm": 0.8543869853019714,
+      "learning_rate": 4.214425647770972e-06,
+      "loss": 0.1403,
+      "step": 5216
+    },
+    {
+      "epoch": 0.8452689565780946,
+      "grad_norm": 0.7991150617599487,
+      "learning_rate": 4.21410734371536e-06,
+      "loss": 0.1359,
+      "step": 5217
+    },
+    {
+      "epoch": 0.8454309786130914,
+      "grad_norm": 0.9248189926147461,
+      "learning_rate": 4.213788987212211e-06,
+      "loss": 0.1542,
+      "step": 5218
+    },
+    {
+      "epoch": 0.8455930006480882,
+      "grad_norm": 0.950682520866394,
+      "learning_rate": 4.213470578271265e-06,
+      "loss": 0.169,
+      "step": 5219
+    },
+    {
+      "epoch": 0.8457550226830849,
+      "grad_norm": 0.9078273177146912,
+      "learning_rate": 4.213152116902267e-06,
+      "loss": 0.1633,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8459170447180817,
+      "grad_norm": 0.8999932408332825,
+      "learning_rate": 4.21283360311496e-06,
+      "loss": 0.1556,
+      "step": 5221
+    },
+    {
+      "epoch": 0.8460790667530784,
+      "grad_norm": 0.7840633988380432,
+      "learning_rate": 4.212515036919089e-06,
+      "loss": 0.1242,
+      "step": 5222
+    },
+    {
+      "epoch": 0.8462410887880751,
+      "grad_norm": 0.8409323692321777,
+      "learning_rate": 4.212196418324402e-06,
+      "loss": 0.1342,
+      "step": 5223
+    },
+    {
+      "epoch": 0.846403110823072,
+      "grad_norm": 0.7986196279525757,
+      "learning_rate": 4.211877747340649e-06,
+      "loss": 0.1334,
+      "step": 5224
+    },
+    {
+      "epoch": 0.8465651328580687,
+      "grad_norm": 0.8227624297142029,
+      "learning_rate": 4.211559023977579e-06,
+      "loss": 0.1352,
+      "step": 5225
+    },
+    {
+      "epoch": 0.8467271548930655,
+      "grad_norm": 0.7546223402023315,
+      "learning_rate": 4.211240248244945e-06,
+      "loss": 0.1311,
+      "step": 5226
+    },
+    {
+      "epoch": 0.8468891769280622,
+      "grad_norm": 0.7613492012023926,
+      "learning_rate": 4.2109214201525e-06,
+      "loss": 0.1243,
+      "step": 5227
+    },
+    {
+      "epoch": 0.847051198963059,
+      "grad_norm": 0.9470206499099731,
+      "learning_rate": 4.21060253971e-06,
+      "loss": 0.1578,
+      "step": 5228
+    },
+    {
+      "epoch": 0.8472132209980557,
+      "grad_norm": 0.716661274433136,
+      "learning_rate": 4.210283606927203e-06,
+      "loss": 0.1045,
+      "step": 5229
+    },
+    {
+      "epoch": 0.8473752430330524,
+      "grad_norm": 0.8112010359764099,
+      "learning_rate": 4.2099646218138655e-06,
+      "loss": 0.1199,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8475372650680493,
+      "grad_norm": 0.9128946661949158,
+      "learning_rate": 4.209645584379748e-06,
+      "loss": 0.1431,
+      "step": 5231
+    },
+    {
+      "epoch": 0.847699287103046,
+      "grad_norm": 1.0120484828948975,
+      "learning_rate": 4.209326494634614e-06,
+      "loss": 0.1603,
+      "step": 5232
+    },
+    {
+      "epoch": 0.8478613091380428,
+      "grad_norm": 0.7939407229423523,
+      "learning_rate": 4.209007352588226e-06,
+      "loss": 0.1269,
+      "step": 5233
+    },
+    {
+      "epoch": 0.8480233311730395,
+      "grad_norm": 0.8421609401702881,
+      "learning_rate": 4.208688158250348e-06,
+      "loss": 0.1258,
+      "step": 5234
+    },
+    {
+      "epoch": 0.8481853532080363,
+      "grad_norm": 0.8701285719871521,
+      "learning_rate": 4.208368911630747e-06,
+      "loss": 0.1282,
+      "step": 5235
+    },
+    {
+      "epoch": 0.848347375243033,
+      "grad_norm": 0.9212756752967834,
+      "learning_rate": 4.2080496127391914e-06,
+      "loss": 0.13,
+      "step": 5236
+    },
+    {
+      "epoch": 0.8485093972780298,
+      "grad_norm": 0.8270485997200012,
+      "learning_rate": 4.207730261585452e-06,
+      "loss": 0.1358,
+      "step": 5237
+    },
+    {
+      "epoch": 0.8486714193130266,
+      "grad_norm": 0.8947330713272095,
+      "learning_rate": 4.207410858179298e-06,
+      "loss": 0.1462,
+      "step": 5238
+    },
+    {
+      "epoch": 0.8488334413480233,
+      "grad_norm": 0.7907639741897583,
+      "learning_rate": 4.207091402530504e-06,
+      "loss": 0.1296,
+      "step": 5239
+    },
+    {
+      "epoch": 0.8489954633830201,
+      "grad_norm": 0.956722617149353,
+      "learning_rate": 4.206771894648846e-06,
+      "loss": 0.157,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8491574854180168,
+      "grad_norm": 0.8899962306022644,
+      "learning_rate": 4.206452334544096e-06,
+      "loss": 0.1576,
+      "step": 5241
+    },
+    {
+      "epoch": 0.8493195074530137,
+      "grad_norm": 0.8392747640609741,
+      "learning_rate": 4.206132722226035e-06,
+      "loss": 0.132,
+      "step": 5242
+    },
+    {
+      "epoch": 0.8494815294880104,
+      "grad_norm": 0.9162569642066956,
+      "learning_rate": 4.205813057704441e-06,
+      "loss": 0.1577,
+      "step": 5243
+    },
+    {
+      "epoch": 0.8496435515230071,
+      "grad_norm": 0.9247191548347473,
+      "learning_rate": 4.205493340989096e-06,
+      "loss": 0.1536,
+      "step": 5244
+    },
+    {
+      "epoch": 0.8498055735580039,
+      "grad_norm": 0.8115431666374207,
+      "learning_rate": 4.2051735720897815e-06,
+      "loss": 0.1302,
+      "step": 5245
+    },
+    {
+      "epoch": 0.8499675955930006,
+      "grad_norm": 0.8264175653457642,
+      "learning_rate": 4.204853751016282e-06,
+      "loss": 0.1469,
+      "step": 5246
+    },
+    {
+      "epoch": 0.8501296176279974,
+      "grad_norm": 0.8077431321144104,
+      "learning_rate": 4.2045338777783844e-06,
+      "loss": 0.141,
+      "step": 5247
+    },
+    {
+      "epoch": 0.8502916396629941,
+      "grad_norm": 0.9603082537651062,
+      "learning_rate": 4.204213952385875e-06,
+      "loss": 0.1644,
+      "step": 5248
+    },
+    {
+      "epoch": 0.850453661697991,
+      "grad_norm": 0.8778827786445618,
+      "learning_rate": 4.2038939748485416e-06,
+      "loss": 0.1475,
+      "step": 5249
+    },
+    {
+      "epoch": 0.8506156837329877,
+      "grad_norm": 0.7682041525840759,
+      "learning_rate": 4.203573945176177e-06,
+      "loss": 0.1167,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8507777057679844,
+      "grad_norm": 0.8808218836784363,
+      "learning_rate": 4.203253863378571e-06,
+      "loss": 0.137,
+      "step": 5251
+    },
+    {
+      "epoch": 0.8509397278029812,
+      "grad_norm": 0.7965406775474548,
+      "learning_rate": 4.202933729465519e-06,
+      "loss": 0.131,
+      "step": 5252
+    },
+    {
+      "epoch": 0.8511017498379779,
+      "grad_norm": 0.8025760650634766,
+      "learning_rate": 4.202613543446817e-06,
+      "loss": 0.1314,
+      "step": 5253
+    },
+    {
+      "epoch": 0.8512637718729748,
+      "grad_norm": 0.9073064923286438,
+      "learning_rate": 4.20229330533226e-06,
+      "loss": 0.1461,
+      "step": 5254
+    },
+    {
+      "epoch": 0.8514257939079715,
+      "grad_norm": 0.90974360704422,
+      "learning_rate": 4.201973015131647e-06,
+      "loss": 0.1437,
+      "step": 5255
+    },
+    {
+      "epoch": 0.8515878159429683,
+      "grad_norm": 0.9711911082267761,
+      "learning_rate": 4.201652672854779e-06,
+      "loss": 0.1507,
+      "step": 5256
+    },
+    {
+      "epoch": 0.851749837977965,
+      "grad_norm": 0.9854370355606079,
+      "learning_rate": 4.2013322785114574e-06,
+      "loss": 0.1656,
+      "step": 5257
+    },
+    {
+      "epoch": 0.8519118600129617,
+      "grad_norm": 0.7474818229675293,
+      "learning_rate": 4.201011832111485e-06,
+      "loss": 0.1203,
+      "step": 5258
+    },
+    {
+      "epoch": 0.8520738820479585,
+      "grad_norm": 0.90165776014328,
+      "learning_rate": 4.200691333664666e-06,
+      "loss": 0.1446,
+      "step": 5259
+    },
+    {
+      "epoch": 0.8522359040829552,
+      "grad_norm": 0.8739867210388184,
+      "learning_rate": 4.2003707831808086e-06,
+      "loss": 0.1355,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8523979261179521,
+      "grad_norm": 0.9299155473709106,
+      "learning_rate": 4.20005018066972e-06,
+      "loss": 0.1332,
+      "step": 5261
+    },
+    {
+      "epoch": 0.8525599481529488,
+      "grad_norm": 0.8256843686103821,
+      "learning_rate": 4.199729526141209e-06,
+      "loss": 0.1293,
+      "step": 5262
+    },
+    {
+      "epoch": 0.8527219701879456,
+      "grad_norm": 0.8616979122161865,
+      "learning_rate": 4.199408819605089e-06,
+      "loss": 0.1399,
+      "step": 5263
+    },
+    {
+      "epoch": 0.8528839922229423,
+      "grad_norm": 0.9190691709518433,
+      "learning_rate": 4.199088061071172e-06,
+      "loss": 0.1357,
+      "step": 5264
+    },
+    {
+      "epoch": 0.8530460142579391,
+      "grad_norm": 0.9460611939430237,
+      "learning_rate": 4.19876725054927e-06,
+      "loss": 0.1527,
+      "step": 5265
+    },
+    {
+      "epoch": 0.8532080362929358,
+      "grad_norm": 0.9079461693763733,
+      "learning_rate": 4.198446388049203e-06,
+      "loss": 0.1596,
+      "step": 5266
+    },
+    {
+      "epoch": 0.8533700583279326,
+      "grad_norm": 0.7783133387565613,
+      "learning_rate": 4.198125473580786e-06,
+      "loss": 0.1377,
+      "step": 5267
+    },
+    {
+      "epoch": 0.8535320803629294,
+      "grad_norm": 0.9109153151512146,
+      "learning_rate": 4.197804507153838e-06,
+      "loss": 0.1472,
+      "step": 5268
+    },
+    {
+      "epoch": 0.8536941023979261,
+      "grad_norm": 0.8502582907676697,
+      "learning_rate": 4.197483488778182e-06,
+      "loss": 0.137,
+      "step": 5269
+    },
+    {
+      "epoch": 0.8538561244329229,
+      "grad_norm": 0.7953091859817505,
+      "learning_rate": 4.197162418463639e-06,
+      "loss": 0.1279,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8540181464679196,
+      "grad_norm": 0.8548662662506104,
+      "learning_rate": 4.196841296220033e-06,
+      "loss": 0.1379,
+      "step": 5271
+    },
+    {
+      "epoch": 0.8541801685029164,
+      "grad_norm": 0.8967195749282837,
+      "learning_rate": 4.1965201220571895e-06,
+      "loss": 0.1413,
+      "step": 5272
+    },
+    {
+      "epoch": 0.8543421905379132,
+      "grad_norm": 0.771815836429596,
+      "learning_rate": 4.1961988959849355e-06,
+      "loss": 0.1281,
+      "step": 5273
+    },
+    {
+      "epoch": 0.8545042125729099,
+      "grad_norm": 1.0291990041732788,
+      "learning_rate": 4.1958776180131e-06,
+      "loss": 0.1764,
+      "step": 5274
+    },
+    {
+      "epoch": 0.8546662346079067,
+      "grad_norm": 0.8116664290428162,
+      "learning_rate": 4.195556288151513e-06,
+      "loss": 0.1252,
+      "step": 5275
+    },
+    {
+      "epoch": 0.8548282566429034,
+      "grad_norm": 0.8838712573051453,
+      "learning_rate": 4.1952349064100074e-06,
+      "loss": 0.1545,
+      "step": 5276
+    },
+    {
+      "epoch": 0.8549902786779002,
+      "grad_norm": 0.8008884787559509,
+      "learning_rate": 4.194913472798415e-06,
+      "loss": 0.1244,
+      "step": 5277
+    },
+    {
+      "epoch": 0.8551523007128969,
+      "grad_norm": 0.890133261680603,
+      "learning_rate": 4.194591987326574e-06,
+      "loss": 0.1484,
+      "step": 5278
+    },
+    {
+      "epoch": 0.8553143227478938,
+      "grad_norm": 0.8935836553573608,
+      "learning_rate": 4.194270450004317e-06,
+      "loss": 0.1325,
+      "step": 5279
+    },
+    {
+      "epoch": 0.8554763447828905,
+      "grad_norm": 0.888762354850769,
+      "learning_rate": 4.193948860841485e-06,
+      "loss": 0.1455,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8556383668178872,
+      "grad_norm": 0.7954049706459045,
+      "learning_rate": 4.193627219847918e-06,
+      "loss": 0.1257,
+      "step": 5281
+    },
+    {
+      "epoch": 0.855800388852884,
+      "grad_norm": 0.8666353821754456,
+      "learning_rate": 4.193305527033456e-06,
+      "loss": 0.1473,
+      "step": 5282
+    },
+    {
+      "epoch": 0.8559624108878807,
+      "grad_norm": 0.8803139328956604,
+      "learning_rate": 4.192983782407941e-06,
+      "loss": 0.1557,
+      "step": 5283
+    },
+    {
+      "epoch": 0.8561244329228775,
+      "grad_norm": 0.81354159116745,
+      "learning_rate": 4.192661985981221e-06,
+      "loss": 0.1376,
+      "step": 5284
+    },
+    {
+      "epoch": 0.8562864549578743,
+      "grad_norm": 0.936042070388794,
+      "learning_rate": 4.19234013776314e-06,
+      "loss": 0.1426,
+      "step": 5285
+    },
+    {
+      "epoch": 0.8564484769928711,
+      "grad_norm": 0.8939327597618103,
+      "learning_rate": 4.192018237763547e-06,
+      "loss": 0.1422,
+      "step": 5286
+    },
+    {
+      "epoch": 0.8566104990278678,
+      "grad_norm": 0.806371808052063,
+      "learning_rate": 4.19169628599229e-06,
+      "loss": 0.1313,
+      "step": 5287
+    },
+    {
+      "epoch": 0.8567725210628645,
+      "grad_norm": 0.9480794072151184,
+      "learning_rate": 4.19137428245922e-06,
+      "loss": 0.1468,
+      "step": 5288
+    },
+    {
+      "epoch": 0.8569345430978613,
+      "grad_norm": 0.8664402365684509,
+      "learning_rate": 4.191052227174189e-06,
+      "loss": 0.1419,
+      "step": 5289
+    },
+    {
+      "epoch": 0.857096565132858,
+      "grad_norm": 0.8385558724403381,
+      "learning_rate": 4.190730120147054e-06,
+      "loss": 0.1322,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8572585871678549,
+      "grad_norm": 0.8843077421188354,
+      "learning_rate": 4.190407961387668e-06,
+      "loss": 0.1497,
+      "step": 5291
+    },
+    {
+      "epoch": 0.8574206092028516,
+      "grad_norm": 0.8863962292671204,
+      "learning_rate": 4.190085750905889e-06,
+      "loss": 0.1397,
+      "step": 5292
+    },
+    {
+      "epoch": 0.8575826312378484,
+      "grad_norm": 0.8218333125114441,
+      "learning_rate": 4.189763488711576e-06,
+      "loss": 0.123,
+      "step": 5293
+    },
+    {
+      "epoch": 0.8577446532728451,
+      "grad_norm": 0.8600748777389526,
+      "learning_rate": 4.189441174814589e-06,
+      "loss": 0.1472,
+      "step": 5294
+    },
+    {
+      "epoch": 0.8579066753078418,
+      "grad_norm": 0.7937850952148438,
+      "learning_rate": 4.189118809224792e-06,
+      "loss": 0.1289,
+      "step": 5295
+    },
+    {
+      "epoch": 0.8580686973428386,
+      "grad_norm": 0.8559263348579407,
+      "learning_rate": 4.188796391952046e-06,
+      "loss": 0.1388,
+      "step": 5296
+    },
+    {
+      "epoch": 0.8582307193778353,
+      "grad_norm": 0.8307374119758606,
+      "learning_rate": 4.1884739230062165e-06,
+      "loss": 0.1346,
+      "step": 5297
+    },
+    {
+      "epoch": 0.8583927414128322,
+      "grad_norm": 0.8185672760009766,
+      "learning_rate": 4.188151402397172e-06,
+      "loss": 0.1309,
+      "step": 5298
+    },
+    {
+      "epoch": 0.8585547634478289,
+      "grad_norm": 0.8880597949028015,
+      "learning_rate": 4.187828830134779e-06,
+      "loss": 0.1561,
+      "step": 5299
+    },
+    {
+      "epoch": 0.8587167854828257,
+      "grad_norm": 0.7549867630004883,
+      "learning_rate": 4.187506206228909e-06,
+      "loss": 0.1112,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8588788075178224,
+      "grad_norm": 0.9183105230331421,
+      "learning_rate": 4.187183530689433e-06,
+      "loss": 0.1465,
+      "step": 5301
+    },
+    {
+      "epoch": 0.8590408295528191,
+      "grad_norm": 0.8254534602165222,
+      "learning_rate": 4.1868608035262225e-06,
+      "loss": 0.1268,
+      "step": 5302
+    },
+    {
+      "epoch": 0.859202851587816,
+      "grad_norm": 0.8823205232620239,
+      "learning_rate": 4.186538024749155e-06,
+      "loss": 0.1542,
+      "step": 5303
+    },
+    {
+      "epoch": 0.8593648736228127,
+      "grad_norm": 0.793454110622406,
+      "learning_rate": 4.186215194368105e-06,
+      "loss": 0.1343,
+      "step": 5304
+    },
+    {
+      "epoch": 0.8595268956578095,
+      "grad_norm": 0.820057213306427,
+      "learning_rate": 4.18589231239295e-06,
+      "loss": 0.1341,
+      "step": 5305
+    },
+    {
+      "epoch": 0.8596889176928062,
+      "grad_norm": 0.7398642301559448,
+      "learning_rate": 4.18556937883357e-06,
+      "loss": 0.1248,
+      "step": 5306
+    },
+    {
+      "epoch": 0.859850939727803,
+      "grad_norm": 0.8161394596099854,
+      "learning_rate": 4.185246393699847e-06,
+      "loss": 0.131,
+      "step": 5307
+    },
+    {
+      "epoch": 0.8600129617627997,
+      "grad_norm": 0.9015643000602722,
+      "learning_rate": 4.184923357001661e-06,
+      "loss": 0.1234,
+      "step": 5308
+    },
+    {
+      "epoch": 0.8601749837977966,
+      "grad_norm": 0.8390319347381592,
+      "learning_rate": 4.184600268748899e-06,
+      "loss": 0.1388,
+      "step": 5309
+    },
+    {
+      "epoch": 0.8603370058327933,
+      "grad_norm": 0.9225760102272034,
+      "learning_rate": 4.184277128951445e-06,
+      "loss": 0.1541,
+      "step": 5310
+    },
+    {
+      "epoch": 0.86049902786779,
+      "grad_norm": 0.8592953085899353,
+      "learning_rate": 4.183953937619187e-06,
+      "loss": 0.1292,
+      "step": 5311
+    },
+    {
+      "epoch": 0.8606610499027868,
+      "grad_norm": 0.9123047590255737,
+      "learning_rate": 4.1836306947620135e-06,
+      "loss": 0.152,
+      "step": 5312
+    },
+    {
+      "epoch": 0.8608230719377835,
+      "grad_norm": 0.903361439704895,
+      "learning_rate": 4.183307400389815e-06,
+      "loss": 0.1498,
+      "step": 5313
+    },
+    {
+      "epoch": 0.8609850939727803,
+      "grad_norm": 0.8942160606384277,
+      "learning_rate": 4.182984054512483e-06,
+      "loss": 0.1397,
+      "step": 5314
+    },
+    {
+      "epoch": 0.861147116007777,
+      "grad_norm": 0.9034626483917236,
+      "learning_rate": 4.1826606571399134e-06,
+      "loss": 0.1574,
+      "step": 5315
+    },
+    {
+      "epoch": 0.8613091380427739,
+      "grad_norm": 0.8866369128227234,
+      "learning_rate": 4.182337208281998e-06,
+      "loss": 0.1503,
+      "step": 5316
+    },
+    {
+      "epoch": 0.8614711600777706,
+      "grad_norm": 0.8173493146896362,
+      "learning_rate": 4.182013707948635e-06,
+      "loss": 0.1335,
+      "step": 5317
+    },
+    {
+      "epoch": 0.8616331821127673,
+      "grad_norm": 0.9902571439743042,
+      "learning_rate": 4.181690156149724e-06,
+      "loss": 0.158,
+      "step": 5318
+    },
+    {
+      "epoch": 0.8617952041477641,
+      "grad_norm": 0.8877143859863281,
+      "learning_rate": 4.181366552895163e-06,
+      "loss": 0.1396,
+      "step": 5319
+    },
+    {
+      "epoch": 0.8619572261827608,
+      "grad_norm": 0.9022709131240845,
+      "learning_rate": 4.1810428981948555e-06,
+      "loss": 0.145,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8621192482177576,
+      "grad_norm": 0.8367879986763,
+      "learning_rate": 4.180719192058702e-06,
+      "loss": 0.135,
+      "step": 5321
+    },
+    {
+      "epoch": 0.8622812702527544,
+      "grad_norm": 0.8607010245323181,
+      "learning_rate": 4.1803954344966095e-06,
+      "loss": 0.1522,
+      "step": 5322
+    },
+    {
+      "epoch": 0.8624432922877512,
+      "grad_norm": 0.9045525193214417,
+      "learning_rate": 4.180071625518482e-06,
+      "loss": 0.1469,
+      "step": 5323
+    },
+    {
+      "epoch": 0.8626053143227479,
+      "grad_norm": 0.7549746632575989,
+      "learning_rate": 4.17974776513423e-06,
+      "loss": 0.1298,
+      "step": 5324
+    },
+    {
+      "epoch": 0.8627673363577446,
+      "grad_norm": 0.8301894068717957,
+      "learning_rate": 4.17942385335376e-06,
+      "loss": 0.13,
+      "step": 5325
+    },
+    {
+      "epoch": 0.8629293583927414,
+      "grad_norm": 0.9236185550689697,
+      "learning_rate": 4.179099890186985e-06,
+      "loss": 0.1521,
+      "step": 5326
+    },
+    {
+      "epoch": 0.8630913804277381,
+      "grad_norm": 0.8897588849067688,
+      "learning_rate": 4.1787758756438166e-06,
+      "loss": 0.1416,
+      "step": 5327
+    },
+    {
+      "epoch": 0.863253402462735,
+      "grad_norm": 0.8857617974281311,
+      "learning_rate": 4.178451809734168e-06,
+      "loss": 0.1507,
+      "step": 5328
+    },
+    {
+      "epoch": 0.8634154244977317,
+      "grad_norm": 0.7978904247283936,
+      "learning_rate": 4.178127692467957e-06,
+      "loss": 0.1303,
+      "step": 5329
+    },
+    {
+      "epoch": 0.8635774465327285,
+      "grad_norm": 0.880963921546936,
+      "learning_rate": 4.1778035238550995e-06,
+      "loss": 0.1519,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8637394685677252,
+      "grad_norm": 0.9408493638038635,
+      "learning_rate": 4.177479303905514e-06,
+      "loss": 0.1644,
+      "step": 5331
+    },
+    {
+      "epoch": 0.8639014906027219,
+      "grad_norm": 0.8565788269042969,
+      "learning_rate": 4.177155032629122e-06,
+      "loss": 0.1322,
+      "step": 5332
+    },
+    {
+      "epoch": 0.8640635126377187,
+      "grad_norm": 0.8621842861175537,
+      "learning_rate": 4.176830710035843e-06,
+      "loss": 0.1456,
+      "step": 5333
+    },
+    {
+      "epoch": 0.8642255346727155,
+      "grad_norm": 1.0310484170913696,
+      "learning_rate": 4.176506336135603e-06,
+      "loss": 0.1557,
+      "step": 5334
+    },
+    {
+      "epoch": 0.8643875567077123,
+      "grad_norm": 0.9071840643882751,
+      "learning_rate": 4.176181910938326e-06,
+      "loss": 0.154,
+      "step": 5335
+    },
+    {
+      "epoch": 0.864549578742709,
+      "grad_norm": 0.8645082116127014,
+      "learning_rate": 4.175857434453939e-06,
+      "loss": 0.1404,
+      "step": 5336
+    },
+    {
+      "epoch": 0.8647116007777058,
+      "grad_norm": 0.9229341149330139,
+      "learning_rate": 4.1755329066923705e-06,
+      "loss": 0.1429,
+      "step": 5337
+    },
+    {
+      "epoch": 0.8648736228127025,
+      "grad_norm": 0.9101964831352234,
+      "learning_rate": 4.175208327663549e-06,
+      "loss": 0.1418,
+      "step": 5338
+    },
+    {
+      "epoch": 0.8650356448476992,
+      "grad_norm": 0.8676378726959229,
+      "learning_rate": 4.1748836973774075e-06,
+      "loss": 0.1481,
+      "step": 5339
+    },
+    {
+      "epoch": 0.8651976668826961,
+      "grad_norm": 0.8613499402999878,
+      "learning_rate": 4.174559015843878e-06,
+      "loss": 0.147,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8653596889176928,
+      "grad_norm": 0.8355089426040649,
+      "learning_rate": 4.174234283072894e-06,
+      "loss": 0.1394,
+      "step": 5341
+    },
+    {
+      "epoch": 0.8655217109526896,
+      "grad_norm": 0.8457769751548767,
+      "learning_rate": 4.173909499074392e-06,
+      "loss": 0.1457,
+      "step": 5342
+    },
+    {
+      "epoch": 0.8656837329876863,
+      "grad_norm": 0.7920171618461609,
+      "learning_rate": 4.173584663858311e-06,
+      "loss": 0.115,
+      "step": 5343
+    },
+    {
+      "epoch": 0.8658457550226831,
+      "grad_norm": 0.7907648682594299,
+      "learning_rate": 4.173259777434589e-06,
+      "loss": 0.1376,
+      "step": 5344
+    },
+    {
+      "epoch": 0.8660077770576798,
+      "grad_norm": 0.8158167004585266,
+      "learning_rate": 4.172934839813168e-06,
+      "loss": 0.1325,
+      "step": 5345
+    },
+    {
+      "epoch": 0.8661697990926766,
+      "grad_norm": 0.8299226760864258,
+      "learning_rate": 4.1726098510039894e-06,
+      "loss": 0.1327,
+      "step": 5346
+    },
+    {
+      "epoch": 0.8663318211276734,
+      "grad_norm": 0.7698670029640198,
+      "learning_rate": 4.172284811016996e-06,
+      "loss": 0.1341,
+      "step": 5347
+    },
+    {
+      "epoch": 0.8664938431626701,
+      "grad_norm": 0.9853094816207886,
+      "learning_rate": 4.171959719862134e-06,
+      "loss": 0.1616,
+      "step": 5348
+    },
+    {
+      "epoch": 0.8666558651976669,
+      "grad_norm": 0.7773309350013733,
+      "learning_rate": 4.171634577549351e-06,
+      "loss": 0.1279,
+      "step": 5349
+    },
+    {
+      "epoch": 0.8668178872326636,
+      "grad_norm": 0.8574365377426147,
+      "learning_rate": 4.171309384088596e-06,
+      "loss": 0.1493,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8669799092676604,
+      "grad_norm": 0.8746230006217957,
+      "learning_rate": 4.170984139489817e-06,
+      "loss": 0.1465,
+      "step": 5351
+    },
+    {
+      "epoch": 0.8671419313026572,
+      "grad_norm": 0.9685440063476562,
+      "learning_rate": 4.170658843762968e-06,
+      "loss": 0.1518,
+      "step": 5352
+    },
+    {
+      "epoch": 0.8673039533376539,
+      "grad_norm": 0.930526077747345,
+      "learning_rate": 4.170333496918001e-06,
+      "loss": 0.1746,
+      "step": 5353
+    },
+    {
+      "epoch": 0.8674659753726507,
+      "grad_norm": 0.8363896608352661,
+      "learning_rate": 4.170008098964871e-06,
+      "loss": 0.1354,
+      "step": 5354
+    },
+    {
+      "epoch": 0.8676279974076474,
+      "grad_norm": 0.8706420063972473,
+      "learning_rate": 4.1696826499135345e-06,
+      "loss": 0.1363,
+      "step": 5355
+    },
+    {
+      "epoch": 0.8677900194426442,
+      "grad_norm": 0.8633265495300293,
+      "learning_rate": 4.169357149773949e-06,
+      "loss": 0.1452,
+      "step": 5356
+    },
+    {
+      "epoch": 0.8679520414776409,
+      "grad_norm": 0.8911048173904419,
+      "learning_rate": 4.169031598556076e-06,
+      "loss": 0.1588,
+      "step": 5357
+    },
+    {
+      "epoch": 0.8681140635126378,
+      "grad_norm": 0.8662182092666626,
+      "learning_rate": 4.168705996269874e-06,
+      "loss": 0.148,
+      "step": 5358
+    },
+    {
+      "epoch": 0.8682760855476345,
+      "grad_norm": 0.9110774397850037,
+      "learning_rate": 4.168380342925307e-06,
+      "loss": 0.1447,
+      "step": 5359
+    },
+    {
+      "epoch": 0.8684381075826313,
+      "grad_norm": 0.7783262729644775,
+      "learning_rate": 4.168054638532338e-06,
+      "loss": 0.1278,
+      "step": 5360
+    },
+    {
+      "epoch": 0.868600129617628,
+      "grad_norm": 0.8996500372886658,
+      "learning_rate": 4.167728883100935e-06,
+      "loss": 0.1546,
+      "step": 5361
+    },
+    {
+      "epoch": 0.8687621516526247,
+      "grad_norm": 0.8864397406578064,
+      "learning_rate": 4.167403076641063e-06,
+      "loss": 0.1348,
+      "step": 5362
+    },
+    {
+      "epoch": 0.8689241736876215,
+      "grad_norm": 0.7758836150169373,
+      "learning_rate": 4.167077219162693e-06,
+      "loss": 0.1296,
+      "step": 5363
+    },
+    {
+      "epoch": 0.8690861957226182,
+      "grad_norm": 0.8843368291854858,
+      "learning_rate": 4.166751310675793e-06,
+      "loss": 0.1553,
+      "step": 5364
+    },
+    {
+      "epoch": 0.8692482177576151,
+      "grad_norm": 0.7979202270507812,
+      "learning_rate": 4.166425351190337e-06,
+      "loss": 0.1344,
+      "step": 5365
+    },
+    {
+      "epoch": 0.8694102397926118,
+      "grad_norm": 0.7076486945152283,
+      "learning_rate": 4.166099340716298e-06,
+      "loss": 0.1014,
+      "step": 5366
+    },
+    {
+      "epoch": 0.8695722618276086,
+      "grad_norm": 0.8106122612953186,
+      "learning_rate": 4.165773279263651e-06,
+      "loss": 0.1328,
+      "step": 5367
+    },
+    {
+      "epoch": 0.8697342838626053,
+      "grad_norm": 0.7798807621002197,
+      "learning_rate": 4.165447166842373e-06,
+      "loss": 0.1404,
+      "step": 5368
+    },
+    {
+      "epoch": 0.869896305897602,
+      "grad_norm": 0.776866614818573,
+      "learning_rate": 4.165121003462441e-06,
+      "loss": 0.1164,
+      "step": 5369
+    },
+    {
+      "epoch": 0.8700583279325989,
+      "grad_norm": 0.8250222206115723,
+      "learning_rate": 4.164794789133837e-06,
+      "loss": 0.1363,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8702203499675956,
+      "grad_norm": 0.8319069147109985,
+      "learning_rate": 4.164468523866541e-06,
+      "loss": 0.1417,
+      "step": 5371
+    },
+    {
+      "epoch": 0.8703823720025924,
+      "grad_norm": 0.919909656047821,
+      "learning_rate": 4.164142207670536e-06,
+      "loss": 0.1611,
+      "step": 5372
+    },
+    {
+      "epoch": 0.8705443940375891,
+      "grad_norm": 0.8057491183280945,
+      "learning_rate": 4.163815840555806e-06,
+      "loss": 0.1266,
+      "step": 5373
+    },
+    {
+      "epoch": 0.8707064160725859,
+      "grad_norm": 0.8016455769538879,
+      "learning_rate": 4.163489422532338e-06,
+      "loss": 0.1256,
+      "step": 5374
+    },
+    {
+      "epoch": 0.8708684381075826,
+      "grad_norm": 0.976750910282135,
+      "learning_rate": 4.1631629536101195e-06,
+      "loss": 0.1645,
+      "step": 5375
+    },
+    {
+      "epoch": 0.8710304601425793,
+      "grad_norm": 0.9348950982093811,
+      "learning_rate": 4.162836433799139e-06,
+      "loss": 0.1425,
+      "step": 5376
+    },
+    {
+      "epoch": 0.8711924821775762,
+      "grad_norm": 0.7779992818832397,
+      "learning_rate": 4.162509863109389e-06,
+      "loss": 0.1247,
+      "step": 5377
+    },
+    {
+      "epoch": 0.8713545042125729,
+      "grad_norm": 0.9029181003570557,
+      "learning_rate": 4.162183241550858e-06,
+      "loss": 0.141,
+      "step": 5378
+    },
+    {
+      "epoch": 0.8715165262475697,
+      "grad_norm": 0.8280603885650635,
+      "learning_rate": 4.1618565691335434e-06,
+      "loss": 0.1285,
+      "step": 5379
+    },
+    {
+      "epoch": 0.8716785482825664,
+      "grad_norm": 0.792360782623291,
+      "learning_rate": 4.161529845867439e-06,
+      "loss": 0.128,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8718405703175632,
+      "grad_norm": 0.9347276091575623,
+      "learning_rate": 4.161203071762543e-06,
+      "loss": 0.1447,
+      "step": 5381
+    },
+    {
+      "epoch": 0.87200259235256,
+      "grad_norm": 0.9125970602035522,
+      "learning_rate": 4.160876246828853e-06,
+      "loss": 0.1381,
+      "step": 5382
+    },
+    {
+      "epoch": 0.8721646143875567,
+      "grad_norm": 0.8670451045036316,
+      "learning_rate": 4.160549371076369e-06,
+      "loss": 0.1382,
+      "step": 5383
+    },
+    {
+      "epoch": 0.8723266364225535,
+      "grad_norm": 0.8932413458824158,
+      "learning_rate": 4.160222444515092e-06,
+      "loss": 0.1445,
+      "step": 5384
+    },
+    {
+      "epoch": 0.8724886584575502,
+      "grad_norm": 0.8334319591522217,
+      "learning_rate": 4.159895467155026e-06,
+      "loss": 0.1404,
+      "step": 5385
+    },
+    {
+      "epoch": 0.872650680492547,
+      "grad_norm": 0.8121306896209717,
+      "learning_rate": 4.159568439006176e-06,
+      "loss": 0.1379,
+      "step": 5386
+    },
+    {
+      "epoch": 0.8728127025275437,
+      "grad_norm": 0.8688573837280273,
+      "learning_rate": 4.159241360078548e-06,
+      "loss": 0.1456,
+      "step": 5387
+    },
+    {
+      "epoch": 0.8729747245625405,
+      "grad_norm": 0.8056637644767761,
+      "learning_rate": 4.1589142303821485e-06,
+      "loss": 0.1377,
+      "step": 5388
+    },
+    {
+      "epoch": 0.8731367465975373,
+      "grad_norm": 0.8120809197425842,
+      "learning_rate": 4.15858704992699e-06,
+      "loss": 0.1306,
+      "step": 5389
+    },
+    {
+      "epoch": 0.873298768632534,
+      "grad_norm": 0.8274014592170715,
+      "learning_rate": 4.158259818723079e-06,
+      "loss": 0.1463,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8734607906675308,
+      "grad_norm": 0.9297308921813965,
+      "learning_rate": 4.157932536780432e-06,
+      "loss": 0.145,
+      "step": 5391
+    },
+    {
+      "epoch": 0.8736228127025275,
+      "grad_norm": 0.8492165803909302,
+      "learning_rate": 4.157605204109062e-06,
+      "loss": 0.1381,
+      "step": 5392
+    },
+    {
+      "epoch": 0.8737848347375243,
+      "grad_norm": 0.8986421823501587,
+      "learning_rate": 4.157277820718983e-06,
+      "loss": 0.145,
+      "step": 5393
+    },
+    {
+      "epoch": 0.873946856772521,
+      "grad_norm": 0.8207853436470032,
+      "learning_rate": 4.156950386620214e-06,
+      "loss": 0.1245,
+      "step": 5394
+    },
+    {
+      "epoch": 0.8741088788075179,
+      "grad_norm": 0.8820725679397583,
+      "learning_rate": 4.156622901822772e-06,
+      "loss": 0.1459,
+      "step": 5395
+    },
+    {
+      "epoch": 0.8742709008425146,
+      "grad_norm": 0.8960903286933899,
+      "learning_rate": 4.156295366336679e-06,
+      "loss": 0.1464,
+      "step": 5396
+    },
+    {
+      "epoch": 0.8744329228775113,
+      "grad_norm": 0.8151302337646484,
+      "learning_rate": 4.1559677801719554e-06,
+      "loss": 0.1344,
+      "step": 5397
+    },
+    {
+      "epoch": 0.8745949449125081,
+      "grad_norm": 0.8475366830825806,
+      "learning_rate": 4.155640143338625e-06,
+      "loss": 0.1303,
+      "step": 5398
+    },
+    {
+      "epoch": 0.8747569669475048,
+      "grad_norm": 0.8408517241477966,
+      "learning_rate": 4.155312455846714e-06,
+      "loss": 0.1476,
+      "step": 5399
+    },
+    {
+      "epoch": 0.8749189889825016,
+      "grad_norm": 0.8452920317649841,
+      "learning_rate": 4.154984717706246e-06,
+      "loss": 0.143,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8750810110174984,
+      "grad_norm": 0.8104416728019714,
+      "learning_rate": 4.154656928927252e-06,
+      "loss": 0.1317,
+      "step": 5401
+    },
+    {
+      "epoch": 0.8752430330524952,
+      "grad_norm": 0.8546898365020752,
+      "learning_rate": 4.15432908951976e-06,
+      "loss": 0.1447,
+      "step": 5402
+    },
+    {
+      "epoch": 0.8754050550874919,
+      "grad_norm": 0.9186674952507019,
+      "learning_rate": 4.1540011994938e-06,
+      "loss": 0.1344,
+      "step": 5403
+    },
+    {
+      "epoch": 0.8755670771224887,
+      "grad_norm": 0.8319907188415527,
+      "learning_rate": 4.153673258859406e-06,
+      "loss": 0.1459,
+      "step": 5404
+    },
+    {
+      "epoch": 0.8757290991574854,
+      "grad_norm": 0.8684608340263367,
+      "learning_rate": 4.153345267626614e-06,
+      "loss": 0.1466,
+      "step": 5405
+    },
+    {
+      "epoch": 0.8758911211924821,
+      "grad_norm": 0.8214244246482849,
+      "learning_rate": 4.153017225805456e-06,
+      "loss": 0.1257,
+      "step": 5406
+    },
+    {
+      "epoch": 0.876053143227479,
+      "grad_norm": 0.9074698090553284,
+      "learning_rate": 4.152689133405971e-06,
+      "loss": 0.1296,
+      "step": 5407
+    },
+    {
+      "epoch": 0.8762151652624757,
+      "grad_norm": 0.8300620317459106,
+      "learning_rate": 4.1523609904382e-06,
+      "loss": 0.1331,
+      "step": 5408
+    },
+    {
+      "epoch": 0.8763771872974725,
+      "grad_norm": 0.8042019009590149,
+      "learning_rate": 4.152032796912179e-06,
+      "loss": 0.1265,
+      "step": 5409
+    },
+    {
+      "epoch": 0.8765392093324692,
+      "grad_norm": 0.8547518253326416,
+      "learning_rate": 4.1517045528379544e-06,
+      "loss": 0.138,
+      "step": 5410
+    },
+    {
+      "epoch": 0.876701231367466,
+      "grad_norm": 0.9264114499092102,
+      "learning_rate": 4.1513762582255655e-06,
+      "loss": 0.1608,
+      "step": 5411
+    },
+    {
+      "epoch": 0.8768632534024627,
+      "grad_norm": 0.7929533123970032,
+      "learning_rate": 4.151047913085061e-06,
+      "loss": 0.129,
+      "step": 5412
+    },
+    {
+      "epoch": 0.8770252754374595,
+      "grad_norm": 0.9087817072868347,
+      "learning_rate": 4.150719517426485e-06,
+      "loss": 0.1529,
+      "step": 5413
+    },
+    {
+      "epoch": 0.8771872974724563,
+      "grad_norm": 0.8429016470909119,
+      "learning_rate": 4.150391071259886e-06,
+      "loss": 0.1428,
+      "step": 5414
+    },
+    {
+      "epoch": 0.877349319507453,
+      "grad_norm": 0.9927331209182739,
+      "learning_rate": 4.1500625745953145e-06,
+      "loss": 0.1493,
+      "step": 5415
+    },
+    {
+      "epoch": 0.8775113415424498,
+      "grad_norm": 0.7533746361732483,
+      "learning_rate": 4.149734027442821e-06,
+      "loss": 0.1202,
+      "step": 5416
+    },
+    {
+      "epoch": 0.8776733635774465,
+      "grad_norm": 0.8377231359481812,
+      "learning_rate": 4.14940542981246e-06,
+      "loss": 0.1399,
+      "step": 5417
+    },
+    {
+      "epoch": 0.8778353856124433,
+      "grad_norm": 1.0180152654647827,
+      "learning_rate": 4.149076781714283e-06,
+      "loss": 0.1536,
+      "step": 5418
+    },
+    {
+      "epoch": 0.87799740764744,
+      "grad_norm": 0.9475023746490479,
+      "learning_rate": 4.148748083158347e-06,
+      "loss": 0.1683,
+      "step": 5419
+    },
+    {
+      "epoch": 0.8781594296824368,
+      "grad_norm": 0.863429844379425,
+      "learning_rate": 4.1484193341547106e-06,
+      "loss": 0.1407,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8783214517174336,
+      "grad_norm": 0.7526499032974243,
+      "learning_rate": 4.14809053471343e-06,
+      "loss": 0.1262,
+      "step": 5421
+    },
+    {
+      "epoch": 0.8784834737524303,
+      "grad_norm": 0.851190984249115,
+      "learning_rate": 4.147761684844569e-06,
+      "loss": 0.1361,
+      "step": 5422
+    },
+    {
+      "epoch": 0.8786454957874271,
+      "grad_norm": 0.823888897895813,
+      "learning_rate": 4.147432784558188e-06,
+      "loss": 0.1298,
+      "step": 5423
+    },
+    {
+      "epoch": 0.8788075178224238,
+      "grad_norm": 0.8875905871391296,
+      "learning_rate": 4.147103833864349e-06,
+      "loss": 0.1347,
+      "step": 5424
+    },
+    {
+      "epoch": 0.8789695398574207,
+      "grad_norm": 0.9177705645561218,
+      "learning_rate": 4.146774832773119e-06,
+      "loss": 0.1559,
+      "step": 5425
+    },
+    {
+      "epoch": 0.8791315618924174,
+      "grad_norm": 0.8365907669067383,
+      "learning_rate": 4.146445781294566e-06,
+      "loss": 0.1295,
+      "step": 5426
+    },
+    {
+      "epoch": 0.8792935839274141,
+      "grad_norm": 0.8464487791061401,
+      "learning_rate": 4.146116679438754e-06,
+      "loss": 0.135,
+      "step": 5427
+    },
+    {
+      "epoch": 0.8794556059624109,
+      "grad_norm": 0.8880855441093445,
+      "learning_rate": 4.145787527215757e-06,
+      "loss": 0.1451,
+      "step": 5428
+    },
+    {
+      "epoch": 0.8796176279974076,
+      "grad_norm": 0.991811215877533,
+      "learning_rate": 4.145458324635643e-06,
+      "loss": 0.1647,
+      "step": 5429
+    },
+    {
+      "epoch": 0.8797796500324044,
+      "grad_norm": 0.7490970492362976,
+      "learning_rate": 4.145129071708487e-06,
+      "loss": 0.1117,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8799416720674011,
+      "grad_norm": 1.0946390628814697,
+      "learning_rate": 4.144799768444362e-06,
+      "loss": 0.1776,
+      "step": 5431
+    },
+    {
+      "epoch": 0.880103694102398,
+      "grad_norm": 0.8403367400169373,
+      "learning_rate": 4.144470414853345e-06,
+      "loss": 0.148,
+      "step": 5432
+    },
+    {
+      "epoch": 0.8802657161373947,
+      "grad_norm": 0.9710992574691772,
+      "learning_rate": 4.1441410109455126e-06,
+      "loss": 0.1562,
+      "step": 5433
+    },
+    {
+      "epoch": 0.8804277381723914,
+      "grad_norm": 0.8550183176994324,
+      "learning_rate": 4.143811556730944e-06,
+      "loss": 0.1501,
+      "step": 5434
+    },
+    {
+      "epoch": 0.8805897602073882,
+      "grad_norm": 0.92138671875,
+      "learning_rate": 4.143482052219719e-06,
+      "loss": 0.1599,
+      "step": 5435
+    },
+    {
+      "epoch": 0.8807517822423849,
+      "grad_norm": 0.8856903910636902,
+      "learning_rate": 4.143152497421922e-06,
+      "loss": 0.1497,
+      "step": 5436
+    },
+    {
+      "epoch": 0.8809138042773818,
+      "grad_norm": 0.7469887733459473,
+      "learning_rate": 4.142822892347634e-06,
+      "loss": 0.1194,
+      "step": 5437
+    },
+    {
+      "epoch": 0.8810758263123785,
+      "grad_norm": 0.8389981389045715,
+      "learning_rate": 4.142493237006941e-06,
+      "loss": 0.1428,
+      "step": 5438
+    },
+    {
+      "epoch": 0.8812378483473753,
+      "grad_norm": 0.7978473901748657,
+      "learning_rate": 4.14216353140993e-06,
+      "loss": 0.1363,
+      "step": 5439
+    },
+    {
+      "epoch": 0.881399870382372,
+      "grad_norm": 0.9018765687942505,
+      "learning_rate": 4.141833775566688e-06,
+      "loss": 0.141,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8815618924173687,
+      "grad_norm": 0.9037135243415833,
+      "learning_rate": 4.141503969487307e-06,
+      "loss": 0.1424,
+      "step": 5441
+    },
+    {
+      "epoch": 0.8817239144523655,
+      "grad_norm": 0.9119111895561218,
+      "learning_rate": 4.1411741131818765e-06,
+      "loss": 0.1625,
+      "step": 5442
+    },
+    {
+      "epoch": 0.8818859364873622,
+      "grad_norm": 0.796134352684021,
+      "learning_rate": 4.140844206660489e-06,
+      "loss": 0.1313,
+      "step": 5443
+    },
+    {
+      "epoch": 0.8820479585223591,
+      "grad_norm": 0.7966582179069519,
+      "learning_rate": 4.14051424993324e-06,
+      "loss": 0.1299,
+      "step": 5444
+    },
+    {
+      "epoch": 0.8822099805573558,
+      "grad_norm": 0.9936891794204712,
+      "learning_rate": 4.140184243010225e-06,
+      "loss": 0.156,
+      "step": 5445
+    },
+    {
+      "epoch": 0.8823720025923526,
+      "grad_norm": 0.7525246739387512,
+      "learning_rate": 4.1398541859015405e-06,
+      "loss": 0.1276,
+      "step": 5446
+    },
+    {
+      "epoch": 0.8825340246273493,
+      "grad_norm": 0.8001253604888916,
+      "learning_rate": 4.139524078617287e-06,
+      "loss": 0.1233,
+      "step": 5447
+    },
+    {
+      "epoch": 0.8826960466623461,
+      "grad_norm": 0.8744903206825256,
+      "learning_rate": 4.139193921167565e-06,
+      "loss": 0.1431,
+      "step": 5448
+    },
+    {
+      "epoch": 0.8828580686973428,
+      "grad_norm": 0.8717635273933411,
+      "learning_rate": 4.138863713562475e-06,
+      "loss": 0.1468,
+      "step": 5449
+    },
+    {
+      "epoch": 0.8830200907323396,
+      "grad_norm": 0.8670765161514282,
+      "learning_rate": 4.138533455812121e-06,
+      "loss": 0.1377,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8831821127673364,
+      "grad_norm": 0.9572997689247131,
+      "learning_rate": 4.1382031479266084e-06,
+      "loss": 0.1574,
+      "step": 5451
+    },
+    {
+      "epoch": 0.8833441348023331,
+      "grad_norm": 0.9534028768539429,
+      "learning_rate": 4.137872789916044e-06,
+      "loss": 0.176,
+      "step": 5452
+    },
+    {
+      "epoch": 0.8835061568373299,
+      "grad_norm": 1.0097438097000122,
+      "learning_rate": 4.137542381790537e-06,
+      "loss": 0.1481,
+      "step": 5453
+    },
+    {
+      "epoch": 0.8836681788723266,
+      "grad_norm": 0.7446777820587158,
+      "learning_rate": 4.137211923560195e-06,
+      "loss": 0.1169,
+      "step": 5454
+    },
+    {
+      "epoch": 0.8838302009073234,
+      "grad_norm": 0.7878828644752502,
+      "learning_rate": 4.13688141523513e-06,
+      "loss": 0.1216,
+      "step": 5455
+    },
+    {
+      "epoch": 0.8839922229423202,
+      "grad_norm": 0.9208990931510925,
+      "learning_rate": 4.136550856825455e-06,
+      "loss": 0.1597,
+      "step": 5456
+    },
+    {
+      "epoch": 0.8841542449773169,
+      "grad_norm": 0.872546911239624,
+      "learning_rate": 4.136220248341284e-06,
+      "loss": 0.1296,
+      "step": 5457
+    },
+    {
+      "epoch": 0.8843162670123137,
+      "grad_norm": 0.9872225522994995,
+      "learning_rate": 4.135889589792733e-06,
+      "loss": 0.1712,
+      "step": 5458
+    },
+    {
+      "epoch": 0.8844782890473104,
+      "grad_norm": 0.8968865871429443,
+      "learning_rate": 4.135558881189919e-06,
+      "loss": 0.1464,
+      "step": 5459
+    },
+    {
+      "epoch": 0.8846403110823072,
+      "grad_norm": 0.883246898651123,
+      "learning_rate": 4.135228122542962e-06,
+      "loss": 0.1519,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8848023331173039,
+      "grad_norm": 0.9164681434631348,
+      "learning_rate": 4.134897313861981e-06,
+      "loss": 0.1489,
+      "step": 5461
+    },
+    {
+      "epoch": 0.8849643551523008,
+      "grad_norm": 0.8244009017944336,
+      "learning_rate": 4.1345664551570985e-06,
+      "loss": 0.1366,
+      "step": 5462
+    },
+    {
+      "epoch": 0.8851263771872975,
+      "grad_norm": 0.8056451082229614,
+      "learning_rate": 4.134235546438439e-06,
+      "loss": 0.1461,
+      "step": 5463
+    },
+    {
+      "epoch": 0.8852883992222942,
+      "grad_norm": 0.8311920762062073,
+      "learning_rate": 4.133904587716126e-06,
+      "loss": 0.1495,
+      "step": 5464
+    },
+    {
+      "epoch": 0.885450421257291,
+      "grad_norm": 0.9182047843933105,
+      "learning_rate": 4.133573579000286e-06,
+      "loss": 0.1392,
+      "step": 5465
+    },
+    {
+      "epoch": 0.8856124432922877,
+      "grad_norm": 0.8886732459068298,
+      "learning_rate": 4.133242520301049e-06,
+      "loss": 0.133,
+      "step": 5466
+    },
+    {
+      "epoch": 0.8857744653272845,
+      "grad_norm": 0.6573583483695984,
+      "learning_rate": 4.1329114116285415e-06,
+      "loss": 0.1026,
+      "step": 5467
+    },
+    {
+      "epoch": 0.8859364873622813,
+      "grad_norm": 0.8312880992889404,
+      "learning_rate": 4.132580252992898e-06,
+      "loss": 0.1333,
+      "step": 5468
+    },
+    {
+      "epoch": 0.8860985093972781,
+      "grad_norm": 0.8526415228843689,
+      "learning_rate": 4.132249044404249e-06,
+      "loss": 0.1405,
+      "step": 5469
+    },
+    {
+      "epoch": 0.8862605314322748,
+      "grad_norm": 0.9522249102592468,
+      "learning_rate": 4.131917785872728e-06,
+      "loss": 0.1444,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8864225534672715,
+      "grad_norm": 0.8345977067947388,
+      "learning_rate": 4.131586477408473e-06,
+      "loss": 0.1337,
+      "step": 5471
+    },
+    {
+      "epoch": 0.8865845755022683,
+      "grad_norm": 0.8408584594726562,
+      "learning_rate": 4.13125511902162e-06,
+      "loss": 0.1444,
+      "step": 5472
+    },
+    {
+      "epoch": 0.886746597537265,
+      "grad_norm": 0.8456746935844421,
+      "learning_rate": 4.1309237107223086e-06,
+      "loss": 0.1427,
+      "step": 5473
+    },
+    {
+      "epoch": 0.8869086195722619,
+      "grad_norm": 0.8888773918151855,
+      "learning_rate": 4.130592252520677e-06,
+      "loss": 0.1442,
+      "step": 5474
+    },
+    {
+      "epoch": 0.8870706416072586,
+      "grad_norm": 0.9684158563613892,
+      "learning_rate": 4.13026074442687e-06,
+      "loss": 0.1564,
+      "step": 5475
+    },
+    {
+      "epoch": 0.8872326636422554,
+      "grad_norm": 0.9485166072845459,
+      "learning_rate": 4.129929186451028e-06,
+      "loss": 0.1569,
+      "step": 5476
+    },
+    {
+      "epoch": 0.8873946856772521,
+      "grad_norm": 0.9848198294639587,
+      "learning_rate": 4.129597578603298e-06,
+      "loss": 0.1365,
+      "step": 5477
+    },
+    {
+      "epoch": 0.8875567077122488,
+      "grad_norm": 0.9694206118583679,
+      "learning_rate": 4.129265920893826e-06,
+      "loss": 0.1594,
+      "step": 5478
+    },
+    {
+      "epoch": 0.8877187297472456,
+      "grad_norm": 0.7698529958724976,
+      "learning_rate": 4.128934213332759e-06,
+      "loss": 0.1278,
+      "step": 5479
+    },
+    {
+      "epoch": 0.8878807517822424,
+      "grad_norm": 0.814104437828064,
+      "learning_rate": 4.128602455930247e-06,
+      "loss": 0.1389,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8880427738172392,
+      "grad_norm": 0.8766177892684937,
+      "learning_rate": 4.128270648696441e-06,
+      "loss": 0.1513,
+      "step": 5481
+    },
+    {
+      "epoch": 0.8882047958522359,
+      "grad_norm": 0.796972393989563,
+      "learning_rate": 4.127938791641493e-06,
+      "loss": 0.1341,
+      "step": 5482
+    },
+    {
+      "epoch": 0.8883668178872327,
+      "grad_norm": 0.8404687643051147,
+      "learning_rate": 4.127606884775559e-06,
+      "loss": 0.1433,
+      "step": 5483
+    },
+    {
+      "epoch": 0.8885288399222294,
+      "grad_norm": 0.8364979028701782,
+      "learning_rate": 4.127274928108792e-06,
+      "loss": 0.1228,
+      "step": 5484
+    },
+    {
+      "epoch": 0.8886908619572261,
+      "grad_norm": 0.872944176197052,
+      "learning_rate": 4.12694292165135e-06,
+      "loss": 0.1582,
+      "step": 5485
+    },
+    {
+      "epoch": 0.888852883992223,
+      "grad_norm": 0.8381401300430298,
+      "learning_rate": 4.126610865413392e-06,
+      "loss": 0.1407,
+      "step": 5486
+    },
+    {
+      "epoch": 0.8890149060272197,
+      "grad_norm": 0.8389954566955566,
+      "learning_rate": 4.126278759405078e-06,
+      "loss": 0.1522,
+      "step": 5487
+    },
+    {
+      "epoch": 0.8891769280622165,
+      "grad_norm": 0.8609594702720642,
+      "learning_rate": 4.125946603636569e-06,
+      "loss": 0.149,
+      "step": 5488
+    },
+    {
+      "epoch": 0.8893389500972132,
+      "grad_norm": 0.9145193099975586,
+      "learning_rate": 4.12561439811803e-06,
+      "loss": 0.1501,
+      "step": 5489
+    },
+    {
+      "epoch": 0.88950097213221,
+      "grad_norm": 0.8599084615707397,
+      "learning_rate": 4.125282142859622e-06,
+      "loss": 0.1438,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8896629941672067,
+      "grad_norm": 0.8774310946464539,
+      "learning_rate": 4.124949837871516e-06,
+      "loss": 0.1433,
+      "step": 5491
+    },
+    {
+      "epoch": 0.8898250162022034,
+      "grad_norm": 0.8296245336532593,
+      "learning_rate": 4.124617483163876e-06,
+      "loss": 0.1127,
+      "step": 5492
+    },
+    {
+      "epoch": 0.8899870382372003,
+      "grad_norm": 0.8327217102050781,
+      "learning_rate": 4.124285078746872e-06,
+      "loss": 0.1337,
+      "step": 5493
+    },
+    {
+      "epoch": 0.890149060272197,
+      "grad_norm": 0.8418365120887756,
+      "learning_rate": 4.123952624630676e-06,
+      "loss": 0.1461,
+      "step": 5494
+    },
+    {
+      "epoch": 0.8903110823071938,
+      "grad_norm": 0.8435823917388916,
+      "learning_rate": 4.123620120825459e-06,
+      "loss": 0.1376,
+      "step": 5495
+    },
+    {
+      "epoch": 0.8904731043421905,
+      "grad_norm": 0.8990997076034546,
+      "learning_rate": 4.123287567341396e-06,
+      "loss": 0.1331,
+      "step": 5496
+    },
+    {
+      "epoch": 0.8906351263771873,
+      "grad_norm": 0.9464835524559021,
+      "learning_rate": 4.122954964188662e-06,
+      "loss": 0.1385,
+      "step": 5497
+    },
+    {
+      "epoch": 0.890797148412184,
+      "grad_norm": 0.8224294781684875,
+      "learning_rate": 4.122622311377433e-06,
+      "loss": 0.1298,
+      "step": 5498
+    },
+    {
+      "epoch": 0.8909591704471809,
+      "grad_norm": 0.8208863735198975,
+      "learning_rate": 4.122289608917888e-06,
+      "loss": 0.1374,
+      "step": 5499
+    },
+    {
+      "epoch": 0.8911211924821776,
+      "grad_norm": 0.7936856746673584,
+      "learning_rate": 4.121956856820207e-06,
+      "loss": 0.1262,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8912832145171743,
+      "grad_norm": 0.8944551348686218,
+      "learning_rate": 4.121624055094571e-06,
+      "loss": 0.1464,
+      "step": 5501
+    },
+    {
+      "epoch": 0.8914452365521711,
+      "grad_norm": 0.8844398260116577,
+      "learning_rate": 4.1212912037511634e-06,
+      "loss": 0.1422,
+      "step": 5502
+    },
+    {
+      "epoch": 0.8916072585871678,
+      "grad_norm": 0.790374755859375,
+      "learning_rate": 4.120958302800169e-06,
+      "loss": 0.1264,
+      "step": 5503
+    },
+    {
+      "epoch": 0.8917692806221647,
+      "grad_norm": 0.838116466999054,
+      "learning_rate": 4.1206253522517725e-06,
+      "loss": 0.1279,
+      "step": 5504
+    },
+    {
+      "epoch": 0.8919313026571614,
+      "grad_norm": 0.9478814601898193,
+      "learning_rate": 4.120292352116162e-06,
+      "loss": 0.1509,
+      "step": 5505
+    },
+    {
+      "epoch": 0.8920933246921582,
+      "grad_norm": 0.8016614317893982,
+      "learning_rate": 4.119959302403527e-06,
+      "loss": 0.1353,
+      "step": 5506
+    },
+    {
+      "epoch": 0.8922553467271549,
+      "grad_norm": 0.7916797399520874,
+      "learning_rate": 4.119626203124056e-06,
+      "loss": 0.1264,
+      "step": 5507
+    },
+    {
+      "epoch": 0.8924173687621516,
+      "grad_norm": 0.8813158273696899,
+      "learning_rate": 4.119293054287945e-06,
+      "loss": 0.1237,
+      "step": 5508
+    },
+    {
+      "epoch": 0.8925793907971484,
+      "grad_norm": 0.8696245551109314,
+      "learning_rate": 4.118959855905383e-06,
+      "loss": 0.1459,
+      "step": 5509
+    },
+    {
+      "epoch": 0.8927414128321451,
+      "grad_norm": 0.7300010919570923,
+      "learning_rate": 4.118626607986569e-06,
+      "loss": 0.1244,
+      "step": 5510
+    },
+    {
+      "epoch": 0.892903434867142,
+      "grad_norm": 0.8156650066375732,
+      "learning_rate": 4.118293310541697e-06,
+      "loss": 0.1353,
+      "step": 5511
+    },
+    {
+      "epoch": 0.8930654569021387,
+      "grad_norm": 0.7121819853782654,
+      "learning_rate": 4.1179599635809654e-06,
+      "loss": 0.1122,
+      "step": 5512
+    },
+    {
+      "epoch": 0.8932274789371355,
+      "grad_norm": 0.8799218535423279,
+      "learning_rate": 4.117626567114575e-06,
+      "loss": 0.1382,
+      "step": 5513
+    },
+    {
+      "epoch": 0.8933895009721322,
+      "grad_norm": 0.863865315914154,
+      "learning_rate": 4.1172931211527254e-06,
+      "loss": 0.1367,
+      "step": 5514
+    },
+    {
+      "epoch": 0.8935515230071289,
+      "grad_norm": 0.9362986087799072,
+      "learning_rate": 4.116959625705621e-06,
+      "loss": 0.151,
+      "step": 5515
+    },
+    {
+      "epoch": 0.8937135450421257,
+      "grad_norm": 1.1234683990478516,
+      "learning_rate": 4.116626080783464e-06,
+      "loss": 0.1598,
+      "step": 5516
+    },
+    {
+      "epoch": 0.8938755670771225,
+      "grad_norm": 0.8734990954399109,
+      "learning_rate": 4.116292486396463e-06,
+      "loss": 0.148,
+      "step": 5517
+    },
+    {
+      "epoch": 0.8940375891121193,
+      "grad_norm": 0.8898338079452515,
+      "learning_rate": 4.1159588425548215e-06,
+      "loss": 0.1485,
+      "step": 5518
+    },
+    {
+      "epoch": 0.894199611147116,
+      "grad_norm": 0.9480844140052795,
+      "learning_rate": 4.1156251492687505e-06,
+      "loss": 0.1463,
+      "step": 5519
+    },
+    {
+      "epoch": 0.8943616331821128,
+      "grad_norm": 0.8957088589668274,
+      "learning_rate": 4.11529140654846e-06,
+      "loss": 0.1609,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8945236552171095,
+      "grad_norm": 0.8064636588096619,
+      "learning_rate": 4.114957614404161e-06,
+      "loss": 0.1282,
+      "step": 5521
+    },
+    {
+      "epoch": 0.8946856772521062,
+      "grad_norm": 0.9250004291534424,
+      "learning_rate": 4.114623772846067e-06,
+      "loss": 0.1473,
+      "step": 5522
+    },
+    {
+      "epoch": 0.8948476992871031,
+      "grad_norm": 0.8078949451446533,
+      "learning_rate": 4.114289881884394e-06,
+      "loss": 0.1319,
+      "step": 5523
+    },
+    {
+      "epoch": 0.8950097213220998,
+      "grad_norm": 0.7498173713684082,
+      "learning_rate": 4.113955941529355e-06,
+      "loss": 0.1088,
+      "step": 5524
+    },
+    {
+      "epoch": 0.8951717433570966,
+      "grad_norm": 0.7994401454925537,
+      "learning_rate": 4.1136219517911715e-06,
+      "loss": 0.1244,
+      "step": 5525
+    },
+    {
+      "epoch": 0.8953337653920933,
+      "grad_norm": 1.1339514255523682,
+      "learning_rate": 4.113287912680061e-06,
+      "loss": 0.1646,
+      "step": 5526
+    },
+    {
+      "epoch": 0.8954957874270901,
+      "grad_norm": 0.9239411950111389,
+      "learning_rate": 4.112953824206244e-06,
+      "loss": 0.1468,
+      "step": 5527
+    },
+    {
+      "epoch": 0.8956578094620868,
+      "grad_norm": 0.9831513166427612,
+      "learning_rate": 4.112619686379944e-06,
+      "loss": 0.1582,
+      "step": 5528
+    },
+    {
+      "epoch": 0.8958198314970836,
+      "grad_norm": 0.8446866869926453,
+      "learning_rate": 4.112285499211383e-06,
+      "loss": 0.1353,
+      "step": 5529
+    },
+    {
+      "epoch": 0.8959818535320804,
+      "grad_norm": 0.8573483228683472,
+      "learning_rate": 4.111951262710788e-06,
+      "loss": 0.1379,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8961438755670771,
+      "grad_norm": 0.9121212959289551,
+      "learning_rate": 4.111616976888385e-06,
+      "loss": 0.1577,
+      "step": 5531
+    },
+    {
+      "epoch": 0.8963058976020739,
+      "grad_norm": 0.8051736950874329,
+      "learning_rate": 4.111282641754403e-06,
+      "loss": 0.1327,
+      "step": 5532
+    },
+    {
+      "epoch": 0.8964679196370706,
+      "grad_norm": 0.7908130288124084,
+      "learning_rate": 4.1109482573190705e-06,
+      "loss": 0.1223,
+      "step": 5533
+    },
+    {
+      "epoch": 0.8966299416720674,
+      "grad_norm": 0.8380170464515686,
+      "learning_rate": 4.110613823592621e-06,
+      "loss": 0.1249,
+      "step": 5534
+    },
+    {
+      "epoch": 0.8967919637070642,
+      "grad_norm": 0.9545575380325317,
+      "learning_rate": 4.110279340585285e-06,
+      "loss": 0.1605,
+      "step": 5535
+    },
+    {
+      "epoch": 0.8969539857420609,
+      "grad_norm": 0.8907868266105652,
+      "learning_rate": 4.109944808307298e-06,
+      "loss": 0.1562,
+      "step": 5536
+    },
+    {
+      "epoch": 0.8971160077770577,
+      "grad_norm": 0.8784735202789307,
+      "learning_rate": 4.109610226768897e-06,
+      "loss": 0.1361,
+      "step": 5537
+    },
+    {
+      "epoch": 0.8972780298120544,
+      "grad_norm": 0.8206544518470764,
+      "learning_rate": 4.109275595980316e-06,
+      "loss": 0.1343,
+      "step": 5538
+    },
+    {
+      "epoch": 0.8974400518470512,
+      "grad_norm": 0.7294670939445496,
+      "learning_rate": 4.108940915951798e-06,
+      "loss": 0.1152,
+      "step": 5539
+    },
+    {
+      "epoch": 0.8976020738820479,
+      "grad_norm": 0.9412605166435242,
+      "learning_rate": 4.108606186693582e-06,
+      "loss": 0.1527,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8977640959170448,
+      "grad_norm": 0.8774265050888062,
+      "learning_rate": 4.1082714082159084e-06,
+      "loss": 0.147,
+      "step": 5541
+    },
+    {
+      "epoch": 0.8979261179520415,
+      "grad_norm": 0.8635554313659668,
+      "learning_rate": 4.1079365805290214e-06,
+      "loss": 0.1484,
+      "step": 5542
+    },
+    {
+      "epoch": 0.8980881399870383,
+      "grad_norm": 0.9252685904502869,
+      "learning_rate": 4.107601703643167e-06,
+      "loss": 0.1404,
+      "step": 5543
+    },
+    {
+      "epoch": 0.898250162022035,
+      "grad_norm": 0.9969372749328613,
+      "learning_rate": 4.10726677756859e-06,
+      "loss": 0.1578,
+      "step": 5544
+    },
+    {
+      "epoch": 0.8984121840570317,
+      "grad_norm": 0.9096860289573669,
+      "learning_rate": 4.1069318023155405e-06,
+      "loss": 0.1433,
+      "step": 5545
+    },
+    {
+      "epoch": 0.8985742060920285,
+      "grad_norm": 0.8402926325798035,
+      "learning_rate": 4.106596777894265e-06,
+      "loss": 0.1256,
+      "step": 5546
+    },
+    {
+      "epoch": 0.8987362281270252,
+      "grad_norm": 0.8750645518302917,
+      "learning_rate": 4.106261704315017e-06,
+      "loss": 0.1416,
+      "step": 5547
+    },
+    {
+      "epoch": 0.8988982501620221,
+      "grad_norm": 0.7757678627967834,
+      "learning_rate": 4.105926581588046e-06,
+      "loss": 0.1162,
+      "step": 5548
+    },
+    {
+      "epoch": 0.8990602721970188,
+      "grad_norm": 1.0043946504592896,
+      "learning_rate": 4.10559140972361e-06,
+      "loss": 0.1457,
+      "step": 5549
+    },
+    {
+      "epoch": 0.8992222942320156,
+      "grad_norm": 0.8587821125984192,
+      "learning_rate": 4.105256188731962e-06,
+      "loss": 0.1499,
+      "step": 5550
+    },
+    {
+      "epoch": 0.8993843162670123,
+      "grad_norm": 0.8745139241218567,
+      "learning_rate": 4.104920918623359e-06,
+      "loss": 0.1486,
+      "step": 5551
+    },
+    {
+      "epoch": 0.899546338302009,
+      "grad_norm": 0.8056247234344482,
+      "learning_rate": 4.104585599408059e-06,
+      "loss": 0.1264,
+      "step": 5552
+    },
+    {
+      "epoch": 0.8997083603370059,
+      "grad_norm": 0.9480019211769104,
+      "learning_rate": 4.104250231096324e-06,
+      "loss": 0.1563,
+      "step": 5553
+    },
+    {
+      "epoch": 0.8998703823720026,
+      "grad_norm": 0.8790834546089172,
+      "learning_rate": 4.1039148136984134e-06,
+      "loss": 0.1446,
+      "step": 5554
+    },
+    {
+      "epoch": 0.9000324044069994,
+      "grad_norm": 0.8392221927642822,
+      "learning_rate": 4.1035793472245905e-06,
+      "loss": 0.1363,
+      "step": 5555
+    },
+    {
+      "epoch": 0.9001944264419961,
+      "grad_norm": 0.9883176684379578,
+      "learning_rate": 4.103243831685121e-06,
+      "loss": 0.1532,
+      "step": 5556
+    },
+    {
+      "epoch": 0.9003564484769929,
+      "grad_norm": 0.9498316049575806,
+      "learning_rate": 4.102908267090269e-06,
+      "loss": 0.1664,
+      "step": 5557
+    },
+    {
+      "epoch": 0.9005184705119896,
+      "grad_norm": 1.109565258026123,
+      "learning_rate": 4.102572653450304e-06,
+      "loss": 0.1679,
+      "step": 5558
+    },
+    {
+      "epoch": 0.9006804925469863,
+      "grad_norm": 1.8112865686416626,
+      "learning_rate": 4.102236990775493e-06,
+      "loss": 0.1594,
+      "step": 5559
+    },
+    {
+      "epoch": 0.9008425145819832,
+      "grad_norm": 0.8903408050537109,
+      "learning_rate": 4.101901279076108e-06,
+      "loss": 0.1408,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9010045366169799,
+      "grad_norm": 0.8736206293106079,
+      "learning_rate": 4.101565518362421e-06,
+      "loss": 0.1407,
+      "step": 5561
+    },
+    {
+      "epoch": 0.9011665586519767,
+      "grad_norm": 0.8691003322601318,
+      "learning_rate": 4.101229708644704e-06,
+      "loss": 0.1545,
+      "step": 5562
+    },
+    {
+      "epoch": 0.9013285806869734,
+      "grad_norm": 0.951155960559845,
+      "learning_rate": 4.100893849933234e-06,
+      "loss": 0.1373,
+      "step": 5563
+    },
+    {
+      "epoch": 0.9014906027219702,
+      "grad_norm": 0.774764358997345,
+      "learning_rate": 4.100557942238284e-06,
+      "loss": 0.1234,
+      "step": 5564
+    },
+    {
+      "epoch": 0.901652624756967,
+      "grad_norm": 0.8135146498680115,
+      "learning_rate": 4.100221985570137e-06,
+      "loss": 0.1403,
+      "step": 5565
+    },
+    {
+      "epoch": 0.9018146467919637,
+      "grad_norm": 0.8030052781105042,
+      "learning_rate": 4.099885979939068e-06,
+      "loss": 0.1311,
+      "step": 5566
+    },
+    {
+      "epoch": 0.9019766688269605,
+      "grad_norm": 0.7571940422058105,
+      "learning_rate": 4.099549925355359e-06,
+      "loss": 0.1109,
+      "step": 5567
+    },
+    {
+      "epoch": 0.9021386908619572,
+      "grad_norm": 0.7888020277023315,
+      "learning_rate": 4.099213821829295e-06,
+      "loss": 0.1194,
+      "step": 5568
+    },
+    {
+      "epoch": 0.902300712896954,
+      "grad_norm": 0.9889339208602905,
+      "learning_rate": 4.098877669371156e-06,
+      "loss": 0.1492,
+      "step": 5569
+    },
+    {
+      "epoch": 0.9024627349319507,
+      "grad_norm": 0.8497052192687988,
+      "learning_rate": 4.098541467991231e-06,
+      "loss": 0.1194,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9026247569669476,
+      "grad_norm": 0.8576518297195435,
+      "learning_rate": 4.098205217699806e-06,
+      "loss": 0.1441,
+      "step": 5571
+    },
+    {
+      "epoch": 0.9027867790019443,
+      "grad_norm": 0.8945521712303162,
+      "learning_rate": 4.097868918507168e-06,
+      "loss": 0.1584,
+      "step": 5572
+    },
+    {
+      "epoch": 0.902948801036941,
+      "grad_norm": 0.9076288342475891,
+      "learning_rate": 4.097532570423608e-06,
+      "loss": 0.1563,
+      "step": 5573
+    },
+    {
+      "epoch": 0.9031108230719378,
+      "grad_norm": 0.8376015424728394,
+      "learning_rate": 4.097196173459417e-06,
+      "loss": 0.1199,
+      "step": 5574
+    },
+    {
+      "epoch": 0.9032728451069345,
+      "grad_norm": 0.9018917679786682,
+      "learning_rate": 4.096859727624889e-06,
+      "loss": 0.1483,
+      "step": 5575
+    },
+    {
+      "epoch": 0.9034348671419313,
+      "grad_norm": 0.8175559639930725,
+      "learning_rate": 4.0965232329303175e-06,
+      "loss": 0.1333,
+      "step": 5576
+    },
+    {
+      "epoch": 0.903596889176928,
+      "grad_norm": 0.8534868955612183,
+      "learning_rate": 4.096186689385997e-06,
+      "loss": 0.136,
+      "step": 5577
+    },
+    {
+      "epoch": 0.9037589112119249,
+      "grad_norm": 0.8894876837730408,
+      "learning_rate": 4.095850097002228e-06,
+      "loss": 0.146,
+      "step": 5578
+    },
+    {
+      "epoch": 0.9039209332469216,
+      "grad_norm": 0.852396547794342,
+      "learning_rate": 4.095513455789307e-06,
+      "loss": 0.1268,
+      "step": 5579
+    },
+    {
+      "epoch": 0.9040829552819183,
+      "grad_norm": 0.9027949571609497,
+      "learning_rate": 4.095176765757537e-06,
+      "loss": 0.1505,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9042449773169151,
+      "grad_norm": 0.9837702512741089,
+      "learning_rate": 4.094840026917217e-06,
+      "loss": 0.154,
+      "step": 5581
+    },
+    {
+      "epoch": 0.9044069993519118,
+      "grad_norm": 0.7658412456512451,
+      "learning_rate": 4.094503239278652e-06,
+      "loss": 0.1292,
+      "step": 5582
+    },
+    {
+      "epoch": 0.9045690213869086,
+      "grad_norm": 0.8777854442596436,
+      "learning_rate": 4.094166402852146e-06,
+      "loss": 0.1327,
+      "step": 5583
+    },
+    {
+      "epoch": 0.9047310434219054,
+      "grad_norm": 0.9733120799064636,
+      "learning_rate": 4.0938295176480055e-06,
+      "loss": 0.1326,
+      "step": 5584
+    },
+    {
+      "epoch": 0.9048930654569022,
+      "grad_norm": 0.9805978536605835,
+      "learning_rate": 4.09349258367654e-06,
+      "loss": 0.1428,
+      "step": 5585
+    },
+    {
+      "epoch": 0.9050550874918989,
+      "grad_norm": 0.8723875284194946,
+      "learning_rate": 4.093155600948057e-06,
+      "loss": 0.1496,
+      "step": 5586
+    },
+    {
+      "epoch": 0.9052171095268956,
+      "grad_norm": 0.8509448766708374,
+      "learning_rate": 4.092818569472869e-06,
+      "loss": 0.1371,
+      "step": 5587
+    },
+    {
+      "epoch": 0.9053791315618924,
+      "grad_norm": 1.0776581764221191,
+      "learning_rate": 4.092481489261285e-06,
+      "loss": 0.1783,
+      "step": 5588
+    },
+    {
+      "epoch": 0.9055411535968891,
+      "grad_norm": 0.8364329934120178,
+      "learning_rate": 4.0921443603236235e-06,
+      "loss": 0.1295,
+      "step": 5589
+    },
+    {
+      "epoch": 0.905703175631886,
+      "grad_norm": 0.8056396245956421,
+      "learning_rate": 4.0918071826701966e-06,
+      "loss": 0.1196,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9058651976668827,
+      "grad_norm": 0.8377984166145325,
+      "learning_rate": 4.0914699563113214e-06,
+      "loss": 0.1338,
+      "step": 5591
+    },
+    {
+      "epoch": 0.9060272197018795,
+      "grad_norm": 0.9143077731132507,
+      "learning_rate": 4.091132681257317e-06,
+      "loss": 0.1505,
+      "step": 5592
+    },
+    {
+      "epoch": 0.9061892417368762,
+      "grad_norm": 0.874538242816925,
+      "learning_rate": 4.0907953575185035e-06,
+      "loss": 0.1465,
+      "step": 5593
+    },
+    {
+      "epoch": 0.906351263771873,
+      "grad_norm": 0.9015555381774902,
+      "learning_rate": 4.090457985105202e-06,
+      "loss": 0.1463,
+      "step": 5594
+    },
+    {
+      "epoch": 0.9065132858068697,
+      "grad_norm": 0.8938778638839722,
+      "learning_rate": 4.090120564027734e-06,
+      "loss": 0.167,
+      "step": 5595
+    },
+    {
+      "epoch": 0.9066753078418665,
+      "grad_norm": 0.7953081130981445,
+      "learning_rate": 4.089783094296425e-06,
+      "loss": 0.1215,
+      "step": 5596
+    },
+    {
+      "epoch": 0.9068373298768633,
+      "grad_norm": 0.9266294836997986,
+      "learning_rate": 4.0894455759216015e-06,
+      "loss": 0.1584,
+      "step": 5597
+    },
+    {
+      "epoch": 0.90699935191186,
+      "grad_norm": 0.9055389165878296,
+      "learning_rate": 4.089108008913589e-06,
+      "loss": 0.1614,
+      "step": 5598
+    },
+    {
+      "epoch": 0.9071613739468568,
+      "grad_norm": 0.9142413139343262,
+      "learning_rate": 4.088770393282717e-06,
+      "loss": 0.1326,
+      "step": 5599
+    },
+    {
+      "epoch": 0.9073233959818535,
+      "grad_norm": 0.922514796257019,
+      "learning_rate": 4.088432729039316e-06,
+      "loss": 0.1363,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9074854180168503,
+      "grad_norm": 0.8099052906036377,
+      "learning_rate": 4.088095016193717e-06,
+      "loss": 0.1321,
+      "step": 5601
+    },
+    {
+      "epoch": 0.907647440051847,
+      "grad_norm": 0.7604457139968872,
+      "learning_rate": 4.087757254756254e-06,
+      "loss": 0.1283,
+      "step": 5602
+    },
+    {
+      "epoch": 0.9078094620868438,
+      "grad_norm": 0.8297195434570312,
+      "learning_rate": 4.087419444737261e-06,
+      "loss": 0.1483,
+      "step": 5603
+    },
+    {
+      "epoch": 0.9079714841218406,
+      "grad_norm": 0.7264284491539001,
+      "learning_rate": 4.087081586147075e-06,
+      "loss": 0.1193,
+      "step": 5604
+    },
+    {
+      "epoch": 0.9081335061568373,
+      "grad_norm": 0.7358347177505493,
+      "learning_rate": 4.086743678996032e-06,
+      "loss": 0.1254,
+      "step": 5605
+    },
+    {
+      "epoch": 0.9082955281918341,
+      "grad_norm": 0.7477496862411499,
+      "learning_rate": 4.086405723294474e-06,
+      "loss": 0.1247,
+      "step": 5606
+    },
+    {
+      "epoch": 0.9084575502268308,
+      "grad_norm": 0.8757005929946899,
+      "learning_rate": 4.086067719052739e-06,
+      "loss": 0.1475,
+      "step": 5607
+    },
+    {
+      "epoch": 0.9086195722618277,
+      "grad_norm": 0.7548813223838806,
+      "learning_rate": 4.0857296662811696e-06,
+      "loss": 0.1146,
+      "step": 5608
+    },
+    {
+      "epoch": 0.9087815942968244,
+      "grad_norm": 0.9473977088928223,
+      "learning_rate": 4.08539156499011e-06,
+      "loss": 0.134,
+      "step": 5609
+    },
+    {
+      "epoch": 0.9089436163318211,
+      "grad_norm": 0.8978647589683533,
+      "learning_rate": 4.085053415189905e-06,
+      "loss": 0.151,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9091056383668179,
+      "grad_norm": 0.8081299662590027,
+      "learning_rate": 4.084715216890902e-06,
+      "loss": 0.1178,
+      "step": 5611
+    },
+    {
+      "epoch": 0.9092676604018146,
+      "grad_norm": 0.8348207473754883,
+      "learning_rate": 4.084376970103448e-06,
+      "loss": 0.1308,
+      "step": 5612
+    },
+    {
+      "epoch": 0.9094296824368114,
+      "grad_norm": 1.0297025442123413,
+      "learning_rate": 4.0840386748378914e-06,
+      "loss": 0.1464,
+      "step": 5613
+    },
+    {
+      "epoch": 0.9095917044718081,
+      "grad_norm": 0.7381079196929932,
+      "learning_rate": 4.0837003311045865e-06,
+      "loss": 0.1112,
+      "step": 5614
+    },
+    {
+      "epoch": 0.909753726506805,
+      "grad_norm": 0.8947216868400574,
+      "learning_rate": 4.083361938913884e-06,
+      "loss": 0.1424,
+      "step": 5615
+    },
+    {
+      "epoch": 0.9099157485418017,
+      "grad_norm": 0.9418050050735474,
+      "learning_rate": 4.083023498276136e-06,
+      "loss": 0.1445,
+      "step": 5616
+    },
+    {
+      "epoch": 0.9100777705767984,
+      "grad_norm": 0.8582934141159058,
+      "learning_rate": 4.0826850092017015e-06,
+      "loss": 0.1446,
+      "step": 5617
+    },
+    {
+      "epoch": 0.9102397926117952,
+      "grad_norm": 0.8298516273498535,
+      "learning_rate": 4.082346471700935e-06,
+      "loss": 0.1382,
+      "step": 5618
+    },
+    {
+      "epoch": 0.9104018146467919,
+      "grad_norm": 0.8987380862236023,
+      "learning_rate": 4.082007885784196e-06,
+      "loss": 0.1405,
+      "step": 5619
+    },
+    {
+      "epoch": 0.9105638366817888,
+      "grad_norm": 0.7453068494796753,
+      "learning_rate": 4.081669251461844e-06,
+      "loss": 0.1175,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9107258587167855,
+      "grad_norm": 0.9048896431922913,
+      "learning_rate": 4.08133056874424e-06,
+      "loss": 0.1491,
+      "step": 5621
+    },
+    {
+      "epoch": 0.9108878807517823,
+      "grad_norm": 0.9545086026191711,
+      "learning_rate": 4.080991837641748e-06,
+      "loss": 0.1519,
+      "step": 5622
+    },
+    {
+      "epoch": 0.911049902786779,
+      "grad_norm": 0.8169971704483032,
+      "learning_rate": 4.0806530581647305e-06,
+      "loss": 0.1244,
+      "step": 5623
+    },
+    {
+      "epoch": 0.9112119248217757,
+      "grad_norm": 0.7835274338722229,
+      "learning_rate": 4.080314230323556e-06,
+      "loss": 0.1404,
+      "step": 5624
+    },
+    {
+      "epoch": 0.9113739468567725,
+      "grad_norm": 0.9390804171562195,
+      "learning_rate": 4.07997535412859e-06,
+      "loss": 0.1438,
+      "step": 5625
+    },
+    {
+      "epoch": 0.9115359688917692,
+      "grad_norm": 0.7569699883460999,
+      "learning_rate": 4.079636429590201e-06,
+      "loss": 0.1186,
+      "step": 5626
+    },
+    {
+      "epoch": 0.9116979909267661,
+      "grad_norm": 0.9515724778175354,
+      "learning_rate": 4.07929745671876e-06,
+      "loss": 0.1582,
+      "step": 5627
+    },
+    {
+      "epoch": 0.9118600129617628,
+      "grad_norm": 0.892797589302063,
+      "learning_rate": 4.07895843552464e-06,
+      "loss": 0.1337,
+      "step": 5628
+    },
+    {
+      "epoch": 0.9120220349967596,
+      "grad_norm": 0.9236721396446228,
+      "learning_rate": 4.078619366018212e-06,
+      "loss": 0.1684,
+      "step": 5629
+    },
+    {
+      "epoch": 0.9121840570317563,
+      "grad_norm": 0.8024287223815918,
+      "learning_rate": 4.078280248209851e-06,
+      "loss": 0.1236,
+      "step": 5630
+    },
+    {
+      "epoch": 0.912346079066753,
+      "grad_norm": 0.8392508029937744,
+      "learning_rate": 4.077941082109934e-06,
+      "loss": 0.1306,
+      "step": 5631
+    },
+    {
+      "epoch": 0.9125081011017498,
+      "grad_norm": 0.9787663221359253,
+      "learning_rate": 4.077601867728839e-06,
+      "loss": 0.1641,
+      "step": 5632
+    },
+    {
+      "epoch": 0.9126701231367466,
+      "grad_norm": 0.7922909259796143,
+      "learning_rate": 4.077262605076943e-06,
+      "loss": 0.1286,
+      "step": 5633
+    },
+    {
+      "epoch": 0.9128321451717434,
+      "grad_norm": 0.7346786260604858,
+      "learning_rate": 4.07692329416463e-06,
+      "loss": 0.1231,
+      "step": 5634
+    },
+    {
+      "epoch": 0.9129941672067401,
+      "grad_norm": 0.8033357262611389,
+      "learning_rate": 4.07658393500228e-06,
+      "loss": 0.1274,
+      "step": 5635
+    },
+    {
+      "epoch": 0.9131561892417369,
+      "grad_norm": 0.8313119411468506,
+      "learning_rate": 4.0762445276002765e-06,
+      "loss": 0.1404,
+      "step": 5636
+    },
+    {
+      "epoch": 0.9133182112767336,
+      "grad_norm": 0.7644381523132324,
+      "learning_rate": 4.075905071969005e-06,
+      "loss": 0.1241,
+      "step": 5637
+    },
+    {
+      "epoch": 0.9134802333117304,
+      "grad_norm": 0.8910484313964844,
+      "learning_rate": 4.075565568118852e-06,
+      "loss": 0.1439,
+      "step": 5638
+    },
+    {
+      "epoch": 0.9136422553467272,
+      "grad_norm": 0.8554601669311523,
+      "learning_rate": 4.075226016060205e-06,
+      "loss": 0.1388,
+      "step": 5639
+    },
+    {
+      "epoch": 0.9138042773817239,
+      "grad_norm": 0.8526726365089417,
+      "learning_rate": 4.074886415803454e-06,
+      "loss": 0.1508,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9139662994167207,
+      "grad_norm": 0.8573095798492432,
+      "learning_rate": 4.07454676735899e-06,
+      "loss": 0.1487,
+      "step": 5641
+    },
+    {
+      "epoch": 0.9141283214517174,
+      "grad_norm": 0.7731853723526001,
+      "learning_rate": 4.074207070737205e-06,
+      "loss": 0.1233,
+      "step": 5642
+    },
+    {
+      "epoch": 0.9142903434867142,
+      "grad_norm": 0.9754795432090759,
+      "learning_rate": 4.073867325948494e-06,
+      "loss": 0.1559,
+      "step": 5643
+    },
+    {
+      "epoch": 0.9144523655217109,
+      "grad_norm": 0.8023756146430969,
+      "learning_rate": 4.07352753300325e-06,
+      "loss": 0.1269,
+      "step": 5644
+    },
+    {
+      "epoch": 0.9146143875567078,
+      "grad_norm": 0.9316949248313904,
+      "learning_rate": 4.073187691911873e-06,
+      "loss": 0.146,
+      "step": 5645
+    },
+    {
+      "epoch": 0.9147764095917045,
+      "grad_norm": 0.8113452196121216,
+      "learning_rate": 4.072847802684758e-06,
+      "loss": 0.1353,
+      "step": 5646
+    },
+    {
+      "epoch": 0.9149384316267012,
+      "grad_norm": 0.8404650092124939,
+      "learning_rate": 4.072507865332308e-06,
+      "loss": 0.1431,
+      "step": 5647
+    },
+    {
+      "epoch": 0.915100453661698,
+      "grad_norm": 0.9183782935142517,
+      "learning_rate": 4.072167879864922e-06,
+      "loss": 0.1453,
+      "step": 5648
+    },
+    {
+      "epoch": 0.9152624756966947,
+      "grad_norm": 0.8436623811721802,
+      "learning_rate": 4.071827846293004e-06,
+      "loss": 0.1267,
+      "step": 5649
+    },
+    {
+      "epoch": 0.9154244977316915,
+      "grad_norm": 0.9088074564933777,
+      "learning_rate": 4.071487764626957e-06,
+      "loss": 0.1451,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9155865197666883,
+      "grad_norm": 0.8326597213745117,
+      "learning_rate": 4.071147634877187e-06,
+      "loss": 0.1361,
+      "step": 5651
+    },
+    {
+      "epoch": 0.9157485418016851,
+      "grad_norm": 0.850050687789917,
+      "learning_rate": 4.070807457054102e-06,
+      "loss": 0.1234,
+      "step": 5652
+    },
+    {
+      "epoch": 0.9159105638366818,
+      "grad_norm": 0.8121395707130432,
+      "learning_rate": 4.07046723116811e-06,
+      "loss": 0.1254,
+      "step": 5653
+    },
+    {
+      "epoch": 0.9160725858716785,
+      "grad_norm": 0.8910734057426453,
+      "learning_rate": 4.070126957229622e-06,
+      "loss": 0.1436,
+      "step": 5654
+    },
+    {
+      "epoch": 0.9162346079066753,
+      "grad_norm": 0.8616095781326294,
+      "learning_rate": 4.0697866352490475e-06,
+      "loss": 0.1444,
+      "step": 5655
+    },
+    {
+      "epoch": 0.916396629941672,
+      "grad_norm": 0.9634042978286743,
+      "learning_rate": 4.069446265236801e-06,
+      "loss": 0.1737,
+      "step": 5656
+    },
+    {
+      "epoch": 0.9165586519766689,
+      "grad_norm": 0.8356743454933167,
+      "learning_rate": 4.0691058472032975e-06,
+      "loss": 0.1345,
+      "step": 5657
+    },
+    {
+      "epoch": 0.9167206740116656,
+      "grad_norm": 0.8691932559013367,
+      "learning_rate": 4.068765381158951e-06,
+      "loss": 0.1398,
+      "step": 5658
+    },
+    {
+      "epoch": 0.9168826960466624,
+      "grad_norm": 0.9177868366241455,
+      "learning_rate": 4.068424867114181e-06,
+      "loss": 0.1573,
+      "step": 5659
+    },
+    {
+      "epoch": 0.9170447180816591,
+      "grad_norm": 0.802828311920166,
+      "learning_rate": 4.068084305079406e-06,
+      "loss": 0.1188,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9172067401166558,
+      "grad_norm": 0.920111358165741,
+      "learning_rate": 4.067743695065045e-06,
+      "loss": 0.1502,
+      "step": 5661
+    },
+    {
+      "epoch": 0.9173687621516526,
+      "grad_norm": 0.872612714767456,
+      "learning_rate": 4.067403037081522e-06,
+      "loss": 0.1551,
+      "step": 5662
+    },
+    {
+      "epoch": 0.9175307841866494,
+      "grad_norm": 0.7936202883720398,
+      "learning_rate": 4.0670623311392575e-06,
+      "loss": 0.1376,
+      "step": 5663
+    },
+    {
+      "epoch": 0.9176928062216462,
+      "grad_norm": 0.8577109575271606,
+      "learning_rate": 4.06672157724868e-06,
+      "loss": 0.1404,
+      "step": 5664
+    },
+    {
+      "epoch": 0.9178548282566429,
+      "grad_norm": 0.7876363396644592,
+      "learning_rate": 4.066380775420211e-06,
+      "loss": 0.1236,
+      "step": 5665
+    },
+    {
+      "epoch": 0.9180168502916397,
+      "grad_norm": 0.9405772089958191,
+      "learning_rate": 4.066039925664283e-06,
+      "loss": 0.1358,
+      "step": 5666
+    },
+    {
+      "epoch": 0.9181788723266364,
+      "grad_norm": 0.8037310242652893,
+      "learning_rate": 4.065699027991322e-06,
+      "loss": 0.1344,
+      "step": 5667
+    },
+    {
+      "epoch": 0.9183408943616331,
+      "grad_norm": 0.9089890718460083,
+      "learning_rate": 4.06535808241176e-06,
+      "loss": 0.1502,
+      "step": 5668
+    },
+    {
+      "epoch": 0.91850291639663,
+      "grad_norm": 0.7645153999328613,
+      "learning_rate": 4.065017088936028e-06,
+      "loss": 0.1121,
+      "step": 5669
+    },
+    {
+      "epoch": 0.9186649384316267,
+      "grad_norm": 0.8740200400352478,
+      "learning_rate": 4.064676047574561e-06,
+      "loss": 0.1576,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9188269604666235,
+      "grad_norm": 0.8704571723937988,
+      "learning_rate": 4.064334958337794e-06,
+      "loss": 0.1325,
+      "step": 5671
+    },
+    {
+      "epoch": 0.9189889825016202,
+      "grad_norm": 0.9229740500450134,
+      "learning_rate": 4.063993821236162e-06,
+      "loss": 0.1428,
+      "step": 5672
+    },
+    {
+      "epoch": 0.919151004536617,
+      "grad_norm": 1.0061372518539429,
+      "learning_rate": 4.063652636280105e-06,
+      "loss": 0.1629,
+      "step": 5673
+    },
+    {
+      "epoch": 0.9193130265716137,
+      "grad_norm": 0.8159903883934021,
+      "learning_rate": 4.063311403480061e-06,
+      "loss": 0.1316,
+      "step": 5674
+    },
+    {
+      "epoch": 0.9194750486066104,
+      "grad_norm": 0.91038578748703,
+      "learning_rate": 4.06297012284647e-06,
+      "loss": 0.1555,
+      "step": 5675
+    },
+    {
+      "epoch": 0.9196370706416073,
+      "grad_norm": 0.9407716989517212,
+      "learning_rate": 4.0626287943897765e-06,
+      "loss": 0.1476,
+      "step": 5676
+    },
+    {
+      "epoch": 0.919799092676604,
+      "grad_norm": 0.8589774370193481,
+      "learning_rate": 4.062287418120423e-06,
+      "loss": 0.1333,
+      "step": 5677
+    },
+    {
+      "epoch": 0.9199611147116008,
+      "grad_norm": 0.8598899841308594,
+      "learning_rate": 4.061945994048855e-06,
+      "loss": 0.1264,
+      "step": 5678
+    },
+    {
+      "epoch": 0.9201231367465975,
+      "grad_norm": 0.8190400004386902,
+      "learning_rate": 4.06160452218552e-06,
+      "loss": 0.1291,
+      "step": 5679
+    },
+    {
+      "epoch": 0.9202851587815943,
+      "grad_norm": 0.9325578808784485,
+      "learning_rate": 4.061263002540865e-06,
+      "loss": 0.15,
+      "step": 5680
+    },
+    {
+      "epoch": 0.920447180816591,
+      "grad_norm": 0.8312631845474243,
+      "learning_rate": 4.060921435125341e-06,
+      "loss": 0.1357,
+      "step": 5681
+    },
+    {
+      "epoch": 0.9206092028515879,
+      "grad_norm": 0.8420397043228149,
+      "learning_rate": 4.060579819949398e-06,
+      "loss": 0.1447,
+      "step": 5682
+    },
+    {
+      "epoch": 0.9207712248865846,
+      "grad_norm": 0.7866024971008301,
+      "learning_rate": 4.06023815702349e-06,
+      "loss": 0.1397,
+      "step": 5683
+    },
+    {
+      "epoch": 0.9209332469215813,
+      "grad_norm": 0.8895429372787476,
+      "learning_rate": 4.059896446358068e-06,
+      "loss": 0.1463,
+      "step": 5684
+    },
+    {
+      "epoch": 0.9210952689565781,
+      "grad_norm": 0.7228549122810364,
+      "learning_rate": 4.059554687963591e-06,
+      "loss": 0.1146,
+      "step": 5685
+    },
+    {
+      "epoch": 0.9212572909915748,
+      "grad_norm": 0.7941901087760925,
+      "learning_rate": 4.059212881850515e-06,
+      "loss": 0.1317,
+      "step": 5686
+    },
+    {
+      "epoch": 0.9214193130265717,
+      "grad_norm": 0.7205981016159058,
+      "learning_rate": 4.058871028029296e-06,
+      "loss": 0.1143,
+      "step": 5687
+    },
+    {
+      "epoch": 0.9215813350615684,
+      "grad_norm": 0.934319257736206,
+      "learning_rate": 4.0585291265103985e-06,
+      "loss": 0.1565,
+      "step": 5688
+    },
+    {
+      "epoch": 0.9217433570965652,
+      "grad_norm": 0.8551682233810425,
+      "learning_rate": 4.05818717730428e-06,
+      "loss": 0.141,
+      "step": 5689
+    },
+    {
+      "epoch": 0.9219053791315619,
+      "grad_norm": 0.9247624278068542,
+      "learning_rate": 4.057845180421405e-06,
+      "loss": 0.1454,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9220674011665586,
+      "grad_norm": 0.8106127381324768,
+      "learning_rate": 4.057503135872237e-06,
+      "loss": 0.1259,
+      "step": 5691
+    },
+    {
+      "epoch": 0.9222294232015554,
+      "grad_norm": 0.8142223954200745,
+      "learning_rate": 4.057161043667243e-06,
+      "loss": 0.1383,
+      "step": 5692
+    },
+    {
+      "epoch": 0.9223914452365521,
+      "grad_norm": 0.7778276801109314,
+      "learning_rate": 4.056818903816888e-06,
+      "loss": 0.1258,
+      "step": 5693
+    },
+    {
+      "epoch": 0.922553467271549,
+      "grad_norm": 0.9407106041908264,
+      "learning_rate": 4.056476716331643e-06,
+      "loss": 0.1488,
+      "step": 5694
+    },
+    {
+      "epoch": 0.9227154893065457,
+      "grad_norm": 0.8241350054740906,
+      "learning_rate": 4.056134481221977e-06,
+      "loss": 0.1358,
+      "step": 5695
+    },
+    {
+      "epoch": 0.9228775113415425,
+      "grad_norm": 0.9014641046524048,
+      "learning_rate": 4.05579219849836e-06,
+      "loss": 0.1459,
+      "step": 5696
+    },
+    {
+      "epoch": 0.9230395333765392,
+      "grad_norm": 0.7756420969963074,
+      "learning_rate": 4.055449868171269e-06,
+      "loss": 0.1265,
+      "step": 5697
+    },
+    {
+      "epoch": 0.9232015554115359,
+      "grad_norm": 0.9437333345413208,
+      "learning_rate": 4.055107490251175e-06,
+      "loss": 0.1625,
+      "step": 5698
+    },
+    {
+      "epoch": 0.9233635774465327,
+      "grad_norm": 0.9008864164352417,
+      "learning_rate": 4.054765064748554e-06,
+      "loss": 0.1386,
+      "step": 5699
+    },
+    {
+      "epoch": 0.9235255994815295,
+      "grad_norm": 0.8397160768508911,
+      "learning_rate": 4.054422591673887e-06,
+      "loss": 0.1367,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9236876215165263,
+      "grad_norm": 0.8157175183296204,
+      "learning_rate": 4.054080071037649e-06,
+      "loss": 0.1287,
+      "step": 5701
+    },
+    {
+      "epoch": 0.923849643551523,
+      "grad_norm": 0.8612614870071411,
+      "learning_rate": 4.0537375028503225e-06,
+      "loss": 0.1411,
+      "step": 5702
+    },
+    {
+      "epoch": 0.9240116655865198,
+      "grad_norm": 0.933879554271698,
+      "learning_rate": 4.053394887122387e-06,
+      "loss": 0.1231,
+      "step": 5703
+    },
+    {
+      "epoch": 0.9241736876215165,
+      "grad_norm": 0.8254159688949585,
+      "learning_rate": 4.053052223864328e-06,
+      "loss": 0.1382,
+      "step": 5704
+    },
+    {
+      "epoch": 0.9243357096565132,
+      "grad_norm": 0.9842225313186646,
+      "learning_rate": 4.052709513086629e-06,
+      "loss": 0.161,
+      "step": 5705
+    },
+    {
+      "epoch": 0.9244977316915101,
+      "grad_norm": 0.8593588471412659,
+      "learning_rate": 4.052366754799776e-06,
+      "loss": 0.1222,
+      "step": 5706
+    },
+    {
+      "epoch": 0.9246597537265068,
+      "grad_norm": 0.7498965859413147,
+      "learning_rate": 4.052023949014257e-06,
+      "loss": 0.1276,
+      "step": 5707
+    },
+    {
+      "epoch": 0.9248217757615036,
+      "grad_norm": 0.9924911856651306,
+      "learning_rate": 4.051681095740561e-06,
+      "loss": 0.1548,
+      "step": 5708
+    },
+    {
+      "epoch": 0.9249837977965003,
+      "grad_norm": 0.8889843821525574,
+      "learning_rate": 4.051338194989179e-06,
+      "loss": 0.1553,
+      "step": 5709
+    },
+    {
+      "epoch": 0.9251458198314971,
+      "grad_norm": 0.8717356324195862,
+      "learning_rate": 4.050995246770602e-06,
+      "loss": 0.146,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9253078418664938,
+      "grad_norm": 0.8971138596534729,
+      "learning_rate": 4.050652251095324e-06,
+      "loss": 0.1565,
+      "step": 5711
+    },
+    {
+      "epoch": 0.9254698639014906,
+      "grad_norm": 0.8394131660461426,
+      "learning_rate": 4.05030920797384e-06,
+      "loss": 0.1485,
+      "step": 5712
+    },
+    {
+      "epoch": 0.9256318859364874,
+      "grad_norm": 0.8194043636322021,
+      "learning_rate": 4.049966117416645e-06,
+      "loss": 0.1376,
+      "step": 5713
+    },
+    {
+      "epoch": 0.9257939079714841,
+      "grad_norm": 0.9140671491622925,
+      "learning_rate": 4.049622979434239e-06,
+      "loss": 0.1417,
+      "step": 5714
+    },
+    {
+      "epoch": 0.9259559300064809,
+      "grad_norm": 0.7816025018692017,
+      "learning_rate": 4.049279794037118e-06,
+      "loss": 0.1382,
+      "step": 5715
+    },
+    {
+      "epoch": 0.9261179520414776,
+      "grad_norm": 0.9249107241630554,
+      "learning_rate": 4.0489365612357854e-06,
+      "loss": 0.1547,
+      "step": 5716
+    },
+    {
+      "epoch": 0.9262799740764744,
+      "grad_norm": 0.8584994673728943,
+      "learning_rate": 4.048593281040743e-06,
+      "loss": 0.129,
+      "step": 5717
+    },
+    {
+      "epoch": 0.9264419961114712,
+      "grad_norm": 0.8270529508590698,
+      "learning_rate": 4.0482499534624934e-06,
+      "loss": 0.1382,
+      "step": 5718
+    },
+    {
+      "epoch": 0.9266040181464679,
+      "grad_norm": 0.8987339735031128,
+      "learning_rate": 4.047906578511542e-06,
+      "loss": 0.1495,
+      "step": 5719
+    },
+    {
+      "epoch": 0.9267660401814647,
+      "grad_norm": 0.7987256050109863,
+      "learning_rate": 4.047563156198394e-06,
+      "loss": 0.1449,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9269280622164614,
+      "grad_norm": 0.746697723865509,
+      "learning_rate": 4.047219686533559e-06,
+      "loss": 0.1149,
+      "step": 5721
+    },
+    {
+      "epoch": 0.9270900842514582,
+      "grad_norm": 0.860031008720398,
+      "learning_rate": 4.046876169527547e-06,
+      "loss": 0.1473,
+      "step": 5722
+    },
+    {
+      "epoch": 0.9272521062864549,
+      "grad_norm": 0.837323784828186,
+      "learning_rate": 4.046532605190866e-06,
+      "loss": 0.1438,
+      "step": 5723
+    },
+    {
+      "epoch": 0.9274141283214518,
+      "grad_norm": 0.8923266530036926,
+      "learning_rate": 4.04618899353403e-06,
+      "loss": 0.1301,
+      "step": 5724
+    },
+    {
+      "epoch": 0.9275761503564485,
+      "grad_norm": 0.9036858677864075,
+      "learning_rate": 4.045845334567553e-06,
+      "loss": 0.14,
+      "step": 5725
+    },
+    {
+      "epoch": 0.9277381723914452,
+      "grad_norm": 0.8130469918251038,
+      "learning_rate": 4.04550162830195e-06,
+      "loss": 0.1394,
+      "step": 5726
+    },
+    {
+      "epoch": 0.927900194426442,
+      "grad_norm": 0.9334059953689575,
+      "learning_rate": 4.045157874747737e-06,
+      "loss": 0.1509,
+      "step": 5727
+    },
+    {
+      "epoch": 0.9280622164614387,
+      "grad_norm": 0.8468704223632812,
+      "learning_rate": 4.044814073915432e-06,
+      "loss": 0.1319,
+      "step": 5728
+    },
+    {
+      "epoch": 0.9282242384964355,
+      "grad_norm": 0.875339150428772,
+      "learning_rate": 4.0444702258155545e-06,
+      "loss": 0.1525,
+      "step": 5729
+    },
+    {
+      "epoch": 0.9283862605314323,
+      "grad_norm": 1.0649487972259521,
+      "learning_rate": 4.044126330458626e-06,
+      "loss": 0.1664,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9285482825664291,
+      "grad_norm": 0.8550140261650085,
+      "learning_rate": 4.043782387855169e-06,
+      "loss": 0.1398,
+      "step": 5731
+    },
+    {
+      "epoch": 0.9287103046014258,
+      "grad_norm": 0.7701947689056396,
+      "learning_rate": 4.0434383980157055e-06,
+      "loss": 0.1186,
+      "step": 5732
+    },
+    {
+      "epoch": 0.9288723266364226,
+      "grad_norm": 0.7681037783622742,
+      "learning_rate": 4.043094360950763e-06,
+      "loss": 0.1364,
+      "step": 5733
+    },
+    {
+      "epoch": 0.9290343486714193,
+      "grad_norm": 0.8084831833839417,
+      "learning_rate": 4.042750276670867e-06,
+      "loss": 0.1258,
+      "step": 5734
+    },
+    {
+      "epoch": 0.929196370706416,
+      "grad_norm": 0.824355959892273,
+      "learning_rate": 4.042406145186546e-06,
+      "loss": 0.1295,
+      "step": 5735
+    },
+    {
+      "epoch": 0.9293583927414129,
+      "grad_norm": 0.9090822339057922,
+      "learning_rate": 4.04206196650833e-06,
+      "loss": 0.1374,
+      "step": 5736
+    },
+    {
+      "epoch": 0.9295204147764096,
+      "grad_norm": 0.7409492135047913,
+      "learning_rate": 4.041717740646749e-06,
+      "loss": 0.1126,
+      "step": 5737
+    },
+    {
+      "epoch": 0.9296824368114064,
+      "grad_norm": 0.8566955327987671,
+      "learning_rate": 4.041373467612337e-06,
+      "loss": 0.1297,
+      "step": 5738
+    },
+    {
+      "epoch": 0.9298444588464031,
+      "grad_norm": 0.8296953439712524,
+      "learning_rate": 4.0410291474156246e-06,
+      "loss": 0.1296,
+      "step": 5739
+    },
+    {
+      "epoch": 0.9300064808813999,
+      "grad_norm": 0.8668819665908813,
+      "learning_rate": 4.0406847800671515e-06,
+      "loss": 0.1438,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9301685029163966,
+      "grad_norm": 0.8064781427383423,
+      "learning_rate": 4.040340365577452e-06,
+      "loss": 0.1289,
+      "step": 5741
+    },
+    {
+      "epoch": 0.9303305249513933,
+      "grad_norm": 0.9188829064369202,
+      "learning_rate": 4.0399959039570646e-06,
+      "loss": 0.1417,
+      "step": 5742
+    },
+    {
+      "epoch": 0.9304925469863902,
+      "grad_norm": 0.9218767881393433,
+      "learning_rate": 4.039651395216529e-06,
+      "loss": 0.1355,
+      "step": 5743
+    },
+    {
+      "epoch": 0.9306545690213869,
+      "grad_norm": 1.0256109237670898,
+      "learning_rate": 4.039306839366387e-06,
+      "loss": 0.1755,
+      "step": 5744
+    },
+    {
+      "epoch": 0.9308165910563837,
+      "grad_norm": 0.7666441202163696,
+      "learning_rate": 4.038962236417181e-06,
+      "loss": 0.1259,
+      "step": 5745
+    },
+    {
+      "epoch": 0.9309786130913804,
+      "grad_norm": 0.7907612323760986,
+      "learning_rate": 4.038617586379455e-06,
+      "loss": 0.1331,
+      "step": 5746
+    },
+    {
+      "epoch": 0.9311406351263772,
+      "grad_norm": 0.8179818391799927,
+      "learning_rate": 4.0382728892637535e-06,
+      "loss": 0.1369,
+      "step": 5747
+    },
+    {
+      "epoch": 0.931302657161374,
+      "grad_norm": 0.8541617393493652,
+      "learning_rate": 4.0379281450806255e-06,
+      "loss": 0.1495,
+      "step": 5748
+    },
+    {
+      "epoch": 0.9314646791963707,
+      "grad_norm": 0.7930795550346375,
+      "learning_rate": 4.037583353840616e-06,
+      "loss": 0.1259,
+      "step": 5749
+    },
+    {
+      "epoch": 0.9316267012313675,
+      "grad_norm": 0.8774799108505249,
+      "learning_rate": 4.037238515554278e-06,
+      "loss": 0.1422,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9317887232663642,
+      "grad_norm": 0.7685874700546265,
+      "learning_rate": 4.03689363023216e-06,
+      "loss": 0.137,
+      "step": 5751
+    },
+    {
+      "epoch": 0.931950745301361,
+      "grad_norm": 0.8230496644973755,
+      "learning_rate": 4.0365486978848176e-06,
+      "loss": 0.1345,
+      "step": 5752
+    },
+    {
+      "epoch": 0.9321127673363577,
+      "grad_norm": 0.7255528569221497,
+      "learning_rate": 4.0362037185228045e-06,
+      "loss": 0.1223,
+      "step": 5753
+    },
+    {
+      "epoch": 0.9322747893713546,
+      "grad_norm": 0.8288103938102722,
+      "learning_rate": 4.035858692156673e-06,
+      "loss": 0.1412,
+      "step": 5754
+    },
+    {
+      "epoch": 0.9324368114063513,
+      "grad_norm": 0.7872039675712585,
+      "learning_rate": 4.035513618796983e-06,
+      "loss": 0.1254,
+      "step": 5755
+    },
+    {
+      "epoch": 0.932598833441348,
+      "grad_norm": 0.7703442573547363,
+      "learning_rate": 4.035168498454292e-06,
+      "loss": 0.1177,
+      "step": 5756
+    },
+    {
+      "epoch": 0.9327608554763448,
+      "grad_norm": 0.7858262658119202,
+      "learning_rate": 4.034823331139161e-06,
+      "loss": 0.123,
+      "step": 5757
+    },
+    {
+      "epoch": 0.9329228775113415,
+      "grad_norm": 0.9467945694923401,
+      "learning_rate": 4.034478116862149e-06,
+      "loss": 0.1372,
+      "step": 5758
+    },
+    {
+      "epoch": 0.9330848995463383,
+      "grad_norm": 0.7315438985824585,
+      "learning_rate": 4.03413285563382e-06,
+      "loss": 0.1219,
+      "step": 5759
+    },
+    {
+      "epoch": 0.933246921581335,
+      "grad_norm": 0.8702850937843323,
+      "learning_rate": 4.033787547464738e-06,
+      "loss": 0.1397,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9334089436163319,
+      "grad_norm": 0.9242250919342041,
+      "learning_rate": 4.03344219236547e-06,
+      "loss": 0.1486,
+      "step": 5761
+    },
+    {
+      "epoch": 0.9335709656513286,
+      "grad_norm": 0.8591980934143066,
+      "learning_rate": 4.033096790346581e-06,
+      "loss": 0.1383,
+      "step": 5762
+    },
+    {
+      "epoch": 0.9337329876863253,
+      "grad_norm": 0.8047747611999512,
+      "learning_rate": 4.03275134141864e-06,
+      "loss": 0.127,
+      "step": 5763
+    },
+    {
+      "epoch": 0.9338950097213221,
+      "grad_norm": 0.8145176768302917,
+      "learning_rate": 4.032405845592218e-06,
+      "loss": 0.1187,
+      "step": 5764
+    },
+    {
+      "epoch": 0.9340570317563188,
+      "grad_norm": 0.8953231573104858,
+      "learning_rate": 4.0320603028778845e-06,
+      "loss": 0.1398,
+      "step": 5765
+    },
+    {
+      "epoch": 0.9342190537913156,
+      "grad_norm": 0.8610212206840515,
+      "learning_rate": 4.0317147132862135e-06,
+      "loss": 0.1355,
+      "step": 5766
+    },
+    {
+      "epoch": 0.9343810758263124,
+      "grad_norm": 0.7924541234970093,
+      "learning_rate": 4.031369076827779e-06,
+      "loss": 0.1243,
+      "step": 5767
+    },
+    {
+      "epoch": 0.9345430978613092,
+      "grad_norm": 0.8468953371047974,
+      "learning_rate": 4.031023393513157e-06,
+      "loss": 0.1312,
+      "step": 5768
+    },
+    {
+      "epoch": 0.9347051198963059,
+      "grad_norm": 0.8163899779319763,
+      "learning_rate": 4.030677663352924e-06,
+      "loss": 0.1254,
+      "step": 5769
+    },
+    {
+      "epoch": 0.9348671419313026,
+      "grad_norm": 0.8453715443611145,
+      "learning_rate": 4.030331886357659e-06,
+      "loss": 0.1306,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9350291639662994,
+      "grad_norm": 0.8978089094161987,
+      "learning_rate": 4.0299860625379405e-06,
+      "loss": 0.1512,
+      "step": 5771
+    },
+    {
+      "epoch": 0.9351911860012961,
+      "grad_norm": 0.8564348816871643,
+      "learning_rate": 4.029640191904352e-06,
+      "loss": 0.147,
+      "step": 5772
+    },
+    {
+      "epoch": 0.935353208036293,
+      "grad_norm": 0.771431028842926,
+      "learning_rate": 4.029294274467475e-06,
+      "loss": 0.1263,
+      "step": 5773
+    },
+    {
+      "epoch": 0.9355152300712897,
+      "grad_norm": 0.7943999767303467,
+      "learning_rate": 4.028948310237893e-06,
+      "loss": 0.1132,
+      "step": 5774
+    },
+    {
+      "epoch": 0.9356772521062865,
+      "grad_norm": 0.8058924078941345,
+      "learning_rate": 4.028602299226194e-06,
+      "loss": 0.1315,
+      "step": 5775
+    },
+    {
+      "epoch": 0.9358392741412832,
+      "grad_norm": 0.9111430048942566,
+      "learning_rate": 4.0282562414429635e-06,
+      "loss": 0.1582,
+      "step": 5776
+    },
+    {
+      "epoch": 0.93600129617628,
+      "grad_norm": 0.8768683075904846,
+      "learning_rate": 4.02791013689879e-06,
+      "loss": 0.147,
+      "step": 5777
+    },
+    {
+      "epoch": 0.9361633182112767,
+      "grad_norm": 0.7665877342224121,
+      "learning_rate": 4.027563985604264e-06,
+      "loss": 0.1359,
+      "step": 5778
+    },
+    {
+      "epoch": 0.9363253402462735,
+      "grad_norm": 0.9200119972229004,
+      "learning_rate": 4.027217787569977e-06,
+      "loss": 0.1459,
+      "step": 5779
+    },
+    {
+      "epoch": 0.9364873622812703,
+      "grad_norm": 0.9305282831192017,
+      "learning_rate": 4.026871542806521e-06,
+      "loss": 0.1408,
+      "step": 5780
+    },
+    {
+      "epoch": 0.936649384316267,
+      "grad_norm": 0.7743881344795227,
+      "learning_rate": 4.026525251324491e-06,
+      "loss": 0.1419,
+      "step": 5781
+    },
+    {
+      "epoch": 0.9368114063512638,
+      "grad_norm": 0.8831008672714233,
+      "learning_rate": 4.026178913134482e-06,
+      "loss": 0.1248,
+      "step": 5782
+    },
+    {
+      "epoch": 0.9369734283862605,
+      "grad_norm": 0.9593657851219177,
+      "learning_rate": 4.025832528247092e-06,
+      "loss": 0.1668,
+      "step": 5783
+    },
+    {
+      "epoch": 0.9371354504212573,
+      "grad_norm": 0.8962870836257935,
+      "learning_rate": 4.02548609667292e-06,
+      "loss": 0.1516,
+      "step": 5784
+    },
+    {
+      "epoch": 0.937297472456254,
+      "grad_norm": 0.8342962265014648,
+      "learning_rate": 4.025139618422563e-06,
+      "loss": 0.1261,
+      "step": 5785
+    },
+    {
+      "epoch": 0.9374594944912508,
+      "grad_norm": 0.8116139769554138,
+      "learning_rate": 4.024793093506626e-06,
+      "loss": 0.1366,
+      "step": 5786
+    },
+    {
+      "epoch": 0.9376215165262476,
+      "grad_norm": 0.8429117798805237,
+      "learning_rate": 4.024446521935709e-06,
+      "loss": 0.1453,
+      "step": 5787
+    },
+    {
+      "epoch": 0.9377835385612443,
+      "grad_norm": 0.8313448429107666,
+      "learning_rate": 4.024099903720419e-06,
+      "loss": 0.1356,
+      "step": 5788
+    },
+    {
+      "epoch": 0.9379455605962411,
+      "grad_norm": 0.8871910572052002,
+      "learning_rate": 4.023753238871359e-06,
+      "loss": 0.1517,
+      "step": 5789
+    },
+    {
+      "epoch": 0.9381075826312378,
+      "grad_norm": 0.7346284985542297,
+      "learning_rate": 4.023406527399137e-06,
+      "loss": 0.1137,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9382696046662347,
+      "grad_norm": 0.7601230144500732,
+      "learning_rate": 4.023059769314363e-06,
+      "loss": 0.127,
+      "step": 5791
+    },
+    {
+      "epoch": 0.9384316267012314,
+      "grad_norm": 0.7658340930938721,
+      "learning_rate": 4.022712964627645e-06,
+      "loss": 0.1183,
+      "step": 5792
+    },
+    {
+      "epoch": 0.9385936487362281,
+      "grad_norm": 0.7574445605278015,
+      "learning_rate": 4.022366113349596e-06,
+      "loss": 0.1352,
+      "step": 5793
+    },
+    {
+      "epoch": 0.9387556707712249,
+      "grad_norm": 0.9390842914581299,
+      "learning_rate": 4.022019215490827e-06,
+      "loss": 0.1511,
+      "step": 5794
+    },
+    {
+      "epoch": 0.9389176928062216,
+      "grad_norm": 0.8839474320411682,
+      "learning_rate": 4.021672271061955e-06,
+      "loss": 0.1477,
+      "step": 5795
+    },
+    {
+      "epoch": 0.9390797148412184,
+      "grad_norm": 0.7684565186500549,
+      "learning_rate": 4.021325280073592e-06,
+      "loss": 0.1171,
+      "step": 5796
+    },
+    {
+      "epoch": 0.9392417368762151,
+      "grad_norm": 0.9148947596549988,
+      "learning_rate": 4.020978242536357e-06,
+      "loss": 0.1434,
+      "step": 5797
+    },
+    {
+      "epoch": 0.939403758911212,
+      "grad_norm": 0.8702596426010132,
+      "learning_rate": 4.0206311584608705e-06,
+      "loss": 0.1391,
+      "step": 5798
+    },
+    {
+      "epoch": 0.9395657809462087,
+      "grad_norm": 0.8113718628883362,
+      "learning_rate": 4.020284027857748e-06,
+      "loss": 0.1407,
+      "step": 5799
+    },
+    {
+      "epoch": 0.9397278029812054,
+      "grad_norm": 1.005882978439331,
+      "learning_rate": 4.019936850737615e-06,
+      "loss": 0.1302,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9398898250162022,
+      "grad_norm": 0.8756170272827148,
+      "learning_rate": 4.019589627111092e-06,
+      "loss": 0.1336,
+      "step": 5801
+    },
+    {
+      "epoch": 0.9400518470511989,
+      "grad_norm": 0.8532119393348694,
+      "learning_rate": 4.019242356988803e-06,
+      "loss": 0.1408,
+      "step": 5802
+    },
+    {
+      "epoch": 0.9402138690861958,
+      "grad_norm": 1.0048043727874756,
+      "learning_rate": 4.018895040381375e-06,
+      "loss": 0.1587,
+      "step": 5803
+    },
+    {
+      "epoch": 0.9403758911211925,
+      "grad_norm": 0.9122523069381714,
+      "learning_rate": 4.018547677299434e-06,
+      "loss": 0.1292,
+      "step": 5804
+    },
+    {
+      "epoch": 0.9405379131561893,
+      "grad_norm": 0.8948111534118652,
+      "learning_rate": 4.018200267753609e-06,
+      "loss": 0.1612,
+      "step": 5805
+    },
+    {
+      "epoch": 0.940699935191186,
+      "grad_norm": 0.7902446985244751,
+      "learning_rate": 4.01785281175453e-06,
+      "loss": 0.1225,
+      "step": 5806
+    },
+    {
+      "epoch": 0.9408619572261827,
+      "grad_norm": 0.9093374013900757,
+      "learning_rate": 4.017505309312829e-06,
+      "loss": 0.1477,
+      "step": 5807
+    },
+    {
+      "epoch": 0.9410239792611795,
+      "grad_norm": 0.9980859756469727,
+      "learning_rate": 4.017157760439136e-06,
+      "loss": 0.1574,
+      "step": 5808
+    },
+    {
+      "epoch": 0.9411860012961762,
+      "grad_norm": 0.8467405438423157,
+      "learning_rate": 4.0168101651440885e-06,
+      "loss": 0.139,
+      "step": 5809
+    },
+    {
+      "epoch": 0.9413480233311731,
+      "grad_norm": 0.7334152460098267,
+      "learning_rate": 4.01646252343832e-06,
+      "loss": 0.1238,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9415100453661698,
+      "grad_norm": 0.8250093460083008,
+      "learning_rate": 4.016114835332467e-06,
+      "loss": 0.1381,
+      "step": 5811
+    },
+    {
+      "epoch": 0.9416720674011666,
+      "grad_norm": 0.8436107039451599,
+      "learning_rate": 4.015767100837171e-06,
+      "loss": 0.1459,
+      "step": 5812
+    },
+    {
+      "epoch": 0.9418340894361633,
+      "grad_norm": 0.8007429242134094,
+      "learning_rate": 4.015419319963069e-06,
+      "loss": 0.1275,
+      "step": 5813
+    },
+    {
+      "epoch": 0.94199611147116,
+      "grad_norm": 0.7847580909729004,
+      "learning_rate": 4.015071492720802e-06,
+      "loss": 0.1295,
+      "step": 5814
+    },
+    {
+      "epoch": 0.9421581335061568,
+      "grad_norm": 0.8911027312278748,
+      "learning_rate": 4.014723619121015e-06,
+      "loss": 0.1504,
+      "step": 5815
+    },
+    {
+      "epoch": 0.9423201555411536,
+      "grad_norm": 0.811799168586731,
+      "learning_rate": 4.014375699174351e-06,
+      "loss": 0.1243,
+      "step": 5816
+    },
+    {
+      "epoch": 0.9424821775761504,
+      "grad_norm": 0.8510453701019287,
+      "learning_rate": 4.014027732891454e-06,
+      "loss": 0.1331,
+      "step": 5817
+    },
+    {
+      "epoch": 0.9426441996111471,
+      "grad_norm": 0.9293267726898193,
+      "learning_rate": 4.013679720282973e-06,
+      "loss": 0.1531,
+      "step": 5818
+    },
+    {
+      "epoch": 0.9428062216461439,
+      "grad_norm": 0.8839837312698364,
+      "learning_rate": 4.013331661359556e-06,
+      "loss": 0.158,
+      "step": 5819
+    },
+    {
+      "epoch": 0.9429682436811406,
+      "grad_norm": 0.7744541764259338,
+      "learning_rate": 4.012983556131852e-06,
+      "loss": 0.1346,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9431302657161373,
+      "grad_norm": 0.8198082447052002,
+      "learning_rate": 4.012635404610512e-06,
+      "loss": 0.1272,
+      "step": 5821
+    },
+    {
+      "epoch": 0.9432922877511342,
+      "grad_norm": 0.9199150204658508,
+      "learning_rate": 4.01228720680619e-06,
+      "loss": 0.145,
+      "step": 5822
+    },
+    {
+      "epoch": 0.9434543097861309,
+      "grad_norm": 0.9087578058242798,
+      "learning_rate": 4.011938962729538e-06,
+      "loss": 0.1471,
+      "step": 5823
+    },
+    {
+      "epoch": 0.9436163318211277,
+      "grad_norm": 0.7970622181892395,
+      "learning_rate": 4.011590672391213e-06,
+      "loss": 0.1335,
+      "step": 5824
+    },
+    {
+      "epoch": 0.9437783538561244,
+      "grad_norm": 0.9035589694976807,
+      "learning_rate": 4.011242335801872e-06,
+      "loss": 0.1493,
+      "step": 5825
+    },
+    {
+      "epoch": 0.9439403758911212,
+      "grad_norm": 0.825301468372345,
+      "learning_rate": 4.010893952972173e-06,
+      "loss": 0.1256,
+      "step": 5826
+    },
+    {
+      "epoch": 0.9441023979261179,
+      "grad_norm": 0.8092053532600403,
+      "learning_rate": 4.010545523912775e-06,
+      "loss": 0.1258,
+      "step": 5827
+    },
+    {
+      "epoch": 0.9442644199611148,
+      "grad_norm": 0.9639272689819336,
+      "learning_rate": 4.010197048634338e-06,
+      "loss": 0.1487,
+      "step": 5828
+    },
+    {
+      "epoch": 0.9444264419961115,
+      "grad_norm": 1.0004661083221436,
+      "learning_rate": 4.009848527147527e-06,
+      "loss": 0.1527,
+      "step": 5829
+    },
+    {
+      "epoch": 0.9445884640311082,
+      "grad_norm": 0.8995679616928101,
+      "learning_rate": 4.009499959463005e-06,
+      "loss": 0.1308,
+      "step": 5830
+    },
+    {
+      "epoch": 0.944750486066105,
+      "grad_norm": 0.9287495613098145,
+      "learning_rate": 4.009151345591437e-06,
+      "loss": 0.1367,
+      "step": 5831
+    },
+    {
+      "epoch": 0.9449125081011017,
+      "grad_norm": 0.9027858376502991,
+      "learning_rate": 4.00880268554349e-06,
+      "loss": 0.1273,
+      "step": 5832
+    },
+    {
+      "epoch": 0.9450745301360985,
+      "grad_norm": 0.8884682655334473,
+      "learning_rate": 4.008453979329832e-06,
+      "loss": 0.144,
+      "step": 5833
+    },
+    {
+      "epoch": 0.9452365521710953,
+      "grad_norm": 0.932540774345398,
+      "learning_rate": 4.008105226961132e-06,
+      "loss": 0.1585,
+      "step": 5834
+    },
+    {
+      "epoch": 0.9453985742060921,
+      "grad_norm": 0.7685552835464478,
+      "learning_rate": 4.0077564284480625e-06,
+      "loss": 0.1161,
+      "step": 5835
+    },
+    {
+      "epoch": 0.9455605962410888,
+      "grad_norm": 0.830932080745697,
+      "learning_rate": 4.007407583801295e-06,
+      "loss": 0.1369,
+      "step": 5836
+    },
+    {
+      "epoch": 0.9457226182760855,
+      "grad_norm": 0.6946200132369995,
+      "learning_rate": 4.007058693031502e-06,
+      "loss": 0.1087,
+      "step": 5837
+    },
+    {
+      "epoch": 0.9458846403110823,
+      "grad_norm": 0.8081919550895691,
+      "learning_rate": 4.006709756149362e-06,
+      "loss": 0.1276,
+      "step": 5838
+    },
+    {
+      "epoch": 0.946046662346079,
+      "grad_norm": 0.9541810154914856,
+      "learning_rate": 4.0063607731655486e-06,
+      "loss": 0.1513,
+      "step": 5839
+    },
+    {
+      "epoch": 0.9462086843810759,
+      "grad_norm": 0.8741765022277832,
+      "learning_rate": 4.006011744090741e-06,
+      "loss": 0.151,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9463707064160726,
+      "grad_norm": 0.8932287096977234,
+      "learning_rate": 4.005662668935618e-06,
+      "loss": 0.1519,
+      "step": 5841
+    },
+    {
+      "epoch": 0.9465327284510694,
+      "grad_norm": 0.7929561138153076,
+      "learning_rate": 4.005313547710861e-06,
+      "loss": 0.1335,
+      "step": 5842
+    },
+    {
+      "epoch": 0.9466947504860661,
+      "grad_norm": 0.8250575661659241,
+      "learning_rate": 4.004964380427153e-06,
+      "loss": 0.1288,
+      "step": 5843
+    },
+    {
+      "epoch": 0.9468567725210628,
+      "grad_norm": 0.8816642761230469,
+      "learning_rate": 4.004615167095176e-06,
+      "loss": 0.144,
+      "step": 5844
+    },
+    {
+      "epoch": 0.9470187945560596,
+      "grad_norm": 0.7737101316452026,
+      "learning_rate": 4.004265907725616e-06,
+      "loss": 0.1375,
+      "step": 5845
+    },
+    {
+      "epoch": 0.9471808165910564,
+      "grad_norm": 0.851394534111023,
+      "learning_rate": 4.003916602329161e-06,
+      "loss": 0.1353,
+      "step": 5846
+    },
+    {
+      "epoch": 0.9473428386260532,
+      "grad_norm": 0.8754749894142151,
+      "learning_rate": 4.003567250916496e-06,
+      "loss": 0.1465,
+      "step": 5847
+    },
+    {
+      "epoch": 0.9475048606610499,
+      "grad_norm": 0.8369944095611572,
+      "learning_rate": 4.0032178534983115e-06,
+      "loss": 0.1441,
+      "step": 5848
+    },
+    {
+      "epoch": 0.9476668826960467,
+      "grad_norm": 0.9141656160354614,
+      "learning_rate": 4.0028684100852986e-06,
+      "loss": 0.1613,
+      "step": 5849
+    },
+    {
+      "epoch": 0.9478289047310434,
+      "grad_norm": 0.8459381461143494,
+      "learning_rate": 4.00251892068815e-06,
+      "loss": 0.1346,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9479909267660401,
+      "grad_norm": 0.8450573086738586,
+      "learning_rate": 4.002169385317558e-06,
+      "loss": 0.1378,
+      "step": 5851
+    },
+    {
+      "epoch": 0.948152948801037,
+      "grad_norm": 0.8298230767250061,
+      "learning_rate": 4.001819803984218e-06,
+      "loss": 0.1431,
+      "step": 5852
+    },
+    {
+      "epoch": 0.9483149708360337,
+      "grad_norm": 0.7699709534645081,
+      "learning_rate": 4.001470176698826e-06,
+      "loss": 0.132,
+      "step": 5853
+    },
+    {
+      "epoch": 0.9484769928710305,
+      "grad_norm": 0.7468587160110474,
+      "learning_rate": 4.00112050347208e-06,
+      "loss": 0.1125,
+      "step": 5854
+    },
+    {
+      "epoch": 0.9486390149060272,
+      "grad_norm": 0.8313590288162231,
+      "learning_rate": 4.00077078431468e-06,
+      "loss": 0.1382,
+      "step": 5855
+    },
+    {
+      "epoch": 0.948801036941024,
+      "grad_norm": 0.9513286352157593,
+      "learning_rate": 4.000421019237326e-06,
+      "loss": 0.1512,
+      "step": 5856
+    },
+    {
+      "epoch": 0.9489630589760207,
+      "grad_norm": 0.9002991914749146,
+      "learning_rate": 4.000071208250719e-06,
+      "loss": 0.1662,
+      "step": 5857
+    },
+    {
+      "epoch": 0.9491250810110174,
+      "grad_norm": 0.9496574401855469,
+      "learning_rate": 3.999721351365563e-06,
+      "loss": 0.1571,
+      "step": 5858
+    },
+    {
+      "epoch": 0.9492871030460143,
+      "grad_norm": 0.9157005548477173,
+      "learning_rate": 3.999371448592563e-06,
+      "loss": 0.1489,
+      "step": 5859
+    },
+    {
+      "epoch": 0.949449125081011,
+      "grad_norm": 0.8769446015357971,
+      "learning_rate": 3.999021499942425e-06,
+      "loss": 0.1484,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9496111471160078,
+      "grad_norm": 0.8674890398979187,
+      "learning_rate": 3.9986715054258575e-06,
+      "loss": 0.1579,
+      "step": 5861
+    },
+    {
+      "epoch": 0.9497731691510045,
+      "grad_norm": 0.766059935092926,
+      "learning_rate": 3.998321465053568e-06,
+      "loss": 0.13,
+      "step": 5862
+    },
+    {
+      "epoch": 0.9499351911860013,
+      "grad_norm": 0.8092472553253174,
+      "learning_rate": 3.997971378836268e-06,
+      "loss": 0.1296,
+      "step": 5863
+    },
+    {
+      "epoch": 0.950097213220998,
+      "grad_norm": 0.8249281048774719,
+      "learning_rate": 3.9976212467846674e-06,
+      "loss": 0.1396,
+      "step": 5864
+    },
+    {
+      "epoch": 0.9502592352559948,
+      "grad_norm": 0.782484769821167,
+      "learning_rate": 3.997271068909483e-06,
+      "loss": 0.1208,
+      "step": 5865
+    },
+    {
+      "epoch": 0.9504212572909916,
+      "grad_norm": 0.7566808462142944,
+      "learning_rate": 3.996920845221425e-06,
+      "loss": 0.1269,
+      "step": 5866
+    },
+    {
+      "epoch": 0.9505832793259883,
+      "grad_norm": 0.7882890105247498,
+      "learning_rate": 3.9965705757312136e-06,
+      "loss": 0.1227,
+      "step": 5867
+    },
+    {
+      "epoch": 0.9507453013609851,
+      "grad_norm": 0.8613394498825073,
+      "learning_rate": 3.996220260449563e-06,
+      "loss": 0.1334,
+      "step": 5868
+    },
+    {
+      "epoch": 0.9509073233959818,
+      "grad_norm": 0.9516828060150146,
+      "learning_rate": 3.9958698993871935e-06,
+      "loss": 0.1641,
+      "step": 5869
+    },
+    {
+      "epoch": 0.9510693454309787,
+      "grad_norm": 0.8244841694831848,
+      "learning_rate": 3.9955194925548245e-06,
+      "loss": 0.1327,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9512313674659754,
+      "grad_norm": 0.7766361832618713,
+      "learning_rate": 3.995169039963179e-06,
+      "loss": 0.1201,
+      "step": 5871
+    },
+    {
+      "epoch": 0.9513933895009722,
+      "grad_norm": 0.8498024344444275,
+      "learning_rate": 3.994818541622979e-06,
+      "loss": 0.1465,
+      "step": 5872
+    },
+    {
+      "epoch": 0.9515554115359689,
+      "grad_norm": 0.9167964458465576,
+      "learning_rate": 3.994467997544948e-06,
+      "loss": 0.1403,
+      "step": 5873
+    },
+    {
+      "epoch": 0.9517174335709656,
+      "grad_norm": 1.072054386138916,
+      "learning_rate": 3.994117407739814e-06,
+      "loss": 0.1786,
+      "step": 5874
+    },
+    {
+      "epoch": 0.9518794556059624,
+      "grad_norm": 0.8586252331733704,
+      "learning_rate": 3.993766772218303e-06,
+      "loss": 0.1465,
+      "step": 5875
+    },
+    {
+      "epoch": 0.9520414776409591,
+      "grad_norm": 0.8002846837043762,
+      "learning_rate": 3.993416090991143e-06,
+      "loss": 0.1172,
+      "step": 5876
+    },
+    {
+      "epoch": 0.952203499675956,
+      "grad_norm": 0.7997404932975769,
+      "learning_rate": 3.9930653640690655e-06,
+      "loss": 0.125,
+      "step": 5877
+    },
+    {
+      "epoch": 0.9523655217109527,
+      "grad_norm": 0.8767850399017334,
+      "learning_rate": 3.992714591462799e-06,
+      "loss": 0.1404,
+      "step": 5878
+    },
+    {
+      "epoch": 0.9525275437459495,
+      "grad_norm": 0.8756327629089355,
+      "learning_rate": 3.992363773183081e-06,
+      "loss": 0.1434,
+      "step": 5879
+    },
+    {
+      "epoch": 0.9526895657809462,
+      "grad_norm": 0.8966729044914246,
+      "learning_rate": 3.992012909240641e-06,
+      "loss": 0.1502,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9528515878159429,
+      "grad_norm": 0.7201352715492249,
+      "learning_rate": 3.991661999646218e-06,
+      "loss": 0.1114,
+      "step": 5881
+    },
+    {
+      "epoch": 0.9530136098509397,
+      "grad_norm": 0.8088727593421936,
+      "learning_rate": 3.991311044410546e-06,
+      "loss": 0.1477,
+      "step": 5882
+    },
+    {
+      "epoch": 0.9531756318859365,
+      "grad_norm": 0.8761907815933228,
+      "learning_rate": 3.9909600435443665e-06,
+      "loss": 0.1541,
+      "step": 5883
+    },
+    {
+      "epoch": 0.9533376539209333,
+      "grad_norm": 0.7636457681655884,
+      "learning_rate": 3.990608997058416e-06,
+      "loss": 0.1187,
+      "step": 5884
+    },
+    {
+      "epoch": 0.95349967595593,
+      "grad_norm": 0.7406367063522339,
+      "learning_rate": 3.9902579049634385e-06,
+      "loss": 0.1194,
+      "step": 5885
+    },
+    {
+      "epoch": 0.9536616979909268,
+      "grad_norm": 0.8607000112533569,
+      "learning_rate": 3.989906767270175e-06,
+      "loss": 0.1342,
+      "step": 5886
+    },
+    {
+      "epoch": 0.9538237200259235,
+      "grad_norm": 0.8230603933334351,
+      "learning_rate": 3.98955558398937e-06,
+      "loss": 0.1441,
+      "step": 5887
+    },
+    {
+      "epoch": 0.9539857420609202,
+      "grad_norm": 0.8925058841705322,
+      "learning_rate": 3.989204355131769e-06,
+      "loss": 0.1424,
+      "step": 5888
+    },
+    {
+      "epoch": 0.9541477640959171,
+      "grad_norm": 0.9012094140052795,
+      "learning_rate": 3.98885308070812e-06,
+      "loss": 0.1487,
+      "step": 5889
+    },
+    {
+      "epoch": 0.9543097861309138,
+      "grad_norm": 0.818138599395752,
+      "learning_rate": 3.988501760729168e-06,
+      "loss": 0.1225,
+      "step": 5890
+    },
+    {
+      "epoch": 0.9544718081659106,
+      "grad_norm": 0.8324544429779053,
+      "learning_rate": 3.988150395205665e-06,
+      "loss": 0.1278,
+      "step": 5891
+    },
+    {
+      "epoch": 0.9546338302009073,
+      "grad_norm": 0.8996634483337402,
+      "learning_rate": 3.98779898414836e-06,
+      "loss": 0.1454,
+      "step": 5892
+    },
+    {
+      "epoch": 0.9547958522359041,
+      "grad_norm": 0.9417580962181091,
+      "learning_rate": 3.987447527568007e-06,
+      "loss": 0.1461,
+      "step": 5893
+    },
+    {
+      "epoch": 0.9549578742709008,
+      "grad_norm": 0.9308149814605713,
+      "learning_rate": 3.98709602547536e-06,
+      "loss": 0.154,
+      "step": 5894
+    },
+    {
+      "epoch": 0.9551198963058976,
+      "grad_norm": 0.8258563876152039,
+      "learning_rate": 3.986744477881172e-06,
+      "loss": 0.1396,
+      "step": 5895
+    },
+    {
+      "epoch": 0.9552819183408944,
+      "grad_norm": 0.9481770396232605,
+      "learning_rate": 3.986392884796202e-06,
+      "loss": 0.1515,
+      "step": 5896
+    },
+    {
+      "epoch": 0.9554439403758911,
+      "grad_norm": 0.8203932046890259,
+      "learning_rate": 3.986041246231206e-06,
+      "loss": 0.1401,
+      "step": 5897
+    },
+    {
+      "epoch": 0.9556059624108879,
+      "grad_norm": 0.8350275158882141,
+      "learning_rate": 3.9856895621969435e-06,
+      "loss": 0.1318,
+      "step": 5898
+    },
+    {
+      "epoch": 0.9557679844458846,
+      "grad_norm": 0.8285270929336548,
+      "learning_rate": 3.985337832704177e-06,
+      "loss": 0.139,
+      "step": 5899
+    },
+    {
+      "epoch": 0.9559300064808814,
+      "grad_norm": 0.7445301413536072,
+      "learning_rate": 3.984986057763667e-06,
+      "loss": 0.1271,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9560920285158782,
+      "grad_norm": 0.8104132413864136,
+      "learning_rate": 3.984634237386177e-06,
+      "loss": 0.1355,
+      "step": 5901
+    },
+    {
+      "epoch": 0.9562540505508749,
+      "grad_norm": 0.9375022649765015,
+      "learning_rate": 3.984282371582472e-06,
+      "loss": 0.1473,
+      "step": 5902
+    },
+    {
+      "epoch": 0.9564160725858717,
+      "grad_norm": 0.7736932635307312,
+      "learning_rate": 3.983930460363318e-06,
+      "loss": 0.1289,
+      "step": 5903
+    },
+    {
+      "epoch": 0.9565780946208684,
+      "grad_norm": 0.8169133067131042,
+      "learning_rate": 3.983578503739483e-06,
+      "loss": 0.1125,
+      "step": 5904
+    },
+    {
+      "epoch": 0.9567401166558652,
+      "grad_norm": 0.9903810620307922,
+      "learning_rate": 3.983226501721736e-06,
+      "loss": 0.1406,
+      "step": 5905
+    },
+    {
+      "epoch": 0.9569021386908619,
+      "grad_norm": 0.8533335328102112,
+      "learning_rate": 3.982874454320849e-06,
+      "loss": 0.135,
+      "step": 5906
+    },
+    {
+      "epoch": 0.9570641607258588,
+      "grad_norm": 0.8362215161323547,
+      "learning_rate": 3.98252236154759e-06,
+      "loss": 0.1378,
+      "step": 5907
+    },
+    {
+      "epoch": 0.9572261827608555,
+      "grad_norm": 0.8589804768562317,
+      "learning_rate": 3.982170223412735e-06,
+      "loss": 0.1484,
+      "step": 5908
+    },
+    {
+      "epoch": 0.9573882047958522,
+      "grad_norm": 0.8295727372169495,
+      "learning_rate": 3.981818039927058e-06,
+      "loss": 0.147,
+      "step": 5909
+    },
+    {
+      "epoch": 0.957550226830849,
+      "grad_norm": 0.9022241234779358,
+      "learning_rate": 3.981465811101335e-06,
+      "loss": 0.1553,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9577122488658457,
+      "grad_norm": 0.8496821522712708,
+      "learning_rate": 3.981113536946344e-06,
+      "loss": 0.1388,
+      "step": 5911
+    },
+    {
+      "epoch": 0.9578742709008425,
+      "grad_norm": 0.7762805819511414,
+      "learning_rate": 3.9807612174728615e-06,
+      "loss": 0.1212,
+      "step": 5912
+    },
+    {
+      "epoch": 0.9580362929358393,
+      "grad_norm": 0.7569793462753296,
+      "learning_rate": 3.9804088526916706e-06,
+      "loss": 0.1175,
+      "step": 5913
+    },
+    {
+      "epoch": 0.9581983149708361,
+      "grad_norm": 1.0253901481628418,
+      "learning_rate": 3.98005644261355e-06,
+      "loss": 0.1636,
+      "step": 5914
+    },
+    {
+      "epoch": 0.9583603370058328,
+      "grad_norm": 0.7959587574005127,
+      "learning_rate": 3.979703987249285e-06,
+      "loss": 0.1359,
+      "step": 5915
+    },
+    {
+      "epoch": 0.9585223590408296,
+      "grad_norm": 0.8208194971084595,
+      "learning_rate": 3.979351486609659e-06,
+      "loss": 0.1304,
+      "step": 5916
+    },
+    {
+      "epoch": 0.9586843810758263,
+      "grad_norm": 0.8786490559577942,
+      "learning_rate": 3.978998940705456e-06,
+      "loss": 0.144,
+      "step": 5917
+    },
+    {
+      "epoch": 0.958846403110823,
+      "grad_norm": 0.8570854663848877,
+      "learning_rate": 3.978646349547466e-06,
+      "loss": 0.1458,
+      "step": 5918
+    },
+    {
+      "epoch": 0.9590084251458199,
+      "grad_norm": 0.8755070567131042,
+      "learning_rate": 3.978293713146475e-06,
+      "loss": 0.1521,
+      "step": 5919
+    },
+    {
+      "epoch": 0.9591704471808166,
+      "grad_norm": 0.8440230488777161,
+      "learning_rate": 3.977941031513275e-06,
+      "loss": 0.1334,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9593324692158134,
+      "grad_norm": 0.7312674522399902,
+      "learning_rate": 3.977588304658654e-06,
+      "loss": 0.1117,
+      "step": 5921
+    },
+    {
+      "epoch": 0.9594944912508101,
+      "grad_norm": 0.9204146265983582,
+      "learning_rate": 3.977235532593408e-06,
+      "loss": 0.1492,
+      "step": 5922
+    },
+    {
+      "epoch": 0.9596565132858069,
+      "grad_norm": 0.7905224561691284,
+      "learning_rate": 3.9768827153283295e-06,
+      "loss": 0.128,
+      "step": 5923
+    },
+    {
+      "epoch": 0.9598185353208036,
+      "grad_norm": 0.8433038592338562,
+      "learning_rate": 3.976529852874214e-06,
+      "loss": 0.1244,
+      "step": 5924
+    },
+    {
+      "epoch": 0.9599805573558003,
+      "grad_norm": 0.8728172183036804,
+      "learning_rate": 3.976176945241857e-06,
+      "loss": 0.1406,
+      "step": 5925
+    },
+    {
+      "epoch": 0.9601425793907972,
+      "grad_norm": 0.9237974286079407,
+      "learning_rate": 3.975823992442058e-06,
+      "loss": 0.1445,
+      "step": 5926
+    },
+    {
+      "epoch": 0.9603046014257939,
+      "grad_norm": 0.8911517262458801,
+      "learning_rate": 3.9754709944856175e-06,
+      "loss": 0.1443,
+      "step": 5927
+    },
+    {
+      "epoch": 0.9604666234607907,
+      "grad_norm": 0.9263610243797302,
+      "learning_rate": 3.975117951383334e-06,
+      "loss": 0.1428,
+      "step": 5928
+    },
+    {
+      "epoch": 0.9606286454957874,
+      "grad_norm": 0.7457212209701538,
+      "learning_rate": 3.974764863146012e-06,
+      "loss": 0.1228,
+      "step": 5929
+    },
+    {
+      "epoch": 0.9607906675307842,
+      "grad_norm": 0.9272271990776062,
+      "learning_rate": 3.974411729784453e-06,
+      "loss": 0.1505,
+      "step": 5930
+    },
+    {
+      "epoch": 0.960952689565781,
+      "grad_norm": 0.9309628009796143,
+      "learning_rate": 3.974058551309463e-06,
+      "loss": 0.1544,
+      "step": 5931
+    },
+    {
+      "epoch": 0.9611147116007777,
+      "grad_norm": 0.8147203326225281,
+      "learning_rate": 3.973705327731849e-06,
+      "loss": 0.1389,
+      "step": 5932
+    },
+    {
+      "epoch": 0.9612767336357745,
+      "grad_norm": 0.7795599699020386,
+      "learning_rate": 3.9733520590624185e-06,
+      "loss": 0.1237,
+      "step": 5933
+    },
+    {
+      "epoch": 0.9614387556707712,
+      "grad_norm": 0.804670512676239,
+      "learning_rate": 3.97299874531198e-06,
+      "loss": 0.139,
+      "step": 5934
+    },
+    {
+      "epoch": 0.961600777705768,
+      "grad_norm": 0.8195658326148987,
+      "learning_rate": 3.972645386491345e-06,
+      "loss": 0.1495,
+      "step": 5935
+    },
+    {
+      "epoch": 0.9617627997407647,
+      "grad_norm": 0.7763250470161438,
+      "learning_rate": 3.972291982611325e-06,
+      "loss": 0.1301,
+      "step": 5936
+    },
+    {
+      "epoch": 0.9619248217757616,
+      "grad_norm": 1.2038568258285522,
+      "learning_rate": 3.971938533682732e-06,
+      "loss": 0.1845,
+      "step": 5937
+    },
+    {
+      "epoch": 0.9620868438107583,
+      "grad_norm": 0.8428332805633545,
+      "learning_rate": 3.971585039716382e-06,
+      "loss": 0.1433,
+      "step": 5938
+    },
+    {
+      "epoch": 0.962248865845755,
+      "grad_norm": 0.8756902813911438,
+      "learning_rate": 3.971231500723093e-06,
+      "loss": 0.148,
+      "step": 5939
+    },
+    {
+      "epoch": 0.9624108878807518,
+      "grad_norm": 0.8229630589485168,
+      "learning_rate": 3.970877916713678e-06,
+      "loss": 0.1428,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9625729099157485,
+      "grad_norm": 0.8338178992271423,
+      "learning_rate": 3.97052428769896e-06,
+      "loss": 0.1336,
+      "step": 5941
+    },
+    {
+      "epoch": 0.9627349319507453,
+      "grad_norm": 0.8867552280426025,
+      "learning_rate": 3.9701706136897564e-06,
+      "loss": 0.1449,
+      "step": 5942
+    },
+    {
+      "epoch": 0.962896953985742,
+      "grad_norm": 0.8758612275123596,
+      "learning_rate": 3.96981689469689e-06,
+      "loss": 0.1592,
+      "step": 5943
+    },
+    {
+      "epoch": 0.9630589760207389,
+      "grad_norm": 0.8010472059249878,
+      "learning_rate": 3.969463130731183e-06,
+      "loss": 0.1295,
+      "step": 5944
+    },
+    {
+      "epoch": 0.9632209980557356,
+      "grad_norm": 0.9243650436401367,
+      "learning_rate": 3.969109321803461e-06,
+      "loss": 0.1523,
+      "step": 5945
+    },
+    {
+      "epoch": 0.9633830200907323,
+      "grad_norm": 0.8313032984733582,
+      "learning_rate": 3.968755467924549e-06,
+      "loss": 0.1382,
+      "step": 5946
+    },
+    {
+      "epoch": 0.9635450421257291,
+      "grad_norm": 0.8543305397033691,
+      "learning_rate": 3.9684015691052736e-06,
+      "loss": 0.143,
+      "step": 5947
+    },
+    {
+      "epoch": 0.9637070641607258,
+      "grad_norm": 0.8567736148834229,
+      "learning_rate": 3.968047625356463e-06,
+      "loss": 0.1359,
+      "step": 5948
+    },
+    {
+      "epoch": 0.9638690861957226,
+      "grad_norm": 0.9071134328842163,
+      "learning_rate": 3.967693636688948e-06,
+      "loss": 0.1325,
+      "step": 5949
+    },
+    {
+      "epoch": 0.9640311082307194,
+      "grad_norm": 0.7577062249183655,
+      "learning_rate": 3.96733960311356e-06,
+      "loss": 0.1227,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9641931302657162,
+      "grad_norm": 0.9051762223243713,
+      "learning_rate": 3.966985524641132e-06,
+      "loss": 0.1367,
+      "step": 5951
+    },
+    {
+      "epoch": 0.9643551523007129,
+      "grad_norm": 0.8882315754890442,
+      "learning_rate": 3.966631401282495e-06,
+      "loss": 0.1543,
+      "step": 5952
+    },
+    {
+      "epoch": 0.9645171743357096,
+      "grad_norm": 0.8317487835884094,
+      "learning_rate": 3.966277233048487e-06,
+      "loss": 0.1269,
+      "step": 5953
+    },
+    {
+      "epoch": 0.9646791963707064,
+      "grad_norm": 0.8450263738632202,
+      "learning_rate": 3.965923019949944e-06,
+      "loss": 0.1332,
+      "step": 5954
+    },
+    {
+      "epoch": 0.9648412184057031,
+      "grad_norm": 0.9014632105827332,
+      "learning_rate": 3.965568761997704e-06,
+      "loss": 0.1534,
+      "step": 5955
+    },
+    {
+      "epoch": 0.9650032404407,
+      "grad_norm": 0.8284904956817627,
+      "learning_rate": 3.965214459202607e-06,
+      "loss": 0.1298,
+      "step": 5956
+    },
+    {
+      "epoch": 0.9651652624756967,
+      "grad_norm": 0.7886707782745361,
+      "learning_rate": 3.964860111575493e-06,
+      "loss": 0.1353,
+      "step": 5957
+    },
+    {
+      "epoch": 0.9653272845106935,
+      "grad_norm": 0.8882688879966736,
+      "learning_rate": 3.964505719127205e-06,
+      "loss": 0.1421,
+      "step": 5958
+    },
+    {
+      "epoch": 0.9654893065456902,
+      "grad_norm": 0.8411626219749451,
+      "learning_rate": 3.964151281868585e-06,
+      "loss": 0.1326,
+      "step": 5959
+    },
+    {
+      "epoch": 0.9656513285806869,
+      "grad_norm": 0.9165163636207581,
+      "learning_rate": 3.963796799810479e-06,
+      "loss": 0.1466,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9658133506156837,
+      "grad_norm": 0.7748314142227173,
+      "learning_rate": 3.963442272963735e-06,
+      "loss": 0.133,
+      "step": 5961
+    },
+    {
+      "epoch": 0.9659753726506805,
+      "grad_norm": 0.8058779835700989,
+      "learning_rate": 3.9630877013391964e-06,
+      "loss": 0.1459,
+      "step": 5962
+    },
+    {
+      "epoch": 0.9661373946856773,
+      "grad_norm": 0.9120404720306396,
+      "learning_rate": 3.962733084947717e-06,
+      "loss": 0.1459,
+      "step": 5963
+    },
+    {
+      "epoch": 0.966299416720674,
+      "grad_norm": 0.8716152906417847,
+      "learning_rate": 3.962378423800143e-06,
+      "loss": 0.1448,
+      "step": 5964
+    },
+    {
+      "epoch": 0.9664614387556708,
+      "grad_norm": 0.8008790612220764,
+      "learning_rate": 3.962023717907329e-06,
+      "loss": 0.131,
+      "step": 5965
+    },
+    {
+      "epoch": 0.9666234607906675,
+      "grad_norm": 0.7553631067276001,
+      "learning_rate": 3.961668967280128e-06,
+      "loss": 0.1304,
+      "step": 5966
+    },
+    {
+      "epoch": 0.9667854828256643,
+      "grad_norm": 0.9899448752403259,
+      "learning_rate": 3.961314171929392e-06,
+      "loss": 0.1312,
+      "step": 5967
+    },
+    {
+      "epoch": 0.9669475048606611,
+      "grad_norm": 0.7708137631416321,
+      "learning_rate": 3.96095933186598e-06,
+      "loss": 0.1328,
+      "step": 5968
+    },
+    {
+      "epoch": 0.9671095268956578,
+      "grad_norm": 0.7999005317687988,
+      "learning_rate": 3.960604447100747e-06,
+      "loss": 0.1344,
+      "step": 5969
+    },
+    {
+      "epoch": 0.9672715489306546,
+      "grad_norm": 0.7382838129997253,
+      "learning_rate": 3.960249517644553e-06,
+      "loss": 0.1226,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9674335709656513,
+      "grad_norm": 1.0125248432159424,
+      "learning_rate": 3.959894543508258e-06,
+      "loss": 0.1487,
+      "step": 5971
+    },
+    {
+      "epoch": 0.9675955930006481,
+      "grad_norm": 0.871058464050293,
+      "learning_rate": 3.959539524702722e-06,
+      "loss": 0.133,
+      "step": 5972
+    },
+    {
+      "epoch": 0.9677576150356448,
+      "grad_norm": 0.7588186264038086,
+      "learning_rate": 3.9591844612388095e-06,
+      "loss": 0.1224,
+      "step": 5973
+    },
+    {
+      "epoch": 0.9679196370706417,
+      "grad_norm": 0.7937440276145935,
+      "learning_rate": 3.958829353127383e-06,
+      "loss": 0.1285,
+      "step": 5974
+    },
+    {
+      "epoch": 0.9680816591056384,
+      "grad_norm": 0.8277894258499146,
+      "learning_rate": 3.958474200379309e-06,
+      "loss": 0.136,
+      "step": 5975
+    },
+    {
+      "epoch": 0.9682436811406351,
+      "grad_norm": 0.8382477760314941,
+      "learning_rate": 3.958119003005453e-06,
+      "loss": 0.1454,
+      "step": 5976
+    },
+    {
+      "epoch": 0.9684057031756319,
+      "grad_norm": 0.8663250207901001,
+      "learning_rate": 3.9577637610166855e-06,
+      "loss": 0.1466,
+      "step": 5977
+    },
+    {
+      "epoch": 0.9685677252106286,
+      "grad_norm": 0.9586002230644226,
+      "learning_rate": 3.9574084744238735e-06,
+      "loss": 0.1648,
+      "step": 5978
+    },
+    {
+      "epoch": 0.9687297472456254,
+      "grad_norm": 0.7380560636520386,
+      "learning_rate": 3.95705314323789e-06,
+      "loss": 0.1182,
+      "step": 5979
+    },
+    {
+      "epoch": 0.9688917692806222,
+      "grad_norm": 0.8138405084609985,
+      "learning_rate": 3.956697767469606e-06,
+      "loss": 0.1381,
+      "step": 5980
+    },
+    {
+      "epoch": 0.969053791315619,
+      "grad_norm": 0.8596041202545166,
+      "learning_rate": 3.956342347129894e-06,
+      "loss": 0.1617,
+      "step": 5981
+    },
+    {
+      "epoch": 0.9692158133506157,
+      "grad_norm": 0.83723384141922,
+      "learning_rate": 3.955986882229632e-06,
+      "loss": 0.1415,
+      "step": 5982
+    },
+    {
+      "epoch": 0.9693778353856124,
+      "grad_norm": 0.8260191679000854,
+      "learning_rate": 3.955631372779694e-06,
+      "loss": 0.1442,
+      "step": 5983
+    },
+    {
+      "epoch": 0.9695398574206092,
+      "grad_norm": 0.776252806186676,
+      "learning_rate": 3.95527581879096e-06,
+      "loss": 0.134,
+      "step": 5984
+    },
+    {
+      "epoch": 0.9697018794556059,
+      "grad_norm": 0.8889530301094055,
+      "learning_rate": 3.954920220274307e-06,
+      "loss": 0.1459,
+      "step": 5985
+    },
+    {
+      "epoch": 0.9698639014906028,
+      "grad_norm": 0.9564645886421204,
+      "learning_rate": 3.954564577240615e-06,
+      "loss": 0.164,
+      "step": 5986
+    },
+    {
+      "epoch": 0.9700259235255995,
+      "grad_norm": 0.7842409610748291,
+      "learning_rate": 3.954208889700768e-06,
+      "loss": 0.1222,
+      "step": 5987
+    },
+    {
+      "epoch": 0.9701879455605963,
+      "grad_norm": 0.8455632328987122,
+      "learning_rate": 3.9538531576656465e-06,
+      "loss": 0.1347,
+      "step": 5988
+    },
+    {
+      "epoch": 0.970349967595593,
+      "grad_norm": 0.688930869102478,
+      "learning_rate": 3.953497381146139e-06,
+      "loss": 0.1216,
+      "step": 5989
+    },
+    {
+      "epoch": 0.9705119896305897,
+      "grad_norm": 1.0175490379333496,
+      "learning_rate": 3.953141560153128e-06,
+      "loss": 0.1671,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9706740116655865,
+      "grad_norm": 0.9077379703521729,
+      "learning_rate": 3.952785694697502e-06,
+      "loss": 0.1554,
+      "step": 5991
+    },
+    {
+      "epoch": 0.9708360337005832,
+      "grad_norm": 0.8699192404747009,
+      "learning_rate": 3.952429784790148e-06,
+      "loss": 0.1375,
+      "step": 5992
+    },
+    {
+      "epoch": 0.9709980557355801,
+      "grad_norm": 0.8834347724914551,
+      "learning_rate": 3.952073830441959e-06,
+      "loss": 0.1395,
+      "step": 5993
+    },
+    {
+      "epoch": 0.9711600777705768,
+      "grad_norm": 0.8686573505401611,
+      "learning_rate": 3.951717831663825e-06,
+      "loss": 0.1358,
+      "step": 5994
+    },
+    {
+      "epoch": 0.9713220998055736,
+      "grad_norm": 1.110762357711792,
+      "learning_rate": 3.951361788466636e-06,
+      "loss": 0.1464,
+      "step": 5995
+    },
+    {
+      "epoch": 0.9714841218405703,
+      "grad_norm": 0.8450319170951843,
+      "learning_rate": 3.951005700861291e-06,
+      "loss": 0.1428,
+      "step": 5996
+    },
+    {
+      "epoch": 0.971646143875567,
+      "grad_norm": 0.9132416844367981,
+      "learning_rate": 3.950649568858682e-06,
+      "loss": 0.1431,
+      "step": 5997
+    },
+    {
+      "epoch": 0.9718081659105638,
+      "grad_norm": 0.695837140083313,
+      "learning_rate": 3.9502933924697076e-06,
+      "loss": 0.1142,
+      "step": 5998
+    },
+    {
+      "epoch": 0.9719701879455606,
+      "grad_norm": 0.8655622005462646,
+      "learning_rate": 3.949937171705264e-06,
+      "loss": 0.1355,
+      "step": 5999
+    },
+    {
+      "epoch": 0.9721322099805574,
+      "grad_norm": 0.7984117865562439,
+      "learning_rate": 3.949580906576252e-06,
+      "loss": 0.1272,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9722942320155541,
+      "grad_norm": 0.8514125347137451,
+      "learning_rate": 3.949224597093572e-06,
+      "loss": 0.1446,
+      "step": 6001
+    },
+    {
+      "epoch": 0.9724562540505509,
+      "grad_norm": 0.9616315960884094,
+      "learning_rate": 3.948868243268127e-06,
+      "loss": 0.1523,
+      "step": 6002
+    },
+    {
+      "epoch": 0.9726182760855476,
+      "grad_norm": 0.9208241105079651,
+      "learning_rate": 3.948511845110819e-06,
+      "loss": 0.1457,
+      "step": 6003
+    },
+    {
+      "epoch": 0.9727802981205443,
+      "grad_norm": 0.7493133544921875,
+      "learning_rate": 3.948155402632554e-06,
+      "loss": 0.1158,
+      "step": 6004
+    },
+    {
+      "epoch": 0.9729423201555412,
+      "grad_norm": 1.0176682472229004,
+      "learning_rate": 3.947798915844239e-06,
+      "loss": 0.1731,
+      "step": 6005
+    },
+    {
+      "epoch": 0.9731043421905379,
+      "grad_norm": 0.8895021080970764,
+      "learning_rate": 3.94744238475678e-06,
+      "loss": 0.1423,
+      "step": 6006
+    },
+    {
+      "epoch": 0.9732663642255347,
+      "grad_norm": 0.7985418438911438,
+      "learning_rate": 3.947085809381087e-06,
+      "loss": 0.1244,
+      "step": 6007
+    },
+    {
+      "epoch": 0.9734283862605314,
+      "grad_norm": 0.7558802366256714,
+      "learning_rate": 3.94672918972807e-06,
+      "loss": 0.1223,
+      "step": 6008
+    },
+    {
+      "epoch": 0.9735904082955282,
+      "grad_norm": 0.8864288330078125,
+      "learning_rate": 3.946372525808641e-06,
+      "loss": 0.1423,
+      "step": 6009
+    },
+    {
+      "epoch": 0.9737524303305249,
+      "grad_norm": 0.9449102282524109,
+      "learning_rate": 3.946015817633714e-06,
+      "loss": 0.1252,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9739144523655218,
+      "grad_norm": 0.7771594524383545,
+      "learning_rate": 3.9456590652142005e-06,
+      "loss": 0.1204,
+      "step": 6011
+    },
+    {
+      "epoch": 0.9740764744005185,
+      "grad_norm": 0.7628992199897766,
+      "learning_rate": 3.945302268561019e-06,
+      "loss": 0.1282,
+      "step": 6012
+    },
+    {
+      "epoch": 0.9742384964355152,
+      "grad_norm": 0.8803264498710632,
+      "learning_rate": 3.944945427685085e-06,
+      "loss": 0.1308,
+      "step": 6013
+    },
+    {
+      "epoch": 0.974400518470512,
+      "grad_norm": 0.7819192409515381,
+      "learning_rate": 3.944588542597319e-06,
+      "loss": 0.1302,
+      "step": 6014
+    },
+    {
+      "epoch": 0.9745625405055087,
+      "grad_norm": 0.9194301962852478,
+      "learning_rate": 3.944231613308637e-06,
+      "loss": 0.1479,
+      "step": 6015
+    },
+    {
+      "epoch": 0.9747245625405055,
+      "grad_norm": 0.9355913996696472,
+      "learning_rate": 3.943874639829964e-06,
+      "loss": 0.1463,
+      "step": 6016
+    },
+    {
+      "epoch": 0.9748865845755023,
+      "grad_norm": 0.8575900793075562,
+      "learning_rate": 3.9435176221722215e-06,
+      "loss": 0.125,
+      "step": 6017
+    },
+    {
+      "epoch": 0.9750486066104991,
+      "grad_norm": 0.9341332912445068,
+      "learning_rate": 3.943160560346332e-06,
+      "loss": 0.1502,
+      "step": 6018
+    },
+    {
+      "epoch": 0.9752106286454958,
+      "grad_norm": 0.9694084525108337,
+      "learning_rate": 3.942803454363224e-06,
+      "loss": 0.1459,
+      "step": 6019
+    },
+    {
+      "epoch": 0.9753726506804925,
+      "grad_norm": 0.7550426721572876,
+      "learning_rate": 3.942446304233819e-06,
+      "loss": 0.1141,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9755346727154893,
+      "grad_norm": 0.890663743019104,
+      "learning_rate": 3.942089109969049e-06,
+      "loss": 0.144,
+      "step": 6021
+    },
+    {
+      "epoch": 0.975696694750486,
+      "grad_norm": 0.7910590171813965,
+      "learning_rate": 3.941731871579842e-06,
+      "loss": 0.1293,
+      "step": 6022
+    },
+    {
+      "epoch": 0.9758587167854829,
+      "grad_norm": 0.9037308096885681,
+      "learning_rate": 3.941374589077128e-06,
+      "loss": 0.1524,
+      "step": 6023
+    },
+    {
+      "epoch": 0.9760207388204796,
+      "grad_norm": 0.9725434184074402,
+      "learning_rate": 3.94101726247184e-06,
+      "loss": 0.1665,
+      "step": 6024
+    },
+    {
+      "epoch": 0.9761827608554764,
+      "grad_norm": 0.8812149167060852,
+      "learning_rate": 3.940659891774912e-06,
+      "loss": 0.139,
+      "step": 6025
+    },
+    {
+      "epoch": 0.9763447828904731,
+      "grad_norm": 0.9266581535339355,
+      "learning_rate": 3.9403024769972766e-06,
+      "loss": 0.1425,
+      "step": 6026
+    },
+    {
+      "epoch": 0.9765068049254698,
+      "grad_norm": 0.7862972617149353,
+      "learning_rate": 3.939945018149871e-06,
+      "loss": 0.1288,
+      "step": 6027
+    },
+    {
+      "epoch": 0.9766688269604666,
+      "grad_norm": 0.7440429925918579,
+      "learning_rate": 3.939587515243632e-06,
+      "loss": 0.1236,
+      "step": 6028
+    },
+    {
+      "epoch": 0.9768308489954634,
+      "grad_norm": 0.9152294397354126,
+      "learning_rate": 3.9392299682894995e-06,
+      "loss": 0.1558,
+      "step": 6029
+    },
+    {
+      "epoch": 0.9769928710304602,
+      "grad_norm": 0.907541811466217,
+      "learning_rate": 3.938872377298413e-06,
+      "loss": 0.1379,
+      "step": 6030
+    },
+    {
+      "epoch": 0.9771548930654569,
+      "grad_norm": 0.8130697011947632,
+      "learning_rate": 3.938514742281313e-06,
+      "loss": 0.1342,
+      "step": 6031
+    },
+    {
+      "epoch": 0.9773169151004537,
+      "grad_norm": 0.7944709658622742,
+      "learning_rate": 3.938157063249144e-06,
+      "loss": 0.1385,
+      "step": 6032
+    },
+    {
+      "epoch": 0.9774789371354504,
+      "grad_norm": 0.8079535365104675,
+      "learning_rate": 3.937799340212849e-06,
+      "loss": 0.1355,
+      "step": 6033
+    },
+    {
+      "epoch": 0.9776409591704471,
+      "grad_norm": 0.8324079513549805,
+      "learning_rate": 3.937441573183373e-06,
+      "loss": 0.1358,
+      "step": 6034
+    },
+    {
+      "epoch": 0.977802981205444,
+      "grad_norm": 0.868264377117157,
+      "learning_rate": 3.937083762171663e-06,
+      "loss": 0.1393,
+      "step": 6035
+    },
+    {
+      "epoch": 0.9779650032404407,
+      "grad_norm": 0.7842541933059692,
+      "learning_rate": 3.936725907188668e-06,
+      "loss": 0.1252,
+      "step": 6036
+    },
+    {
+      "epoch": 0.9781270252754375,
+      "grad_norm": 0.9624860286712646,
+      "learning_rate": 3.936368008245337e-06,
+      "loss": 0.1542,
+      "step": 6037
+    },
+    {
+      "epoch": 0.9782890473104342,
+      "grad_norm": 0.866369366645813,
+      "learning_rate": 3.936010065352622e-06,
+      "loss": 0.1383,
+      "step": 6038
+    },
+    {
+      "epoch": 0.978451069345431,
+      "grad_norm": 0.8309353590011597,
+      "learning_rate": 3.935652078521473e-06,
+      "loss": 0.1299,
+      "step": 6039
+    },
+    {
+      "epoch": 0.9786130913804277,
+      "grad_norm": 0.8431254029273987,
+      "learning_rate": 3.935294047762844e-06,
+      "loss": 0.1346,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9787751134154244,
+      "grad_norm": 0.8997429609298706,
+      "learning_rate": 3.934935973087691e-06,
+      "loss": 0.1351,
+      "step": 6041
+    },
+    {
+      "epoch": 0.9789371354504213,
+      "grad_norm": 0.9118685722351074,
+      "learning_rate": 3.93457785450697e-06,
+      "loss": 0.1442,
+      "step": 6042
+    },
+    {
+      "epoch": 0.979099157485418,
+      "grad_norm": 0.8979122638702393,
+      "learning_rate": 3.934219692031639e-06,
+      "loss": 0.1526,
+      "step": 6043
+    },
+    {
+      "epoch": 0.9792611795204148,
+      "grad_norm": 0.8222776055335999,
+      "learning_rate": 3.933861485672656e-06,
+      "loss": 0.1327,
+      "step": 6044
+    },
+    {
+      "epoch": 0.9794232015554115,
+      "grad_norm": 0.854524552822113,
+      "learning_rate": 3.9335032354409794e-06,
+      "loss": 0.1402,
+      "step": 6045
+    },
+    {
+      "epoch": 0.9795852235904083,
+      "grad_norm": 0.8611454367637634,
+      "learning_rate": 3.933144941347574e-06,
+      "loss": 0.1437,
+      "step": 6046
+    },
+    {
+      "epoch": 0.979747245625405,
+      "grad_norm": 0.8141959309577942,
+      "learning_rate": 3.9327866034034025e-06,
+      "loss": 0.1492,
+      "step": 6047
+    },
+    {
+      "epoch": 0.9799092676604018,
+      "grad_norm": 0.7863458395004272,
+      "learning_rate": 3.932428221619427e-06,
+      "loss": 0.1275,
+      "step": 6048
+    },
+    {
+      "epoch": 0.9800712896953986,
+      "grad_norm": 0.8340588808059692,
+      "learning_rate": 3.9320697960066155e-06,
+      "loss": 0.1455,
+      "step": 6049
+    },
+    {
+      "epoch": 0.9802333117303953,
+      "grad_norm": 0.9195173978805542,
+      "learning_rate": 3.931711326575933e-06,
+      "loss": 0.1282,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9803953337653921,
+      "grad_norm": 0.8525646328926086,
+      "learning_rate": 3.931352813338348e-06,
+      "loss": 0.1431,
+      "step": 6051
+    },
+    {
+      "epoch": 0.9805573558003888,
+      "grad_norm": 0.8366729617118835,
+      "learning_rate": 3.9309942563048315e-06,
+      "loss": 0.1291,
+      "step": 6052
+    },
+    {
+      "epoch": 0.9807193778353857,
+      "grad_norm": 0.8262105584144592,
+      "learning_rate": 3.930635655486353e-06,
+      "loss": 0.1377,
+      "step": 6053
+    },
+    {
+      "epoch": 0.9808813998703824,
+      "grad_norm": 0.9952908158302307,
+      "learning_rate": 3.930277010893887e-06,
+      "loss": 0.1637,
+      "step": 6054
+    },
+    {
+      "epoch": 0.9810434219053791,
+      "grad_norm": 0.8296589851379395,
+      "learning_rate": 3.929918322538404e-06,
+      "loss": 0.1395,
+      "step": 6055
+    },
+    {
+      "epoch": 0.9812054439403759,
+      "grad_norm": 0.8813459873199463,
+      "learning_rate": 3.929559590430881e-06,
+      "loss": 0.1525,
+      "step": 6056
+    },
+    {
+      "epoch": 0.9813674659753726,
+      "grad_norm": 2.0823469161987305,
+      "learning_rate": 3.9292008145822955e-06,
+      "loss": 0.1402,
+      "step": 6057
+    },
+    {
+      "epoch": 0.9815294880103694,
+      "grad_norm": 0.8811355233192444,
+      "learning_rate": 3.928841995003622e-06,
+      "loss": 0.1431,
+      "step": 6058
+    },
+    {
+      "epoch": 0.9816915100453661,
+      "grad_norm": 0.7101410627365112,
+      "learning_rate": 3.928483131705842e-06,
+      "loss": 0.1126,
+      "step": 6059
+    },
+    {
+      "epoch": 0.981853532080363,
+      "grad_norm": 0.7249727249145508,
+      "learning_rate": 3.928124224699935e-06,
+      "loss": 0.1111,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9820155541153597,
+      "grad_norm": 1.166760802268982,
+      "learning_rate": 3.927765273996882e-06,
+      "loss": 0.1405,
+      "step": 6061
+    },
+    {
+      "epoch": 0.9821775761503565,
+      "grad_norm": 0.7682209610939026,
+      "learning_rate": 3.927406279607668e-06,
+      "loss": 0.1365,
+      "step": 6062
+    },
+    {
+      "epoch": 0.9823395981853532,
+      "grad_norm": 0.916122317314148,
+      "learning_rate": 3.927047241543275e-06,
+      "loss": 0.1388,
+      "step": 6063
+    },
+    {
+      "epoch": 0.9825016202203499,
+      "grad_norm": 0.9284025430679321,
+      "learning_rate": 3.92668815981469e-06,
+      "loss": 0.1475,
+      "step": 6064
+    },
+    {
+      "epoch": 0.9826636422553467,
+      "grad_norm": 0.9036963582038879,
+      "learning_rate": 3.9263290344329e-06,
+      "loss": 0.1482,
+      "step": 6065
+    },
+    {
+      "epoch": 0.9828256642903435,
+      "grad_norm": 0.9499265551567078,
+      "learning_rate": 3.925969865408893e-06,
+      "loss": 0.1494,
+      "step": 6066
+    },
+    {
+      "epoch": 0.9829876863253403,
+      "grad_norm": 0.8025641441345215,
+      "learning_rate": 3.925610652753659e-06,
+      "loss": 0.1319,
+      "step": 6067
+    },
+    {
+      "epoch": 0.983149708360337,
+      "grad_norm": 0.840201199054718,
+      "learning_rate": 3.925251396478189e-06,
+      "loss": 0.1483,
+      "step": 6068
+    },
+    {
+      "epoch": 0.9833117303953338,
+      "grad_norm": 0.9733452796936035,
+      "learning_rate": 3.924892096593476e-06,
+      "loss": 0.1581,
+      "step": 6069
+    },
+    {
+      "epoch": 0.9834737524303305,
+      "grad_norm": 0.7932161688804626,
+      "learning_rate": 3.9245327531105115e-06,
+      "loss": 0.1362,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9836357744653272,
+      "grad_norm": 0.9006250500679016,
+      "learning_rate": 3.924173366040294e-06,
+      "loss": 0.1351,
+      "step": 6071
+    },
+    {
+      "epoch": 0.9837977965003241,
+      "grad_norm": 0.913811206817627,
+      "learning_rate": 3.923813935393816e-06,
+      "loss": 0.1392,
+      "step": 6072
+    },
+    {
+      "epoch": 0.9839598185353208,
+      "grad_norm": 0.7927339673042297,
+      "learning_rate": 3.923454461182078e-06,
+      "loss": 0.1271,
+      "step": 6073
+    },
+    {
+      "epoch": 0.9841218405703176,
+      "grad_norm": 0.8521562218666077,
+      "learning_rate": 3.923094943416078e-06,
+      "loss": 0.1226,
+      "step": 6074
+    },
+    {
+      "epoch": 0.9842838626053143,
+      "grad_norm": 1.042307734489441,
+      "learning_rate": 3.922735382106817e-06,
+      "loss": 0.1802,
+      "step": 6075
+    },
+    {
+      "epoch": 0.9844458846403111,
+      "grad_norm": 0.9505312442779541,
+      "learning_rate": 3.922375777265296e-06,
+      "loss": 0.1345,
+      "step": 6076
+    },
+    {
+      "epoch": 0.9846079066753078,
+      "grad_norm": 0.8530185222625732,
+      "learning_rate": 3.922016128902519e-06,
+      "loss": 0.1367,
+      "step": 6077
+    },
+    {
+      "epoch": 0.9847699287103046,
+      "grad_norm": 0.8222798705101013,
+      "learning_rate": 3.921656437029488e-06,
+      "loss": 0.1481,
+      "step": 6078
+    },
+    {
+      "epoch": 0.9849319507453014,
+      "grad_norm": 1.2158775329589844,
+      "learning_rate": 3.921296701657211e-06,
+      "loss": 0.1225,
+      "step": 6079
+    },
+    {
+      "epoch": 0.9850939727802981,
+      "grad_norm": 2.2171168327331543,
+      "learning_rate": 3.9209369227966945e-06,
+      "loss": 0.1395,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9852559948152949,
+      "grad_norm": 0.7408256530761719,
+      "learning_rate": 3.920577100458948e-06,
+      "loss": 0.1185,
+      "step": 6081
+    },
+    {
+      "epoch": 0.9854180168502916,
+      "grad_norm": 0.8751834034919739,
+      "learning_rate": 3.920217234654978e-06,
+      "loss": 0.1487,
+      "step": 6082
+    },
+    {
+      "epoch": 0.9855800388852884,
+      "grad_norm": 0.9367472529411316,
+      "learning_rate": 3.919857325395799e-06,
+      "loss": 0.1458,
+      "step": 6083
+    },
+    {
+      "epoch": 0.9857420609202852,
+      "grad_norm": 0.9146280288696289,
+      "learning_rate": 3.919497372692421e-06,
+      "loss": 0.1509,
+      "step": 6084
+    },
+    {
+      "epoch": 0.9859040829552819,
+      "grad_norm": 0.8548406958580017,
+      "learning_rate": 3.919137376555859e-06,
+      "loss": 0.1375,
+      "step": 6085
+    },
+    {
+      "epoch": 0.9860661049902787,
+      "grad_norm": 0.9017359614372253,
+      "learning_rate": 3.918777336997127e-06,
+      "loss": 0.1334,
+      "step": 6086
+    },
+    {
+      "epoch": 0.9862281270252754,
+      "grad_norm": 0.6784223318099976,
+      "learning_rate": 3.918417254027243e-06,
+      "loss": 0.1084,
+      "step": 6087
+    },
+    {
+      "epoch": 0.9863901490602722,
+      "grad_norm": 0.7751449346542358,
+      "learning_rate": 3.918057127657222e-06,
+      "loss": 0.1193,
+      "step": 6088
+    },
+    {
+      "epoch": 0.9865521710952689,
+      "grad_norm": 0.9636908769607544,
+      "learning_rate": 3.917696957898085e-06,
+      "loss": 0.1814,
+      "step": 6089
+    },
+    {
+      "epoch": 0.9867141931302658,
+      "grad_norm": 0.8328016996383667,
+      "learning_rate": 3.9173367447608525e-06,
+      "loss": 0.1353,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9868762151652625,
+      "grad_norm": 0.9201585054397583,
+      "learning_rate": 3.9169764882565445e-06,
+      "loss": 0.13,
+      "step": 6091
+    },
+    {
+      "epoch": 0.9870382372002592,
+      "grad_norm": 0.8333777189254761,
+      "learning_rate": 3.916616188396185e-06,
+      "loss": 0.1367,
+      "step": 6092
+    },
+    {
+      "epoch": 0.987200259235256,
+      "grad_norm": 0.8652458190917969,
+      "learning_rate": 3.916255845190799e-06,
+      "loss": 0.1338,
+      "step": 6093
+    },
+    {
+      "epoch": 0.9873622812702527,
+      "grad_norm": 0.9521787762641907,
+      "learning_rate": 3.915895458651411e-06,
+      "loss": 0.1585,
+      "step": 6094
+    },
+    {
+      "epoch": 0.9875243033052495,
+      "grad_norm": 0.70457923412323,
+      "learning_rate": 3.915535028789049e-06,
+      "loss": 0.1024,
+      "step": 6095
+    },
+    {
+      "epoch": 0.9876863253402463,
+      "grad_norm": 0.9282441735267639,
+      "learning_rate": 3.9151745556147404e-06,
+      "loss": 0.1636,
+      "step": 6096
+    },
+    {
+      "epoch": 0.9878483473752431,
+      "grad_norm": 0.8370496034622192,
+      "learning_rate": 3.914814039139515e-06,
+      "loss": 0.13,
+      "step": 6097
+    },
+    {
+      "epoch": 0.9880103694102398,
+      "grad_norm": 0.8045435547828674,
+      "learning_rate": 3.914453479374403e-06,
+      "loss": 0.1413,
+      "step": 6098
+    },
+    {
+      "epoch": 0.9881723914452365,
+      "grad_norm": 0.8645136952400208,
+      "learning_rate": 3.914092876330439e-06,
+      "loss": 0.1422,
+      "step": 6099
+    },
+    {
+      "epoch": 0.9883344134802333,
+      "grad_norm": 0.8007838726043701,
+      "learning_rate": 3.913732230018654e-06,
+      "loss": 0.1364,
+      "step": 6100
+    },
+    {
+      "epoch": 0.98849643551523,
+      "grad_norm": 0.9162132143974304,
+      "learning_rate": 3.913371540450084e-06,
+      "loss": 0.1577,
+      "step": 6101
+    },
+    {
+      "epoch": 0.9886584575502269,
+      "grad_norm": 0.8496082425117493,
+      "learning_rate": 3.913010807635765e-06,
+      "loss": 0.1321,
+      "step": 6102
+    },
+    {
+      "epoch": 0.9888204795852236,
+      "grad_norm": 0.7729654312133789,
+      "learning_rate": 3.912650031586734e-06,
+      "loss": 0.1298,
+      "step": 6103
+    },
+    {
+      "epoch": 0.9889825016202204,
+      "grad_norm": 0.7477056980133057,
+      "learning_rate": 3.9122892123140324e-06,
+      "loss": 0.111,
+      "step": 6104
+    },
+    {
+      "epoch": 0.9891445236552171,
+      "grad_norm": 0.937633752822876,
+      "learning_rate": 3.911928349828697e-06,
+      "loss": 0.1427,
+      "step": 6105
+    },
+    {
+      "epoch": 0.9893065456902139,
+      "grad_norm": 0.8032740354537964,
+      "learning_rate": 3.911567444141771e-06,
+      "loss": 0.1345,
+      "step": 6106
+    },
+    {
+      "epoch": 0.9894685677252106,
+      "grad_norm": 0.8660422563552856,
+      "learning_rate": 3.911206495264299e-06,
+      "loss": 0.1322,
+      "step": 6107
+    },
+    {
+      "epoch": 0.9896305897602073,
+      "grad_norm": 0.987167477607727,
+      "learning_rate": 3.910845503207322e-06,
+      "loss": 0.1606,
+      "step": 6108
+    },
+    {
+      "epoch": 0.9897926117952042,
+      "grad_norm": 0.7517658472061157,
+      "learning_rate": 3.910484467981886e-06,
+      "loss": 0.1172,
+      "step": 6109
+    },
+    {
+      "epoch": 0.9899546338302009,
+      "grad_norm": 0.8762201070785522,
+      "learning_rate": 3.9101233895990396e-06,
+      "loss": 0.145,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9901166558651977,
+      "grad_norm": 0.7897788882255554,
+      "learning_rate": 3.9097622680698296e-06,
+      "loss": 0.1191,
+      "step": 6111
+    },
+    {
+      "epoch": 0.9902786779001944,
+      "grad_norm": 0.8457385301589966,
+      "learning_rate": 3.909401103405307e-06,
+      "loss": 0.1304,
+      "step": 6112
+    },
+    {
+      "epoch": 0.9904406999351912,
+      "grad_norm": 0.8108502626419067,
+      "learning_rate": 3.9090398956165194e-06,
+      "loss": 0.1161,
+      "step": 6113
+    },
+    {
+      "epoch": 0.990602721970188,
+      "grad_norm": 0.9371310472488403,
+      "learning_rate": 3.908678644714522e-06,
+      "loss": 0.1596,
+      "step": 6114
+    },
+    {
+      "epoch": 0.9907647440051847,
+      "grad_norm": 0.8349683284759521,
+      "learning_rate": 3.908317350710366e-06,
+      "loss": 0.1356,
+      "step": 6115
+    },
+    {
+      "epoch": 0.9909267660401815,
+      "grad_norm": 0.8057882785797119,
+      "learning_rate": 3.907956013615108e-06,
+      "loss": 0.1202,
+      "step": 6116
+    },
+    {
+      "epoch": 0.9910887880751782,
+      "grad_norm": 0.8097211718559265,
+      "learning_rate": 3.907594633439803e-06,
+      "loss": 0.1269,
+      "step": 6117
+    },
+    {
+      "epoch": 0.991250810110175,
+      "grad_norm": 0.8821548819541931,
+      "learning_rate": 3.907233210195508e-06,
+      "loss": 0.1455,
+      "step": 6118
+    },
+    {
+      "epoch": 0.9914128321451717,
+      "grad_norm": 0.88327556848526,
+      "learning_rate": 3.906871743893283e-06,
+      "loss": 0.143,
+      "step": 6119
+    },
+    {
+      "epoch": 0.9915748541801686,
+      "grad_norm": 0.7924086451530457,
+      "learning_rate": 3.906510234544186e-06,
+      "loss": 0.1278,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9917368762151653,
+      "grad_norm": 0.8668779134750366,
+      "learning_rate": 3.906148682159281e-06,
+      "loss": 0.1289,
+      "step": 6121
+    },
+    {
+      "epoch": 0.991898898250162,
+      "grad_norm": 0.8895964026451111,
+      "learning_rate": 3.905787086749628e-06,
+      "loss": 0.1447,
+      "step": 6122
+    },
+    {
+      "epoch": 0.9920609202851588,
+      "grad_norm": 0.8018289804458618,
+      "learning_rate": 3.905425448326293e-06,
+      "loss": 0.1378,
+      "step": 6123
+    },
+    {
+      "epoch": 0.9922229423201555,
+      "grad_norm": 0.8257730603218079,
+      "learning_rate": 3.90506376690034e-06,
+      "loss": 0.1348,
+      "step": 6124
+    },
+    {
+      "epoch": 0.9923849643551523,
+      "grad_norm": 0.8249266743659973,
+      "learning_rate": 3.9047020424828355e-06,
+      "loss": 0.1194,
+      "step": 6125
+    },
+    {
+      "epoch": 0.992546986390149,
+      "grad_norm": 0.8011137247085571,
+      "learning_rate": 3.904340275084848e-06,
+      "loss": 0.1286,
+      "step": 6126
+    },
+    {
+      "epoch": 0.9927090084251459,
+      "grad_norm": 0.8544657230377197,
+      "learning_rate": 3.903978464717446e-06,
+      "loss": 0.1367,
+      "step": 6127
+    },
+    {
+      "epoch": 0.9928710304601426,
+      "grad_norm": 0.8813843131065369,
+      "learning_rate": 3.9036166113917015e-06,
+      "loss": 0.1451,
+      "step": 6128
+    },
+    {
+      "epoch": 0.9930330524951393,
+      "grad_norm": 0.7664751410484314,
+      "learning_rate": 3.903254715118686e-06,
+      "loss": 0.1218,
+      "step": 6129
+    },
+    {
+      "epoch": 0.9931950745301361,
+      "grad_norm": 0.7715081572532654,
+      "learning_rate": 3.90289277590947e-06,
+      "loss": 0.1336,
+      "step": 6130
+    },
+    {
+      "epoch": 0.9933570965651328,
+      "grad_norm": 0.8148114085197449,
+      "learning_rate": 3.902530793775132e-06,
+      "loss": 0.1296,
+      "step": 6131
+    },
+    {
+      "epoch": 0.9935191186001296,
+      "grad_norm": 0.8578192591667175,
+      "learning_rate": 3.902168768726745e-06,
+      "loss": 0.1366,
+      "step": 6132
+    },
+    {
+      "epoch": 0.9936811406351264,
+      "grad_norm": 0.7358187437057495,
+      "learning_rate": 3.9018067007753865e-06,
+      "loss": 0.1169,
+      "step": 6133
+    },
+    {
+      "epoch": 0.9938431626701232,
+      "grad_norm": 0.7255628108978271,
+      "learning_rate": 3.9014445899321355e-06,
+      "loss": 0.1157,
+      "step": 6134
+    },
+    {
+      "epoch": 0.9940051847051199,
+      "grad_norm": 0.7971196174621582,
+      "learning_rate": 3.901082436208071e-06,
+      "loss": 0.1348,
+      "step": 6135
+    },
+    {
+      "epoch": 0.9941672067401166,
+      "grad_norm": 0.800977349281311,
+      "learning_rate": 3.900720239614275e-06,
+      "loss": 0.1239,
+      "step": 6136
+    },
+    {
+      "epoch": 0.9943292287751134,
+      "grad_norm": 0.8563368320465088,
+      "learning_rate": 3.90035800016183e-06,
+      "loss": 0.1358,
+      "step": 6137
+    },
+    {
+      "epoch": 0.9944912508101101,
+      "grad_norm": 0.7976403832435608,
+      "learning_rate": 3.899995717861818e-06,
+      "loss": 0.1327,
+      "step": 6138
+    },
+    {
+      "epoch": 0.994653272845107,
+      "grad_norm": 0.7980660200119019,
+      "learning_rate": 3.899633392725325e-06,
+      "loss": 0.1272,
+      "step": 6139
+    },
+    {
+      "epoch": 0.9948152948801037,
+      "grad_norm": 0.7953473329544067,
+      "learning_rate": 3.899271024763438e-06,
+      "loss": 0.135,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9949773169151005,
+      "grad_norm": 0.7407287955284119,
+      "learning_rate": 3.898908613987243e-06,
+      "loss": 0.119,
+      "step": 6141
+    },
+    {
+      "epoch": 0.9951393389500972,
+      "grad_norm": 0.8958144187927246,
+      "learning_rate": 3.89854616040783e-06,
+      "loss": 0.1443,
+      "step": 6142
+    },
+    {
+      "epoch": 0.9953013609850939,
+      "grad_norm": 0.7830373644828796,
+      "learning_rate": 3.898183664036289e-06,
+      "loss": 0.1207,
+      "step": 6143
+    },
+    {
+      "epoch": 0.9954633830200907,
+      "grad_norm": 0.7853714227676392,
+      "learning_rate": 3.897821124883711e-06,
+      "loss": 0.1148,
+      "step": 6144
+    },
+    {
+      "epoch": 0.9956254050550875,
+      "grad_norm": 0.9703736305236816,
+      "learning_rate": 3.89745854296119e-06,
+      "loss": 0.1581,
+      "step": 6145
+    },
+    {
+      "epoch": 0.9957874270900843,
+      "grad_norm": 0.6939820051193237,
+      "learning_rate": 3.897095918279818e-06,
+      "loss": 0.11,
+      "step": 6146
+    },
+    {
+      "epoch": 0.995949449125081,
+      "grad_norm": 1.0396939516067505,
+      "learning_rate": 3.896733250850694e-06,
+      "loss": 0.1744,
+      "step": 6147
+    },
+    {
+      "epoch": 0.9961114711600778,
+      "grad_norm": 0.8816990256309509,
+      "learning_rate": 3.896370540684911e-06,
+      "loss": 0.1486,
+      "step": 6148
+    },
+    {
+      "epoch": 0.9962734931950745,
+      "grad_norm": 0.8170563578605652,
+      "learning_rate": 3.896007787793569e-06,
+      "loss": 0.1283,
+      "step": 6149
+    },
+    {
+      "epoch": 0.9964355152300713,
+      "grad_norm": 0.7984526753425598,
+      "learning_rate": 3.895644992187767e-06,
+      "loss": 0.1279,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9965975372650681,
+      "grad_norm": 0.8505795001983643,
+      "learning_rate": 3.895282153878606e-06,
+      "loss": 0.1496,
+      "step": 6151
+    },
+    {
+      "epoch": 0.9967595593000648,
+      "grad_norm": 0.8687652349472046,
+      "learning_rate": 3.894919272877187e-06,
+      "loss": 0.1546,
+      "step": 6152
+    },
+    {
+      "epoch": 0.9969215813350616,
+      "grad_norm": 0.7245486974716187,
+      "learning_rate": 3.894556349194613e-06,
+      "loss": 0.1203,
+      "step": 6153
+    },
+    {
+      "epoch": 0.9970836033700583,
+      "grad_norm": 0.8453713059425354,
+      "learning_rate": 3.894193382841991e-06,
+      "loss": 0.1318,
+      "step": 6154
+    },
+    {
+      "epoch": 0.9972456254050551,
+      "grad_norm": 0.7163641452789307,
+      "learning_rate": 3.893830373830425e-06,
+      "loss": 0.1158,
+      "step": 6155
+    },
+    {
+      "epoch": 0.9974076474400518,
+      "grad_norm": 0.8621683120727539,
+      "learning_rate": 3.893467322171022e-06,
+      "loss": 0.1385,
+      "step": 6156
+    },
+    {
+      "epoch": 0.9975696694750487,
+      "grad_norm": 0.7639792561531067,
+      "learning_rate": 3.893104227874892e-06,
+      "loss": 0.1332,
+      "step": 6157
+    },
+    {
+      "epoch": 0.9977316915100454,
+      "grad_norm": 0.8099455237388611,
+      "learning_rate": 3.892741090953143e-06,
+      "loss": 0.1299,
+      "step": 6158
+    },
+    {
+      "epoch": 0.9978937135450421,
+      "grad_norm": 1.0411269664764404,
+      "learning_rate": 3.892377911416888e-06,
+      "loss": 0.1516,
+      "step": 6159
+    },
+    {
+      "epoch": 0.9980557355800389,
+      "grad_norm": 0.8103548288345337,
+      "learning_rate": 3.892014689277238e-06,
+      "loss": 0.1205,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9982177576150356,
+      "grad_norm": 0.7883999943733215,
+      "learning_rate": 3.891651424545307e-06,
+      "loss": 0.1273,
+      "step": 6161
+    },
+    {
+      "epoch": 0.9983797796500324,
+      "grad_norm": 0.8199430704116821,
+      "learning_rate": 3.891288117232209e-06,
+      "loss": 0.1318,
+      "step": 6162
+    },
+    {
+      "epoch": 0.9985418016850292,
+      "grad_norm": 0.8376681208610535,
+      "learning_rate": 3.890924767349062e-06,
+      "loss": 0.1385,
+      "step": 6163
+    },
+    {
+      "epoch": 0.998703823720026,
+      "grad_norm": 0.8806131482124329,
+      "learning_rate": 3.890561374906985e-06,
+      "loss": 0.1461,
+      "step": 6164
+    },
+    {
+      "epoch": 0.9988658457550227,
+      "grad_norm": 0.7740499377250671,
+      "learning_rate": 3.8901979399170935e-06,
+      "loss": 0.1256,
+      "step": 6165
+    },
+    {
+      "epoch": 0.9990278677900194,
+      "grad_norm": 0.8058324456214905,
+      "learning_rate": 3.889834462390509e-06,
+      "loss": 0.123,
+      "step": 6166
+    },
+    {
+      "epoch": 0.9991898898250162,
+      "grad_norm": 0.8365480303764343,
+      "learning_rate": 3.889470942338354e-06,
+      "loss": 0.1291,
+      "step": 6167
+    },
+    {
+      "epoch": 0.9993519118600129,
+      "grad_norm": 0.8258197903633118,
+      "learning_rate": 3.889107379771749e-06,
+      "loss": 0.1342,
+      "step": 6168
+    },
+    {
+      "epoch": 0.9995139338950098,
+      "grad_norm": 0.8261362314224243,
+      "learning_rate": 3.888743774701822e-06,
+      "loss": 0.1371,
+      "step": 6169
+    },
+    {
+      "epoch": 0.9996759559300065,
+      "grad_norm": 0.8388069868087769,
+      "learning_rate": 3.888380127139695e-06,
+      "loss": 0.1249,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9998379779650033,
+      "grad_norm": 0.9523016214370728,
+      "learning_rate": 3.888016437096497e-06,
+      "loss": 0.1439,
+      "step": 6171
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9021637439727783,
+      "learning_rate": 3.887652704583354e-06,
+      "loss": 0.1394,
+      "step": 6172
+    },
+    {
+      "epoch": 1.0001620220349967,
+      "grad_norm": 0.7438717484474182,
+      "learning_rate": 3.887288929611396e-06,
+      "loss": 0.1082,
+      "step": 6173
+    },
+    {
+      "epoch": 1.0003240440699934,
+      "grad_norm": 0.7213918566703796,
+      "learning_rate": 3.886925112191754e-06,
+      "loss": 0.1127,
+      "step": 6174
+    },
+    {
+      "epoch": 1.0004860661049904,
+      "grad_norm": 0.6617137789726257,
+      "learning_rate": 3.88656125233556e-06,
+      "loss": 0.0903,
+      "step": 6175
+    },
+    {
+      "epoch": 1.000648088139987,
+      "grad_norm": 0.8376255631446838,
+      "learning_rate": 3.886197350053948e-06,
+      "loss": 0.1015,
+      "step": 6176
+    },
+    {
+      "epoch": 1.0008101101749838,
+      "grad_norm": 0.7548001408576965,
+      "learning_rate": 3.88583340535805e-06,
+      "loss": 0.1017,
+      "step": 6177
+    },
+    {
+      "epoch": 1.0009721322099805,
+      "grad_norm": 0.7314820885658264,
+      "learning_rate": 3.885469418259005e-06,
+      "loss": 0.1011,
+      "step": 6178
+    },
+    {
+      "epoch": 1.0011341542449774,
+      "grad_norm": 0.7919092178344727,
+      "learning_rate": 3.885105388767948e-06,
+      "loss": 0.1008,
+      "step": 6179
+    },
+    {
+      "epoch": 1.0012961762799741,
+      "grad_norm": 0.7691347002983093,
+      "learning_rate": 3.8847413168960175e-06,
+      "loss": 0.1067,
+      "step": 6180
+    },
+    {
+      "epoch": 1.0014581983149708,
+      "grad_norm": 0.7515387535095215,
+      "learning_rate": 3.884377202654354e-06,
+      "loss": 0.0979,
+      "step": 6181
+    },
+    {
+      "epoch": 1.0016202203499676,
+      "grad_norm": 0.819832980632782,
+      "learning_rate": 3.884013046054098e-06,
+      "loss": 0.1017,
+      "step": 6182
+    },
+    {
+      "epoch": 1.0017822423849643,
+      "grad_norm": 0.7462260723114014,
+      "learning_rate": 3.883648847106393e-06,
+      "loss": 0.1009,
+      "step": 6183
+    },
+    {
+      "epoch": 1.0019442644199612,
+      "grad_norm": 0.8221039772033691,
+      "learning_rate": 3.8832846058223814e-06,
+      "loss": 0.1054,
+      "step": 6184
+    },
+    {
+      "epoch": 1.002106286454958,
+      "grad_norm": 0.8322626948356628,
+      "learning_rate": 3.882920322213207e-06,
+      "loss": 0.0982,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0022683084899546,
+      "grad_norm": 0.8119961023330688,
+      "learning_rate": 3.882555996290019e-06,
+      "loss": 0.0944,
+      "step": 6186
+    },
+    {
+      "epoch": 1.0024303305249513,
+      "grad_norm": 1.05998957157135,
+      "learning_rate": 3.882191628063962e-06,
+      "loss": 0.1197,
+      "step": 6187
+    },
+    {
+      "epoch": 1.002592352559948,
+      "grad_norm": 0.9906516075134277,
+      "learning_rate": 3.881827217546187e-06,
+      "loss": 0.1164,
+      "step": 6188
+    },
+    {
+      "epoch": 1.002754374594945,
+      "grad_norm": 0.8134022355079651,
+      "learning_rate": 3.881462764747842e-06,
+      "loss": 0.0937,
+      "step": 6189
+    },
+    {
+      "epoch": 1.0029163966299417,
+      "grad_norm": 1.1540935039520264,
+      "learning_rate": 3.881098269680081e-06,
+      "loss": 0.1009,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0030784186649384,
+      "grad_norm": 0.8985638618469238,
+      "learning_rate": 3.880733732354054e-06,
+      "loss": 0.1034,
+      "step": 6191
+    },
+    {
+      "epoch": 1.0032404406999351,
+      "grad_norm": 0.9259430766105652,
+      "learning_rate": 3.880369152780916e-06,
+      "loss": 0.1073,
+      "step": 6192
+    },
+    {
+      "epoch": 1.003402462734932,
+      "grad_norm": 0.8292042016983032,
+      "learning_rate": 3.880004530971823e-06,
+      "loss": 0.1135,
+      "step": 6193
+    },
+    {
+      "epoch": 1.0035644847699288,
+      "grad_norm": 0.7862805128097534,
+      "learning_rate": 3.879639866937931e-06,
+      "loss": 0.0931,
+      "step": 6194
+    },
+    {
+      "epoch": 1.0037265068049255,
+      "grad_norm": 0.7968874573707581,
+      "learning_rate": 3.879275160690397e-06,
+      "loss": 0.1072,
+      "step": 6195
+    },
+    {
+      "epoch": 1.0038885288399222,
+      "grad_norm": 0.7896031141281128,
+      "learning_rate": 3.8789104122403815e-06,
+      "loss": 0.0998,
+      "step": 6196
+    },
+    {
+      "epoch": 1.004050550874919,
+      "grad_norm": 0.8073359727859497,
+      "learning_rate": 3.878545621599043e-06,
+      "loss": 0.1079,
+      "step": 6197
+    },
+    {
+      "epoch": 1.0042125729099158,
+      "grad_norm": 0.871297299861908,
+      "learning_rate": 3.878180788777546e-06,
+      "loss": 0.116,
+      "step": 6198
+    },
+    {
+      "epoch": 1.0043745949449125,
+      "grad_norm": 0.7919323444366455,
+      "learning_rate": 3.877815913787052e-06,
+      "loss": 0.1082,
+      "step": 6199
+    },
+    {
+      "epoch": 1.0045366169799093,
+      "grad_norm": 0.8027322292327881,
+      "learning_rate": 3.877450996638725e-06,
+      "loss": 0.1072,
+      "step": 6200
+    },
+    {
+      "epoch": 1.004698639014906,
+      "grad_norm": 0.7986040115356445,
+      "learning_rate": 3.87708603734373e-06,
+      "loss": 0.1078,
+      "step": 6201
+    },
+    {
+      "epoch": 1.0048606610499027,
+      "grad_norm": 0.7671363949775696,
+      "learning_rate": 3.876721035913236e-06,
+      "loss": 0.1029,
+      "step": 6202
+    },
+    {
+      "epoch": 1.0050226830848996,
+      "grad_norm": 0.7083774209022522,
+      "learning_rate": 3.87635599235841e-06,
+      "loss": 0.0964,
+      "step": 6203
+    },
+    {
+      "epoch": 1.0051847051198963,
+      "grad_norm": 0.7173096537590027,
+      "learning_rate": 3.87599090669042e-06,
+      "loss": 0.0972,
+      "step": 6204
+    },
+    {
+      "epoch": 1.005346727154893,
+      "grad_norm": 0.8262728452682495,
+      "learning_rate": 3.8756257789204384e-06,
+      "loss": 0.1103,
+      "step": 6205
+    },
+    {
+      "epoch": 1.0055087491898898,
+      "grad_norm": 0.783673107624054,
+      "learning_rate": 3.875260609059638e-06,
+      "loss": 0.0982,
+      "step": 6206
+    },
+    {
+      "epoch": 1.0056707712248867,
+      "grad_norm": 0.7423332929611206,
+      "learning_rate": 3.8748953971191895e-06,
+      "loss": 0.089,
+      "step": 6207
+    },
+    {
+      "epoch": 1.0058327932598834,
+      "grad_norm": 0.7428287267684937,
+      "learning_rate": 3.87453014311027e-06,
+      "loss": 0.0988,
+      "step": 6208
+    },
+    {
+      "epoch": 1.00599481529488,
+      "grad_norm": 0.886810302734375,
+      "learning_rate": 3.874164847044054e-06,
+      "loss": 0.1142,
+      "step": 6209
+    },
+    {
+      "epoch": 1.0061568373298768,
+      "grad_norm": 0.8856390714645386,
+      "learning_rate": 3.87379950893172e-06,
+      "loss": 0.1035,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0063188593648735,
+      "grad_norm": 0.7835367321968079,
+      "learning_rate": 3.873434128784444e-06,
+      "loss": 0.1062,
+      "step": 6211
+    },
+    {
+      "epoch": 1.0064808813998705,
+      "grad_norm": 0.9305019974708557,
+      "learning_rate": 3.8730687066134086e-06,
+      "loss": 0.1045,
+      "step": 6212
+    },
+    {
+      "epoch": 1.0066429034348672,
+      "grad_norm": 1.0636266469955444,
+      "learning_rate": 3.872703242429794e-06,
+      "loss": 0.1275,
+      "step": 6213
+    },
+    {
+      "epoch": 1.0068049254698639,
+      "grad_norm": 0.8035116195678711,
+      "learning_rate": 3.8723377362447805e-06,
+      "loss": 0.1081,
+      "step": 6214
+    },
+    {
+      "epoch": 1.0069669475048606,
+      "grad_norm": 0.7572904825210571,
+      "learning_rate": 3.871972188069554e-06,
+      "loss": 0.0901,
+      "step": 6215
+    },
+    {
+      "epoch": 1.0071289695398573,
+      "grad_norm": 0.8394489288330078,
+      "learning_rate": 3.871606597915298e-06,
+      "loss": 0.0988,
+      "step": 6216
+    },
+    {
+      "epoch": 1.0072909915748542,
+      "grad_norm": 0.9537179470062256,
+      "learning_rate": 3.871240965793201e-06,
+      "loss": 0.1117,
+      "step": 6217
+    },
+    {
+      "epoch": 1.007453013609851,
+      "grad_norm": 0.8602259755134583,
+      "learning_rate": 3.870875291714448e-06,
+      "loss": 0.1031,
+      "step": 6218
+    },
+    {
+      "epoch": 1.0076150356448477,
+      "grad_norm": 0.894371747970581,
+      "learning_rate": 3.870509575690228e-06,
+      "loss": 0.117,
+      "step": 6219
+    },
+    {
+      "epoch": 1.0077770576798444,
+      "grad_norm": 0.8317716121673584,
+      "learning_rate": 3.870143817731732e-06,
+      "loss": 0.104,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0079390797148413,
+      "grad_norm": 0.8154535293579102,
+      "learning_rate": 3.86977801785015e-06,
+      "loss": 0.1053,
+      "step": 6221
+    },
+    {
+      "epoch": 1.008101101749838,
+      "grad_norm": 0.9335169196128845,
+      "learning_rate": 3.8694121760566765e-06,
+      "loss": 0.1125,
+      "step": 6222
+    },
+    {
+      "epoch": 1.0082631237848347,
+      "grad_norm": 0.8863696455955505,
+      "learning_rate": 3.869046292362504e-06,
+      "loss": 0.1095,
+      "step": 6223
+    },
+    {
+      "epoch": 1.0084251458198314,
+      "grad_norm": 0.7653059363365173,
+      "learning_rate": 3.868680366778828e-06,
+      "loss": 0.0922,
+      "step": 6224
+    },
+    {
+      "epoch": 1.0085871678548282,
+      "grad_norm": 0.7611356973648071,
+      "learning_rate": 3.868314399316845e-06,
+      "loss": 0.1027,
+      "step": 6225
+    },
+    {
+      "epoch": 1.008749189889825,
+      "grad_norm": 0.9434992671012878,
+      "learning_rate": 3.867948389987752e-06,
+      "loss": 0.1091,
+      "step": 6226
+    },
+    {
+      "epoch": 1.0089112119248218,
+      "grad_norm": 0.8308627605438232,
+      "learning_rate": 3.86758233880275e-06,
+      "loss": 0.1032,
+      "step": 6227
+    },
+    {
+      "epoch": 1.0090732339598185,
+      "grad_norm": 1.1444343328475952,
+      "learning_rate": 3.8672162457730365e-06,
+      "loss": 0.1183,
+      "step": 6228
+    },
+    {
+      "epoch": 1.0092352559948152,
+      "grad_norm": 0.9171835780143738,
+      "learning_rate": 3.866850110909816e-06,
+      "loss": 0.1025,
+      "step": 6229
+    },
+    {
+      "epoch": 1.0093972780298122,
+      "grad_norm": 0.8272443413734436,
+      "learning_rate": 3.866483934224288e-06,
+      "loss": 0.1077,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0095593000648089,
+      "grad_norm": 0.7675562500953674,
+      "learning_rate": 3.866117715727659e-06,
+      "loss": 0.0994,
+      "step": 6231
+    },
+    {
+      "epoch": 1.0097213220998056,
+      "grad_norm": 0.8303428888320923,
+      "learning_rate": 3.865751455431134e-06,
+      "loss": 0.1019,
+      "step": 6232
+    },
+    {
+      "epoch": 1.0098833441348023,
+      "grad_norm": 0.9183821082115173,
+      "learning_rate": 3.86538515334592e-06,
+      "loss": 0.1054,
+      "step": 6233
+    },
+    {
+      "epoch": 1.010045366169799,
+      "grad_norm": 0.8656824827194214,
+      "learning_rate": 3.865018809483224e-06,
+      "loss": 0.1171,
+      "step": 6234
+    },
+    {
+      "epoch": 1.010207388204796,
+      "grad_norm": 0.8201124668121338,
+      "learning_rate": 3.864652423854256e-06,
+      "loss": 0.1118,
+      "step": 6235
+    },
+    {
+      "epoch": 1.0103694102397927,
+      "grad_norm": 0.8729208111763,
+      "learning_rate": 3.864285996470226e-06,
+      "loss": 0.1078,
+      "step": 6236
+    },
+    {
+      "epoch": 1.0105314322747894,
+      "grad_norm": 0.8456924557685852,
+      "learning_rate": 3.863919527342346e-06,
+      "loss": 0.1091,
+      "step": 6237
+    },
+    {
+      "epoch": 1.010693454309786,
+      "grad_norm": 0.9836832284927368,
+      "learning_rate": 3.863553016481829e-06,
+      "loss": 0.1226,
+      "step": 6238
+    },
+    {
+      "epoch": 1.0108554763447828,
+      "grad_norm": 0.7698848843574524,
+      "learning_rate": 3.863186463899891e-06,
+      "loss": 0.1032,
+      "step": 6239
+    },
+    {
+      "epoch": 1.0110174983797797,
+      "grad_norm": 0.8878111839294434,
+      "learning_rate": 3.862819869607743e-06,
+      "loss": 0.1039,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0111795204147764,
+      "grad_norm": 0.7739588618278503,
+      "learning_rate": 3.862453233616608e-06,
+      "loss": 0.1052,
+      "step": 6241
+    },
+    {
+      "epoch": 1.0113415424497731,
+      "grad_norm": 0.7633801698684692,
+      "learning_rate": 3.862086555937699e-06,
+      "loss": 0.0993,
+      "step": 6242
+    },
+    {
+      "epoch": 1.0115035644847699,
+      "grad_norm": 0.8758237957954407,
+      "learning_rate": 3.861719836582239e-06,
+      "loss": 0.1156,
+      "step": 6243
+    },
+    {
+      "epoch": 1.0116655865197668,
+      "grad_norm": 0.8216294050216675,
+      "learning_rate": 3.861353075561446e-06,
+      "loss": 0.1035,
+      "step": 6244
+    },
+    {
+      "epoch": 1.0118276085547635,
+      "grad_norm": 0.9009713530540466,
+      "learning_rate": 3.860986272886545e-06,
+      "loss": 0.1155,
+      "step": 6245
+    },
+    {
+      "epoch": 1.0119896305897602,
+      "grad_norm": 0.8081684112548828,
+      "learning_rate": 3.860619428568756e-06,
+      "loss": 0.1059,
+      "step": 6246
+    },
+    {
+      "epoch": 1.012151652624757,
+      "grad_norm": 0.6876736283302307,
+      "learning_rate": 3.860252542619305e-06,
+      "loss": 0.0884,
+      "step": 6247
+    },
+    {
+      "epoch": 1.0123136746597536,
+      "grad_norm": 0.8331660032272339,
+      "learning_rate": 3.859885615049419e-06,
+      "loss": 0.1155,
+      "step": 6248
+    },
+    {
+      "epoch": 1.0124756966947506,
+      "grad_norm": 0.8493173122406006,
+      "learning_rate": 3.859518645870323e-06,
+      "loss": 0.1121,
+      "step": 6249
+    },
+    {
+      "epoch": 1.0126377187297473,
+      "grad_norm": 0.7302095293998718,
+      "learning_rate": 3.8591516350932476e-06,
+      "loss": 0.0871,
+      "step": 6250
+    },
+    {
+      "epoch": 1.012799740764744,
+      "grad_norm": 0.8454660773277283,
+      "learning_rate": 3.85878458272942e-06,
+      "loss": 0.1086,
+      "step": 6251
+    },
+    {
+      "epoch": 1.0129617627997407,
+      "grad_norm": 0.786479651927948,
+      "learning_rate": 3.8584174887900735e-06,
+      "loss": 0.0989,
+      "step": 6252
+    },
+    {
+      "epoch": 1.0131237848347374,
+      "grad_norm": 0.8221831917762756,
+      "learning_rate": 3.858050353286439e-06,
+      "loss": 0.104,
+      "step": 6253
+    },
+    {
+      "epoch": 1.0132858068697344,
+      "grad_norm": 0.7293557524681091,
+      "learning_rate": 3.8576831762297495e-06,
+      "loss": 0.0906,
+      "step": 6254
+    },
+    {
+      "epoch": 1.013447828904731,
+      "grad_norm": 0.7902601361274719,
+      "learning_rate": 3.85731595763124e-06,
+      "loss": 0.1,
+      "step": 6255
+    },
+    {
+      "epoch": 1.0136098509397278,
+      "grad_norm": 0.8392426371574402,
+      "learning_rate": 3.856948697502148e-06,
+      "loss": 0.1109,
+      "step": 6256
+    },
+    {
+      "epoch": 1.0137718729747245,
+      "grad_norm": 0.7996932864189148,
+      "learning_rate": 3.856581395853709e-06,
+      "loss": 0.1005,
+      "step": 6257
+    },
+    {
+      "epoch": 1.0139338950097214,
+      "grad_norm": 0.7863364815711975,
+      "learning_rate": 3.8562140526971625e-06,
+      "loss": 0.0929,
+      "step": 6258
+    },
+    {
+      "epoch": 1.0140959170447181,
+      "grad_norm": 0.7616758942604065,
+      "learning_rate": 3.855846668043747e-06,
+      "loss": 0.0958,
+      "step": 6259
+    },
+    {
+      "epoch": 1.0142579390797148,
+      "grad_norm": 0.8799887299537659,
+      "learning_rate": 3.855479241904705e-06,
+      "loss": 0.0998,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0144199611147116,
+      "grad_norm": 0.8284506797790527,
+      "learning_rate": 3.855111774291279e-06,
+      "loss": 0.1061,
+      "step": 6261
+    },
+    {
+      "epoch": 1.0145819831497083,
+      "grad_norm": 0.7687853574752808,
+      "learning_rate": 3.8547442652147115e-06,
+      "loss": 0.0868,
+      "step": 6262
+    },
+    {
+      "epoch": 1.0147440051847052,
+      "grad_norm": 0.8260930776596069,
+      "learning_rate": 3.854376714686249e-06,
+      "loss": 0.1064,
+      "step": 6263
+    },
+    {
+      "epoch": 1.014906027219702,
+      "grad_norm": 0.7864606380462646,
+      "learning_rate": 3.854009122717135e-06,
+      "loss": 0.1048,
+      "step": 6264
+    },
+    {
+      "epoch": 1.0150680492546986,
+      "grad_norm": 0.8198474645614624,
+      "learning_rate": 3.853641489318619e-06,
+      "loss": 0.1052,
+      "step": 6265
+    },
+    {
+      "epoch": 1.0152300712896953,
+      "grad_norm": 0.8373472690582275,
+      "learning_rate": 3.8532738145019484e-06,
+      "loss": 0.1046,
+      "step": 6266
+    },
+    {
+      "epoch": 1.0153920933246923,
+      "grad_norm": 0.8073359131813049,
+      "learning_rate": 3.8529060982783756e-06,
+      "loss": 0.0921,
+      "step": 6267
+    },
+    {
+      "epoch": 1.015554115359689,
+      "grad_norm": 1.0620416402816772,
+      "learning_rate": 3.852538340659149e-06,
+      "loss": 0.1189,
+      "step": 6268
+    },
+    {
+      "epoch": 1.0157161373946857,
+      "grad_norm": 0.9277188181877136,
+      "learning_rate": 3.852170541655523e-06,
+      "loss": 0.111,
+      "step": 6269
+    },
+    {
+      "epoch": 1.0158781594296824,
+      "grad_norm": 0.8073578476905823,
+      "learning_rate": 3.85180270127875e-06,
+      "loss": 0.0924,
+      "step": 6270
+    },
+    {
+      "epoch": 1.0160401814646791,
+      "grad_norm": 0.8376033902168274,
+      "learning_rate": 3.8514348195400854e-06,
+      "loss": 0.1046,
+      "step": 6271
+    },
+    {
+      "epoch": 1.016202203499676,
+      "grad_norm": 0.7943440675735474,
+      "learning_rate": 3.851066896450787e-06,
+      "loss": 0.0997,
+      "step": 6272
+    },
+    {
+      "epoch": 1.0163642255346728,
+      "grad_norm": 0.8442563414573669,
+      "learning_rate": 3.85069893202211e-06,
+      "loss": 0.1051,
+      "step": 6273
+    },
+    {
+      "epoch": 1.0165262475696695,
+      "grad_norm": 0.8523194789886475,
+      "learning_rate": 3.850330926265314e-06,
+      "loss": 0.1097,
+      "step": 6274
+    },
+    {
+      "epoch": 1.0166882696046662,
+      "grad_norm": 0.6826653480529785,
+      "learning_rate": 3.849962879191661e-06,
+      "loss": 0.0823,
+      "step": 6275
+    },
+    {
+      "epoch": 1.016850291639663,
+      "grad_norm": 0.782434344291687,
+      "learning_rate": 3.849594790812409e-06,
+      "loss": 0.1043,
+      "step": 6276
+    },
+    {
+      "epoch": 1.0170123136746598,
+      "grad_norm": 0.8440970182418823,
+      "learning_rate": 3.849226661138823e-06,
+      "loss": 0.1049,
+      "step": 6277
+    },
+    {
+      "epoch": 1.0171743357096565,
+      "grad_norm": 0.809670090675354,
+      "learning_rate": 3.848858490182167e-06,
+      "loss": 0.1051,
+      "step": 6278
+    },
+    {
+      "epoch": 1.0173363577446533,
+      "grad_norm": 0.7815456390380859,
+      "learning_rate": 3.848490277953704e-06,
+      "loss": 0.0917,
+      "step": 6279
+    },
+    {
+      "epoch": 1.01749837977965,
+      "grad_norm": 0.8251093626022339,
+      "learning_rate": 3.8481220244647025e-06,
+      "loss": 0.1014,
+      "step": 6280
+    },
+    {
+      "epoch": 1.017660401814647,
+      "grad_norm": 0.7042380571365356,
+      "learning_rate": 3.84775372972643e-06,
+      "loss": 0.078,
+      "step": 6281
+    },
+    {
+      "epoch": 1.0178224238496436,
+      "grad_norm": 0.795159101486206,
+      "learning_rate": 3.847385393750154e-06,
+      "loss": 0.1048,
+      "step": 6282
+    },
+    {
+      "epoch": 1.0179844458846403,
+      "grad_norm": 0.8827878832817078,
+      "learning_rate": 3.847017016547146e-06,
+      "loss": 0.1123,
+      "step": 6283
+    },
+    {
+      "epoch": 1.018146467919637,
+      "grad_norm": 0.7256340980529785,
+      "learning_rate": 3.846648598128677e-06,
+      "loss": 0.0877,
+      "step": 6284
+    },
+    {
+      "epoch": 1.0183084899546337,
+      "grad_norm": 1.0416500568389893,
+      "learning_rate": 3.846280138506019e-06,
+      "loss": 0.1112,
+      "step": 6285
+    },
+    {
+      "epoch": 1.0184705119896307,
+      "grad_norm": 0.8687002062797546,
+      "learning_rate": 3.8459116376904475e-06,
+      "loss": 0.1075,
+      "step": 6286
+    },
+    {
+      "epoch": 1.0186325340246274,
+      "grad_norm": 0.8167076706886292,
+      "learning_rate": 3.845543095693236e-06,
+      "loss": 0.0987,
+      "step": 6287
+    },
+    {
+      "epoch": 1.018794556059624,
+      "grad_norm": 0.7660663723945618,
+      "learning_rate": 3.8451745125256635e-06,
+      "loss": 0.1017,
+      "step": 6288
+    },
+    {
+      "epoch": 1.0189565780946208,
+      "grad_norm": 0.80263751745224,
+      "learning_rate": 3.8448058881990055e-06,
+      "loss": 0.0947,
+      "step": 6289
+    },
+    {
+      "epoch": 1.0191186001296175,
+      "grad_norm": 0.9206402897834778,
+      "learning_rate": 3.8444372227245415e-06,
+      "loss": 0.1107,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0192806221646145,
+      "grad_norm": 0.9525105953216553,
+      "learning_rate": 3.8440685161135514e-06,
+      "loss": 0.1077,
+      "step": 6291
+    },
+    {
+      "epoch": 1.0194426441996112,
+      "grad_norm": 0.7992419004440308,
+      "learning_rate": 3.843699768377318e-06,
+      "loss": 0.1066,
+      "step": 6292
+    },
+    {
+      "epoch": 1.0196046662346079,
+      "grad_norm": 0.7668746709823608,
+      "learning_rate": 3.843330979527124e-06,
+      "loss": 0.0952,
+      "step": 6293
+    },
+    {
+      "epoch": 1.0197666882696046,
+      "grad_norm": 0.8306681513786316,
+      "learning_rate": 3.842962149574252e-06,
+      "loss": 0.1061,
+      "step": 6294
+    },
+    {
+      "epoch": 1.0199287103046015,
+      "grad_norm": 0.9119642376899719,
+      "learning_rate": 3.8425932785299875e-06,
+      "loss": 0.1174,
+      "step": 6295
+    },
+    {
+      "epoch": 1.0200907323395982,
+      "grad_norm": 0.8632789254188538,
+      "learning_rate": 3.842224366405619e-06,
+      "loss": 0.1145,
+      "step": 6296
+    },
+    {
+      "epoch": 1.020252754374595,
+      "grad_norm": 0.8445510864257812,
+      "learning_rate": 3.841855413212432e-06,
+      "loss": 0.1029,
+      "step": 6297
+    },
+    {
+      "epoch": 1.0204147764095917,
+      "grad_norm": 0.8951784372329712,
+      "learning_rate": 3.841486418961717e-06,
+      "loss": 0.1144,
+      "step": 6298
+    },
+    {
+      "epoch": 1.0205767984445884,
+      "grad_norm": 1.5688501596450806,
+      "learning_rate": 3.841117383664763e-06,
+      "loss": 0.0966,
+      "step": 6299
+    },
+    {
+      "epoch": 1.0207388204795853,
+      "grad_norm": 0.9988122582435608,
+      "learning_rate": 3.840748307332865e-06,
+      "loss": 0.1283,
+      "step": 6300
+    },
+    {
+      "epoch": 1.020900842514582,
+      "grad_norm": 0.7491373419761658,
+      "learning_rate": 3.84037918997731e-06,
+      "loss": 0.0965,
+      "step": 6301
+    },
+    {
+      "epoch": 1.0210628645495787,
+      "grad_norm": 0.8068737387657166,
+      "learning_rate": 3.840010031609398e-06,
+      "loss": 0.0956,
+      "step": 6302
+    },
+    {
+      "epoch": 1.0212248865845754,
+      "grad_norm": 0.8668238520622253,
+      "learning_rate": 3.839640832240421e-06,
+      "loss": 0.1107,
+      "step": 6303
+    },
+    {
+      "epoch": 1.0213869086195722,
+      "grad_norm": 0.9297085404396057,
+      "learning_rate": 3.8392715918816755e-06,
+      "loss": 0.1196,
+      "step": 6304
+    },
+    {
+      "epoch": 1.021548930654569,
+      "grad_norm": 0.7360817193984985,
+      "learning_rate": 3.8389023105444625e-06,
+      "loss": 0.0893,
+      "step": 6305
+    },
+    {
+      "epoch": 1.0217109526895658,
+      "grad_norm": 0.8899022340774536,
+      "learning_rate": 3.838532988240077e-06,
+      "loss": 0.1111,
+      "step": 6306
+    },
+    {
+      "epoch": 1.0218729747245625,
+      "grad_norm": 0.6964313387870789,
+      "learning_rate": 3.838163624979822e-06,
+      "loss": 0.0904,
+      "step": 6307
+    },
+    {
+      "epoch": 1.0220349967595592,
+      "grad_norm": 0.8330636024475098,
+      "learning_rate": 3.837794220774998e-06,
+      "loss": 0.0966,
+      "step": 6308
+    },
+    {
+      "epoch": 1.0221970187945562,
+      "grad_norm": 0.9185472726821899,
+      "learning_rate": 3.837424775636908e-06,
+      "loss": 0.1158,
+      "step": 6309
+    },
+    {
+      "epoch": 1.0223590408295529,
+      "grad_norm": 0.8500825762748718,
+      "learning_rate": 3.8370552895768565e-06,
+      "loss": 0.1024,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0225210628645496,
+      "grad_norm": 0.9668059349060059,
+      "learning_rate": 3.836685762606149e-06,
+      "loss": 0.1048,
+      "step": 6311
+    },
+    {
+      "epoch": 1.0226830848995463,
+      "grad_norm": 0.9085898399353027,
+      "learning_rate": 3.836316194736093e-06,
+      "loss": 0.1038,
+      "step": 6312
+    },
+    {
+      "epoch": 1.022845106934543,
+      "grad_norm": 0.829475998878479,
+      "learning_rate": 3.8359465859779934e-06,
+      "loss": 0.1031,
+      "step": 6313
+    },
+    {
+      "epoch": 1.02300712896954,
+      "grad_norm": 0.946996808052063,
+      "learning_rate": 3.835576936343162e-06,
+      "loss": 0.107,
+      "step": 6314
+    },
+    {
+      "epoch": 1.0231691510045366,
+      "grad_norm": 0.9723039865493774,
+      "learning_rate": 3.835207245842908e-06,
+      "loss": 0.1155,
+      "step": 6315
+    },
+    {
+      "epoch": 1.0233311730395334,
+      "grad_norm": 0.9182832837104797,
+      "learning_rate": 3.8348375144885445e-06,
+      "loss": 0.1155,
+      "step": 6316
+    },
+    {
+      "epoch": 1.02349319507453,
+      "grad_norm": 1.0076240301132202,
+      "learning_rate": 3.834467742291382e-06,
+      "loss": 0.12,
+      "step": 6317
+    },
+    {
+      "epoch": 1.023655217109527,
+      "grad_norm": 0.7207759618759155,
+      "learning_rate": 3.834097929262737e-06,
+      "loss": 0.0801,
+      "step": 6318
+    },
+    {
+      "epoch": 1.0238172391445237,
+      "grad_norm": 0.9244574904441833,
+      "learning_rate": 3.833728075413923e-06,
+      "loss": 0.1174,
+      "step": 6319
+    },
+    {
+      "epoch": 1.0239792611795204,
+      "grad_norm": 0.7919167280197144,
+      "learning_rate": 3.833358180756258e-06,
+      "loss": 0.0981,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0241412832145171,
+      "grad_norm": 0.7644730806350708,
+      "learning_rate": 3.832988245301058e-06,
+      "loss": 0.0937,
+      "step": 6321
+    },
+    {
+      "epoch": 1.0243033052495139,
+      "grad_norm": 0.8419947624206543,
+      "learning_rate": 3.832618269059645e-06,
+      "loss": 0.0974,
+      "step": 6322
+    },
+    {
+      "epoch": 1.0244653272845108,
+      "grad_norm": 0.7853063941001892,
+      "learning_rate": 3.832248252043338e-06,
+      "loss": 0.0995,
+      "step": 6323
+    },
+    {
+      "epoch": 1.0246273493195075,
+      "grad_norm": 0.9024096727371216,
+      "learning_rate": 3.831878194263458e-06,
+      "loss": 0.1134,
+      "step": 6324
+    },
+    {
+      "epoch": 1.0247893713545042,
+      "grad_norm": 0.8228088021278381,
+      "learning_rate": 3.831508095731328e-06,
+      "loss": 0.1003,
+      "step": 6325
+    },
+    {
+      "epoch": 1.024951393389501,
+      "grad_norm": 0.798202395439148,
+      "learning_rate": 3.831137956458272e-06,
+      "loss": 0.0996,
+      "step": 6326
+    },
+    {
+      "epoch": 1.0251134154244976,
+      "grad_norm": 1.0323954820632935,
+      "learning_rate": 3.830767776455617e-06,
+      "loss": 0.1177,
+      "step": 6327
+    },
+    {
+      "epoch": 1.0252754374594946,
+      "grad_norm": 0.8482728004455566,
+      "learning_rate": 3.830397555734687e-06,
+      "loss": 0.1058,
+      "step": 6328
+    },
+    {
+      "epoch": 1.0254374594944913,
+      "grad_norm": 0.8211125731468201,
+      "learning_rate": 3.830027294306813e-06,
+      "loss": 0.1168,
+      "step": 6329
+    },
+    {
+      "epoch": 1.025599481529488,
+      "grad_norm": 0.8774312138557434,
+      "learning_rate": 3.8296569921833214e-06,
+      "loss": 0.1134,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0257615035644847,
+      "grad_norm": 0.8460497856140137,
+      "learning_rate": 3.829286649375544e-06,
+      "loss": 0.1064,
+      "step": 6331
+    },
+    {
+      "epoch": 1.0259235255994816,
+      "grad_norm": 0.8636878728866577,
+      "learning_rate": 3.8289162658948114e-06,
+      "loss": 0.0956,
+      "step": 6332
+    },
+    {
+      "epoch": 1.0260855476344783,
+      "grad_norm": 0.8535585403442383,
+      "learning_rate": 3.828545841752457e-06,
+      "loss": 0.1093,
+      "step": 6333
+    },
+    {
+      "epoch": 1.026247569669475,
+      "grad_norm": 0.9328159093856812,
+      "learning_rate": 3.828175376959815e-06,
+      "loss": 0.1123,
+      "step": 6334
+    },
+    {
+      "epoch": 1.0264095917044718,
+      "grad_norm": 0.8551136255264282,
+      "learning_rate": 3.827804871528221e-06,
+      "loss": 0.0988,
+      "step": 6335
+    },
+    {
+      "epoch": 1.0265716137394685,
+      "grad_norm": 0.7710674405097961,
+      "learning_rate": 3.827434325469011e-06,
+      "loss": 0.0962,
+      "step": 6336
+    },
+    {
+      "epoch": 1.0267336357744654,
+      "grad_norm": 0.7422839999198914,
+      "learning_rate": 3.827063738793523e-06,
+      "loss": 0.0941,
+      "step": 6337
+    },
+    {
+      "epoch": 1.0268956578094621,
+      "grad_norm": 0.8090795874595642,
+      "learning_rate": 3.8266931115130955e-06,
+      "loss": 0.0914,
+      "step": 6338
+    },
+    {
+      "epoch": 1.0270576798444588,
+      "grad_norm": 0.8428863286972046,
+      "learning_rate": 3.82632244363907e-06,
+      "loss": 0.1066,
+      "step": 6339
+    },
+    {
+      "epoch": 1.0272197018794555,
+      "grad_norm": 0.8122255802154541,
+      "learning_rate": 3.8259517351827866e-06,
+      "loss": 0.1008,
+      "step": 6340
+    },
+    {
+      "epoch": 1.0273817239144523,
+      "grad_norm": 0.9258431792259216,
+      "learning_rate": 3.8255809861555895e-06,
+      "loss": 0.1109,
+      "step": 6341
+    },
+    {
+      "epoch": 1.0275437459494492,
+      "grad_norm": 0.886329174041748,
+      "learning_rate": 3.825210196568823e-06,
+      "loss": 0.1144,
+      "step": 6342
+    },
+    {
+      "epoch": 1.027705767984446,
+      "grad_norm": 0.9017022848129272,
+      "learning_rate": 3.824839366433829e-06,
+      "loss": 0.1152,
+      "step": 6343
+    },
+    {
+      "epoch": 1.0278677900194426,
+      "grad_norm": 0.8935456871986389,
+      "learning_rate": 3.824468495761958e-06,
+      "loss": 0.0945,
+      "step": 6344
+    },
+    {
+      "epoch": 1.0280298120544393,
+      "grad_norm": 0.8333491683006287,
+      "learning_rate": 3.824097584564556e-06,
+      "loss": 0.1041,
+      "step": 6345
+    },
+    {
+      "epoch": 1.0281918340894363,
+      "grad_norm": 0.7439111471176147,
+      "learning_rate": 3.823726632852972e-06,
+      "loss": 0.0931,
+      "step": 6346
+    },
+    {
+      "epoch": 1.028353856124433,
+      "grad_norm": 0.8706629276275635,
+      "learning_rate": 3.823355640638557e-06,
+      "loss": 0.1069,
+      "step": 6347
+    },
+    {
+      "epoch": 1.0285158781594297,
+      "grad_norm": 0.8351131081581116,
+      "learning_rate": 3.822984607932661e-06,
+      "loss": 0.1065,
+      "step": 6348
+    },
+    {
+      "epoch": 1.0286779001944264,
+      "grad_norm": 0.7511221766471863,
+      "learning_rate": 3.822613534746638e-06,
+      "loss": 0.0905,
+      "step": 6349
+    },
+    {
+      "epoch": 1.028839922229423,
+      "grad_norm": 0.9029641151428223,
+      "learning_rate": 3.8222424210918404e-06,
+      "loss": 0.1051,
+      "step": 6350
+    },
+    {
+      "epoch": 1.02900194426442,
+      "grad_norm": 0.770819902420044,
+      "learning_rate": 3.821871266979626e-06,
+      "loss": 0.0997,
+      "step": 6351
+    },
+    {
+      "epoch": 1.0291639662994168,
+      "grad_norm": 0.9414584040641785,
+      "learning_rate": 3.821500072421349e-06,
+      "loss": 0.1117,
+      "step": 6352
+    },
+    {
+      "epoch": 1.0293259883344135,
+      "grad_norm": 0.836219072341919,
+      "learning_rate": 3.821128837428368e-06,
+      "loss": 0.1091,
+      "step": 6353
+    },
+    {
+      "epoch": 1.0294880103694102,
+      "grad_norm": 0.9013179540634155,
+      "learning_rate": 3.820757562012042e-06,
+      "loss": 0.1025,
+      "step": 6354
+    },
+    {
+      "epoch": 1.029650032404407,
+      "grad_norm": 0.7845829725265503,
+      "learning_rate": 3.82038624618373e-06,
+      "loss": 0.1003,
+      "step": 6355
+    },
+    {
+      "epoch": 1.0298120544394038,
+      "grad_norm": 0.7390191555023193,
+      "learning_rate": 3.820014889954794e-06,
+      "loss": 0.0994,
+      "step": 6356
+    },
+    {
+      "epoch": 1.0299740764744005,
+      "grad_norm": 0.8984519243240356,
+      "learning_rate": 3.819643493336598e-06,
+      "loss": 0.1132,
+      "step": 6357
+    },
+    {
+      "epoch": 1.0301360985093972,
+      "grad_norm": 0.8943284749984741,
+      "learning_rate": 3.819272056340504e-06,
+      "loss": 0.1086,
+      "step": 6358
+    },
+    {
+      "epoch": 1.030298120544394,
+      "grad_norm": 0.9379931688308716,
+      "learning_rate": 3.818900578977877e-06,
+      "loss": 0.1062,
+      "step": 6359
+    },
+    {
+      "epoch": 1.030460142579391,
+      "grad_norm": 0.8385069370269775,
+      "learning_rate": 3.818529061260084e-06,
+      "loss": 0.1045,
+      "step": 6360
+    },
+    {
+      "epoch": 1.0306221646143876,
+      "grad_norm": 0.8240720629692078,
+      "learning_rate": 3.8181575031984935e-06,
+      "loss": 0.0985,
+      "step": 6361
+    },
+    {
+      "epoch": 1.0307841866493843,
+      "grad_norm": 0.8299695253372192,
+      "learning_rate": 3.817785904804473e-06,
+      "loss": 0.1021,
+      "step": 6362
+    },
+    {
+      "epoch": 1.030946208684381,
+      "grad_norm": 0.7998603582382202,
+      "learning_rate": 3.817414266089392e-06,
+      "loss": 0.1041,
+      "step": 6363
+    },
+    {
+      "epoch": 1.0311082307193777,
+      "grad_norm": 0.9111914038658142,
+      "learning_rate": 3.817042587064623e-06,
+      "loss": 0.1054,
+      "step": 6364
+    },
+    {
+      "epoch": 1.0312702527543747,
+      "grad_norm": 0.7303438186645508,
+      "learning_rate": 3.816670867741538e-06,
+      "loss": 0.0943,
+      "step": 6365
+    },
+    {
+      "epoch": 1.0314322747893714,
+      "grad_norm": 0.8028538227081299,
+      "learning_rate": 3.81629910813151e-06,
+      "loss": 0.0959,
+      "step": 6366
+    },
+    {
+      "epoch": 1.031594296824368,
+      "grad_norm": 0.7954698204994202,
+      "learning_rate": 3.815927308245917e-06,
+      "loss": 0.1083,
+      "step": 6367
+    },
+    {
+      "epoch": 1.0317563188593648,
+      "grad_norm": 0.7787458896636963,
+      "learning_rate": 3.815555468096131e-06,
+      "loss": 0.1001,
+      "step": 6368
+    },
+    {
+      "epoch": 1.0319183408943617,
+      "grad_norm": 0.8813876509666443,
+      "learning_rate": 3.815183587693531e-06,
+      "loss": 0.1022,
+      "step": 6369
+    },
+    {
+      "epoch": 1.0320803629293585,
+      "grad_norm": 0.9205014705657959,
+      "learning_rate": 3.814811667049497e-06,
+      "loss": 0.1054,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0322423849643552,
+      "grad_norm": 0.7909181714057922,
+      "learning_rate": 3.8144397061754066e-06,
+      "loss": 0.1005,
+      "step": 6371
+    },
+    {
+      "epoch": 1.0324044069993519,
+      "grad_norm": 0.7991738319396973,
+      "learning_rate": 3.814067705082643e-06,
+      "loss": 0.0988,
+      "step": 6372
+    },
+    {
+      "epoch": 1.0325664290343486,
+      "grad_norm": 0.8876237869262695,
+      "learning_rate": 3.8136956637825878e-06,
+      "loss": 0.1064,
+      "step": 6373
+    },
+    {
+      "epoch": 1.0327284510693455,
+      "grad_norm": 1.0412876605987549,
+      "learning_rate": 3.8133235822866234e-06,
+      "loss": 0.1069,
+      "step": 6374
+    },
+    {
+      "epoch": 1.0328904731043422,
+      "grad_norm": 0.9145109057426453,
+      "learning_rate": 3.812951460606136e-06,
+      "loss": 0.0998,
+      "step": 6375
+    },
+    {
+      "epoch": 1.033052495139339,
+      "grad_norm": 0.7650631070137024,
+      "learning_rate": 3.812579298752511e-06,
+      "loss": 0.0925,
+      "step": 6376
+    },
+    {
+      "epoch": 1.0332145171743357,
+      "grad_norm": 0.7742469310760498,
+      "learning_rate": 3.812207096737137e-06,
+      "loss": 0.0951,
+      "step": 6377
+    },
+    {
+      "epoch": 1.0333765392093324,
+      "grad_norm": 0.900032639503479,
+      "learning_rate": 3.8118348545714e-06,
+      "loss": 0.1075,
+      "step": 6378
+    },
+    {
+      "epoch": 1.0335385612443293,
+      "grad_norm": 0.8306341767311096,
+      "learning_rate": 3.811462572266691e-06,
+      "loss": 0.1033,
+      "step": 6379
+    },
+    {
+      "epoch": 1.033700583279326,
+      "grad_norm": 0.9517765641212463,
+      "learning_rate": 3.8110902498344023e-06,
+      "loss": 0.1076,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0338626053143227,
+      "grad_norm": 0.8399965763092041,
+      "learning_rate": 3.810717887285923e-06,
+      "loss": 0.0931,
+      "step": 6381
+    },
+    {
+      "epoch": 1.0340246273493194,
+      "grad_norm": 1.0168367624282837,
+      "learning_rate": 3.8103454846326493e-06,
+      "loss": 0.103,
+      "step": 6382
+    },
+    {
+      "epoch": 1.0341866493843164,
+      "grad_norm": 0.8459004163742065,
+      "learning_rate": 3.8099730418859743e-06,
+      "loss": 0.1079,
+      "step": 6383
+    },
+    {
+      "epoch": 1.034348671419313,
+      "grad_norm": 0.8548713326454163,
+      "learning_rate": 3.809600559057295e-06,
+      "loss": 0.0932,
+      "step": 6384
+    },
+    {
+      "epoch": 1.0345106934543098,
+      "grad_norm": 0.986228346824646,
+      "learning_rate": 3.809228036158007e-06,
+      "loss": 0.1074,
+      "step": 6385
+    },
+    {
+      "epoch": 1.0346727154893065,
+      "grad_norm": 0.919900119304657,
+      "learning_rate": 3.80885547319951e-06,
+      "loss": 0.1051,
+      "step": 6386
+    },
+    {
+      "epoch": 1.0348347375243032,
+      "grad_norm": 0.8741225004196167,
+      "learning_rate": 3.808482870193202e-06,
+      "loss": 0.1084,
+      "step": 6387
+    },
+    {
+      "epoch": 1.0349967595593002,
+      "grad_norm": 0.898740828037262,
+      "learning_rate": 3.808110227150485e-06,
+      "loss": 0.0994,
+      "step": 6388
+    },
+    {
+      "epoch": 1.0351587815942969,
+      "grad_norm": 0.8765122890472412,
+      "learning_rate": 3.8077375440827613e-06,
+      "loss": 0.1025,
+      "step": 6389
+    },
+    {
+      "epoch": 1.0353208036292936,
+      "grad_norm": 0.8248468637466431,
+      "learning_rate": 3.8073648210014323e-06,
+      "loss": 0.0955,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0354828256642903,
+      "grad_norm": 0.7942231297492981,
+      "learning_rate": 3.8069920579179042e-06,
+      "loss": 0.0982,
+      "step": 6391
+    },
+    {
+      "epoch": 1.035644847699287,
+      "grad_norm": 0.8907127380371094,
+      "learning_rate": 3.806619254843582e-06,
+      "loss": 0.1016,
+      "step": 6392
+    },
+    {
+      "epoch": 1.035806869734284,
+      "grad_norm": 0.9660011529922485,
+      "learning_rate": 3.806246411789872e-06,
+      "loss": 0.1185,
+      "step": 6393
+    },
+    {
+      "epoch": 1.0359688917692806,
+      "grad_norm": 0.7657979130744934,
+      "learning_rate": 3.8058735287681835e-06,
+      "loss": 0.0972,
+      "step": 6394
+    },
+    {
+      "epoch": 1.0361309138042774,
+      "grad_norm": 0.8937506675720215,
+      "learning_rate": 3.8055006057899254e-06,
+      "loss": 0.1065,
+      "step": 6395
+    },
+    {
+      "epoch": 1.036292935839274,
+      "grad_norm": 0.789402425289154,
+      "learning_rate": 3.8051276428665074e-06,
+      "loss": 0.091,
+      "step": 6396
+    },
+    {
+      "epoch": 1.036454957874271,
+      "grad_norm": 0.9826111197471619,
+      "learning_rate": 3.8047546400093425e-06,
+      "loss": 0.1106,
+      "step": 6397
+    },
+    {
+      "epoch": 1.0366169799092677,
+      "grad_norm": 0.8755872249603271,
+      "learning_rate": 3.8043815972298424e-06,
+      "loss": 0.114,
+      "step": 6398
+    },
+    {
+      "epoch": 1.0367790019442644,
+      "grad_norm": 0.8057990074157715,
+      "learning_rate": 3.8040085145394224e-06,
+      "loss": 0.1052,
+      "step": 6399
+    },
+    {
+      "epoch": 1.0369410239792611,
+      "grad_norm": 0.8804965615272522,
+      "learning_rate": 3.8036353919494973e-06,
+      "loss": 0.1134,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0371030460142578,
+      "grad_norm": 0.8010181784629822,
+      "learning_rate": 3.8032622294714837e-06,
+      "loss": 0.1062,
+      "step": 6401
+    },
+    {
+      "epoch": 1.0372650680492548,
+      "grad_norm": 0.7778938412666321,
+      "learning_rate": 3.8028890271168e-06,
+      "loss": 0.0924,
+      "step": 6402
+    },
+    {
+      "epoch": 1.0374270900842515,
+      "grad_norm": 0.7856367826461792,
+      "learning_rate": 3.8025157848968653e-06,
+      "loss": 0.1058,
+      "step": 6403
+    },
+    {
+      "epoch": 1.0375891121192482,
+      "grad_norm": 0.6976110339164734,
+      "learning_rate": 3.8021425028230994e-06,
+      "loss": 0.0898,
+      "step": 6404
+    },
+    {
+      "epoch": 1.037751134154245,
+      "grad_norm": 0.7459751963615417,
+      "learning_rate": 3.8017691809069234e-06,
+      "loss": 0.0889,
+      "step": 6405
+    },
+    {
+      "epoch": 1.0379131561892416,
+      "grad_norm": 0.7587976455688477,
+      "learning_rate": 3.801395819159761e-06,
+      "loss": 0.0969,
+      "step": 6406
+    },
+    {
+      "epoch": 1.0380751782242386,
+      "grad_norm": 0.7939549684524536,
+      "learning_rate": 3.8010224175930366e-06,
+      "loss": 0.0979,
+      "step": 6407
+    },
+    {
+      "epoch": 1.0382372002592353,
+      "grad_norm": 0.7911587953567505,
+      "learning_rate": 3.8006489762181744e-06,
+      "loss": 0.0932,
+      "step": 6408
+    },
+    {
+      "epoch": 1.038399222294232,
+      "grad_norm": 0.9099779725074768,
+      "learning_rate": 3.8002754950466004e-06,
+      "loss": 0.1058,
+      "step": 6409
+    },
+    {
+      "epoch": 1.0385612443292287,
+      "grad_norm": 0.7849684953689575,
+      "learning_rate": 3.7999019740897423e-06,
+      "loss": 0.0961,
+      "step": 6410
+    },
+    {
+      "epoch": 1.0387232663642256,
+      "grad_norm": 0.8940638303756714,
+      "learning_rate": 3.7995284133590317e-06,
+      "loss": 0.0876,
+      "step": 6411
+    },
+    {
+      "epoch": 1.0388852883992223,
+      "grad_norm": 0.9506855607032776,
+      "learning_rate": 3.799154812865894e-06,
+      "loss": 0.1111,
+      "step": 6412
+    },
+    {
+      "epoch": 1.039047310434219,
+      "grad_norm": 0.8368770480155945,
+      "learning_rate": 3.798781172621765e-06,
+      "loss": 0.1079,
+      "step": 6413
+    },
+    {
+      "epoch": 1.0392093324692158,
+      "grad_norm": 0.8906238079071045,
+      "learning_rate": 3.7984074926380733e-06,
+      "loss": 0.1126,
+      "step": 6414
+    },
+    {
+      "epoch": 1.0393713545042125,
+      "grad_norm": 0.950920581817627,
+      "learning_rate": 3.7980337729262555e-06,
+      "loss": 0.1118,
+      "step": 6415
+    },
+    {
+      "epoch": 1.0395333765392094,
+      "grad_norm": 0.9369775652885437,
+      "learning_rate": 3.7976600134977455e-06,
+      "loss": 0.1117,
+      "step": 6416
+    },
+    {
+      "epoch": 1.0396953985742061,
+      "grad_norm": 0.8446056246757507,
+      "learning_rate": 3.7972862143639788e-06,
+      "loss": 0.0977,
+      "step": 6417
+    },
+    {
+      "epoch": 1.0398574206092028,
+      "grad_norm": 0.930582582950592,
+      "learning_rate": 3.7969123755363935e-06,
+      "loss": 0.1135,
+      "step": 6418
+    },
+    {
+      "epoch": 1.0400194426441995,
+      "grad_norm": 0.7642794251441956,
+      "learning_rate": 3.796538497026428e-06,
+      "loss": 0.0935,
+      "step": 6419
+    },
+    {
+      "epoch": 1.0401814646791965,
+      "grad_norm": 0.9868995547294617,
+      "learning_rate": 3.7961645788455225e-06,
+      "loss": 0.1143,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0403434867141932,
+      "grad_norm": 1.0182535648345947,
+      "learning_rate": 3.7957906210051173e-06,
+      "loss": 0.0991,
+      "step": 6421
+    },
+    {
+      "epoch": 1.04050550874919,
+      "grad_norm": 0.9137951731681824,
+      "learning_rate": 3.7954166235166545e-06,
+      "loss": 0.1159,
+      "step": 6422
+    },
+    {
+      "epoch": 1.0406675307841866,
+      "grad_norm": 0.9183018207550049,
+      "learning_rate": 3.795042586391578e-06,
+      "loss": 0.09,
+      "step": 6423
+    },
+    {
+      "epoch": 1.0408295528191833,
+      "grad_norm": 0.856632649898529,
+      "learning_rate": 3.794668509641332e-06,
+      "loss": 0.1037,
+      "step": 6424
+    },
+    {
+      "epoch": 1.0409915748541803,
+      "grad_norm": 0.8316843509674072,
+      "learning_rate": 3.7942943932773636e-06,
+      "loss": 0.1061,
+      "step": 6425
+    },
+    {
+      "epoch": 1.041153596889177,
+      "grad_norm": 0.8080172538757324,
+      "learning_rate": 3.793920237311118e-06,
+      "loss": 0.1112,
+      "step": 6426
+    },
+    {
+      "epoch": 1.0413156189241737,
+      "grad_norm": 0.8021390438079834,
+      "learning_rate": 3.793546041754044e-06,
+      "loss": 0.1035,
+      "step": 6427
+    },
+    {
+      "epoch": 1.0414776409591704,
+      "grad_norm": 0.8733294010162354,
+      "learning_rate": 3.793171806617593e-06,
+      "loss": 0.1044,
+      "step": 6428
+    },
+    {
+      "epoch": 1.041639662994167,
+      "grad_norm": 0.8201737403869629,
+      "learning_rate": 3.7927975319132133e-06,
+      "loss": 0.0958,
+      "step": 6429
+    },
+    {
+      "epoch": 1.041801685029164,
+      "grad_norm": 0.7947893142700195,
+      "learning_rate": 3.7924232176523574e-06,
+      "loss": 0.1048,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0419637070641607,
+      "grad_norm": 0.839570939540863,
+      "learning_rate": 3.7920488638464788e-06,
+      "loss": 0.1012,
+      "step": 6431
+    },
+    {
+      "epoch": 1.0421257290991575,
+      "grad_norm": 0.8176724910736084,
+      "learning_rate": 3.7916744705070318e-06,
+      "loss": 0.0941,
+      "step": 6432
+    },
+    {
+      "epoch": 1.0422877511341542,
+      "grad_norm": 0.7297854423522949,
+      "learning_rate": 3.7913000376454713e-06,
+      "loss": 0.0962,
+      "step": 6433
+    },
+    {
+      "epoch": 1.042449773169151,
+      "grad_norm": 0.7584792375564575,
+      "learning_rate": 3.790925565273255e-06,
+      "loss": 0.0948,
+      "step": 6434
+    },
+    {
+      "epoch": 1.0426117952041478,
+      "grad_norm": 0.7733738422393799,
+      "learning_rate": 3.790551053401841e-06,
+      "loss": 0.0947,
+      "step": 6435
+    },
+    {
+      "epoch": 1.0427738172391445,
+      "grad_norm": 0.8243172764778137,
+      "learning_rate": 3.790176502042686e-06,
+      "loss": 0.1108,
+      "step": 6436
+    },
+    {
+      "epoch": 1.0429358392741412,
+      "grad_norm": 0.8829347491264343,
+      "learning_rate": 3.7898019112072537e-06,
+      "loss": 0.114,
+      "step": 6437
+    },
+    {
+      "epoch": 1.043097861309138,
+      "grad_norm": 0.8396970629692078,
+      "learning_rate": 3.789427280907004e-06,
+      "loss": 0.1037,
+      "step": 6438
+    },
+    {
+      "epoch": 1.0432598833441349,
+      "grad_norm": 0.8729767203330994,
+      "learning_rate": 3.7890526111534e-06,
+      "loss": 0.107,
+      "step": 6439
+    },
+    {
+      "epoch": 1.0434219053791316,
+      "grad_norm": 0.9895196557044983,
+      "learning_rate": 3.7886779019579045e-06,
+      "loss": 0.1178,
+      "step": 6440
+    },
+    {
+      "epoch": 1.0435839274141283,
+      "grad_norm": 0.8271682262420654,
+      "learning_rate": 3.788303153331985e-06,
+      "loss": 0.1127,
+      "step": 6441
+    },
+    {
+      "epoch": 1.043745949449125,
+      "grad_norm": 0.8002727031707764,
+      "learning_rate": 3.787928365287106e-06,
+      "loss": 0.1027,
+      "step": 6442
+    },
+    {
+      "epoch": 1.043907971484122,
+      "grad_norm": 0.788192629814148,
+      "learning_rate": 3.7875535378347356e-06,
+      "loss": 0.1054,
+      "step": 6443
+    },
+    {
+      "epoch": 1.0440699935191187,
+      "grad_norm": 0.8381497263908386,
+      "learning_rate": 3.7871786709863435e-06,
+      "loss": 0.1006,
+      "step": 6444
+    },
+    {
+      "epoch": 1.0442320155541154,
+      "grad_norm": 0.9270142912864685,
+      "learning_rate": 3.7868037647533977e-06,
+      "loss": 0.1152,
+      "step": 6445
+    },
+    {
+      "epoch": 1.044394037589112,
+      "grad_norm": 0.8742038011550903,
+      "learning_rate": 3.7864288191473718e-06,
+      "loss": 0.11,
+      "step": 6446
+    },
+    {
+      "epoch": 1.0445560596241088,
+      "grad_norm": 0.7907035946846008,
+      "learning_rate": 3.786053834179737e-06,
+      "loss": 0.105,
+      "step": 6447
+    },
+    {
+      "epoch": 1.0447180816591057,
+      "grad_norm": 0.7882545590400696,
+      "learning_rate": 3.7856788098619667e-06,
+      "loss": 0.1039,
+      "step": 6448
+    },
+    {
+      "epoch": 1.0448801036941024,
+      "grad_norm": 0.880588710308075,
+      "learning_rate": 3.7853037462055366e-06,
+      "loss": 0.1136,
+      "step": 6449
+    },
+    {
+      "epoch": 1.0450421257290992,
+      "grad_norm": 0.7238618731498718,
+      "learning_rate": 3.7849286432219216e-06,
+      "loss": 0.0888,
+      "step": 6450
+    },
+    {
+      "epoch": 1.0452041477640959,
+      "grad_norm": 0.887286365032196,
+      "learning_rate": 3.7845535009226e-06,
+      "loss": 0.107,
+      "step": 6451
+    },
+    {
+      "epoch": 1.0453661697990926,
+      "grad_norm": 0.8213676810264587,
+      "learning_rate": 3.78417831931905e-06,
+      "loss": 0.1036,
+      "step": 6452
+    },
+    {
+      "epoch": 1.0455281918340895,
+      "grad_norm": 0.8090675473213196,
+      "learning_rate": 3.783803098422751e-06,
+      "loss": 0.0995,
+      "step": 6453
+    },
+    {
+      "epoch": 1.0456902138690862,
+      "grad_norm": 0.8054291605949402,
+      "learning_rate": 3.783427838245184e-06,
+      "loss": 0.1053,
+      "step": 6454
+    },
+    {
+      "epoch": 1.045852235904083,
+      "grad_norm": 0.9021458625793457,
+      "learning_rate": 3.78305253879783e-06,
+      "loss": 0.1142,
+      "step": 6455
+    },
+    {
+      "epoch": 1.0460142579390797,
+      "grad_norm": 0.710974395275116,
+      "learning_rate": 3.7826772000921742e-06,
+      "loss": 0.0924,
+      "step": 6456
+    },
+    {
+      "epoch": 1.0461762799740764,
+      "grad_norm": 0.8471440672874451,
+      "learning_rate": 3.7823018221397e-06,
+      "loss": 0.1111,
+      "step": 6457
+    },
+    {
+      "epoch": 1.0463383020090733,
+      "grad_norm": 0.8694987297058105,
+      "learning_rate": 3.781926404951893e-06,
+      "loss": 0.1016,
+      "step": 6458
+    },
+    {
+      "epoch": 1.04650032404407,
+      "grad_norm": 0.8285094499588013,
+      "learning_rate": 3.78155094854024e-06,
+      "loss": 0.1016,
+      "step": 6459
+    },
+    {
+      "epoch": 1.0466623460790667,
+      "grad_norm": 0.9631251692771912,
+      "learning_rate": 3.7811754529162294e-06,
+      "loss": 0.1083,
+      "step": 6460
+    },
+    {
+      "epoch": 1.0468243681140634,
+      "grad_norm": 0.7672160267829895,
+      "learning_rate": 3.7807999180913514e-06,
+      "loss": 0.0994,
+      "step": 6461
+    },
+    {
+      "epoch": 1.0469863901490604,
+      "grad_norm": 0.8148001432418823,
+      "learning_rate": 3.7804243440770936e-06,
+      "loss": 0.1043,
+      "step": 6462
+    },
+    {
+      "epoch": 1.047148412184057,
+      "grad_norm": 0.9620845317840576,
+      "learning_rate": 3.780048730884951e-06,
+      "loss": 0.1056,
+      "step": 6463
+    },
+    {
+      "epoch": 1.0473104342190538,
+      "grad_norm": 0.840048611164093,
+      "learning_rate": 3.779673078526414e-06,
+      "loss": 0.1031,
+      "step": 6464
+    },
+    {
+      "epoch": 1.0474724562540505,
+      "grad_norm": 1.0370488166809082,
+      "learning_rate": 3.7792973870129773e-06,
+      "loss": 0.1211,
+      "step": 6465
+    },
+    {
+      "epoch": 1.0476344782890472,
+      "grad_norm": 0.7531951665878296,
+      "learning_rate": 3.7789216563561373e-06,
+      "loss": 0.0909,
+      "step": 6466
+    },
+    {
+      "epoch": 1.0477965003240441,
+      "grad_norm": 0.9405876994132996,
+      "learning_rate": 3.7785458865673885e-06,
+      "loss": 0.1147,
+      "step": 6467
+    },
+    {
+      "epoch": 1.0479585223590409,
+      "grad_norm": 1.0181210041046143,
+      "learning_rate": 3.778170077658231e-06,
+      "loss": 0.1209,
+      "step": 6468
+    },
+    {
+      "epoch": 1.0481205443940376,
+      "grad_norm": 0.8447110056877136,
+      "learning_rate": 3.7777942296401606e-06,
+      "loss": 0.1024,
+      "step": 6469
+    },
+    {
+      "epoch": 1.0482825664290343,
+      "grad_norm": 0.8490682244300842,
+      "learning_rate": 3.77741834252468e-06,
+      "loss": 0.1012,
+      "step": 6470
+    },
+    {
+      "epoch": 1.0484445884640312,
+      "grad_norm": 0.8966073393821716,
+      "learning_rate": 3.777042416323289e-06,
+      "loss": 0.104,
+      "step": 6471
+    },
+    {
+      "epoch": 1.048606610499028,
+      "grad_norm": 0.7608264088630676,
+      "learning_rate": 3.7766664510474903e-06,
+      "loss": 0.0988,
+      "step": 6472
+    },
+    {
+      "epoch": 1.0487686325340246,
+      "grad_norm": 0.9018314480781555,
+      "learning_rate": 3.776290446708789e-06,
+      "loss": 0.1082,
+      "step": 6473
+    },
+    {
+      "epoch": 1.0489306545690213,
+      "grad_norm": 0.8391294479370117,
+      "learning_rate": 3.775914403318687e-06,
+      "loss": 0.1015,
+      "step": 6474
+    },
+    {
+      "epoch": 1.049092676604018,
+      "grad_norm": 0.8135133981704712,
+      "learning_rate": 3.7755383208886923e-06,
+      "loss": 0.095,
+      "step": 6475
+    },
+    {
+      "epoch": 1.049254698639015,
+      "grad_norm": 0.8267797231674194,
+      "learning_rate": 3.7751621994303123e-06,
+      "loss": 0.0941,
+      "step": 6476
+    },
+    {
+      "epoch": 1.0494167206740117,
+      "grad_norm": 0.7961505055427551,
+      "learning_rate": 3.774786038955054e-06,
+      "loss": 0.1048,
+      "step": 6477
+    },
+    {
+      "epoch": 1.0495787427090084,
+      "grad_norm": 0.7784950137138367,
+      "learning_rate": 3.7744098394744287e-06,
+      "loss": 0.1042,
+      "step": 6478
+    },
+    {
+      "epoch": 1.0497407647440051,
+      "grad_norm": 0.8187713623046875,
+      "learning_rate": 3.774033600999946e-06,
+      "loss": 0.1076,
+      "step": 6479
+    },
+    {
+      "epoch": 1.0499027867790018,
+      "grad_norm": 0.8849846720695496,
+      "learning_rate": 3.7736573235431174e-06,
+      "loss": 0.1115,
+      "step": 6480
+    },
+    {
+      "epoch": 1.0500648088139988,
+      "grad_norm": 0.830732524394989,
+      "learning_rate": 3.773281007115458e-06,
+      "loss": 0.1076,
+      "step": 6481
+    },
+    {
+      "epoch": 1.0502268308489955,
+      "grad_norm": 0.8834472298622131,
+      "learning_rate": 3.7729046517284805e-06,
+      "loss": 0.1195,
+      "step": 6482
+    },
+    {
+      "epoch": 1.0503888528839922,
+      "grad_norm": 0.8346211910247803,
+      "learning_rate": 3.7725282573937015e-06,
+      "loss": 0.112,
+      "step": 6483
+    },
+    {
+      "epoch": 1.050550874918989,
+      "grad_norm": 0.7828723192214966,
+      "learning_rate": 3.7721518241226375e-06,
+      "loss": 0.0984,
+      "step": 6484
+    },
+    {
+      "epoch": 1.0507128969539858,
+      "grad_norm": 0.8663693070411682,
+      "learning_rate": 3.7717753519268053e-06,
+      "loss": 0.1056,
+      "step": 6485
+    },
+    {
+      "epoch": 1.0508749189889826,
+      "grad_norm": 0.819918155670166,
+      "learning_rate": 3.771398840817725e-06,
+      "loss": 0.0987,
+      "step": 6486
+    },
+    {
+      "epoch": 1.0510369410239793,
+      "grad_norm": 0.7646182179450989,
+      "learning_rate": 3.771022290806917e-06,
+      "loss": 0.0878,
+      "step": 6487
+    },
+    {
+      "epoch": 1.051198963058976,
+      "grad_norm": 0.8001990914344788,
+      "learning_rate": 3.770645701905904e-06,
+      "loss": 0.095,
+      "step": 6488
+    },
+    {
+      "epoch": 1.0513609850939727,
+      "grad_norm": 0.9029562473297119,
+      "learning_rate": 3.770269074126206e-06,
+      "loss": 0.1188,
+      "step": 6489
+    },
+    {
+      "epoch": 1.0515230071289696,
+      "grad_norm": 0.9364879727363586,
+      "learning_rate": 3.7698924074793484e-06,
+      "loss": 0.1128,
+      "step": 6490
+    },
+    {
+      "epoch": 1.0516850291639663,
+      "grad_norm": 0.8705113530158997,
+      "learning_rate": 3.769515701976856e-06,
+      "loss": 0.0952,
+      "step": 6491
+    },
+    {
+      "epoch": 1.051847051198963,
+      "grad_norm": 0.8115032911300659,
+      "learning_rate": 3.7691389576302567e-06,
+      "loss": 0.099,
+      "step": 6492
+    },
+    {
+      "epoch": 1.0520090732339598,
+      "grad_norm": 0.8196799159049988,
+      "learning_rate": 3.7687621744510756e-06,
+      "loss": 0.0995,
+      "step": 6493
+    },
+    {
+      "epoch": 1.0521710952689567,
+      "grad_norm": 0.8506253957748413,
+      "learning_rate": 3.768385352450842e-06,
+      "loss": 0.0996,
+      "step": 6494
+    },
+    {
+      "epoch": 1.0523331173039534,
+      "grad_norm": 0.7713782787322998,
+      "learning_rate": 3.7680084916410876e-06,
+      "loss": 0.0889,
+      "step": 6495
+    },
+    {
+      "epoch": 1.0524951393389501,
+      "grad_norm": 0.8281717896461487,
+      "learning_rate": 3.7676315920333396e-06,
+      "loss": 0.1026,
+      "step": 6496
+    },
+    {
+      "epoch": 1.0526571613739468,
+      "grad_norm": 0.870488703250885,
+      "learning_rate": 3.7672546536391343e-06,
+      "loss": 0.1065,
+      "step": 6497
+    },
+    {
+      "epoch": 1.0528191834089435,
+      "grad_norm": 0.9622962474822998,
+      "learning_rate": 3.7668776764700023e-06,
+      "loss": 0.1097,
+      "step": 6498
+    },
+    {
+      "epoch": 1.0529812054439405,
+      "grad_norm": 0.8624140620231628,
+      "learning_rate": 3.76650066053748e-06,
+      "loss": 0.1093,
+      "step": 6499
+    },
+    {
+      "epoch": 1.0531432274789372,
+      "grad_norm": 0.9349178671836853,
+      "learning_rate": 3.766123605853101e-06,
+      "loss": 0.115,
+      "step": 6500
+    },
+    {
+      "epoch": 1.053305249513934,
+      "grad_norm": 0.8151592016220093,
+      "learning_rate": 3.7657465124284047e-06,
+      "loss": 0.0977,
+      "step": 6501
+    },
+    {
+      "epoch": 1.0534672715489306,
+      "grad_norm": 0.9163764715194702,
+      "learning_rate": 3.765369380274928e-06,
+      "loss": 0.1155,
+      "step": 6502
+    },
+    {
+      "epoch": 1.0536292935839273,
+      "grad_norm": 0.8974437117576599,
+      "learning_rate": 3.76499220940421e-06,
+      "loss": 0.1149,
+      "step": 6503
+    },
+    {
+      "epoch": 1.0537913156189243,
+      "grad_norm": 0.8582158088684082,
+      "learning_rate": 3.7646149998277924e-06,
+      "loss": 0.1047,
+      "step": 6504
+    },
+    {
+      "epoch": 1.053953337653921,
+      "grad_norm": 0.8871607184410095,
+      "learning_rate": 3.7642377515572153e-06,
+      "loss": 0.0993,
+      "step": 6505
+    },
+    {
+      "epoch": 1.0541153596889177,
+      "grad_norm": 0.8530762791633606,
+      "learning_rate": 3.7638604646040232e-06,
+      "loss": 0.1047,
+      "step": 6506
+    },
+    {
+      "epoch": 1.0542773817239144,
+      "grad_norm": 0.8357167840003967,
+      "learning_rate": 3.763483138979759e-06,
+      "loss": 0.111,
+      "step": 6507
+    },
+    {
+      "epoch": 1.054439403758911,
+      "grad_norm": 0.8622068166732788,
+      "learning_rate": 3.763105774695968e-06,
+      "loss": 0.111,
+      "step": 6508
+    },
+    {
+      "epoch": 1.054601425793908,
+      "grad_norm": 0.8668839335441589,
+      "learning_rate": 3.762728371764197e-06,
+      "loss": 0.102,
+      "step": 6509
+    },
+    {
+      "epoch": 1.0547634478289047,
+      "grad_norm": 0.7583625316619873,
+      "learning_rate": 3.7623509301959935e-06,
+      "loss": 0.102,
+      "step": 6510
+    },
+    {
+      "epoch": 1.0549254698639015,
+      "grad_norm": 0.7609004974365234,
+      "learning_rate": 3.761973450002907e-06,
+      "loss": 0.0888,
+      "step": 6511
+    },
+    {
+      "epoch": 1.0550874918988982,
+      "grad_norm": 0.8587732911109924,
+      "learning_rate": 3.7615959311964865e-06,
+      "loss": 0.1126,
+      "step": 6512
+    },
+    {
+      "epoch": 1.055249513933895,
+      "grad_norm": 0.8386414647102356,
+      "learning_rate": 3.7612183737882833e-06,
+      "loss": 0.1107,
+      "step": 6513
+    },
+    {
+      "epoch": 1.0554115359688918,
+      "grad_norm": 0.7540522813796997,
+      "learning_rate": 3.760840777789851e-06,
+      "loss": 0.0933,
+      "step": 6514
+    },
+    {
+      "epoch": 1.0555735580038885,
+      "grad_norm": 0.7534650564193726,
+      "learning_rate": 3.7604631432127413e-06,
+      "loss": 0.0938,
+      "step": 6515
+    },
+    {
+      "epoch": 1.0557355800388852,
+      "grad_norm": 0.831891655921936,
+      "learning_rate": 3.7600854700685095e-06,
+      "loss": 0.108,
+      "step": 6516
+    },
+    {
+      "epoch": 1.055897602073882,
+      "grad_norm": 0.8216171264648438,
+      "learning_rate": 3.7597077583687115e-06,
+      "loss": 0.1109,
+      "step": 6517
+    },
+    {
+      "epoch": 1.0560596241088789,
+      "grad_norm": 0.8234738707542419,
+      "learning_rate": 3.759330008124905e-06,
+      "loss": 0.1066,
+      "step": 6518
+    },
+    {
+      "epoch": 1.0562216461438756,
+      "grad_norm": 0.799088180065155,
+      "learning_rate": 3.7589522193486476e-06,
+      "loss": 0.0963,
+      "step": 6519
+    },
+    {
+      "epoch": 1.0563836681788723,
+      "grad_norm": 0.850976288318634,
+      "learning_rate": 3.7585743920514985e-06,
+      "loss": 0.1049,
+      "step": 6520
+    },
+    {
+      "epoch": 1.056545690213869,
+      "grad_norm": 0.8160610198974609,
+      "learning_rate": 3.7581965262450193e-06,
+      "loss": 0.1048,
+      "step": 6521
+    },
+    {
+      "epoch": 1.056707712248866,
+      "grad_norm": 0.850570023059845,
+      "learning_rate": 3.757818621940771e-06,
+      "loss": 0.1042,
+      "step": 6522
+    },
+    {
+      "epoch": 1.0568697342838627,
+      "grad_norm": 0.7830667495727539,
+      "learning_rate": 3.7574406791503167e-06,
+      "loss": 0.1023,
+      "step": 6523
+    },
+    {
+      "epoch": 1.0570317563188594,
+      "grad_norm": 0.831290066242218,
+      "learning_rate": 3.7570626978852203e-06,
+      "loss": 0.1041,
+      "step": 6524
+    },
+    {
+      "epoch": 1.057193778353856,
+      "grad_norm": 0.9002595543861389,
+      "learning_rate": 3.7566846781570476e-06,
+      "loss": 0.1133,
+      "step": 6525
+    },
+    {
+      "epoch": 1.0573558003888528,
+      "grad_norm": 1.0122647285461426,
+      "learning_rate": 3.7563066199773645e-06,
+      "loss": 0.1241,
+      "step": 6526
+    },
+    {
+      "epoch": 1.0575178224238497,
+      "grad_norm": 0.858625054359436,
+      "learning_rate": 3.75592852335774e-06,
+      "loss": 0.1033,
+      "step": 6527
+    },
+    {
+      "epoch": 1.0576798444588464,
+      "grad_norm": 0.7898378372192383,
+      "learning_rate": 3.7555503883097414e-06,
+      "loss": 0.1033,
+      "step": 6528
+    },
+    {
+      "epoch": 1.0578418664938432,
+      "grad_norm": 0.8522166609764099,
+      "learning_rate": 3.755172214844939e-06,
+      "loss": 0.1046,
+      "step": 6529
+    },
+    {
+      "epoch": 1.0580038885288399,
+      "grad_norm": 0.8669360280036926,
+      "learning_rate": 3.7547940029749054e-06,
+      "loss": 0.1087,
+      "step": 6530
+    },
+    {
+      "epoch": 1.0581659105638366,
+      "grad_norm": 0.8492510914802551,
+      "learning_rate": 3.7544157527112103e-06,
+      "loss": 0.1089,
+      "step": 6531
+    },
+    {
+      "epoch": 1.0583279325988335,
+      "grad_norm": 0.7833178043365479,
+      "learning_rate": 3.75403746406543e-06,
+      "loss": 0.0964,
+      "step": 6532
+    },
+    {
+      "epoch": 1.0584899546338302,
+      "grad_norm": 0.8604576587677002,
+      "learning_rate": 3.7536591370491373e-06,
+      "loss": 0.1123,
+      "step": 6533
+    },
+    {
+      "epoch": 1.058651976668827,
+      "grad_norm": 0.7791811227798462,
+      "learning_rate": 3.7532807716739082e-06,
+      "loss": 0.1048,
+      "step": 6534
+    },
+    {
+      "epoch": 1.0588139987038236,
+      "grad_norm": 0.8196133375167847,
+      "learning_rate": 3.7529023679513217e-06,
+      "loss": 0.109,
+      "step": 6535
+    },
+    {
+      "epoch": 1.0589760207388206,
+      "grad_norm": 0.7942836284637451,
+      "learning_rate": 3.752523925892954e-06,
+      "loss": 0.0982,
+      "step": 6536
+    },
+    {
+      "epoch": 1.0591380427738173,
+      "grad_norm": 0.7868502140045166,
+      "learning_rate": 3.7521454455103857e-06,
+      "loss": 0.1046,
+      "step": 6537
+    },
+    {
+      "epoch": 1.059300064808814,
+      "grad_norm": 0.8380435109138489,
+      "learning_rate": 3.7517669268151967e-06,
+      "loss": 0.1071,
+      "step": 6538
+    },
+    {
+      "epoch": 1.0594620868438107,
+      "grad_norm": 0.6852514743804932,
+      "learning_rate": 3.751388369818969e-06,
+      "loss": 0.0811,
+      "step": 6539
+    },
+    {
+      "epoch": 1.0596241088788074,
+      "grad_norm": 0.7679120302200317,
+      "learning_rate": 3.751009774533285e-06,
+      "loss": 0.0952,
+      "step": 6540
+    },
+    {
+      "epoch": 1.0597861309138044,
+      "grad_norm": 0.8072113394737244,
+      "learning_rate": 3.7506311409697295e-06,
+      "loss": 0.1035,
+      "step": 6541
+    },
+    {
+      "epoch": 1.059948152948801,
+      "grad_norm": 1.01187002658844,
+      "learning_rate": 3.7502524691398877e-06,
+      "loss": 0.131,
+      "step": 6542
+    },
+    {
+      "epoch": 1.0601101749837978,
+      "grad_norm": 0.8866243958473206,
+      "learning_rate": 3.7498737590553465e-06,
+      "loss": 0.1028,
+      "step": 6543
+    },
+    {
+      "epoch": 1.0602721970187945,
+      "grad_norm": 0.9462065696716309,
+      "learning_rate": 3.7494950107276917e-06,
+      "loss": 0.1029,
+      "step": 6544
+    },
+    {
+      "epoch": 1.0604342190537914,
+      "grad_norm": 0.7820582985877991,
+      "learning_rate": 3.749116224168514e-06,
+      "loss": 0.0942,
+      "step": 6545
+    },
+    {
+      "epoch": 1.0605962410887881,
+      "grad_norm": 0.9271798133850098,
+      "learning_rate": 3.7487373993894027e-06,
+      "loss": 0.1145,
+      "step": 6546
+    },
+    {
+      "epoch": 1.0607582631237849,
+      "grad_norm": 0.8751158714294434,
+      "learning_rate": 3.748358536401949e-06,
+      "loss": 0.105,
+      "step": 6547
+    },
+    {
+      "epoch": 1.0609202851587816,
+      "grad_norm": 0.8312302231788635,
+      "learning_rate": 3.7479796352177445e-06,
+      "loss": 0.0998,
+      "step": 6548
+    },
+    {
+      "epoch": 1.0610823071937783,
+      "grad_norm": 0.7504762411117554,
+      "learning_rate": 3.7476006958483835e-06,
+      "loss": 0.0974,
+      "step": 6549
+    },
+    {
+      "epoch": 1.0612443292287752,
+      "grad_norm": 0.8843858242034912,
+      "learning_rate": 3.7472217183054605e-06,
+      "loss": 0.1055,
+      "step": 6550
+    },
+    {
+      "epoch": 1.061406351263772,
+      "grad_norm": 0.9589284658432007,
+      "learning_rate": 3.7468427026005705e-06,
+      "loss": 0.1222,
+      "step": 6551
+    },
+    {
+      "epoch": 1.0615683732987686,
+      "grad_norm": 0.8354429602622986,
+      "learning_rate": 3.7464636487453122e-06,
+      "loss": 0.1088,
+      "step": 6552
+    },
+    {
+      "epoch": 1.0617303953337653,
+      "grad_norm": 0.7289818525314331,
+      "learning_rate": 3.7460845567512817e-06,
+      "loss": 0.0927,
+      "step": 6553
+    },
+    {
+      "epoch": 1.061892417368762,
+      "grad_norm": 0.8219195604324341,
+      "learning_rate": 3.74570542663008e-06,
+      "loss": 0.1049,
+      "step": 6554
+    },
+    {
+      "epoch": 1.062054439403759,
+      "grad_norm": 0.8895604610443115,
+      "learning_rate": 3.745326258393306e-06,
+      "loss": 0.1069,
+      "step": 6555
+    },
+    {
+      "epoch": 1.0622164614387557,
+      "grad_norm": 0.8279933929443359,
+      "learning_rate": 3.744947052052562e-06,
+      "loss": 0.1081,
+      "step": 6556
+    },
+    {
+      "epoch": 1.0623784834737524,
+      "grad_norm": 0.7519248127937317,
+      "learning_rate": 3.744567807619451e-06,
+      "loss": 0.0904,
+      "step": 6557
+    },
+    {
+      "epoch": 1.0625405055087491,
+      "grad_norm": 0.7404701113700867,
+      "learning_rate": 3.7441885251055774e-06,
+      "loss": 0.0901,
+      "step": 6558
+    },
+    {
+      "epoch": 1.0627025275437458,
+      "grad_norm": 0.8219286203384399,
+      "learning_rate": 3.743809204522546e-06,
+      "loss": 0.1059,
+      "step": 6559
+    },
+    {
+      "epoch": 1.0628645495787428,
+      "grad_norm": 0.8403288125991821,
+      "learning_rate": 3.7434298458819622e-06,
+      "loss": 0.1126,
+      "step": 6560
+    },
+    {
+      "epoch": 1.0630265716137395,
+      "grad_norm": 0.742680549621582,
+      "learning_rate": 3.743050449195435e-06,
+      "loss": 0.0932,
+      "step": 6561
+    },
+    {
+      "epoch": 1.0631885936487362,
+      "grad_norm": 0.8612416386604309,
+      "learning_rate": 3.7426710144745717e-06,
+      "loss": 0.1066,
+      "step": 6562
+    },
+    {
+      "epoch": 1.063350615683733,
+      "grad_norm": 0.8878268599510193,
+      "learning_rate": 3.7422915417309825e-06,
+      "loss": 0.1157,
+      "step": 6563
+    },
+    {
+      "epoch": 1.0635126377187298,
+      "grad_norm": 0.8379852771759033,
+      "learning_rate": 3.7419120309762787e-06,
+      "loss": 0.1026,
+      "step": 6564
+    },
+    {
+      "epoch": 1.0636746597537265,
+      "grad_norm": 0.8249161839485168,
+      "learning_rate": 3.7415324822220717e-06,
+      "loss": 0.1052,
+      "step": 6565
+    },
+    {
+      "epoch": 1.0638366817887233,
+      "grad_norm": 0.8621309399604797,
+      "learning_rate": 3.7411528954799752e-06,
+      "loss": 0.1006,
+      "step": 6566
+    },
+    {
+      "epoch": 1.06399870382372,
+      "grad_norm": 0.860681414604187,
+      "learning_rate": 3.740773270761604e-06,
+      "loss": 0.103,
+      "step": 6567
+    },
+    {
+      "epoch": 1.0641607258587167,
+      "grad_norm": 0.995371401309967,
+      "learning_rate": 3.740393608078573e-06,
+      "loss": 0.1238,
+      "step": 6568
+    },
+    {
+      "epoch": 1.0643227478937136,
+      "grad_norm": 0.8341780304908752,
+      "learning_rate": 3.7400139074424997e-06,
+      "loss": 0.1076,
+      "step": 6569
+    },
+    {
+      "epoch": 1.0644847699287103,
+      "grad_norm": 0.8263168931007385,
+      "learning_rate": 3.739634168865001e-06,
+      "loss": 0.1026,
+      "step": 6570
+    },
+    {
+      "epoch": 1.064646791963707,
+      "grad_norm": 0.959525465965271,
+      "learning_rate": 3.7392543923576974e-06,
+      "loss": 0.1143,
+      "step": 6571
+    },
+    {
+      "epoch": 1.0648088139987038,
+      "grad_norm": 0.8506279587745667,
+      "learning_rate": 3.738874577932208e-06,
+      "loss": 0.1033,
+      "step": 6572
+    },
+    {
+      "epoch": 1.0649708360337007,
+      "grad_norm": 0.8389852046966553,
+      "learning_rate": 3.7384947256001534e-06,
+      "loss": 0.1043,
+      "step": 6573
+    },
+    {
+      "epoch": 1.0651328580686974,
+      "grad_norm": 0.7704646587371826,
+      "learning_rate": 3.738114835373159e-06,
+      "loss": 0.0953,
+      "step": 6574
+    },
+    {
+      "epoch": 1.065294880103694,
+      "grad_norm": 0.8785123229026794,
+      "learning_rate": 3.7377349072628457e-06,
+      "loss": 0.1146,
+      "step": 6575
+    },
+    {
+      "epoch": 1.0654569021386908,
+      "grad_norm": 0.7751638889312744,
+      "learning_rate": 3.73735494128084e-06,
+      "loss": 0.0961,
+      "step": 6576
+    },
+    {
+      "epoch": 1.0656189241736875,
+      "grad_norm": 0.8245144486427307,
+      "learning_rate": 3.7369749374387677e-06,
+      "loss": 0.108,
+      "step": 6577
+    },
+    {
+      "epoch": 1.0657809462086845,
+      "grad_norm": 0.8200885653495789,
+      "learning_rate": 3.736594895748255e-06,
+      "loss": 0.0974,
+      "step": 6578
+    },
+    {
+      "epoch": 1.0659429682436812,
+      "grad_norm": 0.8444353342056274,
+      "learning_rate": 3.7362148162209315e-06,
+      "loss": 0.0993,
+      "step": 6579
+    },
+    {
+      "epoch": 1.0661049902786779,
+      "grad_norm": 0.822864830493927,
+      "learning_rate": 3.7358346988684258e-06,
+      "loss": 0.1042,
+      "step": 6580
+    },
+    {
+      "epoch": 1.0662670123136746,
+      "grad_norm": 0.8917192220687866,
+      "learning_rate": 3.73545454370237e-06,
+      "loss": 0.1073,
+      "step": 6581
+    },
+    {
+      "epoch": 1.0664290343486713,
+      "grad_norm": 1.0070949792861938,
+      "learning_rate": 3.735074350734393e-06,
+      "loss": 0.1275,
+      "step": 6582
+    },
+    {
+      "epoch": 1.0665910563836682,
+      "grad_norm": 0.8910343647003174,
+      "learning_rate": 3.7346941199761317e-06,
+      "loss": 0.1139,
+      "step": 6583
+    },
+    {
+      "epoch": 1.066753078418665,
+      "grad_norm": 0.8535321950912476,
+      "learning_rate": 3.734313851439217e-06,
+      "loss": 0.103,
+      "step": 6584
+    },
+    {
+      "epoch": 1.0669151004536617,
+      "grad_norm": 0.8377382159233093,
+      "learning_rate": 3.7339335451352864e-06,
+      "loss": 0.1053,
+      "step": 6585
+    },
+    {
+      "epoch": 1.0670771224886584,
+      "grad_norm": 0.8273501992225647,
+      "learning_rate": 3.7335532010759747e-06,
+      "loss": 0.1034,
+      "step": 6586
+    },
+    {
+      "epoch": 1.0672391445236553,
+      "grad_norm": 0.9124737977981567,
+      "learning_rate": 3.73317281927292e-06,
+      "loss": 0.1059,
+      "step": 6587
+    },
+    {
+      "epoch": 1.067401166558652,
+      "grad_norm": 0.8772713541984558,
+      "learning_rate": 3.732792399737761e-06,
+      "loss": 0.1118,
+      "step": 6588
+    },
+    {
+      "epoch": 1.0675631885936487,
+      "grad_norm": 0.7443407773971558,
+      "learning_rate": 3.7324119424821387e-06,
+      "loss": 0.0956,
+      "step": 6589
+    },
+    {
+      "epoch": 1.0677252106286454,
+      "grad_norm": 0.7781869173049927,
+      "learning_rate": 3.7320314475176933e-06,
+      "loss": 0.0921,
+      "step": 6590
+    },
+    {
+      "epoch": 1.0678872326636422,
+      "grad_norm": 1.039441466331482,
+      "learning_rate": 3.7316509148560664e-06,
+      "loss": 0.1076,
+      "step": 6591
+    },
+    {
+      "epoch": 1.068049254698639,
+      "grad_norm": 0.7879117131233215,
+      "learning_rate": 3.731270344508903e-06,
+      "loss": 0.0951,
+      "step": 6592
+    },
+    {
+      "epoch": 1.0682112767336358,
+      "grad_norm": 0.8069095015525818,
+      "learning_rate": 3.730889736487846e-06,
+      "loss": 0.103,
+      "step": 6593
+    },
+    {
+      "epoch": 1.0683732987686325,
+      "grad_norm": 0.7860719561576843,
+      "learning_rate": 3.7305090908045422e-06,
+      "loss": 0.0989,
+      "step": 6594
+    },
+    {
+      "epoch": 1.0685353208036292,
+      "grad_norm": 0.7485154271125793,
+      "learning_rate": 3.7301284074706372e-06,
+      "loss": 0.0916,
+      "step": 6595
+    },
+    {
+      "epoch": 1.0686973428386262,
+      "grad_norm": 0.8204521536827087,
+      "learning_rate": 3.7297476864977805e-06,
+      "loss": 0.1062,
+      "step": 6596
+    },
+    {
+      "epoch": 1.0688593648736229,
+      "grad_norm": 0.7526237964630127,
+      "learning_rate": 3.72936692789762e-06,
+      "loss": 0.0921,
+      "step": 6597
+    },
+    {
+      "epoch": 1.0690213869086196,
+      "grad_norm": 0.7098335027694702,
+      "learning_rate": 3.7289861316818077e-06,
+      "loss": 0.0926,
+      "step": 6598
+    },
+    {
+      "epoch": 1.0691834089436163,
+      "grad_norm": 0.8343385457992554,
+      "learning_rate": 3.7286052978619926e-06,
+      "loss": 0.1039,
+      "step": 6599
+    },
+    {
+      "epoch": 1.069345430978613,
+      "grad_norm": 0.8440389633178711,
+      "learning_rate": 3.728224426449829e-06,
+      "loss": 0.1092,
+      "step": 6600
+    },
+    {
+      "epoch": 1.06950745301361,
+      "grad_norm": 0.9151831865310669,
+      "learning_rate": 3.72784351745697e-06,
+      "loss": 0.1205,
+      "step": 6601
+    },
+    {
+      "epoch": 1.0696694750486067,
+      "grad_norm": 0.8347547650337219,
+      "learning_rate": 3.7274625708950706e-06,
+      "loss": 0.0992,
+      "step": 6602
+    },
+    {
+      "epoch": 1.0698314970836034,
+      "grad_norm": 0.9815458655357361,
+      "learning_rate": 3.727081586775787e-06,
+      "loss": 0.1116,
+      "step": 6603
+    },
+    {
+      "epoch": 1.0699935191186,
+      "grad_norm": 0.7796982526779175,
+      "learning_rate": 3.7267005651107763e-06,
+      "loss": 0.1061,
+      "step": 6604
+    },
+    {
+      "epoch": 1.0701555411535968,
+      "grad_norm": 0.7754164338111877,
+      "learning_rate": 3.7263195059116973e-06,
+      "loss": 0.1041,
+      "step": 6605
+    },
+    {
+      "epoch": 1.0703175631885937,
+      "grad_norm": 0.8905292749404907,
+      "learning_rate": 3.7259384091902085e-06,
+      "loss": 0.1112,
+      "step": 6606
+    },
+    {
+      "epoch": 1.0704795852235904,
+      "grad_norm": 0.7484425902366638,
+      "learning_rate": 3.7255572749579716e-06,
+      "loss": 0.0919,
+      "step": 6607
+    },
+    {
+      "epoch": 1.0706416072585871,
+      "grad_norm": 0.863379955291748,
+      "learning_rate": 3.7251761032266475e-06,
+      "loss": 0.1023,
+      "step": 6608
+    },
+    {
+      "epoch": 1.0708036292935839,
+      "grad_norm": 0.9683666229248047,
+      "learning_rate": 3.7247948940078996e-06,
+      "loss": 0.1008,
+      "step": 6609
+    },
+    {
+      "epoch": 1.0709656513285806,
+      "grad_norm": 0.9362031817436218,
+      "learning_rate": 3.7244136473133924e-06,
+      "loss": 0.1129,
+      "step": 6610
+    },
+    {
+      "epoch": 1.0711276733635775,
+      "grad_norm": 0.8648001551628113,
+      "learning_rate": 3.72403236315479e-06,
+      "loss": 0.1067,
+      "step": 6611
+    },
+    {
+      "epoch": 1.0712896953985742,
+      "grad_norm": 0.8078235387802124,
+      "learning_rate": 3.7236510415437598e-06,
+      "loss": 0.1048,
+      "step": 6612
+    },
+    {
+      "epoch": 1.071451717433571,
+      "grad_norm": 0.7274784445762634,
+      "learning_rate": 3.7232696824919685e-06,
+      "loss": 0.0971,
+      "step": 6613
+    },
+    {
+      "epoch": 1.0716137394685676,
+      "grad_norm": 0.820236325263977,
+      "learning_rate": 3.7228882860110856e-06,
+      "loss": 0.0984,
+      "step": 6614
+    },
+    {
+      "epoch": 1.0717757615035646,
+      "grad_norm": 0.8109707832336426,
+      "learning_rate": 3.7225068521127793e-06,
+      "loss": 0.1016,
+      "step": 6615
+    },
+    {
+      "epoch": 1.0719377835385613,
+      "grad_norm": 0.8222618103027344,
+      "learning_rate": 3.7221253808087234e-06,
+      "loss": 0.1114,
+      "step": 6616
+    },
+    {
+      "epoch": 1.072099805573558,
+      "grad_norm": 0.7346887588500977,
+      "learning_rate": 3.7217438721105876e-06,
+      "loss": 0.0923,
+      "step": 6617
+    },
+    {
+      "epoch": 1.0722618276085547,
+      "grad_norm": 0.7791684865951538,
+      "learning_rate": 3.721362326030046e-06,
+      "loss": 0.1004,
+      "step": 6618
+    },
+    {
+      "epoch": 1.0724238496435514,
+      "grad_norm": 0.8646988868713379,
+      "learning_rate": 3.7209807425787724e-06,
+      "loss": 0.1047,
+      "step": 6619
+    },
+    {
+      "epoch": 1.0725858716785484,
+      "grad_norm": 0.7894246578216553,
+      "learning_rate": 3.720599121768443e-06,
+      "loss": 0.0985,
+      "step": 6620
+    },
+    {
+      "epoch": 1.072747893713545,
+      "grad_norm": 0.9039403796195984,
+      "learning_rate": 3.720217463610735e-06,
+      "loss": 0.1114,
+      "step": 6621
+    },
+    {
+      "epoch": 1.0729099157485418,
+      "grad_norm": 0.8473367094993591,
+      "learning_rate": 3.7198357681173247e-06,
+      "loss": 0.1048,
+      "step": 6622
+    },
+    {
+      "epoch": 1.0730719377835385,
+      "grad_norm": 0.8533846735954285,
+      "learning_rate": 3.719454035299892e-06,
+      "loss": 0.1001,
+      "step": 6623
+    },
+    {
+      "epoch": 1.0732339598185354,
+      "grad_norm": 0.8822036981582642,
+      "learning_rate": 3.7190722651701166e-06,
+      "loss": 0.1085,
+      "step": 6624
+    },
+    {
+      "epoch": 1.0733959818535321,
+      "grad_norm": 0.74126797914505,
+      "learning_rate": 3.7186904577396805e-06,
+      "loss": 0.0936,
+      "step": 6625
+    },
+    {
+      "epoch": 1.0735580038885288,
+      "grad_norm": 1.021998643875122,
+      "learning_rate": 3.718308613020265e-06,
+      "loss": 0.1299,
+      "step": 6626
+    },
+    {
+      "epoch": 1.0737200259235256,
+      "grad_norm": 0.7277312278747559,
+      "learning_rate": 3.7179267310235544e-06,
+      "loss": 0.0948,
+      "step": 6627
+    },
+    {
+      "epoch": 1.0738820479585223,
+      "grad_norm": 0.8704293966293335,
+      "learning_rate": 3.717544811761233e-06,
+      "loss": 0.1089,
+      "step": 6628
+    },
+    {
+      "epoch": 1.0740440699935192,
+      "grad_norm": 0.7847870588302612,
+      "learning_rate": 3.717162855244988e-06,
+      "loss": 0.0992,
+      "step": 6629
+    },
+    {
+      "epoch": 1.074206092028516,
+      "grad_norm": 0.8332440853118896,
+      "learning_rate": 3.716780861486503e-06,
+      "loss": 0.1022,
+      "step": 6630
+    },
+    {
+      "epoch": 1.0743681140635126,
+      "grad_norm": 0.8781189322471619,
+      "learning_rate": 3.7163988304974704e-06,
+      "loss": 0.1105,
+      "step": 6631
+    },
+    {
+      "epoch": 1.0745301360985093,
+      "grad_norm": 0.9302778244018555,
+      "learning_rate": 3.716016762289576e-06,
+      "loss": 0.1082,
+      "step": 6632
+    },
+    {
+      "epoch": 1.074692158133506,
+      "grad_norm": 0.7837849855422974,
+      "learning_rate": 3.715634656874511e-06,
+      "loss": 0.096,
+      "step": 6633
+    },
+    {
+      "epoch": 1.074854180168503,
+      "grad_norm": 0.7850882411003113,
+      "learning_rate": 3.7152525142639682e-06,
+      "loss": 0.1039,
+      "step": 6634
+    },
+    {
+      "epoch": 1.0750162022034997,
+      "grad_norm": 0.8564865589141846,
+      "learning_rate": 3.7148703344696386e-06,
+      "loss": 0.1117,
+      "step": 6635
+    },
+    {
+      "epoch": 1.0751782242384964,
+      "grad_norm": 0.9498354196548462,
+      "learning_rate": 3.7144881175032178e-06,
+      "loss": 0.1151,
+      "step": 6636
+    },
+    {
+      "epoch": 1.0753402462734931,
+      "grad_norm": 0.949449896812439,
+      "learning_rate": 3.714105863376398e-06,
+      "loss": 0.1191,
+      "step": 6637
+    },
+    {
+      "epoch": 1.07550226830849,
+      "grad_norm": 0.8418267369270325,
+      "learning_rate": 3.713723572100878e-06,
+      "loss": 0.1067,
+      "step": 6638
+    },
+    {
+      "epoch": 1.0756642903434868,
+      "grad_norm": 0.7026755809783936,
+      "learning_rate": 3.713341243688353e-06,
+      "loss": 0.091,
+      "step": 6639
+    },
+    {
+      "epoch": 1.0758263123784835,
+      "grad_norm": 0.7640417218208313,
+      "learning_rate": 3.7129588781505232e-06,
+      "loss": 0.0962,
+      "step": 6640
+    },
+    {
+      "epoch": 1.0759883344134802,
+      "grad_norm": 0.7686159610748291,
+      "learning_rate": 3.7125764754990864e-06,
+      "loss": 0.0942,
+      "step": 6641
+    },
+    {
+      "epoch": 1.076150356448477,
+      "grad_norm": 0.9323800206184387,
+      "learning_rate": 3.7121940357457438e-06,
+      "loss": 0.1188,
+      "step": 6642
+    },
+    {
+      "epoch": 1.0763123784834738,
+      "grad_norm": 0.801908016204834,
+      "learning_rate": 3.7118115589021976e-06,
+      "loss": 0.0964,
+      "step": 6643
+    },
+    {
+      "epoch": 1.0764744005184705,
+      "grad_norm": 0.7992369532585144,
+      "learning_rate": 3.7114290449801493e-06,
+      "loss": 0.0955,
+      "step": 6644
+    },
+    {
+      "epoch": 1.0766364225534673,
+      "grad_norm": 0.8944490551948547,
+      "learning_rate": 3.711046493991305e-06,
+      "loss": 0.1083,
+      "step": 6645
+    },
+    {
+      "epoch": 1.076798444588464,
+      "grad_norm": 0.8392427563667297,
+      "learning_rate": 3.7106639059473675e-06,
+      "loss": 0.1101,
+      "step": 6646
+    },
+    {
+      "epoch": 1.076960466623461,
+      "grad_norm": 0.8066583871841431,
+      "learning_rate": 3.7102812808600452e-06,
+      "loss": 0.0912,
+      "step": 6647
+    },
+    {
+      "epoch": 1.0771224886584576,
+      "grad_norm": 0.8310062885284424,
+      "learning_rate": 3.7098986187410447e-06,
+      "loss": 0.1061,
+      "step": 6648
+    },
+    {
+      "epoch": 1.0772845106934543,
+      "grad_norm": 0.9091915488243103,
+      "learning_rate": 3.7095159196020736e-06,
+      "loss": 0.1213,
+      "step": 6649
+    },
+    {
+      "epoch": 1.077446532728451,
+      "grad_norm": 0.8808692693710327,
+      "learning_rate": 3.7091331834548427e-06,
+      "loss": 0.1142,
+      "step": 6650
+    },
+    {
+      "epoch": 1.0776085547634477,
+      "grad_norm": 0.8686168789863586,
+      "learning_rate": 3.708750410311062e-06,
+      "loss": 0.1112,
+      "step": 6651
+    },
+    {
+      "epoch": 1.0777705767984447,
+      "grad_norm": 0.8946408629417419,
+      "learning_rate": 3.7083676001824443e-06,
+      "loss": 0.1145,
+      "step": 6652
+    },
+    {
+      "epoch": 1.0779325988334414,
+      "grad_norm": 0.8658775091171265,
+      "learning_rate": 3.7079847530807023e-06,
+      "loss": 0.1079,
+      "step": 6653
+    },
+    {
+      "epoch": 1.078094620868438,
+      "grad_norm": 0.8047525882720947,
+      "learning_rate": 3.70760186901755e-06,
+      "loss": 0.0999,
+      "step": 6654
+    },
+    {
+      "epoch": 1.0782566429034348,
+      "grad_norm": 0.7709795236587524,
+      "learning_rate": 3.7072189480047027e-06,
+      "loss": 0.0851,
+      "step": 6655
+    },
+    {
+      "epoch": 1.0784186649384315,
+      "grad_norm": 0.9803235530853271,
+      "learning_rate": 3.706835990053877e-06,
+      "loss": 0.1154,
+      "step": 6656
+    },
+    {
+      "epoch": 1.0785806869734285,
+      "grad_norm": 0.8865209817886353,
+      "learning_rate": 3.7064529951767905e-06,
+      "loss": 0.1043,
+      "step": 6657
+    },
+    {
+      "epoch": 1.0787427090084252,
+      "grad_norm": 0.8521233797073364,
+      "learning_rate": 3.7060699633851615e-06,
+      "loss": 0.1057,
+      "step": 6658
+    },
+    {
+      "epoch": 1.0789047310434219,
+      "grad_norm": 0.8935887217521667,
+      "learning_rate": 3.705686894690711e-06,
+      "loss": 0.1043,
+      "step": 6659
+    },
+    {
+      "epoch": 1.0790667530784186,
+      "grad_norm": 0.7139661908149719,
+      "learning_rate": 3.7053037891051596e-06,
+      "loss": 0.0868,
+      "step": 6660
+    },
+    {
+      "epoch": 1.0792287751134155,
+      "grad_norm": 0.8877597451210022,
+      "learning_rate": 3.7049206466402278e-06,
+      "loss": 0.1123,
+      "step": 6661
+    },
+    {
+      "epoch": 1.0793907971484122,
+      "grad_norm": 0.8162228465080261,
+      "learning_rate": 3.704537467307641e-06,
+      "loss": 0.1032,
+      "step": 6662
+    },
+    {
+      "epoch": 1.079552819183409,
+      "grad_norm": 0.8396819829940796,
+      "learning_rate": 3.704154251119122e-06,
+      "loss": 0.1027,
+      "step": 6663
+    },
+    {
+      "epoch": 1.0797148412184057,
+      "grad_norm": 0.8406090140342712,
+      "learning_rate": 3.7037709980863974e-06,
+      "loss": 0.0978,
+      "step": 6664
+    },
+    {
+      "epoch": 1.0798768632534024,
+      "grad_norm": 0.9074726700782776,
+      "learning_rate": 3.703387708221193e-06,
+      "loss": 0.1035,
+      "step": 6665
+    },
+    {
+      "epoch": 1.0800388852883993,
+      "grad_norm": 0.9996922612190247,
+      "learning_rate": 3.703004381535237e-06,
+      "loss": 0.1203,
+      "step": 6666
+    },
+    {
+      "epoch": 1.080200907323396,
+      "grad_norm": 0.8596282005310059,
+      "learning_rate": 3.7026210180402588e-06,
+      "loss": 0.1132,
+      "step": 6667
+    },
+    {
+      "epoch": 1.0803629293583927,
+      "grad_norm": 0.7845479249954224,
+      "learning_rate": 3.7022376177479863e-06,
+      "loss": 0.0991,
+      "step": 6668
+    },
+    {
+      "epoch": 1.0805249513933894,
+      "grad_norm": 0.8344932794570923,
+      "learning_rate": 3.701854180670153e-06,
+      "loss": 0.1019,
+      "step": 6669
+    },
+    {
+      "epoch": 1.0806869734283864,
+      "grad_norm": 0.8211415410041809,
+      "learning_rate": 3.7014707068184895e-06,
+      "loss": 0.1064,
+      "step": 6670
+    },
+    {
+      "epoch": 1.080848995463383,
+      "grad_norm": 0.7003439664840698,
+      "learning_rate": 3.7010871962047314e-06,
+      "loss": 0.0901,
+      "step": 6671
+    },
+    {
+      "epoch": 1.0810110174983798,
+      "grad_norm": 0.8313621282577515,
+      "learning_rate": 3.70070364884061e-06,
+      "loss": 0.0949,
+      "step": 6672
+    },
+    {
+      "epoch": 1.0811730395333765,
+      "grad_norm": 0.9178413152694702,
+      "learning_rate": 3.7003200647378634e-06,
+      "loss": 0.1111,
+      "step": 6673
+    },
+    {
+      "epoch": 1.0813350615683732,
+      "grad_norm": 0.8344830870628357,
+      "learning_rate": 3.6999364439082274e-06,
+      "loss": 0.1049,
+      "step": 6674
+    },
+    {
+      "epoch": 1.0814970836033702,
+      "grad_norm": 0.8041746616363525,
+      "learning_rate": 3.69955278636344e-06,
+      "loss": 0.1088,
+      "step": 6675
+    },
+    {
+      "epoch": 1.0816591056383669,
+      "grad_norm": 0.8075587153434753,
+      "learning_rate": 3.6991690921152407e-06,
+      "loss": 0.1055,
+      "step": 6676
+    },
+    {
+      "epoch": 1.0818211276733636,
+      "grad_norm": 0.729253351688385,
+      "learning_rate": 3.6987853611753686e-06,
+      "loss": 0.0843,
+      "step": 6677
+    },
+    {
+      "epoch": 1.0819831497083603,
+      "grad_norm": 0.9236080646514893,
+      "learning_rate": 3.698401593555565e-06,
+      "loss": 0.117,
+      "step": 6678
+    },
+    {
+      "epoch": 1.082145171743357,
+      "grad_norm": 0.9156818985939026,
+      "learning_rate": 3.6980177892675735e-06,
+      "loss": 0.1183,
+      "step": 6679
+    },
+    {
+      "epoch": 1.082307193778354,
+      "grad_norm": 0.8869015574455261,
+      "learning_rate": 3.697633948323136e-06,
+      "loss": 0.103,
+      "step": 6680
+    },
+    {
+      "epoch": 1.0824692158133506,
+      "grad_norm": 0.7443458437919617,
+      "learning_rate": 3.6972500707339986e-06,
+      "loss": 0.0958,
+      "step": 6681
+    },
+    {
+      "epoch": 1.0826312378483474,
+      "grad_norm": 0.8404899835586548,
+      "learning_rate": 3.6968661565119062e-06,
+      "loss": 0.1051,
+      "step": 6682
+    },
+    {
+      "epoch": 1.082793259883344,
+      "grad_norm": 0.8908175826072693,
+      "learning_rate": 3.6964822056686057e-06,
+      "loss": 0.108,
+      "step": 6683
+    },
+    {
+      "epoch": 1.0829552819183408,
+      "grad_norm": 0.8125652074813843,
+      "learning_rate": 3.6960982182158458e-06,
+      "loss": 0.1058,
+      "step": 6684
+    },
+    {
+      "epoch": 1.0831173039533377,
+      "grad_norm": 1.0027449131011963,
+      "learning_rate": 3.695714194165374e-06,
+      "loss": 0.1219,
+      "step": 6685
+    },
+    {
+      "epoch": 1.0832793259883344,
+      "grad_norm": 0.8607674837112427,
+      "learning_rate": 3.6953301335289415e-06,
+      "loss": 0.1124,
+      "step": 6686
+    },
+    {
+      "epoch": 1.0834413480233311,
+      "grad_norm": 0.8296065926551819,
+      "learning_rate": 3.6949460363183e-06,
+      "loss": 0.114,
+      "step": 6687
+    },
+    {
+      "epoch": 1.0836033700583279,
+      "grad_norm": 0.923507571220398,
+      "learning_rate": 3.6945619025452006e-06,
+      "loss": 0.1124,
+      "step": 6688
+    },
+    {
+      "epoch": 1.0837653920933248,
+      "grad_norm": 0.8684310913085938,
+      "learning_rate": 3.694177732221399e-06,
+      "loss": 0.1112,
+      "step": 6689
+    },
+    {
+      "epoch": 1.0839274141283215,
+      "grad_norm": 0.7922837138175964,
+      "learning_rate": 3.6937935253586475e-06,
+      "loss": 0.1036,
+      "step": 6690
+    },
+    {
+      "epoch": 1.0840894361633182,
+      "grad_norm": 0.7880176305770874,
+      "learning_rate": 3.693409281968704e-06,
+      "loss": 0.0899,
+      "step": 6691
+    },
+    {
+      "epoch": 1.084251458198315,
+      "grad_norm": 0.8277053833007812,
+      "learning_rate": 3.6930250020633237e-06,
+      "loss": 0.1015,
+      "step": 6692
+    },
+    {
+      "epoch": 1.0844134802333116,
+      "grad_norm": 0.8327934145927429,
+      "learning_rate": 3.692640685654266e-06,
+      "loss": 0.1002,
+      "step": 6693
+    },
+    {
+      "epoch": 1.0845755022683086,
+      "grad_norm": 0.8471828699111938,
+      "learning_rate": 3.692256332753289e-06,
+      "loss": 0.1098,
+      "step": 6694
+    },
+    {
+      "epoch": 1.0847375243033053,
+      "grad_norm": 0.8789288401603699,
+      "learning_rate": 3.691871943372154e-06,
+      "loss": 0.1028,
+      "step": 6695
+    },
+    {
+      "epoch": 1.084899546338302,
+      "grad_norm": 0.8185938000679016,
+      "learning_rate": 3.691487517522621e-06,
+      "loss": 0.1081,
+      "step": 6696
+    },
+    {
+      "epoch": 1.0850615683732987,
+      "grad_norm": 0.8919969797134399,
+      "learning_rate": 3.691103055216454e-06,
+      "loss": 0.1175,
+      "step": 6697
+    },
+    {
+      "epoch": 1.0852235904082956,
+      "grad_norm": 0.8887705206871033,
+      "learning_rate": 3.690718556465416e-06,
+      "loss": 0.1059,
+      "step": 6698
+    },
+    {
+      "epoch": 1.0853856124432923,
+      "grad_norm": 0.8780819773674011,
+      "learning_rate": 3.690334021281271e-06,
+      "loss": 0.0972,
+      "step": 6699
+    },
+    {
+      "epoch": 1.085547634478289,
+      "grad_norm": 0.9211699962615967,
+      "learning_rate": 3.689949449675786e-06,
+      "loss": 0.1204,
+      "step": 6700
+    },
+    {
+      "epoch": 1.0857096565132858,
+      "grad_norm": 0.8538796901702881,
+      "learning_rate": 3.6895648416607273e-06,
+      "loss": 0.1136,
+      "step": 6701
+    },
+    {
+      "epoch": 1.0858716785482825,
+      "grad_norm": 0.9533454775810242,
+      "learning_rate": 3.689180197247863e-06,
+      "loss": 0.1214,
+      "step": 6702
+    },
+    {
+      "epoch": 1.0860337005832794,
+      "grad_norm": 0.7428584098815918,
+      "learning_rate": 3.6887955164489626e-06,
+      "loss": 0.0825,
+      "step": 6703
+    },
+    {
+      "epoch": 1.0861957226182761,
+      "grad_norm": 0.8387787342071533,
+      "learning_rate": 3.688410799275796e-06,
+      "loss": 0.105,
+      "step": 6704
+    },
+    {
+      "epoch": 1.0863577446532728,
+      "grad_norm": 0.9467548131942749,
+      "learning_rate": 3.6880260457401353e-06,
+      "loss": 0.1085,
+      "step": 6705
+    },
+    {
+      "epoch": 1.0865197666882696,
+      "grad_norm": 0.88310307264328,
+      "learning_rate": 3.6876412558537524e-06,
+      "loss": 0.1143,
+      "step": 6706
+    },
+    {
+      "epoch": 1.0866817887232663,
+      "grad_norm": 0.8482463955879211,
+      "learning_rate": 3.6872564296284214e-06,
+      "loss": 0.1039,
+      "step": 6707
+    },
+    {
+      "epoch": 1.0868438107582632,
+      "grad_norm": 0.8076966404914856,
+      "learning_rate": 3.686871567075916e-06,
+      "loss": 0.1022,
+      "step": 6708
+    },
+    {
+      "epoch": 1.08700583279326,
+      "grad_norm": 0.8349097967147827,
+      "learning_rate": 3.686486668208013e-06,
+      "loss": 0.1116,
+      "step": 6709
+    },
+    {
+      "epoch": 1.0871678548282566,
+      "grad_norm": 0.8799702525138855,
+      "learning_rate": 3.6861017330364897e-06,
+      "loss": 0.1052,
+      "step": 6710
+    },
+    {
+      "epoch": 1.0873298768632533,
+      "grad_norm": 1.1798293590545654,
+      "learning_rate": 3.685716761573123e-06,
+      "loss": 0.1036,
+      "step": 6711
+    },
+    {
+      "epoch": 1.0874918988982503,
+      "grad_norm": 0.907839298248291,
+      "learning_rate": 3.685331753829693e-06,
+      "loss": 0.1081,
+      "step": 6712
+    },
+    {
+      "epoch": 1.087653920933247,
+      "grad_norm": 0.8660617470741272,
+      "learning_rate": 3.68494670981798e-06,
+      "loss": 0.1047,
+      "step": 6713
+    },
+    {
+      "epoch": 1.0878159429682437,
+      "grad_norm": 0.8517418503761292,
+      "learning_rate": 3.684561629549765e-06,
+      "loss": 0.0974,
+      "step": 6714
+    },
+    {
+      "epoch": 1.0879779650032404,
+      "grad_norm": 0.862423300743103,
+      "learning_rate": 3.684176513036831e-06,
+      "loss": 0.1084,
+      "step": 6715
+    },
+    {
+      "epoch": 1.088139987038237,
+      "grad_norm": 0.6887312531471252,
+      "learning_rate": 3.6837913602909615e-06,
+      "loss": 0.0862,
+      "step": 6716
+    },
+    {
+      "epoch": 1.088302009073234,
+      "grad_norm": 0.7741922736167908,
+      "learning_rate": 3.683406171323941e-06,
+      "loss": 0.0962,
+      "step": 6717
+    },
+    {
+      "epoch": 1.0884640311082308,
+      "grad_norm": 0.8818541169166565,
+      "learning_rate": 3.6830209461475554e-06,
+      "loss": 0.1033,
+      "step": 6718
+    },
+    {
+      "epoch": 1.0886260531432275,
+      "grad_norm": 0.8155554533004761,
+      "learning_rate": 3.682635684773591e-06,
+      "loss": 0.1041,
+      "step": 6719
+    },
+    {
+      "epoch": 1.0887880751782242,
+      "grad_norm": 0.86274254322052,
+      "learning_rate": 3.6822503872138377e-06,
+      "loss": 0.1046,
+      "step": 6720
+    },
+    {
+      "epoch": 1.0889500972132211,
+      "grad_norm": 0.8824654817581177,
+      "learning_rate": 3.681865053480082e-06,
+      "loss": 0.1147,
+      "step": 6721
+    },
+    {
+      "epoch": 1.0891121192482178,
+      "grad_norm": 0.7932639122009277,
+      "learning_rate": 3.6814796835841172e-06,
+      "loss": 0.1008,
+      "step": 6722
+    },
+    {
+      "epoch": 1.0892741412832145,
+      "grad_norm": 0.7773560285568237,
+      "learning_rate": 3.681094277537732e-06,
+      "loss": 0.1036,
+      "step": 6723
+    },
+    {
+      "epoch": 1.0894361633182112,
+      "grad_norm": 0.7581130862236023,
+      "learning_rate": 3.6807088353527216e-06,
+      "loss": 0.0872,
+      "step": 6724
+    },
+    {
+      "epoch": 1.089598185353208,
+      "grad_norm": 0.8676707148551941,
+      "learning_rate": 3.680323357040877e-06,
+      "loss": 0.1011,
+      "step": 6725
+    },
+    {
+      "epoch": 1.089760207388205,
+      "grad_norm": 0.8844422101974487,
+      "learning_rate": 3.6799378426139942e-06,
+      "loss": 0.1222,
+      "step": 6726
+    },
+    {
+      "epoch": 1.0899222294232016,
+      "grad_norm": 0.7242690920829773,
+      "learning_rate": 3.679552292083869e-06,
+      "loss": 0.0894,
+      "step": 6727
+    },
+    {
+      "epoch": 1.0900842514581983,
+      "grad_norm": 0.7637506127357483,
+      "learning_rate": 3.679166705462298e-06,
+      "loss": 0.0938,
+      "step": 6728
+    },
+    {
+      "epoch": 1.090246273493195,
+      "grad_norm": 0.9128175973892212,
+      "learning_rate": 3.67878108276108e-06,
+      "loss": 0.1092,
+      "step": 6729
+    },
+    {
+      "epoch": 1.0904082955281917,
+      "grad_norm": 0.7904967069625854,
+      "learning_rate": 3.6783954239920118e-06,
+      "loss": 0.0979,
+      "step": 6730
+    },
+    {
+      "epoch": 1.0905703175631887,
+      "grad_norm": 0.827260434627533,
+      "learning_rate": 3.678009729166897e-06,
+      "loss": 0.1045,
+      "step": 6731
+    },
+    {
+      "epoch": 1.0907323395981854,
+      "grad_norm": 0.7866484522819519,
+      "learning_rate": 3.677623998297534e-06,
+      "loss": 0.0973,
+      "step": 6732
+    },
+    {
+      "epoch": 1.090894361633182,
+      "grad_norm": 0.8426867723464966,
+      "learning_rate": 3.677238231395727e-06,
+      "loss": 0.1104,
+      "step": 6733
+    },
+    {
+      "epoch": 1.0910563836681788,
+      "grad_norm": 0.7677156329154968,
+      "learning_rate": 3.6768524284732794e-06,
+      "loss": 0.0942,
+      "step": 6734
+    },
+    {
+      "epoch": 1.0912184057031755,
+      "grad_norm": 1.0549900531768799,
+      "learning_rate": 3.676466589541995e-06,
+      "loss": 0.1023,
+      "step": 6735
+    },
+    {
+      "epoch": 1.0913804277381725,
+      "grad_norm": 0.7646496295928955,
+      "learning_rate": 3.6760807146136796e-06,
+      "loss": 0.0946,
+      "step": 6736
+    },
+    {
+      "epoch": 1.0915424497731692,
+      "grad_norm": 0.8362929224967957,
+      "learning_rate": 3.6756948037001406e-06,
+      "loss": 0.1068,
+      "step": 6737
+    },
+    {
+      "epoch": 1.0917044718081659,
+      "grad_norm": 0.940573513507843,
+      "learning_rate": 3.675308856813186e-06,
+      "loss": 0.1196,
+      "step": 6738
+    },
+    {
+      "epoch": 1.0918664938431626,
+      "grad_norm": 0.9485230445861816,
+      "learning_rate": 3.674922873964625e-06,
+      "loss": 0.1154,
+      "step": 6739
+    },
+    {
+      "epoch": 1.0920285158781595,
+      "grad_norm": 0.8467913866043091,
+      "learning_rate": 3.6745368551662663e-06,
+      "loss": 0.0991,
+      "step": 6740
+    },
+    {
+      "epoch": 1.0921905379131562,
+      "grad_norm": 0.8293401002883911,
+      "learning_rate": 3.6741508004299227e-06,
+      "loss": 0.1055,
+      "step": 6741
+    },
+    {
+      "epoch": 1.092352559948153,
+      "grad_norm": 0.9706871509552002,
+      "learning_rate": 3.6737647097674056e-06,
+      "loss": 0.1058,
+      "step": 6742
+    },
+    {
+      "epoch": 1.0925145819831497,
+      "grad_norm": 0.9424450993537903,
+      "learning_rate": 3.673378583190529e-06,
+      "loss": 0.1142,
+      "step": 6743
+    },
+    {
+      "epoch": 1.0926766040181464,
+      "grad_norm": 0.8551537990570068,
+      "learning_rate": 3.6729924207111077e-06,
+      "loss": 0.1007,
+      "step": 6744
+    },
+    {
+      "epoch": 1.0928386260531433,
+      "grad_norm": 0.8865105509757996,
+      "learning_rate": 3.6726062223409563e-06,
+      "loss": 0.1095,
+      "step": 6745
+    },
+    {
+      "epoch": 1.09300064808814,
+      "grad_norm": 0.7189201712608337,
+      "learning_rate": 3.6722199880918928e-06,
+      "loss": 0.091,
+      "step": 6746
+    },
+    {
+      "epoch": 1.0931626701231367,
+      "grad_norm": 0.8280571699142456,
+      "learning_rate": 3.6718337179757336e-06,
+      "loss": 0.1077,
+      "step": 6747
+    },
+    {
+      "epoch": 1.0933246921581334,
+      "grad_norm": 0.9577053189277649,
+      "learning_rate": 3.6714474120042993e-06,
+      "loss": 0.1168,
+      "step": 6748
+    },
+    {
+      "epoch": 1.0934867141931304,
+      "grad_norm": 0.8207298517227173,
+      "learning_rate": 3.6710610701894088e-06,
+      "loss": 0.1025,
+      "step": 6749
+    },
+    {
+      "epoch": 1.093648736228127,
+      "grad_norm": 0.78386390209198,
+      "learning_rate": 3.6706746925428833e-06,
+      "loss": 0.1061,
+      "step": 6750
+    },
+    {
+      "epoch": 1.0938107582631238,
+      "grad_norm": 0.799771249294281,
+      "learning_rate": 3.6702882790765453e-06,
+      "loss": 0.0973,
+      "step": 6751
+    },
+    {
+      "epoch": 1.0939727802981205,
+      "grad_norm": 0.824006199836731,
+      "learning_rate": 3.6699018298022173e-06,
+      "loss": 0.1105,
+      "step": 6752
+    },
+    {
+      "epoch": 1.0941348023331172,
+      "grad_norm": 0.8404650092124939,
+      "learning_rate": 3.6695153447317254e-06,
+      "loss": 0.1,
+      "step": 6753
+    },
+    {
+      "epoch": 1.0942968243681142,
+      "grad_norm": 0.7487378120422363,
+      "learning_rate": 3.6691288238768928e-06,
+      "loss": 0.0998,
+      "step": 6754
+    },
+    {
+      "epoch": 1.0944588464031109,
+      "grad_norm": 0.9007918834686279,
+      "learning_rate": 3.6687422672495493e-06,
+      "loss": 0.1164,
+      "step": 6755
+    },
+    {
+      "epoch": 1.0946208684381076,
+      "grad_norm": 0.7542750835418701,
+      "learning_rate": 3.6683556748615196e-06,
+      "loss": 0.0965,
+      "step": 6756
+    },
+    {
+      "epoch": 1.0947828904731043,
+      "grad_norm": 0.8000967502593994,
+      "learning_rate": 3.6679690467246337e-06,
+      "loss": 0.1096,
+      "step": 6757
+    },
+    {
+      "epoch": 1.094944912508101,
+      "grad_norm": 0.7331892251968384,
+      "learning_rate": 3.667582382850721e-06,
+      "loss": 0.0891,
+      "step": 6758
+    },
+    {
+      "epoch": 1.095106934543098,
+      "grad_norm": 0.790254533290863,
+      "learning_rate": 3.6671956832516136e-06,
+      "loss": 0.097,
+      "step": 6759
+    },
+    {
+      "epoch": 1.0952689565780946,
+      "grad_norm": 0.7792706489562988,
+      "learning_rate": 3.6668089479391433e-06,
+      "loss": 0.0989,
+      "step": 6760
+    },
+    {
+      "epoch": 1.0954309786130914,
+      "grad_norm": 0.8442437648773193,
+      "learning_rate": 3.6664221769251414e-06,
+      "loss": 0.1006,
+      "step": 6761
+    },
+    {
+      "epoch": 1.095593000648088,
+      "grad_norm": 0.928298830986023,
+      "learning_rate": 3.666035370221445e-06,
+      "loss": 0.103,
+      "step": 6762
+    },
+    {
+      "epoch": 1.095755022683085,
+      "grad_norm": 0.8222261071205139,
+      "learning_rate": 3.6656485278398866e-06,
+      "loss": 0.1058,
+      "step": 6763
+    },
+    {
+      "epoch": 1.0959170447180817,
+      "grad_norm": 0.7384744882583618,
+      "learning_rate": 3.665261649792305e-06,
+      "loss": 0.0896,
+      "step": 6764
+    },
+    {
+      "epoch": 1.0960790667530784,
+      "grad_norm": 0.836329460144043,
+      "learning_rate": 3.664874736090537e-06,
+      "loss": 0.1059,
+      "step": 6765
+    },
+    {
+      "epoch": 1.0962410887880751,
+      "grad_norm": 1.0209158658981323,
+      "learning_rate": 3.66448778674642e-06,
+      "loss": 0.1111,
+      "step": 6766
+    },
+    {
+      "epoch": 1.0964031108230718,
+      "grad_norm": 0.7318083047866821,
+      "learning_rate": 3.6641008017717957e-06,
+      "loss": 0.0875,
+      "step": 6767
+    },
+    {
+      "epoch": 1.0965651328580688,
+      "grad_norm": 0.7760915160179138,
+      "learning_rate": 3.663713781178504e-06,
+      "loss": 0.0986,
+      "step": 6768
+    },
+    {
+      "epoch": 1.0967271548930655,
+      "grad_norm": 0.7997096180915833,
+      "learning_rate": 3.6633267249783854e-06,
+      "loss": 0.0943,
+      "step": 6769
+    },
+    {
+      "epoch": 1.0968891769280622,
+      "grad_norm": 0.8263211846351624,
+      "learning_rate": 3.6629396331832854e-06,
+      "loss": 0.0919,
+      "step": 6770
+    },
+    {
+      "epoch": 1.097051198963059,
+      "grad_norm": 0.9232663512229919,
+      "learning_rate": 3.6625525058050467e-06,
+      "loss": 0.1063,
+      "step": 6771
+    },
+    {
+      "epoch": 1.0972132209980558,
+      "grad_norm": 0.9478903412818909,
+      "learning_rate": 3.6621653428555144e-06,
+      "loss": 0.1076,
+      "step": 6772
+    },
+    {
+      "epoch": 1.0973752430330526,
+      "grad_norm": 0.8772745132446289,
+      "learning_rate": 3.661778144346535e-06,
+      "loss": 0.1131,
+      "step": 6773
+    },
+    {
+      "epoch": 1.0975372650680493,
+      "grad_norm": 0.8307527899742126,
+      "learning_rate": 3.661390910289956e-06,
+      "loss": 0.1062,
+      "step": 6774
+    },
+    {
+      "epoch": 1.097699287103046,
+      "grad_norm": 0.8453426361083984,
+      "learning_rate": 3.661003640697626e-06,
+      "loss": 0.0974,
+      "step": 6775
+    },
+    {
+      "epoch": 1.0978613091380427,
+      "grad_norm": 1.1416118144989014,
+      "learning_rate": 3.6606163355813935e-06,
+      "loss": 0.1092,
+      "step": 6776
+    },
+    {
+      "epoch": 1.0980233311730396,
+      "grad_norm": 0.8578381538391113,
+      "learning_rate": 3.66022899495311e-06,
+      "loss": 0.1045,
+      "step": 6777
+    },
+    {
+      "epoch": 1.0981853532080363,
+      "grad_norm": 0.7609858512878418,
+      "learning_rate": 3.6598416188246265e-06,
+      "loss": 0.0962,
+      "step": 6778
+    },
+    {
+      "epoch": 1.098347375243033,
+      "grad_norm": 0.8034330606460571,
+      "learning_rate": 3.659454207207798e-06,
+      "loss": 0.1032,
+      "step": 6779
+    },
+    {
+      "epoch": 1.0985093972780298,
+      "grad_norm": 0.9241167306900024,
+      "learning_rate": 3.6590667601144748e-06,
+      "loss": 0.1066,
+      "step": 6780
+    },
+    {
+      "epoch": 1.0986714193130265,
+      "grad_norm": 0.7677343487739563,
+      "learning_rate": 3.6586792775565137e-06,
+      "loss": 0.0913,
+      "step": 6781
+    },
+    {
+      "epoch": 1.0988334413480234,
+      "grad_norm": 1.0235810279846191,
+      "learning_rate": 3.6582917595457718e-06,
+      "loss": 0.1226,
+      "step": 6782
+    },
+    {
+      "epoch": 1.0989954633830201,
+      "grad_norm": 0.7913617491722107,
+      "learning_rate": 3.657904206094104e-06,
+      "loss": 0.1079,
+      "step": 6783
+    },
+    {
+      "epoch": 1.0991574854180168,
+      "grad_norm": 0.7789272665977478,
+      "learning_rate": 3.6575166172133703e-06,
+      "loss": 0.0998,
+      "step": 6784
+    },
+    {
+      "epoch": 1.0993195074530135,
+      "grad_norm": 0.7309193015098572,
+      "learning_rate": 3.657128992915428e-06,
+      "loss": 0.0935,
+      "step": 6785
+    },
+    {
+      "epoch": 1.0994815294880103,
+      "grad_norm": 0.9271590113639832,
+      "learning_rate": 3.6567413332121402e-06,
+      "loss": 0.1065,
+      "step": 6786
+    },
+    {
+      "epoch": 1.0996435515230072,
+      "grad_norm": 0.8196094632148743,
+      "learning_rate": 3.6563536381153663e-06,
+      "loss": 0.1061,
+      "step": 6787
+    },
+    {
+      "epoch": 1.099805573558004,
+      "grad_norm": 0.8200319409370422,
+      "learning_rate": 3.655965907636969e-06,
+      "loss": 0.1075,
+      "step": 6788
+    },
+    {
+      "epoch": 1.0999675955930006,
+      "grad_norm": 0.8080103993415833,
+      "learning_rate": 3.6555781417888126e-06,
+      "loss": 0.106,
+      "step": 6789
+    },
+    {
+      "epoch": 1.1001296176279973,
+      "grad_norm": 0.8300761580467224,
+      "learning_rate": 3.6551903405827615e-06,
+      "loss": 0.111,
+      "step": 6790
+    },
+    {
+      "epoch": 1.1002916396629943,
+      "grad_norm": 0.887389600276947,
+      "learning_rate": 3.6548025040306816e-06,
+      "loss": 0.1125,
+      "step": 6791
+    },
+    {
+      "epoch": 1.100453661697991,
+      "grad_norm": 0.6711512207984924,
+      "learning_rate": 3.6544146321444397e-06,
+      "loss": 0.0879,
+      "step": 6792
+    },
+    {
+      "epoch": 1.1006156837329877,
+      "grad_norm": 0.7568984627723694,
+      "learning_rate": 3.654026724935904e-06,
+      "loss": 0.0932,
+      "step": 6793
+    },
+    {
+      "epoch": 1.1007777057679844,
+      "grad_norm": 0.8204541206359863,
+      "learning_rate": 3.653638782416943e-06,
+      "loss": 0.1042,
+      "step": 6794
+    },
+    {
+      "epoch": 1.100939727802981,
+      "grad_norm": 0.8643372058868408,
+      "learning_rate": 3.6532508045994262e-06,
+      "loss": 0.1071,
+      "step": 6795
+    },
+    {
+      "epoch": 1.101101749837978,
+      "grad_norm": 0.8105908036231995,
+      "learning_rate": 3.6528627914952263e-06,
+      "loss": 0.094,
+      "step": 6796
+    },
+    {
+      "epoch": 1.1012637718729748,
+      "grad_norm": 0.8288483023643494,
+      "learning_rate": 3.6524747431162148e-06,
+      "loss": 0.1059,
+      "step": 6797
+    },
+    {
+      "epoch": 1.1014257939079715,
+      "grad_norm": 0.8761540651321411,
+      "learning_rate": 3.652086659474265e-06,
+      "loss": 0.1127,
+      "step": 6798
+    },
+    {
+      "epoch": 1.1015878159429682,
+      "grad_norm": 0.8513060808181763,
+      "learning_rate": 3.651698540581252e-06,
+      "loss": 0.1,
+      "step": 6799
+    },
+    {
+      "epoch": 1.101749837977965,
+      "grad_norm": 0.791142463684082,
+      "learning_rate": 3.6513103864490497e-06,
+      "loss": 0.1036,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1019118600129618,
+      "grad_norm": 0.7884705662727356,
+      "learning_rate": 3.6509221970895365e-06,
+      "loss": 0.0956,
+      "step": 6801
+    },
+    {
+      "epoch": 1.1020738820479585,
+      "grad_norm": 0.8608754277229309,
+      "learning_rate": 3.650533972514589e-06,
+      "loss": 0.1153,
+      "step": 6802
+    },
+    {
+      "epoch": 1.1022359040829552,
+      "grad_norm": 0.7507930397987366,
+      "learning_rate": 3.6501457127360863e-06,
+      "loss": 0.0989,
+      "step": 6803
+    },
+    {
+      "epoch": 1.102397926117952,
+      "grad_norm": 0.9590283036231995,
+      "learning_rate": 3.6497574177659073e-06,
+      "loss": 0.1085,
+      "step": 6804
+    },
+    {
+      "epoch": 1.1025599481529489,
+      "grad_norm": 0.805808424949646,
+      "learning_rate": 3.6493690876159343e-06,
+      "loss": 0.0971,
+      "step": 6805
+    },
+    {
+      "epoch": 1.1027219701879456,
+      "grad_norm": 0.7817717790603638,
+      "learning_rate": 3.6489807222980487e-06,
+      "loss": 0.0987,
+      "step": 6806
+    },
+    {
+      "epoch": 1.1028839922229423,
+      "grad_norm": 0.8280859589576721,
+      "learning_rate": 3.648592321824133e-06,
+      "loss": 0.1153,
+      "step": 6807
+    },
+    {
+      "epoch": 1.103046014257939,
+      "grad_norm": 0.9206019639968872,
+      "learning_rate": 3.648203886206073e-06,
+      "loss": 0.1135,
+      "step": 6808
+    },
+    {
+      "epoch": 1.1032080362929357,
+      "grad_norm": 0.8165472745895386,
+      "learning_rate": 3.647815415455751e-06,
+      "loss": 0.1011,
+      "step": 6809
+    },
+    {
+      "epoch": 1.1033700583279327,
+      "grad_norm": 0.8732919096946716,
+      "learning_rate": 3.6474269095850568e-06,
+      "loss": 0.1107,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1035320803629294,
+      "grad_norm": 0.9100021719932556,
+      "learning_rate": 3.647038368605875e-06,
+      "loss": 0.1159,
+      "step": 6811
+    },
+    {
+      "epoch": 1.103694102397926,
+      "grad_norm": 0.8154706358909607,
+      "learning_rate": 3.646649792530094e-06,
+      "loss": 0.1027,
+      "step": 6812
+    },
+    {
+      "epoch": 1.1038561244329228,
+      "grad_norm": 0.8055473566055298,
+      "learning_rate": 3.646261181369605e-06,
+      "loss": 0.1039,
+      "step": 6813
+    },
+    {
+      "epoch": 1.1040181464679197,
+      "grad_norm": 0.7719459533691406,
+      "learning_rate": 3.645872535136298e-06,
+      "loss": 0.0941,
+      "step": 6814
+    },
+    {
+      "epoch": 1.1041801685029164,
+      "grad_norm": 0.8571450710296631,
+      "learning_rate": 3.6454838538420645e-06,
+      "loss": 0.1105,
+      "step": 6815
+    },
+    {
+      "epoch": 1.1043421905379132,
+      "grad_norm": 0.8617749214172363,
+      "learning_rate": 3.6450951374987958e-06,
+      "loss": 0.1069,
+      "step": 6816
+    },
+    {
+      "epoch": 1.1045042125729099,
+      "grad_norm": 0.7611509561538696,
+      "learning_rate": 3.6447063861183886e-06,
+      "loss": 0.1015,
+      "step": 6817
+    },
+    {
+      "epoch": 1.1046662346079066,
+      "grad_norm": 0.7744086980819702,
+      "learning_rate": 3.6443175997127354e-06,
+      "loss": 0.094,
+      "step": 6818
+    },
+    {
+      "epoch": 1.1048282566429035,
+      "grad_norm": 0.7558098435401917,
+      "learning_rate": 3.6439287782937328e-06,
+      "loss": 0.0985,
+      "step": 6819
+    },
+    {
+      "epoch": 1.1049902786779002,
+      "grad_norm": 0.8518016934394836,
+      "learning_rate": 3.6435399218732776e-06,
+      "loss": 0.1097,
+      "step": 6820
+    },
+    {
+      "epoch": 1.105152300712897,
+      "grad_norm": 0.789505660533905,
+      "learning_rate": 3.6431510304632683e-06,
+      "loss": 0.1019,
+      "step": 6821
+    },
+    {
+      "epoch": 1.1053143227478937,
+      "grad_norm": 0.9358975291252136,
+      "learning_rate": 3.642762104075604e-06,
+      "loss": 0.117,
+      "step": 6822
+    },
+    {
+      "epoch": 1.1054763447828906,
+      "grad_norm": 0.9172407388687134,
+      "learning_rate": 3.642373142722185e-06,
+      "loss": 0.1055,
+      "step": 6823
+    },
+    {
+      "epoch": 1.1056383668178873,
+      "grad_norm": 0.8525999784469604,
+      "learning_rate": 3.641984146414912e-06,
+      "loss": 0.1054,
+      "step": 6824
+    },
+    {
+      "epoch": 1.105800388852884,
+      "grad_norm": 0.8727070093154907,
+      "learning_rate": 3.6415951151656874e-06,
+      "loss": 0.108,
+      "step": 6825
+    },
+    {
+      "epoch": 1.1059624108878807,
+      "grad_norm": 0.8490038514137268,
+      "learning_rate": 3.6412060489864155e-06,
+      "loss": 0.1079,
+      "step": 6826
+    },
+    {
+      "epoch": 1.1061244329228774,
+      "grad_norm": 0.8071814179420471,
+      "learning_rate": 3.6408169478889997e-06,
+      "loss": 0.1039,
+      "step": 6827
+    },
+    {
+      "epoch": 1.1062864549578744,
+      "grad_norm": 0.8197512030601501,
+      "learning_rate": 3.640427811885346e-06,
+      "loss": 0.0976,
+      "step": 6828
+    },
+    {
+      "epoch": 1.106448476992871,
+      "grad_norm": 0.8252851963043213,
+      "learning_rate": 3.640038640987361e-06,
+      "loss": 0.0959,
+      "step": 6829
+    },
+    {
+      "epoch": 1.1066104990278678,
+      "grad_norm": 0.9495880603790283,
+      "learning_rate": 3.639649435206953e-06,
+      "loss": 0.11,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1067725210628645,
+      "grad_norm": 0.9515889286994934,
+      "learning_rate": 3.639260194556029e-06,
+      "loss": 0.1006,
+      "step": 6831
+    },
+    {
+      "epoch": 1.1069345430978612,
+      "grad_norm": 0.9411259889602661,
+      "learning_rate": 3.6388709190465018e-06,
+      "loss": 0.1134,
+      "step": 6832
+    },
+    {
+      "epoch": 1.1070965651328581,
+      "grad_norm": 1.127980351448059,
+      "learning_rate": 3.638481608690279e-06,
+      "loss": 0.1175,
+      "step": 6833
+    },
+    {
+      "epoch": 1.1072585871678549,
+      "grad_norm": 0.776599109172821,
+      "learning_rate": 3.638092263499274e-06,
+      "loss": 0.0971,
+      "step": 6834
+    },
+    {
+      "epoch": 1.1074206092028516,
+      "grad_norm": 1.0031170845031738,
+      "learning_rate": 3.637702883485401e-06,
+      "loss": 0.1137,
+      "step": 6835
+    },
+    {
+      "epoch": 1.1075826312378483,
+      "grad_norm": 0.817516028881073,
+      "learning_rate": 3.6373134686605722e-06,
+      "loss": 0.1019,
+      "step": 6836
+    },
+    {
+      "epoch": 1.107744653272845,
+      "grad_norm": 0.7830043435096741,
+      "learning_rate": 3.636924019036704e-06,
+      "loss": 0.0928,
+      "step": 6837
+    },
+    {
+      "epoch": 1.107906675307842,
+      "grad_norm": 0.7438381314277649,
+      "learning_rate": 3.6365345346257112e-06,
+      "loss": 0.0897,
+      "step": 6838
+    },
+    {
+      "epoch": 1.1080686973428386,
+      "grad_norm": 0.8136735558509827,
+      "learning_rate": 3.636145015439513e-06,
+      "loss": 0.1014,
+      "step": 6839
+    },
+    {
+      "epoch": 1.1082307193778353,
+      "grad_norm": 0.8344704508781433,
+      "learning_rate": 3.635755461490026e-06,
+      "loss": 0.0962,
+      "step": 6840
+    },
+    {
+      "epoch": 1.108392741412832,
+      "grad_norm": 1.0317944288253784,
+      "learning_rate": 3.635365872789171e-06,
+      "loss": 0.1298,
+      "step": 6841
+    },
+    {
+      "epoch": 1.108554763447829,
+      "grad_norm": 0.8629235029220581,
+      "learning_rate": 3.634976249348867e-06,
+      "loss": 0.1053,
+      "step": 6842
+    },
+    {
+      "epoch": 1.1087167854828257,
+      "grad_norm": 0.8366121649742126,
+      "learning_rate": 3.6345865911810373e-06,
+      "loss": 0.1031,
+      "step": 6843
+    },
+    {
+      "epoch": 1.1088788075178224,
+      "grad_norm": 0.7734202146530151,
+      "learning_rate": 3.6341968982976027e-06,
+      "loss": 0.0979,
+      "step": 6844
+    },
+    {
+      "epoch": 1.1090408295528191,
+      "grad_norm": 0.9085724353790283,
+      "learning_rate": 3.633807170710488e-06,
+      "loss": 0.1118,
+      "step": 6845
+    },
+    {
+      "epoch": 1.1092028515878158,
+      "grad_norm": 0.8711262941360474,
+      "learning_rate": 3.6334174084316186e-06,
+      "loss": 0.1035,
+      "step": 6846
+    },
+    {
+      "epoch": 1.1093648736228128,
+      "grad_norm": 0.7932325005531311,
+      "learning_rate": 3.6330276114729185e-06,
+      "loss": 0.1024,
+      "step": 6847
+    },
+    {
+      "epoch": 1.1095268956578095,
+      "grad_norm": 0.8877636194229126,
+      "learning_rate": 3.632637779846315e-06,
+      "loss": 0.0966,
+      "step": 6848
+    },
+    {
+      "epoch": 1.1096889176928062,
+      "grad_norm": 0.7809078693389893,
+      "learning_rate": 3.6322479135637366e-06,
+      "loss": 0.1062,
+      "step": 6849
+    },
+    {
+      "epoch": 1.109850939727803,
+      "grad_norm": 0.7734099626541138,
+      "learning_rate": 3.6318580126371124e-06,
+      "loss": 0.0957,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1100129617627998,
+      "grad_norm": 0.811259925365448,
+      "learning_rate": 3.6314680770783717e-06,
+      "loss": 0.108,
+      "step": 6851
+    },
+    {
+      "epoch": 1.1101749837977966,
+      "grad_norm": 0.9896193742752075,
+      "learning_rate": 3.631078106899446e-06,
+      "loss": 0.1288,
+      "step": 6852
+    },
+    {
+      "epoch": 1.1103370058327933,
+      "grad_norm": 0.7642198801040649,
+      "learning_rate": 3.6306881021122675e-06,
+      "loss": 0.0916,
+      "step": 6853
+    },
+    {
+      "epoch": 1.11049902786779,
+      "grad_norm": 0.774316668510437,
+      "learning_rate": 3.630298062728769e-06,
+      "loss": 0.1009,
+      "step": 6854
+    },
+    {
+      "epoch": 1.1106610499027867,
+      "grad_norm": 0.9205217361450195,
+      "learning_rate": 3.629907988760886e-06,
+      "loss": 0.1038,
+      "step": 6855
+    },
+    {
+      "epoch": 1.1108230719377836,
+      "grad_norm": 0.8705753087997437,
+      "learning_rate": 3.6295178802205515e-06,
+      "loss": 0.1142,
+      "step": 6856
+    },
+    {
+      "epoch": 1.1109850939727803,
+      "grad_norm": 0.8856770992279053,
+      "learning_rate": 3.6291277371197042e-06,
+      "loss": 0.1111,
+      "step": 6857
+    },
+    {
+      "epoch": 1.111147116007777,
+      "grad_norm": 0.7869232296943665,
+      "learning_rate": 3.62873755947028e-06,
+      "loss": 0.1016,
+      "step": 6858
+    },
+    {
+      "epoch": 1.1113091380427738,
+      "grad_norm": 0.8357443809509277,
+      "learning_rate": 3.628347347284218e-06,
+      "loss": 0.1095,
+      "step": 6859
+    },
+    {
+      "epoch": 1.1114711600777705,
+      "grad_norm": 0.8363984227180481,
+      "learning_rate": 3.6279571005734583e-06,
+      "loss": 0.0984,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1116331821127674,
+      "grad_norm": 0.9447827339172363,
+      "learning_rate": 3.627566819349941e-06,
+      "loss": 0.1097,
+      "step": 6861
+    },
+    {
+      "epoch": 1.1117952041477641,
+      "grad_norm": 0.9603884220123291,
+      "learning_rate": 3.6271765036256064e-06,
+      "loss": 0.1168,
+      "step": 6862
+    },
+    {
+      "epoch": 1.1119572261827608,
+      "grad_norm": 0.7766293883323669,
+      "learning_rate": 3.6267861534124e-06,
+      "loss": 0.0919,
+      "step": 6863
+    },
+    {
+      "epoch": 1.1121192482177575,
+      "grad_norm": 0.6767370700836182,
+      "learning_rate": 3.6263957687222633e-06,
+      "loss": 0.0832,
+      "step": 6864
+    },
+    {
+      "epoch": 1.1122812702527545,
+      "grad_norm": 0.8346736431121826,
+      "learning_rate": 3.6260053495671423e-06,
+      "loss": 0.1018,
+      "step": 6865
+    },
+    {
+      "epoch": 1.1124432922877512,
+      "grad_norm": 0.871659517288208,
+      "learning_rate": 3.625614895958982e-06,
+      "loss": 0.1006,
+      "step": 6866
+    },
+    {
+      "epoch": 1.112605314322748,
+      "grad_norm": 0.6866405010223389,
+      "learning_rate": 3.6252244079097296e-06,
+      "loss": 0.0874,
+      "step": 6867
+    },
+    {
+      "epoch": 1.1127673363577446,
+      "grad_norm": 0.8131086230278015,
+      "learning_rate": 3.624833885431334e-06,
+      "loss": 0.1062,
+      "step": 6868
+    },
+    {
+      "epoch": 1.1129293583927413,
+      "grad_norm": 0.9282816052436829,
+      "learning_rate": 3.6244433285357433e-06,
+      "loss": 0.1105,
+      "step": 6869
+    },
+    {
+      "epoch": 1.1130913804277383,
+      "grad_norm": 0.7851607799530029,
+      "learning_rate": 3.624052737234908e-06,
+      "loss": 0.1028,
+      "step": 6870
+    },
+    {
+      "epoch": 1.113253402462735,
+      "grad_norm": 0.7424038648605347,
+      "learning_rate": 3.623662111540779e-06,
+      "loss": 0.0843,
+      "step": 6871
+    },
+    {
+      "epoch": 1.1134154244977317,
+      "grad_norm": 0.8808372616767883,
+      "learning_rate": 3.6232714514653082e-06,
+      "loss": 0.1105,
+      "step": 6872
+    },
+    {
+      "epoch": 1.1135774465327284,
+      "grad_norm": 0.9049409627914429,
+      "learning_rate": 3.62288075702045e-06,
+      "loss": 0.1022,
+      "step": 6873
+    },
+    {
+      "epoch": 1.1137394685677253,
+      "grad_norm": 0.8752559423446655,
+      "learning_rate": 3.6224900282181574e-06,
+      "loss": 0.1034,
+      "step": 6874
+    },
+    {
+      "epoch": 1.113901490602722,
+      "grad_norm": 0.7354977130889893,
+      "learning_rate": 3.622099265070386e-06,
+      "loss": 0.0939,
+      "step": 6875
+    },
+    {
+      "epoch": 1.1140635126377187,
+      "grad_norm": 0.9659753441810608,
+      "learning_rate": 3.6217084675890935e-06,
+      "loss": 0.1085,
+      "step": 6876
+    },
+    {
+      "epoch": 1.1142255346727155,
+      "grad_norm": 0.8423113226890564,
+      "learning_rate": 3.6213176357862364e-06,
+      "loss": 0.1078,
+      "step": 6877
+    },
+    {
+      "epoch": 1.1143875567077122,
+      "grad_norm": 0.8920717239379883,
+      "learning_rate": 3.6209267696737723e-06,
+      "loss": 0.111,
+      "step": 6878
+    },
+    {
+      "epoch": 1.114549578742709,
+      "grad_norm": 0.9556100964546204,
+      "learning_rate": 3.6205358692636618e-06,
+      "loss": 0.1241,
+      "step": 6879
+    },
+    {
+      "epoch": 1.1147116007777058,
+      "grad_norm": 0.8358135223388672,
+      "learning_rate": 3.6201449345678657e-06,
+      "loss": 0.1113,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1148736228127025,
+      "grad_norm": 0.9949773550033569,
+      "learning_rate": 3.6197539655983447e-06,
+      "loss": 0.1158,
+      "step": 6881
+    },
+    {
+      "epoch": 1.1150356448476992,
+      "grad_norm": 0.7863255739212036,
+      "learning_rate": 3.6193629623670627e-06,
+      "loss": 0.101,
+      "step": 6882
+    },
+    {
+      "epoch": 1.115197666882696,
+      "grad_norm": 0.8213348388671875,
+      "learning_rate": 3.6189719248859827e-06,
+      "loss": 0.0921,
+      "step": 6883
+    },
+    {
+      "epoch": 1.1153596889176929,
+      "grad_norm": 0.841783344745636,
+      "learning_rate": 3.6185808531670695e-06,
+      "loss": 0.1075,
+      "step": 6884
+    },
+    {
+      "epoch": 1.1155217109526896,
+      "grad_norm": 0.8018286228179932,
+      "learning_rate": 3.61818974722229e-06,
+      "loss": 0.1009,
+      "step": 6885
+    },
+    {
+      "epoch": 1.1156837329876863,
+      "grad_norm": 0.826497495174408,
+      "learning_rate": 3.617798607063609e-06,
+      "loss": 0.0976,
+      "step": 6886
+    },
+    {
+      "epoch": 1.115845755022683,
+      "grad_norm": 0.796943187713623,
+      "learning_rate": 3.6174074327029964e-06,
+      "loss": 0.1033,
+      "step": 6887
+    },
+    {
+      "epoch": 1.1160077770576797,
+      "grad_norm": 0.8021527528762817,
+      "learning_rate": 3.61701622415242e-06,
+      "loss": 0.1055,
+      "step": 6888
+    },
+    {
+      "epoch": 1.1161697990926767,
+      "grad_norm": 0.7664604187011719,
+      "learning_rate": 3.6166249814238503e-06,
+      "loss": 0.0922,
+      "step": 6889
+    },
+    {
+      "epoch": 1.1163318211276734,
+      "grad_norm": 0.7481946349143982,
+      "learning_rate": 3.616233704529259e-06,
+      "loss": 0.0954,
+      "step": 6890
+    },
+    {
+      "epoch": 1.11649384316267,
+      "grad_norm": 0.7271841168403625,
+      "learning_rate": 3.6158423934806164e-06,
+      "loss": 0.09,
+      "step": 6891
+    },
+    {
+      "epoch": 1.1166558651976668,
+      "grad_norm": 1.0113109350204468,
+      "learning_rate": 3.6154510482898973e-06,
+      "loss": 0.1189,
+      "step": 6892
+    },
+    {
+      "epoch": 1.1168178872326637,
+      "grad_norm": 0.7146120667457581,
+      "learning_rate": 3.6150596689690753e-06,
+      "loss": 0.0896,
+      "step": 6893
+    },
+    {
+      "epoch": 1.1169799092676604,
+      "grad_norm": 0.8736560344696045,
+      "learning_rate": 3.6146682555301266e-06,
+      "loss": 0.111,
+      "step": 6894
+    },
+    {
+      "epoch": 1.1171419313026572,
+      "grad_norm": 0.9498534202575684,
+      "learning_rate": 3.614276807985026e-06,
+      "loss": 0.1128,
+      "step": 6895
+    },
+    {
+      "epoch": 1.1173039533376539,
+      "grad_norm": 0.9466055631637573,
+      "learning_rate": 3.613885326345752e-06,
+      "loss": 0.1196,
+      "step": 6896
+    },
+    {
+      "epoch": 1.1174659753726506,
+      "grad_norm": 0.8748520016670227,
+      "learning_rate": 3.6134938106242823e-06,
+      "loss": 0.1088,
+      "step": 6897
+    },
+    {
+      "epoch": 1.1176279974076475,
+      "grad_norm": 1.0820103883743286,
+      "learning_rate": 3.6131022608325973e-06,
+      "loss": 0.1119,
+      "step": 6898
+    },
+    {
+      "epoch": 1.1177900194426442,
+      "grad_norm": 0.7906978726387024,
+      "learning_rate": 3.6127106769826763e-06,
+      "loss": 0.1029,
+      "step": 6899
+    },
+    {
+      "epoch": 1.117952041477641,
+      "grad_norm": 0.7698807120323181,
+      "learning_rate": 3.6123190590865e-06,
+      "loss": 0.0988,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1181140635126376,
+      "grad_norm": 0.8202066421508789,
+      "learning_rate": 3.6119274071560545e-06,
+      "loss": 0.1104,
+      "step": 6901
+    },
+    {
+      "epoch": 1.1182760855476346,
+      "grad_norm": 0.969008207321167,
+      "learning_rate": 3.6115357212033196e-06,
+      "loss": 0.1047,
+      "step": 6902
+    },
+    {
+      "epoch": 1.1184381075826313,
+      "grad_norm": 0.7530080676078796,
+      "learning_rate": 3.611144001240282e-06,
+      "loss": 0.0888,
+      "step": 6903
+    },
+    {
+      "epoch": 1.118600129617628,
+      "grad_norm": 0.8667441010475159,
+      "learning_rate": 3.610752247278927e-06,
+      "loss": 0.1088,
+      "step": 6904
+    },
+    {
+      "epoch": 1.1187621516526247,
+      "grad_norm": 0.7977768778800964,
+      "learning_rate": 3.610360459331241e-06,
+      "loss": 0.1009,
+      "step": 6905
+    },
+    {
+      "epoch": 1.1189241736876214,
+      "grad_norm": 0.8574357032775879,
+      "learning_rate": 3.609968637409212e-06,
+      "loss": 0.1125,
+      "step": 6906
+    },
+    {
+      "epoch": 1.1190861957226184,
+      "grad_norm": 0.8181487321853638,
+      "learning_rate": 3.609576781524829e-06,
+      "loss": 0.1047,
+      "step": 6907
+    },
+    {
+      "epoch": 1.119248217757615,
+      "grad_norm": 0.8313112258911133,
+      "learning_rate": 3.6091848916900816e-06,
+      "loss": 0.1016,
+      "step": 6908
+    },
+    {
+      "epoch": 1.1194102397926118,
+      "grad_norm": 0.7930647730827332,
+      "learning_rate": 3.6087929679169603e-06,
+      "loss": 0.1034,
+      "step": 6909
+    },
+    {
+      "epoch": 1.1195722618276085,
+      "grad_norm": 0.9460997581481934,
+      "learning_rate": 3.6084010102174576e-06,
+      "loss": 0.1059,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1197342838626052,
+      "grad_norm": 0.9013634920120239,
+      "learning_rate": 3.6080090186035664e-06,
+      "loss": 0.1091,
+      "step": 6911
+    },
+    {
+      "epoch": 1.1198963058976021,
+      "grad_norm": 0.905096173286438,
+      "learning_rate": 3.6076169930872805e-06,
+      "loss": 0.1161,
+      "step": 6912
+    },
+    {
+      "epoch": 1.1200583279325989,
+      "grad_norm": 0.7841644883155823,
+      "learning_rate": 3.607224933680595e-06,
+      "loss": 0.106,
+      "step": 6913
+    },
+    {
+      "epoch": 1.1202203499675956,
+      "grad_norm": 0.8874266743659973,
+      "learning_rate": 3.606832840395506e-06,
+      "loss": 0.1087,
+      "step": 6914
+    },
+    {
+      "epoch": 1.1203823720025923,
+      "grad_norm": 0.7959884405136108,
+      "learning_rate": 3.60644071324401e-06,
+      "loss": 0.1002,
+      "step": 6915
+    },
+    {
+      "epoch": 1.1205443940375892,
+      "grad_norm": 0.8270828127861023,
+      "learning_rate": 3.6060485522381067e-06,
+      "loss": 0.1056,
+      "step": 6916
+    },
+    {
+      "epoch": 1.120706416072586,
+      "grad_norm": 0.7726624011993408,
+      "learning_rate": 3.6056563573897927e-06,
+      "loss": 0.0999,
+      "step": 6917
+    },
+    {
+      "epoch": 1.1208684381075826,
+      "grad_norm": 0.9138970375061035,
+      "learning_rate": 3.605264128711072e-06,
+      "loss": 0.1154,
+      "step": 6918
+    },
+    {
+      "epoch": 1.1210304601425793,
+      "grad_norm": 0.8613343238830566,
+      "learning_rate": 3.604871866213942e-06,
+      "loss": 0.111,
+      "step": 6919
+    },
+    {
+      "epoch": 1.121192482177576,
+      "grad_norm": 0.855797290802002,
+      "learning_rate": 3.6044795699104074e-06,
+      "loss": 0.1185,
+      "step": 6920
+    },
+    {
+      "epoch": 1.121354504212573,
+      "grad_norm": 0.7498341202735901,
+      "learning_rate": 3.6040872398124705e-06,
+      "loss": 0.0892,
+      "step": 6921
+    },
+    {
+      "epoch": 1.1215165262475697,
+      "grad_norm": 0.7906498312950134,
+      "learning_rate": 3.6036948759321357e-06,
+      "loss": 0.0972,
+      "step": 6922
+    },
+    {
+      "epoch": 1.1216785482825664,
+      "grad_norm": 0.8063822388648987,
+      "learning_rate": 3.603302478281409e-06,
+      "loss": 0.1016,
+      "step": 6923
+    },
+    {
+      "epoch": 1.1218405703175631,
+      "grad_norm": 0.7980828881263733,
+      "learning_rate": 3.6029100468722954e-06,
+      "loss": 0.0919,
+      "step": 6924
+    },
+    {
+      "epoch": 1.12200259235256,
+      "grad_norm": 0.7965691685676575,
+      "learning_rate": 3.6025175817168046e-06,
+      "loss": 0.0889,
+      "step": 6925
+    },
+    {
+      "epoch": 1.1221646143875568,
+      "grad_norm": 0.8087196350097656,
+      "learning_rate": 3.602125082826944e-06,
+      "loss": 0.0985,
+      "step": 6926
+    },
+    {
+      "epoch": 1.1223266364225535,
+      "grad_norm": 0.9410887956619263,
+      "learning_rate": 3.6017325502147215e-06,
+      "loss": 0.101,
+      "step": 6927
+    },
+    {
+      "epoch": 1.1224886584575502,
+      "grad_norm": 0.8947740197181702,
+      "learning_rate": 3.60133998389215e-06,
+      "loss": 0.111,
+      "step": 6928
+    },
+    {
+      "epoch": 1.122650680492547,
+      "grad_norm": 0.7792215347290039,
+      "learning_rate": 3.6009473838712405e-06,
+      "loss": 0.0952,
+      "step": 6929
+    },
+    {
+      "epoch": 1.1228127025275438,
+      "grad_norm": 0.9168524146080017,
+      "learning_rate": 3.600554750164005e-06,
+      "loss": 0.1141,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1229747245625405,
+      "grad_norm": 0.8868537545204163,
+      "learning_rate": 3.600162082782457e-06,
+      "loss": 0.1154,
+      "step": 6931
+    },
+    {
+      "epoch": 1.1231367465975373,
+      "grad_norm": 0.8856688737869263,
+      "learning_rate": 3.5997693817386128e-06,
+      "loss": 0.0962,
+      "step": 6932
+    },
+    {
+      "epoch": 1.123298768632534,
+      "grad_norm": 0.8674602508544922,
+      "learning_rate": 3.5993766470444856e-06,
+      "loss": 0.096,
+      "step": 6933
+    },
+    {
+      "epoch": 1.1234607906675307,
+      "grad_norm": 0.8470571041107178,
+      "learning_rate": 3.598983878712094e-06,
+      "loss": 0.1131,
+      "step": 6934
+    },
+    {
+      "epoch": 1.1236228127025276,
+      "grad_norm": 0.9755722284317017,
+      "learning_rate": 3.598591076753455e-06,
+      "loss": 0.1109,
+      "step": 6935
+    },
+    {
+      "epoch": 1.1237848347375243,
+      "grad_norm": 0.8115435838699341,
+      "learning_rate": 3.598198241180588e-06,
+      "loss": 0.1044,
+      "step": 6936
+    },
+    {
+      "epoch": 1.123946856772521,
+      "grad_norm": 0.7542571425437927,
+      "learning_rate": 3.5978053720055117e-06,
+      "loss": 0.097,
+      "step": 6937
+    },
+    {
+      "epoch": 1.1241088788075178,
+      "grad_norm": 0.7188454270362854,
+      "learning_rate": 3.597412469240248e-06,
+      "loss": 0.086,
+      "step": 6938
+    },
+    {
+      "epoch": 1.1242709008425145,
+      "grad_norm": 0.807829737663269,
+      "learning_rate": 3.5970195328968183e-06,
+      "loss": 0.1037,
+      "step": 6939
+    },
+    {
+      "epoch": 1.1244329228775114,
+      "grad_norm": 0.7683031558990479,
+      "learning_rate": 3.5966265629872466e-06,
+      "loss": 0.1012,
+      "step": 6940
+    },
+    {
+      "epoch": 1.124594944912508,
+      "grad_norm": 0.805733323097229,
+      "learning_rate": 3.5962335595235547e-06,
+      "loss": 0.1005,
+      "step": 6941
+    },
+    {
+      "epoch": 1.1247569669475048,
+      "grad_norm": 0.8244421482086182,
+      "learning_rate": 3.595840522517769e-06,
+      "loss": 0.1096,
+      "step": 6942
+    },
+    {
+      "epoch": 1.1249189889825015,
+      "grad_norm": 0.7452391386032104,
+      "learning_rate": 3.5954474519819155e-06,
+      "loss": 0.1078,
+      "step": 6943
+    },
+    {
+      "epoch": 1.1250810110174985,
+      "grad_norm": 0.8935032486915588,
+      "learning_rate": 3.5950543479280205e-06,
+      "loss": 0.1056,
+      "step": 6944
+    },
+    {
+      "epoch": 1.1252430330524952,
+      "grad_norm": 0.7510908246040344,
+      "learning_rate": 3.5946612103681135e-06,
+      "loss": 0.0906,
+      "step": 6945
+    },
+    {
+      "epoch": 1.125405055087492,
+      "grad_norm": 0.7809535264968872,
+      "learning_rate": 3.5942680393142203e-06,
+      "loss": 0.1014,
+      "step": 6946
+    },
+    {
+      "epoch": 1.1255670771224886,
+      "grad_norm": 0.9290381073951721,
+      "learning_rate": 3.5938748347783754e-06,
+      "loss": 0.12,
+      "step": 6947
+    },
+    {
+      "epoch": 1.1257290991574855,
+      "grad_norm": 0.7380486726760864,
+      "learning_rate": 3.593481596772606e-06,
+      "loss": 0.0862,
+      "step": 6948
+    },
+    {
+      "epoch": 1.1258911211924822,
+      "grad_norm": 0.7775296568870544,
+      "learning_rate": 3.593088325308947e-06,
+      "loss": 0.1076,
+      "step": 6949
+    },
+    {
+      "epoch": 1.126053143227479,
+      "grad_norm": 0.8212987184524536,
+      "learning_rate": 3.5926950203994303e-06,
+      "loss": 0.0965,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1262151652624757,
+      "grad_norm": 0.758618175983429,
+      "learning_rate": 3.5923016820560904e-06,
+      "loss": 0.0958,
+      "step": 6951
+    },
+    {
+      "epoch": 1.1263771872974724,
+      "grad_norm": 0.7440875172615051,
+      "learning_rate": 3.5919083102909615e-06,
+      "loss": 0.092,
+      "step": 6952
+    },
+    {
+      "epoch": 1.1265392093324693,
+      "grad_norm": 0.755118727684021,
+      "learning_rate": 3.5915149051160812e-06,
+      "loss": 0.0912,
+      "step": 6953
+    },
+    {
+      "epoch": 1.126701231367466,
+      "grad_norm": 0.9674204587936401,
+      "learning_rate": 3.591121466543487e-06,
+      "loss": 0.119,
+      "step": 6954
+    },
+    {
+      "epoch": 1.1268632534024627,
+      "grad_norm": 0.7956207990646362,
+      "learning_rate": 3.5907279945852145e-06,
+      "loss": 0.0993,
+      "step": 6955
+    },
+    {
+      "epoch": 1.1270252754374595,
+      "grad_norm": 0.8555571436882019,
+      "learning_rate": 3.5903344892533067e-06,
+      "loss": 0.1116,
+      "step": 6956
+    },
+    {
+      "epoch": 1.1271872974724562,
+      "grad_norm": 0.741603434085846,
+      "learning_rate": 3.5899409505598014e-06,
+      "loss": 0.0924,
+      "step": 6957
+    },
+    {
+      "epoch": 1.127349319507453,
+      "grad_norm": 0.8872222900390625,
+      "learning_rate": 3.5895473785167407e-06,
+      "loss": 0.0995,
+      "step": 6958
+    },
+    {
+      "epoch": 1.1275113415424498,
+      "grad_norm": 0.8348520398139954,
+      "learning_rate": 3.589153773136167e-06,
+      "loss": 0.1014,
+      "step": 6959
+    },
+    {
+      "epoch": 1.1276733635774465,
+      "grad_norm": 0.8107140064239502,
+      "learning_rate": 3.5887601344301228e-06,
+      "loss": 0.1088,
+      "step": 6960
+    },
+    {
+      "epoch": 1.1278353856124432,
+      "grad_norm": 0.8019354939460754,
+      "learning_rate": 3.588366462410654e-06,
+      "loss": 0.1078,
+      "step": 6961
+    },
+    {
+      "epoch": 1.12799740764744,
+      "grad_norm": 0.8682474493980408,
+      "learning_rate": 3.587972757089805e-06,
+      "loss": 0.1116,
+      "step": 6962
+    },
+    {
+      "epoch": 1.1281594296824369,
+      "grad_norm": 0.9506396055221558,
+      "learning_rate": 3.5875790184796228e-06,
+      "loss": 0.1215,
+      "step": 6963
+    },
+    {
+      "epoch": 1.1283214517174336,
+      "grad_norm": 0.8694068789482117,
+      "learning_rate": 3.587185246592154e-06,
+      "loss": 0.0923,
+      "step": 6964
+    },
+    {
+      "epoch": 1.1284834737524303,
+      "grad_norm": 0.8598493933677673,
+      "learning_rate": 3.5867914414394478e-06,
+      "loss": 0.1113,
+      "step": 6965
+    },
+    {
+      "epoch": 1.128645495787427,
+      "grad_norm": 0.7127923965454102,
+      "learning_rate": 3.5863976030335535e-06,
+      "loss": 0.0917,
+      "step": 6966
+    },
+    {
+      "epoch": 1.128807517822424,
+      "grad_norm": 0.9294934272766113,
+      "learning_rate": 3.5860037313865216e-06,
+      "loss": 0.1139,
+      "step": 6967
+    },
+    {
+      "epoch": 1.1289695398574207,
+      "grad_norm": 0.7821537256240845,
+      "learning_rate": 3.5856098265104033e-06,
+      "loss": 0.1006,
+      "step": 6968
+    },
+    {
+      "epoch": 1.1291315618924174,
+      "grad_norm": 0.7920688986778259,
+      "learning_rate": 3.5852158884172523e-06,
+      "loss": 0.0979,
+      "step": 6969
+    },
+    {
+      "epoch": 1.129293583927414,
+      "grad_norm": 0.772294819355011,
+      "learning_rate": 3.58482191711912e-06,
+      "loss": 0.0933,
+      "step": 6970
+    },
+    {
+      "epoch": 1.1294556059624108,
+      "grad_norm": 0.7953448295593262,
+      "learning_rate": 3.5844279126280635e-06,
+      "loss": 0.1057,
+      "step": 6971
+    },
+    {
+      "epoch": 1.1296176279974077,
+      "grad_norm": 0.7928208112716675,
+      "learning_rate": 3.5840338749561365e-06,
+      "loss": 0.1052,
+      "step": 6972
+    },
+    {
+      "epoch": 1.1297796500324044,
+      "grad_norm": 0.885570228099823,
+      "learning_rate": 3.5836398041153962e-06,
+      "loss": 0.106,
+      "step": 6973
+    },
+    {
+      "epoch": 1.1299416720674011,
+      "grad_norm": 0.8886401057243347,
+      "learning_rate": 3.5832457001179e-06,
+      "loss": 0.107,
+      "step": 6974
+    },
+    {
+      "epoch": 1.1301036941023979,
+      "grad_norm": 0.6969801783561707,
+      "learning_rate": 3.5828515629757073e-06,
+      "loss": 0.0911,
+      "step": 6975
+    },
+    {
+      "epoch": 1.1302657161373948,
+      "grad_norm": 0.8328415155410767,
+      "learning_rate": 3.582457392700878e-06,
+      "loss": 0.1073,
+      "step": 6976
+    },
+    {
+      "epoch": 1.1304277381723915,
+      "grad_norm": 0.8379804491996765,
+      "learning_rate": 3.5820631893054703e-06,
+      "loss": 0.114,
+      "step": 6977
+    },
+    {
+      "epoch": 1.1305897602073882,
+      "grad_norm": 0.8626766204833984,
+      "learning_rate": 3.5816689528015485e-06,
+      "loss": 0.1088,
+      "step": 6978
+    },
+    {
+      "epoch": 1.130751782242385,
+      "grad_norm": 0.7580254077911377,
+      "learning_rate": 3.5812746832011734e-06,
+      "loss": 0.0952,
+      "step": 6979
+    },
+    {
+      "epoch": 1.1309138042773816,
+      "grad_norm": 0.9701016545295715,
+      "learning_rate": 3.580880380516411e-06,
+      "loss": 0.1154,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1310758263123786,
+      "grad_norm": 0.8925403356552124,
+      "learning_rate": 3.5804860447593236e-06,
+      "loss": 0.1066,
+      "step": 6981
+    },
+    {
+      "epoch": 1.1312378483473753,
+      "grad_norm": 0.861957848072052,
+      "learning_rate": 3.5800916759419784e-06,
+      "loss": 0.0999,
+      "step": 6982
+    },
+    {
+      "epoch": 1.131399870382372,
+      "grad_norm": 0.9451190233230591,
+      "learning_rate": 3.579697274076442e-06,
+      "loss": 0.1148,
+      "step": 6983
+    },
+    {
+      "epoch": 1.1315618924173687,
+      "grad_norm": 0.8173058032989502,
+      "learning_rate": 3.579302839174781e-06,
+      "loss": 0.1155,
+      "step": 6984
+    },
+    {
+      "epoch": 1.1317239144523654,
+      "grad_norm": 0.9785115718841553,
+      "learning_rate": 3.578908371249066e-06,
+      "loss": 0.12,
+      "step": 6985
+    },
+    {
+      "epoch": 1.1318859364873624,
+      "grad_norm": 0.8944734334945679,
+      "learning_rate": 3.578513870311365e-06,
+      "loss": 0.1105,
+      "step": 6986
+    },
+    {
+      "epoch": 1.132047958522359,
+      "grad_norm": 0.8797488808631897,
+      "learning_rate": 3.57811933637375e-06,
+      "loss": 0.1095,
+      "step": 6987
+    },
+    {
+      "epoch": 1.1322099805573558,
+      "grad_norm": 0.8068286180496216,
+      "learning_rate": 3.577724769448292e-06,
+      "loss": 0.1002,
+      "step": 6988
+    },
+    {
+      "epoch": 1.1323720025923525,
+      "grad_norm": 0.993378758430481,
+      "learning_rate": 3.577330169547064e-06,
+      "loss": 0.1254,
+      "step": 6989
+    },
+    {
+      "epoch": 1.1325340246273492,
+      "grad_norm": 0.7981913089752197,
+      "learning_rate": 3.57693553668214e-06,
+      "loss": 0.1031,
+      "step": 6990
+    },
+    {
+      "epoch": 1.1326960466623461,
+      "grad_norm": 0.8486551642417908,
+      "learning_rate": 3.5765408708655946e-06,
+      "loss": 0.1028,
+      "step": 6991
+    },
+    {
+      "epoch": 1.1328580686973428,
+      "grad_norm": 0.8483676910400391,
+      "learning_rate": 3.5761461721095037e-06,
+      "loss": 0.1061,
+      "step": 6992
+    },
+    {
+      "epoch": 1.1330200907323396,
+      "grad_norm": 0.8058338165283203,
+      "learning_rate": 3.5757514404259447e-06,
+      "loss": 0.1053,
+      "step": 6993
+    },
+    {
+      "epoch": 1.1331821127673363,
+      "grad_norm": 0.9978876709938049,
+      "learning_rate": 3.575356675826995e-06,
+      "loss": 0.1242,
+      "step": 6994
+    },
+    {
+      "epoch": 1.1333441348023332,
+      "grad_norm": 0.8573361039161682,
+      "learning_rate": 3.5749618783247326e-06,
+      "loss": 0.1093,
+      "step": 6995
+    },
+    {
+      "epoch": 1.13350615683733,
+      "grad_norm": 0.8889877200126648,
+      "learning_rate": 3.574567047931238e-06,
+      "loss": 0.1067,
+      "step": 6996
+    },
+    {
+      "epoch": 1.1336681788723266,
+      "grad_norm": 0.7973397970199585,
+      "learning_rate": 3.5741721846585926e-06,
+      "loss": 0.0991,
+      "step": 6997
+    },
+    {
+      "epoch": 1.1338302009073233,
+      "grad_norm": 0.8841112852096558,
+      "learning_rate": 3.5737772885188777e-06,
+      "loss": 0.1111,
+      "step": 6998
+    },
+    {
+      "epoch": 1.1339922229423203,
+      "grad_norm": 0.8456192016601562,
+      "learning_rate": 3.5733823595241758e-06,
+      "loss": 0.111,
+      "step": 6999
+    },
+    {
+      "epoch": 1.134154244977317,
+      "grad_norm": 1.4436651468276978,
+      "learning_rate": 3.5729873976865726e-06,
+      "loss": 0.0931,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1343162670123137,
+      "grad_norm": 0.8534703850746155,
+      "learning_rate": 3.5725924030181498e-06,
+      "loss": 0.1073,
+      "step": 7001
+    },
+    {
+      "epoch": 1.1344782890473104,
+      "grad_norm": 0.8066534996032715,
+      "learning_rate": 3.5721973755309963e-06,
+      "loss": 0.0973,
+      "step": 7002
+    },
+    {
+      "epoch": 1.1346403110823071,
+      "grad_norm": 0.8632422685623169,
+      "learning_rate": 3.571802315237197e-06,
+      "loss": 0.1128,
+      "step": 7003
+    },
+    {
+      "epoch": 1.134802333117304,
+      "grad_norm": 0.9532308578491211,
+      "learning_rate": 3.5714072221488414e-06,
+      "loss": 0.1121,
+      "step": 7004
+    },
+    {
+      "epoch": 1.1349643551523008,
+      "grad_norm": 0.8398197889328003,
+      "learning_rate": 3.571012096278017e-06,
+      "loss": 0.1081,
+      "step": 7005
+    },
+    {
+      "epoch": 1.1351263771872975,
+      "grad_norm": 0.7505828142166138,
+      "learning_rate": 3.5706169376368143e-06,
+      "loss": 0.0898,
+      "step": 7006
+    },
+    {
+      "epoch": 1.1352883992222942,
+      "grad_norm": 0.8282100558280945,
+      "learning_rate": 3.570221746237325e-06,
+      "loss": 0.1008,
+      "step": 7007
+    },
+    {
+      "epoch": 1.135450421257291,
+      "grad_norm": 0.8855400681495667,
+      "learning_rate": 3.5698265220916388e-06,
+      "loss": 0.118,
+      "step": 7008
+    },
+    {
+      "epoch": 1.1356124432922878,
+      "grad_norm": 0.8593195080757141,
+      "learning_rate": 3.5694312652118513e-06,
+      "loss": 0.1072,
+      "step": 7009
+    },
+    {
+      "epoch": 1.1357744653272845,
+      "grad_norm": 0.9126600027084351,
+      "learning_rate": 3.5690359756100532e-06,
+      "loss": 0.1101,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1359364873622813,
+      "grad_norm": 0.7796016335487366,
+      "learning_rate": 3.5686406532983432e-06,
+      "loss": 0.1009,
+      "step": 7011
+    },
+    {
+      "epoch": 1.136098509397278,
+      "grad_norm": 0.822391152381897,
+      "learning_rate": 3.5682452982888143e-06,
+      "loss": 0.1099,
+      "step": 7012
+    },
+    {
+      "epoch": 1.1362605314322747,
+      "grad_norm": 0.8462624549865723,
+      "learning_rate": 3.5678499105935642e-06,
+      "loss": 0.1055,
+      "step": 7013
+    },
+    {
+      "epoch": 1.1364225534672716,
+      "grad_norm": 0.8373414278030396,
+      "learning_rate": 3.5674544902246916e-06,
+      "loss": 0.1144,
+      "step": 7014
+    },
+    {
+      "epoch": 1.1365845755022683,
+      "grad_norm": 0.8660956621170044,
+      "learning_rate": 3.567059037194294e-06,
+      "loss": 0.1096,
+      "step": 7015
+    },
+    {
+      "epoch": 1.136746597537265,
+      "grad_norm": 0.841236412525177,
+      "learning_rate": 3.566663551514473e-06,
+      "loss": 0.1058,
+      "step": 7016
+    },
+    {
+      "epoch": 1.1369086195722617,
+      "grad_norm": 0.8246383666992188,
+      "learning_rate": 3.5662680331973282e-06,
+      "loss": 0.1015,
+      "step": 7017
+    },
+    {
+      "epoch": 1.1370706416072587,
+      "grad_norm": 0.8704448342323303,
+      "learning_rate": 3.5658724822549624e-06,
+      "loss": 0.1055,
+      "step": 7018
+    },
+    {
+      "epoch": 1.1372326636422554,
+      "grad_norm": 0.7135106325149536,
+      "learning_rate": 3.565476898699477e-06,
+      "loss": 0.0863,
+      "step": 7019
+    },
+    {
+      "epoch": 1.137394685677252,
+      "grad_norm": 0.7876743674278259,
+      "learning_rate": 3.5650812825429774e-06,
+      "loss": 0.1019,
+      "step": 7020
+    },
+    {
+      "epoch": 1.1375567077122488,
+      "grad_norm": 0.773827850818634,
+      "learning_rate": 3.564685633797568e-06,
+      "loss": 0.0966,
+      "step": 7021
+    },
+    {
+      "epoch": 1.1377187297472457,
+      "grad_norm": 0.8527294397354126,
+      "learning_rate": 3.5642899524753548e-06,
+      "loss": 0.1016,
+      "step": 7022
+    },
+    {
+      "epoch": 1.1378807517822425,
+      "grad_norm": 0.8665226697921753,
+      "learning_rate": 3.5638942385884443e-06,
+      "loss": 0.1061,
+      "step": 7023
+    },
+    {
+      "epoch": 1.1380427738172392,
+      "grad_norm": 0.7744306325912476,
+      "learning_rate": 3.5634984921489455e-06,
+      "loss": 0.0958,
+      "step": 7024
+    },
+    {
+      "epoch": 1.1382047958522359,
+      "grad_norm": 0.7468297481536865,
+      "learning_rate": 3.563102713168966e-06,
+      "loss": 0.0925,
+      "step": 7025
+    },
+    {
+      "epoch": 1.1383668178872326,
+      "grad_norm": 0.7687481641769409,
+      "learning_rate": 3.562706901660616e-06,
+      "loss": 0.0983,
+      "step": 7026
+    },
+    {
+      "epoch": 1.1385288399222295,
+      "grad_norm": 0.899817943572998,
+      "learning_rate": 3.5623110576360065e-06,
+      "loss": 0.1092,
+      "step": 7027
+    },
+    {
+      "epoch": 1.1386908619572262,
+      "grad_norm": 0.8666191101074219,
+      "learning_rate": 3.561915181107249e-06,
+      "loss": 0.1056,
+      "step": 7028
+    },
+    {
+      "epoch": 1.138852883992223,
+      "grad_norm": 0.8024566173553467,
+      "learning_rate": 3.5615192720864572e-06,
+      "loss": 0.1007,
+      "step": 7029
+    },
+    {
+      "epoch": 1.1390149060272197,
+      "grad_norm": 0.7231770753860474,
+      "learning_rate": 3.561123330585744e-06,
+      "loss": 0.0895,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1391769280622164,
+      "grad_norm": 0.9641432762145996,
+      "learning_rate": 3.5607273566172255e-06,
+      "loss": 0.1041,
+      "step": 7031
+    },
+    {
+      "epoch": 1.1393389500972133,
+      "grad_norm": 0.7931097745895386,
+      "learning_rate": 3.560331350193016e-06,
+      "loss": 0.1088,
+      "step": 7032
+    },
+    {
+      "epoch": 1.13950097213221,
+      "grad_norm": 0.8860135674476624,
+      "learning_rate": 3.5599353113252343e-06,
+      "loss": 0.1082,
+      "step": 7033
+    },
+    {
+      "epoch": 1.1396629941672067,
+      "grad_norm": 0.8670450448989868,
+      "learning_rate": 3.5595392400259963e-06,
+      "loss": 0.1046,
+      "step": 7034
+    },
+    {
+      "epoch": 1.1398250162022034,
+      "grad_norm": 0.7805025577545166,
+      "learning_rate": 3.5591431363074214e-06,
+      "loss": 0.0926,
+      "step": 7035
+    },
+    {
+      "epoch": 1.1399870382372002,
+      "grad_norm": 0.8997758626937866,
+      "learning_rate": 3.55874700018163e-06,
+      "loss": 0.1134,
+      "step": 7036
+    },
+    {
+      "epoch": 1.140149060272197,
+      "grad_norm": 0.799332857131958,
+      "learning_rate": 3.558350831660742e-06,
+      "loss": 0.0953,
+      "step": 7037
+    },
+    {
+      "epoch": 1.1403110823071938,
+      "grad_norm": 0.8448926210403442,
+      "learning_rate": 3.5579546307568807e-06,
+      "loss": 0.1027,
+      "step": 7038
+    },
+    {
+      "epoch": 1.1404731043421905,
+      "grad_norm": 0.8926445245742798,
+      "learning_rate": 3.557558397482167e-06,
+      "loss": 0.1064,
+      "step": 7039
+    },
+    {
+      "epoch": 1.1406351263771872,
+      "grad_norm": 0.9030429124832153,
+      "learning_rate": 3.557162131848726e-06,
+      "loss": 0.1143,
+      "step": 7040
+    },
+    {
+      "epoch": 1.140797148412184,
+      "grad_norm": 0.8533268570899963,
+      "learning_rate": 3.556765833868682e-06,
+      "loss": 0.1089,
+      "step": 7041
+    },
+    {
+      "epoch": 1.1409591704471809,
+      "grad_norm": 0.9436481595039368,
+      "learning_rate": 3.5563695035541607e-06,
+      "loss": 0.1096,
+      "step": 7042
+    },
+    {
+      "epoch": 1.1411211924821776,
+      "grad_norm": 0.7506798505783081,
+      "learning_rate": 3.5559731409172887e-06,
+      "loss": 0.0839,
+      "step": 7043
+    },
+    {
+      "epoch": 1.1412832145171743,
+      "grad_norm": 0.858508288860321,
+      "learning_rate": 3.5555767459701946e-06,
+      "loss": 0.1047,
+      "step": 7044
+    },
+    {
+      "epoch": 1.141445236552171,
+      "grad_norm": 0.7885152697563171,
+      "learning_rate": 3.5551803187250062e-06,
+      "loss": 0.1009,
+      "step": 7045
+    },
+    {
+      "epoch": 1.141607258587168,
+      "grad_norm": 0.7987487316131592,
+      "learning_rate": 3.554783859193853e-06,
+      "loss": 0.0999,
+      "step": 7046
+    },
+    {
+      "epoch": 1.1417692806221647,
+      "grad_norm": 0.8181909322738647,
+      "learning_rate": 3.5543873673888676e-06,
+      "loss": 0.0995,
+      "step": 7047
+    },
+    {
+      "epoch": 1.1419313026571614,
+      "grad_norm": 0.8372443914413452,
+      "learning_rate": 3.5539908433221793e-06,
+      "loss": 0.1056,
+      "step": 7048
+    },
+    {
+      "epoch": 1.142093324692158,
+      "grad_norm": 0.851033627986908,
+      "learning_rate": 3.553594287005922e-06,
+      "loss": 0.1099,
+      "step": 7049
+    },
+    {
+      "epoch": 1.142255346727155,
+      "grad_norm": 0.7364274859428406,
+      "learning_rate": 3.553197698452229e-06,
+      "loss": 0.098,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1424173687621517,
+      "grad_norm": 0.8522658348083496,
+      "learning_rate": 3.5528010776732354e-06,
+      "loss": 0.1083,
+      "step": 7051
+    },
+    {
+      "epoch": 1.1425793907971484,
+      "grad_norm": 0.8110232949256897,
+      "learning_rate": 3.5524044246810764e-06,
+      "loss": 0.0991,
+      "step": 7052
+    },
+    {
+      "epoch": 1.1427414128321451,
+      "grad_norm": 0.9086326956748962,
+      "learning_rate": 3.5520077394878883e-06,
+      "loss": 0.1123,
+      "step": 7053
+    },
+    {
+      "epoch": 1.1429034348671419,
+      "grad_norm": 0.9316560626029968,
+      "learning_rate": 3.5516110221058096e-06,
+      "loss": 0.1043,
+      "step": 7054
+    },
+    {
+      "epoch": 1.1430654569021388,
+      "grad_norm": 0.8712682723999023,
+      "learning_rate": 3.551214272546979e-06,
+      "loss": 0.1018,
+      "step": 7055
+    },
+    {
+      "epoch": 1.1432274789371355,
+      "grad_norm": 0.8193374872207642,
+      "learning_rate": 3.550817490823535e-06,
+      "loss": 0.1063,
+      "step": 7056
+    },
+    {
+      "epoch": 1.1433895009721322,
+      "grad_norm": 0.7940148115158081,
+      "learning_rate": 3.550420676947619e-06,
+      "loss": 0.0989,
+      "step": 7057
+    },
+    {
+      "epoch": 1.143551523007129,
+      "grad_norm": 1.0240070819854736,
+      "learning_rate": 3.5500238309313717e-06,
+      "loss": 0.0994,
+      "step": 7058
+    },
+    {
+      "epoch": 1.1437135450421256,
+      "grad_norm": 0.8923311233520508,
+      "learning_rate": 3.549626952786937e-06,
+      "loss": 0.1026,
+      "step": 7059
+    },
+    {
+      "epoch": 1.1438755670771226,
+      "grad_norm": 0.8232337236404419,
+      "learning_rate": 3.5492300425264574e-06,
+      "loss": 0.099,
+      "step": 7060
+    },
+    {
+      "epoch": 1.1440375891121193,
+      "grad_norm": 0.8222522139549255,
+      "learning_rate": 3.548833100162077e-06,
+      "loss": 0.0961,
+      "step": 7061
+    },
+    {
+      "epoch": 1.144199611147116,
+      "grad_norm": 0.8657082319259644,
+      "learning_rate": 3.5484361257059425e-06,
+      "loss": 0.1066,
+      "step": 7062
+    },
+    {
+      "epoch": 1.1443616331821127,
+      "grad_norm": 0.9332712888717651,
+      "learning_rate": 3.548039119170199e-06,
+      "loss": 0.1138,
+      "step": 7063
+    },
+    {
+      "epoch": 1.1445236552171094,
+      "grad_norm": 1.0337094068527222,
+      "learning_rate": 3.5476420805669953e-06,
+      "loss": 0.1053,
+      "step": 7064
+    },
+    {
+      "epoch": 1.1446856772521063,
+      "grad_norm": 0.874484658241272,
+      "learning_rate": 3.5472450099084786e-06,
+      "loss": 0.1125,
+      "step": 7065
+    },
+    {
+      "epoch": 1.144847699287103,
+      "grad_norm": 0.7913826704025269,
+      "learning_rate": 3.5468479072067996e-06,
+      "loss": 0.1014,
+      "step": 7066
+    },
+    {
+      "epoch": 1.1450097213220998,
+      "grad_norm": 0.8043907880783081,
+      "learning_rate": 3.5464507724741076e-06,
+      "loss": 0.1029,
+      "step": 7067
+    },
+    {
+      "epoch": 1.1451717433570965,
+      "grad_norm": 0.833845317363739,
+      "learning_rate": 3.5460536057225542e-06,
+      "loss": 0.1024,
+      "step": 7068
+    },
+    {
+      "epoch": 1.1453337653920934,
+      "grad_norm": 0.8379101753234863,
+      "learning_rate": 3.545656406964292e-06,
+      "loss": 0.1051,
+      "step": 7069
+    },
+    {
+      "epoch": 1.1454957874270901,
+      "grad_norm": 0.7530349493026733,
+      "learning_rate": 3.545259176211474e-06,
+      "loss": 0.0899,
+      "step": 7070
+    },
+    {
+      "epoch": 1.1456578094620868,
+      "grad_norm": 0.7313252687454224,
+      "learning_rate": 3.5448619134762552e-06,
+      "loss": 0.09,
+      "step": 7071
+    },
+    {
+      "epoch": 1.1458198314970836,
+      "grad_norm": 0.8761491179466248,
+      "learning_rate": 3.5444646187707897e-06,
+      "loss": 0.106,
+      "step": 7072
+    },
+    {
+      "epoch": 1.1459818535320805,
+      "grad_norm": 0.8301008343696594,
+      "learning_rate": 3.544067292107235e-06,
+      "loss": 0.1041,
+      "step": 7073
+    },
+    {
+      "epoch": 1.1461438755670772,
+      "grad_norm": 0.8845946788787842,
+      "learning_rate": 3.5436699334977476e-06,
+      "loss": 0.1079,
+      "step": 7074
+    },
+    {
+      "epoch": 1.146305897602074,
+      "grad_norm": 0.9687334299087524,
+      "learning_rate": 3.5432725429544856e-06,
+      "loss": 0.1142,
+      "step": 7075
+    },
+    {
+      "epoch": 1.1464679196370706,
+      "grad_norm": 0.9008753299713135,
+      "learning_rate": 3.5428751204896083e-06,
+      "loss": 0.1243,
+      "step": 7076
+    },
+    {
+      "epoch": 1.1466299416720673,
+      "grad_norm": 0.8190997838973999,
+      "learning_rate": 3.5424776661152766e-06,
+      "loss": 0.1043,
+      "step": 7077
+    },
+    {
+      "epoch": 1.1467919637070643,
+      "grad_norm": 0.7773595452308655,
+      "learning_rate": 3.542080179843651e-06,
+      "loss": 0.0952,
+      "step": 7078
+    },
+    {
+      "epoch": 1.146953985742061,
+      "grad_norm": 0.8167815208435059,
+      "learning_rate": 3.5416826616868938e-06,
+      "loss": 0.1005,
+      "step": 7079
+    },
+    {
+      "epoch": 1.1471160077770577,
+      "grad_norm": 0.881354570388794,
+      "learning_rate": 3.5412851116571673e-06,
+      "loss": 0.1099,
+      "step": 7080
+    },
+    {
+      "epoch": 1.1472780298120544,
+      "grad_norm": 0.9158351421356201,
+      "learning_rate": 3.5408875297666366e-06,
+      "loss": 0.1125,
+      "step": 7081
+    },
+    {
+      "epoch": 1.1474400518470511,
+      "grad_norm": 0.8541279435157776,
+      "learning_rate": 3.5404899160274664e-06,
+      "loss": 0.0947,
+      "step": 7082
+    },
+    {
+      "epoch": 1.147602073882048,
+      "grad_norm": 0.9512801170349121,
+      "learning_rate": 3.5400922704518225e-06,
+      "loss": 0.1134,
+      "step": 7083
+    },
+    {
+      "epoch": 1.1477640959170448,
+      "grad_norm": 0.8026208281517029,
+      "learning_rate": 3.5396945930518722e-06,
+      "loss": 0.1004,
+      "step": 7084
+    },
+    {
+      "epoch": 1.1479261179520415,
+      "grad_norm": 0.8425459861755371,
+      "learning_rate": 3.5392968838397836e-06,
+      "loss": 0.1065,
+      "step": 7085
+    },
+    {
+      "epoch": 1.1480881399870382,
+      "grad_norm": 0.9077960252761841,
+      "learning_rate": 3.538899142827726e-06,
+      "loss": 0.1203,
+      "step": 7086
+    },
+    {
+      "epoch": 1.148250162022035,
+      "grad_norm": 0.8521792888641357,
+      "learning_rate": 3.538501370027867e-06,
+      "loss": 0.1073,
+      "step": 7087
+    },
+    {
+      "epoch": 1.1484121840570318,
+      "grad_norm": 0.7184135913848877,
+      "learning_rate": 3.538103565452381e-06,
+      "loss": 0.0894,
+      "step": 7088
+    },
+    {
+      "epoch": 1.1485742060920285,
+      "grad_norm": 0.8005532622337341,
+      "learning_rate": 3.537705729113437e-06,
+      "loss": 0.1096,
+      "step": 7089
+    },
+    {
+      "epoch": 1.1487362281270252,
+      "grad_norm": 0.8464052081108093,
+      "learning_rate": 3.537307861023209e-06,
+      "loss": 0.1035,
+      "step": 7090
+    },
+    {
+      "epoch": 1.148898250162022,
+      "grad_norm": 0.7864227890968323,
+      "learning_rate": 3.536909961193872e-06,
+      "loss": 0.0955,
+      "step": 7091
+    },
+    {
+      "epoch": 1.1490602721970187,
+      "grad_norm": 0.923495352268219,
+      "learning_rate": 3.536512029637597e-06,
+      "loss": 0.1104,
+      "step": 7092
+    },
+    {
+      "epoch": 1.1492222942320156,
+      "grad_norm": 0.732354998588562,
+      "learning_rate": 3.5361140663665644e-06,
+      "loss": 0.0921,
+      "step": 7093
+    },
+    {
+      "epoch": 1.1493843162670123,
+      "grad_norm": 0.8492215871810913,
+      "learning_rate": 3.5357160713929473e-06,
+      "loss": 0.1059,
+      "step": 7094
+    },
+    {
+      "epoch": 1.149546338302009,
+      "grad_norm": 0.899074137210846,
+      "learning_rate": 3.5353180447289253e-06,
+      "loss": 0.1078,
+      "step": 7095
+    },
+    {
+      "epoch": 1.1497083603370057,
+      "grad_norm": 0.8724895119667053,
+      "learning_rate": 3.534919986386676e-06,
+      "loss": 0.1126,
+      "step": 7096
+    },
+    {
+      "epoch": 1.1498703823720027,
+      "grad_norm": 0.9067556262016296,
+      "learning_rate": 3.534521896378381e-06,
+      "loss": 0.1107,
+      "step": 7097
+    },
+    {
+      "epoch": 1.1500324044069994,
+      "grad_norm": 0.8136131763458252,
+      "learning_rate": 3.5341237747162183e-06,
+      "loss": 0.1047,
+      "step": 7098
+    },
+    {
+      "epoch": 1.150194426441996,
+      "grad_norm": 0.8796071410179138,
+      "learning_rate": 3.533725621412371e-06,
+      "loss": 0.1117,
+      "step": 7099
+    },
+    {
+      "epoch": 1.1503564484769928,
+      "grad_norm": 0.8252900242805481,
+      "learning_rate": 3.533327436479021e-06,
+      "loss": 0.1106,
+      "step": 7100
+    },
+    {
+      "epoch": 1.1505184705119897,
+      "grad_norm": 0.8371667861938477,
+      "learning_rate": 3.5329292199283522e-06,
+      "loss": 0.1012,
+      "step": 7101
+    },
+    {
+      "epoch": 1.1506804925469865,
+      "grad_norm": 0.8886693716049194,
+      "learning_rate": 3.53253097177255e-06,
+      "loss": 0.1118,
+      "step": 7102
+    },
+    {
+      "epoch": 1.1508425145819832,
+      "grad_norm": 0.8799384832382202,
+      "learning_rate": 3.532132692023797e-06,
+      "loss": 0.114,
+      "step": 7103
+    },
+    {
+      "epoch": 1.1510045366169799,
+      "grad_norm": 0.815346896648407,
+      "learning_rate": 3.531734380694282e-06,
+      "loss": 0.1071,
+      "step": 7104
+    },
+    {
+      "epoch": 1.1511665586519766,
+      "grad_norm": 0.8964957594871521,
+      "learning_rate": 3.531336037796192e-06,
+      "loss": 0.1094,
+      "step": 7105
+    },
+    {
+      "epoch": 1.1513285806869735,
+      "grad_norm": 0.8892197012901306,
+      "learning_rate": 3.5309376633417146e-06,
+      "loss": 0.1091,
+      "step": 7106
+    },
+    {
+      "epoch": 1.1514906027219702,
+      "grad_norm": 0.9133774042129517,
+      "learning_rate": 3.5305392573430398e-06,
+      "loss": 0.1154,
+      "step": 7107
+    },
+    {
+      "epoch": 1.151652624756967,
+      "grad_norm": 1.0447365045547485,
+      "learning_rate": 3.530140819812357e-06,
+      "loss": 0.1256,
+      "step": 7108
+    },
+    {
+      "epoch": 1.1518146467919637,
+      "grad_norm": 0.8976051807403564,
+      "learning_rate": 3.5297423507618582e-06,
+      "loss": 0.107,
+      "step": 7109
+    },
+    {
+      "epoch": 1.1519766688269604,
+      "grad_norm": 0.7730125784873962,
+      "learning_rate": 3.5293438502037363e-06,
+      "loss": 0.0979,
+      "step": 7110
+    },
+    {
+      "epoch": 1.1521386908619573,
+      "grad_norm": 0.9154132008552551,
+      "learning_rate": 3.5289453181501832e-06,
+      "loss": 0.1154,
+      "step": 7111
+    },
+    {
+      "epoch": 1.152300712896954,
+      "grad_norm": 0.731338620185852,
+      "learning_rate": 3.5285467546133926e-06,
+      "loss": 0.0924,
+      "step": 7112
+    },
+    {
+      "epoch": 1.1524627349319507,
+      "grad_norm": 0.8675764203071594,
+      "learning_rate": 3.5281481596055613e-06,
+      "loss": 0.1181,
+      "step": 7113
+    },
+    {
+      "epoch": 1.1526247569669474,
+      "grad_norm": 0.7357743382453918,
+      "learning_rate": 3.5277495331388835e-06,
+      "loss": 0.0895,
+      "step": 7114
+    },
+    {
+      "epoch": 1.1527867790019442,
+      "grad_norm": 0.8133664131164551,
+      "learning_rate": 3.527350875225558e-06,
+      "loss": 0.1026,
+      "step": 7115
+    },
+    {
+      "epoch": 1.152948801036941,
+      "grad_norm": 0.8403448462486267,
+      "learning_rate": 3.526952185877781e-06,
+      "loss": 0.1002,
+      "step": 7116
+    },
+    {
+      "epoch": 1.1531108230719378,
+      "grad_norm": 0.7883726358413696,
+      "learning_rate": 3.5265534651077527e-06,
+      "loss": 0.1038,
+      "step": 7117
+    },
+    {
+      "epoch": 1.1532728451069345,
+      "grad_norm": 0.9349051117897034,
+      "learning_rate": 3.526154712927672e-06,
+      "loss": 0.1239,
+      "step": 7118
+    },
+    {
+      "epoch": 1.1534348671419312,
+      "grad_norm": 0.8447469472885132,
+      "learning_rate": 3.525755929349741e-06,
+      "loss": 0.1049,
+      "step": 7119
+    },
+    {
+      "epoch": 1.1535968891769282,
+      "grad_norm": 0.8494859933853149,
+      "learning_rate": 3.525357114386161e-06,
+      "loss": 0.1031,
+      "step": 7120
+    },
+    {
+      "epoch": 1.1537589112119249,
+      "grad_norm": 0.7881488800048828,
+      "learning_rate": 3.5249582680491346e-06,
+      "loss": 0.102,
+      "step": 7121
+    },
+    {
+      "epoch": 1.1539209332469216,
+      "grad_norm": 0.7983292937278748,
+      "learning_rate": 3.524559390350865e-06,
+      "loss": 0.0992,
+      "step": 7122
+    },
+    {
+      "epoch": 1.1540829552819183,
+      "grad_norm": 0.8363736867904663,
+      "learning_rate": 3.5241604813035577e-06,
+      "loss": 0.1017,
+      "step": 7123
+    },
+    {
+      "epoch": 1.1542449773169152,
+      "grad_norm": 0.9091150760650635,
+      "learning_rate": 3.523761540919418e-06,
+      "loss": 0.116,
+      "step": 7124
+    },
+    {
+      "epoch": 1.154406999351912,
+      "grad_norm": 0.8165498971939087,
+      "learning_rate": 3.5233625692106525e-06,
+      "loss": 0.1047,
+      "step": 7125
+    },
+    {
+      "epoch": 1.1545690213869086,
+      "grad_norm": 0.7147705554962158,
+      "learning_rate": 3.5229635661894696e-06,
+      "loss": 0.0823,
+      "step": 7126
+    },
+    {
+      "epoch": 1.1547310434219054,
+      "grad_norm": 0.7142750024795532,
+      "learning_rate": 3.5225645318680768e-06,
+      "loss": 0.0944,
+      "step": 7127
+    },
+    {
+      "epoch": 1.154893065456902,
+      "grad_norm": 0.795220673084259,
+      "learning_rate": 3.5221654662586837e-06,
+      "loss": 0.0923,
+      "step": 7128
+    },
+    {
+      "epoch": 1.155055087491899,
+      "grad_norm": 0.7979345321655273,
+      "learning_rate": 3.5217663693735006e-06,
+      "loss": 0.0994,
+      "step": 7129
+    },
+    {
+      "epoch": 1.1552171095268957,
+      "grad_norm": 0.856704592704773,
+      "learning_rate": 3.521367241224739e-06,
+      "loss": 0.1009,
+      "step": 7130
+    },
+    {
+      "epoch": 1.1553791315618924,
+      "grad_norm": 0.7614887356758118,
+      "learning_rate": 3.5209680818246124e-06,
+      "loss": 0.0994,
+      "step": 7131
+    },
+    {
+      "epoch": 1.1555411535968891,
+      "grad_norm": 0.8549346923828125,
+      "learning_rate": 3.5205688911853326e-06,
+      "loss": 0.098,
+      "step": 7132
+    },
+    {
+      "epoch": 1.1557031756318858,
+      "grad_norm": 0.7560814619064331,
+      "learning_rate": 3.520169669319115e-06,
+      "loss": 0.0975,
+      "step": 7133
+    },
+    {
+      "epoch": 1.1558651976668828,
+      "grad_norm": 0.754621148109436,
+      "learning_rate": 3.5197704162381742e-06,
+      "loss": 0.0918,
+      "step": 7134
+    },
+    {
+      "epoch": 1.1560272197018795,
+      "grad_norm": 0.878743588924408,
+      "learning_rate": 3.5193711319547257e-06,
+      "loss": 0.1169,
+      "step": 7135
+    },
+    {
+      "epoch": 1.1561892417368762,
+      "grad_norm": 0.7919618487358093,
+      "learning_rate": 3.5189718164809884e-06,
+      "loss": 0.1018,
+      "step": 7136
+    },
+    {
+      "epoch": 1.156351263771873,
+      "grad_norm": 0.8777773976325989,
+      "learning_rate": 3.5185724698291784e-06,
+      "loss": 0.1095,
+      "step": 7137
+    },
+    {
+      "epoch": 1.1565132858068696,
+      "grad_norm": 0.9273558855056763,
+      "learning_rate": 3.5181730920115165e-06,
+      "loss": 0.1212,
+      "step": 7138
+    },
+    {
+      "epoch": 1.1566753078418666,
+      "grad_norm": 0.7581186890602112,
+      "learning_rate": 3.5177736830402213e-06,
+      "loss": 0.1012,
+      "step": 7139
+    },
+    {
+      "epoch": 1.1568373298768633,
+      "grad_norm": 0.8174877762794495,
+      "learning_rate": 3.517374242927514e-06,
+      "loss": 0.107,
+      "step": 7140
+    },
+    {
+      "epoch": 1.15699935191186,
+      "grad_norm": 0.8098114132881165,
+      "learning_rate": 3.5169747716856186e-06,
+      "loss": 0.1038,
+      "step": 7141
+    },
+    {
+      "epoch": 1.1571613739468567,
+      "grad_norm": 0.8155732154846191,
+      "learning_rate": 3.516575269326755e-06,
+      "loss": 0.0981,
+      "step": 7142
+    },
+    {
+      "epoch": 1.1573233959818534,
+      "grad_norm": 0.8303235769271851,
+      "learning_rate": 3.5161757358631482e-06,
+      "loss": 0.1058,
+      "step": 7143
+    },
+    {
+      "epoch": 1.1574854180168503,
+      "grad_norm": 0.7622935771942139,
+      "learning_rate": 3.515776171307023e-06,
+      "loss": 0.0971,
+      "step": 7144
+    },
+    {
+      "epoch": 1.157647440051847,
+      "grad_norm": 0.7291949391365051,
+      "learning_rate": 3.5153765756706054e-06,
+      "loss": 0.0866,
+      "step": 7145
+    },
+    {
+      "epoch": 1.1578094620868438,
+      "grad_norm": 0.7386295199394226,
+      "learning_rate": 3.5149769489661216e-06,
+      "loss": 0.095,
+      "step": 7146
+    },
+    {
+      "epoch": 1.1579714841218405,
+      "grad_norm": 0.7064818143844604,
+      "learning_rate": 3.514577291205799e-06,
+      "loss": 0.0891,
+      "step": 7147
+    },
+    {
+      "epoch": 1.1581335061568374,
+      "grad_norm": 0.6918752193450928,
+      "learning_rate": 3.5141776024018676e-06,
+      "loss": 0.0936,
+      "step": 7148
+    },
+    {
+      "epoch": 1.1582955281918341,
+      "grad_norm": 0.750155508518219,
+      "learning_rate": 3.5137778825665542e-06,
+      "loss": 0.0925,
+      "step": 7149
+    },
+    {
+      "epoch": 1.1584575502268308,
+      "grad_norm": 0.7644116878509521,
+      "learning_rate": 3.513378131712092e-06,
+      "loss": 0.0917,
+      "step": 7150
+    },
+    {
+      "epoch": 1.1586195722618275,
+      "grad_norm": 0.8276258707046509,
+      "learning_rate": 3.5129783498507114e-06,
+      "loss": 0.0993,
+      "step": 7151
+    },
+    {
+      "epoch": 1.1587815942968245,
+      "grad_norm": 0.9577162265777588,
+      "learning_rate": 3.5125785369946442e-06,
+      "loss": 0.1086,
+      "step": 7152
+    },
+    {
+      "epoch": 1.1589436163318212,
+      "grad_norm": 0.9136195182800293,
+      "learning_rate": 3.512178693156124e-06,
+      "loss": 0.1052,
+      "step": 7153
+    },
+    {
+      "epoch": 1.159105638366818,
+      "grad_norm": 0.9026337265968323,
+      "learning_rate": 3.5117788183473856e-06,
+      "loss": 0.1135,
+      "step": 7154
+    },
+    {
+      "epoch": 1.1592676604018146,
+      "grad_norm": 0.836702823638916,
+      "learning_rate": 3.511378912580664e-06,
+      "loss": 0.106,
+      "step": 7155
+    },
+    {
+      "epoch": 1.1594296824368113,
+      "grad_norm": 0.7680188417434692,
+      "learning_rate": 3.5109789758681944e-06,
+      "loss": 0.0917,
+      "step": 7156
+    },
+    {
+      "epoch": 1.1595917044718083,
+      "grad_norm": 0.9207019209861755,
+      "learning_rate": 3.5105790082222157e-06,
+      "loss": 0.1087,
+      "step": 7157
+    },
+    {
+      "epoch": 1.159753726506805,
+      "grad_norm": 0.7964971661567688,
+      "learning_rate": 3.5101790096549643e-06,
+      "loss": 0.0916,
+      "step": 7158
+    },
+    {
+      "epoch": 1.1599157485418017,
+      "grad_norm": 0.8500689268112183,
+      "learning_rate": 3.5097789801786797e-06,
+      "loss": 0.1081,
+      "step": 7159
+    },
+    {
+      "epoch": 1.1600777705767984,
+      "grad_norm": 0.8392435908317566,
+      "learning_rate": 3.509378919805602e-06,
+      "loss": 0.095,
+      "step": 7160
+    },
+    {
+      "epoch": 1.160239792611795,
+      "grad_norm": 0.7941946983337402,
+      "learning_rate": 3.508978828547972e-06,
+      "loss": 0.0961,
+      "step": 7161
+    },
+    {
+      "epoch": 1.160401814646792,
+      "grad_norm": 0.800559937953949,
+      "learning_rate": 3.5085787064180317e-06,
+      "loss": 0.1083,
+      "step": 7162
+    },
+    {
+      "epoch": 1.1605638366817888,
+      "grad_norm": 0.8886862993240356,
+      "learning_rate": 3.5081785534280233e-06,
+      "loss": 0.1146,
+      "step": 7163
+    },
+    {
+      "epoch": 1.1607258587167855,
+      "grad_norm": 0.8028414249420166,
+      "learning_rate": 3.5077783695901917e-06,
+      "loss": 0.1006,
+      "step": 7164
+    },
+    {
+      "epoch": 1.1608878807517822,
+      "grad_norm": 0.7642043828964233,
+      "learning_rate": 3.507378154916781e-06,
+      "loss": 0.0978,
+      "step": 7165
+    },
+    {
+      "epoch": 1.1610499027867789,
+      "grad_norm": 0.7976152896881104,
+      "learning_rate": 3.506977909420035e-06,
+      "loss": 0.0988,
+      "step": 7166
+    },
+    {
+      "epoch": 1.1612119248217758,
+      "grad_norm": 0.8192991018295288,
+      "learning_rate": 3.5065776331122027e-06,
+      "loss": 0.1036,
+      "step": 7167
+    },
+    {
+      "epoch": 1.1613739468567725,
+      "grad_norm": 0.7200789451599121,
+      "learning_rate": 3.506177326005531e-06,
+      "loss": 0.0982,
+      "step": 7168
+    },
+    {
+      "epoch": 1.1615359688917692,
+      "grad_norm": 0.7286558151245117,
+      "learning_rate": 3.5057769881122674e-06,
+      "loss": 0.1023,
+      "step": 7169
+    },
+    {
+      "epoch": 1.161697990926766,
+      "grad_norm": 0.9604185223579407,
+      "learning_rate": 3.5053766194446626e-06,
+      "loss": 0.1162,
+      "step": 7170
+    },
+    {
+      "epoch": 1.161860012961763,
+      "grad_norm": 0.8336222171783447,
+      "learning_rate": 3.504976220014965e-06,
+      "loss": 0.098,
+      "step": 7171
+    },
+    {
+      "epoch": 1.1620220349967596,
+      "grad_norm": 0.7740049362182617,
+      "learning_rate": 3.504575789835428e-06,
+      "loss": 0.1082,
+      "step": 7172
+    },
+    {
+      "epoch": 1.1621840570317563,
+      "grad_norm": 0.8649178147315979,
+      "learning_rate": 3.5041753289183033e-06,
+      "loss": 0.107,
+      "step": 7173
+    },
+    {
+      "epoch": 1.162346079066753,
+      "grad_norm": 0.8420133590698242,
+      "learning_rate": 3.503774837275843e-06,
+      "loss": 0.1032,
+      "step": 7174
+    },
+    {
+      "epoch": 1.16250810110175,
+      "grad_norm": 0.8649719953536987,
+      "learning_rate": 3.5033743149203013e-06,
+      "loss": 0.1087,
+      "step": 7175
+    },
+    {
+      "epoch": 1.1626701231367467,
+      "grad_norm": 0.8566991686820984,
+      "learning_rate": 3.5029737618639344e-06,
+      "loss": 0.0972,
+      "step": 7176
+    },
+    {
+      "epoch": 1.1628321451717434,
+      "grad_norm": 0.7619419097900391,
+      "learning_rate": 3.5025731781189976e-06,
+      "loss": 0.0886,
+      "step": 7177
+    },
+    {
+      "epoch": 1.16299416720674,
+      "grad_norm": 0.7711379528045654,
+      "learning_rate": 3.5021725636977466e-06,
+      "loss": 0.0956,
+      "step": 7178
+    },
+    {
+      "epoch": 1.1631561892417368,
+      "grad_norm": 0.9927428364753723,
+      "learning_rate": 3.501771918612442e-06,
+      "loss": 0.1124,
+      "step": 7179
+    },
+    {
+      "epoch": 1.1633182112767337,
+      "grad_norm": 0.776023268699646,
+      "learning_rate": 3.5013712428753392e-06,
+      "loss": 0.0964,
+      "step": 7180
+    },
+    {
+      "epoch": 1.1634802333117304,
+      "grad_norm": 0.9973410367965698,
+      "learning_rate": 3.500970536498701e-06,
+      "loss": 0.1263,
+      "step": 7181
+    },
+    {
+      "epoch": 1.1636422553467272,
+      "grad_norm": 0.8526068329811096,
+      "learning_rate": 3.500569799494786e-06,
+      "loss": 0.1016,
+      "step": 7182
+    },
+    {
+      "epoch": 1.1638042773817239,
+      "grad_norm": 0.8753652572631836,
+      "learning_rate": 3.500169031875857e-06,
+      "loss": 0.0948,
+      "step": 7183
+    },
+    {
+      "epoch": 1.1639662994167206,
+      "grad_norm": 0.8186274170875549,
+      "learning_rate": 3.4997682336541756e-06,
+      "loss": 0.1025,
+      "step": 7184
+    },
+    {
+      "epoch": 1.1641283214517175,
+      "grad_norm": 0.7919983267784119,
+      "learning_rate": 3.4993674048420056e-06,
+      "loss": 0.1053,
+      "step": 7185
+    },
+    {
+      "epoch": 1.1642903434867142,
+      "grad_norm": 0.8245944976806641,
+      "learning_rate": 3.498966545451612e-06,
+      "loss": 0.0985,
+      "step": 7186
+    },
+    {
+      "epoch": 1.164452365521711,
+      "grad_norm": 0.8587673902511597,
+      "learning_rate": 3.4985656554952583e-06,
+      "loss": 0.1062,
+      "step": 7187
+    },
+    {
+      "epoch": 1.1646143875567077,
+      "grad_norm": 0.9467253088951111,
+      "learning_rate": 3.4981647349852137e-06,
+      "loss": 0.1154,
+      "step": 7188
+    },
+    {
+      "epoch": 1.1647764095917044,
+      "grad_norm": 0.8093448877334595,
+      "learning_rate": 3.497763783933743e-06,
+      "loss": 0.1013,
+      "step": 7189
+    },
+    {
+      "epoch": 1.1649384316267013,
+      "grad_norm": 0.8070668578147888,
+      "learning_rate": 3.4973628023531146e-06,
+      "loss": 0.0967,
+      "step": 7190
+    },
+    {
+      "epoch": 1.165100453661698,
+      "grad_norm": 0.8548862934112549,
+      "learning_rate": 3.4969617902555984e-06,
+      "loss": 0.0957,
+      "step": 7191
+    },
+    {
+      "epoch": 1.1652624756966947,
+      "grad_norm": 0.8574987649917603,
+      "learning_rate": 3.496560747653464e-06,
+      "loss": 0.0881,
+      "step": 7192
+    },
+    {
+      "epoch": 1.1654244977316914,
+      "grad_norm": 0.7991284132003784,
+      "learning_rate": 3.496159674558982e-06,
+      "loss": 0.0937,
+      "step": 7193
+    },
+    {
+      "epoch": 1.1655865197666881,
+      "grad_norm": 0.9537588357925415,
+      "learning_rate": 3.4957585709844254e-06,
+      "loss": 0.1166,
+      "step": 7194
+    },
+    {
+      "epoch": 1.165748541801685,
+      "grad_norm": 0.7716676592826843,
+      "learning_rate": 3.495357436942066e-06,
+      "loss": 0.0916,
+      "step": 7195
+    },
+    {
+      "epoch": 1.1659105638366818,
+      "grad_norm": 0.8516098856925964,
+      "learning_rate": 3.494956272444177e-06,
+      "loss": 0.1054,
+      "step": 7196
+    },
+    {
+      "epoch": 1.1660725858716785,
+      "grad_norm": 0.7702400088310242,
+      "learning_rate": 3.4945550775030346e-06,
+      "loss": 0.0919,
+      "step": 7197
+    },
+    {
+      "epoch": 1.1662346079066752,
+      "grad_norm": 0.8230445981025696,
+      "learning_rate": 3.494153852130913e-06,
+      "loss": 0.102,
+      "step": 7198
+    },
+    {
+      "epoch": 1.1663966299416721,
+      "grad_norm": 0.8861101269721985,
+      "learning_rate": 3.49375259634009e-06,
+      "loss": 0.1092,
+      "step": 7199
+    },
+    {
+      "epoch": 1.1665586519766689,
+      "grad_norm": 0.8347850441932678,
+      "learning_rate": 3.4933513101428416e-06,
+      "loss": 0.1064,
+      "step": 7200
+    },
+    {
+      "epoch": 1.1667206740116656,
+      "grad_norm": 0.7879016399383545,
+      "learning_rate": 3.492949993551448e-06,
+      "loss": 0.0927,
+      "step": 7201
+    },
+    {
+      "epoch": 1.1668826960466623,
+      "grad_norm": 0.8113725781440735,
+      "learning_rate": 3.4925486465781865e-06,
+      "loss": 0.0954,
+      "step": 7202
+    },
+    {
+      "epoch": 1.1670447180816592,
+      "grad_norm": 0.709119439125061,
+      "learning_rate": 3.492147269235339e-06,
+      "loss": 0.092,
+      "step": 7203
+    },
+    {
+      "epoch": 1.167206740116656,
+      "grad_norm": 1.0254310369491577,
+      "learning_rate": 3.4917458615351853e-06,
+      "loss": 0.1181,
+      "step": 7204
+    },
+    {
+      "epoch": 1.1673687621516526,
+      "grad_norm": 0.7908617258071899,
+      "learning_rate": 3.4913444234900092e-06,
+      "loss": 0.0991,
+      "step": 7205
+    },
+    {
+      "epoch": 1.1675307841866494,
+      "grad_norm": 0.9039345979690552,
+      "learning_rate": 3.490942955112092e-06,
+      "loss": 0.1172,
+      "step": 7206
+    },
+    {
+      "epoch": 1.167692806221646,
+      "grad_norm": 0.7758670449256897,
+      "learning_rate": 3.4905414564137187e-06,
+      "loss": 0.0894,
+      "step": 7207
+    },
+    {
+      "epoch": 1.167854828256643,
+      "grad_norm": 0.9767671823501587,
+      "learning_rate": 3.490139927407174e-06,
+      "loss": 0.1139,
+      "step": 7208
+    },
+    {
+      "epoch": 1.1680168502916397,
+      "grad_norm": 1.0379667282104492,
+      "learning_rate": 3.489738368104743e-06,
+      "loss": 0.1256,
+      "step": 7209
+    },
+    {
+      "epoch": 1.1681788723266364,
+      "grad_norm": 0.9443620443344116,
+      "learning_rate": 3.4893367785187137e-06,
+      "loss": 0.1217,
+      "step": 7210
+    },
+    {
+      "epoch": 1.1683408943616331,
+      "grad_norm": 0.8020828366279602,
+      "learning_rate": 3.488935158661373e-06,
+      "loss": 0.1089,
+      "step": 7211
+    },
+    {
+      "epoch": 1.1685029163966298,
+      "grad_norm": 0.8911034464836121,
+      "learning_rate": 3.4885335085450095e-06,
+      "loss": 0.1104,
+      "step": 7212
+    },
+    {
+      "epoch": 1.1686649384316268,
+      "grad_norm": 0.8075478672981262,
+      "learning_rate": 3.4881318281819134e-06,
+      "loss": 0.0951,
+      "step": 7213
+    },
+    {
+      "epoch": 1.1688269604666235,
+      "grad_norm": 0.9641019701957703,
+      "learning_rate": 3.4877301175843735e-06,
+      "loss": 0.1115,
+      "step": 7214
+    },
+    {
+      "epoch": 1.1689889825016202,
+      "grad_norm": 0.8437277674674988,
+      "learning_rate": 3.4873283767646828e-06,
+      "loss": 0.1074,
+      "step": 7215
+    },
+    {
+      "epoch": 1.169151004536617,
+      "grad_norm": 0.7787491679191589,
+      "learning_rate": 3.486926605735133e-06,
+      "loss": 0.0975,
+      "step": 7216
+    },
+    {
+      "epoch": 1.1693130265716136,
+      "grad_norm": 0.9039609432220459,
+      "learning_rate": 3.486524804508018e-06,
+      "loss": 0.1196,
+      "step": 7217
+    },
+    {
+      "epoch": 1.1694750486066106,
+      "grad_norm": 0.8110989928245544,
+      "learning_rate": 3.486122973095631e-06,
+      "loss": 0.0996,
+      "step": 7218
+    },
+    {
+      "epoch": 1.1696370706416073,
+      "grad_norm": 0.7526705265045166,
+      "learning_rate": 3.485721111510267e-06,
+      "loss": 0.1038,
+      "step": 7219
+    },
+    {
+      "epoch": 1.169799092676604,
+      "grad_norm": 0.9018827676773071,
+      "learning_rate": 3.4853192197642226e-06,
+      "loss": 0.124,
+      "step": 7220
+    },
+    {
+      "epoch": 1.1699611147116007,
+      "grad_norm": 0.8486402034759521,
+      "learning_rate": 3.4849172978697942e-06,
+      "loss": 0.1047,
+      "step": 7221
+    },
+    {
+      "epoch": 1.1701231367465976,
+      "grad_norm": 0.9394757151603699,
+      "learning_rate": 3.48451534583928e-06,
+      "loss": 0.1075,
+      "step": 7222
+    },
+    {
+      "epoch": 1.1702851587815943,
+      "grad_norm": 0.8895528316497803,
+      "learning_rate": 3.4841133636849787e-06,
+      "loss": 0.1223,
+      "step": 7223
+    },
+    {
+      "epoch": 1.170447180816591,
+      "grad_norm": 0.8455293774604797,
+      "learning_rate": 3.4837113514191907e-06,
+      "loss": 0.1074,
+      "step": 7224
+    },
+    {
+      "epoch": 1.1706092028515878,
+      "grad_norm": 0.9192505478858948,
+      "learning_rate": 3.483309309054216e-06,
+      "loss": 0.1201,
+      "step": 7225
+    },
+    {
+      "epoch": 1.1707712248865847,
+      "grad_norm": 0.9902666211128235,
+      "learning_rate": 3.482907236602354e-06,
+      "loss": 0.1139,
+      "step": 7226
+    },
+    {
+      "epoch": 1.1709332469215814,
+      "grad_norm": 0.8750571608543396,
+      "learning_rate": 3.4825051340759114e-06,
+      "loss": 0.1164,
+      "step": 7227
+    },
+    {
+      "epoch": 1.1710952689565781,
+      "grad_norm": 0.793886125087738,
+      "learning_rate": 3.4821030014871886e-06,
+      "loss": 0.1018,
+      "step": 7228
+    },
+    {
+      "epoch": 1.1712572909915748,
+      "grad_norm": 0.824790358543396,
+      "learning_rate": 3.48170083884849e-06,
+      "loss": 0.1069,
+      "step": 7229
+    },
+    {
+      "epoch": 1.1714193130265715,
+      "grad_norm": 0.8293957114219666,
+      "learning_rate": 3.481298646172122e-06,
+      "loss": 0.1003,
+      "step": 7230
+    },
+    {
+      "epoch": 1.1715813350615685,
+      "grad_norm": 0.8428707718849182,
+      "learning_rate": 3.4808964234703903e-06,
+      "loss": 0.0993,
+      "step": 7231
+    },
+    {
+      "epoch": 1.1717433570965652,
+      "grad_norm": 0.8865581154823303,
+      "learning_rate": 3.480494170755602e-06,
+      "loss": 0.1059,
+      "step": 7232
+    },
+    {
+      "epoch": 1.171905379131562,
+      "grad_norm": 0.8242313265800476,
+      "learning_rate": 3.4800918880400635e-06,
+      "loss": 0.105,
+      "step": 7233
+    },
+    {
+      "epoch": 1.1720674011665586,
+      "grad_norm": 0.843008279800415,
+      "learning_rate": 3.479689575336086e-06,
+      "loss": 0.099,
+      "step": 7234
+    },
+    {
+      "epoch": 1.1722294232015553,
+      "grad_norm": 0.7857580780982971,
+      "learning_rate": 3.479287232655978e-06,
+      "loss": 0.0967,
+      "step": 7235
+    },
+    {
+      "epoch": 1.1723914452365523,
+      "grad_norm": 0.8480691313743591,
+      "learning_rate": 3.4788848600120507e-06,
+      "loss": 0.0949,
+      "step": 7236
+    },
+    {
+      "epoch": 1.172553467271549,
+      "grad_norm": 0.783110499382019,
+      "learning_rate": 3.4784824574166153e-06,
+      "loss": 0.0997,
+      "step": 7237
+    },
+    {
+      "epoch": 1.1727154893065457,
+      "grad_norm": 0.8678748607635498,
+      "learning_rate": 3.4780800248819847e-06,
+      "loss": 0.0947,
+      "step": 7238
+    },
+    {
+      "epoch": 1.1728775113415424,
+      "grad_norm": 0.8238358497619629,
+      "learning_rate": 3.477677562420472e-06,
+      "loss": 0.1107,
+      "step": 7239
+    },
+    {
+      "epoch": 1.173039533376539,
+      "grad_norm": 0.8292809724807739,
+      "learning_rate": 3.4772750700443923e-06,
+      "loss": 0.1047,
+      "step": 7240
+    },
+    {
+      "epoch": 1.173201555411536,
+      "grad_norm": 0.8215393424034119,
+      "learning_rate": 3.47687254776606e-06,
+      "loss": 0.0902,
+      "step": 7241
+    },
+    {
+      "epoch": 1.1733635774465327,
+      "grad_norm": 0.9231199622154236,
+      "learning_rate": 3.476469995597792e-06,
+      "loss": 0.1207,
+      "step": 7242
+    },
+    {
+      "epoch": 1.1735255994815295,
+      "grad_norm": 0.7816325426101685,
+      "learning_rate": 3.476067413551906e-06,
+      "loss": 0.0951,
+      "step": 7243
+    },
+    {
+      "epoch": 1.1736876215165262,
+      "grad_norm": 0.8484027981758118,
+      "learning_rate": 3.4756648016407175e-06,
+      "loss": 0.103,
+      "step": 7244
+    },
+    {
+      "epoch": 1.173849643551523,
+      "grad_norm": 0.7972317337989807,
+      "learning_rate": 3.475262159876548e-06,
+      "loss": 0.1049,
+      "step": 7245
+    },
+    {
+      "epoch": 1.1740116655865198,
+      "grad_norm": 0.8663620948791504,
+      "learning_rate": 3.4748594882717163e-06,
+      "loss": 0.0965,
+      "step": 7246
+    },
+    {
+      "epoch": 1.1741736876215165,
+      "grad_norm": 0.9606160521507263,
+      "learning_rate": 3.4744567868385432e-06,
+      "loss": 0.1068,
+      "step": 7247
+    },
+    {
+      "epoch": 1.1743357096565132,
+      "grad_norm": 0.9100252389907837,
+      "learning_rate": 3.474054055589351e-06,
+      "loss": 0.1114,
+      "step": 7248
+    },
+    {
+      "epoch": 1.17449773169151,
+      "grad_norm": 0.7617812156677246,
+      "learning_rate": 3.473651294536462e-06,
+      "loss": 0.1033,
+      "step": 7249
+    },
+    {
+      "epoch": 1.1746597537265069,
+      "grad_norm": 0.7390443086624146,
+      "learning_rate": 3.473248503692199e-06,
+      "loss": 0.0894,
+      "step": 7250
+    },
+    {
+      "epoch": 1.1748217757615036,
+      "grad_norm": 0.9591291546821594,
+      "learning_rate": 3.4728456830688873e-06,
+      "loss": 0.1099,
+      "step": 7251
+    },
+    {
+      "epoch": 1.1749837977965003,
+      "grad_norm": 0.8032423853874207,
+      "learning_rate": 3.472442832678852e-06,
+      "loss": 0.1055,
+      "step": 7252
+    },
+    {
+      "epoch": 1.175145819831497,
+      "grad_norm": 0.8758203387260437,
+      "learning_rate": 3.472039952534419e-06,
+      "loss": 0.107,
+      "step": 7253
+    },
+    {
+      "epoch": 1.175307841866494,
+      "grad_norm": 0.8825302124023438,
+      "learning_rate": 3.471637042647916e-06,
+      "loss": 0.1115,
+      "step": 7254
+    },
+    {
+      "epoch": 1.1754698639014907,
+      "grad_norm": 0.9467746019363403,
+      "learning_rate": 3.471234103031671e-06,
+      "loss": 0.122,
+      "step": 7255
+    },
+    {
+      "epoch": 1.1756318859364874,
+      "grad_norm": 0.8400065302848816,
+      "learning_rate": 3.470831133698013e-06,
+      "loss": 0.1034,
+      "step": 7256
+    },
+    {
+      "epoch": 1.175793907971484,
+      "grad_norm": 0.8600103259086609,
+      "learning_rate": 3.4704281346592703e-06,
+      "loss": 0.1021,
+      "step": 7257
+    },
+    {
+      "epoch": 1.1759559300064808,
+      "grad_norm": 1.0073620080947876,
+      "learning_rate": 3.470025105927777e-06,
+      "loss": 0.1043,
+      "step": 7258
+    },
+    {
+      "epoch": 1.1761179520414777,
+      "grad_norm": 0.8043482303619385,
+      "learning_rate": 3.4696220475158615e-06,
+      "loss": 0.0953,
+      "step": 7259
+    },
+    {
+      "epoch": 1.1762799740764744,
+      "grad_norm": 0.9232493042945862,
+      "learning_rate": 3.4692189594358578e-06,
+      "loss": 0.1104,
+      "step": 7260
+    },
+    {
+      "epoch": 1.1764419961114712,
+      "grad_norm": 0.8835765719413757,
+      "learning_rate": 3.4688158417000993e-06,
+      "loss": 0.1025,
+      "step": 7261
+    },
+    {
+      "epoch": 1.1766040181464679,
+      "grad_norm": 0.9452476501464844,
+      "learning_rate": 3.468412694320921e-06,
+      "loss": 0.1119,
+      "step": 7262
+    },
+    {
+      "epoch": 1.1767660401814646,
+      "grad_norm": 0.7869393825531006,
+      "learning_rate": 3.468009517310659e-06,
+      "loss": 0.0965,
+      "step": 7263
+    },
+    {
+      "epoch": 1.1769280622164615,
+      "grad_norm": 1.0424165725708008,
+      "learning_rate": 3.467606310681646e-06,
+      "loss": 0.1077,
+      "step": 7264
+    },
+    {
+      "epoch": 1.1770900842514582,
+      "grad_norm": 0.8021036982536316,
+      "learning_rate": 3.4672030744462224e-06,
+      "loss": 0.1049,
+      "step": 7265
+    },
+    {
+      "epoch": 1.177252106286455,
+      "grad_norm": 0.830315113067627,
+      "learning_rate": 3.4667998086167253e-06,
+      "loss": 0.099,
+      "step": 7266
+    },
+    {
+      "epoch": 1.1774141283214516,
+      "grad_norm": 0.93221116065979,
+      "learning_rate": 3.4663965132054943e-06,
+      "loss": 0.1029,
+      "step": 7267
+    },
+    {
+      "epoch": 1.1775761503564484,
+      "grad_norm": 0.9032275080680847,
+      "learning_rate": 3.465993188224868e-06,
+      "loss": 0.1074,
+      "step": 7268
+    },
+    {
+      "epoch": 1.1777381723914453,
+      "grad_norm": 0.8030032515525818,
+      "learning_rate": 3.465589833687188e-06,
+      "loss": 0.0987,
+      "step": 7269
+    },
+    {
+      "epoch": 1.177900194426442,
+      "grad_norm": 0.7802344560623169,
+      "learning_rate": 3.4651864496047952e-06,
+      "loss": 0.0968,
+      "step": 7270
+    },
+    {
+      "epoch": 1.1780622164614387,
+      "grad_norm": 0.816124439239502,
+      "learning_rate": 3.4647830359900335e-06,
+      "loss": 0.1089,
+      "step": 7271
+    },
+    {
+      "epoch": 1.1782242384964354,
+      "grad_norm": 0.6987655758857727,
+      "learning_rate": 3.464379592855246e-06,
+      "loss": 0.0939,
+      "step": 7272
+    },
+    {
+      "epoch": 1.1783862605314324,
+      "grad_norm": 0.9841788411140442,
+      "learning_rate": 3.463976120212776e-06,
+      "loss": 0.1049,
+      "step": 7273
+    },
+    {
+      "epoch": 1.178548282566429,
+      "grad_norm": 0.9158380627632141,
+      "learning_rate": 3.4635726180749698e-06,
+      "loss": 0.1152,
+      "step": 7274
+    },
+    {
+      "epoch": 1.1787103046014258,
+      "grad_norm": 0.7727981805801392,
+      "learning_rate": 3.4631690864541723e-06,
+      "loss": 0.1027,
+      "step": 7275
+    },
+    {
+      "epoch": 1.1788723266364225,
+      "grad_norm": 0.7879157662391663,
+      "learning_rate": 3.4627655253627324e-06,
+      "loss": 0.1041,
+      "step": 7276
+    },
+    {
+      "epoch": 1.1790343486714194,
+      "grad_norm": 0.8352596759796143,
+      "learning_rate": 3.4623619348129973e-06,
+      "loss": 0.1104,
+      "step": 7277
+    },
+    {
+      "epoch": 1.1791963707064161,
+      "grad_norm": 0.8539366126060486,
+      "learning_rate": 3.461958314817316e-06,
+      "loss": 0.1097,
+      "step": 7278
+    },
+    {
+      "epoch": 1.1793583927414129,
+      "grad_norm": 0.8420810699462891,
+      "learning_rate": 3.461554665388038e-06,
+      "loss": 0.111,
+      "step": 7279
+    },
+    {
+      "epoch": 1.1795204147764096,
+      "grad_norm": 0.864422082901001,
+      "learning_rate": 3.4611509865375143e-06,
+      "loss": 0.0987,
+      "step": 7280
+    },
+    {
+      "epoch": 1.1796824368114063,
+      "grad_norm": 0.9048283100128174,
+      "learning_rate": 3.460747278278096e-06,
+      "loss": 0.1139,
+      "step": 7281
+    },
+    {
+      "epoch": 1.1798444588464032,
+      "grad_norm": 0.9229883551597595,
+      "learning_rate": 3.4603435406221356e-06,
+      "loss": 0.1075,
+      "step": 7282
+    },
+    {
+      "epoch": 1.1800064808814,
+      "grad_norm": 0.8494004011154175,
+      "learning_rate": 3.4599397735819877e-06,
+      "loss": 0.1002,
+      "step": 7283
+    },
+    {
+      "epoch": 1.1801685029163966,
+      "grad_norm": 0.926969051361084,
+      "learning_rate": 3.4595359771700055e-06,
+      "loss": 0.1131,
+      "step": 7284
+    },
+    {
+      "epoch": 1.1803305249513933,
+      "grad_norm": 0.9146028757095337,
+      "learning_rate": 3.459132151398544e-06,
+      "loss": 0.1114,
+      "step": 7285
+    },
+    {
+      "epoch": 1.18049254698639,
+      "grad_norm": 0.937514066696167,
+      "learning_rate": 3.4587282962799602e-06,
+      "loss": 0.1124,
+      "step": 7286
+    },
+    {
+      "epoch": 1.180654569021387,
+      "grad_norm": 0.7594295144081116,
+      "learning_rate": 3.4583244118266107e-06,
+      "loss": 0.0934,
+      "step": 7287
+    },
+    {
+      "epoch": 1.1808165910563837,
+      "grad_norm": 0.8070241212844849,
+      "learning_rate": 3.4579204980508525e-06,
+      "loss": 0.0971,
+      "step": 7288
+    },
+    {
+      "epoch": 1.1809786130913804,
+      "grad_norm": 0.8179810047149658,
+      "learning_rate": 3.4575165549650463e-06,
+      "loss": 0.0992,
+      "step": 7289
+    },
+    {
+      "epoch": 1.1811406351263771,
+      "grad_norm": 0.8048136830329895,
+      "learning_rate": 3.45711258258155e-06,
+      "loss": 0.1008,
+      "step": 7290
+    },
+    {
+      "epoch": 1.1813026571613738,
+      "grad_norm": 0.755553662776947,
+      "learning_rate": 3.4567085809127247e-06,
+      "loss": 0.0869,
+      "step": 7291
+    },
+    {
+      "epoch": 1.1814646791963708,
+      "grad_norm": 0.9157342314720154,
+      "learning_rate": 3.4563045499709324e-06,
+      "loss": 0.118,
+      "step": 7292
+    },
+    {
+      "epoch": 1.1816267012313675,
+      "grad_norm": 1.028659462928772,
+      "learning_rate": 3.455900489768535e-06,
+      "loss": 0.1275,
+      "step": 7293
+    },
+    {
+      "epoch": 1.1817887232663642,
+      "grad_norm": 0.9139890074729919,
+      "learning_rate": 3.455496400317896e-06,
+      "loss": 0.1034,
+      "step": 7294
+    },
+    {
+      "epoch": 1.181950745301361,
+      "grad_norm": 0.8444374799728394,
+      "learning_rate": 3.455092281631379e-06,
+      "loss": 0.1173,
+      "step": 7295
+    },
+    {
+      "epoch": 1.1821127673363578,
+      "grad_norm": 0.8052892088890076,
+      "learning_rate": 3.45468813372135e-06,
+      "loss": 0.097,
+      "step": 7296
+    },
+    {
+      "epoch": 1.1822747893713546,
+      "grad_norm": 0.7657241225242615,
+      "learning_rate": 3.4542839566001728e-06,
+      "loss": 0.0919,
+      "step": 7297
+    },
+    {
+      "epoch": 1.1824368114063513,
+      "grad_norm": 0.8102367520332336,
+      "learning_rate": 3.453879750280218e-06,
+      "loss": 0.1024,
+      "step": 7298
+    },
+    {
+      "epoch": 1.182598833441348,
+      "grad_norm": 0.7548478245735168,
+      "learning_rate": 3.45347551477385e-06,
+      "loss": 0.0999,
+      "step": 7299
+    },
+    {
+      "epoch": 1.1827608554763447,
+      "grad_norm": 0.8339917659759521,
+      "learning_rate": 3.4530712500934393e-06,
+      "loss": 0.0953,
+      "step": 7300
+    },
+    {
+      "epoch": 1.1829228775113416,
+      "grad_norm": 0.7205444574356079,
+      "learning_rate": 3.4526669562513536e-06,
+      "loss": 0.0898,
+      "step": 7301
+    },
+    {
+      "epoch": 1.1830848995463383,
+      "grad_norm": 0.8617227077484131,
+      "learning_rate": 3.4522626332599657e-06,
+      "loss": 0.1154,
+      "step": 7302
+    },
+    {
+      "epoch": 1.183246921581335,
+      "grad_norm": 0.7841586470603943,
+      "learning_rate": 3.4518582811316455e-06,
+      "loss": 0.1032,
+      "step": 7303
+    },
+    {
+      "epoch": 1.1834089436163318,
+      "grad_norm": 0.9288532733917236,
+      "learning_rate": 3.451453899878765e-06,
+      "loss": 0.1166,
+      "step": 7304
+    },
+    {
+      "epoch": 1.1835709656513287,
+      "grad_norm": 0.7976135611534119,
+      "learning_rate": 3.4510494895136975e-06,
+      "loss": 0.1023,
+      "step": 7305
+    },
+    {
+      "epoch": 1.1837329876863254,
+      "grad_norm": 0.8904582858085632,
+      "learning_rate": 3.450645050048817e-06,
+      "loss": 0.1189,
+      "step": 7306
+    },
+    {
+      "epoch": 1.1838950097213221,
+      "grad_norm": 0.8351088166236877,
+      "learning_rate": 3.450240581496499e-06,
+      "loss": 0.0989,
+      "step": 7307
+    },
+    {
+      "epoch": 1.1840570317563188,
+      "grad_norm": 0.8244988322257996,
+      "learning_rate": 3.449836083869118e-06,
+      "loss": 0.1055,
+      "step": 7308
+    },
+    {
+      "epoch": 1.1842190537913155,
+      "grad_norm": 0.7345836758613586,
+      "learning_rate": 3.449431557179052e-06,
+      "loss": 0.0927,
+      "step": 7309
+    },
+    {
+      "epoch": 1.1843810758263125,
+      "grad_norm": 0.8357600569725037,
+      "learning_rate": 3.449027001438678e-06,
+      "loss": 0.1007,
+      "step": 7310
+    },
+    {
+      "epoch": 1.1845430978613092,
+      "grad_norm": 0.7918616533279419,
+      "learning_rate": 3.4486224166603743e-06,
+      "loss": 0.0971,
+      "step": 7311
+    },
+    {
+      "epoch": 1.184705119896306,
+      "grad_norm": 0.8694443702697754,
+      "learning_rate": 3.44821780285652e-06,
+      "loss": 0.109,
+      "step": 7312
+    },
+    {
+      "epoch": 1.1848671419313026,
+      "grad_norm": 0.8621920943260193,
+      "learning_rate": 3.447813160039496e-06,
+      "loss": 0.106,
+      "step": 7313
+    },
+    {
+      "epoch": 1.1850291639662993,
+      "grad_norm": 0.8037292957305908,
+      "learning_rate": 3.4474084882216826e-06,
+      "loss": 0.1025,
+      "step": 7314
+    },
+    {
+      "epoch": 1.1851911860012962,
+      "grad_norm": 0.8269169330596924,
+      "learning_rate": 3.447003787415462e-06,
+      "loss": 0.114,
+      "step": 7315
+    },
+    {
+      "epoch": 1.185353208036293,
+      "grad_norm": 0.9500218629837036,
+      "learning_rate": 3.4465990576332177e-06,
+      "loss": 0.114,
+      "step": 7316
+    },
+    {
+      "epoch": 1.1855152300712897,
+      "grad_norm": 0.9417467713356018,
+      "learning_rate": 3.4461942988873322e-06,
+      "loss": 0.1184,
+      "step": 7317
+    },
+    {
+      "epoch": 1.1856772521062864,
+      "grad_norm": 0.8411504030227661,
+      "learning_rate": 3.445789511190192e-06,
+      "loss": 0.1066,
+      "step": 7318
+    },
+    {
+      "epoch": 1.185839274141283,
+      "grad_norm": 0.8924190402030945,
+      "learning_rate": 3.44538469455418e-06,
+      "loss": 0.102,
+      "step": 7319
+    },
+    {
+      "epoch": 1.18600129617628,
+      "grad_norm": 0.977077305316925,
+      "learning_rate": 3.4449798489916856e-06,
+      "loss": 0.1224,
+      "step": 7320
+    },
+    {
+      "epoch": 1.1861633182112767,
+      "grad_norm": 0.949083149433136,
+      "learning_rate": 3.444574974515094e-06,
+      "loss": 0.1256,
+      "step": 7321
+    },
+    {
+      "epoch": 1.1863253402462735,
+      "grad_norm": 0.8046411871910095,
+      "learning_rate": 3.444170071136794e-06,
+      "loss": 0.098,
+      "step": 7322
+    },
+    {
+      "epoch": 1.1864873622812702,
+      "grad_norm": 0.918962299823761,
+      "learning_rate": 3.443765138869175e-06,
+      "loss": 0.1103,
+      "step": 7323
+    },
+    {
+      "epoch": 1.186649384316267,
+      "grad_norm": 0.9088669419288635,
+      "learning_rate": 3.4433601777246263e-06,
+      "loss": 0.1128,
+      "step": 7324
+    },
+    {
+      "epoch": 1.1868114063512638,
+      "grad_norm": 0.8256091475486755,
+      "learning_rate": 3.4429551877155396e-06,
+      "loss": 0.1039,
+      "step": 7325
+    },
+    {
+      "epoch": 1.1869734283862605,
+      "grad_norm": 0.8927512168884277,
+      "learning_rate": 3.442550168854305e-06,
+      "loss": 0.1106,
+      "step": 7326
+    },
+    {
+      "epoch": 1.1871354504212572,
+      "grad_norm": 0.7968646883964539,
+      "learning_rate": 3.442145121153317e-06,
+      "loss": 0.102,
+      "step": 7327
+    },
+    {
+      "epoch": 1.1872974724562542,
+      "grad_norm": 0.8282347321510315,
+      "learning_rate": 3.4417400446249684e-06,
+      "loss": 0.1074,
+      "step": 7328
+    },
+    {
+      "epoch": 1.1874594944912509,
+      "grad_norm": 0.7620131969451904,
+      "learning_rate": 3.4413349392816526e-06,
+      "loss": 0.0978,
+      "step": 7329
+    },
+    {
+      "epoch": 1.1876215165262476,
+      "grad_norm": 0.7769449353218079,
+      "learning_rate": 3.440929805135766e-06,
+      "loss": 0.0909,
+      "step": 7330
+    },
+    {
+      "epoch": 1.1877835385612443,
+      "grad_norm": 0.8219037055969238,
+      "learning_rate": 3.440524642199704e-06,
+      "loss": 0.1035,
+      "step": 7331
+    },
+    {
+      "epoch": 1.187945560596241,
+      "grad_norm": 0.9052240252494812,
+      "learning_rate": 3.440119450485865e-06,
+      "loss": 0.1266,
+      "step": 7332
+    },
+    {
+      "epoch": 1.188107582631238,
+      "grad_norm": 0.8714154958724976,
+      "learning_rate": 3.439714230006645e-06,
+      "loss": 0.1143,
+      "step": 7333
+    },
+    {
+      "epoch": 1.1882696046662347,
+      "grad_norm": 0.7879825234413147,
+      "learning_rate": 3.439308980774444e-06,
+      "loss": 0.0973,
+      "step": 7334
+    },
+    {
+      "epoch": 1.1884316267012314,
+      "grad_norm": 0.8592246770858765,
+      "learning_rate": 3.4389037028016615e-06,
+      "loss": 0.1139,
+      "step": 7335
+    },
+    {
+      "epoch": 1.188593648736228,
+      "grad_norm": 0.7850585579872131,
+      "learning_rate": 3.438498396100697e-06,
+      "loss": 0.1065,
+      "step": 7336
+    },
+    {
+      "epoch": 1.1887556707712248,
+      "grad_norm": 0.8654761910438538,
+      "learning_rate": 3.438093060683953e-06,
+      "loss": 0.1017,
+      "step": 7337
+    },
+    {
+      "epoch": 1.1889176928062217,
+      "grad_norm": 0.6961141228675842,
+      "learning_rate": 3.4376876965638317e-06,
+      "loss": 0.0819,
+      "step": 7338
+    },
+    {
+      "epoch": 1.1890797148412184,
+      "grad_norm": 0.9045152068138123,
+      "learning_rate": 3.4372823037527357e-06,
+      "loss": 0.1117,
+      "step": 7339
+    },
+    {
+      "epoch": 1.1892417368762151,
+      "grad_norm": 0.9146777391433716,
+      "learning_rate": 3.4368768822630705e-06,
+      "loss": 0.1137,
+      "step": 7340
+    },
+    {
+      "epoch": 1.1894037589112119,
+      "grad_norm": 0.7968778014183044,
+      "learning_rate": 3.4364714321072384e-06,
+      "loss": 0.0937,
+      "step": 7341
+    },
+    {
+      "epoch": 1.1895657809462086,
+      "grad_norm": 0.8366844654083252,
+      "learning_rate": 3.4360659532976475e-06,
+      "loss": 0.1011,
+      "step": 7342
+    },
+    {
+      "epoch": 1.1897278029812055,
+      "grad_norm": 0.9418715238571167,
+      "learning_rate": 3.4356604458467034e-06,
+      "loss": 0.1189,
+      "step": 7343
+    },
+    {
+      "epoch": 1.1898898250162022,
+      "grad_norm": 0.8848615288734436,
+      "learning_rate": 3.435254909766814e-06,
+      "loss": 0.1043,
+      "step": 7344
+    },
+    {
+      "epoch": 1.190051847051199,
+      "grad_norm": 0.7343488931655884,
+      "learning_rate": 3.434849345070388e-06,
+      "loss": 0.1001,
+      "step": 7345
+    },
+    {
+      "epoch": 1.1902138690861956,
+      "grad_norm": 0.9168867468833923,
+      "learning_rate": 3.4344437517698336e-06,
+      "loss": 0.1032,
+      "step": 7346
+    },
+    {
+      "epoch": 1.1903758911211926,
+      "grad_norm": 0.7533358931541443,
+      "learning_rate": 3.4340381298775628e-06,
+      "loss": 0.0936,
+      "step": 7347
+    },
+    {
+      "epoch": 1.1905379131561893,
+      "grad_norm": 0.8484786152839661,
+      "learning_rate": 3.433632479405984e-06,
+      "loss": 0.1138,
+      "step": 7348
+    },
+    {
+      "epoch": 1.190699935191186,
+      "grad_norm": 0.7753875255584717,
+      "learning_rate": 3.4332268003675117e-06,
+      "loss": 0.0983,
+      "step": 7349
+    },
+    {
+      "epoch": 1.1908619572261827,
+      "grad_norm": 0.7216082215309143,
+      "learning_rate": 3.4328210927745577e-06,
+      "loss": 0.0931,
+      "step": 7350
+    },
+    {
+      "epoch": 1.1910239792611796,
+      "grad_norm": 0.8276437520980835,
+      "learning_rate": 3.432415356639536e-06,
+      "loss": 0.1084,
+      "step": 7351
+    },
+    {
+      "epoch": 1.1911860012961764,
+      "grad_norm": 0.9757489562034607,
+      "learning_rate": 3.4320095919748596e-06,
+      "loss": 0.1231,
+      "step": 7352
+    },
+    {
+      "epoch": 1.191348023331173,
+      "grad_norm": 0.760451078414917,
+      "learning_rate": 3.4316037987929456e-06,
+      "loss": 0.0991,
+      "step": 7353
+    },
+    {
+      "epoch": 1.1915100453661698,
+      "grad_norm": 0.7524585127830505,
+      "learning_rate": 3.43119797710621e-06,
+      "loss": 0.0946,
+      "step": 7354
+    },
+    {
+      "epoch": 1.1916720674011665,
+      "grad_norm": 0.9145233631134033,
+      "learning_rate": 3.4307921269270694e-06,
+      "loss": 0.1226,
+      "step": 7355
+    },
+    {
+      "epoch": 1.1918340894361634,
+      "grad_norm": 0.8342117667198181,
+      "learning_rate": 3.4303862482679435e-06,
+      "loss": 0.1062,
+      "step": 7356
+    },
+    {
+      "epoch": 1.1919961114711601,
+      "grad_norm": 0.9443584680557251,
+      "learning_rate": 3.429980341141248e-06,
+      "loss": 0.1181,
+      "step": 7357
+    },
+    {
+      "epoch": 1.1921581335061568,
+      "grad_norm": 0.7842267155647278,
+      "learning_rate": 3.429574405559406e-06,
+      "loss": 0.1026,
+      "step": 7358
+    },
+    {
+      "epoch": 1.1923201555411536,
+      "grad_norm": 0.8545981645584106,
+      "learning_rate": 3.429168441534836e-06,
+      "loss": 0.1103,
+      "step": 7359
+    },
+    {
+      "epoch": 1.1924821775761503,
+      "grad_norm": 0.8200758695602417,
+      "learning_rate": 3.4287624490799605e-06,
+      "loss": 0.1054,
+      "step": 7360
+    },
+    {
+      "epoch": 1.1926441996111472,
+      "grad_norm": 0.8782671093940735,
+      "learning_rate": 3.428356428207201e-06,
+      "loss": 0.1025,
+      "step": 7361
+    },
+    {
+      "epoch": 1.192806221646144,
+      "grad_norm": 0.841168999671936,
+      "learning_rate": 3.4279503789289824e-06,
+      "loss": 0.1002,
+      "step": 7362
+    },
+    {
+      "epoch": 1.1929682436811406,
+      "grad_norm": 0.7707106471061707,
+      "learning_rate": 3.427544301257727e-06,
+      "loss": 0.0937,
+      "step": 7363
+    },
+    {
+      "epoch": 1.1931302657161373,
+      "grad_norm": 1.0922257900238037,
+      "learning_rate": 3.4271381952058607e-06,
+      "loss": 0.1023,
+      "step": 7364
+    },
+    {
+      "epoch": 1.193292287751134,
+      "grad_norm": 0.8869131803512573,
+      "learning_rate": 3.4267320607858094e-06,
+      "loss": 0.1113,
+      "step": 7365
+    },
+    {
+      "epoch": 1.193454309786131,
+      "grad_norm": 0.8469724655151367,
+      "learning_rate": 3.42632589801e-06,
+      "loss": 0.1005,
+      "step": 7366
+    },
+    {
+      "epoch": 1.1936163318211277,
+      "grad_norm": 0.9287112951278687,
+      "learning_rate": 3.42591970689086e-06,
+      "loss": 0.1235,
+      "step": 7367
+    },
+    {
+      "epoch": 1.1937783538561244,
+      "grad_norm": 0.7303425669670105,
+      "learning_rate": 3.425513487440817e-06,
+      "loss": 0.0855,
+      "step": 7368
+    },
+    {
+      "epoch": 1.1939403758911211,
+      "grad_norm": 0.7840588688850403,
+      "learning_rate": 3.425107239672301e-06,
+      "loss": 0.1109,
+      "step": 7369
+    },
+    {
+      "epoch": 1.1941023979261178,
+      "grad_norm": 0.9029366970062256,
+      "learning_rate": 3.4247009635977425e-06,
+      "loss": 0.1064,
+      "step": 7370
+    },
+    {
+      "epoch": 1.1942644199611148,
+      "grad_norm": 0.9053102731704712,
+      "learning_rate": 3.4242946592295724e-06,
+      "loss": 0.1151,
+      "step": 7371
+    },
+    {
+      "epoch": 1.1944264419961115,
+      "grad_norm": 0.7992795705795288,
+      "learning_rate": 3.4238883265802215e-06,
+      "loss": 0.1023,
+      "step": 7372
+    },
+    {
+      "epoch": 1.1945884640311082,
+      "grad_norm": 0.7910366058349609,
+      "learning_rate": 3.423481965662125e-06,
+      "loss": 0.1009,
+      "step": 7373
+    },
+    {
+      "epoch": 1.194750486066105,
+      "grad_norm": 0.756036102771759,
+      "learning_rate": 3.4230755764877133e-06,
+      "loss": 0.0944,
+      "step": 7374
+    },
+    {
+      "epoch": 1.1949125081011018,
+      "grad_norm": 0.8976595401763916,
+      "learning_rate": 3.4226691590694244e-06,
+      "loss": 0.1048,
+      "step": 7375
+    },
+    {
+      "epoch": 1.1950745301360985,
+      "grad_norm": 0.8526069521903992,
+      "learning_rate": 3.4222627134196917e-06,
+      "loss": 0.1115,
+      "step": 7376
+    },
+    {
+      "epoch": 1.1952365521710953,
+      "grad_norm": 0.8503440022468567,
+      "learning_rate": 3.4218562395509513e-06,
+      "loss": 0.1178,
+      "step": 7377
+    },
+    {
+      "epoch": 1.195398574206092,
+      "grad_norm": 0.8521484732627869,
+      "learning_rate": 3.4214497374756415e-06,
+      "loss": 0.1027,
+      "step": 7378
+    },
+    {
+      "epoch": 1.195560596241089,
+      "grad_norm": 0.8041279315948486,
+      "learning_rate": 3.421043207206199e-06,
+      "loss": 0.0938,
+      "step": 7379
+    },
+    {
+      "epoch": 1.1957226182760856,
+      "grad_norm": 0.7933258414268494,
+      "learning_rate": 3.4206366487550637e-06,
+      "loss": 0.0986,
+      "step": 7380
+    },
+    {
+      "epoch": 1.1958846403110823,
+      "grad_norm": 0.7659348845481873,
+      "learning_rate": 3.420230062134674e-06,
+      "loss": 0.0948,
+      "step": 7381
+    },
+    {
+      "epoch": 1.196046662346079,
+      "grad_norm": 0.766547679901123,
+      "learning_rate": 3.419823447357472e-06,
+      "loss": 0.0984,
+      "step": 7382
+    },
+    {
+      "epoch": 1.1962086843810757,
+      "grad_norm": 0.9550457000732422,
+      "learning_rate": 3.419416804435899e-06,
+      "loss": 0.1161,
+      "step": 7383
+    },
+    {
+      "epoch": 1.1963707064160727,
+      "grad_norm": 0.8955060243606567,
+      "learning_rate": 3.4190101333823956e-06,
+      "loss": 0.1151,
+      "step": 7384
+    },
+    {
+      "epoch": 1.1965327284510694,
+      "grad_norm": 0.8570713996887207,
+      "learning_rate": 3.4186034342094066e-06,
+      "loss": 0.1072,
+      "step": 7385
+    },
+    {
+      "epoch": 1.196694750486066,
+      "grad_norm": 0.9379814863204956,
+      "learning_rate": 3.4181967069293754e-06,
+      "loss": 0.1144,
+      "step": 7386
+    },
+    {
+      "epoch": 1.1968567725210628,
+      "grad_norm": 0.8272217512130737,
+      "learning_rate": 3.417789951554747e-06,
+      "loss": 0.1065,
+      "step": 7387
+    },
+    {
+      "epoch": 1.1970187945560595,
+      "grad_norm": 0.8084843754768372,
+      "learning_rate": 3.417383168097967e-06,
+      "loss": 0.0922,
+      "step": 7388
+    },
+    {
+      "epoch": 1.1971808165910565,
+      "grad_norm": 0.894687294960022,
+      "learning_rate": 3.4169763565714824e-06,
+      "loss": 0.1148,
+      "step": 7389
+    },
+    {
+      "epoch": 1.1973428386260532,
+      "grad_norm": 0.7344521284103394,
+      "learning_rate": 3.41656951698774e-06,
+      "loss": 0.0919,
+      "step": 7390
+    },
+    {
+      "epoch": 1.1975048606610499,
+      "grad_norm": 0.7177636623382568,
+      "learning_rate": 3.416162649359189e-06,
+      "loss": 0.0873,
+      "step": 7391
+    },
+    {
+      "epoch": 1.1976668826960466,
+      "grad_norm": 0.8915464282035828,
+      "learning_rate": 3.4157557536982773e-06,
+      "loss": 0.1083,
+      "step": 7392
+    },
+    {
+      "epoch": 1.1978289047310433,
+      "grad_norm": 0.8106177449226379,
+      "learning_rate": 3.4153488300174557e-06,
+      "loss": 0.0969,
+      "step": 7393
+    },
+    {
+      "epoch": 1.1979909267660402,
+      "grad_norm": 0.8630214929580688,
+      "learning_rate": 3.414941878329175e-06,
+      "loss": 0.1064,
+      "step": 7394
+    },
+    {
+      "epoch": 1.198152948801037,
+      "grad_norm": 0.9472280144691467,
+      "learning_rate": 3.4145348986458874e-06,
+      "loss": 0.1136,
+      "step": 7395
+    },
+    {
+      "epoch": 1.1983149708360337,
+      "grad_norm": 0.795714795589447,
+      "learning_rate": 3.4141278909800444e-06,
+      "loss": 0.0959,
+      "step": 7396
+    },
+    {
+      "epoch": 1.1984769928710304,
+      "grad_norm": 0.8350076675415039,
+      "learning_rate": 3.4137208553441008e-06,
+      "loss": 0.0958,
+      "step": 7397
+    },
+    {
+      "epoch": 1.1986390149060273,
+      "grad_norm": 0.9593220353126526,
+      "learning_rate": 3.41331379175051e-06,
+      "loss": 0.1109,
+      "step": 7398
+    },
+    {
+      "epoch": 1.198801036941024,
+      "grad_norm": 0.7994752526283264,
+      "learning_rate": 3.4129067002117266e-06,
+      "loss": 0.1018,
+      "step": 7399
+    },
+    {
+      "epoch": 1.1989630589760207,
+      "grad_norm": 0.7729980945587158,
+      "learning_rate": 3.4124995807402082e-06,
+      "loss": 0.0983,
+      "step": 7400
+    },
+    {
+      "epoch": 1.1991250810110174,
+      "grad_norm": 0.8964594602584839,
+      "learning_rate": 3.412092433348411e-06,
+      "loss": 0.1145,
+      "step": 7401
+    },
+    {
+      "epoch": 1.1992871030460144,
+      "grad_norm": 0.8727504014968872,
+      "learning_rate": 3.4116852580487925e-06,
+      "loss": 0.1036,
+      "step": 7402
+    },
+    {
+      "epoch": 1.199449125081011,
+      "grad_norm": 0.8404155373573303,
+      "learning_rate": 3.4112780548538097e-06,
+      "loss": 0.101,
+      "step": 7403
+    },
+    {
+      "epoch": 1.1996111471160078,
+      "grad_norm": 0.7900668382644653,
+      "learning_rate": 3.4108708237759258e-06,
+      "loss": 0.096,
+      "step": 7404
+    },
+    {
+      "epoch": 1.1997731691510045,
+      "grad_norm": 1.0467677116394043,
+      "learning_rate": 3.4104635648275975e-06,
+      "loss": 0.1102,
+      "step": 7405
+    },
+    {
+      "epoch": 1.1999351911860012,
+      "grad_norm": 0.7542744874954224,
+      "learning_rate": 3.4100562780212887e-06,
+      "loss": 0.0906,
+      "step": 7406
+    },
+    {
+      "epoch": 1.2000972132209982,
+      "grad_norm": 0.898759663105011,
+      "learning_rate": 3.409648963369459e-06,
+      "loss": 0.1133,
+      "step": 7407
+    },
+    {
+      "epoch": 1.2002592352559949,
+      "grad_norm": 0.758513867855072,
+      "learning_rate": 3.4092416208845723e-06,
+      "loss": 0.0935,
+      "step": 7408
+    },
+    {
+      "epoch": 1.2004212572909916,
+      "grad_norm": 0.8105527758598328,
+      "learning_rate": 3.408834250579093e-06,
+      "loss": 0.1066,
+      "step": 7409
+    },
+    {
+      "epoch": 1.2005832793259883,
+      "grad_norm": 0.8187546730041504,
+      "learning_rate": 3.4084268524654847e-06,
+      "loss": 0.1023,
+      "step": 7410
+    },
+    {
+      "epoch": 1.200745301360985,
+      "grad_norm": 0.8059993386268616,
+      "learning_rate": 3.4080194265562133e-06,
+      "loss": 0.1112,
+      "step": 7411
+    },
+    {
+      "epoch": 1.200907323395982,
+      "grad_norm": 0.8363770246505737,
+      "learning_rate": 3.407611972863744e-06,
+      "loss": 0.1069,
+      "step": 7412
+    },
+    {
+      "epoch": 1.2010693454309787,
+      "grad_norm": 0.7790720462799072,
+      "learning_rate": 3.407204491400546e-06,
+      "loss": 0.1021,
+      "step": 7413
+    },
+    {
+      "epoch": 1.2012313674659754,
+      "grad_norm": 0.8233287930488586,
+      "learning_rate": 3.406796982179085e-06,
+      "loss": 0.1054,
+      "step": 7414
+    },
+    {
+      "epoch": 1.201393389500972,
+      "grad_norm": 0.8368380665779114,
+      "learning_rate": 3.4063894452118313e-06,
+      "loss": 0.1095,
+      "step": 7415
+    },
+    {
+      "epoch": 1.2015554115359688,
+      "grad_norm": 0.8810364007949829,
+      "learning_rate": 3.4059818805112534e-06,
+      "loss": 0.1126,
+      "step": 7416
+    },
+    {
+      "epoch": 1.2017174335709657,
+      "grad_norm": 0.9108153581619263,
+      "learning_rate": 3.4055742880898223e-06,
+      "loss": 0.1095,
+      "step": 7417
+    },
+    {
+      "epoch": 1.2018794556059624,
+      "grad_norm": 0.7670288681983948,
+      "learning_rate": 3.4051666679600105e-06,
+      "loss": 0.0969,
+      "step": 7418
+    },
+    {
+      "epoch": 1.2020414776409591,
+      "grad_norm": 0.6589247584342957,
+      "learning_rate": 3.404759020134288e-06,
+      "loss": 0.0842,
+      "step": 7419
+    },
+    {
+      "epoch": 1.2022034996759559,
+      "grad_norm": 0.7312752604484558,
+      "learning_rate": 3.40435134462513e-06,
+      "loss": 0.0972,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2023655217109526,
+      "grad_norm": 0.8262923359870911,
+      "learning_rate": 3.403943641445008e-06,
+      "loss": 0.1024,
+      "step": 7421
+    },
+    {
+      "epoch": 1.2025275437459495,
+      "grad_norm": 0.8193771243095398,
+      "learning_rate": 3.403535910606399e-06,
+      "loss": 0.1016,
+      "step": 7422
+    },
+    {
+      "epoch": 1.2026895657809462,
+      "grad_norm": 0.7638645768165588,
+      "learning_rate": 3.4031281521217772e-06,
+      "loss": 0.0947,
+      "step": 7423
+    },
+    {
+      "epoch": 1.202851587815943,
+      "grad_norm": 0.9505733847618103,
+      "learning_rate": 3.4027203660036202e-06,
+      "loss": 0.1089,
+      "step": 7424
+    },
+    {
+      "epoch": 1.2030136098509396,
+      "grad_norm": 0.7956478595733643,
+      "learning_rate": 3.402312552264404e-06,
+      "loss": 0.0925,
+      "step": 7425
+    },
+    {
+      "epoch": 1.2031756318859366,
+      "grad_norm": 0.7707629203796387,
+      "learning_rate": 3.4019047109166077e-06,
+      "loss": 0.0984,
+      "step": 7426
+    },
+    {
+      "epoch": 1.2033376539209333,
+      "grad_norm": 0.8056702613830566,
+      "learning_rate": 3.401496841972709e-06,
+      "loss": 0.1076,
+      "step": 7427
+    },
+    {
+      "epoch": 1.20349967595593,
+      "grad_norm": 0.7400745749473572,
+      "learning_rate": 3.401088945445189e-06,
+      "loss": 0.0873,
+      "step": 7428
+    },
+    {
+      "epoch": 1.2036616979909267,
+      "grad_norm": 0.891022801399231,
+      "learning_rate": 3.400681021346528e-06,
+      "loss": 0.1001,
+      "step": 7429
+    },
+    {
+      "epoch": 1.2038237200259236,
+      "grad_norm": 0.8220235705375671,
+      "learning_rate": 3.4002730696892073e-06,
+      "loss": 0.1081,
+      "step": 7430
+    },
+    {
+      "epoch": 1.2039857420609203,
+      "grad_norm": 0.8286675810813904,
+      "learning_rate": 3.399865090485709e-06,
+      "loss": 0.1008,
+      "step": 7431
+    },
+    {
+      "epoch": 1.204147764095917,
+      "grad_norm": 0.8980154991149902,
+      "learning_rate": 3.3994570837485163e-06,
+      "loss": 0.1087,
+      "step": 7432
+    },
+    {
+      "epoch": 1.2043097861309138,
+      "grad_norm": 1.0518361330032349,
+      "learning_rate": 3.3990490494901148e-06,
+      "loss": 0.1222,
+      "step": 7433
+    },
+    {
+      "epoch": 1.2044718081659105,
+      "grad_norm": 0.9503573179244995,
+      "learning_rate": 3.3986409877229863e-06,
+      "loss": 0.1152,
+      "step": 7434
+    },
+    {
+      "epoch": 1.2046338302009074,
+      "grad_norm": 0.7729112505912781,
+      "learning_rate": 3.3982328984596196e-06,
+      "loss": 0.0995,
+      "step": 7435
+    },
+    {
+      "epoch": 1.2047958522359041,
+      "grad_norm": 0.8068438172340393,
+      "learning_rate": 3.3978247817124986e-06,
+      "loss": 0.0985,
+      "step": 7436
+    },
+    {
+      "epoch": 1.2049578742709008,
+      "grad_norm": 0.8518000245094299,
+      "learning_rate": 3.3974166374941137e-06,
+      "loss": 0.11,
+      "step": 7437
+    },
+    {
+      "epoch": 1.2051198963058976,
+      "grad_norm": 0.9522218704223633,
+      "learning_rate": 3.39700846581695e-06,
+      "loss": 0.1175,
+      "step": 7438
+    },
+    {
+      "epoch": 1.2052819183408943,
+      "grad_norm": 0.724781334400177,
+      "learning_rate": 3.396600266693498e-06,
+      "loss": 0.0842,
+      "step": 7439
+    },
+    {
+      "epoch": 1.2054439403758912,
+      "grad_norm": 0.7208175659179688,
+      "learning_rate": 3.3961920401362488e-06,
+      "loss": 0.0929,
+      "step": 7440
+    },
+    {
+      "epoch": 1.205605962410888,
+      "grad_norm": 0.8319954872131348,
+      "learning_rate": 3.395783786157691e-06,
+      "loss": 0.1001,
+      "step": 7441
+    },
+    {
+      "epoch": 1.2057679844458846,
+      "grad_norm": 0.8171935081481934,
+      "learning_rate": 3.3953755047703174e-06,
+      "loss": 0.104,
+      "step": 7442
+    },
+    {
+      "epoch": 1.2059300064808813,
+      "grad_norm": 0.7243193984031677,
+      "learning_rate": 3.394967195986619e-06,
+      "loss": 0.0957,
+      "step": 7443
+    },
+    {
+      "epoch": 1.206092028515878,
+      "grad_norm": 0.8252148628234863,
+      "learning_rate": 3.394558859819092e-06,
+      "loss": 0.1097,
+      "step": 7444
+    },
+    {
+      "epoch": 1.206254050550875,
+      "grad_norm": 0.8779266476631165,
+      "learning_rate": 3.3941504962802273e-06,
+      "loss": 0.1172,
+      "step": 7445
+    },
+    {
+      "epoch": 1.2064160725858717,
+      "grad_norm": 0.8700554370880127,
+      "learning_rate": 3.393742105382522e-06,
+      "loss": 0.1084,
+      "step": 7446
+    },
+    {
+      "epoch": 1.2065780946208684,
+      "grad_norm": 0.9085803627967834,
+      "learning_rate": 3.393333687138471e-06,
+      "loss": 0.1156,
+      "step": 7447
+    },
+    {
+      "epoch": 1.2067401166558651,
+      "grad_norm": 0.938339352607727,
+      "learning_rate": 3.3929252415605708e-06,
+      "loss": 0.1032,
+      "step": 7448
+    },
+    {
+      "epoch": 1.206902138690862,
+      "grad_norm": 0.8974847197532654,
+      "learning_rate": 3.392516768661319e-06,
+      "loss": 0.11,
+      "step": 7449
+    },
+    {
+      "epoch": 1.2070641607258588,
+      "grad_norm": 0.9363797903060913,
+      "learning_rate": 3.3921082684532143e-06,
+      "loss": 0.112,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2072261827608555,
+      "grad_norm": 0.7975560426712036,
+      "learning_rate": 3.3916997409487552e-06,
+      "loss": 0.1048,
+      "step": 7451
+    },
+    {
+      "epoch": 1.2073882047958522,
+      "grad_norm": 0.7326340675354004,
+      "learning_rate": 3.391291186160441e-06,
+      "loss": 0.0983,
+      "step": 7452
+    },
+    {
+      "epoch": 1.2075502268308491,
+      "grad_norm": 0.8057294487953186,
+      "learning_rate": 3.390882604100775e-06,
+      "loss": 0.0962,
+      "step": 7453
+    },
+    {
+      "epoch": 1.2077122488658458,
+      "grad_norm": 0.7411153316497803,
+      "learning_rate": 3.3904739947822556e-06,
+      "loss": 0.0974,
+      "step": 7454
+    },
+    {
+      "epoch": 1.2078742709008425,
+      "grad_norm": 0.822479248046875,
+      "learning_rate": 3.3900653582173883e-06,
+      "loss": 0.1059,
+      "step": 7455
+    },
+    {
+      "epoch": 1.2080362929358393,
+      "grad_norm": 0.9575856924057007,
+      "learning_rate": 3.3896566944186737e-06,
+      "loss": 0.1204,
+      "step": 7456
+    },
+    {
+      "epoch": 1.208198314970836,
+      "grad_norm": 0.7944145798683167,
+      "learning_rate": 3.3892480033986186e-06,
+      "loss": 0.1,
+      "step": 7457
+    },
+    {
+      "epoch": 1.208360337005833,
+      "grad_norm": 0.6548885107040405,
+      "learning_rate": 3.388839285169725e-06,
+      "loss": 0.0808,
+      "step": 7458
+    },
+    {
+      "epoch": 1.2085223590408296,
+      "grad_norm": 0.9237521886825562,
+      "learning_rate": 3.3884305397445017e-06,
+      "loss": 0.1132,
+      "step": 7459
+    },
+    {
+      "epoch": 1.2086843810758263,
+      "grad_norm": 0.8143603205680847,
+      "learning_rate": 3.3880217671354527e-06,
+      "loss": 0.0988,
+      "step": 7460
+    },
+    {
+      "epoch": 1.208846403110823,
+      "grad_norm": 0.8373053073883057,
+      "learning_rate": 3.3876129673550873e-06,
+      "loss": 0.1062,
+      "step": 7461
+    },
+    {
+      "epoch": 1.2090084251458197,
+      "grad_norm": 0.8747492432594299,
+      "learning_rate": 3.3872041404159124e-06,
+      "loss": 0.0978,
+      "step": 7462
+    },
+    {
+      "epoch": 1.2091704471808167,
+      "grad_norm": 0.9876968860626221,
+      "learning_rate": 3.386795286330438e-06,
+      "loss": 0.1106,
+      "step": 7463
+    },
+    {
+      "epoch": 1.2093324692158134,
+      "grad_norm": 0.8512849807739258,
+      "learning_rate": 3.3863864051111744e-06,
+      "loss": 0.1119,
+      "step": 7464
+    },
+    {
+      "epoch": 1.20949449125081,
+      "grad_norm": 0.816228449344635,
+      "learning_rate": 3.385977496770631e-06,
+      "loss": 0.113,
+      "step": 7465
+    },
+    {
+      "epoch": 1.2096565132858068,
+      "grad_norm": 0.8686400055885315,
+      "learning_rate": 3.385568561321321e-06,
+      "loss": 0.1065,
+      "step": 7466
+    },
+    {
+      "epoch": 1.2098185353208035,
+      "grad_norm": 0.9017913937568665,
+      "learning_rate": 3.385159598775755e-06,
+      "loss": 0.105,
+      "step": 7467
+    },
+    {
+      "epoch": 1.2099805573558005,
+      "grad_norm": 0.8263371586799622,
+      "learning_rate": 3.3847506091464487e-06,
+      "loss": 0.1017,
+      "step": 7468
+    },
+    {
+      "epoch": 1.2101425793907972,
+      "grad_norm": 0.8818573355674744,
+      "learning_rate": 3.3843415924459146e-06,
+      "loss": 0.1109,
+      "step": 7469
+    },
+    {
+      "epoch": 1.2103046014257939,
+      "grad_norm": 0.8400024771690369,
+      "learning_rate": 3.383932548686667e-06,
+      "loss": 0.1062,
+      "step": 7470
+    },
+    {
+      "epoch": 1.2104666234607906,
+      "grad_norm": 0.8324894309043884,
+      "learning_rate": 3.3835234778812232e-06,
+      "loss": 0.1058,
+      "step": 7471
+    },
+    {
+      "epoch": 1.2106286454957873,
+      "grad_norm": 0.8616311550140381,
+      "learning_rate": 3.3831143800420983e-06,
+      "loss": 0.1095,
+      "step": 7472
+    },
+    {
+      "epoch": 1.2107906675307842,
+      "grad_norm": 0.821205735206604,
+      "learning_rate": 3.3827052551818113e-06,
+      "loss": 0.1111,
+      "step": 7473
+    },
+    {
+      "epoch": 1.210952689565781,
+      "grad_norm": 0.794984757900238,
+      "learning_rate": 3.3822961033128793e-06,
+      "loss": 0.1011,
+      "step": 7474
+    },
+    {
+      "epoch": 1.2111147116007777,
+      "grad_norm": 0.7533223032951355,
+      "learning_rate": 3.3818869244478214e-06,
+      "loss": 0.0949,
+      "step": 7475
+    },
+    {
+      "epoch": 1.2112767336357744,
+      "grad_norm": 0.8010703921318054,
+      "learning_rate": 3.3814777185991577e-06,
+      "loss": 0.1074,
+      "step": 7476
+    },
+    {
+      "epoch": 1.2114387556707713,
+      "grad_norm": 0.7717397212982178,
+      "learning_rate": 3.3810684857794097e-06,
+      "loss": 0.0989,
+      "step": 7477
+    },
+    {
+      "epoch": 1.211600777705768,
+      "grad_norm": 0.8146265745162964,
+      "learning_rate": 3.380659226001097e-06,
+      "loss": 0.1,
+      "step": 7478
+    },
+    {
+      "epoch": 1.2117627997407647,
+      "grad_norm": 0.7482508420944214,
+      "learning_rate": 3.380249939276744e-06,
+      "loss": 0.0988,
+      "step": 7479
+    },
+    {
+      "epoch": 1.2119248217757614,
+      "grad_norm": 0.8667011260986328,
+      "learning_rate": 3.3798406256188725e-06,
+      "loss": 0.1127,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2120868438107584,
+      "grad_norm": 0.8829639554023743,
+      "learning_rate": 3.379431285040008e-06,
+      "loss": 0.1111,
+      "step": 7481
+    },
+    {
+      "epoch": 1.212248865845755,
+      "grad_norm": 0.7957995533943176,
+      "learning_rate": 3.3790219175526733e-06,
+      "loss": 0.1036,
+      "step": 7482
+    },
+    {
+      "epoch": 1.2124108878807518,
+      "grad_norm": 0.844648003578186,
+      "learning_rate": 3.3786125231693955e-06,
+      "loss": 0.0985,
+      "step": 7483
+    },
+    {
+      "epoch": 1.2125729099157485,
+      "grad_norm": 0.8586723804473877,
+      "learning_rate": 3.3782031019027006e-06,
+      "loss": 0.1066,
+      "step": 7484
+    },
+    {
+      "epoch": 1.2127349319507452,
+      "grad_norm": 0.8165646195411682,
+      "learning_rate": 3.3777936537651162e-06,
+      "loss": 0.1031,
+      "step": 7485
+    },
+    {
+      "epoch": 1.2128969539857422,
+      "grad_norm": 0.825261652469635,
+      "learning_rate": 3.3773841787691708e-06,
+      "loss": 0.0991,
+      "step": 7486
+    },
+    {
+      "epoch": 1.2130589760207389,
+      "grad_norm": 0.8294909000396729,
+      "learning_rate": 3.3769746769273915e-06,
+      "loss": 0.1069,
+      "step": 7487
+    },
+    {
+      "epoch": 1.2132209980557356,
+      "grad_norm": 0.8192000389099121,
+      "learning_rate": 3.3765651482523097e-06,
+      "loss": 0.0998,
+      "step": 7488
+    },
+    {
+      "epoch": 1.2133830200907323,
+      "grad_norm": 0.7805353999137878,
+      "learning_rate": 3.3761555927564553e-06,
+      "loss": 0.0919,
+      "step": 7489
+    },
+    {
+      "epoch": 1.213545042125729,
+      "grad_norm": 0.8784085512161255,
+      "learning_rate": 3.375746010452361e-06,
+      "loss": 0.117,
+      "step": 7490
+    },
+    {
+      "epoch": 1.213707064160726,
+      "grad_norm": 0.9832420945167542,
+      "learning_rate": 3.375336401352557e-06,
+      "loss": 0.115,
+      "step": 7491
+    },
+    {
+      "epoch": 1.2138690861957226,
+      "grad_norm": 0.8437249064445496,
+      "learning_rate": 3.374926765469578e-06,
+      "loss": 0.1156,
+      "step": 7492
+    },
+    {
+      "epoch": 1.2140311082307194,
+      "grad_norm": 0.8112325668334961,
+      "learning_rate": 3.374517102815958e-06,
+      "loss": 0.0974,
+      "step": 7493
+    },
+    {
+      "epoch": 1.214193130265716,
+      "grad_norm": 0.8753647208213806,
+      "learning_rate": 3.3741074134042297e-06,
+      "loss": 0.1087,
+      "step": 7494
+    },
+    {
+      "epoch": 1.2143551523007128,
+      "grad_norm": 0.9637497067451477,
+      "learning_rate": 3.373697697246931e-06,
+      "loss": 0.1211,
+      "step": 7495
+    },
+    {
+      "epoch": 1.2145171743357097,
+      "grad_norm": 0.8705383539199829,
+      "learning_rate": 3.3732879543565955e-06,
+      "loss": 0.1168,
+      "step": 7496
+    },
+    {
+      "epoch": 1.2146791963707064,
+      "grad_norm": 0.7167220115661621,
+      "learning_rate": 3.372878184745764e-06,
+      "loss": 0.0967,
+      "step": 7497
+    },
+    {
+      "epoch": 1.2148412184057031,
+      "grad_norm": 0.8306147456169128,
+      "learning_rate": 3.3724683884269702e-06,
+      "loss": 0.1044,
+      "step": 7498
+    },
+    {
+      "epoch": 1.2150032404406998,
+      "grad_norm": 0.9196276068687439,
+      "learning_rate": 3.3720585654127564e-06,
+      "loss": 0.1258,
+      "step": 7499
+    },
+    {
+      "epoch": 1.2151652624756968,
+      "grad_norm": 0.8953840732574463,
+      "learning_rate": 3.37164871571566e-06,
+      "loss": 0.1224,
+      "step": 7500
+    },
+    {
+      "epoch": 1.2153272845106935,
+      "grad_norm": 0.8981062173843384,
+      "learning_rate": 3.3712388393482224e-06,
+      "loss": 0.1102,
+      "step": 7501
+    },
+    {
+      "epoch": 1.2154893065456902,
+      "grad_norm": 0.7924917340278625,
+      "learning_rate": 3.370828936322985e-06,
+      "loss": 0.1097,
+      "step": 7502
+    },
+    {
+      "epoch": 1.215651328580687,
+      "grad_norm": 0.806941568851471,
+      "learning_rate": 3.37041900665249e-06,
+      "loss": 0.104,
+      "step": 7503
+    },
+    {
+      "epoch": 1.2158133506156839,
+      "grad_norm": 0.7318122982978821,
+      "learning_rate": 3.3700090503492795e-06,
+      "loss": 0.0876,
+      "step": 7504
+    },
+    {
+      "epoch": 1.2159753726506806,
+      "grad_norm": 0.7857891917228699,
+      "learning_rate": 3.369599067425897e-06,
+      "loss": 0.1066,
+      "step": 7505
+    },
+    {
+      "epoch": 1.2161373946856773,
+      "grad_norm": 0.7979137897491455,
+      "learning_rate": 3.3691890578948876e-06,
+      "loss": 0.1032,
+      "step": 7506
+    },
+    {
+      "epoch": 1.216299416720674,
+      "grad_norm": 0.8069559931755066,
+      "learning_rate": 3.3687790217687966e-06,
+      "loss": 0.1047,
+      "step": 7507
+    },
+    {
+      "epoch": 1.2164614387556707,
+      "grad_norm": 0.9257494211196899,
+      "learning_rate": 3.36836895906017e-06,
+      "loss": 0.1059,
+      "step": 7508
+    },
+    {
+      "epoch": 1.2166234607906676,
+      "grad_norm": 0.8408833146095276,
+      "learning_rate": 3.367958869781554e-06,
+      "loss": 0.0991,
+      "step": 7509
+    },
+    {
+      "epoch": 1.2167854828256643,
+      "grad_norm": 0.7625467777252197,
+      "learning_rate": 3.3675487539454972e-06,
+      "loss": 0.102,
+      "step": 7510
+    },
+    {
+      "epoch": 1.216947504860661,
+      "grad_norm": 0.7763059735298157,
+      "learning_rate": 3.367138611564548e-06,
+      "loss": 0.1047,
+      "step": 7511
+    },
+    {
+      "epoch": 1.2171095268956578,
+      "grad_norm": 0.872353196144104,
+      "learning_rate": 3.3667284426512565e-06,
+      "loss": 0.1122,
+      "step": 7512
+    },
+    {
+      "epoch": 1.2172715489306545,
+      "grad_norm": 0.8491617441177368,
+      "learning_rate": 3.366318247218171e-06,
+      "loss": 0.0998,
+      "step": 7513
+    },
+    {
+      "epoch": 1.2174335709656514,
+      "grad_norm": 0.8177269101142883,
+      "learning_rate": 3.3659080252778446e-06,
+      "loss": 0.1037,
+      "step": 7514
+    },
+    {
+      "epoch": 1.2175955930006481,
+      "grad_norm": 0.7532639503479004,
+      "learning_rate": 3.3654977768428276e-06,
+      "loss": 0.0947,
+      "step": 7515
+    },
+    {
+      "epoch": 1.2177576150356448,
+      "grad_norm": 0.7339494824409485,
+      "learning_rate": 3.365087501925673e-06,
+      "loss": 0.1038,
+      "step": 7516
+    },
+    {
+      "epoch": 1.2179196370706415,
+      "grad_norm": 0.9405337572097778,
+      "learning_rate": 3.364677200538935e-06,
+      "loss": 0.1128,
+      "step": 7517
+    },
+    {
+      "epoch": 1.2180816591056383,
+      "grad_norm": 0.8983637094497681,
+      "learning_rate": 3.3642668726951657e-06,
+      "loss": 0.1095,
+      "step": 7518
+    },
+    {
+      "epoch": 1.2182436811406352,
+      "grad_norm": 0.9077023863792419,
+      "learning_rate": 3.363856518406923e-06,
+      "loss": 0.1199,
+      "step": 7519
+    },
+    {
+      "epoch": 1.218405703175632,
+      "grad_norm": 0.8318829536437988,
+      "learning_rate": 3.36344613768676e-06,
+      "loss": 0.0994,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2185677252106286,
+      "grad_norm": 0.9089265465736389,
+      "learning_rate": 3.3630357305472363e-06,
+      "loss": 0.1143,
+      "step": 7521
+    },
+    {
+      "epoch": 1.2187297472456253,
+      "grad_norm": 0.8133178353309631,
+      "learning_rate": 3.362625297000906e-06,
+      "loss": 0.0978,
+      "step": 7522
+    },
+    {
+      "epoch": 1.2188917692806223,
+      "grad_norm": 0.8779920339584351,
+      "learning_rate": 3.3622148370603305e-06,
+      "loss": 0.0998,
+      "step": 7523
+    },
+    {
+      "epoch": 1.219053791315619,
+      "grad_norm": 0.8766910433769226,
+      "learning_rate": 3.3618043507380673e-06,
+      "loss": 0.1154,
+      "step": 7524
+    },
+    {
+      "epoch": 1.2192158133506157,
+      "grad_norm": 0.8510516285896301,
+      "learning_rate": 3.3613938380466758e-06,
+      "loss": 0.1042,
+      "step": 7525
+    },
+    {
+      "epoch": 1.2193778353856124,
+      "grad_norm": 0.8500891327857971,
+      "learning_rate": 3.3609832989987178e-06,
+      "loss": 0.1009,
+      "step": 7526
+    },
+    {
+      "epoch": 1.219539857420609,
+      "grad_norm": 0.8215943574905396,
+      "learning_rate": 3.360572733606754e-06,
+      "loss": 0.1003,
+      "step": 7527
+    },
+    {
+      "epoch": 1.219701879455606,
+      "grad_norm": 0.8531811833381653,
+      "learning_rate": 3.360162141883348e-06,
+      "loss": 0.1103,
+      "step": 7528
+    },
+    {
+      "epoch": 1.2198639014906028,
+      "grad_norm": 0.7764996290206909,
+      "learning_rate": 3.359751523841062e-06,
+      "loss": 0.0973,
+      "step": 7529
+    },
+    {
+      "epoch": 1.2200259235255995,
+      "grad_norm": 0.7268168926239014,
+      "learning_rate": 3.3593408794924585e-06,
+      "loss": 0.0937,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2201879455605962,
+      "grad_norm": 0.7981647253036499,
+      "learning_rate": 3.358930208850105e-06,
+      "loss": 0.1023,
+      "step": 7531
+    },
+    {
+      "epoch": 1.220349967595593,
+      "grad_norm": 0.7587500214576721,
+      "learning_rate": 3.358519511926565e-06,
+      "loss": 0.1031,
+      "step": 7532
+    },
+    {
+      "epoch": 1.2205119896305898,
+      "grad_norm": 0.8148356676101685,
+      "learning_rate": 3.358108788734406e-06,
+      "loss": 0.0965,
+      "step": 7533
+    },
+    {
+      "epoch": 1.2206740116655865,
+      "grad_norm": 0.7737621665000916,
+      "learning_rate": 3.357698039286194e-06,
+      "loss": 0.0945,
+      "step": 7534
+    },
+    {
+      "epoch": 1.2208360337005832,
+      "grad_norm": 0.8707150816917419,
+      "learning_rate": 3.3572872635944982e-06,
+      "loss": 0.1078,
+      "step": 7535
+    },
+    {
+      "epoch": 1.22099805573558,
+      "grad_norm": 0.8325858116149902,
+      "learning_rate": 3.356876461671887e-06,
+      "loss": 0.1106,
+      "step": 7536
+    },
+    {
+      "epoch": 1.221160077770577,
+      "grad_norm": 0.8393840789794922,
+      "learning_rate": 3.3564656335309293e-06,
+      "loss": 0.1064,
+      "step": 7537
+    },
+    {
+      "epoch": 1.2213220998055736,
+      "grad_norm": 0.8071523308753967,
+      "learning_rate": 3.3560547791841957e-06,
+      "loss": 0.1005,
+      "step": 7538
+    },
+    {
+      "epoch": 1.2214841218405703,
+      "grad_norm": 0.845231831073761,
+      "learning_rate": 3.3556438986442574e-06,
+      "loss": 0.1026,
+      "step": 7539
+    },
+    {
+      "epoch": 1.221646143875567,
+      "grad_norm": 0.8520015478134155,
+      "learning_rate": 3.3552329919236865e-06,
+      "loss": 0.0976,
+      "step": 7540
+    },
+    {
+      "epoch": 1.2218081659105637,
+      "grad_norm": 0.9010183215141296,
+      "learning_rate": 3.3548220590350563e-06,
+      "loss": 0.1105,
+      "step": 7541
+    },
+    {
+      "epoch": 1.2219701879455607,
+      "grad_norm": 0.7583953142166138,
+      "learning_rate": 3.3544110999909385e-06,
+      "loss": 0.0901,
+      "step": 7542
+    },
+    {
+      "epoch": 1.2221322099805574,
+      "grad_norm": 0.7752031683921814,
+      "learning_rate": 3.354000114803909e-06,
+      "loss": 0.099,
+      "step": 7543
+    },
+    {
+      "epoch": 1.222294232015554,
+      "grad_norm": 0.8464565873146057,
+      "learning_rate": 3.3535891034865433e-06,
+      "loss": 0.0952,
+      "step": 7544
+    },
+    {
+      "epoch": 1.2224562540505508,
+      "grad_norm": 0.8120436668395996,
+      "learning_rate": 3.3531780660514164e-06,
+      "loss": 0.098,
+      "step": 7545
+    },
+    {
+      "epoch": 1.2226182760855475,
+      "grad_norm": 0.7880799770355225,
+      "learning_rate": 3.3527670025111046e-06,
+      "loss": 0.1071,
+      "step": 7546
+    },
+    {
+      "epoch": 1.2227802981205445,
+      "grad_norm": 0.8038657903671265,
+      "learning_rate": 3.352355912878187e-06,
+      "loss": 0.0996,
+      "step": 7547
+    },
+    {
+      "epoch": 1.2229423201555412,
+      "grad_norm": 0.7888135313987732,
+      "learning_rate": 3.3519447971652407e-06,
+      "loss": 0.089,
+      "step": 7548
+    },
+    {
+      "epoch": 1.2231043421905379,
+      "grad_norm": 0.7552701830863953,
+      "learning_rate": 3.3515336553848454e-06,
+      "loss": 0.0972,
+      "step": 7549
+    },
+    {
+      "epoch": 1.2232663642255346,
+      "grad_norm": 0.720099151134491,
+      "learning_rate": 3.351122487549582e-06,
+      "loss": 0.0838,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2234283862605315,
+      "grad_norm": 0.8370726108551025,
+      "learning_rate": 3.350711293672029e-06,
+      "loss": 0.1043,
+      "step": 7551
+    },
+    {
+      "epoch": 1.2235904082955282,
+      "grad_norm": 0.8462739586830139,
+      "learning_rate": 3.3503000737647696e-06,
+      "loss": 0.1001,
+      "step": 7552
+    },
+    {
+      "epoch": 1.223752430330525,
+      "grad_norm": 0.8034543991088867,
+      "learning_rate": 3.349888827840385e-06,
+      "loss": 0.0989,
+      "step": 7553
+    },
+    {
+      "epoch": 1.2239144523655217,
+      "grad_norm": 0.9257510304450989,
+      "learning_rate": 3.349477555911459e-06,
+      "loss": 0.1208,
+      "step": 7554
+    },
+    {
+      "epoch": 1.2240764744005186,
+      "grad_norm": 0.79314786195755,
+      "learning_rate": 3.349066257990576e-06,
+      "loss": 0.0993,
+      "step": 7555
+    },
+    {
+      "epoch": 1.2242384964355153,
+      "grad_norm": 0.8348277807235718,
+      "learning_rate": 3.3486549340903196e-06,
+      "loss": 0.096,
+      "step": 7556
+    },
+    {
+      "epoch": 1.224400518470512,
+      "grad_norm": 0.7744476199150085,
+      "learning_rate": 3.3482435842232763e-06,
+      "loss": 0.1027,
+      "step": 7557
+    },
+    {
+      "epoch": 1.2245625405055087,
+      "grad_norm": 0.9177954792976379,
+      "learning_rate": 3.3478322084020322e-06,
+      "loss": 0.1159,
+      "step": 7558
+    },
+    {
+      "epoch": 1.2247245625405054,
+      "grad_norm": 0.8459001779556274,
+      "learning_rate": 3.3474208066391747e-06,
+      "loss": 0.0988,
+      "step": 7559
+    },
+    {
+      "epoch": 1.2248865845755024,
+      "grad_norm": 0.8714538812637329,
+      "learning_rate": 3.34700937894729e-06,
+      "loss": 0.1032,
+      "step": 7560
+    },
+    {
+      "epoch": 1.225048606610499,
+      "grad_norm": 0.887595534324646,
+      "learning_rate": 3.3465979253389685e-06,
+      "loss": 0.1079,
+      "step": 7561
+    },
+    {
+      "epoch": 1.2252106286454958,
+      "grad_norm": 0.8663240075111389,
+      "learning_rate": 3.3461864458267996e-06,
+      "loss": 0.1049,
+      "step": 7562
+    },
+    {
+      "epoch": 1.2253726506804925,
+      "grad_norm": 0.7935128211975098,
+      "learning_rate": 3.3457749404233724e-06,
+      "loss": 0.1011,
+      "step": 7563
+    },
+    {
+      "epoch": 1.2255346727154892,
+      "grad_norm": 0.8313045501708984,
+      "learning_rate": 3.3453634091412795e-06,
+      "loss": 0.093,
+      "step": 7564
+    },
+    {
+      "epoch": 1.2256966947504861,
+      "grad_norm": 0.8548048138618469,
+      "learning_rate": 3.3449518519931123e-06,
+      "loss": 0.1095,
+      "step": 7565
+    },
+    {
+      "epoch": 1.2258587167854829,
+      "grad_norm": 0.8249443173408508,
+      "learning_rate": 3.344540268991462e-06,
+      "loss": 0.1114,
+      "step": 7566
+    },
+    {
+      "epoch": 1.2260207388204796,
+      "grad_norm": 0.850307285785675,
+      "learning_rate": 3.344128660148924e-06,
+      "loss": 0.1191,
+      "step": 7567
+    },
+    {
+      "epoch": 1.2261827608554763,
+      "grad_norm": 0.9916208386421204,
+      "learning_rate": 3.343717025478092e-06,
+      "loss": 0.1222,
+      "step": 7568
+    },
+    {
+      "epoch": 1.226344782890473,
+      "grad_norm": 0.825520932674408,
+      "learning_rate": 3.3433053649915603e-06,
+      "loss": 0.1055,
+      "step": 7569
+    },
+    {
+      "epoch": 1.22650680492547,
+      "grad_norm": 0.7133052349090576,
+      "learning_rate": 3.342893678701925e-06,
+      "loss": 0.0932,
+      "step": 7570
+    },
+    {
+      "epoch": 1.2266688269604666,
+      "grad_norm": 0.7931620478630066,
+      "learning_rate": 3.3424819666217834e-06,
+      "loss": 0.0977,
+      "step": 7571
+    },
+    {
+      "epoch": 1.2268308489954634,
+      "grad_norm": 0.80473393201828,
+      "learning_rate": 3.3420702287637325e-06,
+      "loss": 0.106,
+      "step": 7572
+    },
+    {
+      "epoch": 1.22699287103046,
+      "grad_norm": 1.0347249507904053,
+      "learning_rate": 3.3416584651403696e-06,
+      "loss": 0.132,
+      "step": 7573
+    },
+    {
+      "epoch": 1.227154893065457,
+      "grad_norm": 0.8027249574661255,
+      "learning_rate": 3.341246675764295e-06,
+      "loss": 0.1066,
+      "step": 7574
+    },
+    {
+      "epoch": 1.2273169151004537,
+      "grad_norm": 0.7861801385879517,
+      "learning_rate": 3.3408348606481078e-06,
+      "loss": 0.0994,
+      "step": 7575
+    },
+    {
+      "epoch": 1.2274789371354504,
+      "grad_norm": 0.943520188331604,
+      "learning_rate": 3.3404230198044085e-06,
+      "loss": 0.1147,
+      "step": 7576
+    },
+    {
+      "epoch": 1.2276409591704471,
+      "grad_norm": 0.9658901691436768,
+      "learning_rate": 3.340011153245799e-06,
+      "loss": 0.1152,
+      "step": 7577
+    },
+    {
+      "epoch": 1.2278029812054438,
+      "grad_norm": 0.7559435367584229,
+      "learning_rate": 3.3395992609848804e-06,
+      "loss": 0.0998,
+      "step": 7578
+    },
+    {
+      "epoch": 1.2279650032404408,
+      "grad_norm": 0.8726533651351929,
+      "learning_rate": 3.339187343034257e-06,
+      "loss": 0.1108,
+      "step": 7579
+    },
+    {
+      "epoch": 1.2281270252754375,
+      "grad_norm": 0.7859377264976501,
+      "learning_rate": 3.338775399406531e-06,
+      "loss": 0.102,
+      "step": 7580
+    },
+    {
+      "epoch": 1.2282890473104342,
+      "grad_norm": 0.8078779578208923,
+      "learning_rate": 3.338363430114308e-06,
+      "loss": 0.0931,
+      "step": 7581
+    },
+    {
+      "epoch": 1.228451069345431,
+      "grad_norm": 0.9209697842597961,
+      "learning_rate": 3.3379514351701924e-06,
+      "loss": 0.112,
+      "step": 7582
+    },
+    {
+      "epoch": 1.2286130913804278,
+      "grad_norm": 0.8043226599693298,
+      "learning_rate": 3.3375394145867912e-06,
+      "loss": 0.0952,
+      "step": 7583
+    },
+    {
+      "epoch": 1.2287751134154246,
+      "grad_norm": 0.8135362267494202,
+      "learning_rate": 3.3371273683767102e-06,
+      "loss": 0.1067,
+      "step": 7584
+    },
+    {
+      "epoch": 1.2289371354504213,
+      "grad_norm": 0.7498790621757507,
+      "learning_rate": 3.336715296552558e-06,
+      "loss": 0.0919,
+      "step": 7585
+    },
+    {
+      "epoch": 1.229099157485418,
+      "grad_norm": 0.8408077359199524,
+      "learning_rate": 3.3363031991269423e-06,
+      "loss": 0.101,
+      "step": 7586
+    },
+    {
+      "epoch": 1.2292611795204147,
+      "grad_norm": 0.8685369491577148,
+      "learning_rate": 3.3358910761124724e-06,
+      "loss": 0.1081,
+      "step": 7587
+    },
+    {
+      "epoch": 1.2294232015554116,
+      "grad_norm": 0.9170325994491577,
+      "learning_rate": 3.3354789275217587e-06,
+      "loss": 0.1099,
+      "step": 7588
+    },
+    {
+      "epoch": 1.2295852235904083,
+      "grad_norm": 0.8715429306030273,
+      "learning_rate": 3.3350667533674108e-06,
+      "loss": 0.108,
+      "step": 7589
+    },
+    {
+      "epoch": 1.229747245625405,
+      "grad_norm": 0.870263397693634,
+      "learning_rate": 3.3346545536620425e-06,
+      "loss": 0.1057,
+      "step": 7590
+    },
+    {
+      "epoch": 1.2299092676604018,
+      "grad_norm": 0.7764174938201904,
+      "learning_rate": 3.334242328418264e-06,
+      "loss": 0.1045,
+      "step": 7591
+    },
+    {
+      "epoch": 1.2300712896953985,
+      "grad_norm": 0.8364241719245911,
+      "learning_rate": 3.3338300776486886e-06,
+      "loss": 0.1014,
+      "step": 7592
+    },
+    {
+      "epoch": 1.2302333117303954,
+      "grad_norm": 0.8532553315162659,
+      "learning_rate": 3.333417801365931e-06,
+      "loss": 0.1043,
+      "step": 7593
+    },
+    {
+      "epoch": 1.2303953337653921,
+      "grad_norm": 0.795436441898346,
+      "learning_rate": 3.3330054995826056e-06,
+      "loss": 0.1042,
+      "step": 7594
+    },
+    {
+      "epoch": 1.2305573558003888,
+      "grad_norm": 0.7866604328155518,
+      "learning_rate": 3.332593172311328e-06,
+      "loss": 0.0932,
+      "step": 7595
+    },
+    {
+      "epoch": 1.2307193778353855,
+      "grad_norm": 0.8465556502342224,
+      "learning_rate": 3.3321808195647144e-06,
+      "loss": 0.1037,
+      "step": 7596
+    },
+    {
+      "epoch": 1.2308813998703823,
+      "grad_norm": 0.8134123086929321,
+      "learning_rate": 3.33176844135538e-06,
+      "loss": 0.1038,
+      "step": 7597
+    },
+    {
+      "epoch": 1.2310434219053792,
+      "grad_norm": 0.7845098972320557,
+      "learning_rate": 3.3313560376959456e-06,
+      "loss": 0.103,
+      "step": 7598
+    },
+    {
+      "epoch": 1.231205443940376,
+      "grad_norm": 0.8044149279594421,
+      "learning_rate": 3.330943608599028e-06,
+      "loss": 0.0938,
+      "step": 7599
+    },
+    {
+      "epoch": 1.2313674659753726,
+      "grad_norm": 0.7864116430282593,
+      "learning_rate": 3.3305311540772467e-06,
+      "loss": 0.0998,
+      "step": 7600
+    },
+    {
+      "epoch": 1.2315294880103693,
+      "grad_norm": 0.8183262944221497,
+      "learning_rate": 3.3301186741432217e-06,
+      "loss": 0.0995,
+      "step": 7601
+    },
+    {
+      "epoch": 1.2316915100453663,
+      "grad_norm": 0.9076040983200073,
+      "learning_rate": 3.3297061688095746e-06,
+      "loss": 0.1081,
+      "step": 7602
+    },
+    {
+      "epoch": 1.231853532080363,
+      "grad_norm": 0.6733155250549316,
+      "learning_rate": 3.3292936380889262e-06,
+      "loss": 0.0878,
+      "step": 7603
+    },
+    {
+      "epoch": 1.2320155541153597,
+      "grad_norm": 0.8160568475723267,
+      "learning_rate": 3.3288810819938995e-06,
+      "loss": 0.1036,
+      "step": 7604
+    },
+    {
+      "epoch": 1.2321775761503564,
+      "grad_norm": 0.757789134979248,
+      "learning_rate": 3.3284685005371176e-06,
+      "loss": 0.0993,
+      "step": 7605
+    },
+    {
+      "epoch": 1.2323395981853533,
+      "grad_norm": 0.8649821877479553,
+      "learning_rate": 3.3280558937312037e-06,
+      "loss": 0.1062,
+      "step": 7606
+    },
+    {
+      "epoch": 1.23250162022035,
+      "grad_norm": 0.8108959794044495,
+      "learning_rate": 3.3276432615887843e-06,
+      "loss": 0.0959,
+      "step": 7607
+    },
+    {
+      "epoch": 1.2326636422553467,
+      "grad_norm": 0.7675016522407532,
+      "learning_rate": 3.327230604122484e-06,
+      "loss": 0.0941,
+      "step": 7608
+    },
+    {
+      "epoch": 1.2328256642903435,
+      "grad_norm": 0.8105819225311279,
+      "learning_rate": 3.326817921344928e-06,
+      "loss": 0.1078,
+      "step": 7609
+    },
+    {
+      "epoch": 1.2329876863253402,
+      "grad_norm": 0.7653875946998596,
+      "learning_rate": 3.326405213268745e-06,
+      "loss": 0.0968,
+      "step": 7610
+    },
+    {
+      "epoch": 1.233149708360337,
+      "grad_norm": 0.7907019257545471,
+      "learning_rate": 3.3259924799065628e-06,
+      "loss": 0.1014,
+      "step": 7611
+    },
+    {
+      "epoch": 1.2333117303953338,
+      "grad_norm": 0.9260424971580505,
+      "learning_rate": 3.3255797212710095e-06,
+      "loss": 0.1051,
+      "step": 7612
+    },
+    {
+      "epoch": 1.2334737524303305,
+      "grad_norm": 0.8345485925674438,
+      "learning_rate": 3.325166937374714e-06,
+      "loss": 0.1032,
+      "step": 7613
+    },
+    {
+      "epoch": 1.2336357744653272,
+      "grad_norm": 0.9643160104751587,
+      "learning_rate": 3.3247541282303082e-06,
+      "loss": 0.1185,
+      "step": 7614
+    },
+    {
+      "epoch": 1.233797796500324,
+      "grad_norm": 0.8535207509994507,
+      "learning_rate": 3.3243412938504205e-06,
+      "loss": 0.1094,
+      "step": 7615
+    },
+    {
+      "epoch": 1.2339598185353209,
+      "grad_norm": 0.8478245735168457,
+      "learning_rate": 3.3239284342476852e-06,
+      "loss": 0.1035,
+      "step": 7616
+    },
+    {
+      "epoch": 1.2341218405703176,
+      "grad_norm": 0.8517696261405945,
+      "learning_rate": 3.3235155494347325e-06,
+      "loss": 0.1113,
+      "step": 7617
+    },
+    {
+      "epoch": 1.2342838626053143,
+      "grad_norm": 1.065172791481018,
+      "learning_rate": 3.3231026394241983e-06,
+      "loss": 0.1272,
+      "step": 7618
+    },
+    {
+      "epoch": 1.234445884640311,
+      "grad_norm": 0.9062070250511169,
+      "learning_rate": 3.3226897042287145e-06,
+      "loss": 0.1209,
+      "step": 7619
+    },
+    {
+      "epoch": 1.2346079066753077,
+      "grad_norm": 0.8547554612159729,
+      "learning_rate": 3.3222767438609166e-06,
+      "loss": 0.1091,
+      "step": 7620
+    },
+    {
+      "epoch": 1.2347699287103047,
+      "grad_norm": 0.7908268570899963,
+      "learning_rate": 3.32186375833344e-06,
+      "loss": 0.0999,
+      "step": 7621
+    },
+    {
+      "epoch": 1.2349319507453014,
+      "grad_norm": 0.7787481546401978,
+      "learning_rate": 3.321450747658922e-06,
+      "loss": 0.0974,
+      "step": 7622
+    },
+    {
+      "epoch": 1.235093972780298,
+      "grad_norm": 0.8876661062240601,
+      "learning_rate": 3.321037711849998e-06,
+      "loss": 0.1232,
+      "step": 7623
+    },
+    {
+      "epoch": 1.2352559948152948,
+      "grad_norm": 0.8357555270195007,
+      "learning_rate": 3.3206246509193076e-06,
+      "loss": 0.1095,
+      "step": 7624
+    },
+    {
+      "epoch": 1.2354180168502917,
+      "grad_norm": 0.8862370848655701,
+      "learning_rate": 3.320211564879488e-06,
+      "loss": 0.1157,
+      "step": 7625
+    },
+    {
+      "epoch": 1.2355800388852884,
+      "grad_norm": 0.835319459438324,
+      "learning_rate": 3.3197984537431797e-06,
+      "loss": 0.1024,
+      "step": 7626
+    },
+    {
+      "epoch": 1.2357420609202852,
+      "grad_norm": 0.8422128558158875,
+      "learning_rate": 3.319385317523024e-06,
+      "loss": 0.1071,
+      "step": 7627
+    },
+    {
+      "epoch": 1.2359040829552819,
+      "grad_norm": 0.7388719916343689,
+      "learning_rate": 3.3189721562316585e-06,
+      "loss": 0.0847,
+      "step": 7628
+    },
+    {
+      "epoch": 1.2360661049902788,
+      "grad_norm": 0.7805131673812866,
+      "learning_rate": 3.318558969881728e-06,
+      "loss": 0.0995,
+      "step": 7629
+    },
+    {
+      "epoch": 1.2362281270252755,
+      "grad_norm": 1.0341185331344604,
+      "learning_rate": 3.3181457584858736e-06,
+      "loss": 0.1243,
+      "step": 7630
+    },
+    {
+      "epoch": 1.2363901490602722,
+      "grad_norm": 0.7121812701225281,
+      "learning_rate": 3.3177325220567385e-06,
+      "loss": 0.0965,
+      "step": 7631
+    },
+    {
+      "epoch": 1.236552171095269,
+      "grad_norm": 0.7928653955459595,
+      "learning_rate": 3.3173192606069673e-06,
+      "loss": 0.1043,
+      "step": 7632
+    },
+    {
+      "epoch": 1.2367141931302656,
+      "grad_norm": 0.7301590442657471,
+      "learning_rate": 3.316905974149205e-06,
+      "loss": 0.0967,
+      "step": 7633
+    },
+    {
+      "epoch": 1.2368762151652626,
+      "grad_norm": 0.9105049967765808,
+      "learning_rate": 3.316492662696097e-06,
+      "loss": 0.1122,
+      "step": 7634
+    },
+    {
+      "epoch": 1.2370382372002593,
+      "grad_norm": 0.8258975148200989,
+      "learning_rate": 3.3160793262602882e-06,
+      "loss": 0.1109,
+      "step": 7635
+    },
+    {
+      "epoch": 1.237200259235256,
+      "grad_norm": 0.9241932034492493,
+      "learning_rate": 3.3156659648544276e-06,
+      "loss": 0.1102,
+      "step": 7636
+    },
+    {
+      "epoch": 1.2373622812702527,
+      "grad_norm": 0.8554189205169678,
+      "learning_rate": 3.315252578491162e-06,
+      "loss": 0.1104,
+      "step": 7637
+    },
+    {
+      "epoch": 1.2375243033052494,
+      "grad_norm": 0.9000004529953003,
+      "learning_rate": 3.314839167183141e-06,
+      "loss": 0.1058,
+      "step": 7638
+    },
+    {
+      "epoch": 1.2376863253402464,
+      "grad_norm": 0.7197251319885254,
+      "learning_rate": 3.3144257309430127e-06,
+      "loss": 0.0863,
+      "step": 7639
+    },
+    {
+      "epoch": 1.237848347375243,
+      "grad_norm": 0.948573887348175,
+      "learning_rate": 3.3140122697834287e-06,
+      "loss": 0.1086,
+      "step": 7640
+    },
+    {
+      "epoch": 1.2380103694102398,
+      "grad_norm": 0.8102221488952637,
+      "learning_rate": 3.3135987837170386e-06,
+      "loss": 0.1001,
+      "step": 7641
+    },
+    {
+      "epoch": 1.2381723914452365,
+      "grad_norm": 0.7119641304016113,
+      "learning_rate": 3.3131852727564947e-06,
+      "loss": 0.0848,
+      "step": 7642
+    },
+    {
+      "epoch": 1.2383344134802332,
+      "grad_norm": 0.8429257869720459,
+      "learning_rate": 3.31277173691445e-06,
+      "loss": 0.1068,
+      "step": 7643
+    },
+    {
+      "epoch": 1.2384964355152301,
+      "grad_norm": 0.7876508831977844,
+      "learning_rate": 3.3123581762035557e-06,
+      "loss": 0.0945,
+      "step": 7644
+    },
+    {
+      "epoch": 1.2386584575502269,
+      "grad_norm": 0.9862732291221619,
+      "learning_rate": 3.311944590636468e-06,
+      "loss": 0.1212,
+      "step": 7645
+    },
+    {
+      "epoch": 1.2388204795852236,
+      "grad_norm": 0.8856245279312134,
+      "learning_rate": 3.31153098022584e-06,
+      "loss": 0.1084,
+      "step": 7646
+    },
+    {
+      "epoch": 1.2389825016202203,
+      "grad_norm": 0.9834122657775879,
+      "learning_rate": 3.3111173449843283e-06,
+      "loss": 0.1125,
+      "step": 7647
+    },
+    {
+      "epoch": 1.239144523655217,
+      "grad_norm": 0.7890782356262207,
+      "learning_rate": 3.3107036849245883e-06,
+      "loss": 0.0953,
+      "step": 7648
+    },
+    {
+      "epoch": 1.239306545690214,
+      "grad_norm": 0.8613621592521667,
+      "learning_rate": 3.310290000059278e-06,
+      "loss": 0.1074,
+      "step": 7649
+    },
+    {
+      "epoch": 1.2394685677252106,
+      "grad_norm": 0.8193936944007874,
+      "learning_rate": 3.309876290401054e-06,
+      "loss": 0.1,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2396305897602073,
+      "grad_norm": 0.7670435309410095,
+      "learning_rate": 3.309462555962576e-06,
+      "loss": 0.0993,
+      "step": 7651
+    },
+    {
+      "epoch": 1.239792611795204,
+      "grad_norm": 0.9082610011100769,
+      "learning_rate": 3.309048796756503e-06,
+      "loss": 0.1142,
+      "step": 7652
+    },
+    {
+      "epoch": 1.239954633830201,
+      "grad_norm": 0.9768080115318298,
+      "learning_rate": 3.3086350127954935e-06,
+      "loss": 0.1186,
+      "step": 7653
+    },
+    {
+      "epoch": 1.2401166558651977,
+      "grad_norm": 0.7574945092201233,
+      "learning_rate": 3.3082212040922103e-06,
+      "loss": 0.0913,
+      "step": 7654
+    },
+    {
+      "epoch": 1.2402786779001944,
+      "grad_norm": 0.8311034440994263,
+      "learning_rate": 3.3078073706593133e-06,
+      "loss": 0.1038,
+      "step": 7655
+    },
+    {
+      "epoch": 1.2404406999351911,
+      "grad_norm": 0.7912598252296448,
+      "learning_rate": 3.307393512509466e-06,
+      "loss": 0.1065,
+      "step": 7656
+    },
+    {
+      "epoch": 1.240602721970188,
+      "grad_norm": 0.8380460143089294,
+      "learning_rate": 3.3069796296553316e-06,
+      "loss": 0.1073,
+      "step": 7657
+    },
+    {
+      "epoch": 1.2407647440051848,
+      "grad_norm": 0.8099128007888794,
+      "learning_rate": 3.3065657221095732e-06,
+      "loss": 0.0964,
+      "step": 7658
+    },
+    {
+      "epoch": 1.2409267660401815,
+      "grad_norm": 0.6533287167549133,
+      "learning_rate": 3.306151789884855e-06,
+      "loss": 0.0861,
+      "step": 7659
+    },
+    {
+      "epoch": 1.2410887880751782,
+      "grad_norm": 0.6852719187736511,
+      "learning_rate": 3.3057378329938432e-06,
+      "loss": 0.0837,
+      "step": 7660
+    },
+    {
+      "epoch": 1.241250810110175,
+      "grad_norm": 0.8555914163589478,
+      "learning_rate": 3.305323851449203e-06,
+      "loss": 0.1047,
+      "step": 7661
+    },
+    {
+      "epoch": 1.2414128321451718,
+      "grad_norm": 0.8162693381309509,
+      "learning_rate": 3.304909845263603e-06,
+      "loss": 0.1062,
+      "step": 7662
+    },
+    {
+      "epoch": 1.2415748541801686,
+      "grad_norm": 0.8287134766578674,
+      "learning_rate": 3.3044958144497086e-06,
+      "loss": 0.1038,
+      "step": 7663
+    },
+    {
+      "epoch": 1.2417368762151653,
+      "grad_norm": 0.7691091299057007,
+      "learning_rate": 3.3040817590201897e-06,
+      "loss": 0.0984,
+      "step": 7664
+    },
+    {
+      "epoch": 1.241898898250162,
+      "grad_norm": 0.8376115560531616,
+      "learning_rate": 3.3036676789877146e-06,
+      "loss": 0.1,
+      "step": 7665
+    },
+    {
+      "epoch": 1.2420609202851587,
+      "grad_norm": 0.7699323296546936,
+      "learning_rate": 3.303253574364953e-06,
+      "loss": 0.0892,
+      "step": 7666
+    },
+    {
+      "epoch": 1.2422229423201556,
+      "grad_norm": 0.8108561038970947,
+      "learning_rate": 3.3028394451645766e-06,
+      "loss": 0.1031,
+      "step": 7667
+    },
+    {
+      "epoch": 1.2423849643551523,
+      "grad_norm": 0.8874189257621765,
+      "learning_rate": 3.3024252913992548e-06,
+      "loss": 0.11,
+      "step": 7668
+    },
+    {
+      "epoch": 1.242546986390149,
+      "grad_norm": 0.7574527859687805,
+      "learning_rate": 3.3020111130816623e-06,
+      "loss": 0.0989,
+      "step": 7669
+    },
+    {
+      "epoch": 1.2427090084251458,
+      "grad_norm": 0.9647727608680725,
+      "learning_rate": 3.3015969102244704e-06,
+      "loss": 0.1051,
+      "step": 7670
+    },
+    {
+      "epoch": 1.2428710304601425,
+      "grad_norm": 0.9136444926261902,
+      "learning_rate": 3.3011826828403527e-06,
+      "loss": 0.1138,
+      "step": 7671
+    },
+    {
+      "epoch": 1.2430330524951394,
+      "grad_norm": 0.9177752733230591,
+      "learning_rate": 3.300768430941983e-06,
+      "loss": 0.1238,
+      "step": 7672
+    },
+    {
+      "epoch": 1.2431950745301361,
+      "grad_norm": 0.7569882869720459,
+      "learning_rate": 3.3003541545420377e-06,
+      "loss": 0.1009,
+      "step": 7673
+    },
+    {
+      "epoch": 1.2433570965651328,
+      "grad_norm": 0.7180430889129639,
+      "learning_rate": 3.299939853653192e-06,
+      "loss": 0.0873,
+      "step": 7674
+    },
+    {
+      "epoch": 1.2435191186001295,
+      "grad_norm": 0.7186658978462219,
+      "learning_rate": 3.2995255282881227e-06,
+      "loss": 0.0835,
+      "step": 7675
+    },
+    {
+      "epoch": 1.2436811406351265,
+      "grad_norm": 0.8784341812133789,
+      "learning_rate": 3.299111178459507e-06,
+      "loss": 0.1175,
+      "step": 7676
+    },
+    {
+      "epoch": 1.2438431626701232,
+      "grad_norm": 0.8925139307975769,
+      "learning_rate": 3.2986968041800234e-06,
+      "loss": 0.1186,
+      "step": 7677
+    },
+    {
+      "epoch": 1.24400518470512,
+      "grad_norm": 0.8931295275688171,
+      "learning_rate": 3.29828240546235e-06,
+      "loss": 0.1053,
+      "step": 7678
+    },
+    {
+      "epoch": 1.2441672067401166,
+      "grad_norm": 0.8283528685569763,
+      "learning_rate": 3.297867982319166e-06,
+      "loss": 0.1039,
+      "step": 7679
+    },
+    {
+      "epoch": 1.2443292287751135,
+      "grad_norm": 0.9566637873649597,
+      "learning_rate": 3.297453534763154e-06,
+      "loss": 0.1112,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2444912508101102,
+      "grad_norm": 0.8479618430137634,
+      "learning_rate": 3.2970390628069924e-06,
+      "loss": 0.107,
+      "step": 7681
+    },
+    {
+      "epoch": 1.244653272845107,
+      "grad_norm": 0.7277421951293945,
+      "learning_rate": 3.2966245664633654e-06,
+      "loss": 0.0957,
+      "step": 7682
+    },
+    {
+      "epoch": 1.2448152948801037,
+      "grad_norm": 0.7973564863204956,
+      "learning_rate": 3.296210045744954e-06,
+      "loss": 0.0933,
+      "step": 7683
+    },
+    {
+      "epoch": 1.2449773169151004,
+      "grad_norm": 0.8752889037132263,
+      "learning_rate": 3.295795500664442e-06,
+      "loss": 0.1146,
+      "step": 7684
+    },
+    {
+      "epoch": 1.2451393389500973,
+      "grad_norm": 0.772816002368927,
+      "learning_rate": 3.295380931234513e-06,
+      "loss": 0.0988,
+      "step": 7685
+    },
+    {
+      "epoch": 1.245301360985094,
+      "grad_norm": 1.0557066202163696,
+      "learning_rate": 3.294966337467853e-06,
+      "loss": 0.1175,
+      "step": 7686
+    },
+    {
+      "epoch": 1.2454633830200907,
+      "grad_norm": 0.7834747433662415,
+      "learning_rate": 3.294551719377147e-06,
+      "loss": 0.1024,
+      "step": 7687
+    },
+    {
+      "epoch": 1.2456254050550875,
+      "grad_norm": 0.8686286807060242,
+      "learning_rate": 3.2941370769750804e-06,
+      "loss": 0.1102,
+      "step": 7688
+    },
+    {
+      "epoch": 1.2457874270900842,
+      "grad_norm": 0.8577167391777039,
+      "learning_rate": 3.2937224102743414e-06,
+      "loss": 0.1075,
+      "step": 7689
+    },
+    {
+      "epoch": 1.245949449125081,
+      "grad_norm": 0.712014377117157,
+      "learning_rate": 3.293307719287617e-06,
+      "loss": 0.0949,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2461114711600778,
+      "grad_norm": 0.7292968034744263,
+      "learning_rate": 3.292893004027597e-06,
+      "loss": 0.0953,
+      "step": 7691
+    },
+    {
+      "epoch": 1.2462734931950745,
+      "grad_norm": 0.8187505006790161,
+      "learning_rate": 3.2924782645069684e-06,
+      "loss": 0.0978,
+      "step": 7692
+    },
+    {
+      "epoch": 1.2464355152300712,
+      "grad_norm": 0.8897111415863037,
+      "learning_rate": 3.292063500738424e-06,
+      "loss": 0.1111,
+      "step": 7693
+    },
+    {
+      "epoch": 1.246597537265068,
+      "grad_norm": 0.806907057762146,
+      "learning_rate": 3.291648712734653e-06,
+      "loss": 0.096,
+      "step": 7694
+    },
+    {
+      "epoch": 1.2467595593000649,
+      "grad_norm": 0.8400275111198425,
+      "learning_rate": 3.2912339005083473e-06,
+      "loss": 0.0945,
+      "step": 7695
+    },
+    {
+      "epoch": 1.2469215813350616,
+      "grad_norm": 0.8764445185661316,
+      "learning_rate": 3.290819064072198e-06,
+      "loss": 0.1005,
+      "step": 7696
+    },
+    {
+      "epoch": 1.2470836033700583,
+      "grad_norm": 0.994835615158081,
+      "learning_rate": 3.2904042034389e-06,
+      "loss": 0.1262,
+      "step": 7697
+    },
+    {
+      "epoch": 1.247245625405055,
+      "grad_norm": 0.8427457213401794,
+      "learning_rate": 3.289989318621146e-06,
+      "loss": 0.1038,
+      "step": 7698
+    },
+    {
+      "epoch": 1.2474076474400517,
+      "grad_norm": 0.8053460717201233,
+      "learning_rate": 3.289574409631631e-06,
+      "loss": 0.0978,
+      "step": 7699
+    },
+    {
+      "epoch": 1.2475696694750487,
+      "grad_norm": 0.9428670406341553,
+      "learning_rate": 3.289159476483049e-06,
+      "loss": 0.1128,
+      "step": 7700
+    },
+    {
+      "epoch": 1.2477316915100454,
+      "grad_norm": 0.9878700375556946,
+      "learning_rate": 3.288744519188097e-06,
+      "loss": 0.1245,
+      "step": 7701
+    },
+    {
+      "epoch": 1.247893713545042,
+      "grad_norm": 0.84808748960495,
+      "learning_rate": 3.2883295377594716e-06,
+      "loss": 0.0883,
+      "step": 7702
+    },
+    {
+      "epoch": 1.2480557355800388,
+      "grad_norm": 0.8799005746841431,
+      "learning_rate": 3.2879145322098694e-06,
+      "loss": 0.1154,
+      "step": 7703
+    },
+    {
+      "epoch": 1.2482177576150357,
+      "grad_norm": 0.9856189489364624,
+      "learning_rate": 3.2874995025519897e-06,
+      "loss": 0.115,
+      "step": 7704
+    },
+    {
+      "epoch": 1.2483797796500324,
+      "grad_norm": 0.798912525177002,
+      "learning_rate": 3.2870844487985307e-06,
+      "loss": 0.0959,
+      "step": 7705
+    },
+    {
+      "epoch": 1.2485418016850292,
+      "grad_norm": 0.8618741035461426,
+      "learning_rate": 3.2866693709621933e-06,
+      "loss": 0.1131,
+      "step": 7706
+    },
+    {
+      "epoch": 1.2487038237200259,
+      "grad_norm": 0.8063377737998962,
+      "learning_rate": 3.2862542690556765e-06,
+      "loss": 0.1011,
+      "step": 7707
+    },
+    {
+      "epoch": 1.2488658457550228,
+      "grad_norm": 0.7270522713661194,
+      "learning_rate": 3.285839143091681e-06,
+      "loss": 0.098,
+      "step": 7708
+    },
+    {
+      "epoch": 1.2490278677900195,
+      "grad_norm": 0.7814246416091919,
+      "learning_rate": 3.2854239930829097e-06,
+      "loss": 0.098,
+      "step": 7709
+    },
+    {
+      "epoch": 1.2491898898250162,
+      "grad_norm": 0.7192318439483643,
+      "learning_rate": 3.2850088190420647e-06,
+      "loss": 0.0922,
+      "step": 7710
+    },
+    {
+      "epoch": 1.249351911860013,
+      "grad_norm": 0.7969322204589844,
+      "learning_rate": 3.284593620981851e-06,
+      "loss": 0.1072,
+      "step": 7711
+    },
+    {
+      "epoch": 1.2495139338950096,
+      "grad_norm": 0.8182503581047058,
+      "learning_rate": 3.284178398914969e-06,
+      "loss": 0.1014,
+      "step": 7712
+    },
+    {
+      "epoch": 1.2496759559300066,
+      "grad_norm": 0.9507556557655334,
+      "learning_rate": 3.283763152854127e-06,
+      "loss": 0.1117,
+      "step": 7713
+    },
+    {
+      "epoch": 1.2498379779650033,
+      "grad_norm": 0.7872381210327148,
+      "learning_rate": 3.283347882812028e-06,
+      "loss": 0.0951,
+      "step": 7714
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7765859961509705,
+      "learning_rate": 3.282932588801381e-06,
+      "loss": 0.0978,
+      "step": 7715
+    },
+    {
+      "epoch": 1.2501620220349967,
+      "grad_norm": 0.9354822635650635,
+      "learning_rate": 3.282517270834891e-06,
+      "loss": 0.1167,
+      "step": 7716
+    },
+    {
+      "epoch": 1.2503240440699934,
+      "grad_norm": 0.8700661659240723,
+      "learning_rate": 3.2821019289252654e-06,
+      "loss": 0.1112,
+      "step": 7717
+    },
+    {
+      "epoch": 1.2504860661049904,
+      "grad_norm": 0.8417790532112122,
+      "learning_rate": 3.281686563085214e-06,
+      "loss": 0.1065,
+      "step": 7718
+    },
+    {
+      "epoch": 1.250648088139987,
+      "grad_norm": 0.8531923294067383,
+      "learning_rate": 3.2812711733274453e-06,
+      "loss": 0.1171,
+      "step": 7719
+    },
+    {
+      "epoch": 1.2508101101749838,
+      "grad_norm": 0.8502379059791565,
+      "learning_rate": 3.28085575966467e-06,
+      "loss": 0.1064,
+      "step": 7720
+    },
+    {
+      "epoch": 1.2509721322099805,
+      "grad_norm": 0.9494119882583618,
+      "learning_rate": 3.280440322109597e-06,
+      "loss": 0.1099,
+      "step": 7721
+    },
+    {
+      "epoch": 1.2511341542449772,
+      "grad_norm": 0.7425954341888428,
+      "learning_rate": 3.2800248606749395e-06,
+      "loss": 0.0846,
+      "step": 7722
+    },
+    {
+      "epoch": 1.2512961762799741,
+      "grad_norm": 0.7303003668785095,
+      "learning_rate": 3.2796093753734087e-06,
+      "loss": 0.098,
+      "step": 7723
+    },
+    {
+      "epoch": 1.2514581983149708,
+      "grad_norm": 0.9115877151489258,
+      "learning_rate": 3.2791938662177174e-06,
+      "loss": 0.1093,
+      "step": 7724
+    },
+    {
+      "epoch": 1.2516202203499676,
+      "grad_norm": 0.7848078608512878,
+      "learning_rate": 3.2787783332205796e-06,
+      "loss": 0.0934,
+      "step": 7725
+    },
+    {
+      "epoch": 1.2517822423849643,
+      "grad_norm": 0.8890565037727356,
+      "learning_rate": 3.278362776394709e-06,
+      "loss": 0.1088,
+      "step": 7726
+    },
+    {
+      "epoch": 1.251944264419961,
+      "grad_norm": 1.0119807720184326,
+      "learning_rate": 3.277947195752822e-06,
+      "loss": 0.1263,
+      "step": 7727
+    },
+    {
+      "epoch": 1.252106286454958,
+      "grad_norm": 0.8432857394218445,
+      "learning_rate": 3.277531591307632e-06,
+      "loss": 0.1109,
+      "step": 7728
+    },
+    {
+      "epoch": 1.2522683084899546,
+      "grad_norm": 0.8839412331581116,
+      "learning_rate": 3.2771159630718584e-06,
+      "loss": 0.1058,
+      "step": 7729
+    },
+    {
+      "epoch": 1.2524303305249513,
+      "grad_norm": 0.8032359480857849,
+      "learning_rate": 3.2767003110582164e-06,
+      "loss": 0.0991,
+      "step": 7730
+    },
+    {
+      "epoch": 1.2525923525599483,
+      "grad_norm": 0.7420846819877625,
+      "learning_rate": 3.276284635279424e-06,
+      "loss": 0.0962,
+      "step": 7731
+    },
+    {
+      "epoch": 1.252754374594945,
+      "grad_norm": 0.7983898520469666,
+      "learning_rate": 3.275868935748201e-06,
+      "loss": 0.1003,
+      "step": 7732
+    },
+    {
+      "epoch": 1.2529163966299417,
+      "grad_norm": 0.8384861946105957,
+      "learning_rate": 3.2754532124772653e-06,
+      "loss": 0.1081,
+      "step": 7733
+    },
+    {
+      "epoch": 1.2530784186649384,
+      "grad_norm": 0.8961964845657349,
+      "learning_rate": 3.2750374654793387e-06,
+      "loss": 0.1239,
+      "step": 7734
+    },
+    {
+      "epoch": 1.2532404406999351,
+      "grad_norm": 0.818870484828949,
+      "learning_rate": 3.2746216947671405e-06,
+      "loss": 0.1061,
+      "step": 7735
+    },
+    {
+      "epoch": 1.253402462734932,
+      "grad_norm": 0.8016375303268433,
+      "learning_rate": 3.2742059003533933e-06,
+      "loss": 0.0954,
+      "step": 7736
+    },
+    {
+      "epoch": 1.2535644847699288,
+      "grad_norm": 0.9985673427581787,
+      "learning_rate": 3.2737900822508197e-06,
+      "loss": 0.1111,
+      "step": 7737
+    },
+    {
+      "epoch": 1.2537265068049255,
+      "grad_norm": 0.8439716100692749,
+      "learning_rate": 3.2733742404721413e-06,
+      "loss": 0.1127,
+      "step": 7738
+    },
+    {
+      "epoch": 1.2538885288399222,
+      "grad_norm": 0.7654055953025818,
+      "learning_rate": 3.272958375030083e-06,
+      "loss": 0.1004,
+      "step": 7739
+    },
+    {
+      "epoch": 1.254050550874919,
+      "grad_norm": 0.7196124792098999,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 0.0928,
+      "step": 7740
+    },
+    {
+      "epoch": 1.2542125729099158,
+      "grad_norm": 0.9231361150741577,
+      "learning_rate": 3.272126573206724e-06,
+      "loss": 0.1016,
+      "step": 7741
+    },
+    {
+      "epoch": 1.2543745949449125,
+      "grad_norm": 0.7543466687202454,
+      "learning_rate": 3.2717106368508755e-06,
+      "loss": 0.0929,
+      "step": 7742
+    },
+    {
+      "epoch": 1.2545366169799093,
+      "grad_norm": 0.8236292600631714,
+      "learning_rate": 3.271294676882548e-06,
+      "loss": 0.1005,
+      "step": 7743
+    },
+    {
+      "epoch": 1.254698639014906,
+      "grad_norm": 0.8038961887359619,
+      "learning_rate": 3.27087869331447e-06,
+      "loss": 0.101,
+      "step": 7744
+    },
+    {
+      "epoch": 1.2548606610499027,
+      "grad_norm": 0.8604749441146851,
+      "learning_rate": 3.270462686159369e-06,
+      "loss": 0.1037,
+      "step": 7745
+    },
+    {
+      "epoch": 1.2550226830848996,
+      "grad_norm": 0.744813859462738,
+      "learning_rate": 3.2700466554299755e-06,
+      "loss": 0.0934,
+      "step": 7746
+    },
+    {
+      "epoch": 1.2551847051198963,
+      "grad_norm": 0.8217744827270508,
+      "learning_rate": 3.2696306011390167e-06,
+      "loss": 0.106,
+      "step": 7747
+    },
+    {
+      "epoch": 1.255346727154893,
+      "grad_norm": 0.755609929561615,
+      "learning_rate": 3.2692145232992244e-06,
+      "loss": 0.0933,
+      "step": 7748
+    },
+    {
+      "epoch": 1.2555087491898898,
+      "grad_norm": 0.7776917219161987,
+      "learning_rate": 3.2687984219233295e-06,
+      "loss": 0.1035,
+      "step": 7749
+    },
+    {
+      "epoch": 1.2556707712248865,
+      "grad_norm": 0.8656284809112549,
+      "learning_rate": 3.268382297024063e-06,
+      "loss": 0.111,
+      "step": 7750
+    },
+    {
+      "epoch": 1.2558327932598834,
+      "grad_norm": 0.8041141629219055,
+      "learning_rate": 3.2679661486141577e-06,
+      "loss": 0.1009,
+      "step": 7751
+    },
+    {
+      "epoch": 1.25599481529488,
+      "grad_norm": 0.7365778088569641,
+      "learning_rate": 3.2675499767063464e-06,
+      "loss": 0.0935,
+      "step": 7752
+    },
+    {
+      "epoch": 1.2561568373298768,
+      "grad_norm": 0.8223831057548523,
+      "learning_rate": 3.267133781313364e-06,
+      "loss": 0.1034,
+      "step": 7753
+    },
+    {
+      "epoch": 1.2563188593648738,
+      "grad_norm": 0.8573729395866394,
+      "learning_rate": 3.266717562447944e-06,
+      "loss": 0.1009,
+      "step": 7754
+    },
+    {
+      "epoch": 1.2564808813998705,
+      "grad_norm": 0.9449803829193115,
+      "learning_rate": 3.2663013201228216e-06,
+      "loss": 0.1201,
+      "step": 7755
+    },
+    {
+      "epoch": 1.2566429034348672,
+      "grad_norm": 0.7694631814956665,
+      "learning_rate": 3.2658850543507336e-06,
+      "loss": 0.0973,
+      "step": 7756
+    },
+    {
+      "epoch": 1.2568049254698639,
+      "grad_norm": 1.0734034776687622,
+      "learning_rate": 3.265468765144416e-06,
+      "loss": 0.1069,
+      "step": 7757
+    },
+    {
+      "epoch": 1.2569669475048606,
+      "grad_norm": 0.9590340852737427,
+      "learning_rate": 3.2650524525166064e-06,
+      "loss": 0.1212,
+      "step": 7758
+    },
+    {
+      "epoch": 1.2571289695398575,
+      "grad_norm": 0.7816064357757568,
+      "learning_rate": 3.264636116480044e-06,
+      "loss": 0.1041,
+      "step": 7759
+    },
+    {
+      "epoch": 1.2572909915748542,
+      "grad_norm": 0.745184600353241,
+      "learning_rate": 3.2642197570474665e-06,
+      "loss": 0.0875,
+      "step": 7760
+    },
+    {
+      "epoch": 1.257453013609851,
+      "grad_norm": 0.8741956949234009,
+      "learning_rate": 3.2638033742316137e-06,
+      "loss": 0.1039,
+      "step": 7761
+    },
+    {
+      "epoch": 1.2576150356448477,
+      "grad_norm": 0.9357261061668396,
+      "learning_rate": 3.263386968045226e-06,
+      "loss": 0.1157,
+      "step": 7762
+    },
+    {
+      "epoch": 1.2577770576798444,
+      "grad_norm": 0.773931086063385,
+      "learning_rate": 3.2629705385010445e-06,
+      "loss": 0.0962,
+      "step": 7763
+    },
+    {
+      "epoch": 1.2579390797148413,
+      "grad_norm": 0.8367682695388794,
+      "learning_rate": 3.262554085611811e-06,
+      "loss": 0.1009,
+      "step": 7764
+    },
+    {
+      "epoch": 1.258101101749838,
+      "grad_norm": 0.8734476566314697,
+      "learning_rate": 3.2621376093902675e-06,
+      "loss": 0.1155,
+      "step": 7765
+    },
+    {
+      "epoch": 1.2582631237848347,
+      "grad_norm": 0.9115141034126282,
+      "learning_rate": 3.261721109849158e-06,
+      "loss": 0.1112,
+      "step": 7766
+    },
+    {
+      "epoch": 1.2584251458198314,
+      "grad_norm": 0.7026832103729248,
+      "learning_rate": 3.261304587001225e-06,
+      "loss": 0.0882,
+      "step": 7767
+    },
+    {
+      "epoch": 1.2585871678548282,
+      "grad_norm": 0.7548426389694214,
+      "learning_rate": 3.2608880408592148e-06,
+      "loss": 0.097,
+      "step": 7768
+    },
+    {
+      "epoch": 1.258749189889825,
+      "grad_norm": 0.736499011516571,
+      "learning_rate": 3.2604714714358716e-06,
+      "loss": 0.0918,
+      "step": 7769
+    },
+    {
+      "epoch": 1.2589112119248218,
+      "grad_norm": 0.7644399404525757,
+      "learning_rate": 3.2600548787439413e-06,
+      "loss": 0.094,
+      "step": 7770
+    },
+    {
+      "epoch": 1.2590732339598185,
+      "grad_norm": 0.8792965412139893,
+      "learning_rate": 3.2596382627961714e-06,
+      "loss": 0.1082,
+      "step": 7771
+    },
+    {
+      "epoch": 1.2592352559948152,
+      "grad_norm": 0.8653140664100647,
+      "learning_rate": 3.2592216236053086e-06,
+      "loss": 0.1051,
+      "step": 7772
+    },
+    {
+      "epoch": 1.259397278029812,
+      "grad_norm": 0.898324728012085,
+      "learning_rate": 3.2588049611841023e-06,
+      "loss": 0.107,
+      "step": 7773
+    },
+    {
+      "epoch": 1.2595593000648089,
+      "grad_norm": 0.8365474343299866,
+      "learning_rate": 3.2583882755452994e-06,
+      "loss": 0.102,
+      "step": 7774
+    },
+    {
+      "epoch": 1.2597213220998056,
+      "grad_norm": 0.8603896498680115,
+      "learning_rate": 3.2579715667016516e-06,
+      "loss": 0.1069,
+      "step": 7775
+    },
+    {
+      "epoch": 1.2598833441348023,
+      "grad_norm": 0.8619314432144165,
+      "learning_rate": 3.257554834665907e-06,
+      "loss": 0.1045,
+      "step": 7776
+    },
+    {
+      "epoch": 1.2600453661697992,
+      "grad_norm": 0.7768551111221313,
+      "learning_rate": 3.2571380794508183e-06,
+      "loss": 0.0992,
+      "step": 7777
+    },
+    {
+      "epoch": 1.2602073882047957,
+      "grad_norm": 0.8619160056114197,
+      "learning_rate": 3.2567213010691367e-06,
+      "loss": 0.1118,
+      "step": 7778
+    },
+    {
+      "epoch": 1.2603694102397927,
+      "grad_norm": 0.9219297170639038,
+      "learning_rate": 3.256304499533614e-06,
+      "loss": 0.1117,
+      "step": 7779
+    },
+    {
+      "epoch": 1.2605314322747894,
+      "grad_norm": 0.8448036313056946,
+      "learning_rate": 3.255887674857004e-06,
+      "loss": 0.104,
+      "step": 7780
+    },
+    {
+      "epoch": 1.260693454309786,
+      "grad_norm": 0.80354905128479,
+      "learning_rate": 3.255470827052061e-06,
+      "loss": 0.1035,
+      "step": 7781
+    },
+    {
+      "epoch": 1.260855476344783,
+      "grad_norm": 0.7997376918792725,
+      "learning_rate": 3.2550539561315385e-06,
+      "loss": 0.1034,
+      "step": 7782
+    },
+    {
+      "epoch": 1.2610174983797797,
+      "grad_norm": 0.7824645042419434,
+      "learning_rate": 3.2546370621081912e-06,
+      "loss": 0.0989,
+      "step": 7783
+    },
+    {
+      "epoch": 1.2611795204147764,
+      "grad_norm": 0.757111668586731,
+      "learning_rate": 3.2542201449947774e-06,
+      "loss": 0.1016,
+      "step": 7784
+    },
+    {
+      "epoch": 1.2613415424497731,
+      "grad_norm": 0.8208655714988708,
+      "learning_rate": 3.253803204804052e-06,
+      "loss": 0.1116,
+      "step": 7785
+    },
+    {
+      "epoch": 1.2615035644847699,
+      "grad_norm": 0.817868709564209,
+      "learning_rate": 3.2533862415487723e-06,
+      "loss": 0.1056,
+      "step": 7786
+    },
+    {
+      "epoch": 1.2616655865197668,
+      "grad_norm": 0.7902513146400452,
+      "learning_rate": 3.252969255241697e-06,
+      "loss": 0.1007,
+      "step": 7787
+    },
+    {
+      "epoch": 1.2618276085547635,
+      "grad_norm": 0.8338719010353088,
+      "learning_rate": 3.2525522458955843e-06,
+      "loss": 0.1024,
+      "step": 7788
+    },
+    {
+      "epoch": 1.2619896305897602,
+      "grad_norm": 0.8464298844337463,
+      "learning_rate": 3.2521352135231944e-06,
+      "loss": 0.1108,
+      "step": 7789
+    },
+    {
+      "epoch": 1.262151652624757,
+      "grad_norm": 0.809252142906189,
+      "learning_rate": 3.251718158137287e-06,
+      "loss": 0.1044,
+      "step": 7790
+    },
+    {
+      "epoch": 1.2623136746597536,
+      "grad_norm": 0.6791431903839111,
+      "learning_rate": 3.2513010797506236e-06,
+      "loss": 0.0965,
+      "step": 7791
+    },
+    {
+      "epoch": 1.2624756966947506,
+      "grad_norm": 0.7555492520332336,
+      "learning_rate": 3.2508839783759642e-06,
+      "loss": 0.1014,
+      "step": 7792
+    },
+    {
+      "epoch": 1.2626377187297473,
+      "grad_norm": 0.6891449093818665,
+      "learning_rate": 3.2504668540260732e-06,
+      "loss": 0.088,
+      "step": 7793
+    },
+    {
+      "epoch": 1.262799740764744,
+      "grad_norm": 0.770533561706543,
+      "learning_rate": 3.2500497067137116e-06,
+      "loss": 0.1035,
+      "step": 7794
+    },
+    {
+      "epoch": 1.2629617627997407,
+      "grad_norm": 0.7673627138137817,
+      "learning_rate": 3.2496325364516444e-06,
+      "loss": 0.0951,
+      "step": 7795
+    },
+    {
+      "epoch": 1.2631237848347374,
+      "grad_norm": 0.8320962190628052,
+      "learning_rate": 3.2492153432526356e-06,
+      "loss": 0.1048,
+      "step": 7796
+    },
+    {
+      "epoch": 1.2632858068697344,
+      "grad_norm": 0.8069162368774414,
+      "learning_rate": 3.248798127129451e-06,
+      "loss": 0.1025,
+      "step": 7797
+    },
+    {
+      "epoch": 1.263447828904731,
+      "grad_norm": 0.9537956118583679,
+      "learning_rate": 3.2483808880948552e-06,
+      "loss": 0.1023,
+      "step": 7798
+    },
+    {
+      "epoch": 1.2636098509397278,
+      "grad_norm": 0.9524184465408325,
+      "learning_rate": 3.2479636261616156e-06,
+      "loss": 0.1296,
+      "step": 7799
+    },
+    {
+      "epoch": 1.2637718729747245,
+      "grad_norm": 0.8504661321640015,
+      "learning_rate": 3.2475463413424983e-06,
+      "loss": 0.1071,
+      "step": 7800
+    },
+    {
+      "epoch": 1.2639338950097212,
+      "grad_norm": 0.8363580703735352,
+      "learning_rate": 3.247129033650273e-06,
+      "loss": 0.0971,
+      "step": 7801
+    },
+    {
+      "epoch": 1.2640959170447181,
+      "grad_norm": 0.7588921785354614,
+      "learning_rate": 3.246711703097707e-06,
+      "loss": 0.0923,
+      "step": 7802
+    },
+    {
+      "epoch": 1.2642579390797148,
+      "grad_norm": 0.7907635569572449,
+      "learning_rate": 3.2462943496975696e-06,
+      "loss": 0.1052,
+      "step": 7803
+    },
+    {
+      "epoch": 1.2644199611147116,
+      "grad_norm": 0.8892641067504883,
+      "learning_rate": 3.2458769734626315e-06,
+      "loss": 0.1188,
+      "step": 7804
+    },
+    {
+      "epoch": 1.2645819831497085,
+      "grad_norm": 0.7947612404823303,
+      "learning_rate": 3.245459574405662e-06,
+      "loss": 0.1109,
+      "step": 7805
+    },
+    {
+      "epoch": 1.2647440051847052,
+      "grad_norm": 0.8161318898200989,
+      "learning_rate": 3.245042152539435e-06,
+      "loss": 0.1071,
+      "step": 7806
+    },
+    {
+      "epoch": 1.264906027219702,
+      "grad_norm": 0.8670127987861633,
+      "learning_rate": 3.2446247078767195e-06,
+      "loss": 0.1086,
+      "step": 7807
+    },
+    {
+      "epoch": 1.2650680492546986,
+      "grad_norm": 0.8083024621009827,
+      "learning_rate": 3.2442072404302917e-06,
+      "loss": 0.1044,
+      "step": 7808
+    },
+    {
+      "epoch": 1.2652300712896953,
+      "grad_norm": 0.7659672498703003,
+      "learning_rate": 3.243789750212922e-06,
+      "loss": 0.097,
+      "step": 7809
+    },
+    {
+      "epoch": 1.2653920933246923,
+      "grad_norm": 0.7633987069129944,
+      "learning_rate": 3.243372237237386e-06,
+      "loss": 0.0838,
+      "step": 7810
+    },
+    {
+      "epoch": 1.265554115359689,
+      "grad_norm": 0.7828187942504883,
+      "learning_rate": 3.2429547015164585e-06,
+      "loss": 0.0969,
+      "step": 7811
+    },
+    {
+      "epoch": 1.2657161373946857,
+      "grad_norm": 0.9652653932571411,
+      "learning_rate": 3.2425371430629155e-06,
+      "loss": 0.1092,
+      "step": 7812
+    },
+    {
+      "epoch": 1.2658781594296824,
+      "grad_norm": 0.807366669178009,
+      "learning_rate": 3.242119561889533e-06,
+      "loss": 0.0976,
+      "step": 7813
+    },
+    {
+      "epoch": 1.2660401814646791,
+      "grad_norm": 0.8526018857955933,
+      "learning_rate": 3.241701958009087e-06,
+      "loss": 0.1037,
+      "step": 7814
+    },
+    {
+      "epoch": 1.266202203499676,
+      "grad_norm": 0.9918192625045776,
+      "learning_rate": 3.2412843314343566e-06,
+      "loss": 0.1089,
+      "step": 7815
+    },
+    {
+      "epoch": 1.2663642255346728,
+      "grad_norm": 0.7938478589057922,
+      "learning_rate": 3.2408666821781186e-06,
+      "loss": 0.1041,
+      "step": 7816
+    },
+    {
+      "epoch": 1.2665262475696695,
+      "grad_norm": 0.8537562489509583,
+      "learning_rate": 3.2404490102531536e-06,
+      "loss": 0.1142,
+      "step": 7817
+    },
+    {
+      "epoch": 1.2666882696046662,
+      "grad_norm": 0.858700692653656,
+      "learning_rate": 3.2400313156722414e-06,
+      "loss": 0.1092,
+      "step": 7818
+    },
+    {
+      "epoch": 1.266850291639663,
+      "grad_norm": 0.8112673163414001,
+      "learning_rate": 3.2396135984481607e-06,
+      "loss": 0.097,
+      "step": 7819
+    },
+    {
+      "epoch": 1.2670123136746598,
+      "grad_norm": 0.7495766282081604,
+      "learning_rate": 3.2391958585936946e-06,
+      "loss": 0.1006,
+      "step": 7820
+    },
+    {
+      "epoch": 1.2671743357096565,
+      "grad_norm": 0.8106545805931091,
+      "learning_rate": 3.2387780961216237e-06,
+      "loss": 0.1012,
+      "step": 7821
+    },
+    {
+      "epoch": 1.2673363577446533,
+      "grad_norm": 0.7898968458175659,
+      "learning_rate": 3.2383603110447304e-06,
+      "loss": 0.0889,
+      "step": 7822
+    },
+    {
+      "epoch": 1.26749837977965,
+      "grad_norm": 0.8760552406311035,
+      "learning_rate": 3.237942503375799e-06,
+      "loss": 0.1115,
+      "step": 7823
+    },
+    {
+      "epoch": 1.2676604018146467,
+      "grad_norm": 0.7963511943817139,
+      "learning_rate": 3.2375246731276122e-06,
+      "loss": 0.1045,
+      "step": 7824
+    },
+    {
+      "epoch": 1.2678224238496436,
+      "grad_norm": 0.8547530770301819,
+      "learning_rate": 3.237106820312956e-06,
+      "loss": 0.1028,
+      "step": 7825
+    },
+    {
+      "epoch": 1.2679844458846403,
+      "grad_norm": 0.749346137046814,
+      "learning_rate": 3.236688944944614e-06,
+      "loss": 0.0918,
+      "step": 7826
+    },
+    {
+      "epoch": 1.268146467919637,
+      "grad_norm": 0.6958228945732117,
+      "learning_rate": 3.2362710470353737e-06,
+      "loss": 0.087,
+      "step": 7827
+    },
+    {
+      "epoch": 1.268308489954634,
+      "grad_norm": 0.9344890713691711,
+      "learning_rate": 3.2358531265980207e-06,
+      "loss": 0.1125,
+      "step": 7828
+    },
+    {
+      "epoch": 1.2684705119896305,
+      "grad_norm": 0.8283147215843201,
+      "learning_rate": 3.2354351836453423e-06,
+      "loss": 0.097,
+      "step": 7829
+    },
+    {
+      "epoch": 1.2686325340246274,
+      "grad_norm": 0.9189488291740417,
+      "learning_rate": 3.2350172181901283e-06,
+      "loss": 0.1061,
+      "step": 7830
+    },
+    {
+      "epoch": 1.268794556059624,
+      "grad_norm": 0.8217201232910156,
+      "learning_rate": 3.234599230245165e-06,
+      "loss": 0.1026,
+      "step": 7831
+    },
+    {
+      "epoch": 1.2689565780946208,
+      "grad_norm": 0.9388702511787415,
+      "learning_rate": 3.2341812198232437e-06,
+      "loss": 0.1086,
+      "step": 7832
+    },
+    {
+      "epoch": 1.2691186001296177,
+      "grad_norm": 0.7720819115638733,
+      "learning_rate": 3.2337631869371534e-06,
+      "loss": 0.1009,
+      "step": 7833
+    },
+    {
+      "epoch": 1.2692806221646145,
+      "grad_norm": 0.9018075466156006,
+      "learning_rate": 3.2333451315996857e-06,
+      "loss": 0.1147,
+      "step": 7834
+    },
+    {
+      "epoch": 1.2694426441996112,
+      "grad_norm": 0.7737125754356384,
+      "learning_rate": 3.2329270538236313e-06,
+      "loss": 0.098,
+      "step": 7835
+    },
+    {
+      "epoch": 1.2696046662346079,
+      "grad_norm": 0.9752976298332214,
+      "learning_rate": 3.232508953621782e-06,
+      "loss": 0.1188,
+      "step": 7836
+    },
+    {
+      "epoch": 1.2697666882696046,
+      "grad_norm": 0.8050476908683777,
+      "learning_rate": 3.232090831006932e-06,
+      "loss": 0.1102,
+      "step": 7837
+    },
+    {
+      "epoch": 1.2699287103046015,
+      "grad_norm": 0.8317782878875732,
+      "learning_rate": 3.231672685991874e-06,
+      "loss": 0.0991,
+      "step": 7838
+    },
+    {
+      "epoch": 1.2700907323395982,
+      "grad_norm": 0.8194529414176941,
+      "learning_rate": 3.231254518589403e-06,
+      "loss": 0.1068,
+      "step": 7839
+    },
+    {
+      "epoch": 1.270252754374595,
+      "grad_norm": 0.8640487194061279,
+      "learning_rate": 3.2308363288123128e-06,
+      "loss": 0.1159,
+      "step": 7840
+    },
+    {
+      "epoch": 1.2704147764095917,
+      "grad_norm": 0.7963087558746338,
+      "learning_rate": 3.2304181166733993e-06,
+      "loss": 0.1065,
+      "step": 7841
+    },
+    {
+      "epoch": 1.2705767984445884,
+      "grad_norm": 1.0018774271011353,
+      "learning_rate": 3.2299998821854593e-06,
+      "loss": 0.1209,
+      "step": 7842
+    },
+    {
+      "epoch": 1.2707388204795853,
+      "grad_norm": 0.7300251722335815,
+      "learning_rate": 3.2295816253612897e-06,
+      "loss": 0.0895,
+      "step": 7843
+    },
+    {
+      "epoch": 1.270900842514582,
+      "grad_norm": 0.7629234790802002,
+      "learning_rate": 3.229163346213688e-06,
+      "loss": 0.1015,
+      "step": 7844
+    },
+    {
+      "epoch": 1.2710628645495787,
+      "grad_norm": 0.7023546099662781,
+      "learning_rate": 3.2287450447554526e-06,
+      "loss": 0.0956,
+      "step": 7845
+    },
+    {
+      "epoch": 1.2712248865845754,
+      "grad_norm": 0.8268736600875854,
+      "learning_rate": 3.228326720999382e-06,
+      "loss": 0.1019,
+      "step": 7846
+    },
+    {
+      "epoch": 1.2713869086195722,
+      "grad_norm": 0.7980479001998901,
+      "learning_rate": 3.227908374958276e-06,
+      "loss": 0.1008,
+      "step": 7847
+    },
+    {
+      "epoch": 1.271548930654569,
+      "grad_norm": 0.7699779868125916,
+      "learning_rate": 3.2274900066449355e-06,
+      "loss": 0.096,
+      "step": 7848
+    },
+    {
+      "epoch": 1.2717109526895658,
+      "grad_norm": 0.9146498441696167,
+      "learning_rate": 3.2270716160721612e-06,
+      "loss": 0.1143,
+      "step": 7849
+    },
+    {
+      "epoch": 1.2718729747245625,
+      "grad_norm": 0.8463569283485413,
+      "learning_rate": 3.2266532032527548e-06,
+      "loss": 0.105,
+      "step": 7850
+    },
+    {
+      "epoch": 1.2720349967595592,
+      "grad_norm": 0.8177133202552795,
+      "learning_rate": 3.2262347681995187e-06,
+      "loss": 0.1015,
+      "step": 7851
+    },
+    {
+      "epoch": 1.272197018794556,
+      "grad_norm": 0.8679166436195374,
+      "learning_rate": 3.225816310925257e-06,
+      "loss": 0.0998,
+      "step": 7852
+    },
+    {
+      "epoch": 1.2723590408295529,
+      "grad_norm": 0.7825645804405212,
+      "learning_rate": 3.2253978314427716e-06,
+      "loss": 0.1001,
+      "step": 7853
+    },
+    {
+      "epoch": 1.2725210628645496,
+      "grad_norm": 0.8941231966018677,
+      "learning_rate": 3.224979329764869e-06,
+      "loss": 0.0982,
+      "step": 7854
+    },
+    {
+      "epoch": 1.2726830848995463,
+      "grad_norm": 0.8882502913475037,
+      "learning_rate": 3.2245608059043525e-06,
+      "loss": 0.1052,
+      "step": 7855
+    },
+    {
+      "epoch": 1.2728451069345432,
+      "grad_norm": 0.7035767436027527,
+      "learning_rate": 3.224142259874029e-06,
+      "loss": 0.0808,
+      "step": 7856
+    },
+    {
+      "epoch": 1.27300712896954,
+      "grad_norm": 0.837063193321228,
+      "learning_rate": 3.2237236916867047e-06,
+      "loss": 0.1036,
+      "step": 7857
+    },
+    {
+      "epoch": 1.2731691510045366,
+      "grad_norm": 0.8827104568481445,
+      "learning_rate": 3.223305101355187e-06,
+      "loss": 0.1098,
+      "step": 7858
+    },
+    {
+      "epoch": 1.2733311730395334,
+      "grad_norm": 0.8305638432502747,
+      "learning_rate": 3.2228864888922838e-06,
+      "loss": 0.1051,
+      "step": 7859
+    },
+    {
+      "epoch": 1.27349319507453,
+      "grad_norm": 0.9665209054946899,
+      "learning_rate": 3.2224678543108024e-06,
+      "loss": 0.1178,
+      "step": 7860
+    },
+    {
+      "epoch": 1.273655217109527,
+      "grad_norm": 0.8135932683944702,
+      "learning_rate": 3.222049197623554e-06,
+      "loss": 0.0927,
+      "step": 7861
+    },
+    {
+      "epoch": 1.2738172391445237,
+      "grad_norm": 0.8016318082809448,
+      "learning_rate": 3.221630518843347e-06,
+      "loss": 0.1005,
+      "step": 7862
+    },
+    {
+      "epoch": 1.2739792611795204,
+      "grad_norm": 0.8300334811210632,
+      "learning_rate": 3.2212118179829925e-06,
+      "loss": 0.1057,
+      "step": 7863
+    },
+    {
+      "epoch": 1.2741412832145171,
+      "grad_norm": 0.7906784415245056,
+      "learning_rate": 3.2207930950553017e-06,
+      "loss": 0.1001,
+      "step": 7864
+    },
+    {
+      "epoch": 1.2743033052495139,
+      "grad_norm": 0.8609299063682556,
+      "learning_rate": 3.2203743500730867e-06,
+      "loss": 0.1019,
+      "step": 7865
+    },
+    {
+      "epoch": 1.2744653272845108,
+      "grad_norm": 0.783918023109436,
+      "learning_rate": 3.2199555830491597e-06,
+      "loss": 0.0979,
+      "step": 7866
+    },
+    {
+      "epoch": 1.2746273493195075,
+      "grad_norm": 0.9125205278396606,
+      "learning_rate": 3.219536793996334e-06,
+      "loss": 0.1175,
+      "step": 7867
+    },
+    {
+      "epoch": 1.2747893713545042,
+      "grad_norm": 0.8157711029052734,
+      "learning_rate": 3.2191179829274244e-06,
+      "loss": 0.0942,
+      "step": 7868
+    },
+    {
+      "epoch": 1.274951393389501,
+      "grad_norm": 0.9252278208732605,
+      "learning_rate": 3.218699149855244e-06,
+      "loss": 0.1111,
+      "step": 7869
+    },
+    {
+      "epoch": 1.2751134154244976,
+      "grad_norm": 0.8688713908195496,
+      "learning_rate": 3.2182802947926086e-06,
+      "loss": 0.1111,
+      "step": 7870
+    },
+    {
+      "epoch": 1.2752754374594946,
+      "grad_norm": 0.841914176940918,
+      "learning_rate": 3.217861417752335e-06,
+      "loss": 0.1005,
+      "step": 7871
+    },
+    {
+      "epoch": 1.2754374594944913,
+      "grad_norm": 0.8829323053359985,
+      "learning_rate": 3.2174425187472387e-06,
+      "loss": 0.1121,
+      "step": 7872
+    },
+    {
+      "epoch": 1.275599481529488,
+      "grad_norm": 0.8816058039665222,
+      "learning_rate": 3.2170235977901375e-06,
+      "loss": 0.1215,
+      "step": 7873
+    },
+    {
+      "epoch": 1.2757615035644847,
+      "grad_norm": 0.7793697118759155,
+      "learning_rate": 3.2166046548938497e-06,
+      "loss": 0.1007,
+      "step": 7874
+    },
+    {
+      "epoch": 1.2759235255994814,
+      "grad_norm": 0.7956479787826538,
+      "learning_rate": 3.216185690071193e-06,
+      "loss": 0.1034,
+      "step": 7875
+    },
+    {
+      "epoch": 1.2760855476344783,
+      "grad_norm": 0.80839604139328,
+      "learning_rate": 3.215766703334988e-06,
+      "loss": 0.1001,
+      "step": 7876
+    },
+    {
+      "epoch": 1.276247569669475,
+      "grad_norm": 0.8526255488395691,
+      "learning_rate": 3.215347694698054e-06,
+      "loss": 0.1031,
+      "step": 7877
+    },
+    {
+      "epoch": 1.2764095917044718,
+      "grad_norm": 0.8941551446914673,
+      "learning_rate": 3.214928664173211e-06,
+      "loss": 0.1132,
+      "step": 7878
+    },
+    {
+      "epoch": 1.2765716137394687,
+      "grad_norm": 0.7045158743858337,
+      "learning_rate": 3.2145096117732823e-06,
+      "loss": 0.0901,
+      "step": 7879
+    },
+    {
+      "epoch": 1.2767336357744652,
+      "grad_norm": 0.9706689715385437,
+      "learning_rate": 3.2140905375110875e-06,
+      "loss": 0.1244,
+      "step": 7880
+    },
+    {
+      "epoch": 1.2768956578094621,
+      "grad_norm": 0.9292829036712646,
+      "learning_rate": 3.21367144139945e-06,
+      "loss": 0.1131,
+      "step": 7881
+    },
+    {
+      "epoch": 1.2770576798444588,
+      "grad_norm": 0.8300909996032715,
+      "learning_rate": 3.2132523234511943e-06,
+      "loss": 0.1164,
+      "step": 7882
+    },
+    {
+      "epoch": 1.2772197018794555,
+      "grad_norm": 0.7891339659690857,
+      "learning_rate": 3.2128331836791436e-06,
+      "loss": 0.1007,
+      "step": 7883
+    },
+    {
+      "epoch": 1.2773817239144525,
+      "grad_norm": 0.6846839189529419,
+      "learning_rate": 3.2124140220961215e-06,
+      "loss": 0.0874,
+      "step": 7884
+    },
+    {
+      "epoch": 1.2775437459494492,
+      "grad_norm": 0.8767536282539368,
+      "learning_rate": 3.211994838714955e-06,
+      "loss": 0.1121,
+      "step": 7885
+    },
+    {
+      "epoch": 1.277705767984446,
+      "grad_norm": 0.891893744468689,
+      "learning_rate": 3.2115756335484694e-06,
+      "loss": 0.113,
+      "step": 7886
+    },
+    {
+      "epoch": 1.2778677900194426,
+      "grad_norm": 0.8657979965209961,
+      "learning_rate": 3.2111564066094913e-06,
+      "loss": 0.1176,
+      "step": 7887
+    },
+    {
+      "epoch": 1.2780298120544393,
+      "grad_norm": 0.8780067563056946,
+      "learning_rate": 3.210737157910848e-06,
+      "loss": 0.111,
+      "step": 7888
+    },
+    {
+      "epoch": 1.2781918340894363,
+      "grad_norm": 0.9224578738212585,
+      "learning_rate": 3.2103178874653677e-06,
+      "loss": 0.1151,
+      "step": 7889
+    },
+    {
+      "epoch": 1.278353856124433,
+      "grad_norm": 0.7573640942573547,
+      "learning_rate": 3.2098985952858796e-06,
+      "loss": 0.1087,
+      "step": 7890
+    },
+    {
+      "epoch": 1.2785158781594297,
+      "grad_norm": 0.8327637910842896,
+      "learning_rate": 3.2094792813852116e-06,
+      "loss": 0.1107,
+      "step": 7891
+    },
+    {
+      "epoch": 1.2786779001944264,
+      "grad_norm": 0.9042618870735168,
+      "learning_rate": 3.209059945776195e-06,
+      "loss": 0.1194,
+      "step": 7892
+    },
+    {
+      "epoch": 1.278839922229423,
+      "grad_norm": 0.9340540766716003,
+      "learning_rate": 3.2086405884716592e-06,
+      "loss": 0.1316,
+      "step": 7893
+    },
+    {
+      "epoch": 1.27900194426442,
+      "grad_norm": 0.7579165697097778,
+      "learning_rate": 3.2082212094844374e-06,
+      "loss": 0.099,
+      "step": 7894
+    },
+    {
+      "epoch": 1.2791639662994168,
+      "grad_norm": 0.7476134896278381,
+      "learning_rate": 3.20780180882736e-06,
+      "loss": 0.1061,
+      "step": 7895
+    },
+    {
+      "epoch": 1.2793259883344135,
+      "grad_norm": 0.8922802209854126,
+      "learning_rate": 3.20738238651326e-06,
+      "loss": 0.1033,
+      "step": 7896
+    },
+    {
+      "epoch": 1.2794880103694102,
+      "grad_norm": 0.8457995057106018,
+      "learning_rate": 3.2069629425549705e-06,
+      "loss": 0.1123,
+      "step": 7897
+    },
+    {
+      "epoch": 1.279650032404407,
+      "grad_norm": 0.7586995959281921,
+      "learning_rate": 3.206543476965326e-06,
+      "loss": 0.0922,
+      "step": 7898
+    },
+    {
+      "epoch": 1.2798120544394038,
+      "grad_norm": 0.9203418493270874,
+      "learning_rate": 3.2061239897571613e-06,
+      "loss": 0.1247,
+      "step": 7899
+    },
+    {
+      "epoch": 1.2799740764744005,
+      "grad_norm": 0.8178397417068481,
+      "learning_rate": 3.2057044809433108e-06,
+      "loss": 0.1039,
+      "step": 7900
+    },
+    {
+      "epoch": 1.2801360985093972,
+      "grad_norm": 0.794354259967804,
+      "learning_rate": 3.2052849505366113e-06,
+      "loss": 0.1085,
+      "step": 7901
+    },
+    {
+      "epoch": 1.280298120544394,
+      "grad_norm": 0.7944015264511108,
+      "learning_rate": 3.2048653985498985e-06,
+      "loss": 0.1099,
+      "step": 7902
+    },
+    {
+      "epoch": 1.2804601425793907,
+      "grad_norm": 0.8562796115875244,
+      "learning_rate": 3.2044458249960108e-06,
+      "loss": 0.1018,
+      "step": 7903
+    },
+    {
+      "epoch": 1.2806221646143876,
+      "grad_norm": 0.9842815399169922,
+      "learning_rate": 3.204026229887785e-06,
+      "loss": 0.12,
+      "step": 7904
+    },
+    {
+      "epoch": 1.2807841866493843,
+      "grad_norm": 0.7473559379577637,
+      "learning_rate": 3.2036066132380606e-06,
+      "loss": 0.0911,
+      "step": 7905
+    },
+    {
+      "epoch": 1.280946208684381,
+      "grad_norm": 0.7652568221092224,
+      "learning_rate": 3.203186975059677e-06,
+      "loss": 0.1059,
+      "step": 7906
+    },
+    {
+      "epoch": 1.281108230719378,
+      "grad_norm": 0.8993542790412903,
+      "learning_rate": 3.2027673153654733e-06,
+      "loss": 0.112,
+      "step": 7907
+    },
+    {
+      "epoch": 1.2812702527543747,
+      "grad_norm": 0.8552115559577942,
+      "learning_rate": 3.2023476341682902e-06,
+      "loss": 0.1073,
+      "step": 7908
+    },
+    {
+      "epoch": 1.2814322747893714,
+      "grad_norm": 0.7751871347427368,
+      "learning_rate": 3.2019279314809694e-06,
+      "loss": 0.0883,
+      "step": 7909
+    },
+    {
+      "epoch": 1.281594296824368,
+      "grad_norm": 0.891760528087616,
+      "learning_rate": 3.2015082073163524e-06,
+      "loss": 0.114,
+      "step": 7910
+    },
+    {
+      "epoch": 1.2817563188593648,
+      "grad_norm": 0.8688755035400391,
+      "learning_rate": 3.201088461687282e-06,
+      "loss": 0.1083,
+      "step": 7911
+    },
+    {
+      "epoch": 1.2819183408943617,
+      "grad_norm": 0.8324447274208069,
+      "learning_rate": 3.2006686946066012e-06,
+      "loss": 0.107,
+      "step": 7912
+    },
+    {
+      "epoch": 1.2820803629293585,
+      "grad_norm": 0.779424250125885,
+      "learning_rate": 3.2002489060871534e-06,
+      "loss": 0.0916,
+      "step": 7913
+    },
+    {
+      "epoch": 1.2822423849643552,
+      "grad_norm": 0.7420961260795593,
+      "learning_rate": 3.1998290961417844e-06,
+      "loss": 0.0904,
+      "step": 7914
+    },
+    {
+      "epoch": 1.2824044069993519,
+      "grad_norm": 0.8292026519775391,
+      "learning_rate": 3.199409264783338e-06,
+      "loss": 0.0986,
+      "step": 7915
+    },
+    {
+      "epoch": 1.2825664290343486,
+      "grad_norm": 0.8516577482223511,
+      "learning_rate": 3.1989894120246613e-06,
+      "loss": 0.103,
+      "step": 7916
+    },
+    {
+      "epoch": 1.2827284510693455,
+      "grad_norm": 0.7372321486473083,
+      "learning_rate": 3.1985695378786e-06,
+      "loss": 0.0895,
+      "step": 7917
+    },
+    {
+      "epoch": 1.2828904731043422,
+      "grad_norm": 0.8905530571937561,
+      "learning_rate": 3.1981496423580012e-06,
+      "loss": 0.1178,
+      "step": 7918
+    },
+    {
+      "epoch": 1.283052495139339,
+      "grad_norm": 0.819083571434021,
+      "learning_rate": 3.1977297254757124e-06,
+      "loss": 0.1058,
+      "step": 7919
+    },
+    {
+      "epoch": 1.2832145171743357,
+      "grad_norm": 0.8953108787536621,
+      "learning_rate": 3.1973097872445828e-06,
+      "loss": 0.1166,
+      "step": 7920
+    },
+    {
+      "epoch": 1.2833765392093324,
+      "grad_norm": 0.7964316606521606,
+      "learning_rate": 3.196889827677462e-06,
+      "loss": 0.0898,
+      "step": 7921
+    },
+    {
+      "epoch": 1.2835385612443293,
+      "grad_norm": 0.8181843161582947,
+      "learning_rate": 3.1964698467871976e-06,
+      "loss": 0.0913,
+      "step": 7922
+    },
+    {
+      "epoch": 1.283700583279326,
+      "grad_norm": 0.8924493789672852,
+      "learning_rate": 3.1960498445866423e-06,
+      "loss": 0.1187,
+      "step": 7923
+    },
+    {
+      "epoch": 1.2838626053143227,
+      "grad_norm": 0.9556626677513123,
+      "learning_rate": 3.1956298210886454e-06,
+      "loss": 0.1147,
+      "step": 7924
+    },
+    {
+      "epoch": 1.2840246273493194,
+      "grad_norm": 0.8342770934104919,
+      "learning_rate": 3.1952097763060595e-06,
+      "loss": 0.1027,
+      "step": 7925
+    },
+    {
+      "epoch": 1.2841866493843161,
+      "grad_norm": 0.7476953864097595,
+      "learning_rate": 3.1947897102517374e-06,
+      "loss": 0.0885,
+      "step": 7926
+    },
+    {
+      "epoch": 1.284348671419313,
+      "grad_norm": 0.8038281798362732,
+      "learning_rate": 3.1943696229385307e-06,
+      "loss": 0.1014,
+      "step": 7927
+    },
+    {
+      "epoch": 1.2845106934543098,
+      "grad_norm": 0.9089291095733643,
+      "learning_rate": 3.1939495143792944e-06,
+      "loss": 0.1119,
+      "step": 7928
+    },
+    {
+      "epoch": 1.2846727154893065,
+      "grad_norm": 0.9920743107795715,
+      "learning_rate": 3.193529384586882e-06,
+      "loss": 0.1191,
+      "step": 7929
+    },
+    {
+      "epoch": 1.2848347375243034,
+      "grad_norm": 0.7742721438407898,
+      "learning_rate": 3.1931092335741497e-06,
+      "loss": 0.0964,
+      "step": 7930
+    },
+    {
+      "epoch": 1.2849967595593,
+      "grad_norm": 0.8795700073242188,
+      "learning_rate": 3.1926890613539513e-06,
+      "loss": 0.1061,
+      "step": 7931
+    },
+    {
+      "epoch": 1.2851587815942969,
+      "grad_norm": 0.8437364101409912,
+      "learning_rate": 3.192268867939144e-06,
+      "loss": 0.1071,
+      "step": 7932
+    },
+    {
+      "epoch": 1.2853208036292936,
+      "grad_norm": 0.8689562678337097,
+      "learning_rate": 3.191848653342584e-06,
+      "loss": 0.1101,
+      "step": 7933
+    },
+    {
+      "epoch": 1.2854828256642903,
+      "grad_norm": 0.8216108679771423,
+      "learning_rate": 3.1914284175771303e-06,
+      "loss": 0.1049,
+      "step": 7934
+    },
+    {
+      "epoch": 1.2856448476992872,
+      "grad_norm": 0.8345628380775452,
+      "learning_rate": 3.19100816065564e-06,
+      "loss": 0.0987,
+      "step": 7935
+    },
+    {
+      "epoch": 1.285806869734284,
+      "grad_norm": 0.830443799495697,
+      "learning_rate": 3.1905878825909726e-06,
+      "loss": 0.107,
+      "step": 7936
+    },
+    {
+      "epoch": 1.2859688917692806,
+      "grad_norm": 0.7912009358406067,
+      "learning_rate": 3.190167583395986e-06,
+      "loss": 0.1001,
+      "step": 7937
+    },
+    {
+      "epoch": 1.2861309138042774,
+      "grad_norm": 0.8424001932144165,
+      "learning_rate": 3.189747263083543e-06,
+      "loss": 0.1061,
+      "step": 7938
+    },
+    {
+      "epoch": 1.286292935839274,
+      "grad_norm": 0.8694502115249634,
+      "learning_rate": 3.1893269216665017e-06,
+      "loss": 0.1152,
+      "step": 7939
+    },
+    {
+      "epoch": 1.286454957874271,
+      "grad_norm": 0.8436936736106873,
+      "learning_rate": 3.188906559157725e-06,
+      "loss": 0.1028,
+      "step": 7940
+    },
+    {
+      "epoch": 1.2866169799092677,
+      "grad_norm": 0.7603984475135803,
+      "learning_rate": 3.188486175570075e-06,
+      "loss": 0.0907,
+      "step": 7941
+    },
+    {
+      "epoch": 1.2867790019442644,
+      "grad_norm": 0.8339808583259583,
+      "learning_rate": 3.1880657709164144e-06,
+      "loss": 0.107,
+      "step": 7942
+    },
+    {
+      "epoch": 1.2869410239792611,
+      "grad_norm": 0.8520210385322571,
+      "learning_rate": 3.187645345209606e-06,
+      "loss": 0.1049,
+      "step": 7943
+    },
+    {
+      "epoch": 1.2871030460142578,
+      "grad_norm": 0.8916787505149841,
+      "learning_rate": 3.1872248984625135e-06,
+      "loss": 0.108,
+      "step": 7944
+    },
+    {
+      "epoch": 1.2872650680492548,
+      "grad_norm": 0.9074912667274475,
+      "learning_rate": 3.1868044306880037e-06,
+      "loss": 0.1019,
+      "step": 7945
+    },
+    {
+      "epoch": 1.2874270900842515,
+      "grad_norm": 0.9418996572494507,
+      "learning_rate": 3.1863839418989385e-06,
+      "loss": 0.1114,
+      "step": 7946
+    },
+    {
+      "epoch": 1.2875891121192482,
+      "grad_norm": 0.7757999897003174,
+      "learning_rate": 3.185963432108187e-06,
+      "loss": 0.0916,
+      "step": 7947
+    },
+    {
+      "epoch": 1.287751134154245,
+      "grad_norm": 0.8359844088554382,
+      "learning_rate": 3.185542901328613e-06,
+      "loss": 0.1024,
+      "step": 7948
+    },
+    {
+      "epoch": 1.2879131561892416,
+      "grad_norm": 1.3665879964828491,
+      "learning_rate": 3.185122349573087e-06,
+      "loss": 0.1411,
+      "step": 7949
+    },
+    {
+      "epoch": 1.2880751782242386,
+      "grad_norm": 0.7807172536849976,
+      "learning_rate": 3.184701776854474e-06,
+      "loss": 0.104,
+      "step": 7950
+    },
+    {
+      "epoch": 1.2882372002592353,
+      "grad_norm": 0.7108926177024841,
+      "learning_rate": 3.1842811831856444e-06,
+      "loss": 0.0831,
+      "step": 7951
+    },
+    {
+      "epoch": 1.288399222294232,
+      "grad_norm": 0.7794750332832336,
+      "learning_rate": 3.1838605685794665e-06,
+      "loss": 0.1034,
+      "step": 7952
+    },
+    {
+      "epoch": 1.2885612443292287,
+      "grad_norm": 0.8731608986854553,
+      "learning_rate": 3.183439933048809e-06,
+      "loss": 0.0996,
+      "step": 7953
+    },
+    {
+      "epoch": 1.2887232663642254,
+      "grad_norm": 0.8640766143798828,
+      "learning_rate": 3.1830192766065445e-06,
+      "loss": 0.1089,
+      "step": 7954
+    },
+    {
+      "epoch": 1.2888852883992223,
+      "grad_norm": 0.8059561848640442,
+      "learning_rate": 3.1825985992655422e-06,
+      "loss": 0.0989,
+      "step": 7955
+    },
+    {
+      "epoch": 1.289047310434219,
+      "grad_norm": 0.889413595199585,
+      "learning_rate": 3.1821779010386755e-06,
+      "loss": 0.1129,
+      "step": 7956
+    },
+    {
+      "epoch": 1.2892093324692158,
+      "grad_norm": 0.7468616366386414,
+      "learning_rate": 3.181757181938815e-06,
+      "loss": 0.0968,
+      "step": 7957
+    },
+    {
+      "epoch": 1.2893713545042127,
+      "grad_norm": 0.7996143102645874,
+      "learning_rate": 3.181336441978835e-06,
+      "loss": 0.0968,
+      "step": 7958
+    },
+    {
+      "epoch": 1.2895333765392094,
+      "grad_norm": 0.8068464398384094,
+      "learning_rate": 3.1809156811716084e-06,
+      "loss": 0.0977,
+      "step": 7959
+    },
+    {
+      "epoch": 1.2896953985742061,
+      "grad_norm": 0.7814459800720215,
+      "learning_rate": 3.18049489953001e-06,
+      "loss": 0.1007,
+      "step": 7960
+    },
+    {
+      "epoch": 1.2898574206092028,
+      "grad_norm": 0.7772566080093384,
+      "learning_rate": 3.180074097066914e-06,
+      "loss": 0.0983,
+      "step": 7961
+    },
+    {
+      "epoch": 1.2900194426441995,
+      "grad_norm": 0.7291167974472046,
+      "learning_rate": 3.1796532737951975e-06,
+      "loss": 0.0959,
+      "step": 7962
+    },
+    {
+      "epoch": 1.2901814646791965,
+      "grad_norm": 0.7671595215797424,
+      "learning_rate": 3.1792324297277345e-06,
+      "loss": 0.0991,
+      "step": 7963
+    },
+    {
+      "epoch": 1.2903434867141932,
+      "grad_norm": 0.7343456745147705,
+      "learning_rate": 3.1788115648774033e-06,
+      "loss": 0.0958,
+      "step": 7964
+    },
+    {
+      "epoch": 1.29050550874919,
+      "grad_norm": 0.8339682817459106,
+      "learning_rate": 3.1783906792570805e-06,
+      "loss": 0.1065,
+      "step": 7965
+    },
+    {
+      "epoch": 1.2906675307841866,
+      "grad_norm": 0.9149320125579834,
+      "learning_rate": 3.177969772879645e-06,
+      "loss": 0.1138,
+      "step": 7966
+    },
+    {
+      "epoch": 1.2908295528191833,
+      "grad_norm": 0.8161713480949402,
+      "learning_rate": 3.1775488457579756e-06,
+      "loss": 0.1041,
+      "step": 7967
+    },
+    {
+      "epoch": 1.2909915748541803,
+      "grad_norm": 0.8028397560119629,
+      "learning_rate": 3.1771278979049496e-06,
+      "loss": 0.0971,
+      "step": 7968
+    },
+    {
+      "epoch": 1.291153596889177,
+      "grad_norm": 0.7893111109733582,
+      "learning_rate": 3.1767069293334502e-06,
+      "loss": 0.0984,
+      "step": 7969
+    },
+    {
+      "epoch": 1.2913156189241737,
+      "grad_norm": 0.9668081402778625,
+      "learning_rate": 3.176285940056355e-06,
+      "loss": 0.1123,
+      "step": 7970
+    },
+    {
+      "epoch": 1.2914776409591704,
+      "grad_norm": 0.896835207939148,
+      "learning_rate": 3.1758649300865473e-06,
+      "loss": 0.1062,
+      "step": 7971
+    },
+    {
+      "epoch": 1.291639662994167,
+      "grad_norm": 0.8189420700073242,
+      "learning_rate": 3.1754438994369087e-06,
+      "loss": 0.1035,
+      "step": 7972
+    },
+    {
+      "epoch": 1.291801685029164,
+      "grad_norm": 0.8541057109832764,
+      "learning_rate": 3.1750228481203206e-06,
+      "loss": 0.107,
+      "step": 7973
+    },
+    {
+      "epoch": 1.2919637070641607,
+      "grad_norm": 0.7424408793449402,
+      "learning_rate": 3.174601776149668e-06,
+      "loss": 0.0974,
+      "step": 7974
+    },
+    {
+      "epoch": 1.2921257290991575,
+      "grad_norm": 0.9748353958129883,
+      "learning_rate": 3.174180683537832e-06,
+      "loss": 0.1157,
+      "step": 7975
+    },
+    {
+      "epoch": 1.2922877511341542,
+      "grad_norm": 0.9521284699440002,
+      "learning_rate": 3.1737595702976996e-06,
+      "loss": 0.1038,
+      "step": 7976
+    },
+    {
+      "epoch": 1.2924497731691509,
+      "grad_norm": 0.9242678284645081,
+      "learning_rate": 3.1733384364421536e-06,
+      "loss": 0.1133,
+      "step": 7977
+    },
+    {
+      "epoch": 1.2926117952041478,
+      "grad_norm": 0.8057260513305664,
+      "learning_rate": 3.1729172819840825e-06,
+      "loss": 0.0995,
+      "step": 7978
+    },
+    {
+      "epoch": 1.2927738172391445,
+      "grad_norm": 0.7817633152008057,
+      "learning_rate": 3.17249610693637e-06,
+      "loss": 0.1035,
+      "step": 7979
+    },
+    {
+      "epoch": 1.2929358392741412,
+      "grad_norm": 0.8198245763778687,
+      "learning_rate": 3.1720749113119045e-06,
+      "loss": 0.1063,
+      "step": 7980
+    },
+    {
+      "epoch": 1.2930978613091382,
+      "grad_norm": 0.7432113885879517,
+      "learning_rate": 3.1716536951235727e-06,
+      "loss": 0.0965,
+      "step": 7981
+    },
+    {
+      "epoch": 1.2932598833441347,
+      "grad_norm": 0.7126169800758362,
+      "learning_rate": 3.1712324583842637e-06,
+      "loss": 0.0963,
+      "step": 7982
+    },
+    {
+      "epoch": 1.2934219053791316,
+      "grad_norm": 0.7479450702667236,
+      "learning_rate": 3.1708112011068647e-06,
+      "loss": 0.1006,
+      "step": 7983
+    },
+    {
+      "epoch": 1.2935839274141283,
+      "grad_norm": 0.8944701552391052,
+      "learning_rate": 3.1703899233042675e-06,
+      "loss": 0.1105,
+      "step": 7984
+    },
+    {
+      "epoch": 1.293745949449125,
+      "grad_norm": 0.8011032938957214,
+      "learning_rate": 3.1699686249893614e-06,
+      "loss": 0.1094,
+      "step": 7985
+    },
+    {
+      "epoch": 1.293907971484122,
+      "grad_norm": 0.7931639552116394,
+      "learning_rate": 3.1695473061750353e-06,
+      "loss": 0.1034,
+      "step": 7986
+    },
+    {
+      "epoch": 1.2940699935191187,
+      "grad_norm": 0.8425852060317993,
+      "learning_rate": 3.1691259668741823e-06,
+      "loss": 0.0978,
+      "step": 7987
+    },
+    {
+      "epoch": 1.2942320155541154,
+      "grad_norm": 0.7550665736198425,
+      "learning_rate": 3.1687046070996942e-06,
+      "loss": 0.0942,
+      "step": 7988
+    },
+    {
+      "epoch": 1.294394037589112,
+      "grad_norm": 0.8623692393302917,
+      "learning_rate": 3.168283226864463e-06,
+      "loss": 0.111,
+      "step": 7989
+    },
+    {
+      "epoch": 1.2945560596241088,
+      "grad_norm": 0.8119961619377136,
+      "learning_rate": 3.1678618261813828e-06,
+      "loss": 0.1057,
+      "step": 7990
+    },
+    {
+      "epoch": 1.2947180816591057,
+      "grad_norm": 0.9069503545761108,
+      "learning_rate": 3.1674404050633465e-06,
+      "loss": 0.1106,
+      "step": 7991
+    },
+    {
+      "epoch": 1.2948801036941024,
+      "grad_norm": 0.716511070728302,
+      "learning_rate": 3.167018963523249e-06,
+      "loss": 0.0845,
+      "step": 7992
+    },
+    {
+      "epoch": 1.2950421257290992,
+      "grad_norm": 0.8659084439277649,
+      "learning_rate": 3.166597501573986e-06,
+      "loss": 0.1073,
+      "step": 7993
+    },
+    {
+      "epoch": 1.2952041477640959,
+      "grad_norm": 0.9604227542877197,
+      "learning_rate": 3.1661760192284518e-06,
+      "loss": 0.1231,
+      "step": 7994
+    },
+    {
+      "epoch": 1.2953661697990926,
+      "grad_norm": 0.869018018245697,
+      "learning_rate": 3.165754516499544e-06,
+      "loss": 0.0979,
+      "step": 7995
+    },
+    {
+      "epoch": 1.2955281918340895,
+      "grad_norm": 0.8611443638801575,
+      "learning_rate": 3.165332993400159e-06,
+      "loss": 0.1135,
+      "step": 7996
+    },
+    {
+      "epoch": 1.2956902138690862,
+      "grad_norm": 0.8204544186592102,
+      "learning_rate": 3.1649114499431944e-06,
+      "loss": 0.108,
+      "step": 7997
+    },
+    {
+      "epoch": 1.295852235904083,
+      "grad_norm": 0.9157761931419373,
+      "learning_rate": 3.1644898861415484e-06,
+      "loss": 0.1186,
+      "step": 7998
+    },
+    {
+      "epoch": 1.2960142579390797,
+      "grad_norm": 0.8155969381332397,
+      "learning_rate": 3.1640683020081196e-06,
+      "loss": 0.1094,
+      "step": 7999
+    },
+    {
+      "epoch": 1.2961762799740764,
+      "grad_norm": 0.9700890779495239,
+      "learning_rate": 3.163646697555809e-06,
+      "loss": 0.1176,
+      "step": 8000
+    },
+    {
+      "epoch": 1.2963383020090733,
+      "grad_norm": 0.9052625894546509,
+      "learning_rate": 3.163225072797514e-06,
+      "loss": 0.1113,
+      "step": 8001
+    },
+    {
+      "epoch": 1.29650032404407,
+      "grad_norm": 0.8415457010269165,
+      "learning_rate": 3.1628034277461376e-06,
+      "loss": 0.103,
+      "step": 8002
+    },
+    {
+      "epoch": 1.2966623460790667,
+      "grad_norm": 0.9100468158721924,
+      "learning_rate": 3.1623817624145804e-06,
+      "loss": 0.1048,
+      "step": 8003
+    },
+    {
+      "epoch": 1.2968243681140634,
+      "grad_norm": 0.8173115849494934,
+      "learning_rate": 3.161960076815743e-06,
+      "loss": 0.0973,
+      "step": 8004
+    },
+    {
+      "epoch": 1.2969863901490601,
+      "grad_norm": 0.8509911298751831,
+      "learning_rate": 3.1615383709625303e-06,
+      "loss": 0.1052,
+      "step": 8005
+    },
+    {
+      "epoch": 1.297148412184057,
+      "grad_norm": 0.8229318261146545,
+      "learning_rate": 3.1611166448678445e-06,
+      "loss": 0.1083,
+      "step": 8006
+    },
+    {
+      "epoch": 1.2973104342190538,
+      "grad_norm": 0.7618302702903748,
+      "learning_rate": 3.1606948985445884e-06,
+      "loss": 0.1004,
+      "step": 8007
+    },
+    {
+      "epoch": 1.2974724562540505,
+      "grad_norm": 0.7803520560264587,
+      "learning_rate": 3.1602731320056675e-06,
+      "loss": 0.0968,
+      "step": 8008
+    },
+    {
+      "epoch": 1.2976344782890474,
+      "grad_norm": 1.071578860282898,
+      "learning_rate": 3.1598513452639867e-06,
+      "loss": 0.1186,
+      "step": 8009
+    },
+    {
+      "epoch": 1.2977965003240441,
+      "grad_norm": 0.9254783987998962,
+      "learning_rate": 3.159429538332452e-06,
+      "loss": 0.1081,
+      "step": 8010
+    },
+    {
+      "epoch": 1.2979585223590409,
+      "grad_norm": 0.8896194100379944,
+      "learning_rate": 3.1590077112239685e-06,
+      "loss": 0.1153,
+      "step": 8011
+    },
+    {
+      "epoch": 1.2981205443940376,
+      "grad_norm": 0.8195780515670776,
+      "learning_rate": 3.1585858639514444e-06,
+      "loss": 0.1114,
+      "step": 8012
+    },
+    {
+      "epoch": 1.2982825664290343,
+      "grad_norm": 0.869594931602478,
+      "learning_rate": 3.158163996527786e-06,
+      "loss": 0.104,
+      "step": 8013
+    },
+    {
+      "epoch": 1.2984445884640312,
+      "grad_norm": 0.9179608225822449,
+      "learning_rate": 3.1577421089659023e-06,
+      "loss": 0.1009,
+      "step": 8014
+    },
+    {
+      "epoch": 1.298606610499028,
+      "grad_norm": 0.8011884689331055,
+      "learning_rate": 3.157320201278702e-06,
+      "loss": 0.0954,
+      "step": 8015
+    },
+    {
+      "epoch": 1.2987686325340246,
+      "grad_norm": 0.8461195826530457,
+      "learning_rate": 3.1568982734790943e-06,
+      "loss": 0.108,
+      "step": 8016
+    },
+    {
+      "epoch": 1.2989306545690213,
+      "grad_norm": 0.7561697959899902,
+      "learning_rate": 3.1564763255799886e-06,
+      "loss": 0.0965,
+      "step": 8017
+    },
+    {
+      "epoch": 1.299092676604018,
+      "grad_norm": 0.8390095233917236,
+      "learning_rate": 3.1560543575942958e-06,
+      "loss": 0.1127,
+      "step": 8018
+    },
+    {
+      "epoch": 1.299254698639015,
+      "grad_norm": 0.9161712527275085,
+      "learning_rate": 3.155632369534928e-06,
+      "loss": 0.1163,
+      "step": 8019
+    },
+    {
+      "epoch": 1.2994167206740117,
+      "grad_norm": 0.8636978268623352,
+      "learning_rate": 3.1552103614147955e-06,
+      "loss": 0.1006,
+      "step": 8020
+    },
+    {
+      "epoch": 1.2995787427090084,
+      "grad_norm": 0.7662655115127563,
+      "learning_rate": 3.154788333246812e-06,
+      "loss": 0.093,
+      "step": 8021
+    },
+    {
+      "epoch": 1.2997407647440051,
+      "grad_norm": 0.7853266000747681,
+      "learning_rate": 3.1543662850438905e-06,
+      "loss": 0.0991,
+      "step": 8022
+    },
+    {
+      "epoch": 1.2999027867790018,
+      "grad_norm": 0.8680589199066162,
+      "learning_rate": 3.153944216818943e-06,
+      "loss": 0.1214,
+      "step": 8023
+    },
+    {
+      "epoch": 1.3000648088139988,
+      "grad_norm": 0.7828645706176758,
+      "learning_rate": 3.1535221285848866e-06,
+      "loss": 0.1039,
+      "step": 8024
+    },
+    {
+      "epoch": 1.3002268308489955,
+      "grad_norm": 0.8467226624488831,
+      "learning_rate": 3.1531000203546336e-06,
+      "loss": 0.1025,
+      "step": 8025
+    },
+    {
+      "epoch": 1.3003888528839922,
+      "grad_norm": 0.8064707517623901,
+      "learning_rate": 3.1526778921411006e-06,
+      "loss": 0.1102,
+      "step": 8026
+    },
+    {
+      "epoch": 1.300550874918989,
+      "grad_norm": 0.8857673406600952,
+      "learning_rate": 3.152255743957203e-06,
+      "loss": 0.1082,
+      "step": 8027
+    },
+    {
+      "epoch": 1.3007128969539856,
+      "grad_norm": 0.7488471865653992,
+      "learning_rate": 3.151833575815859e-06,
+      "loss": 0.1013,
+      "step": 8028
+    },
+    {
+      "epoch": 1.3008749189889826,
+      "grad_norm": 0.7051052451133728,
+      "learning_rate": 3.1514113877299844e-06,
+      "loss": 0.087,
+      "step": 8029
+    },
+    {
+      "epoch": 1.3010369410239793,
+      "grad_norm": 0.8191677927970886,
+      "learning_rate": 3.1509891797124977e-06,
+      "loss": 0.1059,
+      "step": 8030
+    },
+    {
+      "epoch": 1.301198963058976,
+      "grad_norm": 0.8188061118125916,
+      "learning_rate": 3.150566951776318e-06,
+      "loss": 0.1143,
+      "step": 8031
+    },
+    {
+      "epoch": 1.301360985093973,
+      "grad_norm": 0.9264690279960632,
+      "learning_rate": 3.150144703934363e-06,
+      "loss": 0.1195,
+      "step": 8032
+    },
+    {
+      "epoch": 1.3015230071289696,
+      "grad_norm": 0.786266565322876,
+      "learning_rate": 3.1497224361995544e-06,
+      "loss": 0.1015,
+      "step": 8033
+    },
+    {
+      "epoch": 1.3016850291639663,
+      "grad_norm": 0.9195615649223328,
+      "learning_rate": 3.149300148584811e-06,
+      "loss": 0.1219,
+      "step": 8034
+    },
+    {
+      "epoch": 1.301847051198963,
+      "grad_norm": 0.8377419114112854,
+      "learning_rate": 3.1488778411030547e-06,
+      "loss": 0.1083,
+      "step": 8035
+    },
+    {
+      "epoch": 1.3020090732339598,
+      "grad_norm": 0.7512086629867554,
+      "learning_rate": 3.1484555137672063e-06,
+      "loss": 0.0884,
+      "step": 8036
+    },
+    {
+      "epoch": 1.3021710952689567,
+      "grad_norm": 0.7718841433525085,
+      "learning_rate": 3.148033166590188e-06,
+      "loss": 0.0941,
+      "step": 8037
+    },
+    {
+      "epoch": 1.3023331173039534,
+      "grad_norm": 0.7764375805854797,
+      "learning_rate": 3.147610799584924e-06,
+      "loss": 0.1025,
+      "step": 8038
+    },
+    {
+      "epoch": 1.3024951393389501,
+      "grad_norm": 0.9073190689086914,
+      "learning_rate": 3.147188412764336e-06,
+      "loss": 0.1038,
+      "step": 8039
+    },
+    {
+      "epoch": 1.3026571613739468,
+      "grad_norm": 0.794582724571228,
+      "learning_rate": 3.1467660061413497e-06,
+      "loss": 0.0986,
+      "step": 8040
+    },
+    {
+      "epoch": 1.3028191834089435,
+      "grad_norm": 0.8445225358009338,
+      "learning_rate": 3.1463435797288876e-06,
+      "loss": 0.11,
+      "step": 8041
+    },
+    {
+      "epoch": 1.3029812054439405,
+      "grad_norm": 0.8097845315933228,
+      "learning_rate": 3.1459211335398765e-06,
+      "loss": 0.1001,
+      "step": 8042
+    },
+    {
+      "epoch": 1.3031432274789372,
+      "grad_norm": 0.8664547801017761,
+      "learning_rate": 3.1454986675872417e-06,
+      "loss": 0.1093,
+      "step": 8043
+    },
+    {
+      "epoch": 1.303305249513934,
+      "grad_norm": 0.9166635274887085,
+      "learning_rate": 3.14507618188391e-06,
+      "loss": 0.1105,
+      "step": 8044
+    },
+    {
+      "epoch": 1.3034672715489306,
+      "grad_norm": 0.8189054727554321,
+      "learning_rate": 3.1446536764428083e-06,
+      "loss": 0.1063,
+      "step": 8045
+    },
+    {
+      "epoch": 1.3036292935839273,
+      "grad_norm": 0.8049901723861694,
+      "learning_rate": 3.144231151276864e-06,
+      "loss": 0.1105,
+      "step": 8046
+    },
+    {
+      "epoch": 1.3037913156189243,
+      "grad_norm": 0.8540433049201965,
+      "learning_rate": 3.1438086063990054e-06,
+      "loss": 0.1084,
+      "step": 8047
+    },
+    {
+      "epoch": 1.303953337653921,
+      "grad_norm": 0.8205851912498474,
+      "learning_rate": 3.143386041822162e-06,
+      "loss": 0.0978,
+      "step": 8048
+    },
+    {
+      "epoch": 1.3041153596889177,
+      "grad_norm": 0.9036335945129395,
+      "learning_rate": 3.1429634575592617e-06,
+      "loss": 0.1028,
+      "step": 8049
+    },
+    {
+      "epoch": 1.3042773817239144,
+      "grad_norm": 0.8568320870399475,
+      "learning_rate": 3.142540853623236e-06,
+      "loss": 0.1099,
+      "step": 8050
+    },
+    {
+      "epoch": 1.304439403758911,
+      "grad_norm": 0.9346566796302795,
+      "learning_rate": 3.1421182300270146e-06,
+      "loss": 0.1128,
+      "step": 8051
+    },
+    {
+      "epoch": 1.304601425793908,
+      "grad_norm": 0.8423227667808533,
+      "learning_rate": 3.14169558678353e-06,
+      "loss": 0.1033,
+      "step": 8052
+    },
+    {
+      "epoch": 1.3047634478289047,
+      "grad_norm": 0.8017847537994385,
+      "learning_rate": 3.1412729239057133e-06,
+      "loss": 0.0991,
+      "step": 8053
+    },
+    {
+      "epoch": 1.3049254698639015,
+      "grad_norm": 0.9350404739379883,
+      "learning_rate": 3.1408502414064963e-06,
+      "loss": 0.1108,
+      "step": 8054
+    },
+    {
+      "epoch": 1.3050874918988984,
+      "grad_norm": 0.7774401307106018,
+      "learning_rate": 3.140427539298814e-06,
+      "loss": 0.0955,
+      "step": 8055
+    },
+    {
+      "epoch": 1.3052495139338949,
+      "grad_norm": 0.7763550281524658,
+      "learning_rate": 3.140004817595597e-06,
+      "loss": 0.0892,
+      "step": 8056
+    },
+    {
+      "epoch": 1.3054115359688918,
+      "grad_norm": 0.8174300193786621,
+      "learning_rate": 3.139582076309783e-06,
+      "loss": 0.1017,
+      "step": 8057
+    },
+    {
+      "epoch": 1.3055735580038885,
+      "grad_norm": 0.7830010056495667,
+      "learning_rate": 3.1391593154543043e-06,
+      "loss": 0.0931,
+      "step": 8058
+    },
+    {
+      "epoch": 1.3057355800388852,
+      "grad_norm": 0.8983414769172668,
+      "learning_rate": 3.1387365350420973e-06,
+      "loss": 0.1045,
+      "step": 8059
+    },
+    {
+      "epoch": 1.3058976020738822,
+      "grad_norm": 0.8986459374427795,
+      "learning_rate": 3.138313735086099e-06,
+      "loss": 0.1145,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3060596241088789,
+      "grad_norm": 0.924191415309906,
+      "learning_rate": 3.137890915599243e-06,
+      "loss": 0.1091,
+      "step": 8061
+    },
+    {
+      "epoch": 1.3062216461438756,
+      "grad_norm": 0.9674059152603149,
+      "learning_rate": 3.137468076594471e-06,
+      "loss": 0.1148,
+      "step": 8062
+    },
+    {
+      "epoch": 1.3063836681788723,
+      "grad_norm": 0.8017002940177917,
+      "learning_rate": 3.1370452180847165e-06,
+      "loss": 0.108,
+      "step": 8063
+    },
+    {
+      "epoch": 1.306545690213869,
+      "grad_norm": 0.8276110291481018,
+      "learning_rate": 3.1366223400829215e-06,
+      "loss": 0.1081,
+      "step": 8064
+    },
+    {
+      "epoch": 1.306707712248866,
+      "grad_norm": 0.8368842005729675,
+      "learning_rate": 3.136199442602023e-06,
+      "loss": 0.1065,
+      "step": 8065
+    },
+    {
+      "epoch": 1.3068697342838627,
+      "grad_norm": 0.8028512597084045,
+      "learning_rate": 3.135776525654961e-06,
+      "loss": 0.0926,
+      "step": 8066
+    },
+    {
+      "epoch": 1.3070317563188594,
+      "grad_norm": 0.7708948850631714,
+      "learning_rate": 3.135353589254676e-06,
+      "loss": 0.0956,
+      "step": 8067
+    },
+    {
+      "epoch": 1.307193778353856,
+      "grad_norm": 0.7509301900863647,
+      "learning_rate": 3.1349306334141084e-06,
+      "loss": 0.1024,
+      "step": 8068
+    },
+    {
+      "epoch": 1.3073558003888528,
+      "grad_norm": 0.8279680609703064,
+      "learning_rate": 3.1345076581462007e-06,
+      "loss": 0.1069,
+      "step": 8069
+    },
+    {
+      "epoch": 1.3075178224238497,
+      "grad_norm": 0.763145387172699,
+      "learning_rate": 3.134084663463894e-06,
+      "loss": 0.097,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3076798444588464,
+      "grad_norm": 0.8883261680603027,
+      "learning_rate": 3.1336616493801305e-06,
+      "loss": 0.1074,
+      "step": 8071
+    },
+    {
+      "epoch": 1.3078418664938432,
+      "grad_norm": 0.9554191827774048,
+      "learning_rate": 3.1332386159078536e-06,
+      "loss": 0.1057,
+      "step": 8072
+    },
+    {
+      "epoch": 1.3080038885288399,
+      "grad_norm": 0.9759199023246765,
+      "learning_rate": 3.132815563060008e-06,
+      "loss": 0.1251,
+      "step": 8073
+    },
+    {
+      "epoch": 1.3081659105638366,
+      "grad_norm": 0.7938733696937561,
+      "learning_rate": 3.132392490849537e-06,
+      "loss": 0.0819,
+      "step": 8074
+    },
+    {
+      "epoch": 1.3083279325988335,
+      "grad_norm": 0.7426988482475281,
+      "learning_rate": 3.1319693992893874e-06,
+      "loss": 0.089,
+      "step": 8075
+    },
+    {
+      "epoch": 1.3084899546338302,
+      "grad_norm": 0.8131446838378906,
+      "learning_rate": 3.1315462883925026e-06,
+      "loss": 0.0958,
+      "step": 8076
+    },
+    {
+      "epoch": 1.308651976668827,
+      "grad_norm": 0.9161220192909241,
+      "learning_rate": 3.1311231581718303e-06,
+      "loss": 0.1206,
+      "step": 8077
+    },
+    {
+      "epoch": 1.3088139987038236,
+      "grad_norm": 0.834787905216217,
+      "learning_rate": 3.1307000086403162e-06,
+      "loss": 0.1093,
+      "step": 8078
+    },
+    {
+      "epoch": 1.3089760207388204,
+      "grad_norm": 1.0066626071929932,
+      "learning_rate": 3.1302768398109077e-06,
+      "loss": 0.1017,
+      "step": 8079
+    },
+    {
+      "epoch": 1.3091380427738173,
+      "grad_norm": 0.7865386605262756,
+      "learning_rate": 3.1298536516965537e-06,
+      "loss": 0.1025,
+      "step": 8080
+    },
+    {
+      "epoch": 1.309300064808814,
+      "grad_norm": 0.8171322345733643,
+      "learning_rate": 3.129430444310202e-06,
+      "loss": 0.1062,
+      "step": 8081
+    },
+    {
+      "epoch": 1.3094620868438107,
+      "grad_norm": 0.850459635257721,
+      "learning_rate": 3.129007217664802e-06,
+      "loss": 0.107,
+      "step": 8082
+    },
+    {
+      "epoch": 1.3096241088788076,
+      "grad_norm": 0.8040875792503357,
+      "learning_rate": 3.128583971773303e-06,
+      "loss": 0.1006,
+      "step": 8083
+    },
+    {
+      "epoch": 1.3097861309138044,
+      "grad_norm": 0.7739570736885071,
+      "learning_rate": 3.1281607066486565e-06,
+      "loss": 0.0991,
+      "step": 8084
+    },
+    {
+      "epoch": 1.309948152948801,
+      "grad_norm": 0.7959015369415283,
+      "learning_rate": 3.127737422303811e-06,
+      "loss": 0.097,
+      "step": 8085
+    },
+    {
+      "epoch": 1.3101101749837978,
+      "grad_norm": 0.8150349259376526,
+      "learning_rate": 3.127314118751721e-06,
+      "loss": 0.0953,
+      "step": 8086
+    },
+    {
+      "epoch": 1.3102721970187945,
+      "grad_norm": 0.8278794884681702,
+      "learning_rate": 3.1268907960053356e-06,
+      "loss": 0.0999,
+      "step": 8087
+    },
+    {
+      "epoch": 1.3104342190537914,
+      "grad_norm": 0.7229242324829102,
+      "learning_rate": 3.12646745407761e-06,
+      "loss": 0.0914,
+      "step": 8088
+    },
+    {
+      "epoch": 1.3105962410887881,
+      "grad_norm": 0.9541682004928589,
+      "learning_rate": 3.126044092981496e-06,
+      "loss": 0.1125,
+      "step": 8089
+    },
+    {
+      "epoch": 1.3107582631237849,
+      "grad_norm": 0.8512731790542603,
+      "learning_rate": 3.1256207127299475e-06,
+      "loss": 0.1065,
+      "step": 8090
+    },
+    {
+      "epoch": 1.3109202851587816,
+      "grad_norm": 0.9198841452598572,
+      "learning_rate": 3.12519731333592e-06,
+      "loss": 0.1117,
+      "step": 8091
+    },
+    {
+      "epoch": 1.3110823071937783,
+      "grad_norm": 0.7064549326896667,
+      "learning_rate": 3.124773894812367e-06,
+      "loss": 0.0941,
+      "step": 8092
+    },
+    {
+      "epoch": 1.3112443292287752,
+      "grad_norm": 0.806461751461029,
+      "learning_rate": 3.124350457172245e-06,
+      "loss": 0.0987,
+      "step": 8093
+    },
+    {
+      "epoch": 1.311406351263772,
+      "grad_norm": 0.7722499370574951,
+      "learning_rate": 3.123927000428509e-06,
+      "loss": 0.1029,
+      "step": 8094
+    },
+    {
+      "epoch": 1.3115683732987686,
+      "grad_norm": 0.9062454104423523,
+      "learning_rate": 3.123503524594118e-06,
+      "loss": 0.1205,
+      "step": 8095
+    },
+    {
+      "epoch": 1.3117303953337653,
+      "grad_norm": 0.9097535610198975,
+      "learning_rate": 3.123080029682027e-06,
+      "loss": 0.1028,
+      "step": 8096
+    },
+    {
+      "epoch": 1.311892417368762,
+      "grad_norm": 0.9619364738464355,
+      "learning_rate": 3.1226565157051953e-06,
+      "loss": 0.116,
+      "step": 8097
+    },
+    {
+      "epoch": 1.312054439403759,
+      "grad_norm": 0.845770001411438,
+      "learning_rate": 3.1222329826765806e-06,
+      "loss": 0.1112,
+      "step": 8098
+    },
+    {
+      "epoch": 1.3122164614387557,
+      "grad_norm": 0.7698118686676025,
+      "learning_rate": 3.121809430609143e-06,
+      "loss": 0.101,
+      "step": 8099
+    },
+    {
+      "epoch": 1.3123784834737524,
+      "grad_norm": 0.9154854416847229,
+      "learning_rate": 3.121385859515842e-06,
+      "loss": 0.101,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3125405055087491,
+      "grad_norm": 0.9142758250236511,
+      "learning_rate": 3.1209622694096362e-06,
+      "loss": 0.1169,
+      "step": 8101
+    },
+    {
+      "epoch": 1.3127025275437458,
+      "grad_norm": 0.8378267288208008,
+      "learning_rate": 3.1205386603034886e-06,
+      "loss": 0.1112,
+      "step": 8102
+    },
+    {
+      "epoch": 1.3128645495787428,
+      "grad_norm": 0.8595819473266602,
+      "learning_rate": 3.1201150322103593e-06,
+      "loss": 0.1037,
+      "step": 8103
+    },
+    {
+      "epoch": 1.3130265716137395,
+      "grad_norm": 0.9312635660171509,
+      "learning_rate": 3.1196913851432108e-06,
+      "loss": 0.1021,
+      "step": 8104
+    },
+    {
+      "epoch": 1.3131885936487362,
+      "grad_norm": 0.7517058253288269,
+      "learning_rate": 3.119267719115005e-06,
+      "loss": 0.0987,
+      "step": 8105
+    },
+    {
+      "epoch": 1.3133506156837331,
+      "grad_norm": 0.7930306196212769,
+      "learning_rate": 3.1188440341387063e-06,
+      "loss": 0.0988,
+      "step": 8106
+    },
+    {
+      "epoch": 1.3135126377187296,
+      "grad_norm": 0.692436933517456,
+      "learning_rate": 3.1184203302272775e-06,
+      "loss": 0.0828,
+      "step": 8107
+    },
+    {
+      "epoch": 1.3136746597537265,
+      "grad_norm": 0.7574872970581055,
+      "learning_rate": 3.1179966073936837e-06,
+      "loss": 0.0962,
+      "step": 8108
+    },
+    {
+      "epoch": 1.3138366817887233,
+      "grad_norm": 0.758668065071106,
+      "learning_rate": 3.1175728656508874e-06,
+      "loss": 0.1004,
+      "step": 8109
+    },
+    {
+      "epoch": 1.31399870382372,
+      "grad_norm": 0.7493375539779663,
+      "learning_rate": 3.117149105011858e-06,
+      "loss": 0.0977,
+      "step": 8110
+    },
+    {
+      "epoch": 1.314160725858717,
+      "grad_norm": 0.7759285569190979,
+      "learning_rate": 3.1167253254895584e-06,
+      "loss": 0.1007,
+      "step": 8111
+    },
+    {
+      "epoch": 1.3143227478937136,
+      "grad_norm": 0.8017150163650513,
+      "learning_rate": 3.1163015270969567e-06,
+      "loss": 0.101,
+      "step": 8112
+    },
+    {
+      "epoch": 1.3144847699287103,
+      "grad_norm": 0.9174550771713257,
+      "learning_rate": 3.1158777098470194e-06,
+      "loss": 0.1181,
+      "step": 8113
+    },
+    {
+      "epoch": 1.314646791963707,
+      "grad_norm": 1.0454412698745728,
+      "learning_rate": 3.115453873752714e-06,
+      "loss": 0.1151,
+      "step": 8114
+    },
+    {
+      "epoch": 1.3148088139987038,
+      "grad_norm": 0.9942188262939453,
+      "learning_rate": 3.115030018827011e-06,
+      "loss": 0.1252,
+      "step": 8115
+    },
+    {
+      "epoch": 1.3149708360337007,
+      "grad_norm": 0.8930821418762207,
+      "learning_rate": 3.114606145082876e-06,
+      "loss": 0.1013,
+      "step": 8116
+    },
+    {
+      "epoch": 1.3151328580686974,
+      "grad_norm": 0.8020307421684265,
+      "learning_rate": 3.1141822525332815e-06,
+      "loss": 0.0981,
+      "step": 8117
+    },
+    {
+      "epoch": 1.315294880103694,
+      "grad_norm": 0.9933139681816101,
+      "learning_rate": 3.1137583411911954e-06,
+      "loss": 0.1177,
+      "step": 8118
+    },
+    {
+      "epoch": 1.3154569021386908,
+      "grad_norm": 0.9036347270011902,
+      "learning_rate": 3.113334411069591e-06,
+      "loss": 0.1096,
+      "step": 8119
+    },
+    {
+      "epoch": 1.3156189241736875,
+      "grad_norm": 0.874993622303009,
+      "learning_rate": 3.1129104621814365e-06,
+      "loss": 0.1011,
+      "step": 8120
+    },
+    {
+      "epoch": 1.3157809462086845,
+      "grad_norm": 0.870242178440094,
+      "learning_rate": 3.112486494539705e-06,
+      "loss": 0.1122,
+      "step": 8121
+    },
+    {
+      "epoch": 1.3159429682436812,
+      "grad_norm": 0.7542531490325928,
+      "learning_rate": 3.1120625081573696e-06,
+      "loss": 0.0923,
+      "step": 8122
+    },
+    {
+      "epoch": 1.3161049902786779,
+      "grad_norm": 0.8048637509346008,
+      "learning_rate": 3.111638503047402e-06,
+      "loss": 0.0955,
+      "step": 8123
+    },
+    {
+      "epoch": 1.3162670123136746,
+      "grad_norm": 0.7695650458335876,
+      "learning_rate": 3.1112144792227774e-06,
+      "loss": 0.0995,
+      "step": 8124
+    },
+    {
+      "epoch": 1.3164290343486713,
+      "grad_norm": 0.8860523700714111,
+      "learning_rate": 3.110790436696468e-06,
+      "loss": 0.113,
+      "step": 8125
+    },
+    {
+      "epoch": 1.3165910563836682,
+      "grad_norm": 0.7619615197181702,
+      "learning_rate": 3.1103663754814493e-06,
+      "loss": 0.1018,
+      "step": 8126
+    },
+    {
+      "epoch": 1.316753078418665,
+      "grad_norm": 1.0961564779281616,
+      "learning_rate": 3.1099422955906965e-06,
+      "loss": 0.1428,
+      "step": 8127
+    },
+    {
+      "epoch": 1.3169151004536617,
+      "grad_norm": 0.8678660988807678,
+      "learning_rate": 3.109518197037186e-06,
+      "loss": 0.1052,
+      "step": 8128
+    },
+    {
+      "epoch": 1.3170771224886584,
+      "grad_norm": 0.8276517987251282,
+      "learning_rate": 3.109094079833893e-06,
+      "loss": 0.0979,
+      "step": 8129
+    },
+    {
+      "epoch": 1.317239144523655,
+      "grad_norm": 0.8221682906150818,
+      "learning_rate": 3.1086699439937957e-06,
+      "loss": 0.101,
+      "step": 8130
+    },
+    {
+      "epoch": 1.317401166558652,
+      "grad_norm": 0.7636736631393433,
+      "learning_rate": 3.1082457895298705e-06,
+      "loss": 0.0981,
+      "step": 8131
+    },
+    {
+      "epoch": 1.3175631885936487,
+      "grad_norm": 0.7403510212898254,
+      "learning_rate": 3.1078216164550966e-06,
+      "loss": 0.1002,
+      "step": 8132
+    },
+    {
+      "epoch": 1.3177252106286454,
+      "grad_norm": 0.9044564962387085,
+      "learning_rate": 3.1073974247824523e-06,
+      "loss": 0.1244,
+      "step": 8133
+    },
+    {
+      "epoch": 1.3178872326636424,
+      "grad_norm": 0.7922056913375854,
+      "learning_rate": 3.1069732145249166e-06,
+      "loss": 0.1003,
+      "step": 8134
+    },
+    {
+      "epoch": 1.318049254698639,
+      "grad_norm": 0.8038402795791626,
+      "learning_rate": 3.106548985695469e-06,
+      "loss": 0.1019,
+      "step": 8135
+    },
+    {
+      "epoch": 1.3182112767336358,
+      "grad_norm": 0.7442052364349365,
+      "learning_rate": 3.1061247383070905e-06,
+      "loss": 0.0865,
+      "step": 8136
+    },
+    {
+      "epoch": 1.3183732987686325,
+      "grad_norm": 0.6717336177825928,
+      "learning_rate": 3.105700472372762e-06,
+      "loss": 0.0862,
+      "step": 8137
+    },
+    {
+      "epoch": 1.3185353208036292,
+      "grad_norm": 0.7845094799995422,
+      "learning_rate": 3.1052761879054637e-06,
+      "loss": 0.1059,
+      "step": 8138
+    },
+    {
+      "epoch": 1.3186973428386262,
+      "grad_norm": 0.9265906810760498,
+      "learning_rate": 3.1048518849181795e-06,
+      "loss": 0.1059,
+      "step": 8139
+    },
+    {
+      "epoch": 1.3188593648736229,
+      "grad_norm": 0.7679669260978699,
+      "learning_rate": 3.1044275634238913e-06,
+      "loss": 0.097,
+      "step": 8140
+    },
+    {
+      "epoch": 1.3190213869086196,
+      "grad_norm": 0.7996553182601929,
+      "learning_rate": 3.1040032234355827e-06,
+      "loss": 0.0952,
+      "step": 8141
+    },
+    {
+      "epoch": 1.3191834089436163,
+      "grad_norm": 0.8960670232772827,
+      "learning_rate": 3.103578864966237e-06,
+      "loss": 0.1098,
+      "step": 8142
+    },
+    {
+      "epoch": 1.319345430978613,
+      "grad_norm": 0.8924859166145325,
+      "learning_rate": 3.1031544880288384e-06,
+      "loss": 0.1093,
+      "step": 8143
+    },
+    {
+      "epoch": 1.31950745301361,
+      "grad_norm": 0.874876856803894,
+      "learning_rate": 3.1027300926363723e-06,
+      "loss": 0.1064,
+      "step": 8144
+    },
+    {
+      "epoch": 1.3196694750486067,
+      "grad_norm": 1.0655916929244995,
+      "learning_rate": 3.1023056788018234e-06,
+      "loss": 0.1288,
+      "step": 8145
+    },
+    {
+      "epoch": 1.3198314970836034,
+      "grad_norm": 0.8831046223640442,
+      "learning_rate": 3.1018812465381796e-06,
+      "loss": 0.1017,
+      "step": 8146
+    },
+    {
+      "epoch": 1.3199935191186,
+      "grad_norm": 0.7947402596473694,
+      "learning_rate": 3.1014567958584246e-06,
+      "loss": 0.0977,
+      "step": 8147
+    },
+    {
+      "epoch": 1.3201555411535968,
+      "grad_norm": 0.945785641670227,
+      "learning_rate": 3.1010323267755486e-06,
+      "loss": 0.1148,
+      "step": 8148
+    },
+    {
+      "epoch": 1.3203175631885937,
+      "grad_norm": 0.8066475987434387,
+      "learning_rate": 3.1006078393025366e-06,
+      "loss": 0.1053,
+      "step": 8149
+    },
+    {
+      "epoch": 1.3204795852235904,
+      "grad_norm": 0.8218502998352051,
+      "learning_rate": 3.100183333452379e-06,
+      "loss": 0.0961,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3206416072585871,
+      "grad_norm": 0.7845996618270874,
+      "learning_rate": 3.0997588092380636e-06,
+      "loss": 0.1035,
+      "step": 8151
+    },
+    {
+      "epoch": 1.3208036292935839,
+      "grad_norm": 0.7798967957496643,
+      "learning_rate": 3.0993342666725803e-06,
+      "loss": 0.095,
+      "step": 8152
+    },
+    {
+      "epoch": 1.3209656513285806,
+      "grad_norm": 0.8794483542442322,
+      "learning_rate": 3.0989097057689175e-06,
+      "loss": 0.1131,
+      "step": 8153
+    },
+    {
+      "epoch": 1.3211276733635775,
+      "grad_norm": 0.8552145957946777,
+      "learning_rate": 3.0984851265400683e-06,
+      "loss": 0.0966,
+      "step": 8154
+    },
+    {
+      "epoch": 1.3212896953985742,
+      "grad_norm": 0.8511651158332825,
+      "learning_rate": 3.098060528999023e-06,
+      "loss": 0.102,
+      "step": 8155
+    },
+    {
+      "epoch": 1.321451717433571,
+      "grad_norm": 0.8631334900856018,
+      "learning_rate": 3.097635913158772e-06,
+      "loss": 0.1088,
+      "step": 8156
+    },
+    {
+      "epoch": 1.3216137394685679,
+      "grad_norm": 0.6932099461555481,
+      "learning_rate": 3.0972112790323076e-06,
+      "loss": 0.0885,
+      "step": 8157
+    },
+    {
+      "epoch": 1.3217757615035644,
+      "grad_norm": 0.8453887104988098,
+      "learning_rate": 3.096786626632624e-06,
+      "loss": 0.111,
+      "step": 8158
+    },
+    {
+      "epoch": 1.3219377835385613,
+      "grad_norm": 0.8994348049163818,
+      "learning_rate": 3.0963619559727143e-06,
+      "loss": 0.1123,
+      "step": 8159
+    },
+    {
+      "epoch": 1.322099805573558,
+      "grad_norm": 0.869174599647522,
+      "learning_rate": 3.0959372670655714e-06,
+      "loss": 0.1109,
+      "step": 8160
+    },
+    {
+      "epoch": 1.3222618276085547,
+      "grad_norm": 0.7786162495613098,
+      "learning_rate": 3.09551255992419e-06,
+      "loss": 0.0986,
+      "step": 8161
+    },
+    {
+      "epoch": 1.3224238496435516,
+      "grad_norm": 0.8360916972160339,
+      "learning_rate": 3.0950878345615654e-06,
+      "loss": 0.1046,
+      "step": 8162
+    },
+    {
+      "epoch": 1.3225858716785484,
+      "grad_norm": 0.8206416368484497,
+      "learning_rate": 3.0946630909906943e-06,
+      "loss": 0.1086,
+      "step": 8163
+    },
+    {
+      "epoch": 1.322747893713545,
+      "grad_norm": 0.904058575630188,
+      "learning_rate": 3.0942383292245704e-06,
+      "loss": 0.1145,
+      "step": 8164
+    },
+    {
+      "epoch": 1.3229099157485418,
+      "grad_norm": 0.8536134362220764,
+      "learning_rate": 3.0938135492761923e-06,
+      "loss": 0.1122,
+      "step": 8165
+    },
+    {
+      "epoch": 1.3230719377835385,
+      "grad_norm": 0.7881593108177185,
+      "learning_rate": 3.0933887511585564e-06,
+      "loss": 0.0954,
+      "step": 8166
+    },
+    {
+      "epoch": 1.3232339598185354,
+      "grad_norm": 0.8875266909599304,
+      "learning_rate": 3.0929639348846604e-06,
+      "loss": 0.1132,
+      "step": 8167
+    },
+    {
+      "epoch": 1.3233959818535321,
+      "grad_norm": 0.8193500638008118,
+      "learning_rate": 3.0925391004675037e-06,
+      "loss": 0.1013,
+      "step": 8168
+    },
+    {
+      "epoch": 1.3235580038885288,
+      "grad_norm": 0.8682994842529297,
+      "learning_rate": 3.0921142479200833e-06,
+      "loss": 0.1098,
+      "step": 8169
+    },
+    {
+      "epoch": 1.3237200259235256,
+      "grad_norm": 0.7923477292060852,
+      "learning_rate": 3.0916893772554006e-06,
+      "loss": 0.0961,
+      "step": 8170
+    },
+    {
+      "epoch": 1.3238820479585223,
+      "grad_norm": 0.8617781400680542,
+      "learning_rate": 3.0912644884864547e-06,
+      "loss": 0.1053,
+      "step": 8171
+    },
+    {
+      "epoch": 1.3240440699935192,
+      "grad_norm": 0.7619832158088684,
+      "learning_rate": 3.0908395816262466e-06,
+      "loss": 0.1016,
+      "step": 8172
+    },
+    {
+      "epoch": 1.324206092028516,
+      "grad_norm": 0.8054799437522888,
+      "learning_rate": 3.0904146566877762e-06,
+      "loss": 0.1082,
+      "step": 8173
+    },
+    {
+      "epoch": 1.3243681140635126,
+      "grad_norm": 0.8365015387535095,
+      "learning_rate": 3.0899897136840468e-06,
+      "loss": 0.1075,
+      "step": 8174
+    },
+    {
+      "epoch": 1.3245301360985093,
+      "grad_norm": 0.753804087638855,
+      "learning_rate": 3.0895647526280598e-06,
+      "loss": 0.0947,
+      "step": 8175
+    },
+    {
+      "epoch": 1.324692158133506,
+      "grad_norm": 0.9248984456062317,
+      "learning_rate": 3.0891397735328176e-06,
+      "loss": 0.1242,
+      "step": 8176
+    },
+    {
+      "epoch": 1.324854180168503,
+      "grad_norm": 0.7784146070480347,
+      "learning_rate": 3.088714776411325e-06,
+      "loss": 0.1127,
+      "step": 8177
+    },
+    {
+      "epoch": 1.3250162022034997,
+      "grad_norm": 0.8793583512306213,
+      "learning_rate": 3.088289761276584e-06,
+      "loss": 0.09,
+      "step": 8178
+    },
+    {
+      "epoch": 1.3251782242384964,
+      "grad_norm": 0.7790801525115967,
+      "learning_rate": 3.0878647281416007e-06,
+      "loss": 0.0928,
+      "step": 8179
+    },
+    {
+      "epoch": 1.3253402462734931,
+      "grad_norm": 0.8331507444381714,
+      "learning_rate": 3.0874396770193785e-06,
+      "loss": 0.1003,
+      "step": 8180
+    },
+    {
+      "epoch": 1.3255022683084898,
+      "grad_norm": 0.8366468548774719,
+      "learning_rate": 3.0870146079229245e-06,
+      "loss": 0.1221,
+      "step": 8181
+    },
+    {
+      "epoch": 1.3256642903434868,
+      "grad_norm": 0.8266817331314087,
+      "learning_rate": 3.0865895208652436e-06,
+      "loss": 0.1096,
+      "step": 8182
+    },
+    {
+      "epoch": 1.3258263123784835,
+      "grad_norm": 0.8993229866027832,
+      "learning_rate": 3.086164415859343e-06,
+      "loss": 0.1211,
+      "step": 8183
+    },
+    {
+      "epoch": 1.3259883344134802,
+      "grad_norm": 0.8876615762710571,
+      "learning_rate": 3.0857392929182296e-06,
+      "loss": 0.1092,
+      "step": 8184
+    },
+    {
+      "epoch": 1.3261503564484771,
+      "grad_norm": 0.7923975586891174,
+      "learning_rate": 3.085314152054911e-06,
+      "loss": 0.1,
+      "step": 8185
+    },
+    {
+      "epoch": 1.3263123784834738,
+      "grad_norm": 0.7979410886764526,
+      "learning_rate": 3.084888993282397e-06,
+      "loss": 0.1007,
+      "step": 8186
+    },
+    {
+      "epoch": 1.3264744005184705,
+      "grad_norm": 0.9142311215400696,
+      "learning_rate": 3.0844638166136943e-06,
+      "loss": 0.1154,
+      "step": 8187
+    },
+    {
+      "epoch": 1.3266364225534673,
+      "grad_norm": 0.8483611941337585,
+      "learning_rate": 3.0840386220618137e-06,
+      "loss": 0.1061,
+      "step": 8188
+    },
+    {
+      "epoch": 1.326798444588464,
+      "grad_norm": 0.9049779772758484,
+      "learning_rate": 3.0836134096397642e-06,
+      "loss": 0.1039,
+      "step": 8189
+    },
+    {
+      "epoch": 1.326960466623461,
+      "grad_norm": 0.8768754601478577,
+      "learning_rate": 3.083188179360556e-06,
+      "loss": 0.1141,
+      "step": 8190
+    },
+    {
+      "epoch": 1.3271224886584576,
+      "grad_norm": 0.7958570718765259,
+      "learning_rate": 3.082762931237202e-06,
+      "loss": 0.1006,
+      "step": 8191
+    },
+    {
+      "epoch": 1.3272845106934543,
+      "grad_norm": 0.8159347176551819,
+      "learning_rate": 3.0823376652827123e-06,
+      "loss": 0.1026,
+      "step": 8192
+    },
+    {
+      "epoch": 1.327446532728451,
+      "grad_norm": 0.8997524976730347,
+      "learning_rate": 3.081912381510099e-06,
+      "loss": 0.1096,
+      "step": 8193
+    },
+    {
+      "epoch": 1.3276085547634477,
+      "grad_norm": 0.8138853311538696,
+      "learning_rate": 3.0814870799323748e-06,
+      "loss": 0.1031,
+      "step": 8194
+    },
+    {
+      "epoch": 1.3277705767984447,
+      "grad_norm": 0.8049913048744202,
+      "learning_rate": 3.0810617605625538e-06,
+      "loss": 0.0991,
+      "step": 8195
+    },
+    {
+      "epoch": 1.3279325988334414,
+      "grad_norm": 0.7610349059104919,
+      "learning_rate": 3.080636423413649e-06,
+      "loss": 0.1038,
+      "step": 8196
+    },
+    {
+      "epoch": 1.328094620868438,
+      "grad_norm": 0.8136640191078186,
+      "learning_rate": 3.0802110684986742e-06,
+      "loss": 0.101,
+      "step": 8197
+    },
+    {
+      "epoch": 1.3282566429034348,
+      "grad_norm": 0.8479365706443787,
+      "learning_rate": 3.079785695830645e-06,
+      "loss": 0.1213,
+      "step": 8198
+    },
+    {
+      "epoch": 1.3284186649384315,
+      "grad_norm": 0.8573151230812073,
+      "learning_rate": 3.0793603054225767e-06,
+      "loss": 0.1135,
+      "step": 8199
+    },
+    {
+      "epoch": 1.3285806869734285,
+      "grad_norm": 0.8498213887214661,
+      "learning_rate": 3.0789348972874844e-06,
+      "loss": 0.1053,
+      "step": 8200
+    },
+    {
+      "epoch": 1.3287427090084252,
+      "grad_norm": 0.764045774936676,
+      "learning_rate": 3.078509471438386e-06,
+      "loss": 0.1018,
+      "step": 8201
+    },
+    {
+      "epoch": 1.3289047310434219,
+      "grad_norm": 0.7893896102905273,
+      "learning_rate": 3.0780840278882974e-06,
+      "loss": 0.0983,
+      "step": 8202
+    },
+    {
+      "epoch": 1.3290667530784186,
+      "grad_norm": 0.7424217462539673,
+      "learning_rate": 3.0776585666502367e-06,
+      "loss": 0.093,
+      "step": 8203
+    },
+    {
+      "epoch": 1.3292287751134153,
+      "grad_norm": 0.8020110726356506,
+      "learning_rate": 3.077233087737222e-06,
+      "loss": 0.1035,
+      "step": 8204
+    },
+    {
+      "epoch": 1.3293907971484122,
+      "grad_norm": 0.7891095280647278,
+      "learning_rate": 3.0768075911622712e-06,
+      "loss": 0.1097,
+      "step": 8205
+    },
+    {
+      "epoch": 1.329552819183409,
+      "grad_norm": 0.769567608833313,
+      "learning_rate": 3.0763820769384038e-06,
+      "loss": 0.0996,
+      "step": 8206
+    },
+    {
+      "epoch": 1.3297148412184057,
+      "grad_norm": 0.8365778923034668,
+      "learning_rate": 3.07595654507864e-06,
+      "loss": 0.0994,
+      "step": 8207
+    },
+    {
+      "epoch": 1.3298768632534026,
+      "grad_norm": 0.8200099468231201,
+      "learning_rate": 3.0755309955960007e-06,
+      "loss": 0.104,
+      "step": 8208
+    },
+    {
+      "epoch": 1.330038885288399,
+      "grad_norm": 0.7496293783187866,
+      "learning_rate": 3.0751054285035037e-06,
+      "loss": 0.1011,
+      "step": 8209
+    },
+    {
+      "epoch": 1.330200907323396,
+      "grad_norm": 0.9090161919593811,
+      "learning_rate": 3.074679843814174e-06,
+      "loss": 0.1109,
+      "step": 8210
+    },
+    {
+      "epoch": 1.3303629293583927,
+      "grad_norm": 0.8106396198272705,
+      "learning_rate": 3.0742542415410307e-06,
+      "loss": 0.0995,
+      "step": 8211
+    },
+    {
+      "epoch": 1.3305249513933894,
+      "grad_norm": 0.9422435164451599,
+      "learning_rate": 3.073828621697098e-06,
+      "loss": 0.1133,
+      "step": 8212
+    },
+    {
+      "epoch": 1.3306869734283864,
+      "grad_norm": 0.8159134387969971,
+      "learning_rate": 3.0734029842953976e-06,
+      "loss": 0.1045,
+      "step": 8213
+    },
+    {
+      "epoch": 1.330848995463383,
+      "grad_norm": 0.8325510621070862,
+      "learning_rate": 3.072977329348954e-06,
+      "loss": 0.1119,
+      "step": 8214
+    },
+    {
+      "epoch": 1.3310110174983798,
+      "grad_norm": 0.8749605417251587,
+      "learning_rate": 3.07255165687079e-06,
+      "loss": 0.1091,
+      "step": 8215
+    },
+    {
+      "epoch": 1.3311730395333765,
+      "grad_norm": 0.8472917079925537,
+      "learning_rate": 3.072125966873932e-06,
+      "loss": 0.11,
+      "step": 8216
+    },
+    {
+      "epoch": 1.3313350615683732,
+      "grad_norm": 0.9041043519973755,
+      "learning_rate": 3.0717002593714027e-06,
+      "loss": 0.1122,
+      "step": 8217
+    },
+    {
+      "epoch": 1.3314970836033702,
+      "grad_norm": 0.9191260933876038,
+      "learning_rate": 3.0712745343762295e-06,
+      "loss": 0.1157,
+      "step": 8218
+    },
+    {
+      "epoch": 1.3316591056383669,
+      "grad_norm": 0.8039606809616089,
+      "learning_rate": 3.070848791901438e-06,
+      "loss": 0.1042,
+      "step": 8219
+    },
+    {
+      "epoch": 1.3318211276733636,
+      "grad_norm": 0.8702144026756287,
+      "learning_rate": 3.0704230319600547e-06,
+      "loss": 0.1087,
+      "step": 8220
+    },
+    {
+      "epoch": 1.3319831497083603,
+      "grad_norm": 0.7728626132011414,
+      "learning_rate": 3.0699972545651067e-06,
+      "loss": 0.0971,
+      "step": 8221
+    },
+    {
+      "epoch": 1.332145171743357,
+      "grad_norm": 0.7619557976722717,
+      "learning_rate": 3.069571459729623e-06,
+      "loss": 0.092,
+      "step": 8222
+    },
+    {
+      "epoch": 1.332307193778354,
+      "grad_norm": 0.7611822485923767,
+      "learning_rate": 3.069145647466631e-06,
+      "loss": 0.1038,
+      "step": 8223
+    },
+    {
+      "epoch": 1.3324692158133506,
+      "grad_norm": 0.7991523146629333,
+      "learning_rate": 3.068719817789158e-06,
+      "loss": 0.1018,
+      "step": 8224
+    },
+    {
+      "epoch": 1.3326312378483474,
+      "grad_norm": 0.7888767719268799,
+      "learning_rate": 3.0682939707102366e-06,
+      "loss": 0.0988,
+      "step": 8225
+    },
+    {
+      "epoch": 1.332793259883344,
+      "grad_norm": 0.7734952569007874,
+      "learning_rate": 3.067868106242894e-06,
+      "loss": 0.0969,
+      "step": 8226
+    },
+    {
+      "epoch": 1.3329552819183408,
+      "grad_norm": 0.848461925983429,
+      "learning_rate": 3.0674422244001616e-06,
+      "loss": 0.1006,
+      "step": 8227
+    },
+    {
+      "epoch": 1.3331173039533377,
+      "grad_norm": 0.7484978437423706,
+      "learning_rate": 3.0670163251950703e-06,
+      "loss": 0.0981,
+      "step": 8228
+    },
+    {
+      "epoch": 1.3332793259883344,
+      "grad_norm": 0.7978498935699463,
+      "learning_rate": 3.0665904086406516e-06,
+      "loss": 0.1004,
+      "step": 8229
+    },
+    {
+      "epoch": 1.3334413480233311,
+      "grad_norm": 0.8219764828681946,
+      "learning_rate": 3.0661644747499385e-06,
+      "loss": 0.0978,
+      "step": 8230
+    },
+    {
+      "epoch": 1.3336033700583279,
+      "grad_norm": 0.910052478313446,
+      "learning_rate": 3.065738523535961e-06,
+      "loss": 0.1144,
+      "step": 8231
+    },
+    {
+      "epoch": 1.3337653920933246,
+      "grad_norm": 0.7298288941383362,
+      "learning_rate": 3.0653125550117547e-06,
+      "loss": 0.0957,
+      "step": 8232
+    },
+    {
+      "epoch": 1.3339274141283215,
+      "grad_norm": 0.7882039546966553,
+      "learning_rate": 3.064886569190352e-06,
+      "loss": 0.101,
+      "step": 8233
+    },
+    {
+      "epoch": 1.3340894361633182,
+      "grad_norm": 0.8351643085479736,
+      "learning_rate": 3.0644605660847875e-06,
+      "loss": 0.1027,
+      "step": 8234
+    },
+    {
+      "epoch": 1.334251458198315,
+      "grad_norm": 0.8841098546981812,
+      "learning_rate": 3.0640345457080955e-06,
+      "loss": 0.1066,
+      "step": 8235
+    },
+    {
+      "epoch": 1.3344134802333119,
+      "grad_norm": 0.9468151330947876,
+      "learning_rate": 3.0636085080733113e-06,
+      "loss": 0.1148,
+      "step": 8236
+    },
+    {
+      "epoch": 1.3345755022683086,
+      "grad_norm": 0.7528395056724548,
+      "learning_rate": 3.0631824531934707e-06,
+      "loss": 0.0928,
+      "step": 8237
+    },
+    {
+      "epoch": 1.3347375243033053,
+      "grad_norm": 0.81801837682724,
+      "learning_rate": 3.0627563810816097e-06,
+      "loss": 0.0954,
+      "step": 8238
+    },
+    {
+      "epoch": 1.334899546338302,
+      "grad_norm": 0.9190750122070312,
+      "learning_rate": 3.0623302917507657e-06,
+      "loss": 0.1095,
+      "step": 8239
+    },
+    {
+      "epoch": 1.3350615683732987,
+      "grad_norm": 0.8107596039772034,
+      "learning_rate": 3.0619041852139746e-06,
+      "loss": 0.1023,
+      "step": 8240
+    },
+    {
+      "epoch": 1.3352235904082956,
+      "grad_norm": 0.8642371296882629,
+      "learning_rate": 3.0614780614842764e-06,
+      "loss": 0.1095,
+      "step": 8241
+    },
+    {
+      "epoch": 1.3353856124432923,
+      "grad_norm": 0.92511385679245,
+      "learning_rate": 3.061051920574708e-06,
+      "loss": 0.1094,
+      "step": 8242
+    },
+    {
+      "epoch": 1.335547634478289,
+      "grad_norm": 0.87492835521698,
+      "learning_rate": 3.0606257624983082e-06,
+      "loss": 0.1073,
+      "step": 8243
+    },
+    {
+      "epoch": 1.3357096565132858,
+      "grad_norm": 0.7280792593955994,
+      "learning_rate": 3.0601995872681167e-06,
+      "loss": 0.0947,
+      "step": 8244
+    },
+    {
+      "epoch": 1.3358716785482825,
+      "grad_norm": 0.7862234711647034,
+      "learning_rate": 3.0597733948971737e-06,
+      "loss": 0.092,
+      "step": 8245
+    },
+    {
+      "epoch": 1.3360337005832794,
+      "grad_norm": 0.9407415390014648,
+      "learning_rate": 3.0593471853985197e-06,
+      "loss": 0.1128,
+      "step": 8246
+    },
+    {
+      "epoch": 1.3361957226182761,
+      "grad_norm": 0.7237008213996887,
+      "learning_rate": 3.0589209587851954e-06,
+      "loss": 0.0908,
+      "step": 8247
+    },
+    {
+      "epoch": 1.3363577446532728,
+      "grad_norm": 0.8024490475654602,
+      "learning_rate": 3.058494715070242e-06,
+      "loss": 0.096,
+      "step": 8248
+    },
+    {
+      "epoch": 1.3365197666882696,
+      "grad_norm": 0.7283364534378052,
+      "learning_rate": 3.0580684542667016e-06,
+      "loss": 0.0916,
+      "step": 8249
+    },
+    {
+      "epoch": 1.3366817887232663,
+      "grad_norm": 0.8358945250511169,
+      "learning_rate": 3.0576421763876174e-06,
+      "loss": 0.0954,
+      "step": 8250
+    },
+    {
+      "epoch": 1.3368438107582632,
+      "grad_norm": 0.9684699773788452,
+      "learning_rate": 3.0572158814460323e-06,
+      "loss": 0.126,
+      "step": 8251
+    },
+    {
+      "epoch": 1.33700583279326,
+      "grad_norm": 0.7843111157417297,
+      "learning_rate": 3.056789569454989e-06,
+      "loss": 0.099,
+      "step": 8252
+    },
+    {
+      "epoch": 1.3371678548282566,
+      "grad_norm": 0.7540812492370605,
+      "learning_rate": 3.056363240427533e-06,
+      "loss": 0.0931,
+      "step": 8253
+    },
+    {
+      "epoch": 1.3373298768632533,
+      "grad_norm": 0.9374097585678101,
+      "learning_rate": 3.055936894376708e-06,
+      "loss": 0.1093,
+      "step": 8254
+    },
+    {
+      "epoch": 1.33749189889825,
+      "grad_norm": 0.8523105978965759,
+      "learning_rate": 3.0555105313155587e-06,
+      "loss": 0.102,
+      "step": 8255
+    },
+    {
+      "epoch": 1.337653920933247,
+      "grad_norm": 0.914612352848053,
+      "learning_rate": 3.055084151257133e-06,
+      "loss": 0.1071,
+      "step": 8256
+    },
+    {
+      "epoch": 1.3378159429682437,
+      "grad_norm": 0.8243370056152344,
+      "learning_rate": 3.0546577542144734e-06,
+      "loss": 0.1144,
+      "step": 8257
+    },
+    {
+      "epoch": 1.3379779650032404,
+      "grad_norm": 0.7740969061851501,
+      "learning_rate": 3.054231340200631e-06,
+      "loss": 0.0894,
+      "step": 8258
+    },
+    {
+      "epoch": 1.3381399870382373,
+      "grad_norm": 0.6962242722511292,
+      "learning_rate": 3.053804909228649e-06,
+      "loss": 0.0889,
+      "step": 8259
+    },
+    {
+      "epoch": 1.3383020090732338,
+      "grad_norm": 0.7744547128677368,
+      "learning_rate": 3.053378461311578e-06,
+      "loss": 0.0924,
+      "step": 8260
+    },
+    {
+      "epoch": 1.3384640311082308,
+      "grad_norm": 0.8196523189544678,
+      "learning_rate": 3.052951996462465e-06,
+      "loss": 0.0982,
+      "step": 8261
+    },
+    {
+      "epoch": 1.3386260531432275,
+      "grad_norm": 0.7271378040313721,
+      "learning_rate": 3.0525255146943582e-06,
+      "loss": 0.0969,
+      "step": 8262
+    },
+    {
+      "epoch": 1.3387880751782242,
+      "grad_norm": 0.7864364981651306,
+      "learning_rate": 3.052099016020309e-06,
+      "loss": 0.0988,
+      "step": 8263
+    },
+    {
+      "epoch": 1.3389500972132211,
+      "grad_norm": 0.7747987508773804,
+      "learning_rate": 3.0516725004533648e-06,
+      "loss": 0.0925,
+      "step": 8264
+    },
+    {
+      "epoch": 1.3391121192482178,
+      "grad_norm": 0.8016791343688965,
+      "learning_rate": 3.0512459680065785e-06,
+      "loss": 0.1004,
+      "step": 8265
+    },
+    {
+      "epoch": 1.3392741412832145,
+      "grad_norm": 0.8657497763633728,
+      "learning_rate": 3.0508194186929983e-06,
+      "loss": 0.1072,
+      "step": 8266
+    },
+    {
+      "epoch": 1.3394361633182112,
+      "grad_norm": 0.8073965311050415,
+      "learning_rate": 3.0503928525256775e-06,
+      "loss": 0.098,
+      "step": 8267
+    },
+    {
+      "epoch": 1.339598185353208,
+      "grad_norm": 0.7711490988731384,
+      "learning_rate": 3.0499662695176675e-06,
+      "loss": 0.1004,
+      "step": 8268
+    },
+    {
+      "epoch": 1.339760207388205,
+      "grad_norm": 0.7585821151733398,
+      "learning_rate": 3.04953966968202e-06,
+      "loss": 0.0958,
+      "step": 8269
+    },
+    {
+      "epoch": 1.3399222294232016,
+      "grad_norm": 0.7577686309814453,
+      "learning_rate": 3.0491130530317887e-06,
+      "loss": 0.0968,
+      "step": 8270
+    },
+    {
+      "epoch": 1.3400842514581983,
+      "grad_norm": 0.7601026296615601,
+      "learning_rate": 3.048686419580027e-06,
+      "loss": 0.0991,
+      "step": 8271
+    },
+    {
+      "epoch": 1.340246273493195,
+      "grad_norm": 0.9418452382087708,
+      "learning_rate": 3.0482597693397887e-06,
+      "loss": 0.1168,
+      "step": 8272
+    },
+    {
+      "epoch": 1.3404082955281917,
+      "grad_norm": 0.8291217684745789,
+      "learning_rate": 3.047833102324128e-06,
+      "loss": 0.1024,
+      "step": 8273
+    },
+    {
+      "epoch": 1.3405703175631887,
+      "grad_norm": 0.8256798982620239,
+      "learning_rate": 3.0474064185461e-06,
+      "loss": 0.1015,
+      "step": 8274
+    },
+    {
+      "epoch": 1.3407323395981854,
+      "grad_norm": 0.823760986328125,
+      "learning_rate": 3.0469797180187606e-06,
+      "loss": 0.1101,
+      "step": 8275
+    },
+    {
+      "epoch": 1.340894361633182,
+      "grad_norm": 0.7810642123222351,
+      "learning_rate": 3.0465530007551646e-06,
+      "loss": 0.1018,
+      "step": 8276
+    },
+    {
+      "epoch": 1.3410563836681788,
+      "grad_norm": 0.8536295890808105,
+      "learning_rate": 3.04612626676837e-06,
+      "loss": 0.1047,
+      "step": 8277
+    },
+    {
+      "epoch": 1.3412184057031755,
+      "grad_norm": 0.837985098361969,
+      "learning_rate": 3.0456995160714344e-06,
+      "loss": 0.1103,
+      "step": 8278
+    },
+    {
+      "epoch": 1.3413804277381725,
+      "grad_norm": 0.765397846698761,
+      "learning_rate": 3.0452727486774118e-06,
+      "loss": 0.0972,
+      "step": 8279
+    },
+    {
+      "epoch": 1.3415424497731692,
+      "grad_norm": 0.8630374073982239,
+      "learning_rate": 3.044845964599365e-06,
+      "loss": 0.0982,
+      "step": 8280
+    },
+    {
+      "epoch": 1.3417044718081659,
+      "grad_norm": 0.7347887754440308,
+      "learning_rate": 3.044419163850349e-06,
+      "loss": 0.0935,
+      "step": 8281
+    },
+    {
+      "epoch": 1.3418664938431626,
+      "grad_norm": 0.773932158946991,
+      "learning_rate": 3.043992346443424e-06,
+      "loss": 0.0996,
+      "step": 8282
+    },
+    {
+      "epoch": 1.3420285158781593,
+      "grad_norm": 0.77265864610672,
+      "learning_rate": 3.04356551239165e-06,
+      "loss": 0.0913,
+      "step": 8283
+    },
+    {
+      "epoch": 1.3421905379131562,
+      "grad_norm": 0.7659427523612976,
+      "learning_rate": 3.043138661708086e-06,
+      "loss": 0.1001,
+      "step": 8284
+    },
+    {
+      "epoch": 1.342352559948153,
+      "grad_norm": 0.8944420218467712,
+      "learning_rate": 3.0427117944057943e-06,
+      "loss": 0.1002,
+      "step": 8285
+    },
+    {
+      "epoch": 1.3425145819831497,
+      "grad_norm": 0.8467785716056824,
+      "learning_rate": 3.042284910497834e-06,
+      "loss": 0.0995,
+      "step": 8286
+    },
+    {
+      "epoch": 1.3426766040181466,
+      "grad_norm": 0.9520701766014099,
+      "learning_rate": 3.0418580099972687e-06,
+      "loss": 0.1151,
+      "step": 8287
+    },
+    {
+      "epoch": 1.3428386260531433,
+      "grad_norm": 0.7847069501876831,
+      "learning_rate": 3.0414310929171587e-06,
+      "loss": 0.1078,
+      "step": 8288
+    },
+    {
+      "epoch": 1.34300064808814,
+      "grad_norm": 0.9613496661186218,
+      "learning_rate": 3.0410041592705687e-06,
+      "loss": 0.1154,
+      "step": 8289
+    },
+    {
+      "epoch": 1.3431626701231367,
+      "grad_norm": 0.8269860148429871,
+      "learning_rate": 3.04057720907056e-06,
+      "loss": 0.096,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3433246921581334,
+      "grad_norm": 0.8524847626686096,
+      "learning_rate": 3.0401502423301966e-06,
+      "loss": 0.111,
+      "step": 8291
+    },
+    {
+      "epoch": 1.3434867141931304,
+      "grad_norm": 0.7680644392967224,
+      "learning_rate": 3.039723259062543e-06,
+      "loss": 0.0969,
+      "step": 8292
+    },
+    {
+      "epoch": 1.343648736228127,
+      "grad_norm": 0.8680287003517151,
+      "learning_rate": 3.0392962592806635e-06,
+      "loss": 0.1138,
+      "step": 8293
+    },
+    {
+      "epoch": 1.3438107582631238,
+      "grad_norm": 0.8054776191711426,
+      "learning_rate": 3.0388692429976247e-06,
+      "loss": 0.0998,
+      "step": 8294
+    },
+    {
+      "epoch": 1.3439727802981205,
+      "grad_norm": 0.8584299087524414,
+      "learning_rate": 3.03844221022649e-06,
+      "loss": 0.1017,
+      "step": 8295
+    },
+    {
+      "epoch": 1.3441348023331172,
+      "grad_norm": 0.9022834300994873,
+      "learning_rate": 3.038015160980327e-06,
+      "loss": 0.1122,
+      "step": 8296
+    },
+    {
+      "epoch": 1.3442968243681142,
+      "grad_norm": 0.9513327479362488,
+      "learning_rate": 3.037588095272202e-06,
+      "loss": 0.1153,
+      "step": 8297
+    },
+    {
+      "epoch": 1.3444588464031109,
+      "grad_norm": 0.7637044787406921,
+      "learning_rate": 3.0371610131151823e-06,
+      "loss": 0.1051,
+      "step": 8298
+    },
+    {
+      "epoch": 1.3446208684381076,
+      "grad_norm": 0.8572631478309631,
+      "learning_rate": 3.0367339145223352e-06,
+      "loss": 0.1051,
+      "step": 8299
+    },
+    {
+      "epoch": 1.3447828904731043,
+      "grad_norm": 0.7625088095664978,
+      "learning_rate": 3.0363067995067297e-06,
+      "loss": 0.1026,
+      "step": 8300
+    },
+    {
+      "epoch": 1.344944912508101,
+      "grad_norm": 0.9297232627868652,
+      "learning_rate": 3.0358796680814333e-06,
+      "loss": 0.1002,
+      "step": 8301
+    },
+    {
+      "epoch": 1.345106934543098,
+      "grad_norm": 0.8700084090232849,
+      "learning_rate": 3.035452520259517e-06,
+      "loss": 0.1198,
+      "step": 8302
+    },
+    {
+      "epoch": 1.3452689565780946,
+      "grad_norm": 0.8522663116455078,
+      "learning_rate": 3.035025356054049e-06,
+      "loss": 0.106,
+      "step": 8303
+    },
+    {
+      "epoch": 1.3454309786130914,
+      "grad_norm": 0.7052852511405945,
+      "learning_rate": 3.034598175478099e-06,
+      "loss": 0.0937,
+      "step": 8304
+    },
+    {
+      "epoch": 1.345593000648088,
+      "grad_norm": 0.8086321353912354,
+      "learning_rate": 3.034170978544739e-06,
+      "loss": 0.1022,
+      "step": 8305
+    },
+    {
+      "epoch": 1.3457550226830848,
+      "grad_norm": 0.8114489912986755,
+      "learning_rate": 3.03374376526704e-06,
+      "loss": 0.1035,
+      "step": 8306
+    },
+    {
+      "epoch": 1.3459170447180817,
+      "grad_norm": 0.8548650741577148,
+      "learning_rate": 3.033316535658073e-06,
+      "loss": 0.1087,
+      "step": 8307
+    },
+    {
+      "epoch": 1.3460790667530784,
+      "grad_norm": 1.0227341651916504,
+      "learning_rate": 3.0328892897309105e-06,
+      "loss": 0.1139,
+      "step": 8308
+    },
+    {
+      "epoch": 1.3462410887880751,
+      "grad_norm": 0.9773898720741272,
+      "learning_rate": 3.032462027498626e-06,
+      "loss": 0.1205,
+      "step": 8309
+    },
+    {
+      "epoch": 1.346403110823072,
+      "grad_norm": 0.853813111782074,
+      "learning_rate": 3.0320347489742905e-06,
+      "loss": 0.1066,
+      "step": 8310
+    },
+    {
+      "epoch": 1.3465651328580686,
+      "grad_norm": 0.7821661829948425,
+      "learning_rate": 3.0316074541709813e-06,
+      "loss": 0.105,
+      "step": 8311
+    },
+    {
+      "epoch": 1.3467271548930655,
+      "grad_norm": 0.7272818684577942,
+      "learning_rate": 3.031180143101769e-06,
+      "loss": 0.0946,
+      "step": 8312
+    },
+    {
+      "epoch": 1.3468891769280622,
+      "grad_norm": 0.8688091039657593,
+      "learning_rate": 3.0307528157797306e-06,
+      "loss": 0.1046,
+      "step": 8313
+    },
+    {
+      "epoch": 1.347051198963059,
+      "grad_norm": 0.8635636568069458,
+      "learning_rate": 3.03032547221794e-06,
+      "loss": 0.1076,
+      "step": 8314
+    },
+    {
+      "epoch": 1.3472132209980558,
+      "grad_norm": 0.8457701802253723,
+      "learning_rate": 3.029898112429473e-06,
+      "loss": 0.1164,
+      "step": 8315
+    },
+    {
+      "epoch": 1.3473752430330526,
+      "grad_norm": 0.8496267199516296,
+      "learning_rate": 3.0294707364274066e-06,
+      "loss": 0.1043,
+      "step": 8316
+    },
+    {
+      "epoch": 1.3475372650680493,
+      "grad_norm": 0.7697789669036865,
+      "learning_rate": 3.0290433442248163e-06,
+      "loss": 0.0969,
+      "step": 8317
+    },
+    {
+      "epoch": 1.347699287103046,
+      "grad_norm": 0.8936092257499695,
+      "learning_rate": 3.028615935834781e-06,
+      "loss": 0.1069,
+      "step": 8318
+    },
+    {
+      "epoch": 1.3478613091380427,
+      "grad_norm": 0.7584058046340942,
+      "learning_rate": 3.028188511270376e-06,
+      "loss": 0.0984,
+      "step": 8319
+    },
+    {
+      "epoch": 1.3480233311730396,
+      "grad_norm": 0.7890931963920593,
+      "learning_rate": 3.027761070544682e-06,
+      "loss": 0.1024,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3481853532080363,
+      "grad_norm": 0.7327829599380493,
+      "learning_rate": 3.027333613670775e-06,
+      "loss": 0.09,
+      "step": 8321
+    },
+    {
+      "epoch": 1.348347375243033,
+      "grad_norm": 0.7482281923294067,
+      "learning_rate": 3.026906140661737e-06,
+      "loss": 0.102,
+      "step": 8322
+    },
+    {
+      "epoch": 1.3485093972780298,
+      "grad_norm": 0.7569482922554016,
+      "learning_rate": 3.0264786515306453e-06,
+      "loss": 0.096,
+      "step": 8323
+    },
+    {
+      "epoch": 1.3486714193130265,
+      "grad_norm": 0.7618415355682373,
+      "learning_rate": 3.026051146290581e-06,
+      "loss": 0.0961,
+      "step": 8324
+    },
+    {
+      "epoch": 1.3488334413480234,
+      "grad_norm": 0.8373454213142395,
+      "learning_rate": 3.0256236249546256e-06,
+      "loss": 0.103,
+      "step": 8325
+    },
+    {
+      "epoch": 1.3489954633830201,
+      "grad_norm": 0.9048294425010681,
+      "learning_rate": 3.025196087535858e-06,
+      "loss": 0.111,
+      "step": 8326
+    },
+    {
+      "epoch": 1.3491574854180168,
+      "grad_norm": 0.8331298232078552,
+      "learning_rate": 3.024768534047362e-06,
+      "loss": 0.1084,
+      "step": 8327
+    },
+    {
+      "epoch": 1.3493195074530135,
+      "grad_norm": 0.8242894411087036,
+      "learning_rate": 3.024340964502218e-06,
+      "loss": 0.1079,
+      "step": 8328
+    },
+    {
+      "epoch": 1.3494815294880103,
+      "grad_norm": 0.7231031060218811,
+      "learning_rate": 3.0239133789135094e-06,
+      "loss": 0.0966,
+      "step": 8329
+    },
+    {
+      "epoch": 1.3496435515230072,
+      "grad_norm": 0.7646691799163818,
+      "learning_rate": 3.0234857772943197e-06,
+      "loss": 0.0967,
+      "step": 8330
+    },
+    {
+      "epoch": 1.349805573558004,
+      "grad_norm": 0.7422172427177429,
+      "learning_rate": 3.023058159657732e-06,
+      "loss": 0.0914,
+      "step": 8331
+    },
+    {
+      "epoch": 1.3499675955930006,
+      "grad_norm": 0.7762441635131836,
+      "learning_rate": 3.0226305260168298e-06,
+      "loss": 0.0945,
+      "step": 8332
+    },
+    {
+      "epoch": 1.3501296176279973,
+      "grad_norm": 0.9666473865509033,
+      "learning_rate": 3.0222028763846994e-06,
+      "loss": 0.1287,
+      "step": 8333
+    },
+    {
+      "epoch": 1.350291639662994,
+      "grad_norm": 0.8605021834373474,
+      "learning_rate": 3.0217752107744237e-06,
+      "loss": 0.1062,
+      "step": 8334
+    },
+    {
+      "epoch": 1.350453661697991,
+      "grad_norm": 0.877007246017456,
+      "learning_rate": 3.0213475291990897e-06,
+      "loss": 0.1046,
+      "step": 8335
+    },
+    {
+      "epoch": 1.3506156837329877,
+      "grad_norm": 0.7591307163238525,
+      "learning_rate": 3.0209198316717825e-06,
+      "loss": 0.0909,
+      "step": 8336
+    },
+    {
+      "epoch": 1.3507777057679844,
+      "grad_norm": 0.9145157337188721,
+      "learning_rate": 3.020492118205589e-06,
+      "loss": 0.1147,
+      "step": 8337
+    },
+    {
+      "epoch": 1.3509397278029813,
+      "grad_norm": 0.8037732243537903,
+      "learning_rate": 3.0200643888135973e-06,
+      "loss": 0.1012,
+      "step": 8338
+    },
+    {
+      "epoch": 1.351101749837978,
+      "grad_norm": 0.9152517318725586,
+      "learning_rate": 3.0196366435088926e-06,
+      "loss": 0.115,
+      "step": 8339
+    },
+    {
+      "epoch": 1.3512637718729748,
+      "grad_norm": 0.868812620639801,
+      "learning_rate": 3.019208882304565e-06,
+      "loss": 0.1139,
+      "step": 8340
+    },
+    {
+      "epoch": 1.3514257939079715,
+      "grad_norm": 0.7427473068237305,
+      "learning_rate": 3.018781105213701e-06,
+      "loss": 0.0896,
+      "step": 8341
+    },
+    {
+      "epoch": 1.3515878159429682,
+      "grad_norm": 0.7403958439826965,
+      "learning_rate": 3.0183533122493917e-06,
+      "loss": 0.0943,
+      "step": 8342
+    },
+    {
+      "epoch": 1.351749837977965,
+      "grad_norm": 0.8998346328735352,
+      "learning_rate": 3.017925503424725e-06,
+      "loss": 0.1174,
+      "step": 8343
+    },
+    {
+      "epoch": 1.3519118600129618,
+      "grad_norm": 0.8514041304588318,
+      "learning_rate": 3.017497678752791e-06,
+      "loss": 0.1092,
+      "step": 8344
+    },
+    {
+      "epoch": 1.3520738820479585,
+      "grad_norm": 0.8314045667648315,
+      "learning_rate": 3.0170698382466805e-06,
+      "loss": 0.1074,
+      "step": 8345
+    },
+    {
+      "epoch": 1.3522359040829552,
+      "grad_norm": 0.9838075041770935,
+      "learning_rate": 3.016641981919485e-06,
+      "loss": 0.111,
+      "step": 8346
+    },
+    {
+      "epoch": 1.352397926117952,
+      "grad_norm": 0.9298703670501709,
+      "learning_rate": 3.0162141097842943e-06,
+      "loss": 0.1102,
+      "step": 8347
+    },
+    {
+      "epoch": 1.3525599481529489,
+      "grad_norm": 0.7722398638725281,
+      "learning_rate": 3.0157862218542004e-06,
+      "loss": 0.0921,
+      "step": 8348
+    },
+    {
+      "epoch": 1.3527219701879456,
+      "grad_norm": 0.9279158711433411,
+      "learning_rate": 3.015358318142298e-06,
+      "loss": 0.1113,
+      "step": 8349
+    },
+    {
+      "epoch": 1.3528839922229423,
+      "grad_norm": 0.8341646194458008,
+      "learning_rate": 3.0149303986616772e-06,
+      "loss": 0.1125,
+      "step": 8350
+    },
+    {
+      "epoch": 1.353046014257939,
+      "grad_norm": 0.8737770915031433,
+      "learning_rate": 3.0145024634254323e-06,
+      "loss": 0.1119,
+      "step": 8351
+    },
+    {
+      "epoch": 1.3532080362929357,
+      "grad_norm": 0.8188356757164001,
+      "learning_rate": 3.014074512446657e-06,
+      "loss": 0.0972,
+      "step": 8352
+    },
+    {
+      "epoch": 1.3533700583279327,
+      "grad_norm": 0.8271098136901855,
+      "learning_rate": 3.0136465457384454e-06,
+      "loss": 0.1017,
+      "step": 8353
+    },
+    {
+      "epoch": 1.3535320803629294,
+      "grad_norm": 0.852016031742096,
+      "learning_rate": 3.0132185633138934e-06,
+      "loss": 0.115,
+      "step": 8354
+    },
+    {
+      "epoch": 1.353694102397926,
+      "grad_norm": 0.8249271512031555,
+      "learning_rate": 3.0127905651860946e-06,
+      "loss": 0.0977,
+      "step": 8355
+    },
+    {
+      "epoch": 1.3538561244329228,
+      "grad_norm": 0.7700305581092834,
+      "learning_rate": 3.0123625513681463e-06,
+      "loss": 0.1055,
+      "step": 8356
+    },
+    {
+      "epoch": 1.3540181464679195,
+      "grad_norm": 0.9442518353462219,
+      "learning_rate": 3.0119345218731433e-06,
+      "loss": 0.1065,
+      "step": 8357
+    },
+    {
+      "epoch": 1.3541801685029164,
+      "grad_norm": 0.8089423179626465,
+      "learning_rate": 3.0115064767141827e-06,
+      "loss": 0.0996,
+      "step": 8358
+    },
+    {
+      "epoch": 1.3543421905379132,
+      "grad_norm": 0.9584595561027527,
+      "learning_rate": 3.0110784159043614e-06,
+      "loss": 0.1212,
+      "step": 8359
+    },
+    {
+      "epoch": 1.3545042125729099,
+      "grad_norm": 0.7914947867393494,
+      "learning_rate": 3.0106503394567775e-06,
+      "loss": 0.1007,
+      "step": 8360
+    },
+    {
+      "epoch": 1.3546662346079068,
+      "grad_norm": 0.7712557911872864,
+      "learning_rate": 3.0102222473845296e-06,
+      "loss": 0.1009,
+      "step": 8361
+    },
+    {
+      "epoch": 1.3548282566429035,
+      "grad_norm": 0.9304143190383911,
+      "learning_rate": 3.0097941397007156e-06,
+      "loss": 0.1097,
+      "step": 8362
+    },
+    {
+      "epoch": 1.3549902786779002,
+      "grad_norm": 0.7734033465385437,
+      "learning_rate": 3.0093660164184333e-06,
+      "loss": 0.0984,
+      "step": 8363
+    },
+    {
+      "epoch": 1.355152300712897,
+      "grad_norm": 0.7931289672851562,
+      "learning_rate": 3.008937877550785e-06,
+      "loss": 0.1012,
+      "step": 8364
+    },
+    {
+      "epoch": 1.3553143227478937,
+      "grad_norm": 0.8496924638748169,
+      "learning_rate": 3.008509723110869e-06,
+      "loss": 0.1057,
+      "step": 8365
+    },
+    {
+      "epoch": 1.3554763447828906,
+      "grad_norm": 0.8549742102622986,
+      "learning_rate": 3.008081553111786e-06,
+      "loss": 0.1145,
+      "step": 8366
+    },
+    {
+      "epoch": 1.3556383668178873,
+      "grad_norm": 0.8030797243118286,
+      "learning_rate": 3.007653367566636e-06,
+      "loss": 0.1055,
+      "step": 8367
+    },
+    {
+      "epoch": 1.355800388852884,
+      "grad_norm": 0.9936630725860596,
+      "learning_rate": 3.0072251664885222e-06,
+      "loss": 0.1309,
+      "step": 8368
+    },
+    {
+      "epoch": 1.3559624108878807,
+      "grad_norm": 0.8453781604766846,
+      "learning_rate": 3.0067969498905463e-06,
+      "loss": 0.0999,
+      "step": 8369
+    },
+    {
+      "epoch": 1.3561244329228774,
+      "grad_norm": 0.8510198593139648,
+      "learning_rate": 3.006368717785809e-06,
+      "loss": 0.103,
+      "step": 8370
+    },
+    {
+      "epoch": 1.3562864549578744,
+      "grad_norm": 0.770231306552887,
+      "learning_rate": 3.0059404701874157e-06,
+      "loss": 0.0957,
+      "step": 8371
+    },
+    {
+      "epoch": 1.356448476992871,
+      "grad_norm": 0.7896842956542969,
+      "learning_rate": 3.005512207108467e-06,
+      "loss": 0.0984,
+      "step": 8372
+    },
+    {
+      "epoch": 1.3566104990278678,
+      "grad_norm": 0.7394090294837952,
+      "learning_rate": 3.005083928562069e-06,
+      "loss": 0.094,
+      "step": 8373
+    },
+    {
+      "epoch": 1.3567725210628645,
+      "grad_norm": 0.8859434127807617,
+      "learning_rate": 3.004655634561325e-06,
+      "loss": 0.0986,
+      "step": 8374
+    },
+    {
+      "epoch": 1.3569345430978612,
+      "grad_norm": 1.0016928911209106,
+      "learning_rate": 3.004227325119339e-06,
+      "loss": 0.1174,
+      "step": 8375
+    },
+    {
+      "epoch": 1.3570965651328581,
+      "grad_norm": 0.7929229736328125,
+      "learning_rate": 3.003799000249218e-06,
+      "loss": 0.1043,
+      "step": 8376
+    },
+    {
+      "epoch": 1.3572585871678549,
+      "grad_norm": 0.8156124949455261,
+      "learning_rate": 3.0033706599640665e-06,
+      "loss": 0.106,
+      "step": 8377
+    },
+    {
+      "epoch": 1.3574206092028516,
+      "grad_norm": 0.8548935055732727,
+      "learning_rate": 3.002942304276991e-06,
+      "loss": 0.111,
+      "step": 8378
+    },
+    {
+      "epoch": 1.3575826312378483,
+      "grad_norm": 0.8992919325828552,
+      "learning_rate": 3.0025139332010976e-06,
+      "loss": 0.0987,
+      "step": 8379
+    },
+    {
+      "epoch": 1.357744653272845,
+      "grad_norm": 0.8693149089813232,
+      "learning_rate": 3.002085546749495e-06,
+      "loss": 0.1025,
+      "step": 8380
+    },
+    {
+      "epoch": 1.357906675307842,
+      "grad_norm": 0.8654988408088684,
+      "learning_rate": 3.0016571449352882e-06,
+      "loss": 0.0976,
+      "step": 8381
+    },
+    {
+      "epoch": 1.3580686973428386,
+      "grad_norm": 0.8404430747032166,
+      "learning_rate": 3.001228727771588e-06,
+      "loss": 0.1063,
+      "step": 8382
+    },
+    {
+      "epoch": 1.3582307193778353,
+      "grad_norm": 0.9719839692115784,
+      "learning_rate": 3.0008002952715008e-06,
+      "loss": 0.121,
+      "step": 8383
+    },
+    {
+      "epoch": 1.3583927414128323,
+      "grad_norm": 0.9208077788352966,
+      "learning_rate": 3.000371847448137e-06,
+      "loss": 0.11,
+      "step": 8384
+    },
+    {
+      "epoch": 1.3585547634478288,
+      "grad_norm": 0.7984682321548462,
+      "learning_rate": 2.9999433843146055e-06,
+      "loss": 0.0949,
+      "step": 8385
+    },
+    {
+      "epoch": 1.3587167854828257,
+      "grad_norm": 0.8321592807769775,
+      "learning_rate": 2.9995149058840157e-06,
+      "loss": 0.1012,
+      "step": 8386
+    },
+    {
+      "epoch": 1.3588788075178224,
+      "grad_norm": 0.8269400596618652,
+      "learning_rate": 2.9990864121694795e-06,
+      "loss": 0.1062,
+      "step": 8387
+    },
+    {
+      "epoch": 1.3590408295528191,
+      "grad_norm": 0.862639844417572,
+      "learning_rate": 2.998657903184107e-06,
+      "loss": 0.0967,
+      "step": 8388
+    },
+    {
+      "epoch": 1.359202851587816,
+      "grad_norm": 0.8576428890228271,
+      "learning_rate": 2.9982293789410083e-06,
+      "loss": 0.0987,
+      "step": 8389
+    },
+    {
+      "epoch": 1.3593648736228128,
+      "grad_norm": 0.9003967642784119,
+      "learning_rate": 2.9978008394532966e-06,
+      "loss": 0.1054,
+      "step": 8390
+    },
+    {
+      "epoch": 1.3595268956578095,
+      "grad_norm": 0.8770749568939209,
+      "learning_rate": 2.997372284734084e-06,
+      "loss": 0.1018,
+      "step": 8391
+    },
+    {
+      "epoch": 1.3596889176928062,
+      "grad_norm": 0.8922051191329956,
+      "learning_rate": 2.996943714796483e-06,
+      "loss": 0.1078,
+      "step": 8392
+    },
+    {
+      "epoch": 1.359850939727803,
+      "grad_norm": 0.8432496190071106,
+      "learning_rate": 2.9965151296536076e-06,
+      "loss": 0.1048,
+      "step": 8393
+    },
+    {
+      "epoch": 1.3600129617627998,
+      "grad_norm": 0.8538945913314819,
+      "learning_rate": 2.9960865293185697e-06,
+      "loss": 0.1141,
+      "step": 8394
+    },
+    {
+      "epoch": 1.3601749837977966,
+      "grad_norm": 0.9138311147689819,
+      "learning_rate": 2.9956579138044857e-06,
+      "loss": 0.1115,
+      "step": 8395
+    },
+    {
+      "epoch": 1.3603370058327933,
+      "grad_norm": 0.8055441379547119,
+      "learning_rate": 2.995229283124468e-06,
+      "loss": 0.0974,
+      "step": 8396
+    },
+    {
+      "epoch": 1.36049902786779,
+      "grad_norm": 0.7506464719772339,
+      "learning_rate": 2.9948006372916332e-06,
+      "loss": 0.0969,
+      "step": 8397
+    },
+    {
+      "epoch": 1.3606610499027867,
+      "grad_norm": 0.7606310248374939,
+      "learning_rate": 2.994371976319096e-06,
+      "loss": 0.1064,
+      "step": 8398
+    },
+    {
+      "epoch": 1.3608230719377836,
+      "grad_norm": 0.8419187068939209,
+      "learning_rate": 2.993943300219973e-06,
+      "loss": 0.1104,
+      "step": 8399
+    },
+    {
+      "epoch": 1.3609850939727803,
+      "grad_norm": 0.7976471781730652,
+      "learning_rate": 2.993514609007381e-06,
+      "loss": 0.0947,
+      "step": 8400
+    },
+    {
+      "epoch": 1.361147116007777,
+      "grad_norm": 0.935541033744812,
+      "learning_rate": 2.993085902694434e-06,
+      "loss": 0.1214,
+      "step": 8401
+    },
+    {
+      "epoch": 1.3613091380427738,
+      "grad_norm": 1.0250697135925293,
+      "learning_rate": 2.992657181294254e-06,
+      "loss": 0.1132,
+      "step": 8402
+    },
+    {
+      "epoch": 1.3614711600777705,
+      "grad_norm": 0.7330549955368042,
+      "learning_rate": 2.9922284448199548e-06,
+      "loss": 0.0954,
+      "step": 8403
+    },
+    {
+      "epoch": 1.3616331821127674,
+      "grad_norm": 0.7346470355987549,
+      "learning_rate": 2.9917996932846572e-06,
+      "loss": 0.0933,
+      "step": 8404
+    },
+    {
+      "epoch": 1.3617952041477641,
+      "grad_norm": 0.8301489353179932,
+      "learning_rate": 2.991370926701479e-06,
+      "loss": 0.0975,
+      "step": 8405
+    },
+    {
+      "epoch": 1.3619572261827608,
+      "grad_norm": 0.8129703402519226,
+      "learning_rate": 2.99094214508354e-06,
+      "loss": 0.0998,
+      "step": 8406
+    },
+    {
+      "epoch": 1.3621192482177575,
+      "grad_norm": 0.9655615091323853,
+      "learning_rate": 2.9905133484439585e-06,
+      "loss": 0.1182,
+      "step": 8407
+    },
+    {
+      "epoch": 1.3622812702527543,
+      "grad_norm": 0.8164478540420532,
+      "learning_rate": 2.990084536795856e-06,
+      "loss": 0.1012,
+      "step": 8408
+    },
+    {
+      "epoch": 1.3624432922877512,
+      "grad_norm": 0.7988752722740173,
+      "learning_rate": 2.989655710152353e-06,
+      "loss": 0.095,
+      "step": 8409
+    },
+    {
+      "epoch": 1.362605314322748,
+      "grad_norm": 0.7468502521514893,
+      "learning_rate": 2.989226868526569e-06,
+      "loss": 0.09,
+      "step": 8410
+    },
+    {
+      "epoch": 1.3627673363577446,
+      "grad_norm": 0.816019594669342,
+      "learning_rate": 2.9887980119316284e-06,
+      "loss": 0.0967,
+      "step": 8411
+    },
+    {
+      "epoch": 1.3629293583927415,
+      "grad_norm": 0.7134393453598022,
+      "learning_rate": 2.98836914038065e-06,
+      "loss": 0.093,
+      "step": 8412
+    },
+    {
+      "epoch": 1.3630913804277383,
+      "grad_norm": 0.8280972242355347,
+      "learning_rate": 2.9879402538867584e-06,
+      "loss": 0.0959,
+      "step": 8413
+    },
+    {
+      "epoch": 1.363253402462735,
+      "grad_norm": 0.9556366801261902,
+      "learning_rate": 2.987511352463076e-06,
+      "loss": 0.1118,
+      "step": 8414
+    },
+    {
+      "epoch": 1.3634154244977317,
+      "grad_norm": 0.9054288268089294,
+      "learning_rate": 2.9870824361227257e-06,
+      "loss": 0.1048,
+      "step": 8415
+    },
+    {
+      "epoch": 1.3635774465327284,
+      "grad_norm": 0.9046202898025513,
+      "learning_rate": 2.9866535048788314e-06,
+      "loss": 0.1089,
+      "step": 8416
+    },
+    {
+      "epoch": 1.3637394685677253,
+      "grad_norm": 0.8369488716125488,
+      "learning_rate": 2.986224558744519e-06,
+      "loss": 0.1042,
+      "step": 8417
+    },
+    {
+      "epoch": 1.363901490602722,
+      "grad_norm": 0.710746705532074,
+      "learning_rate": 2.9857955977329095e-06,
+      "loss": 0.0917,
+      "step": 8418
+    },
+    {
+      "epoch": 1.3640635126377187,
+      "grad_norm": 0.8769986629486084,
+      "learning_rate": 2.985366621857132e-06,
+      "loss": 0.0997,
+      "step": 8419
+    },
+    {
+      "epoch": 1.3642255346727155,
+      "grad_norm": 0.8238828182220459,
+      "learning_rate": 2.9849376311303095e-06,
+      "loss": 0.1029,
+      "step": 8420
+    },
+    {
+      "epoch": 1.3643875567077122,
+      "grad_norm": 0.7568957209587097,
+      "learning_rate": 2.9845086255655692e-06,
+      "loss": 0.0914,
+      "step": 8421
+    },
+    {
+      "epoch": 1.364549578742709,
+      "grad_norm": 0.7604775428771973,
+      "learning_rate": 2.984079605176038e-06,
+      "loss": 0.1007,
+      "step": 8422
+    },
+    {
+      "epoch": 1.3647116007777058,
+      "grad_norm": 0.9038381576538086,
+      "learning_rate": 2.9836505699748414e-06,
+      "loss": 0.1094,
+      "step": 8423
+    },
+    {
+      "epoch": 1.3648736228127025,
+      "grad_norm": 0.9362145066261292,
+      "learning_rate": 2.9832215199751085e-06,
+      "loss": 0.1154,
+      "step": 8424
+    },
+    {
+      "epoch": 1.3650356448476992,
+      "grad_norm": 0.8526574969291687,
+      "learning_rate": 2.9827924551899657e-06,
+      "loss": 0.1054,
+      "step": 8425
+    },
+    {
+      "epoch": 1.365197666882696,
+      "grad_norm": 0.8114335536956787,
+      "learning_rate": 2.9823633756325433e-06,
+      "loss": 0.1028,
+      "step": 8426
+    },
+    {
+      "epoch": 1.3653596889176929,
+      "grad_norm": 0.814687967300415,
+      "learning_rate": 2.9819342813159674e-06,
+      "loss": 0.0995,
+      "step": 8427
+    },
+    {
+      "epoch": 1.3655217109526896,
+      "grad_norm": 0.7649007439613342,
+      "learning_rate": 2.9815051722533707e-06,
+      "loss": 0.0909,
+      "step": 8428
+    },
+    {
+      "epoch": 1.3656837329876863,
+      "grad_norm": 0.9378827214241028,
+      "learning_rate": 2.9810760484578794e-06,
+      "loss": 0.1072,
+      "step": 8429
+    },
+    {
+      "epoch": 1.365845755022683,
+      "grad_norm": 0.841758668422699,
+      "learning_rate": 2.9806469099426254e-06,
+      "loss": 0.1043,
+      "step": 8430
+    },
+    {
+      "epoch": 1.3660077770576797,
+      "grad_norm": 0.7740828990936279,
+      "learning_rate": 2.98021775672074e-06,
+      "loss": 0.0929,
+      "step": 8431
+    },
+    {
+      "epoch": 1.3661697990926767,
+      "grad_norm": 0.7625635266304016,
+      "learning_rate": 2.9797885888053517e-06,
+      "loss": 0.0944,
+      "step": 8432
+    },
+    {
+      "epoch": 1.3663318211276734,
+      "grad_norm": 0.8351330757141113,
+      "learning_rate": 2.9793594062095955e-06,
+      "loss": 0.1061,
+      "step": 8433
+    },
+    {
+      "epoch": 1.36649384316267,
+      "grad_norm": 0.791860044002533,
+      "learning_rate": 2.9789302089466e-06,
+      "loss": 0.0981,
+      "step": 8434
+    },
+    {
+      "epoch": 1.366655865197667,
+      "grad_norm": 0.8011177778244019,
+      "learning_rate": 2.9785009970294997e-06,
+      "loss": 0.1084,
+      "step": 8435
+    },
+    {
+      "epoch": 1.3668178872326635,
+      "grad_norm": 0.9865897297859192,
+      "learning_rate": 2.978071770471427e-06,
+      "loss": 0.1206,
+      "step": 8436
+    },
+    {
+      "epoch": 1.3669799092676604,
+      "grad_norm": 0.9098086357116699,
+      "learning_rate": 2.977642529285515e-06,
+      "loss": 0.1129,
+      "step": 8437
+    },
+    {
+      "epoch": 1.3671419313026572,
+      "grad_norm": 0.9035070538520813,
+      "learning_rate": 2.9772132734848974e-06,
+      "loss": 0.111,
+      "step": 8438
+    },
+    {
+      "epoch": 1.3673039533376539,
+      "grad_norm": 0.8441570401191711,
+      "learning_rate": 2.9767840030827082e-06,
+      "loss": 0.1142,
+      "step": 8439
+    },
+    {
+      "epoch": 1.3674659753726508,
+      "grad_norm": 0.857961118221283,
+      "learning_rate": 2.9763547180920825e-06,
+      "loss": 0.108,
+      "step": 8440
+    },
+    {
+      "epoch": 1.3676279974076475,
+      "grad_norm": 0.7608411312103271,
+      "learning_rate": 2.9759254185261555e-06,
+      "loss": 0.0939,
+      "step": 8441
+    },
+    {
+      "epoch": 1.3677900194426442,
+      "grad_norm": 0.8588743209838867,
+      "learning_rate": 2.9754961043980623e-06,
+      "loss": 0.1139,
+      "step": 8442
+    },
+    {
+      "epoch": 1.367952041477641,
+      "grad_norm": 0.771369993686676,
+      "learning_rate": 2.9750667757209385e-06,
+      "loss": 0.1037,
+      "step": 8443
+    },
+    {
+      "epoch": 1.3681140635126376,
+      "grad_norm": 0.8773581385612488,
+      "learning_rate": 2.9746374325079213e-06,
+      "loss": 0.1133,
+      "step": 8444
+    },
+    {
+      "epoch": 1.3682760855476346,
+      "grad_norm": 0.798323392868042,
+      "learning_rate": 2.9742080747721473e-06,
+      "loss": 0.1001,
+      "step": 8445
+    },
+    {
+      "epoch": 1.3684381075826313,
+      "grad_norm": 0.826897144317627,
+      "learning_rate": 2.973778702526754e-06,
+      "loss": 0.1104,
+      "step": 8446
+    },
+    {
+      "epoch": 1.368600129617628,
+      "grad_norm": 0.8716105222702026,
+      "learning_rate": 2.973349315784878e-06,
+      "loss": 0.1109,
+      "step": 8447
+    },
+    {
+      "epoch": 1.3687621516526247,
+      "grad_norm": 0.9297152757644653,
+      "learning_rate": 2.97291991455966e-06,
+      "loss": 0.1176,
+      "step": 8448
+    },
+    {
+      "epoch": 1.3689241736876214,
+      "grad_norm": 0.8426374197006226,
+      "learning_rate": 2.9724904988642357e-06,
+      "loss": 0.0943,
+      "step": 8449
+    },
+    {
+      "epoch": 1.3690861957226184,
+      "grad_norm": 0.7848527431488037,
+      "learning_rate": 2.9720610687117462e-06,
+      "loss": 0.1075,
+      "step": 8450
+    },
+    {
+      "epoch": 1.369248217757615,
+      "grad_norm": 0.822970986366272,
+      "learning_rate": 2.9716316241153303e-06,
+      "loss": 0.1023,
+      "step": 8451
+    },
+    {
+      "epoch": 1.3694102397926118,
+      "grad_norm": 0.8297960162162781,
+      "learning_rate": 2.971202165088128e-06,
+      "loss": 0.1,
+      "step": 8452
+    },
+    {
+      "epoch": 1.3695722618276085,
+      "grad_norm": 0.799751341342926,
+      "learning_rate": 2.9707726916432793e-06,
+      "loss": 0.1051,
+      "step": 8453
+    },
+    {
+      "epoch": 1.3697342838626052,
+      "grad_norm": 0.8356010317802429,
+      "learning_rate": 2.9703432037939255e-06,
+      "loss": 0.1089,
+      "step": 8454
+    },
+    {
+      "epoch": 1.3698963058976021,
+      "grad_norm": 0.8152604699134827,
+      "learning_rate": 2.969913701553209e-06,
+      "loss": 0.1082,
+      "step": 8455
+    },
+    {
+      "epoch": 1.3700583279325989,
+      "grad_norm": 0.8205832839012146,
+      "learning_rate": 2.9694841849342688e-06,
+      "loss": 0.1016,
+      "step": 8456
+    },
+    {
+      "epoch": 1.3702203499675956,
+      "grad_norm": 0.9030212163925171,
+      "learning_rate": 2.9690546539502496e-06,
+      "loss": 0.1063,
+      "step": 8457
+    },
+    {
+      "epoch": 1.3703823720025923,
+      "grad_norm": 0.8552389144897461,
+      "learning_rate": 2.9686251086142927e-06,
+      "loss": 0.1158,
+      "step": 8458
+    },
+    {
+      "epoch": 1.370544394037589,
+      "grad_norm": 0.8218772411346436,
+      "learning_rate": 2.968195548939542e-06,
+      "loss": 0.1007,
+      "step": 8459
+    },
+    {
+      "epoch": 1.370706416072586,
+      "grad_norm": 0.8865529894828796,
+      "learning_rate": 2.9677659749391404e-06,
+      "loss": 0.1133,
+      "step": 8460
+    },
+    {
+      "epoch": 1.3708684381075826,
+      "grad_norm": 0.8135964274406433,
+      "learning_rate": 2.967336386626232e-06,
+      "loss": 0.0994,
+      "step": 8461
+    },
+    {
+      "epoch": 1.3710304601425793,
+      "grad_norm": 0.7750198841094971,
+      "learning_rate": 2.9669067840139603e-06,
+      "loss": 0.1008,
+      "step": 8462
+    },
+    {
+      "epoch": 1.3711924821775763,
+      "grad_norm": 0.7373397946357727,
+      "learning_rate": 2.966477167115472e-06,
+      "loss": 0.0951,
+      "step": 8463
+    },
+    {
+      "epoch": 1.371354504212573,
+      "grad_norm": 0.8936923146247864,
+      "learning_rate": 2.9660475359439113e-06,
+      "loss": 0.1062,
+      "step": 8464
+    },
+    {
+      "epoch": 1.3715165262475697,
+      "grad_norm": 0.8021456003189087,
+      "learning_rate": 2.9656178905124222e-06,
+      "loss": 0.102,
+      "step": 8465
+    },
+    {
+      "epoch": 1.3716785482825664,
+      "grad_norm": 0.921617865562439,
+      "learning_rate": 2.965188230834154e-06,
+      "loss": 0.1193,
+      "step": 8466
+    },
+    {
+      "epoch": 1.3718405703175631,
+      "grad_norm": 0.8052314519882202,
+      "learning_rate": 2.9647585569222516e-06,
+      "loss": 0.0948,
+      "step": 8467
+    },
+    {
+      "epoch": 1.37200259235256,
+      "grad_norm": 1.5328847169876099,
+      "learning_rate": 2.9643288687898614e-06,
+      "loss": 0.1096,
+      "step": 8468
+    },
+    {
+      "epoch": 1.3721646143875568,
+      "grad_norm": 0.905802309513092,
+      "learning_rate": 2.9638991664501314e-06,
+      "loss": 0.1174,
+      "step": 8469
+    },
+    {
+      "epoch": 1.3723266364225535,
+      "grad_norm": 0.80458664894104,
+      "learning_rate": 2.96346944991621e-06,
+      "loss": 0.0972,
+      "step": 8470
+    },
+    {
+      "epoch": 1.3724886584575502,
+      "grad_norm": 0.8991355299949646,
+      "learning_rate": 2.9630397192012445e-06,
+      "loss": 0.1096,
+      "step": 8471
+    },
+    {
+      "epoch": 1.372650680492547,
+      "grad_norm": 1.1109977960586548,
+      "learning_rate": 2.962609974318385e-06,
+      "loss": 0.1276,
+      "step": 8472
+    },
+    {
+      "epoch": 1.3728127025275438,
+      "grad_norm": 0.8459230065345764,
+      "learning_rate": 2.962180215280779e-06,
+      "loss": 0.107,
+      "step": 8473
+    },
+    {
+      "epoch": 1.3729747245625405,
+      "grad_norm": 0.8833880424499512,
+      "learning_rate": 2.961750442101577e-06,
+      "loss": 0.106,
+      "step": 8474
+    },
+    {
+      "epoch": 1.3731367465975373,
+      "grad_norm": 0.8358442783355713,
+      "learning_rate": 2.9613206547939287e-06,
+      "loss": 0.1001,
+      "step": 8475
+    },
+    {
+      "epoch": 1.373298768632534,
+      "grad_norm": 0.8312876224517822,
+      "learning_rate": 2.9608908533709852e-06,
+      "loss": 0.104,
+      "step": 8476
+    },
+    {
+      "epoch": 1.3734607906675307,
+      "grad_norm": 0.7915255427360535,
+      "learning_rate": 2.9604610378458965e-06,
+      "loss": 0.0953,
+      "step": 8477
+    },
+    {
+      "epoch": 1.3736228127025276,
+      "grad_norm": 0.7739593386650085,
+      "learning_rate": 2.9600312082318144e-06,
+      "loss": 0.1022,
+      "step": 8478
+    },
+    {
+      "epoch": 1.3737848347375243,
+      "grad_norm": 0.9240787029266357,
+      "learning_rate": 2.9596013645418913e-06,
+      "loss": 0.1201,
+      "step": 8479
+    },
+    {
+      "epoch": 1.373946856772521,
+      "grad_norm": 0.7890738248825073,
+      "learning_rate": 2.9591715067892777e-06,
+      "loss": 0.0993,
+      "step": 8480
+    },
+    {
+      "epoch": 1.3741088788075178,
+      "grad_norm": 0.7773252725601196,
+      "learning_rate": 2.9587416349871277e-06,
+      "loss": 0.095,
+      "step": 8481
+    },
+    {
+      "epoch": 1.3742709008425145,
+      "grad_norm": 0.888715386390686,
+      "learning_rate": 2.958311749148594e-06,
+      "loss": 0.1149,
+      "step": 8482
+    },
+    {
+      "epoch": 1.3744329228775114,
+      "grad_norm": 0.8670759797096252,
+      "learning_rate": 2.9578818492868293e-06,
+      "loss": 0.1132,
+      "step": 8483
+    },
+    {
+      "epoch": 1.374594944912508,
+      "grad_norm": 0.8122788667678833,
+      "learning_rate": 2.9574519354149884e-06,
+      "loss": 0.0999,
+      "step": 8484
+    },
+    {
+      "epoch": 1.3747569669475048,
+      "grad_norm": 0.9208338260650635,
+      "learning_rate": 2.9570220075462254e-06,
+      "loss": 0.1139,
+      "step": 8485
+    },
+    {
+      "epoch": 1.3749189889825018,
+      "grad_norm": 0.7937371134757996,
+      "learning_rate": 2.9565920656936947e-06,
+      "loss": 0.1054,
+      "step": 8486
+    },
+    {
+      "epoch": 1.3750810110174982,
+      "grad_norm": 0.9310660362243652,
+      "learning_rate": 2.956162109870551e-06,
+      "loss": 0.1085,
+      "step": 8487
+    },
+    {
+      "epoch": 1.3752430330524952,
+      "grad_norm": 0.8265377283096313,
+      "learning_rate": 2.9557321400899524e-06,
+      "loss": 0.1031,
+      "step": 8488
+    },
+    {
+      "epoch": 1.375405055087492,
+      "grad_norm": 0.7675789594650269,
+      "learning_rate": 2.9553021563650514e-06,
+      "loss": 0.0969,
+      "step": 8489
+    },
+    {
+      "epoch": 1.3755670771224886,
+      "grad_norm": 0.8487404584884644,
+      "learning_rate": 2.9548721587090075e-06,
+      "loss": 0.1223,
+      "step": 8490
+    },
+    {
+      "epoch": 1.3757290991574855,
+      "grad_norm": 0.8002118468284607,
+      "learning_rate": 2.9544421471349753e-06,
+      "loss": 0.0937,
+      "step": 8491
+    },
+    {
+      "epoch": 1.3758911211924822,
+      "grad_norm": 0.7949121594429016,
+      "learning_rate": 2.954012121656114e-06,
+      "loss": 0.0965,
+      "step": 8492
+    },
+    {
+      "epoch": 1.376053143227479,
+      "grad_norm": 0.8120603561401367,
+      "learning_rate": 2.9535820822855797e-06,
+      "loss": 0.0894,
+      "step": 8493
+    },
+    {
+      "epoch": 1.3762151652624757,
+      "grad_norm": 0.7125270366668701,
+      "learning_rate": 2.9531520290365316e-06,
+      "loss": 0.0855,
+      "step": 8494
+    },
+    {
+      "epoch": 1.3763771872974724,
+      "grad_norm": 0.819965124130249,
+      "learning_rate": 2.9527219619221293e-06,
+      "loss": 0.0934,
+      "step": 8495
+    },
+    {
+      "epoch": 1.3765392093324693,
+      "grad_norm": 0.8064921498298645,
+      "learning_rate": 2.952291880955529e-06,
+      "loss": 0.0961,
+      "step": 8496
+    },
+    {
+      "epoch": 1.376701231367466,
+      "grad_norm": 0.7966877818107605,
+      "learning_rate": 2.9518617861498924e-06,
+      "loss": 0.0986,
+      "step": 8497
+    },
+    {
+      "epoch": 1.3768632534024627,
+      "grad_norm": 0.8604177236557007,
+      "learning_rate": 2.9514316775183777e-06,
+      "loss": 0.1135,
+      "step": 8498
+    },
+    {
+      "epoch": 1.3770252754374595,
+      "grad_norm": 0.8813721537590027,
+      "learning_rate": 2.9510015550741467e-06,
+      "loss": 0.108,
+      "step": 8499
+    },
+    {
+      "epoch": 1.3771872974724562,
+      "grad_norm": 0.8800192475318909,
+      "learning_rate": 2.950571418830359e-06,
+      "loss": 0.1143,
+      "step": 8500
+    },
+    {
+      "epoch": 1.377349319507453,
+      "grad_norm": 0.8808828592300415,
+      "learning_rate": 2.950141268800177e-06,
+      "loss": 0.1095,
+      "step": 8501
+    },
+    {
+      "epoch": 1.3775113415424498,
+      "grad_norm": 0.7128430008888245,
+      "learning_rate": 2.949711104996761e-06,
+      "loss": 0.0887,
+      "step": 8502
+    },
+    {
+      "epoch": 1.3776733635774465,
+      "grad_norm": 0.7968690395355225,
+      "learning_rate": 2.9492809274332745e-06,
+      "loss": 0.0957,
+      "step": 8503
+    },
+    {
+      "epoch": 1.3778353856124432,
+      "grad_norm": 0.8926577568054199,
+      "learning_rate": 2.948850736122878e-06,
+      "loss": 0.1194,
+      "step": 8504
+    },
+    {
+      "epoch": 1.37799740764744,
+      "grad_norm": 0.9188405871391296,
+      "learning_rate": 2.948420531078735e-06,
+      "loss": 0.1107,
+      "step": 8505
+    },
+    {
+      "epoch": 1.3781594296824369,
+      "grad_norm": 0.7922679781913757,
+      "learning_rate": 2.947990312314009e-06,
+      "loss": 0.0989,
+      "step": 8506
+    },
+    {
+      "epoch": 1.3783214517174336,
+      "grad_norm": 0.9483537077903748,
+      "learning_rate": 2.9475600798418636e-06,
+      "loss": 0.1224,
+      "step": 8507
+    },
+    {
+      "epoch": 1.3784834737524303,
+      "grad_norm": 0.9121973514556885,
+      "learning_rate": 2.9471298336754633e-06,
+      "loss": 0.1141,
+      "step": 8508
+    },
+    {
+      "epoch": 1.378645495787427,
+      "grad_norm": 0.8246192932128906,
+      "learning_rate": 2.9466995738279715e-06,
+      "loss": 0.1031,
+      "step": 8509
+    },
+    {
+      "epoch": 1.3788075178224237,
+      "grad_norm": 0.8705335259437561,
+      "learning_rate": 2.9462693003125544e-06,
+      "loss": 0.1033,
+      "step": 8510
+    },
+    {
+      "epoch": 1.3789695398574207,
+      "grad_norm": 0.7904585003852844,
+      "learning_rate": 2.9458390131423754e-06,
+      "loss": 0.0992,
+      "step": 8511
+    },
+    {
+      "epoch": 1.3791315618924174,
+      "grad_norm": 0.8808677196502686,
+      "learning_rate": 2.945408712330603e-06,
+      "loss": 0.1059,
+      "step": 8512
+    },
+    {
+      "epoch": 1.379293583927414,
+      "grad_norm": 1.1480685472488403,
+      "learning_rate": 2.944978397890401e-06,
+      "loss": 0.1,
+      "step": 8513
+    },
+    {
+      "epoch": 1.379455605962411,
+      "grad_norm": 0.8541058301925659,
+      "learning_rate": 2.944548069834937e-06,
+      "loss": 0.1072,
+      "step": 8514
+    },
+    {
+      "epoch": 1.3796176279974077,
+      "grad_norm": 0.7569074630737305,
+      "learning_rate": 2.9441177281773783e-06,
+      "loss": 0.1054,
+      "step": 8515
+    },
+    {
+      "epoch": 1.3797796500324044,
+      "grad_norm": 0.8691943287849426,
+      "learning_rate": 2.943687372930891e-06,
+      "loss": 0.1116,
+      "step": 8516
+    },
+    {
+      "epoch": 1.3799416720674011,
+      "grad_norm": 0.8004394173622131,
+      "learning_rate": 2.943257004108645e-06,
+      "loss": 0.1001,
+      "step": 8517
+    },
+    {
+      "epoch": 1.3801036941023979,
+      "grad_norm": 0.8240222930908203,
+      "learning_rate": 2.942826621723806e-06,
+      "loss": 0.1008,
+      "step": 8518
+    },
+    {
+      "epoch": 1.3802657161373948,
+      "grad_norm": 0.8319725394248962,
+      "learning_rate": 2.942396225789545e-06,
+      "loss": 0.1021,
+      "step": 8519
+    },
+    {
+      "epoch": 1.3804277381723915,
+      "grad_norm": 0.8528062701225281,
+      "learning_rate": 2.9419658163190295e-06,
+      "loss": 0.1136,
+      "step": 8520
+    },
+    {
+      "epoch": 1.3805897602073882,
+      "grad_norm": 0.7287147641181946,
+      "learning_rate": 2.9415353933254297e-06,
+      "loss": 0.0952,
+      "step": 8521
+    },
+    {
+      "epoch": 1.380751782242385,
+      "grad_norm": 0.8255961537361145,
+      "learning_rate": 2.9411049568219153e-06,
+      "loss": 0.1072,
+      "step": 8522
+    },
+    {
+      "epoch": 1.3809138042773816,
+      "grad_norm": 0.7438040971755981,
+      "learning_rate": 2.940674506821657e-06,
+      "loss": 0.0972,
+      "step": 8523
+    },
+    {
+      "epoch": 1.3810758263123786,
+      "grad_norm": 0.7560163736343384,
+      "learning_rate": 2.9402440433378247e-06,
+      "loss": 0.0906,
+      "step": 8524
+    },
+    {
+      "epoch": 1.3812378483473753,
+      "grad_norm": 0.952171802520752,
+      "learning_rate": 2.9398135663835904e-06,
+      "loss": 0.1007,
+      "step": 8525
+    },
+    {
+      "epoch": 1.381399870382372,
+      "grad_norm": 0.7956331968307495,
+      "learning_rate": 2.939383075972125e-06,
+      "loss": 0.1024,
+      "step": 8526
+    },
+    {
+      "epoch": 1.3815618924173687,
+      "grad_norm": 0.8440989255905151,
+      "learning_rate": 2.9389525721166013e-06,
+      "loss": 0.1028,
+      "step": 8527
+    },
+    {
+      "epoch": 1.3817239144523654,
+      "grad_norm": 1.5592060089111328,
+      "learning_rate": 2.9385220548301906e-06,
+      "loss": 0.1023,
+      "step": 8528
+    },
+    {
+      "epoch": 1.3818859364873624,
+      "grad_norm": 0.8761277794837952,
+      "learning_rate": 2.9380915241260665e-06,
+      "loss": 0.0997,
+      "step": 8529
+    },
+    {
+      "epoch": 1.382047958522359,
+      "grad_norm": 0.8244863748550415,
+      "learning_rate": 2.937660980017402e-06,
+      "loss": 0.1096,
+      "step": 8530
+    },
+    {
+      "epoch": 1.3822099805573558,
+      "grad_norm": 0.7663264870643616,
+      "learning_rate": 2.9372304225173703e-06,
+      "loss": 0.0926,
+      "step": 8531
+    },
+    {
+      "epoch": 1.3823720025923525,
+      "grad_norm": 0.8507534861564636,
+      "learning_rate": 2.936799851639146e-06,
+      "loss": 0.1017,
+      "step": 8532
+    },
+    {
+      "epoch": 1.3825340246273492,
+      "grad_norm": 0.9161567091941833,
+      "learning_rate": 2.936369267395903e-06,
+      "loss": 0.1143,
+      "step": 8533
+    },
+    {
+      "epoch": 1.3826960466623461,
+      "grad_norm": 0.8694104552268982,
+      "learning_rate": 2.9359386698008172e-06,
+      "loss": 0.1074,
+      "step": 8534
+    },
+    {
+      "epoch": 1.3828580686973428,
+      "grad_norm": 0.853909969329834,
+      "learning_rate": 2.9355080588670626e-06,
+      "loss": 0.1031,
+      "step": 8535
+    },
+    {
+      "epoch": 1.3830200907323396,
+      "grad_norm": 0.7998418211936951,
+      "learning_rate": 2.935077434607815e-06,
+      "loss": 0.1022,
+      "step": 8536
+    },
+    {
+      "epoch": 1.3831821127673365,
+      "grad_norm": 0.9292730689048767,
+      "learning_rate": 2.934646797036251e-06,
+      "loss": 0.1037,
+      "step": 8537
+    },
+    {
+      "epoch": 1.383344134802333,
+      "grad_norm": 0.7879573106765747,
+      "learning_rate": 2.9342161461655468e-06,
+      "loss": 0.095,
+      "step": 8538
+    },
+    {
+      "epoch": 1.38350615683733,
+      "grad_norm": 0.9311516284942627,
+      "learning_rate": 2.9337854820088797e-06,
+      "loss": 0.1086,
+      "step": 8539
+    },
+    {
+      "epoch": 1.3836681788723266,
+      "grad_norm": 0.8644829392433167,
+      "learning_rate": 2.9333548045794253e-06,
+      "loss": 0.1038,
+      "step": 8540
+    },
+    {
+      "epoch": 1.3838302009073233,
+      "grad_norm": 0.8739936947822571,
+      "learning_rate": 2.9329241138903642e-06,
+      "loss": 0.1093,
+      "step": 8541
+    },
+    {
+      "epoch": 1.3839922229423203,
+      "grad_norm": 0.7543818354606628,
+      "learning_rate": 2.9324934099548713e-06,
+      "loss": 0.0904,
+      "step": 8542
+    },
+    {
+      "epoch": 1.384154244977317,
+      "grad_norm": 0.7458750009536743,
+      "learning_rate": 2.9320626927861283e-06,
+      "loss": 0.0913,
+      "step": 8543
+    },
+    {
+      "epoch": 1.3843162670123137,
+      "grad_norm": 0.8349683284759521,
+      "learning_rate": 2.931631962397311e-06,
+      "loss": 0.1058,
+      "step": 8544
+    },
+    {
+      "epoch": 1.3844782890473104,
+      "grad_norm": 1.0070006847381592,
+      "learning_rate": 2.9312012188016014e-06,
+      "loss": 0.1281,
+      "step": 8545
+    },
+    {
+      "epoch": 1.3846403110823071,
+      "grad_norm": 0.8791117072105408,
+      "learning_rate": 2.9307704620121775e-06,
+      "loss": 0.1125,
+      "step": 8546
+    },
+    {
+      "epoch": 1.384802333117304,
+      "grad_norm": 0.8105389475822449,
+      "learning_rate": 2.9303396920422196e-06,
+      "loss": 0.1058,
+      "step": 8547
+    },
+    {
+      "epoch": 1.3849643551523008,
+      "grad_norm": 0.8246057629585266,
+      "learning_rate": 2.9299089089049092e-06,
+      "loss": 0.1031,
+      "step": 8548
+    },
+    {
+      "epoch": 1.3851263771872975,
+      "grad_norm": 0.9281556606292725,
+      "learning_rate": 2.9294781126134254e-06,
+      "loss": 0.1114,
+      "step": 8549
+    },
+    {
+      "epoch": 1.3852883992222942,
+      "grad_norm": 0.885179340839386,
+      "learning_rate": 2.929047303180952e-06,
+      "loss": 0.1128,
+      "step": 8550
+    },
+    {
+      "epoch": 1.385450421257291,
+      "grad_norm": 0.7432131767272949,
+      "learning_rate": 2.9286164806206683e-06,
+      "loss": 0.0927,
+      "step": 8551
+    },
+    {
+      "epoch": 1.3856124432922878,
+      "grad_norm": 0.7628340721130371,
+      "learning_rate": 2.9281856449457587e-06,
+      "loss": 0.1045,
+      "step": 8552
+    },
+    {
+      "epoch": 1.3857744653272845,
+      "grad_norm": 0.8845008015632629,
+      "learning_rate": 2.9277547961694037e-06,
+      "loss": 0.1161,
+      "step": 8553
+    },
+    {
+      "epoch": 1.3859364873622813,
+      "grad_norm": 0.7263888120651245,
+      "learning_rate": 2.927323934304787e-06,
+      "loss": 0.095,
+      "step": 8554
+    },
+    {
+      "epoch": 1.386098509397278,
+      "grad_norm": 0.9359971284866333,
+      "learning_rate": 2.9268930593650926e-06,
+      "loss": 0.1342,
+      "step": 8555
+    },
+    {
+      "epoch": 1.3862605314322747,
+      "grad_norm": 0.8538017272949219,
+      "learning_rate": 2.926462171363503e-06,
+      "loss": 0.1154,
+      "step": 8556
+    },
+    {
+      "epoch": 1.3864225534672716,
+      "grad_norm": 0.8223465085029602,
+      "learning_rate": 2.9260312703132037e-06,
+      "loss": 0.1037,
+      "step": 8557
+    },
+    {
+      "epoch": 1.3865845755022683,
+      "grad_norm": 0.7046705484390259,
+      "learning_rate": 2.9256003562273784e-06,
+      "loss": 0.0974,
+      "step": 8558
+    },
+    {
+      "epoch": 1.386746597537265,
+      "grad_norm": 0.8450808525085449,
+      "learning_rate": 2.9251694291192113e-06,
+      "loss": 0.1041,
+      "step": 8559
+    },
+    {
+      "epoch": 1.3869086195722617,
+      "grad_norm": 0.7554501295089722,
+      "learning_rate": 2.924738489001889e-06,
+      "loss": 0.0989,
+      "step": 8560
+    },
+    {
+      "epoch": 1.3870706416072585,
+      "grad_norm": 0.7961752414703369,
+      "learning_rate": 2.924307535888597e-06,
+      "loss": 0.0983,
+      "step": 8561
+    },
+    {
+      "epoch": 1.3872326636422554,
+      "grad_norm": 0.7975520491600037,
+      "learning_rate": 2.923876569792521e-06,
+      "loss": 0.1012,
+      "step": 8562
+    },
+    {
+      "epoch": 1.387394685677252,
+      "grad_norm": 0.980924129486084,
+      "learning_rate": 2.923445590726848e-06,
+      "loss": 0.1208,
+      "step": 8563
+    },
+    {
+      "epoch": 1.3875567077122488,
+      "grad_norm": 0.92117840051651,
+      "learning_rate": 2.923014598704764e-06,
+      "loss": 0.1097,
+      "step": 8564
+    },
+    {
+      "epoch": 1.3877187297472457,
+      "grad_norm": 0.801445484161377,
+      "learning_rate": 2.922583593739458e-06,
+      "loss": 0.103,
+      "step": 8565
+    },
+    {
+      "epoch": 1.3878807517822425,
+      "grad_norm": 0.9321786761283875,
+      "learning_rate": 2.9221525758441155e-06,
+      "loss": 0.1197,
+      "step": 8566
+    },
+    {
+      "epoch": 1.3880427738172392,
+      "grad_norm": 0.7319613099098206,
+      "learning_rate": 2.921721545031927e-06,
+      "loss": 0.0961,
+      "step": 8567
+    },
+    {
+      "epoch": 1.3882047958522359,
+      "grad_norm": 0.7948114275932312,
+      "learning_rate": 2.9212905013160784e-06,
+      "loss": 0.1095,
+      "step": 8568
+    },
+    {
+      "epoch": 1.3883668178872326,
+      "grad_norm": 0.7754814624786377,
+      "learning_rate": 2.920859444709761e-06,
+      "loss": 0.0997,
+      "step": 8569
+    },
+    {
+      "epoch": 1.3885288399222295,
+      "grad_norm": 0.9290817975997925,
+      "learning_rate": 2.920428375226163e-06,
+      "loss": 0.101,
+      "step": 8570
+    },
+    {
+      "epoch": 1.3886908619572262,
+      "grad_norm": 0.761486291885376,
+      "learning_rate": 2.919997292878474e-06,
+      "loss": 0.0954,
+      "step": 8571
+    },
+    {
+      "epoch": 1.388852883992223,
+      "grad_norm": 0.672540009021759,
+      "learning_rate": 2.9195661976798838e-06,
+      "loss": 0.0868,
+      "step": 8572
+    },
+    {
+      "epoch": 1.3890149060272197,
+      "grad_norm": 0.8108746409416199,
+      "learning_rate": 2.919135089643583e-06,
+      "loss": 0.0953,
+      "step": 8573
+    },
+    {
+      "epoch": 1.3891769280622164,
+      "grad_norm": 1.304598093032837,
+      "learning_rate": 2.918703968782764e-06,
+      "loss": 0.0915,
+      "step": 8574
+    },
+    {
+      "epoch": 1.3893389500972133,
+      "grad_norm": 0.9010568261146545,
+      "learning_rate": 2.918272835110616e-06,
+      "loss": 0.1142,
+      "step": 8575
+    },
+    {
+      "epoch": 1.38950097213221,
+      "grad_norm": 0.7341850399971008,
+      "learning_rate": 2.9178416886403318e-06,
+      "loss": 0.0943,
+      "step": 8576
+    },
+    {
+      "epoch": 1.3896629941672067,
+      "grad_norm": 0.8488278388977051,
+      "learning_rate": 2.9174105293851025e-06,
+      "loss": 0.1099,
+      "step": 8577
+    },
+    {
+      "epoch": 1.3898250162022034,
+      "grad_norm": 0.9017966389656067,
+      "learning_rate": 2.916979357358121e-06,
+      "loss": 0.1107,
+      "step": 8578
+    },
+    {
+      "epoch": 1.3899870382372002,
+      "grad_norm": 0.7202101945877075,
+      "learning_rate": 2.916548172572581e-06,
+      "loss": 0.098,
+      "step": 8579
+    },
+    {
+      "epoch": 1.390149060272197,
+      "grad_norm": 0.857900857925415,
+      "learning_rate": 2.9161169750416746e-06,
+      "loss": 0.1077,
+      "step": 8580
+    },
+    {
+      "epoch": 1.3903110823071938,
+      "grad_norm": 0.8433865904808044,
+      "learning_rate": 2.9156857647785964e-06,
+      "loss": 0.1083,
+      "step": 8581
+    },
+    {
+      "epoch": 1.3904731043421905,
+      "grad_norm": 0.8308601975440979,
+      "learning_rate": 2.915254541796539e-06,
+      "loss": 0.1088,
+      "step": 8582
+    },
+    {
+      "epoch": 1.3906351263771872,
+      "grad_norm": 0.77713942527771,
+      "learning_rate": 2.9148233061086973e-06,
+      "loss": 0.1047,
+      "step": 8583
+    },
+    {
+      "epoch": 1.390797148412184,
+      "grad_norm": 0.6983367204666138,
+      "learning_rate": 2.914392057728267e-06,
+      "loss": 0.0904,
+      "step": 8584
+    },
+    {
+      "epoch": 1.3909591704471809,
+      "grad_norm": 0.844786524772644,
+      "learning_rate": 2.913960796668442e-06,
+      "loss": 0.1129,
+      "step": 8585
+    },
+    {
+      "epoch": 1.3911211924821776,
+      "grad_norm": 0.756144106388092,
+      "learning_rate": 2.913529522942418e-06,
+      "loss": 0.1032,
+      "step": 8586
+    },
+    {
+      "epoch": 1.3912832145171743,
+      "grad_norm": 0.7644606828689575,
+      "learning_rate": 2.9130982365633926e-06,
+      "loss": 0.1047,
+      "step": 8587
+    },
+    {
+      "epoch": 1.3914452365521712,
+      "grad_norm": 0.8337986469268799,
+      "learning_rate": 2.9126669375445595e-06,
+      "loss": 0.1056,
+      "step": 8588
+    },
+    {
+      "epoch": 1.3916072585871677,
+      "grad_norm": 0.7386319041252136,
+      "learning_rate": 2.912235625899118e-06,
+      "loss": 0.1003,
+      "step": 8589
+    },
+    {
+      "epoch": 1.3917692806221647,
+      "grad_norm": 0.8051424026489258,
+      "learning_rate": 2.911804301640263e-06,
+      "loss": 0.1058,
+      "step": 8590
+    },
+    {
+      "epoch": 1.3919313026571614,
+      "grad_norm": 0.8822247982025146,
+      "learning_rate": 2.9113729647811935e-06,
+      "loss": 0.1107,
+      "step": 8591
+    },
+    {
+      "epoch": 1.392093324692158,
+      "grad_norm": 0.8278679251670837,
+      "learning_rate": 2.910941615335106e-06,
+      "loss": 0.115,
+      "step": 8592
+    },
+    {
+      "epoch": 1.392255346727155,
+      "grad_norm": 0.8865328431129456,
+      "learning_rate": 2.9105102533152e-06,
+      "loss": 0.1151,
+      "step": 8593
+    },
+    {
+      "epoch": 1.3924173687621517,
+      "grad_norm": 0.8504626750946045,
+      "learning_rate": 2.9100788787346746e-06,
+      "loss": 0.1131,
+      "step": 8594
+    },
+    {
+      "epoch": 1.3925793907971484,
+      "grad_norm": 0.8320386409759521,
+      "learning_rate": 2.9096474916067264e-06,
+      "loss": 0.1116,
+      "step": 8595
+    },
+    {
+      "epoch": 1.3927414128321451,
+      "grad_norm": 0.8168127536773682,
+      "learning_rate": 2.9092160919445566e-06,
+      "loss": 0.1058,
+      "step": 8596
+    },
+    {
+      "epoch": 1.3929034348671419,
+      "grad_norm": 0.8525434136390686,
+      "learning_rate": 2.9087846797613645e-06,
+      "loss": 0.0983,
+      "step": 8597
+    },
+    {
+      "epoch": 1.3930654569021388,
+      "grad_norm": 0.8819026947021484,
+      "learning_rate": 2.9083532550703515e-06,
+      "loss": 0.1126,
+      "step": 8598
+    },
+    {
+      "epoch": 1.3932274789371355,
+      "grad_norm": 0.7245476841926575,
+      "learning_rate": 2.907921817884716e-06,
+      "loss": 0.0919,
+      "step": 8599
+    },
+    {
+      "epoch": 1.3933895009721322,
+      "grad_norm": 0.7397482395172119,
+      "learning_rate": 2.9074903682176607e-06,
+      "loss": 0.1035,
+      "step": 8600
+    },
+    {
+      "epoch": 1.393551523007129,
+      "grad_norm": 0.7988397479057312,
+      "learning_rate": 2.907058906082386e-06,
+      "loss": 0.1043,
+      "step": 8601
+    },
+    {
+      "epoch": 1.3937135450421256,
+      "grad_norm": 0.9405267238616943,
+      "learning_rate": 2.906627431492094e-06,
+      "loss": 0.1193,
+      "step": 8602
+    },
+    {
+      "epoch": 1.3938755670771226,
+      "grad_norm": 0.766861617565155,
+      "learning_rate": 2.9061959444599867e-06,
+      "loss": 0.0918,
+      "step": 8603
+    },
+    {
+      "epoch": 1.3940375891121193,
+      "grad_norm": 0.8565654754638672,
+      "learning_rate": 2.9057644449992655e-06,
+      "loss": 0.1066,
+      "step": 8604
+    },
+    {
+      "epoch": 1.394199611147116,
+      "grad_norm": 0.8554398417472839,
+      "learning_rate": 2.9053329331231356e-06,
+      "loss": 0.1046,
+      "step": 8605
+    },
+    {
+      "epoch": 1.3943616331821127,
+      "grad_norm": 0.8772569298744202,
+      "learning_rate": 2.904901408844798e-06,
+      "loss": 0.1049,
+      "step": 8606
+    },
+    {
+      "epoch": 1.3945236552171094,
+      "grad_norm": 0.8543737530708313,
+      "learning_rate": 2.904469872177458e-06,
+      "loss": 0.1047,
+      "step": 8607
+    },
+    {
+      "epoch": 1.3946856772521063,
+      "grad_norm": 0.8698646426200867,
+      "learning_rate": 2.9040383231343173e-06,
+      "loss": 0.1063,
+      "step": 8608
+    },
+    {
+      "epoch": 1.394847699287103,
+      "grad_norm": 0.8685057163238525,
+      "learning_rate": 2.9036067617285825e-06,
+      "loss": 0.1025,
+      "step": 8609
+    },
+    {
+      "epoch": 1.3950097213220998,
+      "grad_norm": 0.7601488828659058,
+      "learning_rate": 2.903175187973457e-06,
+      "loss": 0.0914,
+      "step": 8610
+    },
+    {
+      "epoch": 1.3951717433570965,
+      "grad_norm": 0.7056039571762085,
+      "learning_rate": 2.902743601882147e-06,
+      "loss": 0.0907,
+      "step": 8611
+    },
+    {
+      "epoch": 1.3953337653920932,
+      "grad_norm": 0.9211429357528687,
+      "learning_rate": 2.9023120034678575e-06,
+      "loss": 0.1088,
+      "step": 8612
+    },
+    {
+      "epoch": 1.3954957874270901,
+      "grad_norm": 0.8788020610809326,
+      "learning_rate": 2.9018803927437946e-06,
+      "loss": 0.117,
+      "step": 8613
+    },
+    {
+      "epoch": 1.3956578094620868,
+      "grad_norm": 0.7978447675704956,
+      "learning_rate": 2.901448769723163e-06,
+      "loss": 0.1123,
+      "step": 8614
+    },
+    {
+      "epoch": 1.3958198314970836,
+      "grad_norm": 0.8556854724884033,
+      "learning_rate": 2.901017134419171e-06,
+      "loss": 0.1007,
+      "step": 8615
+    },
+    {
+      "epoch": 1.3959818535320805,
+      "grad_norm": 0.8050543069839478,
+      "learning_rate": 2.900585486845026e-06,
+      "loss": 0.0957,
+      "step": 8616
+    },
+    {
+      "epoch": 1.3961438755670772,
+      "grad_norm": 0.8602048754692078,
+      "learning_rate": 2.900153827013933e-06,
+      "loss": 0.1133,
+      "step": 8617
+    },
+    {
+      "epoch": 1.396305897602074,
+      "grad_norm": 0.8299257159233093,
+      "learning_rate": 2.8997221549391025e-06,
+      "loss": 0.112,
+      "step": 8618
+    },
+    {
+      "epoch": 1.3964679196370706,
+      "grad_norm": 0.9464213848114014,
+      "learning_rate": 2.8992904706337406e-06,
+      "loss": 0.1197,
+      "step": 8619
+    },
+    {
+      "epoch": 1.3966299416720673,
+      "grad_norm": 0.7960741519927979,
+      "learning_rate": 2.8988587741110575e-06,
+      "loss": 0.097,
+      "step": 8620
+    },
+    {
+      "epoch": 1.3967919637070643,
+      "grad_norm": 0.7059590220451355,
+      "learning_rate": 2.89842706538426e-06,
+      "loss": 0.0874,
+      "step": 8621
+    },
+    {
+      "epoch": 1.396953985742061,
+      "grad_norm": 0.7564810514450073,
+      "learning_rate": 2.8979953444665585e-06,
+      "loss": 0.0873,
+      "step": 8622
+    },
+    {
+      "epoch": 1.3971160077770577,
+      "grad_norm": 0.7459419965744019,
+      "learning_rate": 2.8975636113711637e-06,
+      "loss": 0.0916,
+      "step": 8623
+    },
+    {
+      "epoch": 1.3972780298120544,
+      "grad_norm": 0.8232574462890625,
+      "learning_rate": 2.8971318661112836e-06,
+      "loss": 0.0907,
+      "step": 8624
+    },
+    {
+      "epoch": 1.3974400518470511,
+      "grad_norm": 0.8205267190933228,
+      "learning_rate": 2.89670010870013e-06,
+      "loss": 0.1081,
+      "step": 8625
+    },
+    {
+      "epoch": 1.397602073882048,
+      "grad_norm": 0.8244900703430176,
+      "learning_rate": 2.896268339150912e-06,
+      "loss": 0.0983,
+      "step": 8626
+    },
+    {
+      "epoch": 1.3977640959170448,
+      "grad_norm": 0.7982784509658813,
+      "learning_rate": 2.8958365574768434e-06,
+      "loss": 0.0898,
+      "step": 8627
+    },
+    {
+      "epoch": 1.3979261179520415,
+      "grad_norm": 0.8250177502632141,
+      "learning_rate": 2.895404763691132e-06,
+      "loss": 0.1079,
+      "step": 8628
+    },
+    {
+      "epoch": 1.3980881399870382,
+      "grad_norm": 0.828636646270752,
+      "learning_rate": 2.8949729578069936e-06,
+      "loss": 0.105,
+      "step": 8629
+    },
+    {
+      "epoch": 1.398250162022035,
+      "grad_norm": 0.9336274266242981,
+      "learning_rate": 2.894541139837638e-06,
+      "loss": 0.1116,
+      "step": 8630
+    },
+    {
+      "epoch": 1.3984121840570318,
+      "grad_norm": 0.760558545589447,
+      "learning_rate": 2.8941093097962776e-06,
+      "loss": 0.0934,
+      "step": 8631
+    },
+    {
+      "epoch": 1.3985742060920285,
+      "grad_norm": 0.7663840055465698,
+      "learning_rate": 2.8936774676961264e-06,
+      "loss": 0.1026,
+      "step": 8632
+    },
+    {
+      "epoch": 1.3987362281270252,
+      "grad_norm": 0.8325403928756714,
+      "learning_rate": 2.893245613550397e-06,
+      "loss": 0.1097,
+      "step": 8633
+    },
+    {
+      "epoch": 1.398898250162022,
+      "grad_norm": 0.8603909611701965,
+      "learning_rate": 2.892813747372305e-06,
+      "loss": 0.1057,
+      "step": 8634
+    },
+    {
+      "epoch": 1.3990602721970187,
+      "grad_norm": 0.859589159488678,
+      "learning_rate": 2.892381869175061e-06,
+      "loss": 0.1124,
+      "step": 8635
+    },
+    {
+      "epoch": 1.3992222942320156,
+      "grad_norm": 0.9289984703063965,
+      "learning_rate": 2.891949978971883e-06,
+      "loss": 0.1186,
+      "step": 8636
+    },
+    {
+      "epoch": 1.3993843162670123,
+      "grad_norm": 0.7763188481330872,
+      "learning_rate": 2.891518076775983e-06,
+      "loss": 0.0958,
+      "step": 8637
+    },
+    {
+      "epoch": 1.399546338302009,
+      "grad_norm": 0.8290835618972778,
+      "learning_rate": 2.8910861626005774e-06,
+      "loss": 0.112,
+      "step": 8638
+    },
+    {
+      "epoch": 1.399708360337006,
+      "grad_norm": 0.7143067121505737,
+      "learning_rate": 2.890654236458882e-06,
+      "loss": 0.0856,
+      "step": 8639
+    },
+    {
+      "epoch": 1.3998703823720027,
+      "grad_norm": 0.8053386211395264,
+      "learning_rate": 2.890222298364112e-06,
+      "loss": 0.0909,
+      "step": 8640
+    },
+    {
+      "epoch": 1.4000324044069994,
+      "grad_norm": 0.9519616961479187,
+      "learning_rate": 2.8897903483294844e-06,
+      "loss": 0.1239,
+      "step": 8641
+    },
+    {
+      "epoch": 1.400194426441996,
+      "grad_norm": 0.7467250227928162,
+      "learning_rate": 2.8893583863682157e-06,
+      "loss": 0.0984,
+      "step": 8642
+    },
+    {
+      "epoch": 1.4003564484769928,
+      "grad_norm": 0.7794899940490723,
+      "learning_rate": 2.8889264124935217e-06,
+      "loss": 0.1038,
+      "step": 8643
+    },
+    {
+      "epoch": 1.4005184705119897,
+      "grad_norm": 0.8558558821678162,
+      "learning_rate": 2.888494426718621e-06,
+      "loss": 0.1056,
+      "step": 8644
+    },
+    {
+      "epoch": 1.4006804925469865,
+      "grad_norm": 0.9650840759277344,
+      "learning_rate": 2.888062429056731e-06,
+      "loss": 0.1189,
+      "step": 8645
+    },
+    {
+      "epoch": 1.4008425145819832,
+      "grad_norm": 1.1352195739746094,
+      "learning_rate": 2.8876304195210697e-06,
+      "loss": 0.1153,
+      "step": 8646
+    },
+    {
+      "epoch": 1.4010045366169799,
+      "grad_norm": 0.755615770816803,
+      "learning_rate": 2.8871983981248556e-06,
+      "loss": 0.098,
+      "step": 8647
+    },
+    {
+      "epoch": 1.4011665586519766,
+      "grad_norm": 0.7980273365974426,
+      "learning_rate": 2.8867663648813077e-06,
+      "loss": 0.1014,
+      "step": 8648
+    },
+    {
+      "epoch": 1.4013285806869735,
+      "grad_norm": 0.9927170872688293,
+      "learning_rate": 2.8863343198036453e-06,
+      "loss": 0.116,
+      "step": 8649
+    },
+    {
+      "epoch": 1.4014906027219702,
+      "grad_norm": 0.8513253331184387,
+      "learning_rate": 2.885902262905087e-06,
+      "loss": 0.1084,
+      "step": 8650
+    },
+    {
+      "epoch": 1.401652624756967,
+      "grad_norm": 0.849332332611084,
+      "learning_rate": 2.885470194198854e-06,
+      "loss": 0.1006,
+      "step": 8651
+    },
+    {
+      "epoch": 1.4018146467919637,
+      "grad_norm": 0.7961606979370117,
+      "learning_rate": 2.885038113698165e-06,
+      "loss": 0.0955,
+      "step": 8652
+    },
+    {
+      "epoch": 1.4019766688269604,
+      "grad_norm": 0.7978127002716064,
+      "learning_rate": 2.8846060214162426e-06,
+      "loss": 0.1055,
+      "step": 8653
+    },
+    {
+      "epoch": 1.4021386908619573,
+      "grad_norm": 0.9275472164154053,
+      "learning_rate": 2.8841739173663057e-06,
+      "loss": 0.1165,
+      "step": 8654
+    },
+    {
+      "epoch": 1.402300712896954,
+      "grad_norm": 0.9332693815231323,
+      "learning_rate": 2.883741801561577e-06,
+      "loss": 0.1121,
+      "step": 8655
+    },
+    {
+      "epoch": 1.4024627349319507,
+      "grad_norm": 0.7103630304336548,
+      "learning_rate": 2.883309674015278e-06,
+      "loss": 0.0933,
+      "step": 8656
+    },
+    {
+      "epoch": 1.4026247569669474,
+      "grad_norm": 0.8611851930618286,
+      "learning_rate": 2.8828775347406295e-06,
+      "loss": 0.1044,
+      "step": 8657
+    },
+    {
+      "epoch": 1.4027867790019442,
+      "grad_norm": 0.7644463181495667,
+      "learning_rate": 2.8824453837508563e-06,
+      "loss": 0.0959,
+      "step": 8658
+    },
+    {
+      "epoch": 1.402948801036941,
+      "grad_norm": 0.7625099420547485,
+      "learning_rate": 2.882013221059179e-06,
+      "loss": 0.0867,
+      "step": 8659
+    },
+    {
+      "epoch": 1.4031108230719378,
+      "grad_norm": 0.9080460667610168,
+      "learning_rate": 2.8815810466788225e-06,
+      "loss": 0.109,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4032728451069345,
+      "grad_norm": 0.8725658059120178,
+      "learning_rate": 2.881148860623009e-06,
+      "loss": 0.1114,
+      "step": 8661
+    },
+    {
+      "epoch": 1.4034348671419314,
+      "grad_norm": 0.9554914832115173,
+      "learning_rate": 2.8807166629049623e-06,
+      "loss": 0.1124,
+      "step": 8662
+    },
+    {
+      "epoch": 1.403596889176928,
+      "grad_norm": 0.9919114708900452,
+      "learning_rate": 2.880284453537907e-06,
+      "loss": 0.1289,
+      "step": 8663
+    },
+    {
+      "epoch": 1.4037589112119249,
+      "grad_norm": 0.8158717751502991,
+      "learning_rate": 2.8798522325350683e-06,
+      "loss": 0.1049,
+      "step": 8664
+    },
+    {
+      "epoch": 1.4039209332469216,
+      "grad_norm": 0.8903204202651978,
+      "learning_rate": 2.8794199999096708e-06,
+      "loss": 0.105,
+      "step": 8665
+    },
+    {
+      "epoch": 1.4040829552819183,
+      "grad_norm": 0.7267638444900513,
+      "learning_rate": 2.8789877556749383e-06,
+      "loss": 0.0889,
+      "step": 8666
+    },
+    {
+      "epoch": 1.4042449773169152,
+      "grad_norm": 0.7867768406867981,
+      "learning_rate": 2.8785554998440983e-06,
+      "loss": 0.0943,
+      "step": 8667
+    },
+    {
+      "epoch": 1.404406999351912,
+      "grad_norm": 0.7844392657279968,
+      "learning_rate": 2.8781232324303758e-06,
+      "loss": 0.1015,
+      "step": 8668
+    },
+    {
+      "epoch": 1.4045690213869086,
+      "grad_norm": 0.8047603964805603,
+      "learning_rate": 2.8776909534469976e-06,
+      "loss": 0.1025,
+      "step": 8669
+    },
+    {
+      "epoch": 1.4047310434219054,
+      "grad_norm": 0.85113525390625,
+      "learning_rate": 2.8772586629071902e-06,
+      "loss": 0.1092,
+      "step": 8670
+    },
+    {
+      "epoch": 1.404893065456902,
+      "grad_norm": 0.7649945020675659,
+      "learning_rate": 2.8768263608241805e-06,
+      "loss": 0.1005,
+      "step": 8671
+    },
+    {
+      "epoch": 1.405055087491899,
+      "grad_norm": 0.8279699683189392,
+      "learning_rate": 2.876394047211196e-06,
+      "loss": 0.1035,
+      "step": 8672
+    },
+    {
+      "epoch": 1.4052171095268957,
+      "grad_norm": 0.8674699664115906,
+      "learning_rate": 2.8759617220814654e-06,
+      "loss": 0.1056,
+      "step": 8673
+    },
+    {
+      "epoch": 1.4053791315618924,
+      "grad_norm": 0.7891427278518677,
+      "learning_rate": 2.875529385448215e-06,
+      "loss": 0.0974,
+      "step": 8674
+    },
+    {
+      "epoch": 1.4055411535968891,
+      "grad_norm": 0.7829881906509399,
+      "learning_rate": 2.8750970373246745e-06,
+      "loss": 0.096,
+      "step": 8675
+    },
+    {
+      "epoch": 1.4057031756318858,
+      "grad_norm": 0.8128500580787659,
+      "learning_rate": 2.8746646777240724e-06,
+      "loss": 0.1085,
+      "step": 8676
+    },
+    {
+      "epoch": 1.4058651976668828,
+      "grad_norm": 0.9597805142402649,
+      "learning_rate": 2.874232306659638e-06,
+      "loss": 0.12,
+      "step": 8677
+    },
+    {
+      "epoch": 1.4060272197018795,
+      "grad_norm": 0.7913134694099426,
+      "learning_rate": 2.8737999241446e-06,
+      "loss": 0.102,
+      "step": 8678
+    },
+    {
+      "epoch": 1.4061892417368762,
+      "grad_norm": 0.8597573637962341,
+      "learning_rate": 2.8733675301921893e-06,
+      "loss": 0.1114,
+      "step": 8679
+    },
+    {
+      "epoch": 1.406351263771873,
+      "grad_norm": 0.8230311274528503,
+      "learning_rate": 2.8729351248156364e-06,
+      "loss": 0.1063,
+      "step": 8680
+    },
+    {
+      "epoch": 1.4065132858068696,
+      "grad_norm": 0.9289247989654541,
+      "learning_rate": 2.87250270802817e-06,
+      "loss": 0.1094,
+      "step": 8681
+    },
+    {
+      "epoch": 1.4066753078418666,
+      "grad_norm": 0.647177517414093,
+      "learning_rate": 2.872070279843023e-06,
+      "loss": 0.0788,
+      "step": 8682
+    },
+    {
+      "epoch": 1.4068373298768633,
+      "grad_norm": 0.850144624710083,
+      "learning_rate": 2.871637840273425e-06,
+      "loss": 0.1101,
+      "step": 8683
+    },
+    {
+      "epoch": 1.40699935191186,
+      "grad_norm": 0.8800421357154846,
+      "learning_rate": 2.8712053893326088e-06,
+      "loss": 0.1125,
+      "step": 8684
+    },
+    {
+      "epoch": 1.4071613739468567,
+      "grad_norm": 0.9055140614509583,
+      "learning_rate": 2.8707729270338058e-06,
+      "loss": 0.1144,
+      "step": 8685
+    },
+    {
+      "epoch": 1.4073233959818534,
+      "grad_norm": 0.9264971017837524,
+      "learning_rate": 2.8703404533902492e-06,
+      "loss": 0.1058,
+      "step": 8686
+    },
+    {
+      "epoch": 1.4074854180168503,
+      "grad_norm": 0.8147197365760803,
+      "learning_rate": 2.869907968415171e-06,
+      "loss": 0.1054,
+      "step": 8687
+    },
+    {
+      "epoch": 1.407647440051847,
+      "grad_norm": 0.9164337515830994,
+      "learning_rate": 2.8694754721218027e-06,
+      "loss": 0.1209,
+      "step": 8688
+    },
+    {
+      "epoch": 1.4078094620868438,
+      "grad_norm": 0.8125103116035461,
+      "learning_rate": 2.8690429645233808e-06,
+      "loss": 0.0983,
+      "step": 8689
+    },
+    {
+      "epoch": 1.4079714841218407,
+      "grad_norm": 0.7318618297576904,
+      "learning_rate": 2.8686104456331356e-06,
+      "loss": 0.0946,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4081335061568374,
+      "grad_norm": 0.7763205170631409,
+      "learning_rate": 2.868177915464304e-06,
+      "loss": 0.0966,
+      "step": 8691
+    },
+    {
+      "epoch": 1.4082955281918341,
+      "grad_norm": 0.7962520122528076,
+      "learning_rate": 2.8677453740301185e-06,
+      "loss": 0.0979,
+      "step": 8692
+    },
+    {
+      "epoch": 1.4084575502268308,
+      "grad_norm": 0.8198848366737366,
+      "learning_rate": 2.867312821343815e-06,
+      "loss": 0.0941,
+      "step": 8693
+    },
+    {
+      "epoch": 1.4086195722618275,
+      "grad_norm": 0.8297473192214966,
+      "learning_rate": 2.8668802574186277e-06,
+      "loss": 0.1059,
+      "step": 8694
+    },
+    {
+      "epoch": 1.4087815942968245,
+      "grad_norm": 0.9335203766822815,
+      "learning_rate": 2.866447682267792e-06,
+      "loss": 0.1144,
+      "step": 8695
+    },
+    {
+      "epoch": 1.4089436163318212,
+      "grad_norm": 0.8151853084564209,
+      "learning_rate": 2.8660150959045456e-06,
+      "loss": 0.1041,
+      "step": 8696
+    },
+    {
+      "epoch": 1.409105638366818,
+      "grad_norm": 0.8563622832298279,
+      "learning_rate": 2.8655824983421217e-06,
+      "loss": 0.1103,
+      "step": 8697
+    },
+    {
+      "epoch": 1.4092676604018146,
+      "grad_norm": 0.8096403479576111,
+      "learning_rate": 2.865149889593758e-06,
+      "loss": 0.1094,
+      "step": 8698
+    },
+    {
+      "epoch": 1.4094296824368113,
+      "grad_norm": 0.859600841999054,
+      "learning_rate": 2.8647172696726917e-06,
+      "loss": 0.1041,
+      "step": 8699
+    },
+    {
+      "epoch": 1.4095917044718083,
+      "grad_norm": 0.8782287836074829,
+      "learning_rate": 2.8642846385921593e-06,
+      "loss": 0.1014,
+      "step": 8700
+    },
+    {
+      "epoch": 1.409753726506805,
+      "grad_norm": 0.8271836638450623,
+      "learning_rate": 2.8638519963653987e-06,
+      "loss": 0.1045,
+      "step": 8701
+    },
+    {
+      "epoch": 1.4099157485418017,
+      "grad_norm": 0.7663862705230713,
+      "learning_rate": 2.863419343005647e-06,
+      "loss": 0.1,
+      "step": 8702
+    },
+    {
+      "epoch": 1.4100777705767984,
+      "grad_norm": 0.9826388955116272,
+      "learning_rate": 2.8629866785261435e-06,
+      "loss": 0.1191,
+      "step": 8703
+    },
+    {
+      "epoch": 1.410239792611795,
+      "grad_norm": 0.7487360239028931,
+      "learning_rate": 2.8625540029401262e-06,
+      "loss": 0.0963,
+      "step": 8704
+    },
+    {
+      "epoch": 1.410401814646792,
+      "grad_norm": 0.8938003778457642,
+      "learning_rate": 2.862121316260833e-06,
+      "loss": 0.107,
+      "step": 8705
+    },
+    {
+      "epoch": 1.4105638366817888,
+      "grad_norm": 0.8720892071723938,
+      "learning_rate": 2.8616886185015046e-06,
+      "loss": 0.1147,
+      "step": 8706
+    },
+    {
+      "epoch": 1.4107258587167855,
+      "grad_norm": 0.8877646327018738,
+      "learning_rate": 2.8612559096753797e-06,
+      "loss": 0.1134,
+      "step": 8707
+    },
+    {
+      "epoch": 1.4108878807517822,
+      "grad_norm": 0.8477475643157959,
+      "learning_rate": 2.860823189795697e-06,
+      "loss": 0.1173,
+      "step": 8708
+    },
+    {
+      "epoch": 1.4110499027867789,
+      "grad_norm": 0.8832873106002808,
+      "learning_rate": 2.8603904588756996e-06,
+      "loss": 0.1127,
+      "step": 8709
+    },
+    {
+      "epoch": 1.4112119248217758,
+      "grad_norm": 0.8365153074264526,
+      "learning_rate": 2.859957716928625e-06,
+      "loss": 0.1029,
+      "step": 8710
+    },
+    {
+      "epoch": 1.4113739468567725,
+      "grad_norm": 0.861546516418457,
+      "learning_rate": 2.8595249639677164e-06,
+      "loss": 0.1085,
+      "step": 8711
+    },
+    {
+      "epoch": 1.4115359688917692,
+      "grad_norm": 0.8814380168914795,
+      "learning_rate": 2.8590922000062125e-06,
+      "loss": 0.1142,
+      "step": 8712
+    },
+    {
+      "epoch": 1.4116979909267662,
+      "grad_norm": 0.8180842995643616,
+      "learning_rate": 2.858659425057357e-06,
+      "loss": 0.1038,
+      "step": 8713
+    },
+    {
+      "epoch": 1.4118600129617627,
+      "grad_norm": 0.7602577209472656,
+      "learning_rate": 2.858226639134391e-06,
+      "loss": 0.0989,
+      "step": 8714
+    },
+    {
+      "epoch": 1.4120220349967596,
+      "grad_norm": 0.8542532324790955,
+      "learning_rate": 2.8577938422505573e-06,
+      "loss": 0.1079,
+      "step": 8715
+    },
+    {
+      "epoch": 1.4121840570317563,
+      "grad_norm": 0.7752845883369446,
+      "learning_rate": 2.8573610344190978e-06,
+      "loss": 0.1,
+      "step": 8716
+    },
+    {
+      "epoch": 1.412346079066753,
+      "grad_norm": 0.8004235029220581,
+      "learning_rate": 2.8569282156532548e-06,
+      "loss": 0.1044,
+      "step": 8717
+    },
+    {
+      "epoch": 1.41250810110175,
+      "grad_norm": 0.904018759727478,
+      "learning_rate": 2.8564953859662725e-06,
+      "loss": 0.1214,
+      "step": 8718
+    },
+    {
+      "epoch": 1.4126701231367467,
+      "grad_norm": 0.8129547238349915,
+      "learning_rate": 2.8560625453713935e-06,
+      "loss": 0.1058,
+      "step": 8719
+    },
+    {
+      "epoch": 1.4128321451717434,
+      "grad_norm": 0.8800033926963806,
+      "learning_rate": 2.8556296938818632e-06,
+      "loss": 0.0912,
+      "step": 8720
+    },
+    {
+      "epoch": 1.41299416720674,
+      "grad_norm": 0.7877295017242432,
+      "learning_rate": 2.8551968315109246e-06,
+      "loss": 0.0965,
+      "step": 8721
+    },
+    {
+      "epoch": 1.4131561892417368,
+      "grad_norm": 0.9171632528305054,
+      "learning_rate": 2.8547639582718223e-06,
+      "loss": 0.1216,
+      "step": 8722
+    },
+    {
+      "epoch": 1.4133182112767337,
+      "grad_norm": 0.766700804233551,
+      "learning_rate": 2.8543310741778013e-06,
+      "loss": 0.0901,
+      "step": 8723
+    },
+    {
+      "epoch": 1.4134802333117304,
+      "grad_norm": 0.8369571566581726,
+      "learning_rate": 2.853898179242107e-06,
+      "loss": 0.1022,
+      "step": 8724
+    },
+    {
+      "epoch": 1.4136422553467272,
+      "grad_norm": 0.7450268268585205,
+      "learning_rate": 2.853465273477985e-06,
+      "loss": 0.0961,
+      "step": 8725
+    },
+    {
+      "epoch": 1.4138042773817239,
+      "grad_norm": 0.8542775511741638,
+      "learning_rate": 2.8530323568986805e-06,
+      "loss": 0.1118,
+      "step": 8726
+    },
+    {
+      "epoch": 1.4139662994167206,
+      "grad_norm": 0.7954638004302979,
+      "learning_rate": 2.852599429517441e-06,
+      "loss": 0.104,
+      "step": 8727
+    },
+    {
+      "epoch": 1.4141283214517175,
+      "grad_norm": 0.7494431138038635,
+      "learning_rate": 2.8521664913475123e-06,
+      "loss": 0.0874,
+      "step": 8728
+    },
+    {
+      "epoch": 1.4142903434867142,
+      "grad_norm": 0.8307495713233948,
+      "learning_rate": 2.8517335424021404e-06,
+      "loss": 0.0994,
+      "step": 8729
+    },
+    {
+      "epoch": 1.414452365521711,
+      "grad_norm": 0.8603584170341492,
+      "learning_rate": 2.8513005826945733e-06,
+      "loss": 0.11,
+      "step": 8730
+    },
+    {
+      "epoch": 1.4146143875567077,
+      "grad_norm": 0.9381308555603027,
+      "learning_rate": 2.850867612238059e-06,
+      "loss": 0.1049,
+      "step": 8731
+    },
+    {
+      "epoch": 1.4147764095917044,
+      "grad_norm": 0.7319527864456177,
+      "learning_rate": 2.8504346310458446e-06,
+      "loss": 0.0898,
+      "step": 8732
+    },
+    {
+      "epoch": 1.4149384316267013,
+      "grad_norm": 0.8726157546043396,
+      "learning_rate": 2.850001639131179e-06,
+      "loss": 0.1157,
+      "step": 8733
+    },
+    {
+      "epoch": 1.415100453661698,
+      "grad_norm": 0.6778660416603088,
+      "learning_rate": 2.8495686365073096e-06,
+      "loss": 0.088,
+      "step": 8734
+    },
+    {
+      "epoch": 1.4152624756966947,
+      "grad_norm": 0.9084062576293945,
+      "learning_rate": 2.849135623187486e-06,
+      "loss": 0.1115,
+      "step": 8735
+    },
+    {
+      "epoch": 1.4154244977316914,
+      "grad_norm": 0.8225180506706238,
+      "learning_rate": 2.848702599184957e-06,
+      "loss": 0.1022,
+      "step": 8736
+    },
+    {
+      "epoch": 1.4155865197666881,
+      "grad_norm": 0.7956673502922058,
+      "learning_rate": 2.8482695645129725e-06,
+      "loss": 0.0988,
+      "step": 8737
+    },
+    {
+      "epoch": 1.415748541801685,
+      "grad_norm": 0.710880696773529,
+      "learning_rate": 2.8478365191847824e-06,
+      "loss": 0.0868,
+      "step": 8738
+    },
+    {
+      "epoch": 1.4159105638366818,
+      "grad_norm": 0.7995860576629639,
+      "learning_rate": 2.8474034632136365e-06,
+      "loss": 0.1029,
+      "step": 8739
+    },
+    {
+      "epoch": 1.4160725858716785,
+      "grad_norm": 0.8467408418655396,
+      "learning_rate": 2.8469703966127853e-06,
+      "loss": 0.1085,
+      "step": 8740
+    },
+    {
+      "epoch": 1.4162346079066754,
+      "grad_norm": 0.8162935972213745,
+      "learning_rate": 2.8465373193954794e-06,
+      "loss": 0.1065,
+      "step": 8741
+    },
+    {
+      "epoch": 1.4163966299416721,
+      "grad_norm": 0.8570012450218201,
+      "learning_rate": 2.8461042315749706e-06,
+      "loss": 0.1045,
+      "step": 8742
+    },
+    {
+      "epoch": 1.4165586519766689,
+      "grad_norm": 0.7713600397109985,
+      "learning_rate": 2.8456711331645085e-06,
+      "loss": 0.0956,
+      "step": 8743
+    },
+    {
+      "epoch": 1.4167206740116656,
+      "grad_norm": 0.7573341131210327,
+      "learning_rate": 2.845238024177348e-06,
+      "loss": 0.098,
+      "step": 8744
+    },
+    {
+      "epoch": 1.4168826960466623,
+      "grad_norm": 0.8098436594009399,
+      "learning_rate": 2.8448049046267377e-06,
+      "loss": 0.0979,
+      "step": 8745
+    },
+    {
+      "epoch": 1.4170447180816592,
+      "grad_norm": 0.8652111887931824,
+      "learning_rate": 2.8443717745259335e-06,
+      "loss": 0.1117,
+      "step": 8746
+    },
+    {
+      "epoch": 1.417206740116656,
+      "grad_norm": 0.8840187788009644,
+      "learning_rate": 2.843938633888186e-06,
+      "loss": 0.1069,
+      "step": 8747
+    },
+    {
+      "epoch": 1.4173687621516526,
+      "grad_norm": 0.7981458902359009,
+      "learning_rate": 2.8435054827267476e-06,
+      "loss": 0.1045,
+      "step": 8748
+    },
+    {
+      "epoch": 1.4175307841866494,
+      "grad_norm": 0.8698453903198242,
+      "learning_rate": 2.843072321054873e-06,
+      "loss": 0.1061,
+      "step": 8749
+    },
+    {
+      "epoch": 1.417692806221646,
+      "grad_norm": 0.9237022995948792,
+      "learning_rate": 2.8426391488858163e-06,
+      "loss": 0.1168,
+      "step": 8750
+    },
+    {
+      "epoch": 1.417854828256643,
+      "grad_norm": 0.8793653249740601,
+      "learning_rate": 2.8422059662328306e-06,
+      "loss": 0.117,
+      "step": 8751
+    },
+    {
+      "epoch": 1.4180168502916397,
+      "grad_norm": 0.7878406047821045,
+      "learning_rate": 2.8417727731091705e-06,
+      "loss": 0.1027,
+      "step": 8752
+    },
+    {
+      "epoch": 1.4181788723266364,
+      "grad_norm": 0.9198635816574097,
+      "learning_rate": 2.84133956952809e-06,
+      "loss": 0.1156,
+      "step": 8753
+    },
+    {
+      "epoch": 1.4183408943616331,
+      "grad_norm": 0.8705915212631226,
+      "learning_rate": 2.840906355502845e-06,
+      "loss": 0.1141,
+      "step": 8754
+    },
+    {
+      "epoch": 1.4185029163966298,
+      "grad_norm": 0.8760009407997131,
+      "learning_rate": 2.8404731310466904e-06,
+      "loss": 0.1222,
+      "step": 8755
+    },
+    {
+      "epoch": 1.4186649384316268,
+      "grad_norm": 0.8203396201133728,
+      "learning_rate": 2.840039896172882e-06,
+      "loss": 0.1032,
+      "step": 8756
+    },
+    {
+      "epoch": 1.4188269604666235,
+      "grad_norm": 0.7818074226379395,
+      "learning_rate": 2.8396066508946757e-06,
+      "loss": 0.1038,
+      "step": 8757
+    },
+    {
+      "epoch": 1.4189889825016202,
+      "grad_norm": 0.7406920194625854,
+      "learning_rate": 2.8391733952253277e-06,
+      "loss": 0.0913,
+      "step": 8758
+    },
+    {
+      "epoch": 1.419151004536617,
+      "grad_norm": 0.8644748330116272,
+      "learning_rate": 2.8387401291780953e-06,
+      "loss": 0.1071,
+      "step": 8759
+    },
+    {
+      "epoch": 1.4193130265716136,
+      "grad_norm": 0.8011223077774048,
+      "learning_rate": 2.838306852766234e-06,
+      "loss": 0.1088,
+      "step": 8760
+    },
+    {
+      "epoch": 1.4194750486066106,
+      "grad_norm": 0.9043474197387695,
+      "learning_rate": 2.8378735660030015e-06,
+      "loss": 0.1187,
+      "step": 8761
+    },
+    {
+      "epoch": 1.4196370706416073,
+      "grad_norm": 0.7650865316390991,
+      "learning_rate": 2.8374402689016557e-06,
+      "loss": 0.0959,
+      "step": 8762
+    },
+    {
+      "epoch": 1.419799092676604,
+      "grad_norm": 0.7926430702209473,
+      "learning_rate": 2.8370069614754543e-06,
+      "loss": 0.094,
+      "step": 8763
+    },
+    {
+      "epoch": 1.419961114711601,
+      "grad_norm": 0.8434385061264038,
+      "learning_rate": 2.8365736437376555e-06,
+      "loss": 0.1034,
+      "step": 8764
+    },
+    {
+      "epoch": 1.4201231367465974,
+      "grad_norm": 0.7807652950286865,
+      "learning_rate": 2.836140315701517e-06,
+      "loss": 0.1004,
+      "step": 8765
+    },
+    {
+      "epoch": 1.4202851587815943,
+      "grad_norm": 0.9462375640869141,
+      "learning_rate": 2.8357069773802996e-06,
+      "loss": 0.1111,
+      "step": 8766
+    },
+    {
+      "epoch": 1.420447180816591,
+      "grad_norm": 0.8678317666053772,
+      "learning_rate": 2.8352736287872593e-06,
+      "loss": 0.1101,
+      "step": 8767
+    },
+    {
+      "epoch": 1.4206092028515878,
+      "grad_norm": 0.8266333937644958,
+      "learning_rate": 2.834840269935659e-06,
+      "loss": 0.1046,
+      "step": 8768
+    },
+    {
+      "epoch": 1.4207712248865847,
+      "grad_norm": 0.8634325265884399,
+      "learning_rate": 2.8344069008387565e-06,
+      "loss": 0.1141,
+      "step": 8769
+    },
+    {
+      "epoch": 1.4209332469215814,
+      "grad_norm": 0.7722366452217102,
+      "learning_rate": 2.833973521509812e-06,
+      "loss": 0.0944,
+      "step": 8770
+    },
+    {
+      "epoch": 1.4210952689565781,
+      "grad_norm": 0.7650349736213684,
+      "learning_rate": 2.8335401319620855e-06,
+      "loss": 0.1033,
+      "step": 8771
+    },
+    {
+      "epoch": 1.4212572909915748,
+      "grad_norm": 0.94795161485672,
+      "learning_rate": 2.833106732208838e-06,
+      "loss": 0.1154,
+      "step": 8772
+    },
+    {
+      "epoch": 1.4214193130265715,
+      "grad_norm": 0.8152647614479065,
+      "learning_rate": 2.832673322263331e-06,
+      "loss": 0.1029,
+      "step": 8773
+    },
+    {
+      "epoch": 1.4215813350615685,
+      "grad_norm": 0.843647301197052,
+      "learning_rate": 2.8322399021388248e-06,
+      "loss": 0.1097,
+      "step": 8774
+    },
+    {
+      "epoch": 1.4217433570965652,
+      "grad_norm": 0.8536216616630554,
+      "learning_rate": 2.8318064718485826e-06,
+      "loss": 0.1011,
+      "step": 8775
+    },
+    {
+      "epoch": 1.421905379131562,
+      "grad_norm": 0.8077113628387451,
+      "learning_rate": 2.8313730314058645e-06,
+      "loss": 0.103,
+      "step": 8776
+    },
+    {
+      "epoch": 1.4220674011665586,
+      "grad_norm": 0.964287281036377,
+      "learning_rate": 2.830939580823934e-06,
+      "loss": 0.1125,
+      "step": 8777
+    },
+    {
+      "epoch": 1.4222294232015553,
+      "grad_norm": 0.8470258712768555,
+      "learning_rate": 2.830506120116053e-06,
+      "loss": 0.1101,
+      "step": 8778
+    },
+    {
+      "epoch": 1.4223914452365523,
+      "grad_norm": 1.022905707359314,
+      "learning_rate": 2.8300726492954845e-06,
+      "loss": 0.1195,
+      "step": 8779
+    },
+    {
+      "epoch": 1.422553467271549,
+      "grad_norm": 0.8365886211395264,
+      "learning_rate": 2.8296391683754916e-06,
+      "loss": 0.107,
+      "step": 8780
+    },
+    {
+      "epoch": 1.4227154893065457,
+      "grad_norm": 0.8074929118156433,
+      "learning_rate": 2.829205677369338e-06,
+      "loss": 0.1062,
+      "step": 8781
+    },
+    {
+      "epoch": 1.4228775113415424,
+      "grad_norm": 0.8210521340370178,
+      "learning_rate": 2.8287721762902877e-06,
+      "loss": 0.1056,
+      "step": 8782
+    },
+    {
+      "epoch": 1.423039533376539,
+      "grad_norm": 0.845281183719635,
+      "learning_rate": 2.8283386651516037e-06,
+      "loss": 0.1036,
+      "step": 8783
+    },
+    {
+      "epoch": 1.423201555411536,
+      "grad_norm": 0.7844056487083435,
+      "learning_rate": 2.8279051439665516e-06,
+      "loss": 0.0968,
+      "step": 8784
+    },
+    {
+      "epoch": 1.4233635774465327,
+      "grad_norm": 0.8035227656364441,
+      "learning_rate": 2.8274716127483955e-06,
+      "loss": 0.1052,
+      "step": 8785
+    },
+    {
+      "epoch": 1.4235255994815295,
+      "grad_norm": 0.8737231492996216,
+      "learning_rate": 2.8270380715104e-06,
+      "loss": 0.114,
+      "step": 8786
+    },
+    {
+      "epoch": 1.4236876215165262,
+      "grad_norm": 0.8196536302566528,
+      "learning_rate": 2.8266045202658316e-06,
+      "loss": 0.1088,
+      "step": 8787
+    },
+    {
+      "epoch": 1.4238496435515229,
+      "grad_norm": 0.7950606942176819,
+      "learning_rate": 2.826170959027956e-06,
+      "loss": 0.0977,
+      "step": 8788
+    },
+    {
+      "epoch": 1.4240116655865198,
+      "grad_norm": 0.8498073816299438,
+      "learning_rate": 2.8257373878100363e-06,
+      "loss": 0.1049,
+      "step": 8789
+    },
+    {
+      "epoch": 1.4241736876215165,
+      "grad_norm": 0.8204236030578613,
+      "learning_rate": 2.8253038066253423e-06,
+      "loss": 0.1023,
+      "step": 8790
+    },
+    {
+      "epoch": 1.4243357096565132,
+      "grad_norm": 0.9035735130310059,
+      "learning_rate": 2.8248702154871387e-06,
+      "loss": 0.1169,
+      "step": 8791
+    },
+    {
+      "epoch": 1.4244977316915102,
+      "grad_norm": 0.9391461610794067,
+      "learning_rate": 2.8244366144086926e-06,
+      "loss": 0.1122,
+      "step": 8792
+    },
+    {
+      "epoch": 1.4246597537265069,
+      "grad_norm": 0.8815183639526367,
+      "learning_rate": 2.824003003403271e-06,
+      "loss": 0.1041,
+      "step": 8793
+    },
+    {
+      "epoch": 1.4248217757615036,
+      "grad_norm": 0.7553703784942627,
+      "learning_rate": 2.823569382484142e-06,
+      "loss": 0.0949,
+      "step": 8794
+    },
+    {
+      "epoch": 1.4249837977965003,
+      "grad_norm": 0.8771671652793884,
+      "learning_rate": 2.823135751664573e-06,
+      "loss": 0.1142,
+      "step": 8795
+    },
+    {
+      "epoch": 1.425145819831497,
+      "grad_norm": 0.845828652381897,
+      "learning_rate": 2.822702110957831e-06,
+      "loss": 0.1074,
+      "step": 8796
+    },
+    {
+      "epoch": 1.425307841866494,
+      "grad_norm": 0.823070764541626,
+      "learning_rate": 2.8222684603771867e-06,
+      "loss": 0.1042,
+      "step": 8797
+    },
+    {
+      "epoch": 1.4254698639014907,
+      "grad_norm": 0.8956811428070068,
+      "learning_rate": 2.8218347999359066e-06,
+      "loss": 0.1038,
+      "step": 8798
+    },
+    {
+      "epoch": 1.4256318859364874,
+      "grad_norm": 0.9194256067276001,
+      "learning_rate": 2.821401129647261e-06,
+      "loss": 0.1194,
+      "step": 8799
+    },
+    {
+      "epoch": 1.425793907971484,
+      "grad_norm": 0.787661612033844,
+      "learning_rate": 2.8209674495245177e-06,
+      "loss": 0.0979,
+      "step": 8800
+    },
+    {
+      "epoch": 1.4259559300064808,
+      "grad_norm": 0.8394095301628113,
+      "learning_rate": 2.820533759580948e-06,
+      "loss": 0.1078,
+      "step": 8801
+    },
+    {
+      "epoch": 1.4261179520414777,
+      "grad_norm": 0.7722134590148926,
+      "learning_rate": 2.82010005982982e-06,
+      "loss": 0.1064,
+      "step": 8802
+    },
+    {
+      "epoch": 1.4262799740764744,
+      "grad_norm": 0.841996967792511,
+      "learning_rate": 2.8196663502844057e-06,
+      "loss": 0.0996,
+      "step": 8803
+    },
+    {
+      "epoch": 1.4264419961114712,
+      "grad_norm": 0.9296277165412903,
+      "learning_rate": 2.819232630957975e-06,
+      "loss": 0.1196,
+      "step": 8804
+    },
+    {
+      "epoch": 1.4266040181464679,
+      "grad_norm": 0.8732765316963196,
+      "learning_rate": 2.8187989018637967e-06,
+      "loss": 0.1097,
+      "step": 8805
+    },
+    {
+      "epoch": 1.4267660401814646,
+      "grad_norm": 0.9206928014755249,
+      "learning_rate": 2.818365163015145e-06,
+      "loss": 0.1229,
+      "step": 8806
+    },
+    {
+      "epoch": 1.4269280622164615,
+      "grad_norm": 0.853338897228241,
+      "learning_rate": 2.817931414425289e-06,
+      "loss": 0.0942,
+      "step": 8807
+    },
+    {
+      "epoch": 1.4270900842514582,
+      "grad_norm": 0.8418228030204773,
+      "learning_rate": 2.8174976561075013e-06,
+      "loss": 0.1125,
+      "step": 8808
+    },
+    {
+      "epoch": 1.427252106286455,
+      "grad_norm": 0.9058745503425598,
+      "learning_rate": 2.8170638880750534e-06,
+      "loss": 0.1103,
+      "step": 8809
+    },
+    {
+      "epoch": 1.4274141283214516,
+      "grad_norm": 0.8129554986953735,
+      "learning_rate": 2.816630110341218e-06,
+      "loss": 0.1046,
+      "step": 8810
+    },
+    {
+      "epoch": 1.4275761503564484,
+      "grad_norm": 0.7814062237739563,
+      "learning_rate": 2.8161963229192677e-06,
+      "loss": 0.103,
+      "step": 8811
+    },
+    {
+      "epoch": 1.4277381723914453,
+      "grad_norm": 0.7614774703979492,
+      "learning_rate": 2.8157625258224746e-06,
+      "loss": 0.0985,
+      "step": 8812
+    },
+    {
+      "epoch": 1.427900194426442,
+      "grad_norm": 0.8251438140869141,
+      "learning_rate": 2.8153287190641133e-06,
+      "loss": 0.105,
+      "step": 8813
+    },
+    {
+      "epoch": 1.4280622164614387,
+      "grad_norm": 0.8020395040512085,
+      "learning_rate": 2.814894902657456e-06,
+      "loss": 0.1027,
+      "step": 8814
+    },
+    {
+      "epoch": 1.4282242384964356,
+      "grad_norm": 0.7605652213096619,
+      "learning_rate": 2.8144610766157758e-06,
+      "loss": 0.0941,
+      "step": 8815
+    },
+    {
+      "epoch": 1.4283862605314321,
+      "grad_norm": 0.7266712784767151,
+      "learning_rate": 2.814027240952348e-06,
+      "loss": 0.085,
+      "step": 8816
+    },
+    {
+      "epoch": 1.428548282566429,
+      "grad_norm": 0.8249565958976746,
+      "learning_rate": 2.813593395680447e-06,
+      "loss": 0.1041,
+      "step": 8817
+    },
+    {
+      "epoch": 1.4287103046014258,
+      "grad_norm": 1.0275297164916992,
+      "learning_rate": 2.8131595408133467e-06,
+      "loss": 0.1263,
+      "step": 8818
+    },
+    {
+      "epoch": 1.4288723266364225,
+      "grad_norm": 1.5727760791778564,
+      "learning_rate": 2.812725676364322e-06,
+      "loss": 0.1006,
+      "step": 8819
+    },
+    {
+      "epoch": 1.4290343486714194,
+      "grad_norm": 0.8301246762275696,
+      "learning_rate": 2.8122918023466485e-06,
+      "loss": 0.1036,
+      "step": 8820
+    },
+    {
+      "epoch": 1.4291963707064161,
+      "grad_norm": 0.8229403495788574,
+      "learning_rate": 2.811857918773602e-06,
+      "loss": 0.0966,
+      "step": 8821
+    },
+    {
+      "epoch": 1.4293583927414129,
+      "grad_norm": 0.794026255607605,
+      "learning_rate": 2.811424025658458e-06,
+      "loss": 0.1077,
+      "step": 8822
+    },
+    {
+      "epoch": 1.4295204147764096,
+      "grad_norm": 0.6794479489326477,
+      "learning_rate": 2.810990123014492e-06,
+      "loss": 0.0826,
+      "step": 8823
+    },
+    {
+      "epoch": 1.4296824368114063,
+      "grad_norm": 0.759669303894043,
+      "learning_rate": 2.8105562108549807e-06,
+      "loss": 0.0996,
+      "step": 8824
+    },
+    {
+      "epoch": 1.4298444588464032,
+      "grad_norm": 0.8474135398864746,
+      "learning_rate": 2.8101222891932013e-06,
+      "loss": 0.1085,
+      "step": 8825
+    },
+    {
+      "epoch": 1.4300064808814,
+      "grad_norm": 0.9885803461074829,
+      "learning_rate": 2.80968835804243e-06,
+      "loss": 0.121,
+      "step": 8826
+    },
+    {
+      "epoch": 1.4301685029163966,
+      "grad_norm": 0.8411709070205688,
+      "learning_rate": 2.809254417415944e-06,
+      "loss": 0.1085,
+      "step": 8827
+    },
+    {
+      "epoch": 1.4303305249513933,
+      "grad_norm": 1.034460425376892,
+      "learning_rate": 2.808820467327022e-06,
+      "loss": 0.1044,
+      "step": 8828
+    },
+    {
+      "epoch": 1.43049254698639,
+      "grad_norm": 0.9514307975769043,
+      "learning_rate": 2.8083865077889404e-06,
+      "loss": 0.1225,
+      "step": 8829
+    },
+    {
+      "epoch": 1.430654569021387,
+      "grad_norm": 0.7484989762306213,
+      "learning_rate": 2.8079525388149787e-06,
+      "loss": 0.0957,
+      "step": 8830
+    },
+    {
+      "epoch": 1.4308165910563837,
+      "grad_norm": 0.8544020652770996,
+      "learning_rate": 2.807518560418414e-06,
+      "loss": 0.1155,
+      "step": 8831
+    },
+    {
+      "epoch": 1.4309786130913804,
+      "grad_norm": 0.8093070387840271,
+      "learning_rate": 2.8070845726125257e-06,
+      "loss": 0.1042,
+      "step": 8832
+    },
+    {
+      "epoch": 1.4311406351263771,
+      "grad_norm": 0.7702214121818542,
+      "learning_rate": 2.806650575410592e-06,
+      "loss": 0.0942,
+      "step": 8833
+    },
+    {
+      "epoch": 1.4313026571613738,
+      "grad_norm": 0.793833315372467,
+      "learning_rate": 2.8062165688258934e-06,
+      "loss": 0.0903,
+      "step": 8834
+    },
+    {
+      "epoch": 1.4314646791963708,
+      "grad_norm": 0.9053971171379089,
+      "learning_rate": 2.8057825528717093e-06,
+      "loss": 0.1259,
+      "step": 8835
+    },
+    {
+      "epoch": 1.4316267012313675,
+      "grad_norm": 0.9572237730026245,
+      "learning_rate": 2.8053485275613177e-06,
+      "loss": 0.111,
+      "step": 8836
+    },
+    {
+      "epoch": 1.4317887232663642,
+      "grad_norm": 0.7796591520309448,
+      "learning_rate": 2.804914492908001e-06,
+      "loss": 0.0959,
+      "step": 8837
+    },
+    {
+      "epoch": 1.431950745301361,
+      "grad_norm": 0.733897864818573,
+      "learning_rate": 2.804480448925039e-06,
+      "loss": 0.0852,
+      "step": 8838
+    },
+    {
+      "epoch": 1.4321127673363576,
+      "grad_norm": 0.7287041544914246,
+      "learning_rate": 2.8040463956257113e-06,
+      "loss": 0.0934,
+      "step": 8839
+    },
+    {
+      "epoch": 1.4322747893713546,
+      "grad_norm": 0.8480597138404846,
+      "learning_rate": 2.8036123330233e-06,
+      "loss": 0.1066,
+      "step": 8840
+    },
+    {
+      "epoch": 1.4324368114063513,
+      "grad_norm": 0.8591471314430237,
+      "learning_rate": 2.8031782611310863e-06,
+      "loss": 0.1044,
+      "step": 8841
+    },
+    {
+      "epoch": 1.432598833441348,
+      "grad_norm": 0.7007931470870972,
+      "learning_rate": 2.802744179962351e-06,
+      "loss": 0.0905,
+      "step": 8842
+    },
+    {
+      "epoch": 1.432760855476345,
+      "grad_norm": 0.7390332818031311,
+      "learning_rate": 2.802310089530377e-06,
+      "loss": 0.0923,
+      "step": 8843
+    },
+    {
+      "epoch": 1.4329228775113416,
+      "grad_norm": 0.8909687995910645,
+      "learning_rate": 2.801875989848446e-06,
+      "loss": 0.1139,
+      "step": 8844
+    },
+    {
+      "epoch": 1.4330848995463383,
+      "grad_norm": 0.795418918132782,
+      "learning_rate": 2.801441880929839e-06,
+      "loss": 0.103,
+      "step": 8845
+    },
+    {
+      "epoch": 1.433246921581335,
+      "grad_norm": 0.7816963195800781,
+      "learning_rate": 2.8010077627878414e-06,
+      "loss": 0.1,
+      "step": 8846
+    },
+    {
+      "epoch": 1.4334089436163318,
+      "grad_norm": 0.8099395632743835,
+      "learning_rate": 2.8005736354357338e-06,
+      "loss": 0.094,
+      "step": 8847
+    },
+    {
+      "epoch": 1.4335709656513287,
+      "grad_norm": 0.7562690377235413,
+      "learning_rate": 2.8001394988868003e-06,
+      "loss": 0.0969,
+      "step": 8848
+    },
+    {
+      "epoch": 1.4337329876863254,
+      "grad_norm": 0.9833892583847046,
+      "learning_rate": 2.7997053531543246e-06,
+      "loss": 0.1045,
+      "step": 8849
+    },
+    {
+      "epoch": 1.4338950097213221,
+      "grad_norm": 0.8875625133514404,
+      "learning_rate": 2.7992711982515908e-06,
+      "loss": 0.1067,
+      "step": 8850
+    },
+    {
+      "epoch": 1.4340570317563188,
+      "grad_norm": 0.7625311613082886,
+      "learning_rate": 2.7988370341918814e-06,
+      "loss": 0.0928,
+      "step": 8851
+    },
+    {
+      "epoch": 1.4342190537913155,
+      "grad_norm": 0.8266885876655579,
+      "learning_rate": 2.798402860988483e-06,
+      "loss": 0.1055,
+      "step": 8852
+    },
+    {
+      "epoch": 1.4343810758263125,
+      "grad_norm": 0.8304306864738464,
+      "learning_rate": 2.7979686786546784e-06,
+      "loss": 0.1017,
+      "step": 8853
+    },
+    {
+      "epoch": 1.4345430978613092,
+      "grad_norm": 0.8403946161270142,
+      "learning_rate": 2.797534487203755e-06,
+      "loss": 0.0894,
+      "step": 8854
+    },
+    {
+      "epoch": 1.434705119896306,
+      "grad_norm": 0.7953592538833618,
+      "learning_rate": 2.7971002866489944e-06,
+      "loss": 0.098,
+      "step": 8855
+    },
+    {
+      "epoch": 1.4348671419313026,
+      "grad_norm": 0.7994452714920044,
+      "learning_rate": 2.7966660770036845e-06,
+      "loss": 0.1017,
+      "step": 8856
+    },
+    {
+      "epoch": 1.4350291639662993,
+      "grad_norm": 0.8061219453811646,
+      "learning_rate": 2.7962318582811113e-06,
+      "loss": 0.1053,
+      "step": 8857
+    },
+    {
+      "epoch": 1.4351911860012962,
+      "grad_norm": 0.9328574538230896,
+      "learning_rate": 2.795797630494559e-06,
+      "loss": 0.1206,
+      "step": 8858
+    },
+    {
+      "epoch": 1.435353208036293,
+      "grad_norm": 0.8536462187767029,
+      "learning_rate": 2.795363393657316e-06,
+      "loss": 0.1042,
+      "step": 8859
+    },
+    {
+      "epoch": 1.4355152300712897,
+      "grad_norm": 0.7858737111091614,
+      "learning_rate": 2.7949291477826666e-06,
+      "loss": 0.0977,
+      "step": 8860
+    },
+    {
+      "epoch": 1.4356772521062864,
+      "grad_norm": 0.887958824634552,
+      "learning_rate": 2.7944948928839007e-06,
+      "loss": 0.1106,
+      "step": 8861
+    },
+    {
+      "epoch": 1.435839274141283,
+      "grad_norm": 0.817033588886261,
+      "learning_rate": 2.7940606289743026e-06,
+      "loss": 0.0977,
+      "step": 8862
+    },
+    {
+      "epoch": 1.43600129617628,
+      "grad_norm": 0.8055084943771362,
+      "learning_rate": 2.793626356067161e-06,
+      "loss": 0.0928,
+      "step": 8863
+    },
+    {
+      "epoch": 1.4361633182112767,
+      "grad_norm": 0.8124540448188782,
+      "learning_rate": 2.793192074175764e-06,
+      "loss": 0.0977,
+      "step": 8864
+    },
+    {
+      "epoch": 1.4363253402462735,
+      "grad_norm": 0.9017152786254883,
+      "learning_rate": 2.7927577833133984e-06,
+      "loss": 0.1147,
+      "step": 8865
+    },
+    {
+      "epoch": 1.4364873622812704,
+      "grad_norm": 0.8836957812309265,
+      "learning_rate": 2.792323483493354e-06,
+      "loss": 0.1094,
+      "step": 8866
+    },
+    {
+      "epoch": 1.4366493843162669,
+      "grad_norm": 0.8717904686927795,
+      "learning_rate": 2.791889174728918e-06,
+      "loss": 0.1102,
+      "step": 8867
+    },
+    {
+      "epoch": 1.4368114063512638,
+      "grad_norm": 0.7438794374465942,
+      "learning_rate": 2.791454857033379e-06,
+      "loss": 0.0957,
+      "step": 8868
+    },
+    {
+      "epoch": 1.4369734283862605,
+      "grad_norm": 0.7357698678970337,
+      "learning_rate": 2.7910205304200273e-06,
+      "loss": 0.0928,
+      "step": 8869
+    },
+    {
+      "epoch": 1.4371354504212572,
+      "grad_norm": 0.8071755170822144,
+      "learning_rate": 2.790586194902151e-06,
+      "loss": 0.0974,
+      "step": 8870
+    },
+    {
+      "epoch": 1.4372974724562542,
+      "grad_norm": 0.8086465001106262,
+      "learning_rate": 2.790151850493041e-06,
+      "loss": 0.1087,
+      "step": 8871
+    },
+    {
+      "epoch": 1.4374594944912509,
+      "grad_norm": 0.7857526540756226,
+      "learning_rate": 2.789717497205986e-06,
+      "loss": 0.0962,
+      "step": 8872
+    },
+    {
+      "epoch": 1.4376215165262476,
+      "grad_norm": 0.8768362998962402,
+      "learning_rate": 2.789283135054277e-06,
+      "loss": 0.0975,
+      "step": 8873
+    },
+    {
+      "epoch": 1.4377835385612443,
+      "grad_norm": 0.8221284747123718,
+      "learning_rate": 2.7888487640512046e-06,
+      "loss": 0.0953,
+      "step": 8874
+    },
+    {
+      "epoch": 1.437945560596241,
+      "grad_norm": 0.9030830264091492,
+      "learning_rate": 2.7884143842100573e-06,
+      "loss": 0.1122,
+      "step": 8875
+    },
+    {
+      "epoch": 1.438107582631238,
+      "grad_norm": 0.9299379587173462,
+      "learning_rate": 2.78797999554413e-06,
+      "loss": 0.1183,
+      "step": 8876
+    },
+    {
+      "epoch": 1.4382696046662347,
+      "grad_norm": 0.7504619359970093,
+      "learning_rate": 2.7875455980667106e-06,
+      "loss": 0.0942,
+      "step": 8877
+    },
+    {
+      "epoch": 1.4384316267012314,
+      "grad_norm": 0.7830450534820557,
+      "learning_rate": 2.787111191791092e-06,
+      "loss": 0.0983,
+      "step": 8878
+    },
+    {
+      "epoch": 1.438593648736228,
+      "grad_norm": 0.7624955177307129,
+      "learning_rate": 2.786676776730566e-06,
+      "loss": 0.0961,
+      "step": 8879
+    },
+    {
+      "epoch": 1.4387556707712248,
+      "grad_norm": 0.7818362712860107,
+      "learning_rate": 2.7862423528984233e-06,
+      "loss": 0.1087,
+      "step": 8880
+    },
+    {
+      "epoch": 1.4389176928062217,
+      "grad_norm": 0.8774171471595764,
+      "learning_rate": 2.7858079203079587e-06,
+      "loss": 0.1192,
+      "step": 8881
+    },
+    {
+      "epoch": 1.4390797148412184,
+      "grad_norm": 0.7342621088027954,
+      "learning_rate": 2.7853734789724618e-06,
+      "loss": 0.0933,
+      "step": 8882
+    },
+    {
+      "epoch": 1.4392417368762151,
+      "grad_norm": 0.7560212016105652,
+      "learning_rate": 2.7849390289052287e-06,
+      "loss": 0.0867,
+      "step": 8883
+    },
+    {
+      "epoch": 1.4394037589112119,
+      "grad_norm": 1.23517644405365,
+      "learning_rate": 2.7845045701195494e-06,
+      "loss": 0.1019,
+      "step": 8884
+    },
+    {
+      "epoch": 1.4395657809462086,
+      "grad_norm": 0.9435912370681763,
+      "learning_rate": 2.78407010262872e-06,
+      "loss": 0.1217,
+      "step": 8885
+    },
+    {
+      "epoch": 1.4397278029812055,
+      "grad_norm": 0.9272345304489136,
+      "learning_rate": 2.7836356264460316e-06,
+      "loss": 0.1167,
+      "step": 8886
+    },
+    {
+      "epoch": 1.4398898250162022,
+      "grad_norm": 0.8792025446891785,
+      "learning_rate": 2.7832011415847802e-06,
+      "loss": 0.1165,
+      "step": 8887
+    },
+    {
+      "epoch": 1.440051847051199,
+      "grad_norm": 0.7143024206161499,
+      "learning_rate": 2.7827666480582593e-06,
+      "loss": 0.0933,
+      "step": 8888
+    },
+    {
+      "epoch": 1.4402138690861956,
+      "grad_norm": 0.8933820724487305,
+      "learning_rate": 2.782332145879763e-06,
+      "loss": 0.1105,
+      "step": 8889
+    },
+    {
+      "epoch": 1.4403758911211924,
+      "grad_norm": 1.0869182348251343,
+      "learning_rate": 2.7818976350625864e-06,
+      "loss": 0.0966,
+      "step": 8890
+    },
+    {
+      "epoch": 1.4405379131561893,
+      "grad_norm": 0.7515729665756226,
+      "learning_rate": 2.781463115620024e-06,
+      "loss": 0.0954,
+      "step": 8891
+    },
+    {
+      "epoch": 1.440699935191186,
+      "grad_norm": 0.6556647419929504,
+      "learning_rate": 2.781028587565372e-06,
+      "loss": 0.0812,
+      "step": 8892
+    },
+    {
+      "epoch": 1.4408619572261827,
+      "grad_norm": 0.8055543899536133,
+      "learning_rate": 2.780594050911925e-06,
+      "loss": 0.1076,
+      "step": 8893
+    },
+    {
+      "epoch": 1.4410239792611796,
+      "grad_norm": 0.8170286417007446,
+      "learning_rate": 2.780159505672979e-06,
+      "loss": 0.1098,
+      "step": 8894
+    },
+    {
+      "epoch": 1.4411860012961764,
+      "grad_norm": 0.8746091723442078,
+      "learning_rate": 2.7797249518618304e-06,
+      "loss": 0.1001,
+      "step": 8895
+    },
+    {
+      "epoch": 1.441348023331173,
+      "grad_norm": 0.8125444650650024,
+      "learning_rate": 2.7792903894917746e-06,
+      "loss": 0.1104,
+      "step": 8896
+    },
+    {
+      "epoch": 1.4415100453661698,
+      "grad_norm": 0.8494876623153687,
+      "learning_rate": 2.778855818576109e-06,
+      "loss": 0.1037,
+      "step": 8897
+    },
+    {
+      "epoch": 1.4416720674011665,
+      "grad_norm": 0.8670206069946289,
+      "learning_rate": 2.7784212391281307e-06,
+      "loss": 0.1096,
+      "step": 8898
+    },
+    {
+      "epoch": 1.4418340894361634,
+      "grad_norm": 0.7742645144462585,
+      "learning_rate": 2.777986651161136e-06,
+      "loss": 0.0943,
+      "step": 8899
+    },
+    {
+      "epoch": 1.4419961114711601,
+      "grad_norm": 0.8516318798065186,
+      "learning_rate": 2.7775520546884216e-06,
+      "loss": 0.1088,
+      "step": 8900
+    },
+    {
+      "epoch": 1.4421581335061568,
+      "grad_norm": 0.8558171987533569,
+      "learning_rate": 2.7771174497232867e-06,
+      "loss": 0.0966,
+      "step": 8901
+    },
+    {
+      "epoch": 1.4423201555411536,
+      "grad_norm": 0.8049555420875549,
+      "learning_rate": 2.7766828362790283e-06,
+      "loss": 0.1069,
+      "step": 8902
+    },
+    {
+      "epoch": 1.4424821775761503,
+      "grad_norm": 0.8404254913330078,
+      "learning_rate": 2.776248214368945e-06,
+      "loss": 0.1062,
+      "step": 8903
+    },
+    {
+      "epoch": 1.4426441996111472,
+      "grad_norm": 0.8144481182098389,
+      "learning_rate": 2.7758135840063344e-06,
+      "loss": 0.1033,
+      "step": 8904
+    },
+    {
+      "epoch": 1.442806221646144,
+      "grad_norm": 0.8399234414100647,
+      "learning_rate": 2.7753789452044965e-06,
+      "loss": 0.1119,
+      "step": 8905
+    },
+    {
+      "epoch": 1.4429682436811406,
+      "grad_norm": 0.8735939264297485,
+      "learning_rate": 2.7749442979767276e-06,
+      "loss": 0.1109,
+      "step": 8906
+    },
+    {
+      "epoch": 1.4431302657161373,
+      "grad_norm": 0.7629880309104919,
+      "learning_rate": 2.7745096423363304e-06,
+      "loss": 0.096,
+      "step": 8907
+    },
+    {
+      "epoch": 1.443292287751134,
+      "grad_norm": 0.9925363659858704,
+      "learning_rate": 2.7740749782966016e-06,
+      "loss": 0.1199,
+      "step": 8908
+    },
+    {
+      "epoch": 1.443454309786131,
+      "grad_norm": 0.8804534077644348,
+      "learning_rate": 2.7736403058708418e-06,
+      "loss": 0.1041,
+      "step": 8909
+    },
+    {
+      "epoch": 1.4436163318211277,
+      "grad_norm": 0.8817947506904602,
+      "learning_rate": 2.7732056250723505e-06,
+      "loss": 0.1021,
+      "step": 8910
+    },
+    {
+      "epoch": 1.4437783538561244,
+      "grad_norm": 0.8750995397567749,
+      "learning_rate": 2.7727709359144285e-06,
+      "loss": 0.1109,
+      "step": 8911
+    },
+    {
+      "epoch": 1.4439403758911211,
+      "grad_norm": 0.8417905569076538,
+      "learning_rate": 2.7723362384103757e-06,
+      "loss": 0.0968,
+      "step": 8912
+    },
+    {
+      "epoch": 1.4441023979261178,
+      "grad_norm": 0.8828087449073792,
+      "learning_rate": 2.771901532573493e-06,
+      "loss": 0.1105,
+      "step": 8913
+    },
+    {
+      "epoch": 1.4442644199611148,
+      "grad_norm": 0.8812400102615356,
+      "learning_rate": 2.771466818417082e-06,
+      "loss": 0.1137,
+      "step": 8914
+    },
+    {
+      "epoch": 1.4444264419961115,
+      "grad_norm": 1.0078636407852173,
+      "learning_rate": 2.7710320959544425e-06,
+      "loss": 0.1276,
+      "step": 8915
+    },
+    {
+      "epoch": 1.4445884640311082,
+      "grad_norm": 0.8423323035240173,
+      "learning_rate": 2.7705973651988777e-06,
+      "loss": 0.0995,
+      "step": 8916
+    },
+    {
+      "epoch": 1.4447504860661051,
+      "grad_norm": 0.6651060581207275,
+      "learning_rate": 2.7701626261636878e-06,
+      "loss": 0.0942,
+      "step": 8917
+    },
+    {
+      "epoch": 1.4449125081011016,
+      "grad_norm": 0.8124906420707703,
+      "learning_rate": 2.769727878862175e-06,
+      "loss": 0.0977,
+      "step": 8918
+    },
+    {
+      "epoch": 1.4450745301360985,
+      "grad_norm": 0.784057080745697,
+      "learning_rate": 2.7692931233076424e-06,
+      "loss": 0.0979,
+      "step": 8919
+    },
+    {
+      "epoch": 1.4452365521710953,
+      "grad_norm": 0.825549840927124,
+      "learning_rate": 2.768858359513392e-06,
+      "loss": 0.1119,
+      "step": 8920
+    },
+    {
+      "epoch": 1.445398574206092,
+      "grad_norm": 0.7953062653541565,
+      "learning_rate": 2.7684235874927264e-06,
+      "loss": 0.1072,
+      "step": 8921
+    },
+    {
+      "epoch": 1.445560596241089,
+      "grad_norm": 0.8686825633049011,
+      "learning_rate": 2.767988807258948e-06,
+      "loss": 0.103,
+      "step": 8922
+    },
+    {
+      "epoch": 1.4457226182760856,
+      "grad_norm": 0.8071175813674927,
+      "learning_rate": 2.7675540188253606e-06,
+      "loss": 0.1091,
+      "step": 8923
+    },
+    {
+      "epoch": 1.4458846403110823,
+      "grad_norm": 0.8495711088180542,
+      "learning_rate": 2.7671192222052685e-06,
+      "loss": 0.1146,
+      "step": 8924
+    },
+    {
+      "epoch": 1.446046662346079,
+      "grad_norm": 0.7676787376403809,
+      "learning_rate": 2.7666844174119738e-06,
+      "loss": 0.0902,
+      "step": 8925
+    },
+    {
+      "epoch": 1.4462086843810757,
+      "grad_norm": 0.950420618057251,
+      "learning_rate": 2.7662496044587817e-06,
+      "loss": 0.1138,
+      "step": 8926
+    },
+    {
+      "epoch": 1.4463707064160727,
+      "grad_norm": 0.7517897486686707,
+      "learning_rate": 2.765814783358996e-06,
+      "loss": 0.094,
+      "step": 8927
+    },
+    {
+      "epoch": 1.4465327284510694,
+      "grad_norm": 0.8687278032302856,
+      "learning_rate": 2.765379954125921e-06,
+      "loss": 0.1039,
+      "step": 8928
+    },
+    {
+      "epoch": 1.446694750486066,
+      "grad_norm": 0.9656805992126465,
+      "learning_rate": 2.764945116772862e-06,
+      "loss": 0.1127,
+      "step": 8929
+    },
+    {
+      "epoch": 1.4468567725210628,
+      "grad_norm": 0.9119703769683838,
+      "learning_rate": 2.764510271313123e-06,
+      "loss": 0.117,
+      "step": 8930
+    },
+    {
+      "epoch": 1.4470187945560595,
+      "grad_norm": 0.7118834853172302,
+      "learning_rate": 2.7640754177600105e-06,
+      "loss": 0.0926,
+      "step": 8931
+    },
+    {
+      "epoch": 1.4471808165910565,
+      "grad_norm": 0.7804062962532043,
+      "learning_rate": 2.7636405561268286e-06,
+      "loss": 0.1061,
+      "step": 8932
+    },
+    {
+      "epoch": 1.4473428386260532,
+      "grad_norm": 0.8037486672401428,
+      "learning_rate": 2.763205686426884e-06,
+      "loss": 0.1024,
+      "step": 8933
+    },
+    {
+      "epoch": 1.4475048606610499,
+      "grad_norm": 0.9153349995613098,
+      "learning_rate": 2.7627708086734827e-06,
+      "loss": 0.1165,
+      "step": 8934
+    },
+    {
+      "epoch": 1.4476668826960466,
+      "grad_norm": 0.8715000748634338,
+      "learning_rate": 2.7623359228799295e-06,
+      "loss": 0.1049,
+      "step": 8935
+    },
+    {
+      "epoch": 1.4478289047310433,
+      "grad_norm": 1.018729329109192,
+      "learning_rate": 2.7619010290595333e-06,
+      "loss": 0.1107,
+      "step": 8936
+    },
+    {
+      "epoch": 1.4479909267660402,
+      "grad_norm": 0.7773792743682861,
+      "learning_rate": 2.761466127225598e-06,
+      "loss": 0.0909,
+      "step": 8937
+    },
+    {
+      "epoch": 1.448152948801037,
+      "grad_norm": 0.722775399684906,
+      "learning_rate": 2.7610312173914334e-06,
+      "loss": 0.0939,
+      "step": 8938
+    },
+    {
+      "epoch": 1.4483149708360337,
+      "grad_norm": 0.8534601926803589,
+      "learning_rate": 2.760596299570344e-06,
+      "loss": 0.0987,
+      "step": 8939
+    },
+    {
+      "epoch": 1.4484769928710304,
+      "grad_norm": 0.81373131275177,
+      "learning_rate": 2.760161373775639e-06,
+      "loss": 0.101,
+      "step": 8940
+    },
+    {
+      "epoch": 1.448639014906027,
+      "grad_norm": 0.8647781014442444,
+      "learning_rate": 2.7597264400206255e-06,
+      "loss": 0.1064,
+      "step": 8941
+    },
+    {
+      "epoch": 1.448801036941024,
+      "grad_norm": 0.9041435122489929,
+      "learning_rate": 2.7592914983186113e-06,
+      "loss": 0.104,
+      "step": 8942
+    },
+    {
+      "epoch": 1.4489630589760207,
+      "grad_norm": 0.7579626441001892,
+      "learning_rate": 2.7588565486829054e-06,
+      "loss": 0.1023,
+      "step": 8943
+    },
+    {
+      "epoch": 1.4491250810110174,
+      "grad_norm": 0.7907978892326355,
+      "learning_rate": 2.758421591126814e-06,
+      "loss": 0.0998,
+      "step": 8944
+    },
+    {
+      "epoch": 1.4492871030460144,
+      "grad_norm": 0.7959480285644531,
+      "learning_rate": 2.757986625663649e-06,
+      "loss": 0.1035,
+      "step": 8945
+    },
+    {
+      "epoch": 1.449449125081011,
+      "grad_norm": 0.8328735828399658,
+      "learning_rate": 2.757551652306717e-06,
+      "loss": 0.1081,
+      "step": 8946
+    },
+    {
+      "epoch": 1.4496111471160078,
+      "grad_norm": 0.8220754265785217,
+      "learning_rate": 2.757116671069327e-06,
+      "loss": 0.0985,
+      "step": 8947
+    },
+    {
+      "epoch": 1.4497731691510045,
+      "grad_norm": 0.7394413352012634,
+      "learning_rate": 2.7566816819647897e-06,
+      "loss": 0.0931,
+      "step": 8948
+    },
+    {
+      "epoch": 1.4499351911860012,
+      "grad_norm": 0.8674811124801636,
+      "learning_rate": 2.756246685006414e-06,
+      "loss": 0.1102,
+      "step": 8949
+    },
+    {
+      "epoch": 1.4500972132209982,
+      "grad_norm": 0.944241464138031,
+      "learning_rate": 2.7558116802075095e-06,
+      "loss": 0.1129,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4502592352559949,
+      "grad_norm": 0.8318021297454834,
+      "learning_rate": 2.755376667581387e-06,
+      "loss": 0.1066,
+      "step": 8951
+    },
+    {
+      "epoch": 1.4504212572909916,
+      "grad_norm": 0.7419362664222717,
+      "learning_rate": 2.754941647141357e-06,
+      "loss": 0.0957,
+      "step": 8952
+    },
+    {
+      "epoch": 1.4505832793259883,
+      "grad_norm": 0.7914148569107056,
+      "learning_rate": 2.754506618900729e-06,
+      "loss": 0.1047,
+      "step": 8953
+    },
+    {
+      "epoch": 1.450745301360985,
+      "grad_norm": 0.8463177680969238,
+      "learning_rate": 2.754071582872814e-06,
+      "loss": 0.1088,
+      "step": 8954
+    },
+    {
+      "epoch": 1.450907323395982,
+      "grad_norm": 0.9250964522361755,
+      "learning_rate": 2.753636539070924e-06,
+      "loss": 0.118,
+      "step": 8955
+    },
+    {
+      "epoch": 1.4510693454309787,
+      "grad_norm": 0.8688222169876099,
+      "learning_rate": 2.753201487508369e-06,
+      "loss": 0.0961,
+      "step": 8956
+    },
+    {
+      "epoch": 1.4512313674659754,
+      "grad_norm": 0.8172348737716675,
+      "learning_rate": 2.752766428198462e-06,
+      "loss": 0.1063,
+      "step": 8957
+    },
+    {
+      "epoch": 1.451393389500972,
+      "grad_norm": 0.8788450360298157,
+      "learning_rate": 2.7523313611545133e-06,
+      "loss": 0.1075,
+      "step": 8958
+    },
+    {
+      "epoch": 1.4515554115359688,
+      "grad_norm": 0.8048200011253357,
+      "learning_rate": 2.7518962863898356e-06,
+      "loss": 0.1012,
+      "step": 8959
+    },
+    {
+      "epoch": 1.4517174335709657,
+      "grad_norm": 0.8715628981590271,
+      "learning_rate": 2.7514612039177422e-06,
+      "loss": 0.1027,
+      "step": 8960
+    },
+    {
+      "epoch": 1.4518794556059624,
+      "grad_norm": 0.9554488658905029,
+      "learning_rate": 2.7510261137515437e-06,
+      "loss": 0.1079,
+      "step": 8961
+    },
+    {
+      "epoch": 1.4520414776409591,
+      "grad_norm": 0.908086895942688,
+      "learning_rate": 2.7505910159045534e-06,
+      "loss": 0.1199,
+      "step": 8962
+    },
+    {
+      "epoch": 1.4522034996759559,
+      "grad_norm": 0.7679954171180725,
+      "learning_rate": 2.750155910390085e-06,
+      "loss": 0.0917,
+      "step": 8963
+    },
+    {
+      "epoch": 1.4523655217109526,
+      "grad_norm": 0.8922917246818542,
+      "learning_rate": 2.74972079722145e-06,
+      "loss": 0.1107,
+      "step": 8964
+    },
+    {
+      "epoch": 1.4525275437459495,
+      "grad_norm": 0.8348422050476074,
+      "learning_rate": 2.7492856764119644e-06,
+      "loss": 0.1075,
+      "step": 8965
+    },
+    {
+      "epoch": 1.4526895657809462,
+      "grad_norm": 0.852973997592926,
+      "learning_rate": 2.7488505479749395e-06,
+      "loss": 0.1131,
+      "step": 8966
+    },
+    {
+      "epoch": 1.452851587815943,
+      "grad_norm": 0.8772859573364258,
+      "learning_rate": 2.7484154119236906e-06,
+      "loss": 0.1085,
+      "step": 8967
+    },
+    {
+      "epoch": 1.4530136098509399,
+      "grad_norm": 0.8057224154472351,
+      "learning_rate": 2.74798026827153e-06,
+      "loss": 0.1,
+      "step": 8968
+    },
+    {
+      "epoch": 1.4531756318859366,
+      "grad_norm": 0.8312748670578003,
+      "learning_rate": 2.7475451170317748e-06,
+      "loss": 0.1064,
+      "step": 8969
+    },
+    {
+      "epoch": 1.4533376539209333,
+      "grad_norm": 0.8106637001037598,
+      "learning_rate": 2.747109958217737e-06,
+      "loss": 0.106,
+      "step": 8970
+    },
+    {
+      "epoch": 1.45349967595593,
+      "grad_norm": 0.8003639578819275,
+      "learning_rate": 2.7466747918427326e-06,
+      "loss": 0.0978,
+      "step": 8971
+    },
+    {
+      "epoch": 1.4536616979909267,
+      "grad_norm": 0.8317378759384155,
+      "learning_rate": 2.746239617920077e-06,
+      "loss": 0.0995,
+      "step": 8972
+    },
+    {
+      "epoch": 1.4538237200259236,
+      "grad_norm": 0.8461683988571167,
+      "learning_rate": 2.7458044364630844e-06,
+      "loss": 0.0959,
+      "step": 8973
+    },
+    {
+      "epoch": 1.4539857420609203,
+      "grad_norm": 0.9303465485572815,
+      "learning_rate": 2.745369247485072e-06,
+      "loss": 0.1011,
+      "step": 8974
+    },
+    {
+      "epoch": 1.454147764095917,
+      "grad_norm": 0.7978488802909851,
+      "learning_rate": 2.7449340509993526e-06,
+      "loss": 0.0928,
+      "step": 8975
+    },
+    {
+      "epoch": 1.4543097861309138,
+      "grad_norm": 0.7847487330436707,
+      "learning_rate": 2.7444988470192457e-06,
+      "loss": 0.0917,
+      "step": 8976
+    },
+    {
+      "epoch": 1.4544718081659105,
+      "grad_norm": 0.7997845411300659,
+      "learning_rate": 2.744063635558065e-06,
+      "loss": 0.0966,
+      "step": 8977
+    },
+    {
+      "epoch": 1.4546338302009074,
+      "grad_norm": 0.7787443995475769,
+      "learning_rate": 2.743628416629128e-06,
+      "loss": 0.0986,
+      "step": 8978
+    },
+    {
+      "epoch": 1.4547958522359041,
+      "grad_norm": 0.9163244366645813,
+      "learning_rate": 2.7431931902457504e-06,
+      "loss": 0.1012,
+      "step": 8979
+    },
+    {
+      "epoch": 1.4549578742709008,
+      "grad_norm": 0.7863480448722839,
+      "learning_rate": 2.7427579564212496e-06,
+      "loss": 0.095,
+      "step": 8980
+    },
+    {
+      "epoch": 1.4551198963058976,
+      "grad_norm": 0.8067967295646667,
+      "learning_rate": 2.7423227151689436e-06,
+      "loss": 0.0963,
+      "step": 8981
+    },
+    {
+      "epoch": 1.4552819183408943,
+      "grad_norm": 0.8217342495918274,
+      "learning_rate": 2.7418874665021483e-06,
+      "loss": 0.1071,
+      "step": 8982
+    },
+    {
+      "epoch": 1.4554439403758912,
+      "grad_norm": 0.754622757434845,
+      "learning_rate": 2.7414522104341827e-06,
+      "loss": 0.0997,
+      "step": 8983
+    },
+    {
+      "epoch": 1.455605962410888,
+      "grad_norm": 0.8360654711723328,
+      "learning_rate": 2.7410169469783632e-06,
+      "loss": 0.1013,
+      "step": 8984
+    },
+    {
+      "epoch": 1.4557679844458846,
+      "grad_norm": 0.8221597075462341,
+      "learning_rate": 2.740581676148008e-06,
+      "loss": 0.101,
+      "step": 8985
+    },
+    {
+      "epoch": 1.4559300064808813,
+      "grad_norm": 0.9259410500526428,
+      "learning_rate": 2.7401463979564365e-06,
+      "loss": 0.1103,
+      "step": 8986
+    },
+    {
+      "epoch": 1.456092028515878,
+      "grad_norm": 0.8288673162460327,
+      "learning_rate": 2.739711112416966e-06,
+      "loss": 0.0889,
+      "step": 8987
+    },
+    {
+      "epoch": 1.456254050550875,
+      "grad_norm": 0.8852518796920776,
+      "learning_rate": 2.7392758195429153e-06,
+      "loss": 0.1054,
+      "step": 8988
+    },
+    {
+      "epoch": 1.4564160725858717,
+      "grad_norm": 0.7813482284545898,
+      "learning_rate": 2.738840519347604e-06,
+      "loss": 0.104,
+      "step": 8989
+    },
+    {
+      "epoch": 1.4565780946208684,
+      "grad_norm": 0.753773033618927,
+      "learning_rate": 2.73840521184435e-06,
+      "loss": 0.0911,
+      "step": 8990
+    },
+    {
+      "epoch": 1.4567401166558653,
+      "grad_norm": 0.8681461215019226,
+      "learning_rate": 2.737969897046475e-06,
+      "loss": 0.107,
+      "step": 8991
+    },
+    {
+      "epoch": 1.4569021386908618,
+      "grad_norm": 0.8012184500694275,
+      "learning_rate": 2.737534574967295e-06,
+      "loss": 0.1021,
+      "step": 8992
+    },
+    {
+      "epoch": 1.4570641607258588,
+      "grad_norm": 0.9006675481796265,
+      "learning_rate": 2.7370992456201333e-06,
+      "loss": 0.1089,
+      "step": 8993
+    },
+    {
+      "epoch": 1.4572261827608555,
+      "grad_norm": 0.8169563412666321,
+      "learning_rate": 2.7366639090183076e-06,
+      "loss": 0.1066,
+      "step": 8994
+    },
+    {
+      "epoch": 1.4573882047958522,
+      "grad_norm": 0.9039627313613892,
+      "learning_rate": 2.7362285651751396e-06,
+      "loss": 0.1084,
+      "step": 8995
+    },
+    {
+      "epoch": 1.4575502268308491,
+      "grad_norm": 1.059561014175415,
+      "learning_rate": 2.7357932141039494e-06,
+      "loss": 0.1334,
+      "step": 8996
+    },
+    {
+      "epoch": 1.4577122488658458,
+      "grad_norm": 0.8857733011245728,
+      "learning_rate": 2.7353578558180566e-06,
+      "loss": 0.1026,
+      "step": 8997
+    },
+    {
+      "epoch": 1.4578742709008425,
+      "grad_norm": 0.9009446501731873,
+      "learning_rate": 2.7349224903307836e-06,
+      "loss": 0.1097,
+      "step": 8998
+    },
+    {
+      "epoch": 1.4580362929358393,
+      "grad_norm": 0.794425368309021,
+      "learning_rate": 2.7344871176554498e-06,
+      "loss": 0.0953,
+      "step": 8999
+    },
+    {
+      "epoch": 1.458198314970836,
+      "grad_norm": 0.752216637134552,
+      "learning_rate": 2.734051737805379e-06,
+      "loss": 0.1114,
+      "step": 9000
+    },
+    {
+      "epoch": 1.458360337005833,
+      "grad_norm": 0.8077507019042969,
+      "learning_rate": 2.733616350793891e-06,
+      "loss": 0.0994,
+      "step": 9001
+    },
+    {
+      "epoch": 1.4585223590408296,
+      "grad_norm": 0.8075626492500305,
+      "learning_rate": 2.733180956634308e-06,
+      "loss": 0.1092,
+      "step": 9002
+    },
+    {
+      "epoch": 1.4586843810758263,
+      "grad_norm": 0.8624060153961182,
+      "learning_rate": 2.7327455553399523e-06,
+      "loss": 0.1134,
+      "step": 9003
+    },
+    {
+      "epoch": 1.458846403110823,
+      "grad_norm": 0.8211466073989868,
+      "learning_rate": 2.7323101469241454e-06,
+      "loss": 0.102,
+      "step": 9004
+    },
+    {
+      "epoch": 1.4590084251458197,
+      "grad_norm": 0.838489830493927,
+      "learning_rate": 2.7318747314002108e-06,
+      "loss": 0.1028,
+      "step": 9005
+    },
+    {
+      "epoch": 1.4591704471808167,
+      "grad_norm": 0.7504817247390747,
+      "learning_rate": 2.7314393087814693e-06,
+      "loss": 0.103,
+      "step": 9006
+    },
+    {
+      "epoch": 1.4593324692158134,
+      "grad_norm": 0.8312764167785645,
+      "learning_rate": 2.731003879081246e-06,
+      "loss": 0.1052,
+      "step": 9007
+    },
+    {
+      "epoch": 1.45949449125081,
+      "grad_norm": 0.7648634910583496,
+      "learning_rate": 2.7305684423128633e-06,
+      "loss": 0.0971,
+      "step": 9008
+    },
+    {
+      "epoch": 1.4596565132858068,
+      "grad_norm": 0.7204511165618896,
+      "learning_rate": 2.7301329984896435e-06,
+      "loss": 0.088,
+      "step": 9009
+    },
+    {
+      "epoch": 1.4598185353208035,
+      "grad_norm": 0.8412272334098816,
+      "learning_rate": 2.729697547624911e-06,
+      "loss": 0.1028,
+      "step": 9010
+    },
+    {
+      "epoch": 1.4599805573558005,
+      "grad_norm": 0.7355968952178955,
+      "learning_rate": 2.7292620897319892e-06,
+      "loss": 0.0942,
+      "step": 9011
+    },
+    {
+      "epoch": 1.4601425793907972,
+      "grad_norm": 0.8538913726806641,
+      "learning_rate": 2.7288266248242025e-06,
+      "loss": 0.0954,
+      "step": 9012
+    },
+    {
+      "epoch": 1.4603046014257939,
+      "grad_norm": 0.7657922506332397,
+      "learning_rate": 2.7283911529148753e-06,
+      "loss": 0.1011,
+      "step": 9013
+    },
+    {
+      "epoch": 1.4604666234607906,
+      "grad_norm": 0.951962947845459,
+      "learning_rate": 2.7279556740173306e-06,
+      "loss": 0.1025,
+      "step": 9014
+    },
+    {
+      "epoch": 1.4606286454957873,
+      "grad_norm": 0.7168994545936584,
+      "learning_rate": 2.727520188144895e-06,
+      "loss": 0.0845,
+      "step": 9015
+    },
+    {
+      "epoch": 1.4607906675307842,
+      "grad_norm": 0.7955917716026306,
+      "learning_rate": 2.7270846953108913e-06,
+      "loss": 0.0934,
+      "step": 9016
+    },
+    {
+      "epoch": 1.460952689565781,
+      "grad_norm": 0.8339834213256836,
+      "learning_rate": 2.7266491955286457e-06,
+      "loss": 0.1052,
+      "step": 9017
+    },
+    {
+      "epoch": 1.4611147116007777,
+      "grad_norm": 0.9418988823890686,
+      "learning_rate": 2.7262136888114833e-06,
+      "loss": 0.1146,
+      "step": 9018
+    },
+    {
+      "epoch": 1.4612767336357746,
+      "grad_norm": 0.7261226177215576,
+      "learning_rate": 2.725778175172729e-06,
+      "loss": 0.0869,
+      "step": 9019
+    },
+    {
+      "epoch": 1.4614387556707713,
+      "grad_norm": 1.1076301336288452,
+      "learning_rate": 2.72534265462571e-06,
+      "loss": 0.1158,
+      "step": 9020
+    },
+    {
+      "epoch": 1.461600777705768,
+      "grad_norm": 0.8176051378250122,
+      "learning_rate": 2.7249071271837503e-06,
+      "loss": 0.1012,
+      "step": 9021
+    },
+    {
+      "epoch": 1.4617627997407647,
+      "grad_norm": 0.7100493311882019,
+      "learning_rate": 2.7244715928601774e-06,
+      "loss": 0.088,
+      "step": 9022
+    },
+    {
+      "epoch": 1.4619248217757614,
+      "grad_norm": 0.8428002595901489,
+      "learning_rate": 2.7240360516683155e-06,
+      "loss": 0.0962,
+      "step": 9023
+    },
+    {
+      "epoch": 1.4620868438107584,
+      "grad_norm": 0.8279069066047668,
+      "learning_rate": 2.723600503621494e-06,
+      "loss": 0.1123,
+      "step": 9024
+    },
+    {
+      "epoch": 1.462248865845755,
+      "grad_norm": 0.8103442788124084,
+      "learning_rate": 2.723164948733038e-06,
+      "loss": 0.1006,
+      "step": 9025
+    },
+    {
+      "epoch": 1.4624108878807518,
+      "grad_norm": 0.8227192163467407,
+      "learning_rate": 2.7227293870162742e-06,
+      "loss": 0.1054,
+      "step": 9026
+    },
+    {
+      "epoch": 1.4625729099157485,
+      "grad_norm": 0.7976576089859009,
+      "learning_rate": 2.7222938184845304e-06,
+      "loss": 0.0935,
+      "step": 9027
+    },
+    {
+      "epoch": 1.4627349319507452,
+      "grad_norm": 0.8542677164077759,
+      "learning_rate": 2.721858243151133e-06,
+      "loss": 0.1126,
+      "step": 9028
+    },
+    {
+      "epoch": 1.4628969539857422,
+      "grad_norm": 0.913292407989502,
+      "learning_rate": 2.7214226610294114e-06,
+      "loss": 0.1131,
+      "step": 9029
+    },
+    {
+      "epoch": 1.4630589760207389,
+      "grad_norm": 0.9596574902534485,
+      "learning_rate": 2.7209870721326915e-06,
+      "loss": 0.1056,
+      "step": 9030
+    },
+    {
+      "epoch": 1.4632209980557356,
+      "grad_norm": 0.8938264846801758,
+      "learning_rate": 2.7205514764743025e-06,
+      "loss": 0.1035,
+      "step": 9031
+    },
+    {
+      "epoch": 1.4633830200907323,
+      "grad_norm": 0.7663517594337463,
+      "learning_rate": 2.7201158740675714e-06,
+      "loss": 0.0973,
+      "step": 9032
+    },
+    {
+      "epoch": 1.463545042125729,
+      "grad_norm": 0.8649742007255554,
+      "learning_rate": 2.7196802649258273e-06,
+      "loss": 0.0919,
+      "step": 9033
+    },
+    {
+      "epoch": 1.463707064160726,
+      "grad_norm": 0.9753785729408264,
+      "learning_rate": 2.719244649062399e-06,
+      "loss": 0.1166,
+      "step": 9034
+    },
+    {
+      "epoch": 1.4638690861957226,
+      "grad_norm": 0.7970572710037231,
+      "learning_rate": 2.7188090264906147e-06,
+      "loss": 0.0942,
+      "step": 9035
+    },
+    {
+      "epoch": 1.4640311082307194,
+      "grad_norm": 0.8879532217979431,
+      "learning_rate": 2.718373397223804e-06,
+      "loss": 0.1078,
+      "step": 9036
+    },
+    {
+      "epoch": 1.464193130265716,
+      "grad_norm": 0.7902569770812988,
+      "learning_rate": 2.7179377612752954e-06,
+      "loss": 0.1028,
+      "step": 9037
+    },
+    {
+      "epoch": 1.4643551523007128,
+      "grad_norm": 0.7548561096191406,
+      "learning_rate": 2.71750211865842e-06,
+      "loss": 0.0948,
+      "step": 9038
+    },
+    {
+      "epoch": 1.4645171743357097,
+      "grad_norm": 0.7446286082267761,
+      "learning_rate": 2.7170664693865045e-06,
+      "loss": 0.0925,
+      "step": 9039
+    },
+    {
+      "epoch": 1.4646791963707064,
+      "grad_norm": 0.7989212274551392,
+      "learning_rate": 2.7166308134728814e-06,
+      "loss": 0.0981,
+      "step": 9040
+    },
+    {
+      "epoch": 1.4648412184057031,
+      "grad_norm": 0.8018403053283691,
+      "learning_rate": 2.7161951509308785e-06,
+      "loss": 0.0979,
+      "step": 9041
+    },
+    {
+      "epoch": 1.4650032404407,
+      "grad_norm": 0.7907342314720154,
+      "learning_rate": 2.715759481773828e-06,
+      "loss": 0.1051,
+      "step": 9042
+    },
+    {
+      "epoch": 1.4651652624756966,
+      "grad_norm": 0.8176236152648926,
+      "learning_rate": 2.7153238060150592e-06,
+      "loss": 0.1035,
+      "step": 9043
+    },
+    {
+      "epoch": 1.4653272845106935,
+      "grad_norm": 0.8785366415977478,
+      "learning_rate": 2.7148881236679035e-06,
+      "loss": 0.1067,
+      "step": 9044
+    },
+    {
+      "epoch": 1.4654893065456902,
+      "grad_norm": 0.8045976161956787,
+      "learning_rate": 2.7144524347456906e-06,
+      "loss": 0.1027,
+      "step": 9045
+    },
+    {
+      "epoch": 1.465651328580687,
+      "grad_norm": 0.7513163685798645,
+      "learning_rate": 2.7140167392617527e-06,
+      "loss": 0.0968,
+      "step": 9046
+    },
+    {
+      "epoch": 1.4658133506156839,
+      "grad_norm": 0.9477952122688293,
+      "learning_rate": 2.7135810372294204e-06,
+      "loss": 0.1219,
+      "step": 9047
+    },
+    {
+      "epoch": 1.4659753726506806,
+      "grad_norm": 0.7233965992927551,
+      "learning_rate": 2.7131453286620253e-06,
+      "loss": 0.0966,
+      "step": 9048
+    },
+    {
+      "epoch": 1.4661373946856773,
+      "grad_norm": 0.8609042763710022,
+      "learning_rate": 2.7127096135728987e-06,
+      "loss": 0.1081,
+      "step": 9049
+    },
+    {
+      "epoch": 1.466299416720674,
+      "grad_norm": 0.7274024486541748,
+      "learning_rate": 2.712273891975372e-06,
+      "loss": 0.097,
+      "step": 9050
+    },
+    {
+      "epoch": 1.4664614387556707,
+      "grad_norm": 0.8457092046737671,
+      "learning_rate": 2.7118381638827795e-06,
+      "loss": 0.1131,
+      "step": 9051
+    },
+    {
+      "epoch": 1.4666234607906676,
+      "grad_norm": 0.8323071599006653,
+      "learning_rate": 2.7114024293084502e-06,
+      "loss": 0.1064,
+      "step": 9052
+    },
+    {
+      "epoch": 1.4667854828256643,
+      "grad_norm": 0.6933931112289429,
+      "learning_rate": 2.710966688265719e-06,
+      "loss": 0.0793,
+      "step": 9053
+    },
+    {
+      "epoch": 1.466947504860661,
+      "grad_norm": 0.8105365633964539,
+      "learning_rate": 2.710530940767917e-06,
+      "loss": 0.0991,
+      "step": 9054
+    },
+    {
+      "epoch": 1.4671095268956578,
+      "grad_norm": 0.7702953815460205,
+      "learning_rate": 2.7100951868283785e-06,
+      "loss": 0.1003,
+      "step": 9055
+    },
+    {
+      "epoch": 1.4672715489306545,
+      "grad_norm": 0.8761657476425171,
+      "learning_rate": 2.7096594264604357e-06,
+      "loss": 0.1091,
+      "step": 9056
+    },
+    {
+      "epoch": 1.4674335709656514,
+      "grad_norm": 0.8410671353340149,
+      "learning_rate": 2.709223659677421e-06,
+      "loss": 0.1059,
+      "step": 9057
+    },
+    {
+      "epoch": 1.4675955930006481,
+      "grad_norm": 0.7424046993255615,
+      "learning_rate": 2.7087878864926696e-06,
+      "loss": 0.0873,
+      "step": 9058
+    },
+    {
+      "epoch": 1.4677576150356448,
+      "grad_norm": 0.8431706428527832,
+      "learning_rate": 2.7083521069195134e-06,
+      "loss": 0.1037,
+      "step": 9059
+    },
+    {
+      "epoch": 1.4679196370706415,
+      "grad_norm": 0.7804480195045471,
+      "learning_rate": 2.707916320971288e-06,
+      "loss": 0.0958,
+      "step": 9060
+    },
+    {
+      "epoch": 1.4680816591056383,
+      "grad_norm": 0.7562078237533569,
+      "learning_rate": 2.707480528661325e-06,
+      "loss": 0.0974,
+      "step": 9061
+    },
+    {
+      "epoch": 1.4682436811406352,
+      "grad_norm": 0.7641584277153015,
+      "learning_rate": 2.7070447300029607e-06,
+      "loss": 0.095,
+      "step": 9062
+    },
+    {
+      "epoch": 1.468405703175632,
+      "grad_norm": 0.9363652467727661,
+      "learning_rate": 2.7066089250095284e-06,
+      "loss": 0.1097,
+      "step": 9063
+    },
+    {
+      "epoch": 1.4685677252106286,
+      "grad_norm": 1.033353567123413,
+      "learning_rate": 2.706173113694363e-06,
+      "loss": 0.1097,
+      "step": 9064
+    },
+    {
+      "epoch": 1.4687297472456253,
+      "grad_norm": 0.8495287299156189,
+      "learning_rate": 2.705737296070799e-06,
+      "loss": 0.1042,
+      "step": 9065
+    },
+    {
+      "epoch": 1.468891769280622,
+      "grad_norm": 0.7492153644561768,
+      "learning_rate": 2.705301472152172e-06,
+      "loss": 0.0886,
+      "step": 9066
+    },
+    {
+      "epoch": 1.469053791315619,
+      "grad_norm": 0.9077339768409729,
+      "learning_rate": 2.7048656419518168e-06,
+      "loss": 0.1071,
+      "step": 9067
+    },
+    {
+      "epoch": 1.4692158133506157,
+      "grad_norm": 0.9256390333175659,
+      "learning_rate": 2.7044298054830687e-06,
+      "loss": 0.1117,
+      "step": 9068
+    },
+    {
+      "epoch": 1.4693778353856124,
+      "grad_norm": 0.8640137910842896,
+      "learning_rate": 2.703993962759263e-06,
+      "loss": 0.1049,
+      "step": 9069
+    },
+    {
+      "epoch": 1.4695398574206093,
+      "grad_norm": 0.8645317554473877,
+      "learning_rate": 2.703558113793736e-06,
+      "loss": 0.1036,
+      "step": 9070
+    },
+    {
+      "epoch": 1.469701879455606,
+      "grad_norm": 0.8858664631843567,
+      "learning_rate": 2.703122258599823e-06,
+      "loss": 0.1113,
+      "step": 9071
+    },
+    {
+      "epoch": 1.4698639014906028,
+      "grad_norm": 0.856381893157959,
+      "learning_rate": 2.7026863971908607e-06,
+      "loss": 0.0984,
+      "step": 9072
+    },
+    {
+      "epoch": 1.4700259235255995,
+      "grad_norm": 0.8103935718536377,
+      "learning_rate": 2.702250529580185e-06,
+      "loss": 0.1053,
+      "step": 9073
+    },
+    {
+      "epoch": 1.4701879455605962,
+      "grad_norm": 0.8574365377426147,
+      "learning_rate": 2.7018146557811325e-06,
+      "loss": 0.0997,
+      "step": 9074
+    },
+    {
+      "epoch": 1.470349967595593,
+      "grad_norm": 0.8630319833755493,
+      "learning_rate": 2.70137877580704e-06,
+      "loss": 0.1014,
+      "step": 9075
+    },
+    {
+      "epoch": 1.4705119896305898,
+      "grad_norm": 0.8821179866790771,
+      "learning_rate": 2.7009428896712443e-06,
+      "loss": 0.1108,
+      "step": 9076
+    },
+    {
+      "epoch": 1.4706740116655865,
+      "grad_norm": 0.8524888157844543,
+      "learning_rate": 2.7005069973870823e-06,
+      "loss": 0.1072,
+      "step": 9077
+    },
+    {
+      "epoch": 1.4708360337005832,
+      "grad_norm": 0.790708601474762,
+      "learning_rate": 2.700071098967892e-06,
+      "loss": 0.0959,
+      "step": 9078
+    },
+    {
+      "epoch": 1.47099805573558,
+      "grad_norm": 0.7711350917816162,
+      "learning_rate": 2.6996351944270096e-06,
+      "loss": 0.0944,
+      "step": 9079
+    },
+    {
+      "epoch": 1.471160077770577,
+      "grad_norm": 0.862248957157135,
+      "learning_rate": 2.699199283777773e-06,
+      "loss": 0.1142,
+      "step": 9080
+    },
+    {
+      "epoch": 1.4713220998055736,
+      "grad_norm": 0.7187860608100891,
+      "learning_rate": 2.698763367033521e-06,
+      "loss": 0.0918,
+      "step": 9081
+    },
+    {
+      "epoch": 1.4714841218405703,
+      "grad_norm": 0.7436991333961487,
+      "learning_rate": 2.6983274442075914e-06,
+      "loss": 0.0882,
+      "step": 9082
+    },
+    {
+      "epoch": 1.471646143875567,
+      "grad_norm": 0.7763605117797852,
+      "learning_rate": 2.6978915153133207e-06,
+      "loss": 0.1046,
+      "step": 9083
+    },
+    {
+      "epoch": 1.4718081659105637,
+      "grad_norm": 0.8496381044387817,
+      "learning_rate": 2.69745558036405e-06,
+      "loss": 0.1087,
+      "step": 9084
+    },
+    {
+      "epoch": 1.4719701879455607,
+      "grad_norm": 0.8233307003974915,
+      "learning_rate": 2.6970196393731146e-06,
+      "loss": 0.1056,
+      "step": 9085
+    },
+    {
+      "epoch": 1.4721322099805574,
+      "grad_norm": 0.806120753288269,
+      "learning_rate": 2.6965836923538568e-06,
+      "loss": 0.1013,
+      "step": 9086
+    },
+    {
+      "epoch": 1.472294232015554,
+      "grad_norm": 0.7680891156196594,
+      "learning_rate": 2.696147739319613e-06,
+      "loss": 0.0902,
+      "step": 9087
+    },
+    {
+      "epoch": 1.4724562540505508,
+      "grad_norm": 0.8429840803146362,
+      "learning_rate": 2.695711780283723e-06,
+      "loss": 0.1087,
+      "step": 9088
+    },
+    {
+      "epoch": 1.4726182760855475,
+      "grad_norm": 0.8204013109207153,
+      "learning_rate": 2.695275815259526e-06,
+      "loss": 0.1044,
+      "step": 9089
+    },
+    {
+      "epoch": 1.4727802981205445,
+      "grad_norm": 0.7914881110191345,
+      "learning_rate": 2.694839844260361e-06,
+      "loss": 0.1021,
+      "step": 9090
+    },
+    {
+      "epoch": 1.4729423201555412,
+      "grad_norm": 0.8489986658096313,
+      "learning_rate": 2.69440386729957e-06,
+      "loss": 0.1064,
+      "step": 9091
+    },
+    {
+      "epoch": 1.4731043421905379,
+      "grad_norm": 0.8216438293457031,
+      "learning_rate": 2.6939678843904897e-06,
+      "loss": 0.1061,
+      "step": 9092
+    },
+    {
+      "epoch": 1.4732663642255348,
+      "grad_norm": 0.8662644028663635,
+      "learning_rate": 2.6935318955464624e-06,
+      "loss": 0.1127,
+      "step": 9093
+    },
+    {
+      "epoch": 1.4734283862605313,
+      "grad_norm": 0.9734071493148804,
+      "learning_rate": 2.6930959007808268e-06,
+      "loss": 0.127,
+      "step": 9094
+    },
+    {
+      "epoch": 1.4735904082955282,
+      "grad_norm": 0.9388130307197571,
+      "learning_rate": 2.692659900106924e-06,
+      "loss": 0.111,
+      "step": 9095
+    },
+    {
+      "epoch": 1.473752430330525,
+      "grad_norm": 0.8008855581283569,
+      "learning_rate": 2.6922238935380946e-06,
+      "loss": 0.1069,
+      "step": 9096
+    },
+    {
+      "epoch": 1.4739144523655217,
+      "grad_norm": 0.8032642602920532,
+      "learning_rate": 2.691787881087679e-06,
+      "loss": 0.0985,
+      "step": 9097
+    },
+    {
+      "epoch": 1.4740764744005186,
+      "grad_norm": 0.7424758672714233,
+      "learning_rate": 2.691351862769018e-06,
+      "loss": 0.1027,
+      "step": 9098
+    },
+    {
+      "epoch": 1.4742384964355153,
+      "grad_norm": 0.8054472208023071,
+      "learning_rate": 2.6909158385954544e-06,
+      "loss": 0.105,
+      "step": 9099
+    },
+    {
+      "epoch": 1.474400518470512,
+      "grad_norm": 0.7597915530204773,
+      "learning_rate": 2.6904798085803276e-06,
+      "loss": 0.0986,
+      "step": 9100
+    },
+    {
+      "epoch": 1.4745625405055087,
+      "grad_norm": 0.7698268294334412,
+      "learning_rate": 2.6900437727369793e-06,
+      "loss": 0.0944,
+      "step": 9101
+    },
+    {
+      "epoch": 1.4747245625405054,
+      "grad_norm": 0.7732874155044556,
+      "learning_rate": 2.689607731078751e-06,
+      "loss": 0.1,
+      "step": 9102
+    },
+    {
+      "epoch": 1.4748865845755024,
+      "grad_norm": 0.7452662587165833,
+      "learning_rate": 2.6891716836189857e-06,
+      "loss": 0.0938,
+      "step": 9103
+    },
+    {
+      "epoch": 1.475048606610499,
+      "grad_norm": 0.7833341956138611,
+      "learning_rate": 2.688735630371024e-06,
+      "loss": 0.1036,
+      "step": 9104
+    },
+    {
+      "epoch": 1.4752106286454958,
+      "grad_norm": 0.7782242894172668,
+      "learning_rate": 2.6882995713482097e-06,
+      "loss": 0.0966,
+      "step": 9105
+    },
+    {
+      "epoch": 1.4753726506804925,
+      "grad_norm": 0.7549180388450623,
+      "learning_rate": 2.6878635065638843e-06,
+      "loss": 0.1022,
+      "step": 9106
+    },
+    {
+      "epoch": 1.4755346727154892,
+      "grad_norm": 0.8578351736068726,
+      "learning_rate": 2.687427436031389e-06,
+      "loss": 0.1039,
+      "step": 9107
+    },
+    {
+      "epoch": 1.4756966947504861,
+      "grad_norm": 0.7809894680976868,
+      "learning_rate": 2.6869913597640686e-06,
+      "loss": 0.101,
+      "step": 9108
+    },
+    {
+      "epoch": 1.4758587167854829,
+      "grad_norm": 0.7138332724571228,
+      "learning_rate": 2.6865552777752644e-06,
+      "loss": 0.0917,
+      "step": 9109
+    },
+    {
+      "epoch": 1.4760207388204796,
+      "grad_norm": 0.8516620397567749,
+      "learning_rate": 2.6861191900783213e-06,
+      "loss": 0.1023,
+      "step": 9110
+    },
+    {
+      "epoch": 1.4761827608554763,
+      "grad_norm": 0.8884662389755249,
+      "learning_rate": 2.6856830966865804e-06,
+      "loss": 0.1067,
+      "step": 9111
+    },
+    {
+      "epoch": 1.476344782890473,
+      "grad_norm": 0.9057988524436951,
+      "learning_rate": 2.685246997613386e-06,
+      "loss": 0.1059,
+      "step": 9112
+    },
+    {
+      "epoch": 1.47650680492547,
+      "grad_norm": 0.7630183696746826,
+      "learning_rate": 2.684810892872083e-06,
+      "loss": 0.1064,
+      "step": 9113
+    },
+    {
+      "epoch": 1.4766688269604666,
+      "grad_norm": 0.751277506351471,
+      "learning_rate": 2.6843747824760125e-06,
+      "loss": 0.0947,
+      "step": 9114
+    },
+    {
+      "epoch": 1.4768308489954634,
+      "grad_norm": 0.8783870935440063,
+      "learning_rate": 2.683938666438521e-06,
+      "loss": 0.1032,
+      "step": 9115
+    },
+    {
+      "epoch": 1.47699287103046,
+      "grad_norm": 0.8420021533966064,
+      "learning_rate": 2.6835025447729495e-06,
+      "loss": 0.0947,
+      "step": 9116
+    },
+    {
+      "epoch": 1.4771548930654568,
+      "grad_norm": 0.7928836941719055,
+      "learning_rate": 2.6830664174926465e-06,
+      "loss": 0.1003,
+      "step": 9117
+    },
+    {
+      "epoch": 1.4773169151004537,
+      "grad_norm": 0.7881826162338257,
+      "learning_rate": 2.682630284610953e-06,
+      "loss": 0.1031,
+      "step": 9118
+    },
+    {
+      "epoch": 1.4774789371354504,
+      "grad_norm": 0.7746166586875916,
+      "learning_rate": 2.682194146141215e-06,
+      "loss": 0.0952,
+      "step": 9119
+    },
+    {
+      "epoch": 1.4776409591704471,
+      "grad_norm": 0.8245477676391602,
+      "learning_rate": 2.6817580020967767e-06,
+      "loss": 0.1023,
+      "step": 9120
+    },
+    {
+      "epoch": 1.477802981205444,
+      "grad_norm": 0.8022861480712891,
+      "learning_rate": 2.6813218524909836e-06,
+      "loss": 0.0994,
+      "step": 9121
+    },
+    {
+      "epoch": 1.4779650032404408,
+      "grad_norm": 0.8441992402076721,
+      "learning_rate": 2.680885697337181e-06,
+      "loss": 0.0999,
+      "step": 9122
+    },
+    {
+      "epoch": 1.4781270252754375,
+      "grad_norm": 0.7806877493858337,
+      "learning_rate": 2.6804495366487132e-06,
+      "loss": 0.1001,
+      "step": 9123
+    },
+    {
+      "epoch": 1.4782890473104342,
+      "grad_norm": 0.6983300447463989,
+      "learning_rate": 2.6800133704389263e-06,
+      "loss": 0.0843,
+      "step": 9124
+    },
+    {
+      "epoch": 1.478451069345431,
+      "grad_norm": 0.8451159596443176,
+      "learning_rate": 2.679577198721166e-06,
+      "loss": 0.1103,
+      "step": 9125
+    },
+    {
+      "epoch": 1.4786130913804278,
+      "grad_norm": 0.7998707890510559,
+      "learning_rate": 2.6791410215087783e-06,
+      "loss": 0.0962,
+      "step": 9126
+    },
+    {
+      "epoch": 1.4787751134154246,
+      "grad_norm": 0.8116523027420044,
+      "learning_rate": 2.678704838815108e-06,
+      "loss": 0.1025,
+      "step": 9127
+    },
+    {
+      "epoch": 1.4789371354504213,
+      "grad_norm": 0.7786074876785278,
+      "learning_rate": 2.678268650653503e-06,
+      "loss": 0.0974,
+      "step": 9128
+    },
+    {
+      "epoch": 1.479099157485418,
+      "grad_norm": 0.8149200081825256,
+      "learning_rate": 2.6778324570373083e-06,
+      "loss": 0.1001,
+      "step": 9129
+    },
+    {
+      "epoch": 1.4792611795204147,
+      "grad_norm": 0.8207477331161499,
+      "learning_rate": 2.6773962579798713e-06,
+      "loss": 0.0947,
+      "step": 9130
+    },
+    {
+      "epoch": 1.4794232015554116,
+      "grad_norm": 0.8058983683586121,
+      "learning_rate": 2.676960053494538e-06,
+      "loss": 0.107,
+      "step": 9131
+    },
+    {
+      "epoch": 1.4795852235904083,
+      "grad_norm": 0.9266466498374939,
+      "learning_rate": 2.6765238435946543e-06,
+      "loss": 0.1122,
+      "step": 9132
+    },
+    {
+      "epoch": 1.479747245625405,
+      "grad_norm": 0.9578770995140076,
+      "learning_rate": 2.676087628293569e-06,
+      "loss": 0.1173,
+      "step": 9133
+    },
+    {
+      "epoch": 1.4799092676604018,
+      "grad_norm": 0.8516537547111511,
+      "learning_rate": 2.675651407604628e-06,
+      "loss": 0.1006,
+      "step": 9134
+    },
+    {
+      "epoch": 1.4800712896953985,
+      "grad_norm": 0.8952048420906067,
+      "learning_rate": 2.6752151815411797e-06,
+      "loss": 0.1053,
+      "step": 9135
+    },
+    {
+      "epoch": 1.4802333117303954,
+      "grad_norm": 0.694008469581604,
+      "learning_rate": 2.67477895011657e-06,
+      "loss": 0.0859,
+      "step": 9136
+    },
+    {
+      "epoch": 1.4803953337653921,
+      "grad_norm": 0.8128898739814758,
+      "learning_rate": 2.674342713344148e-06,
+      "loss": 0.1016,
+      "step": 9137
+    },
+    {
+      "epoch": 1.4805573558003888,
+      "grad_norm": 0.8021711111068726,
+      "learning_rate": 2.6739064712372596e-06,
+      "loss": 0.1034,
+      "step": 9138
+    },
+    {
+      "epoch": 1.4807193778353855,
+      "grad_norm": 0.9167004823684692,
+      "learning_rate": 2.6734702238092557e-06,
+      "loss": 0.1221,
+      "step": 9139
+    },
+    {
+      "epoch": 1.4808813998703823,
+      "grad_norm": 0.8511152863502502,
+      "learning_rate": 2.6730339710734815e-06,
+      "loss": 0.1096,
+      "step": 9140
+    },
+    {
+      "epoch": 1.4810434219053792,
+      "grad_norm": 0.9012813568115234,
+      "learning_rate": 2.6725977130432877e-06,
+      "loss": 0.114,
+      "step": 9141
+    },
+    {
+      "epoch": 1.481205443940376,
+      "grad_norm": 0.7799967527389526,
+      "learning_rate": 2.672161449732021e-06,
+      "loss": 0.0945,
+      "step": 9142
+    },
+    {
+      "epoch": 1.4813674659753726,
+      "grad_norm": 0.8400177359580994,
+      "learning_rate": 2.6717251811530304e-06,
+      "loss": 0.1052,
+      "step": 9143
+    },
+    {
+      "epoch": 1.4815294880103695,
+      "grad_norm": 0.7064500451087952,
+      "learning_rate": 2.671288907319666e-06,
+      "loss": 0.0897,
+      "step": 9144
+    },
+    {
+      "epoch": 1.481691510045366,
+      "grad_norm": 0.8531193733215332,
+      "learning_rate": 2.6708526282452734e-06,
+      "loss": 0.1076,
+      "step": 9145
+    },
+    {
+      "epoch": 1.481853532080363,
+      "grad_norm": 0.8008085489273071,
+      "learning_rate": 2.670416343943205e-06,
+      "loss": 0.1068,
+      "step": 9146
+    },
+    {
+      "epoch": 1.4820155541153597,
+      "grad_norm": 0.7878074645996094,
+      "learning_rate": 2.669980054426809e-06,
+      "loss": 0.0997,
+      "step": 9147
+    },
+    {
+      "epoch": 1.4821775761503564,
+      "grad_norm": 0.8357486724853516,
+      "learning_rate": 2.669543759709434e-06,
+      "loss": 0.1048,
+      "step": 9148
+    },
+    {
+      "epoch": 1.4823395981853533,
+      "grad_norm": 0.7349844574928284,
+      "learning_rate": 2.669107459804431e-06,
+      "loss": 0.0923,
+      "step": 9149
+    },
+    {
+      "epoch": 1.48250162022035,
+      "grad_norm": 0.7902621030807495,
+      "learning_rate": 2.668671154725149e-06,
+      "loss": 0.1033,
+      "step": 9150
+    },
+    {
+      "epoch": 1.4826636422553467,
+      "grad_norm": 0.7515067458152771,
+      "learning_rate": 2.6682348444849373e-06,
+      "loss": 0.0931,
+      "step": 9151
+    },
+    {
+      "epoch": 1.4828256642903435,
+      "grad_norm": 0.798224151134491,
+      "learning_rate": 2.6677985290971464e-06,
+      "loss": 0.102,
+      "step": 9152
+    },
+    {
+      "epoch": 1.4829876863253402,
+      "grad_norm": 0.8462631702423096,
+      "learning_rate": 2.6673622085751275e-06,
+      "loss": 0.0963,
+      "step": 9153
+    },
+    {
+      "epoch": 1.483149708360337,
+      "grad_norm": 0.7932848334312439,
+      "learning_rate": 2.666925882932229e-06,
+      "loss": 0.1096,
+      "step": 9154
+    },
+    {
+      "epoch": 1.4833117303953338,
+      "grad_norm": 0.9925400614738464,
+      "learning_rate": 2.666489552181803e-06,
+      "loss": 0.0995,
+      "step": 9155
+    },
+    {
+      "epoch": 1.4834737524303305,
+      "grad_norm": 0.8233938813209534,
+      "learning_rate": 2.6660532163371995e-06,
+      "loss": 0.097,
+      "step": 9156
+    },
+    {
+      "epoch": 1.4836357744653272,
+      "grad_norm": 0.819572925567627,
+      "learning_rate": 2.6656168754117697e-06,
+      "loss": 0.1022,
+      "step": 9157
+    },
+    {
+      "epoch": 1.483797796500324,
+      "grad_norm": 0.7211984992027283,
+      "learning_rate": 2.665180529418863e-06,
+      "loss": 0.09,
+      "step": 9158
+    },
+    {
+      "epoch": 1.4839598185353209,
+      "grad_norm": 0.7187896966934204,
+      "learning_rate": 2.6647441783718335e-06,
+      "loss": 0.0809,
+      "step": 9159
+    },
+    {
+      "epoch": 1.4841218405703176,
+      "grad_norm": 0.7895532250404358,
+      "learning_rate": 2.6643078222840295e-06,
+      "loss": 0.1055,
+      "step": 9160
+    },
+    {
+      "epoch": 1.4842838626053143,
+      "grad_norm": 0.8991692066192627,
+      "learning_rate": 2.663871461168805e-06,
+      "loss": 0.1035,
+      "step": 9161
+    },
+    {
+      "epoch": 1.484445884640311,
+      "grad_norm": 0.9207788109779358,
+      "learning_rate": 2.6634350950395096e-06,
+      "loss": 0.1109,
+      "step": 9162
+    },
+    {
+      "epoch": 1.4846079066753077,
+      "grad_norm": 0.8399395942687988,
+      "learning_rate": 2.6629987239094956e-06,
+      "loss": 0.1089,
+      "step": 9163
+    },
+    {
+      "epoch": 1.4847699287103047,
+      "grad_norm": 0.8203895688056946,
+      "learning_rate": 2.662562347792116e-06,
+      "loss": 0.0992,
+      "step": 9164
+    },
+    {
+      "epoch": 1.4849319507453014,
+      "grad_norm": 0.7525299787521362,
+      "learning_rate": 2.662125966700721e-06,
+      "loss": 0.0946,
+      "step": 9165
+    },
+    {
+      "epoch": 1.485093972780298,
+      "grad_norm": 0.9387903213500977,
+      "learning_rate": 2.6616895806486644e-06,
+      "loss": 0.118,
+      "step": 9166
+    },
+    {
+      "epoch": 1.4852559948152948,
+      "grad_norm": 0.8505684733390808,
+      "learning_rate": 2.661253189649297e-06,
+      "loss": 0.1121,
+      "step": 9167
+    },
+    {
+      "epoch": 1.4854180168502915,
+      "grad_norm": 0.8152345418930054,
+      "learning_rate": 2.6608167937159735e-06,
+      "loss": 0.1032,
+      "step": 9168
+    },
+    {
+      "epoch": 1.4855800388852884,
+      "grad_norm": 0.7628380060195923,
+      "learning_rate": 2.6603803928620436e-06,
+      "loss": 0.1009,
+      "step": 9169
+    },
+    {
+      "epoch": 1.4857420609202852,
+      "grad_norm": 0.8258995413780212,
+      "learning_rate": 2.6599439871008636e-06,
+      "loss": 0.1027,
+      "step": 9170
+    },
+    {
+      "epoch": 1.4859040829552819,
+      "grad_norm": 0.7622812986373901,
+      "learning_rate": 2.6595075764457838e-06,
+      "loss": 0.0986,
+      "step": 9171
+    },
+    {
+      "epoch": 1.4860661049902788,
+      "grad_norm": 0.7521347999572754,
+      "learning_rate": 2.659071160910158e-06,
+      "loss": 0.0953,
+      "step": 9172
+    },
+    {
+      "epoch": 1.4862281270252755,
+      "grad_norm": 0.7840893864631653,
+      "learning_rate": 2.6586347405073398e-06,
+      "loss": 0.0931,
+      "step": 9173
+    },
+    {
+      "epoch": 1.4863901490602722,
+      "grad_norm": 0.7666772603988647,
+      "learning_rate": 2.6581983152506825e-06,
+      "loss": 0.0981,
+      "step": 9174
+    },
+    {
+      "epoch": 1.486552171095269,
+      "grad_norm": 0.8099150061607361,
+      "learning_rate": 2.65776188515354e-06,
+      "loss": 0.0986,
+      "step": 9175
+    },
+    {
+      "epoch": 1.4867141931302656,
+      "grad_norm": 0.8243248462677002,
+      "learning_rate": 2.6573254502292644e-06,
+      "loss": 0.1039,
+      "step": 9176
+    },
+    {
+      "epoch": 1.4868762151652626,
+      "grad_norm": 0.799664318561554,
+      "learning_rate": 2.6568890104912123e-06,
+      "loss": 0.1003,
+      "step": 9177
+    },
+    {
+      "epoch": 1.4870382372002593,
+      "grad_norm": 0.7624625563621521,
+      "learning_rate": 2.656452565952735e-06,
+      "loss": 0.1001,
+      "step": 9178
+    },
+    {
+      "epoch": 1.487200259235256,
+      "grad_norm": 0.835109531879425,
+      "learning_rate": 2.656016116627188e-06,
+      "loss": 0.1002,
+      "step": 9179
+    },
+    {
+      "epoch": 1.4873622812702527,
+      "grad_norm": 0.7921520471572876,
+      "learning_rate": 2.6555796625279257e-06,
+      "loss": 0.0996,
+      "step": 9180
+    },
+    {
+      "epoch": 1.4875243033052494,
+      "grad_norm": 0.8584132790565491,
+      "learning_rate": 2.6551432036683017e-06,
+      "loss": 0.1059,
+      "step": 9181
+    },
+    {
+      "epoch": 1.4876863253402464,
+      "grad_norm": 0.8519357442855835,
+      "learning_rate": 2.6547067400616717e-06,
+      "loss": 0.0943,
+      "step": 9182
+    },
+    {
+      "epoch": 1.487848347375243,
+      "grad_norm": 0.8248621225357056,
+      "learning_rate": 2.654270271721389e-06,
+      "loss": 0.0998,
+      "step": 9183
+    },
+    {
+      "epoch": 1.4880103694102398,
+      "grad_norm": 0.8023591041564941,
+      "learning_rate": 2.6538337986608105e-06,
+      "loss": 0.0943,
+      "step": 9184
+    },
+    {
+      "epoch": 1.4881723914452365,
+      "grad_norm": 0.8533332347869873,
+      "learning_rate": 2.6533973208932893e-06,
+      "loss": 0.1109,
+      "step": 9185
+    },
+    {
+      "epoch": 1.4883344134802332,
+      "grad_norm": 0.7940263748168945,
+      "learning_rate": 2.6529608384321815e-06,
+      "loss": 0.1073,
+      "step": 9186
+    },
+    {
+      "epoch": 1.4884964355152301,
+      "grad_norm": 0.8785202503204346,
+      "learning_rate": 2.6525243512908423e-06,
+      "loss": 0.1016,
+      "step": 9187
+    },
+    {
+      "epoch": 1.4886584575502269,
+      "grad_norm": 0.7625306844711304,
+      "learning_rate": 2.6520878594826268e-06,
+      "loss": 0.0974,
+      "step": 9188
+    },
+    {
+      "epoch": 1.4888204795852236,
+      "grad_norm": 0.8531996607780457,
+      "learning_rate": 2.6516513630208906e-06,
+      "loss": 0.1111,
+      "step": 9189
+    },
+    {
+      "epoch": 1.4889825016202203,
+      "grad_norm": 0.8253282904624939,
+      "learning_rate": 2.651214861918991e-06,
+      "loss": 0.1047,
+      "step": 9190
+    },
+    {
+      "epoch": 1.489144523655217,
+      "grad_norm": 0.8038217425346375,
+      "learning_rate": 2.6507783561902804e-06,
+      "loss": 0.1079,
+      "step": 9191
+    },
+    {
+      "epoch": 1.489306545690214,
+      "grad_norm": 0.9079601168632507,
+      "learning_rate": 2.6503418458481188e-06,
+      "loss": 0.1085,
+      "step": 9192
+    },
+    {
+      "epoch": 1.4894685677252106,
+      "grad_norm": 0.7232383489608765,
+      "learning_rate": 2.6499053309058593e-06,
+      "loss": 0.0884,
+      "step": 9193
+    },
+    {
+      "epoch": 1.4896305897602073,
+      "grad_norm": 0.7744395136833191,
+      "learning_rate": 2.649468811376861e-06,
+      "loss": 0.099,
+      "step": 9194
+    },
+    {
+      "epoch": 1.4897926117952043,
+      "grad_norm": 0.8593305349349976,
+      "learning_rate": 2.649032287274477e-06,
+      "loss": 0.1131,
+      "step": 9195
+    },
+    {
+      "epoch": 1.4899546338302008,
+      "grad_norm": 0.7993839383125305,
+      "learning_rate": 2.6485957586120664e-06,
+      "loss": 0.1102,
+      "step": 9196
+    },
+    {
+      "epoch": 1.4901166558651977,
+      "grad_norm": 0.7705836296081543,
+      "learning_rate": 2.6481592254029854e-06,
+      "loss": 0.0971,
+      "step": 9197
+    },
+    {
+      "epoch": 1.4902786779001944,
+      "grad_norm": 0.816331148147583,
+      "learning_rate": 2.6477226876605903e-06,
+      "loss": 0.1041,
+      "step": 9198
+    },
+    {
+      "epoch": 1.4904406999351911,
+      "grad_norm": 0.853085458278656,
+      "learning_rate": 2.647286145398239e-06,
+      "loss": 0.1112,
+      "step": 9199
+    },
+    {
+      "epoch": 1.490602721970188,
+      "grad_norm": 0.7312716841697693,
+      "learning_rate": 2.646849598629287e-06,
+      "loss": 0.0957,
+      "step": 9200
+    },
+    {
+      "epoch": 1.4907647440051848,
+      "grad_norm": 0.8262644410133362,
+      "learning_rate": 2.646413047367094e-06,
+      "loss": 0.107,
+      "step": 9201
+    },
+    {
+      "epoch": 1.4909267660401815,
+      "grad_norm": 0.8195114731788635,
+      "learning_rate": 2.645976491625015e-06,
+      "loss": 0.1021,
+      "step": 9202
+    },
+    {
+      "epoch": 1.4910887880751782,
+      "grad_norm": 0.813556432723999,
+      "learning_rate": 2.6455399314164095e-06,
+      "loss": 0.1105,
+      "step": 9203
+    },
+    {
+      "epoch": 1.491250810110175,
+      "grad_norm": 0.8205865025520325,
+      "learning_rate": 2.645103366754633e-06,
+      "loss": 0.1,
+      "step": 9204
+    },
+    {
+      "epoch": 1.4914128321451718,
+      "grad_norm": 0.8455175161361694,
+      "learning_rate": 2.644666797653046e-06,
+      "loss": 0.1036,
+      "step": 9205
+    },
+    {
+      "epoch": 1.4915748541801686,
+      "grad_norm": 0.7989739775657654,
+      "learning_rate": 2.6442302241250047e-06,
+      "loss": 0.1039,
+      "step": 9206
+    },
+    {
+      "epoch": 1.4917368762151653,
+      "grad_norm": 0.8118696808815002,
+      "learning_rate": 2.643793646183867e-06,
+      "loss": 0.0953,
+      "step": 9207
+    },
+    {
+      "epoch": 1.491898898250162,
+      "grad_norm": 0.7257483005523682,
+      "learning_rate": 2.6433570638429923e-06,
+      "loss": 0.0943,
+      "step": 9208
+    },
+    {
+      "epoch": 1.4920609202851587,
+      "grad_norm": 0.6711858510971069,
+      "learning_rate": 2.6429204771157384e-06,
+      "loss": 0.0901,
+      "step": 9209
+    },
+    {
+      "epoch": 1.4922229423201556,
+      "grad_norm": 0.7894310355186462,
+      "learning_rate": 2.6424838860154633e-06,
+      "loss": 0.104,
+      "step": 9210
+    },
+    {
+      "epoch": 1.4923849643551523,
+      "grad_norm": 0.7803335785865784,
+      "learning_rate": 2.6420472905555262e-06,
+      "loss": 0.0985,
+      "step": 9211
+    },
+    {
+      "epoch": 1.492546986390149,
+      "grad_norm": 0.7969242334365845,
+      "learning_rate": 2.641610690749286e-06,
+      "loss": 0.1104,
+      "step": 9212
+    },
+    {
+      "epoch": 1.4927090084251458,
+      "grad_norm": 0.9152027368545532,
+      "learning_rate": 2.641174086610101e-06,
+      "loss": 0.1166,
+      "step": 9213
+    },
+    {
+      "epoch": 1.4928710304601425,
+      "grad_norm": 0.883071780204773,
+      "learning_rate": 2.640737478151331e-06,
+      "loss": 0.114,
+      "step": 9214
+    },
+    {
+      "epoch": 1.4930330524951394,
+      "grad_norm": 0.7122350335121155,
+      "learning_rate": 2.6403008653863343e-06,
+      "loss": 0.0903,
+      "step": 9215
+    },
+    {
+      "epoch": 1.4931950745301361,
+      "grad_norm": 0.8668462634086609,
+      "learning_rate": 2.6398642483284716e-06,
+      "loss": 0.1001,
+      "step": 9216
+    },
+    {
+      "epoch": 1.4933570965651328,
+      "grad_norm": 0.8878929615020752,
+      "learning_rate": 2.6394276269911003e-06,
+      "loss": 0.1123,
+      "step": 9217
+    },
+    {
+      "epoch": 1.4935191186001295,
+      "grad_norm": 0.7911863327026367,
+      "learning_rate": 2.6389910013875814e-06,
+      "loss": 0.1051,
+      "step": 9218
+    },
+    {
+      "epoch": 1.4936811406351262,
+      "grad_norm": 0.9753386378288269,
+      "learning_rate": 2.638554371531274e-06,
+      "loss": 0.0956,
+      "step": 9219
+    },
+    {
+      "epoch": 1.4938431626701232,
+      "grad_norm": 0.7793456315994263,
+      "learning_rate": 2.638117737435538e-06,
+      "loss": 0.0986,
+      "step": 9220
+    },
+    {
+      "epoch": 1.49400518470512,
+      "grad_norm": 0.8406729698181152,
+      "learning_rate": 2.637681099113735e-06,
+      "loss": 0.1014,
+      "step": 9221
+    },
+    {
+      "epoch": 1.4941672067401166,
+      "grad_norm": 0.7772307395935059,
+      "learning_rate": 2.637244456579221e-06,
+      "loss": 0.0949,
+      "step": 9222
+    },
+    {
+      "epoch": 1.4943292287751135,
+      "grad_norm": 0.8179559111595154,
+      "learning_rate": 2.636807809845361e-06,
+      "loss": 0.1041,
+      "step": 9223
+    },
+    {
+      "epoch": 1.4944912508101102,
+      "grad_norm": 0.8753211498260498,
+      "learning_rate": 2.6363711589255115e-06,
+      "loss": 0.1051,
+      "step": 9224
+    },
+    {
+      "epoch": 1.494653272845107,
+      "grad_norm": 0.8662494421005249,
+      "learning_rate": 2.6359345038330352e-06,
+      "loss": 0.1087,
+      "step": 9225
+    },
+    {
+      "epoch": 1.4948152948801037,
+      "grad_norm": 0.7673994898796082,
+      "learning_rate": 2.6354978445812923e-06,
+      "loss": 0.0913,
+      "step": 9226
+    },
+    {
+      "epoch": 1.4949773169151004,
+      "grad_norm": 0.8224777579307556,
+      "learning_rate": 2.6350611811836428e-06,
+      "loss": 0.1055,
+      "step": 9227
+    },
+    {
+      "epoch": 1.4951393389500973,
+      "grad_norm": 0.8199198246002197,
+      "learning_rate": 2.6346245136534483e-06,
+      "loss": 0.0981,
+      "step": 9228
+    },
+    {
+      "epoch": 1.495301360985094,
+      "grad_norm": 0.7925064563751221,
+      "learning_rate": 2.6341878420040694e-06,
+      "loss": 0.0915,
+      "step": 9229
+    },
+    {
+      "epoch": 1.4954633830200907,
+      "grad_norm": 0.7900672554969788,
+      "learning_rate": 2.6337511662488678e-06,
+      "loss": 0.1045,
+      "step": 9230
+    },
+    {
+      "epoch": 1.4956254050550875,
+      "grad_norm": 0.8776764869689941,
+      "learning_rate": 2.6333144864012027e-06,
+      "loss": 0.1074,
+      "step": 9231
+    },
+    {
+      "epoch": 1.4957874270900842,
+      "grad_norm": 0.9464653730392456,
+      "learning_rate": 2.6328778024744384e-06,
+      "loss": 0.1074,
+      "step": 9232
+    },
+    {
+      "epoch": 1.495949449125081,
+      "grad_norm": 0.7300053834915161,
+      "learning_rate": 2.6324411144819345e-06,
+      "loss": 0.0959,
+      "step": 9233
+    },
+    {
+      "epoch": 1.4961114711600778,
+      "grad_norm": 0.793613851070404,
+      "learning_rate": 2.6320044224370526e-06,
+      "loss": 0.1007,
+      "step": 9234
+    },
+    {
+      "epoch": 1.4962734931950745,
+      "grad_norm": 0.7878672480583191,
+      "learning_rate": 2.631567726353155e-06,
+      "loss": 0.1016,
+      "step": 9235
+    },
+    {
+      "epoch": 1.4964355152300712,
+      "grad_norm": 0.7945536971092224,
+      "learning_rate": 2.6311310262436035e-06,
+      "loss": 0.0929,
+      "step": 9236
+    },
+    {
+      "epoch": 1.496597537265068,
+      "grad_norm": 0.9132923483848572,
+      "learning_rate": 2.6306943221217597e-06,
+      "loss": 0.114,
+      "step": 9237
+    },
+    {
+      "epoch": 1.4967595593000649,
+      "grad_norm": 0.9040144681930542,
+      "learning_rate": 2.6302576140009866e-06,
+      "loss": 0.117,
+      "step": 9238
+    },
+    {
+      "epoch": 1.4969215813350616,
+      "grad_norm": 0.8708134889602661,
+      "learning_rate": 2.629820901894645e-06,
+      "loss": 0.1075,
+      "step": 9239
+    },
+    {
+      "epoch": 1.4970836033700583,
+      "grad_norm": 0.7671794295310974,
+      "learning_rate": 2.6293841858160983e-06,
+      "loss": 0.0955,
+      "step": 9240
+    },
+    {
+      "epoch": 1.497245625405055,
+      "grad_norm": 0.7187214493751526,
+      "learning_rate": 2.6289474657787084e-06,
+      "loss": 0.0901,
+      "step": 9241
+    },
+    {
+      "epoch": 1.4974076474400517,
+      "grad_norm": 0.8010789752006531,
+      "learning_rate": 2.6285107417958385e-06,
+      "loss": 0.1064,
+      "step": 9242
+    },
+    {
+      "epoch": 1.4975696694750487,
+      "grad_norm": 0.7064031958580017,
+      "learning_rate": 2.6280740138808503e-06,
+      "loss": 0.0918,
+      "step": 9243
+    },
+    {
+      "epoch": 1.4977316915100454,
+      "grad_norm": 0.7605817317962646,
+      "learning_rate": 2.6276372820471073e-06,
+      "loss": 0.0905,
+      "step": 9244
+    },
+    {
+      "epoch": 1.497893713545042,
+      "grad_norm": 0.8160539269447327,
+      "learning_rate": 2.6272005463079732e-06,
+      "loss": 0.1117,
+      "step": 9245
+    },
+    {
+      "epoch": 1.498055735580039,
+      "grad_norm": 0.7572857141494751,
+      "learning_rate": 2.6267638066768087e-06,
+      "loss": 0.0911,
+      "step": 9246
+    },
+    {
+      "epoch": 1.4982177576150357,
+      "grad_norm": 0.7898995876312256,
+      "learning_rate": 2.6263270631669796e-06,
+      "loss": 0.0998,
+      "step": 9247
+    },
+    {
+      "epoch": 1.4983797796500324,
+      "grad_norm": 0.8465975522994995,
+      "learning_rate": 2.625890315791848e-06,
+      "loss": 0.1061,
+      "step": 9248
+    },
+    {
+      "epoch": 1.4985418016850292,
+      "grad_norm": 0.7732163667678833,
+      "learning_rate": 2.6254535645647772e-06,
+      "loss": 0.0983,
+      "step": 9249
+    },
+    {
+      "epoch": 1.4987038237200259,
+      "grad_norm": 0.7840855717658997,
+      "learning_rate": 2.625016809499131e-06,
+      "loss": 0.1021,
+      "step": 9250
+    },
+    {
+      "epoch": 1.4988658457550228,
+      "grad_norm": 0.7952522039413452,
+      "learning_rate": 2.6245800506082727e-06,
+      "loss": 0.0972,
+      "step": 9251
+    },
+    {
+      "epoch": 1.4990278677900195,
+      "grad_norm": 0.7457720637321472,
+      "learning_rate": 2.6241432879055667e-06,
+      "loss": 0.0979,
+      "step": 9252
+    },
+    {
+      "epoch": 1.4991898898250162,
+      "grad_norm": 0.8276432752609253,
+      "learning_rate": 2.6237065214043754e-06,
+      "loss": 0.108,
+      "step": 9253
+    },
+    {
+      "epoch": 1.499351911860013,
+      "grad_norm": 0.7755012512207031,
+      "learning_rate": 2.6232697511180654e-06,
+      "loss": 0.0939,
+      "step": 9254
+    },
+    {
+      "epoch": 1.4995139338950096,
+      "grad_norm": 0.7345067858695984,
+      "learning_rate": 2.622832977059998e-06,
+      "loss": 0.0905,
+      "step": 9255
+    },
+    {
+      "epoch": 1.4996759559300066,
+      "grad_norm": 0.8585684299468994,
+      "learning_rate": 2.6223961992435406e-06,
+      "loss": 0.1049,
+      "step": 9256
+    },
+    {
+      "epoch": 1.4998379779650033,
+      "grad_norm": 0.729276180267334,
+      "learning_rate": 2.621959417682054e-06,
+      "loss": 0.0873,
+      "step": 9257
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8658743500709534,
+      "learning_rate": 2.6215226323889048e-06,
+      "loss": 0.1089,
+      "step": 9258
+    },
+    {
+      "epoch": 1.5001620220349967,
+      "grad_norm": 0.7544455528259277,
+      "learning_rate": 2.621085843377457e-06,
+      "loss": 0.0943,
+      "step": 9259
+    },
+    {
+      "epoch": 1.5003240440699934,
+      "grad_norm": 0.7996721863746643,
+      "learning_rate": 2.620649050661076e-06,
+      "loss": 0.0983,
+      "step": 9260
+    },
+    {
+      "epoch": 1.5004860661049904,
+      "grad_norm": 0.8732225894927979,
+      "learning_rate": 2.620212254253126e-06,
+      "loss": 0.1048,
+      "step": 9261
+    },
+    {
+      "epoch": 1.500648088139987,
+      "grad_norm": 0.8265805244445801,
+      "learning_rate": 2.6197754541669714e-06,
+      "loss": 0.1041,
+      "step": 9262
+    },
+    {
+      "epoch": 1.5008101101749838,
+      "grad_norm": 0.8330001831054688,
+      "learning_rate": 2.6193386504159777e-06,
+      "loss": 0.1099,
+      "step": 9263
+    },
+    {
+      "epoch": 1.5009721322099807,
+      "grad_norm": 0.780525267124176,
+      "learning_rate": 2.6189018430135106e-06,
+      "loss": 0.0999,
+      "step": 9264
+    },
+    {
+      "epoch": 1.5011341542449772,
+      "grad_norm": 0.9457441568374634,
+      "learning_rate": 2.618465031972935e-06,
+      "loss": 0.1189,
+      "step": 9265
+    },
+    {
+      "epoch": 1.5012961762799741,
+      "grad_norm": 0.8373966813087463,
+      "learning_rate": 2.6180282173076156e-06,
+      "loss": 0.1104,
+      "step": 9266
+    },
+    {
+      "epoch": 1.5014581983149708,
+      "grad_norm": 0.7619836330413818,
+      "learning_rate": 2.6175913990309184e-06,
+      "loss": 0.0991,
+      "step": 9267
+    },
+    {
+      "epoch": 1.5016202203499676,
+      "grad_norm": 0.7752346992492676,
+      "learning_rate": 2.6171545771562085e-06,
+      "loss": 0.0955,
+      "step": 9268
+    },
+    {
+      "epoch": 1.5017822423849645,
+      "grad_norm": 0.8263670206069946,
+      "learning_rate": 2.6167177516968536e-06,
+      "loss": 0.0997,
+      "step": 9269
+    },
+    {
+      "epoch": 1.501944264419961,
+      "grad_norm": 0.796293318271637,
+      "learning_rate": 2.6162809226662167e-06,
+      "loss": 0.1056,
+      "step": 9270
+    },
+    {
+      "epoch": 1.502106286454958,
+      "grad_norm": 0.7782325148582458,
+      "learning_rate": 2.615844090077665e-06,
+      "loss": 0.0962,
+      "step": 9271
+    },
+    {
+      "epoch": 1.5022683084899546,
+      "grad_norm": 0.9106411337852478,
+      "learning_rate": 2.6154072539445645e-06,
+      "loss": 0.1008,
+      "step": 9272
+    },
+    {
+      "epoch": 1.5024303305249513,
+      "grad_norm": 0.8080524802207947,
+      "learning_rate": 2.614970414280281e-06,
+      "loss": 0.1053,
+      "step": 9273
+    },
+    {
+      "epoch": 1.5025923525599483,
+      "grad_norm": 0.9011566638946533,
+      "learning_rate": 2.6145335710981817e-06,
+      "loss": 0.1223,
+      "step": 9274
+    },
+    {
+      "epoch": 1.5027543745949448,
+      "grad_norm": 0.7807307243347168,
+      "learning_rate": 2.6140967244116322e-06,
+      "loss": 0.0926,
+      "step": 9275
+    },
+    {
+      "epoch": 1.5029163966299417,
+      "grad_norm": 0.8064615726470947,
+      "learning_rate": 2.613659874233999e-06,
+      "loss": 0.0976,
+      "step": 9276
+    },
+    {
+      "epoch": 1.5030784186649384,
+      "grad_norm": 0.8541350960731506,
+      "learning_rate": 2.6132230205786483e-06,
+      "loss": 0.1034,
+      "step": 9277
+    },
+    {
+      "epoch": 1.5032404406999351,
+      "grad_norm": 0.8118453621864319,
+      "learning_rate": 2.612786163458948e-06,
+      "loss": 0.1009,
+      "step": 9278
+    },
+    {
+      "epoch": 1.503402462734932,
+      "grad_norm": 0.852788507938385,
+      "learning_rate": 2.6123493028882634e-06,
+      "loss": 0.1146,
+      "step": 9279
+    },
+    {
+      "epoch": 1.5035644847699285,
+      "grad_norm": 0.7566162347793579,
+      "learning_rate": 2.611912438879962e-06,
+      "loss": 0.0975,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5037265068049255,
+      "grad_norm": 0.9245677590370178,
+      "learning_rate": 2.611475571447411e-06,
+      "loss": 0.1078,
+      "step": 9281
+    },
+    {
+      "epoch": 1.5038885288399222,
+      "grad_norm": 0.9976536631584167,
+      "learning_rate": 2.611038700603977e-06,
+      "loss": 0.116,
+      "step": 9282
+    },
+    {
+      "epoch": 1.504050550874919,
+      "grad_norm": 0.8083858489990234,
+      "learning_rate": 2.6106018263630283e-06,
+      "loss": 0.1036,
+      "step": 9283
+    },
+    {
+      "epoch": 1.5042125729099158,
+      "grad_norm": 0.7877292037010193,
+      "learning_rate": 2.6101649487379304e-06,
+      "loss": 0.0926,
+      "step": 9284
+    },
+    {
+      "epoch": 1.5043745949449125,
+      "grad_norm": 0.9315115809440613,
+      "learning_rate": 2.609728067742053e-06,
+      "loss": 0.1173,
+      "step": 9285
+    },
+    {
+      "epoch": 1.5045366169799093,
+      "grad_norm": 0.8206225037574768,
+      "learning_rate": 2.6092911833887602e-06,
+      "loss": 0.1077,
+      "step": 9286
+    },
+    {
+      "epoch": 1.5046986390149062,
+      "grad_norm": 0.805716335773468,
+      "learning_rate": 2.6088542956914233e-06,
+      "loss": 0.1041,
+      "step": 9287
+    },
+    {
+      "epoch": 1.5048606610499027,
+      "grad_norm": 0.9114910364151001,
+      "learning_rate": 2.6084174046634075e-06,
+      "loss": 0.1181,
+      "step": 9288
+    },
+    {
+      "epoch": 1.5050226830848996,
+      "grad_norm": 0.8441805243492126,
+      "learning_rate": 2.607980510318082e-06,
+      "loss": 0.1069,
+      "step": 9289
+    },
+    {
+      "epoch": 1.5051847051198963,
+      "grad_norm": 0.8596044182777405,
+      "learning_rate": 2.607543612668814e-06,
+      "loss": 0.101,
+      "step": 9290
+    },
+    {
+      "epoch": 1.505346727154893,
+      "grad_norm": 0.8306862711906433,
+      "learning_rate": 2.6071067117289717e-06,
+      "loss": 0.1047,
+      "step": 9291
+    },
+    {
+      "epoch": 1.50550874918989,
+      "grad_norm": 0.7705062031745911,
+      "learning_rate": 2.6066698075119237e-06,
+      "loss": 0.0974,
+      "step": 9292
+    },
+    {
+      "epoch": 1.5056707712248865,
+      "grad_norm": 0.8047751784324646,
+      "learning_rate": 2.606232900031037e-06,
+      "loss": 0.1029,
+      "step": 9293
+    },
+    {
+      "epoch": 1.5058327932598834,
+      "grad_norm": 0.7562994360923767,
+      "learning_rate": 2.605795989299681e-06,
+      "loss": 0.0957,
+      "step": 9294
+    },
+    {
+      "epoch": 1.50599481529488,
+      "grad_norm": 0.9327423572540283,
+      "learning_rate": 2.6053590753312237e-06,
+      "loss": 0.0989,
+      "step": 9295
+    },
+    {
+      "epoch": 1.5061568373298768,
+      "grad_norm": 0.7977178692817688,
+      "learning_rate": 2.604922158139033e-06,
+      "loss": 0.1054,
+      "step": 9296
+    },
+    {
+      "epoch": 1.5063188593648738,
+      "grad_norm": 0.7948964238166809,
+      "learning_rate": 2.6044852377364794e-06,
+      "loss": 0.0938,
+      "step": 9297
+    },
+    {
+      "epoch": 1.5064808813998702,
+      "grad_norm": 0.8069312572479248,
+      "learning_rate": 2.6040483141369293e-06,
+      "loss": 0.0934,
+      "step": 9298
+    },
+    {
+      "epoch": 1.5066429034348672,
+      "grad_norm": 0.850750744342804,
+      "learning_rate": 2.6036113873537526e-06,
+      "loss": 0.0958,
+      "step": 9299
+    },
+    {
+      "epoch": 1.5068049254698639,
+      "grad_norm": 0.8110750317573547,
+      "learning_rate": 2.603174457400319e-06,
+      "loss": 0.1068,
+      "step": 9300
+    },
+    {
+      "epoch": 1.5069669475048606,
+      "grad_norm": 0.8430596590042114,
+      "learning_rate": 2.602737524289996e-06,
+      "loss": 0.0992,
+      "step": 9301
+    },
+    {
+      "epoch": 1.5071289695398575,
+      "grad_norm": 0.8320333361625671,
+      "learning_rate": 2.602300588036154e-06,
+      "loss": 0.1031,
+      "step": 9302
+    },
+    {
+      "epoch": 1.507290991574854,
+      "grad_norm": 0.8486113548278809,
+      "learning_rate": 2.6018636486521615e-06,
+      "loss": 0.1004,
+      "step": 9303
+    },
+    {
+      "epoch": 1.507453013609851,
+      "grad_norm": 0.7198436260223389,
+      "learning_rate": 2.6014267061513875e-06,
+      "loss": 0.0936,
+      "step": 9304
+    },
+    {
+      "epoch": 1.5076150356448477,
+      "grad_norm": 0.7572947144508362,
+      "learning_rate": 2.6009897605472022e-06,
+      "loss": 0.094,
+      "step": 9305
+    },
+    {
+      "epoch": 1.5077770576798444,
+      "grad_norm": 0.8611274361610413,
+      "learning_rate": 2.6005528118529738e-06,
+      "loss": 0.1026,
+      "step": 9306
+    },
+    {
+      "epoch": 1.5079390797148413,
+      "grad_norm": 0.904730498790741,
+      "learning_rate": 2.6001158600820735e-06,
+      "loss": 0.1111,
+      "step": 9307
+    },
+    {
+      "epoch": 1.508101101749838,
+      "grad_norm": 0.8339022994041443,
+      "learning_rate": 2.5996789052478693e-06,
+      "loss": 0.1016,
+      "step": 9308
+    },
+    {
+      "epoch": 1.5082631237848347,
+      "grad_norm": 0.8966616988182068,
+      "learning_rate": 2.5992419473637327e-06,
+      "loss": 0.1131,
+      "step": 9309
+    },
+    {
+      "epoch": 1.5084251458198314,
+      "grad_norm": 0.7851499319076538,
+      "learning_rate": 2.5988049864430314e-06,
+      "loss": 0.1015,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5085871678548282,
+      "grad_norm": 0.8004862666130066,
+      "learning_rate": 2.598368022499138e-06,
+      "loss": 0.0977,
+      "step": 9311
+    },
+    {
+      "epoch": 1.508749189889825,
+      "grad_norm": 0.7532079219818115,
+      "learning_rate": 2.597931055545421e-06,
+      "loss": 0.0939,
+      "step": 9312
+    },
+    {
+      "epoch": 1.5089112119248218,
+      "grad_norm": 0.8671373128890991,
+      "learning_rate": 2.59749408559525e-06,
+      "loss": 0.1082,
+      "step": 9313
+    },
+    {
+      "epoch": 1.5090732339598185,
+      "grad_norm": 0.7479811906814575,
+      "learning_rate": 2.597057112661997e-06,
+      "loss": 0.0937,
+      "step": 9314
+    },
+    {
+      "epoch": 1.5092352559948155,
+      "grad_norm": 0.7849185466766357,
+      "learning_rate": 2.59662013675903e-06,
+      "loss": 0.1082,
+      "step": 9315
+    },
+    {
+      "epoch": 1.509397278029812,
+      "grad_norm": 0.7421824336051941,
+      "learning_rate": 2.5961831578997214e-06,
+      "loss": 0.1007,
+      "step": 9316
+    },
+    {
+      "epoch": 1.5095593000648089,
+      "grad_norm": 0.752606213092804,
+      "learning_rate": 2.5957461760974407e-06,
+      "loss": 0.0929,
+      "step": 9317
+    },
+    {
+      "epoch": 1.5097213220998056,
+      "grad_norm": 0.8576077222824097,
+      "learning_rate": 2.5953091913655586e-06,
+      "loss": 0.1081,
+      "step": 9318
+    },
+    {
+      "epoch": 1.5098833441348023,
+      "grad_norm": 0.7942414879798889,
+      "learning_rate": 2.594872203717446e-06,
+      "loss": 0.1102,
+      "step": 9319
+    },
+    {
+      "epoch": 1.5100453661697992,
+      "grad_norm": 0.7361174821853638,
+      "learning_rate": 2.594435213166473e-06,
+      "loss": 0.0973,
+      "step": 9320
+    },
+    {
+      "epoch": 1.5102073882047957,
+      "grad_norm": 0.7839969992637634,
+      "learning_rate": 2.5939982197260115e-06,
+      "loss": 0.0995,
+      "step": 9321
+    },
+    {
+      "epoch": 1.5103694102397927,
+      "grad_norm": 0.9291492104530334,
+      "learning_rate": 2.593561223409432e-06,
+      "loss": 0.1089,
+      "step": 9322
+    },
+    {
+      "epoch": 1.5105314322747894,
+      "grad_norm": 0.8989511132240295,
+      "learning_rate": 2.5931242242301054e-06,
+      "loss": 0.1078,
+      "step": 9323
+    },
+    {
+      "epoch": 1.510693454309786,
+      "grad_norm": 0.8241267800331116,
+      "learning_rate": 2.592687222201403e-06,
+      "loss": 0.1092,
+      "step": 9324
+    },
+    {
+      "epoch": 1.510855476344783,
+      "grad_norm": 0.7994397878646851,
+      "learning_rate": 2.592250217336696e-06,
+      "loss": 0.1023,
+      "step": 9325
+    },
+    {
+      "epoch": 1.5110174983797795,
+      "grad_norm": 0.777235746383667,
+      "learning_rate": 2.5918132096493552e-06,
+      "loss": 0.0913,
+      "step": 9326
+    },
+    {
+      "epoch": 1.5111795204147764,
+      "grad_norm": 0.7653077840805054,
+      "learning_rate": 2.5913761991527527e-06,
+      "loss": 0.0998,
+      "step": 9327
+    },
+    {
+      "epoch": 1.5113415424497731,
+      "grad_norm": 0.9247298836708069,
+      "learning_rate": 2.5909391858602596e-06,
+      "loss": 0.114,
+      "step": 9328
+    },
+    {
+      "epoch": 1.5115035644847699,
+      "grad_norm": 0.8177992701530457,
+      "learning_rate": 2.590502169785247e-06,
+      "loss": 0.0999,
+      "step": 9329
+    },
+    {
+      "epoch": 1.5116655865197668,
+      "grad_norm": 0.7962645292282104,
+      "learning_rate": 2.5900651509410875e-06,
+      "loss": 0.1064,
+      "step": 9330
+    },
+    {
+      "epoch": 1.5118276085547635,
+      "grad_norm": 0.8084893822669983,
+      "learning_rate": 2.589628129341153e-06,
+      "loss": 0.1103,
+      "step": 9331
+    },
+    {
+      "epoch": 1.5119896305897602,
+      "grad_norm": 0.7229411005973816,
+      "learning_rate": 2.5891911049988133e-06,
+      "loss": 0.089,
+      "step": 9332
+    },
+    {
+      "epoch": 1.512151652624757,
+      "grad_norm": 1.190307855606079,
+      "learning_rate": 2.588754077927442e-06,
+      "loss": 0.1183,
+      "step": 9333
+    },
+    {
+      "epoch": 1.5123136746597536,
+      "grad_norm": 0.746012806892395,
+      "learning_rate": 2.5883170481404112e-06,
+      "loss": 0.0924,
+      "step": 9334
+    },
+    {
+      "epoch": 1.5124756966947506,
+      "grad_norm": 0.8450835347175598,
+      "learning_rate": 2.5878800156510925e-06,
+      "loss": 0.1091,
+      "step": 9335
+    },
+    {
+      "epoch": 1.5126377187297473,
+      "grad_norm": 0.7555860280990601,
+      "learning_rate": 2.587442980472858e-06,
+      "loss": 0.0875,
+      "step": 9336
+    },
+    {
+      "epoch": 1.512799740764744,
+      "grad_norm": 0.8183492422103882,
+      "learning_rate": 2.5870059426190787e-06,
+      "loss": 0.0995,
+      "step": 9337
+    },
+    {
+      "epoch": 1.512961762799741,
+      "grad_norm": 0.8046535849571228,
+      "learning_rate": 2.5865689021031292e-06,
+      "loss": 0.0967,
+      "step": 9338
+    },
+    {
+      "epoch": 1.5131237848347374,
+      "grad_norm": 0.8389936685562134,
+      "learning_rate": 2.5861318589383806e-06,
+      "loss": 0.0875,
+      "step": 9339
+    },
+    {
+      "epoch": 1.5132858068697344,
+      "grad_norm": 0.9016577005386353,
+      "learning_rate": 2.5856948131382055e-06,
+      "loss": 0.1099,
+      "step": 9340
+    },
+    {
+      "epoch": 1.513447828904731,
+      "grad_norm": 0.8551927804946899,
+      "learning_rate": 2.585257764715976e-06,
+      "loss": 0.1044,
+      "step": 9341
+    },
+    {
+      "epoch": 1.5136098509397278,
+      "grad_norm": 0.8470696210861206,
+      "learning_rate": 2.584820713685066e-06,
+      "loss": 0.1048,
+      "step": 9342
+    },
+    {
+      "epoch": 1.5137718729747247,
+      "grad_norm": 0.9146429896354675,
+      "learning_rate": 2.5843836600588474e-06,
+      "loss": 0.0846,
+      "step": 9343
+    },
+    {
+      "epoch": 1.5139338950097212,
+      "grad_norm": 0.8819512128829956,
+      "learning_rate": 2.5839466038506927e-06,
+      "loss": 0.1131,
+      "step": 9344
+    },
+    {
+      "epoch": 1.5140959170447181,
+      "grad_norm": 0.7611178755760193,
+      "learning_rate": 2.583509545073975e-06,
+      "loss": 0.0911,
+      "step": 9345
+    },
+    {
+      "epoch": 1.5142579390797148,
+      "grad_norm": 0.8343874216079712,
+      "learning_rate": 2.5830724837420675e-06,
+      "loss": 0.1025,
+      "step": 9346
+    },
+    {
+      "epoch": 1.5144199611147116,
+      "grad_norm": 0.8855875134468079,
+      "learning_rate": 2.5826354198683433e-06,
+      "loss": 0.1137,
+      "step": 9347
+    },
+    {
+      "epoch": 1.5145819831497085,
+      "grad_norm": 0.8919705152511597,
+      "learning_rate": 2.582198353466175e-06,
+      "loss": 0.1163,
+      "step": 9348
+    },
+    {
+      "epoch": 1.514744005184705,
+      "grad_norm": 0.7063683271408081,
+      "learning_rate": 2.5817612845489355e-06,
+      "loss": 0.0853,
+      "step": 9349
+    },
+    {
+      "epoch": 1.514906027219702,
+      "grad_norm": 0.9223026633262634,
+      "learning_rate": 2.5813242131299986e-06,
+      "loss": 0.1116,
+      "step": 9350
+    },
+    {
+      "epoch": 1.5150680492546986,
+      "grad_norm": 0.7677808403968811,
+      "learning_rate": 2.580887139222738e-06,
+      "loss": 0.0966,
+      "step": 9351
+    },
+    {
+      "epoch": 1.5152300712896953,
+      "grad_norm": 0.8146395087242126,
+      "learning_rate": 2.5804500628405265e-06,
+      "loss": 0.0944,
+      "step": 9352
+    },
+    {
+      "epoch": 1.5153920933246923,
+      "grad_norm": 0.804042637348175,
+      "learning_rate": 2.580012983996738e-06,
+      "loss": 0.1033,
+      "step": 9353
+    },
+    {
+      "epoch": 1.5155541153596888,
+      "grad_norm": 0.9203242659568787,
+      "learning_rate": 2.5795759027047457e-06,
+      "loss": 0.1108,
+      "step": 9354
+    },
+    {
+      "epoch": 1.5157161373946857,
+      "grad_norm": 0.8421667218208313,
+      "learning_rate": 2.5791388189779237e-06,
+      "loss": 0.0992,
+      "step": 9355
+    },
+    {
+      "epoch": 1.5158781594296824,
+      "grad_norm": 0.8935291767120361,
+      "learning_rate": 2.578701732829645e-06,
+      "loss": 0.1045,
+      "step": 9356
+    },
+    {
+      "epoch": 1.5160401814646791,
+      "grad_norm": 0.8606316447257996,
+      "learning_rate": 2.5782646442732833e-06,
+      "loss": 0.1096,
+      "step": 9357
+    },
+    {
+      "epoch": 1.516202203499676,
+      "grad_norm": 0.8459725975990295,
+      "learning_rate": 2.5778275533222135e-06,
+      "loss": 0.1099,
+      "step": 9358
+    },
+    {
+      "epoch": 1.5163642255346728,
+      "grad_norm": 0.7800662517547607,
+      "learning_rate": 2.5773904599898087e-06,
+      "loss": 0.0971,
+      "step": 9359
+    },
+    {
+      "epoch": 1.5165262475696695,
+      "grad_norm": 0.8285083770751953,
+      "learning_rate": 2.5769533642894433e-06,
+      "loss": 0.1019,
+      "step": 9360
+    },
+    {
+      "epoch": 1.5166882696046662,
+      "grad_norm": 0.8594668507575989,
+      "learning_rate": 2.5765162662344906e-06,
+      "loss": 0.1067,
+      "step": 9361
+    },
+    {
+      "epoch": 1.516850291639663,
+      "grad_norm": 0.7574124932289124,
+      "learning_rate": 2.576079165838326e-06,
+      "loss": 0.0894,
+      "step": 9362
+    },
+    {
+      "epoch": 1.5170123136746598,
+      "grad_norm": 0.8705188632011414,
+      "learning_rate": 2.575642063114322e-06,
+      "loss": 0.1104,
+      "step": 9363
+    },
+    {
+      "epoch": 1.5171743357096565,
+      "grad_norm": 0.8245108723640442,
+      "learning_rate": 2.5752049580758555e-06,
+      "loss": 0.104,
+      "step": 9364
+    },
+    {
+      "epoch": 1.5173363577446533,
+      "grad_norm": 0.9124253988265991,
+      "learning_rate": 2.5747678507362977e-06,
+      "loss": 0.1215,
+      "step": 9365
+    },
+    {
+      "epoch": 1.5174983797796502,
+      "grad_norm": 0.8030577898025513,
+      "learning_rate": 2.5743307411090255e-06,
+      "loss": 0.1064,
+      "step": 9366
+    },
+    {
+      "epoch": 1.5176604018146467,
+      "grad_norm": 0.7944320440292358,
+      "learning_rate": 2.5738936292074122e-06,
+      "loss": 0.0991,
+      "step": 9367
+    },
+    {
+      "epoch": 1.5178224238496436,
+      "grad_norm": 0.7724756598472595,
+      "learning_rate": 2.5734565150448325e-06,
+      "loss": 0.1032,
+      "step": 9368
+    },
+    {
+      "epoch": 1.5179844458846403,
+      "grad_norm": 0.7421293258666992,
+      "learning_rate": 2.5730193986346623e-06,
+      "loss": 0.0941,
+      "step": 9369
+    },
+    {
+      "epoch": 1.518146467919637,
+      "grad_norm": 0.7390490770339966,
+      "learning_rate": 2.5725822799902738e-06,
+      "loss": 0.0941,
+      "step": 9370
+    },
+    {
+      "epoch": 1.518308489954634,
+      "grad_norm": 0.8915257453918457,
+      "learning_rate": 2.572145159125044e-06,
+      "loss": 0.1067,
+      "step": 9371
+    },
+    {
+      "epoch": 1.5184705119896305,
+      "grad_norm": 0.8176239132881165,
+      "learning_rate": 2.5717080360523464e-06,
+      "loss": 0.1001,
+      "step": 9372
+    },
+    {
+      "epoch": 1.5186325340246274,
+      "grad_norm": 0.7526306509971619,
+      "learning_rate": 2.571270910785557e-06,
+      "loss": 0.1013,
+      "step": 9373
+    },
+    {
+      "epoch": 1.518794556059624,
+      "grad_norm": 0.7733591794967651,
+      "learning_rate": 2.57083378333805e-06,
+      "loss": 0.1024,
+      "step": 9374
+    },
+    {
+      "epoch": 1.5189565780946208,
+      "grad_norm": 0.7548816204071045,
+      "learning_rate": 2.5703966537232006e-06,
+      "loss": 0.104,
+      "step": 9375
+    },
+    {
+      "epoch": 1.5191186001296177,
+      "grad_norm": 0.8029460906982422,
+      "learning_rate": 2.5699595219543838e-06,
+      "loss": 0.1031,
+      "step": 9376
+    },
+    {
+      "epoch": 1.5192806221646142,
+      "grad_norm": 0.912986695766449,
+      "learning_rate": 2.569522388044975e-06,
+      "loss": 0.1102,
+      "step": 9377
+    },
+    {
+      "epoch": 1.5194426441996112,
+      "grad_norm": 0.8626551032066345,
+      "learning_rate": 2.5690852520083496e-06,
+      "loss": 0.1074,
+      "step": 9378
+    },
+    {
+      "epoch": 1.5196046662346079,
+      "grad_norm": 0.7937400937080383,
+      "learning_rate": 2.5686481138578824e-06,
+      "loss": 0.0928,
+      "step": 9379
+    },
+    {
+      "epoch": 1.5197666882696046,
+      "grad_norm": 0.7942370772361755,
+      "learning_rate": 2.5682109736069492e-06,
+      "loss": 0.101,
+      "step": 9380
+    },
+    {
+      "epoch": 1.5199287103046015,
+      "grad_norm": 0.8187708854675293,
+      "learning_rate": 2.5677738312689248e-06,
+      "loss": 0.1009,
+      "step": 9381
+    },
+    {
+      "epoch": 1.5200907323395982,
+      "grad_norm": 1.0334949493408203,
+      "learning_rate": 2.5673366868571858e-06,
+      "loss": 0.1277,
+      "step": 9382
+    },
+    {
+      "epoch": 1.520252754374595,
+      "grad_norm": 0.7634156942367554,
+      "learning_rate": 2.5668995403851065e-06,
+      "loss": 0.0979,
+      "step": 9383
+    },
+    {
+      "epoch": 1.5204147764095917,
+      "grad_norm": 0.7355403304100037,
+      "learning_rate": 2.566462391866064e-06,
+      "loss": 0.0939,
+      "step": 9384
+    },
+    {
+      "epoch": 1.5205767984445884,
+      "grad_norm": 0.8203917145729065,
+      "learning_rate": 2.5660252413134323e-06,
+      "loss": 0.103,
+      "step": 9385
+    },
+    {
+      "epoch": 1.5207388204795853,
+      "grad_norm": 0.8019030690193176,
+      "learning_rate": 2.5655880887405893e-06,
+      "loss": 0.1089,
+      "step": 9386
+    },
+    {
+      "epoch": 1.520900842514582,
+      "grad_norm": 0.8758410215377808,
+      "learning_rate": 2.565150934160908e-06,
+      "loss": 0.1171,
+      "step": 9387
+    },
+    {
+      "epoch": 1.5210628645495787,
+      "grad_norm": 0.8551106452941895,
+      "learning_rate": 2.564713777587767e-06,
+      "loss": 0.1006,
+      "step": 9388
+    },
+    {
+      "epoch": 1.5212248865845757,
+      "grad_norm": 0.7470712065696716,
+      "learning_rate": 2.5642766190345396e-06,
+      "loss": 0.0948,
+      "step": 9389
+    },
+    {
+      "epoch": 1.5213869086195722,
+      "grad_norm": 0.845240592956543,
+      "learning_rate": 2.5638394585146044e-06,
+      "loss": 0.0975,
+      "step": 9390
+    },
+    {
+      "epoch": 1.521548930654569,
+      "grad_norm": 0.9042078256607056,
+      "learning_rate": 2.5634022960413362e-06,
+      "loss": 0.1057,
+      "step": 9391
+    },
+    {
+      "epoch": 1.5217109526895658,
+      "grad_norm": 0.844704270362854,
+      "learning_rate": 2.56296513162811e-06,
+      "loss": 0.1008,
+      "step": 9392
+    },
+    {
+      "epoch": 1.5218729747245625,
+      "grad_norm": 0.8592380881309509,
+      "learning_rate": 2.5625279652883043e-06,
+      "loss": 0.1054,
+      "step": 9393
+    },
+    {
+      "epoch": 1.5220349967595594,
+      "grad_norm": 0.7636672258377075,
+      "learning_rate": 2.5620907970352937e-06,
+      "loss": 0.0947,
+      "step": 9394
+    },
+    {
+      "epoch": 1.522197018794556,
+      "grad_norm": 0.887532114982605,
+      "learning_rate": 2.5616536268824555e-06,
+      "loss": 0.1053,
+      "step": 9395
+    },
+    {
+      "epoch": 1.5223590408295529,
+      "grad_norm": 0.9102559089660645,
+      "learning_rate": 2.561216454843165e-06,
+      "loss": 0.1074,
+      "step": 9396
+    },
+    {
+      "epoch": 1.5225210628645496,
+      "grad_norm": 0.9450837969779968,
+      "learning_rate": 2.560779280930799e-06,
+      "loss": 0.1085,
+      "step": 9397
+    },
+    {
+      "epoch": 1.5226830848995463,
+      "grad_norm": 0.7672545313835144,
+      "learning_rate": 2.5603421051587344e-06,
+      "loss": 0.0921,
+      "step": 9398
+    },
+    {
+      "epoch": 1.5228451069345432,
+      "grad_norm": 0.8284415006637573,
+      "learning_rate": 2.559904927540347e-06,
+      "loss": 0.1048,
+      "step": 9399
+    },
+    {
+      "epoch": 1.5230071289695397,
+      "grad_norm": 0.856656014919281,
+      "learning_rate": 2.5594677480890152e-06,
+      "loss": 0.1047,
+      "step": 9400
+    },
+    {
+      "epoch": 1.5231691510045366,
+      "grad_norm": 0.791228711605072,
+      "learning_rate": 2.559030566818112e-06,
+      "loss": 0.1041,
+      "step": 9401
+    },
+    {
+      "epoch": 1.5233311730395334,
+      "grad_norm": 0.8752877116203308,
+      "learning_rate": 2.558593383741018e-06,
+      "loss": 0.0996,
+      "step": 9402
+    },
+    {
+      "epoch": 1.52349319507453,
+      "grad_norm": 0.8415939211845398,
+      "learning_rate": 2.558156198871108e-06,
+      "loss": 0.0964,
+      "step": 9403
+    },
+    {
+      "epoch": 1.523655217109527,
+      "grad_norm": 0.8422085046768188,
+      "learning_rate": 2.5577190122217583e-06,
+      "loss": 0.1056,
+      "step": 9404
+    },
+    {
+      "epoch": 1.5238172391445235,
+      "grad_norm": 0.7601490616798401,
+      "learning_rate": 2.557281823806347e-06,
+      "loss": 0.0986,
+      "step": 9405
+    },
+    {
+      "epoch": 1.5239792611795204,
+      "grad_norm": 0.8581531047821045,
+      "learning_rate": 2.55684463363825e-06,
+      "loss": 0.1018,
+      "step": 9406
+    },
+    {
+      "epoch": 1.5241412832145171,
+      "grad_norm": 0.7569452524185181,
+      "learning_rate": 2.5564074417308454e-06,
+      "loss": 0.0993,
+      "step": 9407
+    },
+    {
+      "epoch": 1.5243033052495139,
+      "grad_norm": 0.7859471440315247,
+      "learning_rate": 2.5559702480975094e-06,
+      "loss": 0.0949,
+      "step": 9408
+    },
+    {
+      "epoch": 1.5244653272845108,
+      "grad_norm": 0.7991012930870056,
+      "learning_rate": 2.5555330527516197e-06,
+      "loss": 0.1051,
+      "step": 9409
+    },
+    {
+      "epoch": 1.5246273493195075,
+      "grad_norm": 0.8928097486495972,
+      "learning_rate": 2.5550958557065523e-06,
+      "loss": 0.0995,
+      "step": 9410
+    },
+    {
+      "epoch": 1.5247893713545042,
+      "grad_norm": 1.0804157257080078,
+      "learning_rate": 2.554658656975686e-06,
+      "loss": 0.1198,
+      "step": 9411
+    },
+    {
+      "epoch": 1.524951393389501,
+      "grad_norm": 0.8879733681678772,
+      "learning_rate": 2.554221456572396e-06,
+      "loss": 0.1064,
+      "step": 9412
+    },
+    {
+      "epoch": 1.5251134154244976,
+      "grad_norm": 0.8105085492134094,
+      "learning_rate": 2.553784254510061e-06,
+      "loss": 0.104,
+      "step": 9413
+    },
+    {
+      "epoch": 1.5252754374594946,
+      "grad_norm": 0.7311967015266418,
+      "learning_rate": 2.553347050802058e-06,
+      "loss": 0.0945,
+      "step": 9414
+    },
+    {
+      "epoch": 1.5254374594944913,
+      "grad_norm": 0.8524883985519409,
+      "learning_rate": 2.5529098454617644e-06,
+      "loss": 0.1129,
+      "step": 9415
+    },
+    {
+      "epoch": 1.525599481529488,
+      "grad_norm": 0.7317702174186707,
+      "learning_rate": 2.552472638502557e-06,
+      "loss": 0.0935,
+      "step": 9416
+    },
+    {
+      "epoch": 1.525761503564485,
+      "grad_norm": 0.8908723592758179,
+      "learning_rate": 2.5520354299378145e-06,
+      "loss": 0.1069,
+      "step": 9417
+    },
+    {
+      "epoch": 1.5259235255994814,
+      "grad_norm": 0.8989855051040649,
+      "learning_rate": 2.5515982197809142e-06,
+      "loss": 0.1121,
+      "step": 9418
+    },
+    {
+      "epoch": 1.5260855476344783,
+      "grad_norm": 0.9153575897216797,
+      "learning_rate": 2.5511610080452322e-06,
+      "loss": 0.1078,
+      "step": 9419
+    },
+    {
+      "epoch": 1.526247569669475,
+      "grad_norm": 1.041412353515625,
+      "learning_rate": 2.5507237947441478e-06,
+      "loss": 0.1138,
+      "step": 9420
+    },
+    {
+      "epoch": 1.5264095917044718,
+      "grad_norm": 0.935231626033783,
+      "learning_rate": 2.5502865798910377e-06,
+      "loss": 0.1152,
+      "step": 9421
+    },
+    {
+      "epoch": 1.5265716137394687,
+      "grad_norm": 0.7345618009567261,
+      "learning_rate": 2.5498493634992803e-06,
+      "loss": 0.0955,
+      "step": 9422
+    },
+    {
+      "epoch": 1.5267336357744652,
+      "grad_norm": 0.8384798765182495,
+      "learning_rate": 2.5494121455822526e-06,
+      "loss": 0.1117,
+      "step": 9423
+    },
+    {
+      "epoch": 1.5268956578094621,
+      "grad_norm": 0.8286236524581909,
+      "learning_rate": 2.5489749261533333e-06,
+      "loss": 0.104,
+      "step": 9424
+    },
+    {
+      "epoch": 1.5270576798444588,
+      "grad_norm": 0.7071642875671387,
+      "learning_rate": 2.5485377052258987e-06,
+      "loss": 0.0943,
+      "step": 9425
+    },
+    {
+      "epoch": 1.5272197018794555,
+      "grad_norm": 0.7644874453544617,
+      "learning_rate": 2.548100482813329e-06,
+      "loss": 0.0962,
+      "step": 9426
+    },
+    {
+      "epoch": 1.5273817239144525,
+      "grad_norm": 0.8506636619567871,
+      "learning_rate": 2.547663258929001e-06,
+      "loss": 0.1085,
+      "step": 9427
+    },
+    {
+      "epoch": 1.527543745949449,
+      "grad_norm": 0.7493010759353638,
+      "learning_rate": 2.5472260335862915e-06,
+      "loss": 0.0899,
+      "step": 9428
+    },
+    {
+      "epoch": 1.527705767984446,
+      "grad_norm": 0.7209338545799255,
+      "learning_rate": 2.5467888067985803e-06,
+      "loss": 0.0993,
+      "step": 9429
+    },
+    {
+      "epoch": 1.5278677900194426,
+      "grad_norm": 0.9093368053436279,
+      "learning_rate": 2.546351578579245e-06,
+      "loss": 0.1087,
+      "step": 9430
+    },
+    {
+      "epoch": 1.5280298120544393,
+      "grad_norm": 0.7698819637298584,
+      "learning_rate": 2.545914348941664e-06,
+      "loss": 0.0959,
+      "step": 9431
+    },
+    {
+      "epoch": 1.5281918340894363,
+      "grad_norm": 0.8163896799087524,
+      "learning_rate": 2.545477117899213e-06,
+      "loss": 0.0969,
+      "step": 9432
+    },
+    {
+      "epoch": 1.528353856124433,
+      "grad_norm": 0.7964690327644348,
+      "learning_rate": 2.5450398854652747e-06,
+      "loss": 0.1028,
+      "step": 9433
+    },
+    {
+      "epoch": 1.5285158781594297,
+      "grad_norm": 0.894778311252594,
+      "learning_rate": 2.5446026516532235e-06,
+      "loss": 0.111,
+      "step": 9434
+    },
+    {
+      "epoch": 1.5286779001944264,
+      "grad_norm": 0.903476893901825,
+      "learning_rate": 2.5441654164764396e-06,
+      "loss": 0.1154,
+      "step": 9435
+    },
+    {
+      "epoch": 1.528839922229423,
+      "grad_norm": 0.8074314594268799,
+      "learning_rate": 2.5437281799483005e-06,
+      "loss": 0.1042,
+      "step": 9436
+    },
+    {
+      "epoch": 1.52900194426442,
+      "grad_norm": 0.844743013381958,
+      "learning_rate": 2.543290942082185e-06,
+      "loss": 0.1131,
+      "step": 9437
+    },
+    {
+      "epoch": 1.5291639662994168,
+      "grad_norm": 0.8503636121749878,
+      "learning_rate": 2.542853702891471e-06,
+      "loss": 0.1046,
+      "step": 9438
+    },
+    {
+      "epoch": 1.5293259883344135,
+      "grad_norm": 0.8960352540016174,
+      "learning_rate": 2.542416462389539e-06,
+      "loss": 0.1142,
+      "step": 9439
+    },
+    {
+      "epoch": 1.5294880103694104,
+      "grad_norm": 0.7933899164199829,
+      "learning_rate": 2.541979220589765e-06,
+      "loss": 0.0976,
+      "step": 9440
+    },
+    {
+      "epoch": 1.529650032404407,
+      "grad_norm": 0.8900126814842224,
+      "learning_rate": 2.5415419775055277e-06,
+      "loss": 0.1091,
+      "step": 9441
+    },
+    {
+      "epoch": 1.5298120544394038,
+      "grad_norm": 0.7652503848075867,
+      "learning_rate": 2.541104733150207e-06,
+      "loss": 0.0999,
+      "step": 9442
+    },
+    {
+      "epoch": 1.5299740764744005,
+      "grad_norm": 0.818130373954773,
+      "learning_rate": 2.5406674875371807e-06,
+      "loss": 0.1101,
+      "step": 9443
+    },
+    {
+      "epoch": 1.5301360985093972,
+      "grad_norm": 0.7544717788696289,
+      "learning_rate": 2.540230240679828e-06,
+      "loss": 0.0881,
+      "step": 9444
+    },
+    {
+      "epoch": 1.5302981205443942,
+      "grad_norm": 0.856606662273407,
+      "learning_rate": 2.539792992591527e-06,
+      "loss": 0.1124,
+      "step": 9445
+    },
+    {
+      "epoch": 1.5304601425793907,
+      "grad_norm": 0.7668637633323669,
+      "learning_rate": 2.5393557432856575e-06,
+      "loss": 0.0951,
+      "step": 9446
+    },
+    {
+      "epoch": 1.5306221646143876,
+      "grad_norm": 0.8169387578964233,
+      "learning_rate": 2.538918492775596e-06,
+      "loss": 0.1015,
+      "step": 9447
+    },
+    {
+      "epoch": 1.5307841866493843,
+      "grad_norm": 0.7206966876983643,
+      "learning_rate": 2.5384812410747244e-06,
+      "loss": 0.0873,
+      "step": 9448
+    },
+    {
+      "epoch": 1.530946208684381,
+      "grad_norm": 0.803081750869751,
+      "learning_rate": 2.5380439881964185e-06,
+      "loss": 0.093,
+      "step": 9449
+    },
+    {
+      "epoch": 1.531108230719378,
+      "grad_norm": 0.9361199736595154,
+      "learning_rate": 2.53760673415406e-06,
+      "loss": 0.1114,
+      "step": 9450
+    },
+    {
+      "epoch": 1.5312702527543745,
+      "grad_norm": 0.7436367273330688,
+      "learning_rate": 2.537169478961026e-06,
+      "loss": 0.0945,
+      "step": 9451
+    },
+    {
+      "epoch": 1.5314322747893714,
+      "grad_norm": 0.8524910807609558,
+      "learning_rate": 2.5367322226306956e-06,
+      "loss": 0.1094,
+      "step": 9452
+    },
+    {
+      "epoch": 1.531594296824368,
+      "grad_norm": 0.8166900873184204,
+      "learning_rate": 2.5362949651764484e-06,
+      "loss": 0.1032,
+      "step": 9453
+    },
+    {
+      "epoch": 1.5317563188593648,
+      "grad_norm": 0.8203127384185791,
+      "learning_rate": 2.5358577066116622e-06,
+      "loss": 0.1051,
+      "step": 9454
+    },
+    {
+      "epoch": 1.5319183408943617,
+      "grad_norm": 0.7983186841011047,
+      "learning_rate": 2.5354204469497185e-06,
+      "loss": 0.1018,
+      "step": 9455
+    },
+    {
+      "epoch": 1.5320803629293582,
+      "grad_norm": 0.7971822023391724,
+      "learning_rate": 2.534983186203993e-06,
+      "loss": 0.1028,
+      "step": 9456
+    },
+    {
+      "epoch": 1.5322423849643552,
+      "grad_norm": 0.7759663462638855,
+      "learning_rate": 2.5345459243878684e-06,
+      "loss": 0.0967,
+      "step": 9457
+    },
+    {
+      "epoch": 1.5324044069993519,
+      "grad_norm": 0.8662815093994141,
+      "learning_rate": 2.5341086615147207e-06,
+      "loss": 0.1036,
+      "step": 9458
+    },
+    {
+      "epoch": 1.5325664290343486,
+      "grad_norm": 0.8577237725257874,
+      "learning_rate": 2.5336713975979315e-06,
+      "loss": 0.1091,
+      "step": 9459
+    },
+    {
+      "epoch": 1.5327284510693455,
+      "grad_norm": 0.713832676410675,
+      "learning_rate": 2.5332341326508786e-06,
+      "loss": 0.0945,
+      "step": 9460
+    },
+    {
+      "epoch": 1.5328904731043422,
+      "grad_norm": 0.7215037941932678,
+      "learning_rate": 2.532796866686942e-06,
+      "loss": 0.0919,
+      "step": 9461
+    },
+    {
+      "epoch": 1.533052495139339,
+      "grad_norm": 0.7585532665252686,
+      "learning_rate": 2.5323595997195005e-06,
+      "loss": 0.098,
+      "step": 9462
+    },
+    {
+      "epoch": 1.5332145171743357,
+      "grad_norm": 0.7682742476463318,
+      "learning_rate": 2.5319223317619333e-06,
+      "loss": 0.0952,
+      "step": 9463
+    },
+    {
+      "epoch": 1.5333765392093324,
+      "grad_norm": 0.7284139394760132,
+      "learning_rate": 2.53148506282762e-06,
+      "loss": 0.0969,
+      "step": 9464
+    },
+    {
+      "epoch": 1.5335385612443293,
+      "grad_norm": 0.8855341076850891,
+      "learning_rate": 2.5310477929299402e-06,
+      "loss": 0.1052,
+      "step": 9465
+    },
+    {
+      "epoch": 1.533700583279326,
+      "grad_norm": 0.7966269254684448,
+      "learning_rate": 2.530610522082273e-06,
+      "loss": 0.1066,
+      "step": 9466
+    },
+    {
+      "epoch": 1.5338626053143227,
+      "grad_norm": 0.8642690181732178,
+      "learning_rate": 2.5301732502979977e-06,
+      "loss": 0.1016,
+      "step": 9467
+    },
+    {
+      "epoch": 1.5340246273493197,
+      "grad_norm": 1.0195221900939941,
+      "learning_rate": 2.529735977590494e-06,
+      "loss": 0.1063,
+      "step": 9468
+    },
+    {
+      "epoch": 1.5341866493843161,
+      "grad_norm": 0.9482093453407288,
+      "learning_rate": 2.5292987039731415e-06,
+      "loss": 0.116,
+      "step": 9469
+    },
+    {
+      "epoch": 1.534348671419313,
+      "grad_norm": 0.8184714317321777,
+      "learning_rate": 2.52886142945932e-06,
+      "loss": 0.0935,
+      "step": 9470
+    },
+    {
+      "epoch": 1.5345106934543098,
+      "grad_norm": 0.9412416219711304,
+      "learning_rate": 2.5284241540624077e-06,
+      "loss": 0.1175,
+      "step": 9471
+    },
+    {
+      "epoch": 1.5346727154893065,
+      "grad_norm": 0.9039766192436218,
+      "learning_rate": 2.527986877795786e-06,
+      "loss": 0.1045,
+      "step": 9472
+    },
+    {
+      "epoch": 1.5348347375243034,
+      "grad_norm": 0.7086098790168762,
+      "learning_rate": 2.527549600672833e-06,
+      "loss": 0.0866,
+      "step": 9473
+    },
+    {
+      "epoch": 1.5349967595593,
+      "grad_norm": 0.7739736437797546,
+      "learning_rate": 2.527112322706929e-06,
+      "loss": 0.1019,
+      "step": 9474
+    },
+    {
+      "epoch": 1.5351587815942969,
+      "grad_norm": 0.8489776253700256,
+      "learning_rate": 2.5266750439114533e-06,
+      "loss": 0.1016,
+      "step": 9475
+    },
+    {
+      "epoch": 1.5353208036292936,
+      "grad_norm": 0.8796210885047913,
+      "learning_rate": 2.526237764299786e-06,
+      "loss": 0.1076,
+      "step": 9476
+    },
+    {
+      "epoch": 1.5354828256642903,
+      "grad_norm": 0.7349861264228821,
+      "learning_rate": 2.525800483885307e-06,
+      "loss": 0.0936,
+      "step": 9477
+    },
+    {
+      "epoch": 1.5356448476992872,
+      "grad_norm": 0.833838701248169,
+      "learning_rate": 2.5253632026813945e-06,
+      "loss": 0.1022,
+      "step": 9478
+    },
+    {
+      "epoch": 1.5358068697342837,
+      "grad_norm": 0.7381966710090637,
+      "learning_rate": 2.524925920701431e-06,
+      "loss": 0.0966,
+      "step": 9479
+    },
+    {
+      "epoch": 1.5359688917692806,
+      "grad_norm": 0.8320499658584595,
+      "learning_rate": 2.524488637958793e-06,
+      "loss": 0.1105,
+      "step": 9480
+    },
+    {
+      "epoch": 1.5361309138042774,
+      "grad_norm": 0.9006655216217041,
+      "learning_rate": 2.5240513544668634e-06,
+      "loss": 0.11,
+      "step": 9481
+    },
+    {
+      "epoch": 1.536292935839274,
+      "grad_norm": 0.910051703453064,
+      "learning_rate": 2.5236140702390194e-06,
+      "loss": 0.1015,
+      "step": 9482
+    },
+    {
+      "epoch": 1.536454957874271,
+      "grad_norm": 0.788629949092865,
+      "learning_rate": 2.5231767852886424e-06,
+      "loss": 0.0979,
+      "step": 9483
+    },
+    {
+      "epoch": 1.5366169799092677,
+      "grad_norm": 0.8495368957519531,
+      "learning_rate": 2.522739499629112e-06,
+      "loss": 0.0949,
+      "step": 9484
+    },
+    {
+      "epoch": 1.5367790019442644,
+      "grad_norm": 0.7806404232978821,
+      "learning_rate": 2.522302213273808e-06,
+      "loss": 0.1007,
+      "step": 9485
+    },
+    {
+      "epoch": 1.5369410239792611,
+      "grad_norm": 0.8518389463424683,
+      "learning_rate": 2.5218649262361104e-06,
+      "loss": 0.1034,
+      "step": 9486
+    },
+    {
+      "epoch": 1.5371030460142578,
+      "grad_norm": 0.9559544324874878,
+      "learning_rate": 2.521427638529398e-06,
+      "loss": 0.1126,
+      "step": 9487
+    },
+    {
+      "epoch": 1.5372650680492548,
+      "grad_norm": 0.7953001260757446,
+      "learning_rate": 2.520990350167053e-06,
+      "loss": 0.0949,
+      "step": 9488
+    },
+    {
+      "epoch": 1.5374270900842515,
+      "grad_norm": 0.8259668946266174,
+      "learning_rate": 2.5205530611624537e-06,
+      "loss": 0.0983,
+      "step": 9489
+    },
+    {
+      "epoch": 1.5375891121192482,
+      "grad_norm": 0.9413937330245972,
+      "learning_rate": 2.5201157715289796e-06,
+      "loss": 0.1171,
+      "step": 9490
+    },
+    {
+      "epoch": 1.5377511341542451,
+      "grad_norm": 0.7779377698898315,
+      "learning_rate": 2.5196784812800125e-06,
+      "loss": 0.1009,
+      "step": 9491
+    },
+    {
+      "epoch": 1.5379131561892416,
+      "grad_norm": 0.7650463581085205,
+      "learning_rate": 2.519241190428931e-06,
+      "loss": 0.0964,
+      "step": 9492
+    },
+    {
+      "epoch": 1.5380751782242386,
+      "grad_norm": 0.8391408920288086,
+      "learning_rate": 2.5188038989891154e-06,
+      "loss": 0.1014,
+      "step": 9493
+    },
+    {
+      "epoch": 1.5382372002592353,
+      "grad_norm": 0.9325923323631287,
+      "learning_rate": 2.518366606973947e-06,
+      "loss": 0.1169,
+      "step": 9494
+    },
+    {
+      "epoch": 1.538399222294232,
+      "grad_norm": 0.8795773983001709,
+      "learning_rate": 2.5179293143968048e-06,
+      "loss": 0.1103,
+      "step": 9495
+    },
+    {
+      "epoch": 1.538561244329229,
+      "grad_norm": 0.9822404384613037,
+      "learning_rate": 2.517492021271068e-06,
+      "loss": 0.1189,
+      "step": 9496
+    },
+    {
+      "epoch": 1.5387232663642254,
+      "grad_norm": 0.8007500171661377,
+      "learning_rate": 2.517054727610118e-06,
+      "loss": 0.1077,
+      "step": 9497
+    },
+    {
+      "epoch": 1.5388852883992223,
+      "grad_norm": 0.7420747876167297,
+      "learning_rate": 2.5166174334273347e-06,
+      "loss": 0.0917,
+      "step": 9498
+    },
+    {
+      "epoch": 1.539047310434219,
+      "grad_norm": 0.7478721141815186,
+      "learning_rate": 2.5161801387360986e-06,
+      "loss": 0.1011,
+      "step": 9499
+    },
+    {
+      "epoch": 1.5392093324692158,
+      "grad_norm": 0.7703060507774353,
+      "learning_rate": 2.5157428435497887e-06,
+      "loss": 0.0919,
+      "step": 9500
+    },
+    {
+      "epoch": 1.5393713545042127,
+      "grad_norm": 0.8613007068634033,
+      "learning_rate": 2.515305547881787e-06,
+      "loss": 0.1087,
+      "step": 9501
+    },
+    {
+      "epoch": 1.5395333765392092,
+      "grad_norm": 0.7977644205093384,
+      "learning_rate": 2.5148682517454707e-06,
+      "loss": 0.1016,
+      "step": 9502
+    },
+    {
+      "epoch": 1.5396953985742061,
+      "grad_norm": 0.8988528251647949,
+      "learning_rate": 2.5144309551542233e-06,
+      "loss": 0.1072,
+      "step": 9503
+    },
+    {
+      "epoch": 1.5398574206092028,
+      "grad_norm": 0.7992510795593262,
+      "learning_rate": 2.5139936581214235e-06,
+      "loss": 0.0993,
+      "step": 9504
+    },
+    {
+      "epoch": 1.5400194426441995,
+      "grad_norm": 0.7492799758911133,
+      "learning_rate": 2.513556360660451e-06,
+      "loss": 0.0916,
+      "step": 9505
+    },
+    {
+      "epoch": 1.5401814646791965,
+      "grad_norm": 0.6518522500991821,
+      "learning_rate": 2.5131190627846875e-06,
+      "loss": 0.0852,
+      "step": 9506
+    },
+    {
+      "epoch": 1.540343486714193,
+      "grad_norm": 0.8497022390365601,
+      "learning_rate": 2.512681764507512e-06,
+      "loss": 0.109,
+      "step": 9507
+    },
+    {
+      "epoch": 1.54050550874919,
+      "grad_norm": 0.756066083908081,
+      "learning_rate": 2.512244465842305e-06,
+      "loss": 0.1012,
+      "step": 9508
+    },
+    {
+      "epoch": 1.5406675307841866,
+      "grad_norm": 0.8988474607467651,
+      "learning_rate": 2.511807166802447e-06,
+      "loss": 0.1147,
+      "step": 9509
+    },
+    {
+      "epoch": 1.5408295528191833,
+      "grad_norm": 0.7836807370185852,
+      "learning_rate": 2.5113698674013186e-06,
+      "loss": 0.0989,
+      "step": 9510
+    },
+    {
+      "epoch": 1.5409915748541803,
+      "grad_norm": 0.8765679597854614,
+      "learning_rate": 2.510932567652299e-06,
+      "loss": 0.1124,
+      "step": 9511
+    },
+    {
+      "epoch": 1.541153596889177,
+      "grad_norm": 0.8468899130821228,
+      "learning_rate": 2.5104952675687706e-06,
+      "loss": 0.1,
+      "step": 9512
+    },
+    {
+      "epoch": 1.5413156189241737,
+      "grad_norm": 0.8771848082542419,
+      "learning_rate": 2.5100579671641114e-06,
+      "loss": 0.1102,
+      "step": 9513
+    },
+    {
+      "epoch": 1.5414776409591704,
+      "grad_norm": 0.8875617980957031,
+      "learning_rate": 2.509620666451703e-06,
+      "loss": 0.1088,
+      "step": 9514
+    },
+    {
+      "epoch": 1.541639662994167,
+      "grad_norm": 0.7843181490898132,
+      "learning_rate": 2.5091833654449254e-06,
+      "loss": 0.0942,
+      "step": 9515
+    },
+    {
+      "epoch": 1.541801685029164,
+      "grad_norm": 0.9153252243995667,
+      "learning_rate": 2.5087460641571594e-06,
+      "loss": 0.1122,
+      "step": 9516
+    },
+    {
+      "epoch": 1.5419637070641607,
+      "grad_norm": 0.8474258184432983,
+      "learning_rate": 2.5083087626017847e-06,
+      "loss": 0.1033,
+      "step": 9517
+    },
+    {
+      "epoch": 1.5421257290991575,
+      "grad_norm": 0.8334153294563293,
+      "learning_rate": 2.5078714607921825e-06,
+      "loss": 0.106,
+      "step": 9518
+    },
+    {
+      "epoch": 1.5422877511341544,
+      "grad_norm": 0.8846608400344849,
+      "learning_rate": 2.507434158741732e-06,
+      "loss": 0.1001,
+      "step": 9519
+    },
+    {
+      "epoch": 1.5424497731691509,
+      "grad_norm": 0.8367900848388672,
+      "learning_rate": 2.506996856463814e-06,
+      "loss": 0.1098,
+      "step": 9520
+    },
+    {
+      "epoch": 1.5426117952041478,
+      "grad_norm": 0.7622178792953491,
+      "learning_rate": 2.5065595539718098e-06,
+      "loss": 0.0984,
+      "step": 9521
+    },
+    {
+      "epoch": 1.5427738172391445,
+      "grad_norm": 0.7733889818191528,
+      "learning_rate": 2.506122251279099e-06,
+      "loss": 0.1089,
+      "step": 9522
+    },
+    {
+      "epoch": 1.5429358392741412,
+      "grad_norm": 0.9063085913658142,
+      "learning_rate": 2.5056849483990614e-06,
+      "loss": 0.1027,
+      "step": 9523
+    },
+    {
+      "epoch": 1.5430978613091382,
+      "grad_norm": 0.9877629280090332,
+      "learning_rate": 2.5052476453450788e-06,
+      "loss": 0.1207,
+      "step": 9524
+    },
+    {
+      "epoch": 1.5432598833441347,
+      "grad_norm": 0.8588782548904419,
+      "learning_rate": 2.5048103421305313e-06,
+      "loss": 0.1079,
+      "step": 9525
+    },
+    {
+      "epoch": 1.5434219053791316,
+      "grad_norm": 0.7452812194824219,
+      "learning_rate": 2.504373038768799e-06,
+      "loss": 0.0914,
+      "step": 9526
+    },
+    {
+      "epoch": 1.5435839274141283,
+      "grad_norm": 0.8148205876350403,
+      "learning_rate": 2.5039357352732613e-06,
+      "loss": 0.109,
+      "step": 9527
+    },
+    {
+      "epoch": 1.543745949449125,
+      "grad_norm": 0.9041098356246948,
+      "learning_rate": 2.5034984316573003e-06,
+      "loss": 0.1128,
+      "step": 9528
+    },
+    {
+      "epoch": 1.543907971484122,
+      "grad_norm": 0.8906643986701965,
+      "learning_rate": 2.5030611279342955e-06,
+      "loss": 0.123,
+      "step": 9529
+    },
+    {
+      "epoch": 1.5440699935191184,
+      "grad_norm": 0.8250147700309753,
+      "learning_rate": 2.5026238241176283e-06,
+      "loss": 0.1044,
+      "step": 9530
+    },
+    {
+      "epoch": 1.5442320155541154,
+      "grad_norm": 0.8400077819824219,
+      "learning_rate": 2.502186520220677e-06,
+      "loss": 0.1025,
+      "step": 9531
+    },
+    {
+      "epoch": 1.544394037589112,
+      "grad_norm": 0.8428972959518433,
+      "learning_rate": 2.5017492162568246e-06,
+      "loss": 0.0909,
+      "step": 9532
+    },
+    {
+      "epoch": 1.5445560596241088,
+      "grad_norm": 0.7809198498725891,
+      "learning_rate": 2.5013119122394495e-06,
+      "loss": 0.0961,
+      "step": 9533
+    },
+    {
+      "epoch": 1.5447180816591057,
+      "grad_norm": 0.9663342237472534,
+      "learning_rate": 2.5008746081819345e-06,
+      "loss": 0.1164,
+      "step": 9534
+    },
+    {
+      "epoch": 1.5448801036941024,
+      "grad_norm": 0.9518924355506897,
+      "learning_rate": 2.5004373040976574e-06,
+      "loss": 0.1091,
+      "step": 9535
+    },
+    {
+      "epoch": 1.5450421257290992,
+      "grad_norm": 0.8073295950889587,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1027,
+      "step": 9536
+    },
+    {
+      "epoch": 1.5452041477640959,
+      "grad_norm": 0.8112584948539734,
+      "learning_rate": 2.499562695902343e-06,
+      "loss": 0.0952,
+      "step": 9537
+    },
+    {
+      "epoch": 1.5453661697990926,
+      "grad_norm": 0.8796746730804443,
+      "learning_rate": 2.4991253918180668e-06,
+      "loss": 0.1113,
+      "step": 9538
+    },
+    {
+      "epoch": 1.5455281918340895,
+      "grad_norm": 0.7987788915634155,
+      "learning_rate": 2.4986880877605504e-06,
+      "loss": 0.0995,
+      "step": 9539
+    },
+    {
+      "epoch": 1.5456902138690862,
+      "grad_norm": 0.924967348575592,
+      "learning_rate": 2.498250783743176e-06,
+      "loss": 0.1205,
+      "step": 9540
+    },
+    {
+      "epoch": 1.545852235904083,
+      "grad_norm": 0.9094009399414062,
+      "learning_rate": 2.497813479779324e-06,
+      "loss": 0.1179,
+      "step": 9541
+    },
+    {
+      "epoch": 1.5460142579390799,
+      "grad_norm": 0.8155476450920105,
+      "learning_rate": 2.4973761758823734e-06,
+      "loss": 0.1004,
+      "step": 9542
+    },
+    {
+      "epoch": 1.5461762799740764,
+      "grad_norm": 0.7931246161460876,
+      "learning_rate": 2.4969388720657058e-06,
+      "loss": 0.103,
+      "step": 9543
+    },
+    {
+      "epoch": 1.5463383020090733,
+      "grad_norm": 0.8180692791938782,
+      "learning_rate": 2.4965015683427005e-06,
+      "loss": 0.1055,
+      "step": 9544
+    },
+    {
+      "epoch": 1.54650032404407,
+      "grad_norm": 0.7871511578559875,
+      "learning_rate": 2.4960642647267395e-06,
+      "loss": 0.0943,
+      "step": 9545
+    },
+    {
+      "epoch": 1.5466623460790667,
+      "grad_norm": 0.8144401907920837,
+      "learning_rate": 2.4956269612312025e-06,
+      "loss": 0.1029,
+      "step": 9546
+    },
+    {
+      "epoch": 1.5468243681140637,
+      "grad_norm": 0.7799917459487915,
+      "learning_rate": 2.49518965786947e-06,
+      "loss": 0.1033,
+      "step": 9547
+    },
+    {
+      "epoch": 1.5469863901490601,
+      "grad_norm": 0.7506991028785706,
+      "learning_rate": 2.494752354654921e-06,
+      "loss": 0.1017,
+      "step": 9548
+    },
+    {
+      "epoch": 1.547148412184057,
+      "grad_norm": 0.80491042137146,
+      "learning_rate": 2.4943150516009386e-06,
+      "loss": 0.0994,
+      "step": 9549
+    },
+    {
+      "epoch": 1.5473104342190538,
+      "grad_norm": 0.7802474498748779,
+      "learning_rate": 2.4938777487209022e-06,
+      "loss": 0.1002,
+      "step": 9550
+    },
+    {
+      "epoch": 1.5474724562540505,
+      "grad_norm": 0.9787641167640686,
+      "learning_rate": 2.493440446028191e-06,
+      "loss": 0.1154,
+      "step": 9551
+    },
+    {
+      "epoch": 1.5476344782890474,
+      "grad_norm": 0.9254834055900574,
+      "learning_rate": 2.493003143536187e-06,
+      "loss": 0.118,
+      "step": 9552
+    },
+    {
+      "epoch": 1.547796500324044,
+      "grad_norm": 0.8243023157119751,
+      "learning_rate": 2.492565841258268e-06,
+      "loss": 0.1069,
+      "step": 9553
+    },
+    {
+      "epoch": 1.5479585223590409,
+      "grad_norm": 0.8549127578735352,
+      "learning_rate": 2.4921285392078184e-06,
+      "loss": 0.1025,
+      "step": 9554
+    },
+    {
+      "epoch": 1.5481205443940376,
+      "grad_norm": 0.8300154805183411,
+      "learning_rate": 2.4916912373982157e-06,
+      "loss": 0.1019,
+      "step": 9555
+    },
+    {
+      "epoch": 1.5482825664290343,
+      "grad_norm": 0.7590198516845703,
+      "learning_rate": 2.491253935842842e-06,
+      "loss": 0.0964,
+      "step": 9556
+    },
+    {
+      "epoch": 1.5484445884640312,
+      "grad_norm": 0.899855375289917,
+      "learning_rate": 2.490816634555075e-06,
+      "loss": 0.1141,
+      "step": 9557
+    },
+    {
+      "epoch": 1.5486066104990277,
+      "grad_norm": 0.7786848545074463,
+      "learning_rate": 2.490379333548297e-06,
+      "loss": 0.1019,
+      "step": 9558
+    },
+    {
+      "epoch": 1.5487686325340246,
+      "grad_norm": 0.7648515105247498,
+      "learning_rate": 2.489942032835889e-06,
+      "loss": 0.0955,
+      "step": 9559
+    },
+    {
+      "epoch": 1.5489306545690213,
+      "grad_norm": 1.0073561668395996,
+      "learning_rate": 2.4895047324312303e-06,
+      "loss": 0.1208,
+      "step": 9560
+    },
+    {
+      "epoch": 1.549092676604018,
+      "grad_norm": 0.7584699392318726,
+      "learning_rate": 2.4890674323477016e-06,
+      "loss": 0.0963,
+      "step": 9561
+    },
+    {
+      "epoch": 1.549254698639015,
+      "grad_norm": 0.757125735282898,
+      "learning_rate": 2.4886301325986827e-06,
+      "loss": 0.104,
+      "step": 9562
+    },
+    {
+      "epoch": 1.5494167206740117,
+      "grad_norm": 0.7259702682495117,
+      "learning_rate": 2.4881928331975534e-06,
+      "loss": 0.0944,
+      "step": 9563
+    },
+    {
+      "epoch": 1.5495787427090084,
+      "grad_norm": 0.786858320236206,
+      "learning_rate": 2.4877555341576955e-06,
+      "loss": 0.0979,
+      "step": 9564
+    },
+    {
+      "epoch": 1.5497407647440054,
+      "grad_norm": 0.7230560183525085,
+      "learning_rate": 2.487318235492489e-06,
+      "loss": 0.0913,
+      "step": 9565
+    },
+    {
+      "epoch": 1.5499027867790018,
+      "grad_norm": 0.7546066641807556,
+      "learning_rate": 2.4868809372153137e-06,
+      "loss": 0.0931,
+      "step": 9566
+    },
+    {
+      "epoch": 1.5500648088139988,
+      "grad_norm": 0.8449350595474243,
+      "learning_rate": 2.48644363933955e-06,
+      "loss": 0.1074,
+      "step": 9567
+    },
+    {
+      "epoch": 1.5502268308489955,
+      "grad_norm": 0.8704739212989807,
+      "learning_rate": 2.4860063418785773e-06,
+      "loss": 0.1049,
+      "step": 9568
+    },
+    {
+      "epoch": 1.5503888528839922,
+      "grad_norm": 0.7751917839050293,
+      "learning_rate": 2.485569044845777e-06,
+      "loss": 0.0923,
+      "step": 9569
+    },
+    {
+      "epoch": 1.5505508749189891,
+      "grad_norm": 0.7371068596839905,
+      "learning_rate": 2.4851317482545297e-06,
+      "loss": 0.0853,
+      "step": 9570
+    },
+    {
+      "epoch": 1.5507128969539856,
+      "grad_norm": 0.8631237745285034,
+      "learning_rate": 2.4846944521182144e-06,
+      "loss": 0.1069,
+      "step": 9571
+    },
+    {
+      "epoch": 1.5508749189889826,
+      "grad_norm": 0.8396697044372559,
+      "learning_rate": 2.4842571564502117e-06,
+      "loss": 0.1098,
+      "step": 9572
+    },
+    {
+      "epoch": 1.5510369410239793,
+      "grad_norm": 0.7827080488204956,
+      "learning_rate": 2.4838198612639018e-06,
+      "loss": 0.1022,
+      "step": 9573
+    },
+    {
+      "epoch": 1.551198963058976,
+      "grad_norm": 0.7799431681632996,
+      "learning_rate": 2.4833825665726657e-06,
+      "loss": 0.0985,
+      "step": 9574
+    },
+    {
+      "epoch": 1.551360985093973,
+      "grad_norm": 0.8115366101264954,
+      "learning_rate": 2.4829452723898824e-06,
+      "loss": 0.1082,
+      "step": 9575
+    },
+    {
+      "epoch": 1.5515230071289694,
+      "grad_norm": 0.7954683899879456,
+      "learning_rate": 2.482507978728933e-06,
+      "loss": 0.1012,
+      "step": 9576
+    },
+    {
+      "epoch": 1.5516850291639663,
+      "grad_norm": 0.8462672233581543,
+      "learning_rate": 2.482070685603196e-06,
+      "loss": 0.1127,
+      "step": 9577
+    },
+    {
+      "epoch": 1.551847051198963,
+      "grad_norm": 0.7555440664291382,
+      "learning_rate": 2.4816333930260535e-06,
+      "loss": 0.1004,
+      "step": 9578
+    },
+    {
+      "epoch": 1.5520090732339598,
+      "grad_norm": 0.7217851281166077,
+      "learning_rate": 2.481196101010885e-06,
+      "loss": 0.0943,
+      "step": 9579
+    },
+    {
+      "epoch": 1.5521710952689567,
+      "grad_norm": 0.8560697436332703,
+      "learning_rate": 2.4807588095710696e-06,
+      "loss": 0.1011,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5523331173039532,
+      "grad_norm": 0.8755968809127808,
+      "learning_rate": 2.4803215187199883e-06,
+      "loss": 0.0974,
+      "step": 9581
+    },
+    {
+      "epoch": 1.5524951393389501,
+      "grad_norm": 0.7938658595085144,
+      "learning_rate": 2.4798842284710203e-06,
+      "loss": 0.1006,
+      "step": 9582
+    },
+    {
+      "epoch": 1.5526571613739468,
+      "grad_norm": 0.8489831686019897,
+      "learning_rate": 2.4794469388375476e-06,
+      "loss": 0.0943,
+      "step": 9583
+    },
+    {
+      "epoch": 1.5528191834089435,
+      "grad_norm": 0.963909924030304,
+      "learning_rate": 2.4790096498329477e-06,
+      "loss": 0.1159,
+      "step": 9584
+    },
+    {
+      "epoch": 1.5529812054439405,
+      "grad_norm": 0.8376120924949646,
+      "learning_rate": 2.4785723614706025e-06,
+      "loss": 0.1055,
+      "step": 9585
+    },
+    {
+      "epoch": 1.5531432274789372,
+      "grad_norm": 0.9356231689453125,
+      "learning_rate": 2.478135073763891e-06,
+      "loss": 0.1079,
+      "step": 9586
+    },
+    {
+      "epoch": 1.553305249513934,
+      "grad_norm": 0.904419481754303,
+      "learning_rate": 2.477697786726192e-06,
+      "loss": 0.112,
+      "step": 9587
+    },
+    {
+      "epoch": 1.5534672715489306,
+      "grad_norm": 0.7490609288215637,
+      "learning_rate": 2.4772605003708885e-06,
+      "loss": 0.0986,
+      "step": 9588
+    },
+    {
+      "epoch": 1.5536292935839273,
+      "grad_norm": 0.7533844113349915,
+      "learning_rate": 2.476823214711358e-06,
+      "loss": 0.0991,
+      "step": 9589
+    },
+    {
+      "epoch": 1.5537913156189243,
+      "grad_norm": 0.8684905767440796,
+      "learning_rate": 2.476385929760981e-06,
+      "loss": 0.1085,
+      "step": 9590
+    },
+    {
+      "epoch": 1.553953337653921,
+      "grad_norm": 0.8226639032363892,
+      "learning_rate": 2.475948645533138e-06,
+      "loss": 0.1083,
+      "step": 9591
+    },
+    {
+      "epoch": 1.5541153596889177,
+      "grad_norm": 0.8018856644630432,
+      "learning_rate": 2.475511362041207e-06,
+      "loss": 0.1005,
+      "step": 9592
+    },
+    {
+      "epoch": 1.5542773817239146,
+      "grad_norm": 0.6188633441925049,
+      "learning_rate": 2.47507407929857e-06,
+      "loss": 0.0776,
+      "step": 9593
+    },
+    {
+      "epoch": 1.554439403758911,
+      "grad_norm": 0.7864198088645935,
+      "learning_rate": 2.4746367973186063e-06,
+      "loss": 0.0957,
+      "step": 9594
+    },
+    {
+      "epoch": 1.554601425793908,
+      "grad_norm": 0.7992034554481506,
+      "learning_rate": 2.474199516114694e-06,
+      "loss": 0.099,
+      "step": 9595
+    },
+    {
+      "epoch": 1.5547634478289047,
+      "grad_norm": 0.849092423915863,
+      "learning_rate": 2.473762235700214e-06,
+      "loss": 0.1055,
+      "step": 9596
+    },
+    {
+      "epoch": 1.5549254698639015,
+      "grad_norm": 0.709578275680542,
+      "learning_rate": 2.473324956088547e-06,
+      "loss": 0.0904,
+      "step": 9597
+    },
+    {
+      "epoch": 1.5550874918988984,
+      "grad_norm": 0.7516464591026306,
+      "learning_rate": 2.472887677293072e-06,
+      "loss": 0.0981,
+      "step": 9598
+    },
+    {
+      "epoch": 1.5552495139338949,
+      "grad_norm": 0.7759402394294739,
+      "learning_rate": 2.472450399327168e-06,
+      "loss": 0.0952,
+      "step": 9599
+    },
+    {
+      "epoch": 1.5554115359688918,
+      "grad_norm": 0.7998428344726562,
+      "learning_rate": 2.4720131222042156e-06,
+      "loss": 0.1004,
+      "step": 9600
+    },
+    {
+      "epoch": 1.5555735580038885,
+      "grad_norm": 0.8491755127906799,
+      "learning_rate": 2.4715758459375923e-06,
+      "loss": 0.1057,
+      "step": 9601
+    },
+    {
+      "epoch": 1.5557355800388852,
+      "grad_norm": 0.8094433546066284,
+      "learning_rate": 2.4711385705406805e-06,
+      "loss": 0.0951,
+      "step": 9602
+    },
+    {
+      "epoch": 1.5558976020738822,
+      "grad_norm": 0.7939477562904358,
+      "learning_rate": 2.4707012960268594e-06,
+      "loss": 0.1063,
+      "step": 9603
+    },
+    {
+      "epoch": 1.5560596241088787,
+      "grad_norm": 1.0064921379089355,
+      "learning_rate": 2.4702640224095066e-06,
+      "loss": 0.1098,
+      "step": 9604
+    },
+    {
+      "epoch": 1.5562216461438756,
+      "grad_norm": 0.7896994948387146,
+      "learning_rate": 2.4698267497020035e-06,
+      "loss": 0.1027,
+      "step": 9605
+    },
+    {
+      "epoch": 1.5563836681788723,
+      "grad_norm": 0.9108518362045288,
+      "learning_rate": 2.469389477917727e-06,
+      "loss": 0.115,
+      "step": 9606
+    },
+    {
+      "epoch": 1.556545690213869,
+      "grad_norm": 0.7819544076919556,
+      "learning_rate": 2.4689522070700606e-06,
+      "loss": 0.0948,
+      "step": 9607
+    },
+    {
+      "epoch": 1.556707712248866,
+      "grad_norm": 0.7864106297492981,
+      "learning_rate": 2.4685149371723806e-06,
+      "loss": 0.1019,
+      "step": 9608
+    },
+    {
+      "epoch": 1.5568697342838627,
+      "grad_norm": 0.7853366136550903,
+      "learning_rate": 2.468077668238068e-06,
+      "loss": 0.1051,
+      "step": 9609
+    },
+    {
+      "epoch": 1.5570317563188594,
+      "grad_norm": 0.7699379324913025,
+      "learning_rate": 2.467640400280501e-06,
+      "loss": 0.0979,
+      "step": 9610
+    },
+    {
+      "epoch": 1.557193778353856,
+      "grad_norm": 0.7062657475471497,
+      "learning_rate": 2.4672031333130584e-06,
+      "loss": 0.0885,
+      "step": 9611
+    },
+    {
+      "epoch": 1.5573558003888528,
+      "grad_norm": 0.8259475231170654,
+      "learning_rate": 2.466765867349122e-06,
+      "loss": 0.0942,
+      "step": 9612
+    },
+    {
+      "epoch": 1.5575178224238497,
+      "grad_norm": 0.875845193862915,
+      "learning_rate": 2.466328602402069e-06,
+      "loss": 0.1193,
+      "step": 9613
+    },
+    {
+      "epoch": 1.5576798444588464,
+      "grad_norm": 0.7817234396934509,
+      "learning_rate": 2.46589133848528e-06,
+      "loss": 0.0976,
+      "step": 9614
+    },
+    {
+      "epoch": 1.5578418664938432,
+      "grad_norm": 0.8780922889709473,
+      "learning_rate": 2.465454075612132e-06,
+      "loss": 0.1128,
+      "step": 9615
+    },
+    {
+      "epoch": 1.55800388852884,
+      "grad_norm": 0.7593608498573303,
+      "learning_rate": 2.465016813796007e-06,
+      "loss": 0.1004,
+      "step": 9616
+    },
+    {
+      "epoch": 1.5581659105638366,
+      "grad_norm": 0.9103496074676514,
+      "learning_rate": 2.4645795530502823e-06,
+      "loss": 0.1147,
+      "step": 9617
+    },
+    {
+      "epoch": 1.5583279325988335,
+      "grad_norm": 0.7615711688995361,
+      "learning_rate": 2.464142293388338e-06,
+      "loss": 0.0939,
+      "step": 9618
+    },
+    {
+      "epoch": 1.5584899546338302,
+      "grad_norm": 0.7604543566703796,
+      "learning_rate": 2.463705034823553e-06,
+      "loss": 0.0899,
+      "step": 9619
+    },
+    {
+      "epoch": 1.558651976668827,
+      "grad_norm": 0.7992773056030273,
+      "learning_rate": 2.4632677773693048e-06,
+      "loss": 0.0994,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5588139987038239,
+      "grad_norm": 0.7582528591156006,
+      "learning_rate": 2.462830521038975e-06,
+      "loss": 0.0977,
+      "step": 9621
+    },
+    {
+      "epoch": 1.5589760207388204,
+      "grad_norm": 0.7414041757583618,
+      "learning_rate": 2.4623932658459406e-06,
+      "loss": 0.0937,
+      "step": 9622
+    },
+    {
+      "epoch": 1.5591380427738173,
+      "grad_norm": 0.8438722491264343,
+      "learning_rate": 2.461956011803582e-06,
+      "loss": 0.1067,
+      "step": 9623
+    },
+    {
+      "epoch": 1.559300064808814,
+      "grad_norm": 0.8454424142837524,
+      "learning_rate": 2.461518758925277e-06,
+      "loss": 0.107,
+      "step": 9624
+    },
+    {
+      "epoch": 1.5594620868438107,
+      "grad_norm": 0.7657983303070068,
+      "learning_rate": 2.461081507224404e-06,
+      "loss": 0.0993,
+      "step": 9625
+    },
+    {
+      "epoch": 1.5596241088788076,
+      "grad_norm": 0.8123411536216736,
+      "learning_rate": 2.4606442567143434e-06,
+      "loss": 0.0954,
+      "step": 9626
+    },
+    {
+      "epoch": 1.5597861309138041,
+      "grad_norm": 0.8028444051742554,
+      "learning_rate": 2.460207007408474e-06,
+      "loss": 0.0998,
+      "step": 9627
+    },
+    {
+      "epoch": 1.559948152948801,
+      "grad_norm": 0.8465455174446106,
+      "learning_rate": 2.4597697593201728e-06,
+      "loss": 0.1055,
+      "step": 9628
+    },
+    {
+      "epoch": 1.5601101749837978,
+      "grad_norm": 0.7659565210342407,
+      "learning_rate": 2.4593325124628206e-06,
+      "loss": 0.0974,
+      "step": 9629
+    },
+    {
+      "epoch": 1.5602721970187945,
+      "grad_norm": 0.7613688707351685,
+      "learning_rate": 2.4588952668497937e-06,
+      "loss": 0.0921,
+      "step": 9630
+    },
+    {
+      "epoch": 1.5604342190537914,
+      "grad_norm": 0.8725959062576294,
+      "learning_rate": 2.458458022494473e-06,
+      "loss": 0.0952,
+      "step": 9631
+    },
+    {
+      "epoch": 1.560596241088788,
+      "grad_norm": 0.7647036910057068,
+      "learning_rate": 2.4580207794102364e-06,
+      "loss": 0.1001,
+      "step": 9632
+    },
+    {
+      "epoch": 1.5607582631237849,
+      "grad_norm": 0.7943955659866333,
+      "learning_rate": 2.4575835376104624e-06,
+      "loss": 0.098,
+      "step": 9633
+    },
+    {
+      "epoch": 1.5609202851587816,
+      "grad_norm": 0.7511798739433289,
+      "learning_rate": 2.4571462971085293e-06,
+      "loss": 0.0927,
+      "step": 9634
+    },
+    {
+      "epoch": 1.5610823071937783,
+      "grad_norm": 0.8319833874702454,
+      "learning_rate": 2.456709057917815e-06,
+      "loss": 0.0968,
+      "step": 9635
+    },
+    {
+      "epoch": 1.5612443292287752,
+      "grad_norm": 0.8254125118255615,
+      "learning_rate": 2.4562718200517003e-06,
+      "loss": 0.0988,
+      "step": 9636
+    },
+    {
+      "epoch": 1.561406351263772,
+      "grad_norm": 0.7782612442970276,
+      "learning_rate": 2.4558345835235613e-06,
+      "loss": 0.1042,
+      "step": 9637
+    },
+    {
+      "epoch": 1.5615683732987686,
+      "grad_norm": 0.8092712163925171,
+      "learning_rate": 2.4553973483467778e-06,
+      "loss": 0.1003,
+      "step": 9638
+    },
+    {
+      "epoch": 1.5617303953337653,
+      "grad_norm": 0.8870753049850464,
+      "learning_rate": 2.454960114534726e-06,
+      "loss": 0.1047,
+      "step": 9639
+    },
+    {
+      "epoch": 1.561892417368762,
+      "grad_norm": 0.7420865297317505,
+      "learning_rate": 2.454522882100787e-06,
+      "loss": 0.0899,
+      "step": 9640
+    },
+    {
+      "epoch": 1.562054439403759,
+      "grad_norm": 0.8217720985412598,
+      "learning_rate": 2.4540856510583374e-06,
+      "loss": 0.1038,
+      "step": 9641
+    },
+    {
+      "epoch": 1.5622164614387557,
+      "grad_norm": 0.7875351309776306,
+      "learning_rate": 2.453648421420756e-06,
+      "loss": 0.0947,
+      "step": 9642
+    },
+    {
+      "epoch": 1.5623784834737524,
+      "grad_norm": 0.8513997793197632,
+      "learning_rate": 2.4532111932014205e-06,
+      "loss": 0.1039,
+      "step": 9643
+    },
+    {
+      "epoch": 1.5625405055087493,
+      "grad_norm": 0.814285159111023,
+      "learning_rate": 2.4527739664137085e-06,
+      "loss": 0.0976,
+      "step": 9644
+    },
+    {
+      "epoch": 1.5627025275437458,
+      "grad_norm": 0.8172099590301514,
+      "learning_rate": 2.4523367410710004e-06,
+      "loss": 0.096,
+      "step": 9645
+    },
+    {
+      "epoch": 1.5628645495787428,
+      "grad_norm": 0.7031353116035461,
+      "learning_rate": 2.4518995171866717e-06,
+      "loss": 0.0891,
+      "step": 9646
+    },
+    {
+      "epoch": 1.5630265716137395,
+      "grad_norm": 0.9233510494232178,
+      "learning_rate": 2.451462294774102e-06,
+      "loss": 0.1044,
+      "step": 9647
+    },
+    {
+      "epoch": 1.5631885936487362,
+      "grad_norm": 0.9054529070854187,
+      "learning_rate": 2.451025073846668e-06,
+      "loss": 0.1108,
+      "step": 9648
+    },
+    {
+      "epoch": 1.5633506156837331,
+      "grad_norm": 0.8287831544876099,
+      "learning_rate": 2.450587854417748e-06,
+      "loss": 0.1046,
+      "step": 9649
+    },
+    {
+      "epoch": 1.5635126377187296,
+      "grad_norm": 0.9027771949768066,
+      "learning_rate": 2.45015063650072e-06,
+      "loss": 0.1089,
+      "step": 9650
+    },
+    {
+      "epoch": 1.5636746597537265,
+      "grad_norm": 0.7133320569992065,
+      "learning_rate": 2.449713420108963e-06,
+      "loss": 0.0931,
+      "step": 9651
+    },
+    {
+      "epoch": 1.5638366817887233,
+      "grad_norm": 0.8660880327224731,
+      "learning_rate": 2.449276205255853e-06,
+      "loss": 0.1106,
+      "step": 9652
+    },
+    {
+      "epoch": 1.56399870382372,
+      "grad_norm": 0.779906690120697,
+      "learning_rate": 2.448838991954769e-06,
+      "loss": 0.0977,
+      "step": 9653
+    },
+    {
+      "epoch": 1.564160725858717,
+      "grad_norm": 0.9731188416481018,
+      "learning_rate": 2.448401780219087e-06,
+      "loss": 0.1092,
+      "step": 9654
+    },
+    {
+      "epoch": 1.5643227478937134,
+      "grad_norm": 0.8192594647407532,
+      "learning_rate": 2.447964570062186e-06,
+      "loss": 0.0919,
+      "step": 9655
+    },
+    {
+      "epoch": 1.5644847699287103,
+      "grad_norm": 0.8036348223686218,
+      "learning_rate": 2.4475273614974437e-06,
+      "loss": 0.0945,
+      "step": 9656
+    },
+    {
+      "epoch": 1.564646791963707,
+      "grad_norm": 0.8575658202171326,
+      "learning_rate": 2.447090154538237e-06,
+      "loss": 0.1013,
+      "step": 9657
+    },
+    {
+      "epoch": 1.5648088139987038,
+      "grad_norm": 0.7713079452514648,
+      "learning_rate": 2.4466529491979437e-06,
+      "loss": 0.1026,
+      "step": 9658
+    },
+    {
+      "epoch": 1.5649708360337007,
+      "grad_norm": 0.7967318892478943,
+      "learning_rate": 2.4462157454899393e-06,
+      "loss": 0.0993,
+      "step": 9659
+    },
+    {
+      "epoch": 1.5651328580686974,
+      "grad_norm": 0.8018705248832703,
+      "learning_rate": 2.445778543427605e-06,
+      "loss": 0.1044,
+      "step": 9660
+    },
+    {
+      "epoch": 1.565294880103694,
+      "grad_norm": 0.7897207140922546,
+      "learning_rate": 2.445341343024315e-06,
+      "loss": 0.1001,
+      "step": 9661
+    },
+    {
+      "epoch": 1.5654569021386908,
+      "grad_norm": 0.818270206451416,
+      "learning_rate": 2.4449041442934485e-06,
+      "loss": 0.092,
+      "step": 9662
+    },
+    {
+      "epoch": 1.5656189241736875,
+      "grad_norm": 0.9617789387702942,
+      "learning_rate": 2.4444669472483807e-06,
+      "loss": 0.1061,
+      "step": 9663
+    },
+    {
+      "epoch": 1.5657809462086845,
+      "grad_norm": 0.7740570902824402,
+      "learning_rate": 2.4440297519024906e-06,
+      "loss": 0.1002,
+      "step": 9664
+    },
+    {
+      "epoch": 1.5659429682436812,
+      "grad_norm": 0.8566815853118896,
+      "learning_rate": 2.443592558269155e-06,
+      "loss": 0.1084,
+      "step": 9665
+    },
+    {
+      "epoch": 1.5661049902786779,
+      "grad_norm": 0.8553325533866882,
+      "learning_rate": 2.4431553663617502e-06,
+      "loss": 0.1038,
+      "step": 9666
+    },
+    {
+      "epoch": 1.5662670123136748,
+      "grad_norm": 0.7744235396385193,
+      "learning_rate": 2.4427181761936535e-06,
+      "loss": 0.0937,
+      "step": 9667
+    },
+    {
+      "epoch": 1.5664290343486713,
+      "grad_norm": 0.8114905953407288,
+      "learning_rate": 2.4422809877782417e-06,
+      "loss": 0.1026,
+      "step": 9668
+    },
+    {
+      "epoch": 1.5665910563836682,
+      "grad_norm": 0.8174008727073669,
+      "learning_rate": 2.4418438011288926e-06,
+      "loss": 0.0945,
+      "step": 9669
+    },
+    {
+      "epoch": 1.566753078418665,
+      "grad_norm": 0.9269281625747681,
+      "learning_rate": 2.4414066162589823e-06,
+      "loss": 0.1222,
+      "step": 9670
+    },
+    {
+      "epoch": 1.5669151004536617,
+      "grad_norm": 0.7899512052536011,
+      "learning_rate": 2.4409694331818884e-06,
+      "loss": 0.1009,
+      "step": 9671
+    },
+    {
+      "epoch": 1.5670771224886586,
+      "grad_norm": 0.7267419099807739,
+      "learning_rate": 2.4405322519109864e-06,
+      "loss": 0.0993,
+      "step": 9672
+    },
+    {
+      "epoch": 1.567239144523655,
+      "grad_norm": 0.8098709583282471,
+      "learning_rate": 2.4400950724596527e-06,
+      "loss": 0.0992,
+      "step": 9673
+    },
+    {
+      "epoch": 1.567401166558652,
+      "grad_norm": 0.7339199185371399,
+      "learning_rate": 2.4396578948412664e-06,
+      "loss": 0.1005,
+      "step": 9674
+    },
+    {
+      "epoch": 1.5675631885936487,
+      "grad_norm": 0.7503772377967834,
+      "learning_rate": 2.4392207190692015e-06,
+      "loss": 0.1032,
+      "step": 9675
+    },
+    {
+      "epoch": 1.5677252106286454,
+      "grad_norm": 0.8887887001037598,
+      "learning_rate": 2.4387835451568355e-06,
+      "loss": 0.12,
+      "step": 9676
+    },
+    {
+      "epoch": 1.5678872326636424,
+      "grad_norm": 0.8450667262077332,
+      "learning_rate": 2.4383463731175457e-06,
+      "loss": 0.108,
+      "step": 9677
+    },
+    {
+      "epoch": 1.5680492546986389,
+      "grad_norm": 0.9332994818687439,
+      "learning_rate": 2.4379092029647067e-06,
+      "loss": 0.1086,
+      "step": 9678
+    },
+    {
+      "epoch": 1.5682112767336358,
+      "grad_norm": 0.7733569741249084,
+      "learning_rate": 2.437472034711696e-06,
+      "loss": 0.0988,
+      "step": 9679
+    },
+    {
+      "epoch": 1.5683732987686325,
+      "grad_norm": 0.806662917137146,
+      "learning_rate": 2.4370348683718906e-06,
+      "loss": 0.0925,
+      "step": 9680
+    },
+    {
+      "epoch": 1.5685353208036292,
+      "grad_norm": 0.8055416941642761,
+      "learning_rate": 2.436597703958665e-06,
+      "loss": 0.108,
+      "step": 9681
+    },
+    {
+      "epoch": 1.5686973428386262,
+      "grad_norm": 0.7006134390830994,
+      "learning_rate": 2.436160541485396e-06,
+      "loss": 0.0885,
+      "step": 9682
+    },
+    {
+      "epoch": 1.5688593648736227,
+      "grad_norm": 0.8832897543907166,
+      "learning_rate": 2.4357233809654608e-06,
+      "loss": 0.1098,
+      "step": 9683
+    },
+    {
+      "epoch": 1.5690213869086196,
+      "grad_norm": 0.7593520879745483,
+      "learning_rate": 2.4352862224122344e-06,
+      "loss": 0.0944,
+      "step": 9684
+    },
+    {
+      "epoch": 1.5691834089436163,
+      "grad_norm": 0.8768290877342224,
+      "learning_rate": 2.4348490658390924e-06,
+      "loss": 0.114,
+      "step": 9685
+    },
+    {
+      "epoch": 1.569345430978613,
+      "grad_norm": 0.7737889289855957,
+      "learning_rate": 2.4344119112594124e-06,
+      "loss": 0.0968,
+      "step": 9686
+    },
+    {
+      "epoch": 1.56950745301361,
+      "grad_norm": 0.6443475484848022,
+      "learning_rate": 2.4339747586865677e-06,
+      "loss": 0.0738,
+      "step": 9687
+    },
+    {
+      "epoch": 1.5696694750486067,
+      "grad_norm": 0.7420233488082886,
+      "learning_rate": 2.4335376081339364e-06,
+      "loss": 0.0994,
+      "step": 9688
+    },
+    {
+      "epoch": 1.5698314970836034,
+      "grad_norm": 0.8285709023475647,
+      "learning_rate": 2.433100459614894e-06,
+      "loss": 0.1032,
+      "step": 9689
+    },
+    {
+      "epoch": 1.5699935191186,
+      "grad_norm": 0.8334299921989441,
+      "learning_rate": 2.4326633131428147e-06,
+      "loss": 0.1027,
+      "step": 9690
+    },
+    {
+      "epoch": 1.5701555411535968,
+      "grad_norm": 0.7420817613601685,
+      "learning_rate": 2.432226168731076e-06,
+      "loss": 0.0909,
+      "step": 9691
+    },
+    {
+      "epoch": 1.5703175631885937,
+      "grad_norm": 0.7946280241012573,
+      "learning_rate": 2.4317890263930516e-06,
+      "loss": 0.0953,
+      "step": 9692
+    },
+    {
+      "epoch": 1.5704795852235904,
+      "grad_norm": 0.8967946767807007,
+      "learning_rate": 2.431351886142118e-06,
+      "loss": 0.1133,
+      "step": 9693
+    },
+    {
+      "epoch": 1.5706416072585871,
+      "grad_norm": 0.7146013379096985,
+      "learning_rate": 2.430914747991651e-06,
+      "loss": 0.0835,
+      "step": 9694
+    },
+    {
+      "epoch": 1.570803629293584,
+      "grad_norm": 0.9583059549331665,
+      "learning_rate": 2.430477611955026e-06,
+      "loss": 0.1078,
+      "step": 9695
+    },
+    {
+      "epoch": 1.5709656513285806,
+      "grad_norm": 0.8162189722061157,
+      "learning_rate": 2.430040478045617e-06,
+      "loss": 0.1036,
+      "step": 9696
+    },
+    {
+      "epoch": 1.5711276733635775,
+      "grad_norm": 0.8183581829071045,
+      "learning_rate": 2.4296033462768e-06,
+      "loss": 0.0958,
+      "step": 9697
+    },
+    {
+      "epoch": 1.5712896953985742,
+      "grad_norm": 0.7942171096801758,
+      "learning_rate": 2.429166216661951e-06,
+      "loss": 0.0945,
+      "step": 9698
+    },
+    {
+      "epoch": 1.571451717433571,
+      "grad_norm": 0.9565669298171997,
+      "learning_rate": 2.4287290892144434e-06,
+      "loss": 0.1132,
+      "step": 9699
+    },
+    {
+      "epoch": 1.5716137394685679,
+      "grad_norm": 0.7852692604064941,
+      "learning_rate": 2.4282919639476544e-06,
+      "loss": 0.0944,
+      "step": 9700
+    },
+    {
+      "epoch": 1.5717757615035644,
+      "grad_norm": 0.8304145336151123,
+      "learning_rate": 2.427854840874957e-06,
+      "loss": 0.092,
+      "step": 9701
+    },
+    {
+      "epoch": 1.5719377835385613,
+      "grad_norm": 0.869964599609375,
+      "learning_rate": 2.4274177200097266e-06,
+      "loss": 0.1011,
+      "step": 9702
+    },
+    {
+      "epoch": 1.572099805573558,
+      "grad_norm": 0.9880481958389282,
+      "learning_rate": 2.4269806013653385e-06,
+      "loss": 0.1098,
+      "step": 9703
+    },
+    {
+      "epoch": 1.5722618276085547,
+      "grad_norm": 0.8023558259010315,
+      "learning_rate": 2.426543484955168e-06,
+      "loss": 0.1075,
+      "step": 9704
+    },
+    {
+      "epoch": 1.5724238496435516,
+      "grad_norm": 0.8949816823005676,
+      "learning_rate": 2.426106370792588e-06,
+      "loss": 0.1046,
+      "step": 9705
+    },
+    {
+      "epoch": 1.5725858716785481,
+      "grad_norm": 1.0093536376953125,
+      "learning_rate": 2.425669258890975e-06,
+      "loss": 0.1205,
+      "step": 9706
+    },
+    {
+      "epoch": 1.572747893713545,
+      "grad_norm": 0.8207192420959473,
+      "learning_rate": 2.4252321492637027e-06,
+      "loss": 0.1094,
+      "step": 9707
+    },
+    {
+      "epoch": 1.5729099157485418,
+      "grad_norm": 0.9045940637588501,
+      "learning_rate": 2.4247950419241457e-06,
+      "loss": 0.1129,
+      "step": 9708
+    },
+    {
+      "epoch": 1.5730719377835385,
+      "grad_norm": 0.9074610471725464,
+      "learning_rate": 2.4243579368856787e-06,
+      "loss": 0.1103,
+      "step": 9709
+    },
+    {
+      "epoch": 1.5732339598185354,
+      "grad_norm": 0.7920365333557129,
+      "learning_rate": 2.4239208341616755e-06,
+      "loss": 0.099,
+      "step": 9710
+    },
+    {
+      "epoch": 1.5733959818535321,
+      "grad_norm": 0.9034141898155212,
+      "learning_rate": 2.4234837337655098e-06,
+      "loss": 0.1067,
+      "step": 9711
+    },
+    {
+      "epoch": 1.5735580038885288,
+      "grad_norm": 0.7888884544372559,
+      "learning_rate": 2.4230466357105575e-06,
+      "loss": 0.1013,
+      "step": 9712
+    },
+    {
+      "epoch": 1.5737200259235256,
+      "grad_norm": 0.8596574068069458,
+      "learning_rate": 2.422609540010192e-06,
+      "loss": 0.1108,
+      "step": 9713
+    },
+    {
+      "epoch": 1.5738820479585223,
+      "grad_norm": 0.9732898473739624,
+      "learning_rate": 2.4221724466777874e-06,
+      "loss": 0.0983,
+      "step": 9714
+    },
+    {
+      "epoch": 1.5740440699935192,
+      "grad_norm": 0.8367780447006226,
+      "learning_rate": 2.421735355726718e-06,
+      "loss": 0.1024,
+      "step": 9715
+    },
+    {
+      "epoch": 1.574206092028516,
+      "grad_norm": 0.7838643193244934,
+      "learning_rate": 2.421298267170356e-06,
+      "loss": 0.0972,
+      "step": 9716
+    },
+    {
+      "epoch": 1.5743681140635126,
+      "grad_norm": 0.9189515709877014,
+      "learning_rate": 2.420861181022077e-06,
+      "loss": 0.1078,
+      "step": 9717
+    },
+    {
+      "epoch": 1.5745301360985096,
+      "grad_norm": 0.812330424785614,
+      "learning_rate": 2.420424097295255e-06,
+      "loss": 0.1065,
+      "step": 9718
+    },
+    {
+      "epoch": 1.574692158133506,
+      "grad_norm": 0.7901645302772522,
+      "learning_rate": 2.419987016003263e-06,
+      "loss": 0.0998,
+      "step": 9719
+    },
+    {
+      "epoch": 1.574854180168503,
+      "grad_norm": 0.9493116736412048,
+      "learning_rate": 2.419549937159474e-06,
+      "loss": 0.1202,
+      "step": 9720
+    },
+    {
+      "epoch": 1.5750162022034997,
+      "grad_norm": 0.873347282409668,
+      "learning_rate": 2.419112860777262e-06,
+      "loss": 0.0935,
+      "step": 9721
+    },
+    {
+      "epoch": 1.5751782242384964,
+      "grad_norm": 0.7348765730857849,
+      "learning_rate": 2.418675786870002e-06,
+      "loss": 0.0989,
+      "step": 9722
+    },
+    {
+      "epoch": 1.5753402462734933,
+      "grad_norm": 0.9190308451652527,
+      "learning_rate": 2.4182387154510653e-06,
+      "loss": 0.1148,
+      "step": 9723
+    },
+    {
+      "epoch": 1.5755022683084898,
+      "grad_norm": 0.8104141354560852,
+      "learning_rate": 2.4178016465338266e-06,
+      "loss": 0.102,
+      "step": 9724
+    },
+    {
+      "epoch": 1.5756642903434868,
+      "grad_norm": 0.826193630695343,
+      "learning_rate": 2.417364580131658e-06,
+      "loss": 0.1042,
+      "step": 9725
+    },
+    {
+      "epoch": 1.5758263123784835,
+      "grad_norm": 0.7655068635940552,
+      "learning_rate": 2.416927516257933e-06,
+      "loss": 0.0976,
+      "step": 9726
+    },
+    {
+      "epoch": 1.5759883344134802,
+      "grad_norm": 0.8392152190208435,
+      "learning_rate": 2.416490454926026e-06,
+      "loss": 0.1102,
+      "step": 9727
+    },
+    {
+      "epoch": 1.5761503564484771,
+      "grad_norm": 0.7975762486457825,
+      "learning_rate": 2.416053396149308e-06,
+      "loss": 0.1035,
+      "step": 9728
+    },
+    {
+      "epoch": 1.5763123784834736,
+      "grad_norm": 0.8951084613800049,
+      "learning_rate": 2.4156163399411534e-06,
+      "loss": 0.0971,
+      "step": 9729
+    },
+    {
+      "epoch": 1.5764744005184705,
+      "grad_norm": 0.8403456211090088,
+      "learning_rate": 2.415179286314934e-06,
+      "loss": 0.1045,
+      "step": 9730
+    },
+    {
+      "epoch": 1.5766364225534673,
+      "grad_norm": 0.8686217665672302,
+      "learning_rate": 2.414742235284024e-06,
+      "loss": 0.1051,
+      "step": 9731
+    },
+    {
+      "epoch": 1.576798444588464,
+      "grad_norm": 0.810461699962616,
+      "learning_rate": 2.414305186861795e-06,
+      "loss": 0.107,
+      "step": 9732
+    },
+    {
+      "epoch": 1.576960466623461,
+      "grad_norm": 0.8022754788398743,
+      "learning_rate": 2.4138681410616206e-06,
+      "loss": 0.1036,
+      "step": 9733
+    },
+    {
+      "epoch": 1.5771224886584574,
+      "grad_norm": 0.8617199063301086,
+      "learning_rate": 2.4134310978968716e-06,
+      "loss": 0.112,
+      "step": 9734
+    },
+    {
+      "epoch": 1.5772845106934543,
+      "grad_norm": 0.8537782430648804,
+      "learning_rate": 2.4129940573809213e-06,
+      "loss": 0.1043,
+      "step": 9735
+    },
+    {
+      "epoch": 1.577446532728451,
+      "grad_norm": 0.7416249513626099,
+      "learning_rate": 2.412557019527143e-06,
+      "loss": 0.0956,
+      "step": 9736
+    },
+    {
+      "epoch": 1.5776085547634477,
+      "grad_norm": 0.8986727595329285,
+      "learning_rate": 2.4121199843489084e-06,
+      "loss": 0.1091,
+      "step": 9737
+    },
+    {
+      "epoch": 1.5777705767984447,
+      "grad_norm": 0.7465632557868958,
+      "learning_rate": 2.4116829518595896e-06,
+      "loss": 0.0912,
+      "step": 9738
+    },
+    {
+      "epoch": 1.5779325988334414,
+      "grad_norm": 0.7991532683372498,
+      "learning_rate": 2.4112459220725588e-06,
+      "loss": 0.0968,
+      "step": 9739
+    },
+    {
+      "epoch": 1.578094620868438,
+      "grad_norm": 0.7432222366333008,
+      "learning_rate": 2.410808895001187e-06,
+      "loss": 0.0916,
+      "step": 9740
+    },
+    {
+      "epoch": 1.5782566429034348,
+      "grad_norm": 0.8202318549156189,
+      "learning_rate": 2.4103718706588477e-06,
+      "loss": 0.106,
+      "step": 9741
+    },
+    {
+      "epoch": 1.5784186649384315,
+      "grad_norm": 0.8392171263694763,
+      "learning_rate": 2.409934849058913e-06,
+      "loss": 0.1007,
+      "step": 9742
+    },
+    {
+      "epoch": 1.5785806869734285,
+      "grad_norm": 0.8795257806777954,
+      "learning_rate": 2.4094978302147533e-06,
+      "loss": 0.106,
+      "step": 9743
+    },
+    {
+      "epoch": 1.5787427090084252,
+      "grad_norm": 0.8620293140411377,
+      "learning_rate": 2.4090608141397417e-06,
+      "loss": 0.1091,
+      "step": 9744
+    },
+    {
+      "epoch": 1.5789047310434219,
+      "grad_norm": 0.7913221120834351,
+      "learning_rate": 2.4086238008472473e-06,
+      "loss": 0.1002,
+      "step": 9745
+    },
+    {
+      "epoch": 1.5790667530784188,
+      "grad_norm": 0.8350681662559509,
+      "learning_rate": 2.408186790350645e-06,
+      "loss": 0.0996,
+      "step": 9746
+    },
+    {
+      "epoch": 1.5792287751134153,
+      "grad_norm": 0.6608853936195374,
+      "learning_rate": 2.4077497826633045e-06,
+      "loss": 0.0797,
+      "step": 9747
+    },
+    {
+      "epoch": 1.5793907971484122,
+      "grad_norm": 0.8556117415428162,
+      "learning_rate": 2.4073127777985982e-06,
+      "loss": 0.1055,
+      "step": 9748
+    },
+    {
+      "epoch": 1.579552819183409,
+      "grad_norm": 0.8157343864440918,
+      "learning_rate": 2.4068757757698954e-06,
+      "loss": 0.0863,
+      "step": 9749
+    },
+    {
+      "epoch": 1.5797148412184057,
+      "grad_norm": 0.8553090691566467,
+      "learning_rate": 2.406438776590568e-06,
+      "loss": 0.0932,
+      "step": 9750
+    },
+    {
+      "epoch": 1.5798768632534026,
+      "grad_norm": 0.8190301656723022,
+      "learning_rate": 2.406001780273989e-06,
+      "loss": 0.1047,
+      "step": 9751
+    },
+    {
+      "epoch": 1.580038885288399,
+      "grad_norm": 0.7428902983665466,
+      "learning_rate": 2.4055647868335273e-06,
+      "loss": 0.0918,
+      "step": 9752
+    },
+    {
+      "epoch": 1.580200907323396,
+      "grad_norm": 0.7674239277839661,
+      "learning_rate": 2.4051277962825555e-06,
+      "loss": 0.0956,
+      "step": 9753
+    },
+    {
+      "epoch": 1.5803629293583927,
+      "grad_norm": 0.7433567047119141,
+      "learning_rate": 2.404690808634442e-06,
+      "loss": 0.0922,
+      "step": 9754
+    },
+    {
+      "epoch": 1.5805249513933894,
+      "grad_norm": 0.8673844337463379,
+      "learning_rate": 2.40425382390256e-06,
+      "loss": 0.1007,
+      "step": 9755
+    },
+    {
+      "epoch": 1.5806869734283864,
+      "grad_norm": 0.8878071904182434,
+      "learning_rate": 2.4038168421002795e-06,
+      "loss": 0.1117,
+      "step": 9756
+    },
+    {
+      "epoch": 1.5808489954633829,
+      "grad_norm": 0.81288743019104,
+      "learning_rate": 2.403379863240971e-06,
+      "loss": 0.104,
+      "step": 9757
+    },
+    {
+      "epoch": 1.5810110174983798,
+      "grad_norm": 0.8590624332427979,
+      "learning_rate": 2.4029428873380044e-06,
+      "loss": 0.1045,
+      "step": 9758
+    },
+    {
+      "epoch": 1.5811730395333765,
+      "grad_norm": 0.8249778151512146,
+      "learning_rate": 2.40250591440475e-06,
+      "loss": 0.0995,
+      "step": 9759
+    },
+    {
+      "epoch": 1.5813350615683732,
+      "grad_norm": 0.747291624546051,
+      "learning_rate": 2.4020689444545796e-06,
+      "loss": 0.0925,
+      "step": 9760
+    },
+    {
+      "epoch": 1.5814970836033702,
+      "grad_norm": 0.8458421230316162,
+      "learning_rate": 2.4016319775008623e-06,
+      "loss": 0.1052,
+      "step": 9761
+    },
+    {
+      "epoch": 1.5816591056383669,
+      "grad_norm": 0.7993534803390503,
+      "learning_rate": 2.401195013556969e-06,
+      "loss": 0.1025,
+      "step": 9762
+    },
+    {
+      "epoch": 1.5818211276733636,
+      "grad_norm": 0.8762085437774658,
+      "learning_rate": 2.4007580526362685e-06,
+      "loss": 0.1053,
+      "step": 9763
+    },
+    {
+      "epoch": 1.5819831497083603,
+      "grad_norm": 0.9038422107696533,
+      "learning_rate": 2.400321094752131e-06,
+      "loss": 0.1019,
+      "step": 9764
+    },
+    {
+      "epoch": 1.582145171743357,
+      "grad_norm": 0.793109655380249,
+      "learning_rate": 2.399884139917927e-06,
+      "loss": 0.0976,
+      "step": 9765
+    },
+    {
+      "epoch": 1.582307193778354,
+      "grad_norm": 0.8598572611808777,
+      "learning_rate": 2.399447188147027e-06,
+      "loss": 0.103,
+      "step": 9766
+    },
+    {
+      "epoch": 1.5824692158133506,
+      "grad_norm": 0.7793141007423401,
+      "learning_rate": 2.399010239452799e-06,
+      "loss": 0.0916,
+      "step": 9767
+    },
+    {
+      "epoch": 1.5826312378483474,
+      "grad_norm": 0.8122705817222595,
+      "learning_rate": 2.3985732938486137e-06,
+      "loss": 0.0946,
+      "step": 9768
+    },
+    {
+      "epoch": 1.5827932598833443,
+      "grad_norm": 0.8980499505996704,
+      "learning_rate": 2.3981363513478394e-06,
+      "loss": 0.1131,
+      "step": 9769
+    },
+    {
+      "epoch": 1.5829552819183408,
+      "grad_norm": 0.8115260004997253,
+      "learning_rate": 2.3976994119638464e-06,
+      "loss": 0.1007,
+      "step": 9770
+    },
+    {
+      "epoch": 1.5831173039533377,
+      "grad_norm": 0.766045868396759,
+      "learning_rate": 2.3972624757100044e-06,
+      "loss": 0.1038,
+      "step": 9771
+    },
+    {
+      "epoch": 1.5832793259883344,
+      "grad_norm": 0.8186202049255371,
+      "learning_rate": 2.3968255425996817e-06,
+      "loss": 0.1021,
+      "step": 9772
+    },
+    {
+      "epoch": 1.5834413480233311,
+      "grad_norm": 0.7019743919372559,
+      "learning_rate": 2.396388612646247e-06,
+      "loss": 0.0887,
+      "step": 9773
+    },
+    {
+      "epoch": 1.583603370058328,
+      "grad_norm": 0.9179200530052185,
+      "learning_rate": 2.3959516858630707e-06,
+      "loss": 0.1108,
+      "step": 9774
+    },
+    {
+      "epoch": 1.5837653920933246,
+      "grad_norm": 0.8982585668563843,
+      "learning_rate": 2.395514762263522e-06,
+      "loss": 0.1048,
+      "step": 9775
+    },
+    {
+      "epoch": 1.5839274141283215,
+      "grad_norm": 0.8416269421577454,
+      "learning_rate": 2.3950778418609676e-06,
+      "loss": 0.1036,
+      "step": 9776
+    },
+    {
+      "epoch": 1.5840894361633182,
+      "grad_norm": 0.6837180256843567,
+      "learning_rate": 2.3946409246687775e-06,
+      "loss": 0.083,
+      "step": 9777
+    },
+    {
+      "epoch": 1.584251458198315,
+      "grad_norm": 0.7735916376113892,
+      "learning_rate": 2.39420401070032e-06,
+      "loss": 0.0908,
+      "step": 9778
+    },
+    {
+      "epoch": 1.5844134802333119,
+      "grad_norm": 0.894906759262085,
+      "learning_rate": 2.3937670999689634e-06,
+      "loss": 0.1112,
+      "step": 9779
+    },
+    {
+      "epoch": 1.5845755022683083,
+      "grad_norm": 0.8407071232795715,
+      "learning_rate": 2.3933301924880768e-06,
+      "loss": 0.1083,
+      "step": 9780
+    },
+    {
+      "epoch": 1.5847375243033053,
+      "grad_norm": 0.7255405187606812,
+      "learning_rate": 2.392893288271029e-06,
+      "loss": 0.088,
+      "step": 9781
+    },
+    {
+      "epoch": 1.584899546338302,
+      "grad_norm": 0.8209105134010315,
+      "learning_rate": 2.3924563873311868e-06,
+      "loss": 0.104,
+      "step": 9782
+    },
+    {
+      "epoch": 1.5850615683732987,
+      "grad_norm": 0.8289636373519897,
+      "learning_rate": 2.3920194896819183e-06,
+      "loss": 0.1019,
+      "step": 9783
+    },
+    {
+      "epoch": 1.5852235904082956,
+      "grad_norm": 0.8864776492118835,
+      "learning_rate": 2.391582595336593e-06,
+      "loss": 0.1073,
+      "step": 9784
+    },
+    {
+      "epoch": 1.5853856124432921,
+      "grad_norm": 0.8953683376312256,
+      "learning_rate": 2.391145704308577e-06,
+      "loss": 0.1121,
+      "step": 9785
+    },
+    {
+      "epoch": 1.585547634478289,
+      "grad_norm": 0.8129318356513977,
+      "learning_rate": 2.3907088166112406e-06,
+      "loss": 0.1012,
+      "step": 9786
+    },
+    {
+      "epoch": 1.5857096565132858,
+      "grad_norm": 0.7446659207344055,
+      "learning_rate": 2.3902719322579487e-06,
+      "loss": 0.0929,
+      "step": 9787
+    },
+    {
+      "epoch": 1.5858716785482825,
+      "grad_norm": 0.8847704529762268,
+      "learning_rate": 2.3898350512620696e-06,
+      "loss": 0.1074,
+      "step": 9788
+    },
+    {
+      "epoch": 1.5860337005832794,
+      "grad_norm": 0.8319103121757507,
+      "learning_rate": 2.389398173636972e-06,
+      "loss": 0.1071,
+      "step": 9789
+    },
+    {
+      "epoch": 1.5861957226182761,
+      "grad_norm": 0.884433388710022,
+      "learning_rate": 2.3889612993960233e-06,
+      "loss": 0.1163,
+      "step": 9790
+    },
+    {
+      "epoch": 1.5863577446532728,
+      "grad_norm": 0.813884973526001,
+      "learning_rate": 2.3885244285525892e-06,
+      "loss": 0.1017,
+      "step": 9791
+    },
+    {
+      "epoch": 1.5865197666882696,
+      "grad_norm": 0.8731319308280945,
+      "learning_rate": 2.3880875611200387e-06,
+      "loss": 0.1089,
+      "step": 9792
+    },
+    {
+      "epoch": 1.5866817887232663,
+      "grad_norm": 0.7939242720603943,
+      "learning_rate": 2.387650697111737e-06,
+      "loss": 0.1039,
+      "step": 9793
+    },
+    {
+      "epoch": 1.5868438107582632,
+      "grad_norm": 0.7616491317749023,
+      "learning_rate": 2.3872138365410525e-06,
+      "loss": 0.1019,
+      "step": 9794
+    },
+    {
+      "epoch": 1.58700583279326,
+      "grad_norm": 0.8765544295310974,
+      "learning_rate": 2.386776979421352e-06,
+      "loss": 0.1133,
+      "step": 9795
+    },
+    {
+      "epoch": 1.5871678548282566,
+      "grad_norm": 0.8053047060966492,
+      "learning_rate": 2.3863401257660016e-06,
+      "loss": 0.0989,
+      "step": 9796
+    },
+    {
+      "epoch": 1.5873298768632536,
+      "grad_norm": 0.7080039381980896,
+      "learning_rate": 2.3859032755883677e-06,
+      "loss": 0.0945,
+      "step": 9797
+    },
+    {
+      "epoch": 1.58749189889825,
+      "grad_norm": 0.7942770719528198,
+      "learning_rate": 2.3854664289018182e-06,
+      "loss": 0.1005,
+      "step": 9798
+    },
+    {
+      "epoch": 1.587653920933247,
+      "grad_norm": 0.8453302383422852,
+      "learning_rate": 2.3850295857197193e-06,
+      "loss": 0.1095,
+      "step": 9799
+    },
+    {
+      "epoch": 1.5878159429682437,
+      "grad_norm": 0.9450112581253052,
+      "learning_rate": 2.3845927460554363e-06,
+      "loss": 0.1154,
+      "step": 9800
+    },
+    {
+      "epoch": 1.5879779650032404,
+      "grad_norm": 0.8308724164962769,
+      "learning_rate": 2.3841559099223363e-06,
+      "loss": 0.1083,
+      "step": 9801
+    },
+    {
+      "epoch": 1.5881399870382373,
+      "grad_norm": 0.8302947878837585,
+      "learning_rate": 2.383719077333784e-06,
+      "loss": 0.0977,
+      "step": 9802
+    },
+    {
+      "epoch": 1.5883020090732338,
+      "grad_norm": 0.8785433173179626,
+      "learning_rate": 2.3832822483031477e-06,
+      "loss": 0.1076,
+      "step": 9803
+    },
+    {
+      "epoch": 1.5884640311082308,
+      "grad_norm": 0.7806341648101807,
+      "learning_rate": 2.382845422843792e-06,
+      "loss": 0.1086,
+      "step": 9804
+    },
+    {
+      "epoch": 1.5886260531432275,
+      "grad_norm": 0.737848699092865,
+      "learning_rate": 2.382408600969083e-06,
+      "loss": 0.0928,
+      "step": 9805
+    },
+    {
+      "epoch": 1.5887880751782242,
+      "grad_norm": 0.8146355748176575,
+      "learning_rate": 2.381971782692386e-06,
+      "loss": 0.1076,
+      "step": 9806
+    },
+    {
+      "epoch": 1.5889500972132211,
+      "grad_norm": 0.7324455380439758,
+      "learning_rate": 2.3815349680270654e-06,
+      "loss": 0.0893,
+      "step": 9807
+    },
+    {
+      "epoch": 1.5891121192482176,
+      "grad_norm": 0.781540036201477,
+      "learning_rate": 2.3810981569864898e-06,
+      "loss": 0.097,
+      "step": 9808
+    },
+    {
+      "epoch": 1.5892741412832145,
+      "grad_norm": 0.7014777660369873,
+      "learning_rate": 2.3806613495840227e-06,
+      "loss": 0.0886,
+      "step": 9809
+    },
+    {
+      "epoch": 1.5894361633182112,
+      "grad_norm": 0.8267843723297119,
+      "learning_rate": 2.38022454583303e-06,
+      "loss": 0.1021,
+      "step": 9810
+    },
+    {
+      "epoch": 1.589598185353208,
+      "grad_norm": 0.7938618063926697,
+      "learning_rate": 2.379787745746875e-06,
+      "loss": 0.0953,
+      "step": 9811
+    },
+    {
+      "epoch": 1.589760207388205,
+      "grad_norm": 0.9035655856132507,
+      "learning_rate": 2.379350949338924e-06,
+      "loss": 0.1138,
+      "step": 9812
+    },
+    {
+      "epoch": 1.5899222294232016,
+      "grad_norm": 0.7727793455123901,
+      "learning_rate": 2.3789141566225437e-06,
+      "loss": 0.0966,
+      "step": 9813
+    },
+    {
+      "epoch": 1.5900842514581983,
+      "grad_norm": 1.0545909404754639,
+      "learning_rate": 2.378477367611096e-06,
+      "loss": 0.12,
+      "step": 9814
+    },
+    {
+      "epoch": 1.590246273493195,
+      "grad_norm": 0.8968018889427185,
+      "learning_rate": 2.378040582317947e-06,
+      "loss": 0.109,
+      "step": 9815
+    },
+    {
+      "epoch": 1.5904082955281917,
+      "grad_norm": 0.7647299766540527,
+      "learning_rate": 2.377603800756461e-06,
+      "loss": 0.0939,
+      "step": 9816
+    },
+    {
+      "epoch": 1.5905703175631887,
+      "grad_norm": 0.8128365278244019,
+      "learning_rate": 2.377167022940002e-06,
+      "loss": 0.1,
+      "step": 9817
+    },
+    {
+      "epoch": 1.5907323395981854,
+      "grad_norm": 0.7933886051177979,
+      "learning_rate": 2.376730248881935e-06,
+      "loss": 0.1007,
+      "step": 9818
+    },
+    {
+      "epoch": 1.590894361633182,
+      "grad_norm": 0.8362367153167725,
+      "learning_rate": 2.376293478595625e-06,
+      "loss": 0.1051,
+      "step": 9819
+    },
+    {
+      "epoch": 1.591056383668179,
+      "grad_norm": 0.7798475027084351,
+      "learning_rate": 2.3758567120944345e-06,
+      "loss": 0.0938,
+      "step": 9820
+    },
+    {
+      "epoch": 1.5912184057031755,
+      "grad_norm": 0.8067512512207031,
+      "learning_rate": 2.3754199493917277e-06,
+      "loss": 0.0933,
+      "step": 9821
+    },
+    {
+      "epoch": 1.5913804277381725,
+      "grad_norm": 0.7828888893127441,
+      "learning_rate": 2.3749831905008704e-06,
+      "loss": 0.1024,
+      "step": 9822
+    },
+    {
+      "epoch": 1.5915424497731692,
+      "grad_norm": 0.7774432897567749,
+      "learning_rate": 2.3745464354352236e-06,
+      "loss": 0.0959,
+      "step": 9823
+    },
+    {
+      "epoch": 1.5917044718081659,
+      "grad_norm": 0.7905421257019043,
+      "learning_rate": 2.374109684208153e-06,
+      "loss": 0.0848,
+      "step": 9824
+    },
+    {
+      "epoch": 1.5918664938431628,
+      "grad_norm": 0.9159008264541626,
+      "learning_rate": 2.3736729368330212e-06,
+      "loss": 0.1114,
+      "step": 9825
+    },
+    {
+      "epoch": 1.5920285158781593,
+      "grad_norm": 0.9248816967010498,
+      "learning_rate": 2.3732361933231917e-06,
+      "loss": 0.0909,
+      "step": 9826
+    },
+    {
+      "epoch": 1.5921905379131562,
+      "grad_norm": 0.7748779654502869,
+      "learning_rate": 2.3727994536920276e-06,
+      "loss": 0.0993,
+      "step": 9827
+    },
+    {
+      "epoch": 1.592352559948153,
+      "grad_norm": 0.8885847926139832,
+      "learning_rate": 2.3723627179528935e-06,
+      "loss": 0.1007,
+      "step": 9828
+    },
+    {
+      "epoch": 1.5925145819831497,
+      "grad_norm": 0.7419458031654358,
+      "learning_rate": 2.3719259861191506e-06,
+      "loss": 0.0938,
+      "step": 9829
+    },
+    {
+      "epoch": 1.5926766040181466,
+      "grad_norm": 0.7908516526222229,
+      "learning_rate": 2.371489258204163e-06,
+      "loss": 0.0994,
+      "step": 9830
+    },
+    {
+      "epoch": 1.592838626053143,
+      "grad_norm": 0.8385130763053894,
+      "learning_rate": 2.3710525342212925e-06,
+      "loss": 0.1031,
+      "step": 9831
+    },
+    {
+      "epoch": 1.59300064808814,
+      "grad_norm": 0.7185788750648499,
+      "learning_rate": 2.3706158141839025e-06,
+      "loss": 0.0933,
+      "step": 9832
+    },
+    {
+      "epoch": 1.5931626701231367,
+      "grad_norm": 0.7869822382926941,
+      "learning_rate": 2.3701790981053556e-06,
+      "loss": 0.0997,
+      "step": 9833
+    },
+    {
+      "epoch": 1.5933246921581334,
+      "grad_norm": 0.881678581237793,
+      "learning_rate": 2.3697423859990147e-06,
+      "loss": 0.1069,
+      "step": 9834
+    },
+    {
+      "epoch": 1.5934867141931304,
+      "grad_norm": 0.9156317710876465,
+      "learning_rate": 2.3693056778782407e-06,
+      "loss": 0.1083,
+      "step": 9835
+    },
+    {
+      "epoch": 1.5936487362281269,
+      "grad_norm": 0.9538189768791199,
+      "learning_rate": 2.3688689737563965e-06,
+      "loss": 0.0896,
+      "step": 9836
+    },
+    {
+      "epoch": 1.5938107582631238,
+      "grad_norm": 0.7372860312461853,
+      "learning_rate": 2.3684322736468457e-06,
+      "loss": 0.0993,
+      "step": 9837
+    },
+    {
+      "epoch": 1.5939727802981205,
+      "grad_norm": 0.8044807314872742,
+      "learning_rate": 2.367995577562948e-06,
+      "loss": 0.1033,
+      "step": 9838
+    },
+    {
+      "epoch": 1.5941348023331172,
+      "grad_norm": 0.8724144697189331,
+      "learning_rate": 2.3675588855180668e-06,
+      "loss": 0.1125,
+      "step": 9839
+    },
+    {
+      "epoch": 1.5942968243681142,
+      "grad_norm": 0.7166265249252319,
+      "learning_rate": 2.3671221975255616e-06,
+      "loss": 0.0858,
+      "step": 9840
+    },
+    {
+      "epoch": 1.5944588464031109,
+      "grad_norm": 0.9747242331504822,
+      "learning_rate": 2.3666855135987972e-06,
+      "loss": 0.1067,
+      "step": 9841
+    },
+    {
+      "epoch": 1.5946208684381076,
+      "grad_norm": 0.8712856769561768,
+      "learning_rate": 2.366248833751133e-06,
+      "loss": 0.1019,
+      "step": 9842
+    },
+    {
+      "epoch": 1.5947828904731045,
+      "grad_norm": 0.8115395903587341,
+      "learning_rate": 2.3658121579959314e-06,
+      "loss": 0.0982,
+      "step": 9843
+    },
+    {
+      "epoch": 1.594944912508101,
+      "grad_norm": 0.9539070129394531,
+      "learning_rate": 2.365375486346552e-06,
+      "loss": 0.1215,
+      "step": 9844
+    },
+    {
+      "epoch": 1.595106934543098,
+      "grad_norm": 0.870464026927948,
+      "learning_rate": 2.3649388188163572e-06,
+      "loss": 0.1112,
+      "step": 9845
+    },
+    {
+      "epoch": 1.5952689565780946,
+      "grad_norm": 0.7583725452423096,
+      "learning_rate": 2.3645021554187086e-06,
+      "loss": 0.0935,
+      "step": 9846
+    },
+    {
+      "epoch": 1.5954309786130914,
+      "grad_norm": 0.8613632321357727,
+      "learning_rate": 2.364065496166965e-06,
+      "loss": 0.1064,
+      "step": 9847
+    },
+    {
+      "epoch": 1.5955930006480883,
+      "grad_norm": 0.8895267844200134,
+      "learning_rate": 2.3636288410744894e-06,
+      "loss": 0.1075,
+      "step": 9848
+    },
+    {
+      "epoch": 1.5957550226830848,
+      "grad_norm": 0.7960240840911865,
+      "learning_rate": 2.36319219015464e-06,
+      "loss": 0.0979,
+      "step": 9849
+    },
+    {
+      "epoch": 1.5959170447180817,
+      "grad_norm": 0.7139129042625427,
+      "learning_rate": 2.3627555434207787e-06,
+      "loss": 0.0916,
+      "step": 9850
+    },
+    {
+      "epoch": 1.5960790667530784,
+      "grad_norm": 0.9061529636383057,
+      "learning_rate": 2.3623189008862664e-06,
+      "loss": 0.1055,
+      "step": 9851
+    },
+    {
+      "epoch": 1.5962410887880751,
+      "grad_norm": 0.8803797364234924,
+      "learning_rate": 2.3618822625644624e-06,
+      "loss": 0.1099,
+      "step": 9852
+    },
+    {
+      "epoch": 1.596403110823072,
+      "grad_norm": 1.062333106994629,
+      "learning_rate": 2.3614456284687267e-06,
+      "loss": 0.1194,
+      "step": 9853
+    },
+    {
+      "epoch": 1.5965651328580686,
+      "grad_norm": 0.7200803756713867,
+      "learning_rate": 2.36100899861242e-06,
+      "loss": 0.0902,
+      "step": 9854
+    },
+    {
+      "epoch": 1.5967271548930655,
+      "grad_norm": 0.7464329600334167,
+      "learning_rate": 2.3605723730089e-06,
+      "loss": 0.0964,
+      "step": 9855
+    },
+    {
+      "epoch": 1.5968891769280622,
+      "grad_norm": 0.743965744972229,
+      "learning_rate": 2.3601357516715297e-06,
+      "loss": 0.0935,
+      "step": 9856
+    },
+    {
+      "epoch": 1.597051198963059,
+      "grad_norm": 0.8007190227508545,
+      "learning_rate": 2.3596991346136666e-06,
+      "loss": 0.1018,
+      "step": 9857
+    },
+    {
+      "epoch": 1.5972132209980558,
+      "grad_norm": 0.7653955817222595,
+      "learning_rate": 2.35926252184867e-06,
+      "loss": 0.0963,
+      "step": 9858
+    },
+    {
+      "epoch": 1.5973752430330523,
+      "grad_norm": 0.9055894017219543,
+      "learning_rate": 2.3588259133898995e-06,
+      "loss": 0.1082,
+      "step": 9859
+    },
+    {
+      "epoch": 1.5975372650680493,
+      "grad_norm": 0.8475822806358337,
+      "learning_rate": 2.3583893092507144e-06,
+      "loss": 0.1045,
+      "step": 9860
+    },
+    {
+      "epoch": 1.597699287103046,
+      "grad_norm": 0.8215590119361877,
+      "learning_rate": 2.357952709444474e-06,
+      "loss": 0.1003,
+      "step": 9861
+    },
+    {
+      "epoch": 1.5978613091380427,
+      "grad_norm": 0.7612400650978088,
+      "learning_rate": 2.3575161139845375e-06,
+      "loss": 0.0869,
+      "step": 9862
+    },
+    {
+      "epoch": 1.5980233311730396,
+      "grad_norm": 0.7190806865692139,
+      "learning_rate": 2.357079522884263e-06,
+      "loss": 0.0876,
+      "step": 9863
+    },
+    {
+      "epoch": 1.5981853532080363,
+      "grad_norm": 0.8134793043136597,
+      "learning_rate": 2.356642936157008e-06,
+      "loss": 0.0957,
+      "step": 9864
+    },
+    {
+      "epoch": 1.598347375243033,
+      "grad_norm": 0.8630086779594421,
+      "learning_rate": 2.3562063538161332e-06,
+      "loss": 0.1002,
+      "step": 9865
+    },
+    {
+      "epoch": 1.5985093972780298,
+      "grad_norm": 0.7363705635070801,
+      "learning_rate": 2.3557697758749966e-06,
+      "loss": 0.0914,
+      "step": 9866
+    },
+    {
+      "epoch": 1.5986714193130265,
+      "grad_norm": 0.9773649573326111,
+      "learning_rate": 2.355333202346955e-06,
+      "loss": 0.1158,
+      "step": 9867
+    },
+    {
+      "epoch": 1.5988334413480234,
+      "grad_norm": 0.8616942167282104,
+      "learning_rate": 2.3548966332453673e-06,
+      "loss": 0.1058,
+      "step": 9868
+    },
+    {
+      "epoch": 1.5989954633830201,
+      "grad_norm": 0.78786700963974,
+      "learning_rate": 2.354460068583591e-06,
+      "loss": 0.1064,
+      "step": 9869
+    },
+    {
+      "epoch": 1.5991574854180168,
+      "grad_norm": 0.8264607787132263,
+      "learning_rate": 2.3540235083749853e-06,
+      "loss": 0.0996,
+      "step": 9870
+    },
+    {
+      "epoch": 1.5993195074530138,
+      "grad_norm": 0.8713717460632324,
+      "learning_rate": 2.3535869526329067e-06,
+      "loss": 0.0998,
+      "step": 9871
+    },
+    {
+      "epoch": 1.5994815294880103,
+      "grad_norm": 0.8369178175926208,
+      "learning_rate": 2.3531504013707134e-06,
+      "loss": 0.1026,
+      "step": 9872
+    },
+    {
+      "epoch": 1.5996435515230072,
+      "grad_norm": 0.8778187036514282,
+      "learning_rate": 2.3527138546017623e-06,
+      "loss": 0.1061,
+      "step": 9873
+    },
+    {
+      "epoch": 1.599805573558004,
+      "grad_norm": 0.7916249632835388,
+      "learning_rate": 2.35227731233941e-06,
+      "loss": 0.0948,
+      "step": 9874
+    },
+    {
+      "epoch": 1.5999675955930006,
+      "grad_norm": 0.8178213238716125,
+      "learning_rate": 2.3518407745970155e-06,
+      "loss": 0.1131,
+      "step": 9875
+    },
+    {
+      "epoch": 1.6001296176279975,
+      "grad_norm": 0.7969405055046082,
+      "learning_rate": 2.3514042413879344e-06,
+      "loss": 0.0874,
+      "step": 9876
+    },
+    {
+      "epoch": 1.600291639662994,
+      "grad_norm": 0.9230097532272339,
+      "learning_rate": 2.3509677127255233e-06,
+      "loss": 0.1043,
+      "step": 9877
+    },
+    {
+      "epoch": 1.600453661697991,
+      "grad_norm": 0.8090982437133789,
+      "learning_rate": 2.350531188623141e-06,
+      "loss": 0.0954,
+      "step": 9878
+    },
+    {
+      "epoch": 1.6006156837329877,
+      "grad_norm": 0.8020190000534058,
+      "learning_rate": 2.3500946690941407e-06,
+      "loss": 0.1007,
+      "step": 9879
+    },
+    {
+      "epoch": 1.6007777057679844,
+      "grad_norm": 0.8556835651397705,
+      "learning_rate": 2.349658154151882e-06,
+      "loss": 0.105,
+      "step": 9880
+    },
+    {
+      "epoch": 1.6009397278029813,
+      "grad_norm": 0.7550347447395325,
+      "learning_rate": 2.34922164380972e-06,
+      "loss": 0.0951,
+      "step": 9881
+    },
+    {
+      "epoch": 1.6011017498379778,
+      "grad_norm": 0.8687736988067627,
+      "learning_rate": 2.3487851380810106e-06,
+      "loss": 0.1147,
+      "step": 9882
+    },
+    {
+      "epoch": 1.6012637718729748,
+      "grad_norm": 0.8163692951202393,
+      "learning_rate": 2.3483486369791106e-06,
+      "loss": 0.0996,
+      "step": 9883
+    },
+    {
+      "epoch": 1.6014257939079715,
+      "grad_norm": 0.9547244310379028,
+      "learning_rate": 2.3479121405173736e-06,
+      "loss": 0.1209,
+      "step": 9884
+    },
+    {
+      "epoch": 1.6015878159429682,
+      "grad_norm": 0.7805865406990051,
+      "learning_rate": 2.3474756487091586e-06,
+      "loss": 0.1022,
+      "step": 9885
+    },
+    {
+      "epoch": 1.601749837977965,
+      "grad_norm": 0.7808001041412354,
+      "learning_rate": 2.347039161567819e-06,
+      "loss": 0.1002,
+      "step": 9886
+    },
+    {
+      "epoch": 1.6019118600129616,
+      "grad_norm": 0.8721002340316772,
+      "learning_rate": 2.346602679106712e-06,
+      "loss": 0.1171,
+      "step": 9887
+    },
+    {
+      "epoch": 1.6020738820479585,
+      "grad_norm": 0.8056678175926208,
+      "learning_rate": 2.34616620133919e-06,
+      "loss": 0.0971,
+      "step": 9888
+    },
+    {
+      "epoch": 1.6022359040829552,
+      "grad_norm": 0.7318929433822632,
+      "learning_rate": 2.345729728278611e-06,
+      "loss": 0.1024,
+      "step": 9889
+    },
+    {
+      "epoch": 1.602397926117952,
+      "grad_norm": 0.7396207451820374,
+      "learning_rate": 2.345293259938329e-06,
+      "loss": 0.0944,
+      "step": 9890
+    },
+    {
+      "epoch": 1.6025599481529489,
+      "grad_norm": 0.8219144344329834,
+      "learning_rate": 2.3448567963316987e-06,
+      "loss": 0.1008,
+      "step": 9891
+    },
+    {
+      "epoch": 1.6027219701879456,
+      "grad_norm": 0.7780699133872986,
+      "learning_rate": 2.3444203374720755e-06,
+      "loss": 0.0938,
+      "step": 9892
+    },
+    {
+      "epoch": 1.6028839922229423,
+      "grad_norm": 0.7455930709838867,
+      "learning_rate": 2.3439838833728122e-06,
+      "loss": 0.0948,
+      "step": 9893
+    },
+    {
+      "epoch": 1.6030460142579392,
+      "grad_norm": 0.8056654334068298,
+      "learning_rate": 2.3435474340472657e-06,
+      "loss": 0.098,
+      "step": 9894
+    },
+    {
+      "epoch": 1.6032080362929357,
+      "grad_norm": 0.7863870859146118,
+      "learning_rate": 2.3431109895087886e-06,
+      "loss": 0.1008,
+      "step": 9895
+    },
+    {
+      "epoch": 1.6033700583279327,
+      "grad_norm": 0.7881834506988525,
+      "learning_rate": 2.3426745497707364e-06,
+      "loss": 0.1083,
+      "step": 9896
+    },
+    {
+      "epoch": 1.6035320803629294,
+      "grad_norm": 0.8019961714744568,
+      "learning_rate": 2.3422381148464614e-06,
+      "loss": 0.0984,
+      "step": 9897
+    },
+    {
+      "epoch": 1.603694102397926,
+      "grad_norm": 0.8999032974243164,
+      "learning_rate": 2.341801684749318e-06,
+      "loss": 0.1112,
+      "step": 9898
+    },
+    {
+      "epoch": 1.603856124432923,
+      "grad_norm": 0.7355244755744934,
+      "learning_rate": 2.341365259492661e-06,
+      "loss": 0.0951,
+      "step": 9899
+    },
+    {
+      "epoch": 1.6040181464679195,
+      "grad_norm": 0.8507843017578125,
+      "learning_rate": 2.3409288390898427e-06,
+      "loss": 0.1107,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6041801685029164,
+      "grad_norm": 0.7582950592041016,
+      "learning_rate": 2.3404924235542175e-06,
+      "loss": 0.0901,
+      "step": 9901
+    },
+    {
+      "epoch": 1.6043421905379132,
+      "grad_norm": 0.9285005927085876,
+      "learning_rate": 2.3400560128991377e-06,
+      "loss": 0.1125,
+      "step": 9902
+    },
+    {
+      "epoch": 1.6045042125729099,
+      "grad_norm": 0.7797400951385498,
+      "learning_rate": 2.3396196071379563e-06,
+      "loss": 0.1004,
+      "step": 9903
+    },
+    {
+      "epoch": 1.6046662346079068,
+      "grad_norm": 0.8635115623474121,
+      "learning_rate": 2.3391832062840273e-06,
+      "loss": 0.1106,
+      "step": 9904
+    },
+    {
+      "epoch": 1.6048282566429033,
+      "grad_norm": 0.7101535797119141,
+      "learning_rate": 2.3387468103507037e-06,
+      "loss": 0.0905,
+      "step": 9905
+    },
+    {
+      "epoch": 1.6049902786779002,
+      "grad_norm": 0.8763110041618347,
+      "learning_rate": 2.338310419351337e-06,
+      "loss": 0.1038,
+      "step": 9906
+    },
+    {
+      "epoch": 1.605152300712897,
+      "grad_norm": 0.6906706094741821,
+      "learning_rate": 2.3378740332992794e-06,
+      "loss": 0.0851,
+      "step": 9907
+    },
+    {
+      "epoch": 1.6053143227478937,
+      "grad_norm": 0.7445785403251648,
+      "learning_rate": 2.3374376522078852e-06,
+      "loss": 0.0896,
+      "step": 9908
+    },
+    {
+      "epoch": 1.6054763447828906,
+      "grad_norm": 0.9417932629585266,
+      "learning_rate": 2.337001276090505e-06,
+      "loss": 0.1059,
+      "step": 9909
+    },
+    {
+      "epoch": 1.605638366817887,
+      "grad_norm": 0.8038756847381592,
+      "learning_rate": 2.3365649049604917e-06,
+      "loss": 0.0981,
+      "step": 9910
+    },
+    {
+      "epoch": 1.605800388852884,
+      "grad_norm": 0.903634786605835,
+      "learning_rate": 2.3361285388311963e-06,
+      "loss": 0.1137,
+      "step": 9911
+    },
+    {
+      "epoch": 1.6059624108878807,
+      "grad_norm": 1.037934422492981,
+      "learning_rate": 2.3356921777159705e-06,
+      "loss": 0.1227,
+      "step": 9912
+    },
+    {
+      "epoch": 1.6061244329228774,
+      "grad_norm": 0.8287703990936279,
+      "learning_rate": 2.335255821628167e-06,
+      "loss": 0.0974,
+      "step": 9913
+    },
+    {
+      "epoch": 1.6062864549578744,
+      "grad_norm": 0.8133599758148193,
+      "learning_rate": 2.334819470581137e-06,
+      "loss": 0.0945,
+      "step": 9914
+    },
+    {
+      "epoch": 1.606448476992871,
+      "grad_norm": 0.8264515995979309,
+      "learning_rate": 2.3343831245882316e-06,
+      "loss": 0.108,
+      "step": 9915
+    },
+    {
+      "epoch": 1.6066104990278678,
+      "grad_norm": 0.7875205874443054,
+      "learning_rate": 2.3339467836628018e-06,
+      "loss": 0.0986,
+      "step": 9916
+    },
+    {
+      "epoch": 1.6067725210628645,
+      "grad_norm": 0.784372091293335,
+      "learning_rate": 2.333510447818198e-06,
+      "loss": 0.1,
+      "step": 9917
+    },
+    {
+      "epoch": 1.6069345430978612,
+      "grad_norm": 0.9532830715179443,
+      "learning_rate": 2.3330741170677713e-06,
+      "loss": 0.1321,
+      "step": 9918
+    },
+    {
+      "epoch": 1.6070965651328581,
+      "grad_norm": 0.8498433828353882,
+      "learning_rate": 2.3326377914248733e-06,
+      "loss": 0.1047,
+      "step": 9919
+    },
+    {
+      "epoch": 1.6072585871678549,
+      "grad_norm": 0.9842066168785095,
+      "learning_rate": 2.3322014709028545e-06,
+      "loss": 0.1192,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6074206092028516,
+      "grad_norm": 0.6417335867881775,
+      "learning_rate": 2.3317651555150636e-06,
+      "loss": 0.0823,
+      "step": 9921
+    },
+    {
+      "epoch": 1.6075826312378485,
+      "grad_norm": 0.8965574502944946,
+      "learning_rate": 2.3313288452748515e-06,
+      "loss": 0.1156,
+      "step": 9922
+    },
+    {
+      "epoch": 1.607744653272845,
+      "grad_norm": 0.810722291469574,
+      "learning_rate": 2.3308925401955694e-06,
+      "loss": 0.0982,
+      "step": 9923
+    },
+    {
+      "epoch": 1.607906675307842,
+      "grad_norm": 0.8201885223388672,
+      "learning_rate": 2.3304562402905662e-06,
+      "loss": 0.1068,
+      "step": 9924
+    },
+    {
+      "epoch": 1.6080686973428386,
+      "grad_norm": 0.8500423431396484,
+      "learning_rate": 2.3300199455731922e-06,
+      "loss": 0.1053,
+      "step": 9925
+    },
+    {
+      "epoch": 1.6082307193778353,
+      "grad_norm": 0.8203928470611572,
+      "learning_rate": 2.329583656056796e-06,
+      "loss": 0.0957,
+      "step": 9926
+    },
+    {
+      "epoch": 1.6083927414128323,
+      "grad_norm": 0.7739825248718262,
+      "learning_rate": 2.329147371754727e-06,
+      "loss": 0.1001,
+      "step": 9927
+    },
+    {
+      "epoch": 1.6085547634478288,
+      "grad_norm": 0.8251372575759888,
+      "learning_rate": 2.3287110926803354e-06,
+      "loss": 0.1153,
+      "step": 9928
+    },
+    {
+      "epoch": 1.6087167854828257,
+      "grad_norm": 0.7434987425804138,
+      "learning_rate": 2.3282748188469704e-06,
+      "loss": 0.0985,
+      "step": 9929
+    },
+    {
+      "epoch": 1.6088788075178224,
+      "grad_norm": 1.1180413961410522,
+      "learning_rate": 2.32783855026798e-06,
+      "loss": 0.1098,
+      "step": 9930
+    },
+    {
+      "epoch": 1.6090408295528191,
+      "grad_norm": 0.8994811177253723,
+      "learning_rate": 2.3274022869567123e-06,
+      "loss": 0.1136,
+      "step": 9931
+    },
+    {
+      "epoch": 1.609202851587816,
+      "grad_norm": 0.8187955021858215,
+      "learning_rate": 2.3269660289265184e-06,
+      "loss": 0.1092,
+      "step": 9932
+    },
+    {
+      "epoch": 1.6093648736228126,
+      "grad_norm": 0.894491970539093,
+      "learning_rate": 2.3265297761907447e-06,
+      "loss": 0.1171,
+      "step": 9933
+    },
+    {
+      "epoch": 1.6095268956578095,
+      "grad_norm": 0.9280018210411072,
+      "learning_rate": 2.3260935287627408e-06,
+      "loss": 0.1186,
+      "step": 9934
+    },
+    {
+      "epoch": 1.6096889176928062,
+      "grad_norm": 0.9082358479499817,
+      "learning_rate": 2.3256572866558533e-06,
+      "loss": 0.1149,
+      "step": 9935
+    },
+    {
+      "epoch": 1.609850939727803,
+      "grad_norm": 0.8570298552513123,
+      "learning_rate": 2.3252210498834306e-06,
+      "loss": 0.1134,
+      "step": 9936
+    },
+    {
+      "epoch": 1.6100129617627998,
+      "grad_norm": 0.8263987898826599,
+      "learning_rate": 2.3247848184588208e-06,
+      "loss": 0.0962,
+      "step": 9937
+    },
+    {
+      "epoch": 1.6101749837977966,
+      "grad_norm": 0.8012087345123291,
+      "learning_rate": 2.3243485923953725e-06,
+      "loss": 0.1005,
+      "step": 9938
+    },
+    {
+      "epoch": 1.6103370058327933,
+      "grad_norm": 0.8303671479225159,
+      "learning_rate": 2.323912371706432e-06,
+      "loss": 0.1127,
+      "step": 9939
+    },
+    {
+      "epoch": 1.61049902786779,
+      "grad_norm": 0.9363792538642883,
+      "learning_rate": 2.323476156405347e-06,
+      "loss": 0.1046,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6106610499027867,
+      "grad_norm": 0.8461704254150391,
+      "learning_rate": 2.323039946505463e-06,
+      "loss": 0.1076,
+      "step": 9941
+    },
+    {
+      "epoch": 1.6108230719377836,
+      "grad_norm": 0.7788037061691284,
+      "learning_rate": 2.3226037420201296e-06,
+      "loss": 0.0963,
+      "step": 9942
+    },
+    {
+      "epoch": 1.6109850939727803,
+      "grad_norm": 0.9676366448402405,
+      "learning_rate": 2.3221675429626925e-06,
+      "loss": 0.1045,
+      "step": 9943
+    },
+    {
+      "epoch": 1.611147116007777,
+      "grad_norm": 0.9113138318061829,
+      "learning_rate": 2.3217313493464977e-06,
+      "loss": 0.108,
+      "step": 9944
+    },
+    {
+      "epoch": 1.611309138042774,
+      "grad_norm": 0.6841544508934021,
+      "learning_rate": 2.3212951611848927e-06,
+      "loss": 0.0867,
+      "step": 9945
+    },
+    {
+      "epoch": 1.6114711600777705,
+      "grad_norm": 0.7147634625434875,
+      "learning_rate": 2.320858978491222e-06,
+      "loss": 0.0938,
+      "step": 9946
+    },
+    {
+      "epoch": 1.6116331821127674,
+      "grad_norm": 0.8967075347900391,
+      "learning_rate": 2.3204228012788346e-06,
+      "loss": 0.1185,
+      "step": 9947
+    },
+    {
+      "epoch": 1.6117952041477641,
+      "grad_norm": 0.7635133266448975,
+      "learning_rate": 2.319986629561074e-06,
+      "loss": 0.0941,
+      "step": 9948
+    },
+    {
+      "epoch": 1.6119572261827608,
+      "grad_norm": 0.9004403352737427,
+      "learning_rate": 2.319550463351288e-06,
+      "loss": 0.1137,
+      "step": 9949
+    },
+    {
+      "epoch": 1.6121192482177578,
+      "grad_norm": 0.7615978121757507,
+      "learning_rate": 2.3191143026628206e-06,
+      "loss": 0.0906,
+      "step": 9950
+    },
+    {
+      "epoch": 1.6122812702527543,
+      "grad_norm": 0.8137632608413696,
+      "learning_rate": 2.3186781475090168e-06,
+      "loss": 0.0992,
+      "step": 9951
+    },
+    {
+      "epoch": 1.6124432922877512,
+      "grad_norm": 0.8467097282409668,
+      "learning_rate": 2.318241997903224e-06,
+      "loss": 0.1082,
+      "step": 9952
+    },
+    {
+      "epoch": 1.612605314322748,
+      "grad_norm": 0.8358951210975647,
+      "learning_rate": 2.317805853858786e-06,
+      "loss": 0.1054,
+      "step": 9953
+    },
+    {
+      "epoch": 1.6127673363577446,
+      "grad_norm": 0.8378834128379822,
+      "learning_rate": 2.3173697153890486e-06,
+      "loss": 0.1095,
+      "step": 9954
+    },
+    {
+      "epoch": 1.6129293583927415,
+      "grad_norm": 0.7304298281669617,
+      "learning_rate": 2.316933582507354e-06,
+      "loss": 0.0923,
+      "step": 9955
+    },
+    {
+      "epoch": 1.613091380427738,
+      "grad_norm": 0.7708770632743835,
+      "learning_rate": 2.31649745522705e-06,
+      "loss": 0.0938,
+      "step": 9956
+    },
+    {
+      "epoch": 1.613253402462735,
+      "grad_norm": 0.8786473870277405,
+      "learning_rate": 2.31606133356148e-06,
+      "loss": 0.111,
+      "step": 9957
+    },
+    {
+      "epoch": 1.6134154244977317,
+      "grad_norm": 0.724189281463623,
+      "learning_rate": 2.3156252175239883e-06,
+      "loss": 0.0913,
+      "step": 9958
+    },
+    {
+      "epoch": 1.6135774465327284,
+      "grad_norm": 0.6984865665435791,
+      "learning_rate": 2.3151891071279183e-06,
+      "loss": 0.0833,
+      "step": 9959
+    },
+    {
+      "epoch": 1.6137394685677253,
+      "grad_norm": 0.833328902721405,
+      "learning_rate": 2.3147530023866136e-06,
+      "loss": 0.1095,
+      "step": 9960
+    },
+    {
+      "epoch": 1.6139014906027218,
+      "grad_norm": 0.8859132528305054,
+      "learning_rate": 2.3143169033134204e-06,
+      "loss": 0.1241,
+      "step": 9961
+    },
+    {
+      "epoch": 1.6140635126377187,
+      "grad_norm": 0.8652672171592712,
+      "learning_rate": 2.3138808099216796e-06,
+      "loss": 0.1024,
+      "step": 9962
+    },
+    {
+      "epoch": 1.6142255346727155,
+      "grad_norm": 0.7666911482810974,
+      "learning_rate": 2.313444722224736e-06,
+      "loss": 0.0994,
+      "step": 9963
+    },
+    {
+      "epoch": 1.6143875567077122,
+      "grad_norm": 0.6945036053657532,
+      "learning_rate": 2.3130086402359327e-06,
+      "loss": 0.0851,
+      "step": 9964
+    },
+    {
+      "epoch": 1.614549578742709,
+      "grad_norm": 0.8811757564544678,
+      "learning_rate": 2.3125725639686116e-06,
+      "loss": 0.11,
+      "step": 9965
+    },
+    {
+      "epoch": 1.6147116007777058,
+      "grad_norm": 0.7738437056541443,
+      "learning_rate": 2.312136493436117e-06,
+      "loss": 0.0957,
+      "step": 9966
+    },
+    {
+      "epoch": 1.6148736228127025,
+      "grad_norm": 0.8249509334564209,
+      "learning_rate": 2.311700428651791e-06,
+      "loss": 0.1109,
+      "step": 9967
+    },
+    {
+      "epoch": 1.6150356448476992,
+      "grad_norm": 0.8863063454627991,
+      "learning_rate": 2.311264369628976e-06,
+      "loss": 0.1033,
+      "step": 9968
+    },
+    {
+      "epoch": 1.615197666882696,
+      "grad_norm": 0.7870522141456604,
+      "learning_rate": 2.3108283163810155e-06,
+      "loss": 0.0972,
+      "step": 9969
+    },
+    {
+      "epoch": 1.6153596889176929,
+      "grad_norm": 0.8956674337387085,
+      "learning_rate": 2.3103922689212494e-06,
+      "loss": 0.1076,
+      "step": 9970
+    },
+    {
+      "epoch": 1.6155217109526896,
+      "grad_norm": 0.7711880207061768,
+      "learning_rate": 2.3099562272630216e-06,
+      "loss": 0.0966,
+      "step": 9971
+    },
+    {
+      "epoch": 1.6156837329876863,
+      "grad_norm": 0.8216359615325928,
+      "learning_rate": 2.3095201914196732e-06,
+      "loss": 0.1039,
+      "step": 9972
+    },
+    {
+      "epoch": 1.6158457550226832,
+      "grad_norm": 0.7300341725349426,
+      "learning_rate": 2.309084161404547e-06,
+      "loss": 0.0933,
+      "step": 9973
+    },
+    {
+      "epoch": 1.6160077770576797,
+      "grad_norm": 0.783735454082489,
+      "learning_rate": 2.308648137230982e-06,
+      "loss": 0.0987,
+      "step": 9974
+    },
+    {
+      "epoch": 1.6161697990926767,
+      "grad_norm": 0.8075628876686096,
+      "learning_rate": 2.3082121189123213e-06,
+      "loss": 0.1045,
+      "step": 9975
+    },
+    {
+      "epoch": 1.6163318211276734,
+      "grad_norm": 0.8673627972602844,
+      "learning_rate": 2.3077761064619062e-06,
+      "loss": 0.1028,
+      "step": 9976
+    },
+    {
+      "epoch": 1.61649384316267,
+      "grad_norm": 0.8210455179214478,
+      "learning_rate": 2.307340099893077e-06,
+      "loss": 0.1024,
+      "step": 9977
+    },
+    {
+      "epoch": 1.616655865197667,
+      "grad_norm": 0.803597092628479,
+      "learning_rate": 2.3069040992191745e-06,
+      "loss": 0.0922,
+      "step": 9978
+    },
+    {
+      "epoch": 1.6168178872326635,
+      "grad_norm": 0.8294203877449036,
+      "learning_rate": 2.3064681044535385e-06,
+      "loss": 0.1002,
+      "step": 9979
+    },
+    {
+      "epoch": 1.6169799092676604,
+      "grad_norm": 0.7242622971534729,
+      "learning_rate": 2.3060321156095107e-06,
+      "loss": 0.0904,
+      "step": 9980
+    },
+    {
+      "epoch": 1.6171419313026572,
+      "grad_norm": 0.6747761964797974,
+      "learning_rate": 2.305596132700431e-06,
+      "loss": 0.0841,
+      "step": 9981
+    },
+    {
+      "epoch": 1.6173039533376539,
+      "grad_norm": 0.7614217400550842,
+      "learning_rate": 2.3051601557396393e-06,
+      "loss": 0.0929,
+      "step": 9982
+    },
+    {
+      "epoch": 1.6174659753726508,
+      "grad_norm": 0.8482229113578796,
+      "learning_rate": 2.304724184740475e-06,
+      "loss": 0.1065,
+      "step": 9983
+    },
+    {
+      "epoch": 1.6176279974076473,
+      "grad_norm": 0.8954651951789856,
+      "learning_rate": 2.3042882197162776e-06,
+      "loss": 0.1151,
+      "step": 9984
+    },
+    {
+      "epoch": 1.6177900194426442,
+      "grad_norm": 0.9330970048904419,
+      "learning_rate": 2.3038522606803882e-06,
+      "loss": 0.1117,
+      "step": 9985
+    },
+    {
+      "epoch": 1.617952041477641,
+      "grad_norm": 0.8899834156036377,
+      "learning_rate": 2.303416307646144e-06,
+      "loss": 0.106,
+      "step": 9986
+    },
+    {
+      "epoch": 1.6181140635126376,
+      "grad_norm": 0.8274690508842468,
+      "learning_rate": 2.302980360626886e-06,
+      "loss": 0.1017,
+      "step": 9987
+    },
+    {
+      "epoch": 1.6182760855476346,
+      "grad_norm": 0.8075361251831055,
+      "learning_rate": 2.3025444196359513e-06,
+      "loss": 0.1055,
+      "step": 9988
+    },
+    {
+      "epoch": 1.6184381075826313,
+      "grad_norm": 0.8113299012184143,
+      "learning_rate": 2.3021084846866793e-06,
+      "loss": 0.1039,
+      "step": 9989
+    },
+    {
+      "epoch": 1.618600129617628,
+      "grad_norm": 0.7555704712867737,
+      "learning_rate": 2.3016725557924095e-06,
+      "loss": 0.0869,
+      "step": 9990
+    },
+    {
+      "epoch": 1.6187621516526247,
+      "grad_norm": 0.8604114055633545,
+      "learning_rate": 2.3012366329664794e-06,
+      "loss": 0.109,
+      "step": 9991
+    },
+    {
+      "epoch": 1.6189241736876214,
+      "grad_norm": 0.8491653800010681,
+      "learning_rate": 2.3008007162222273e-06,
+      "loss": 0.1071,
+      "step": 9992
+    },
+    {
+      "epoch": 1.6190861957226184,
+      "grad_norm": 0.7831771969795227,
+      "learning_rate": 2.3003648055729917e-06,
+      "loss": 0.0946,
+      "step": 9993
+    },
+    {
+      "epoch": 1.619248217757615,
+      "grad_norm": 0.8524945378303528,
+      "learning_rate": 2.2999289010321092e-06,
+      "loss": 0.1057,
+      "step": 9994
+    },
+    {
+      "epoch": 1.6194102397926118,
+      "grad_norm": 0.7014240622520447,
+      "learning_rate": 2.299493002612918e-06,
+      "loss": 0.092,
+      "step": 9995
+    },
+    {
+      "epoch": 1.6195722618276087,
+      "grad_norm": 0.8205536007881165,
+      "learning_rate": 2.299057110328757e-06,
+      "loss": 0.1018,
+      "step": 9996
+    },
+    {
+      "epoch": 1.6197342838626052,
+      "grad_norm": 0.810691773891449,
+      "learning_rate": 2.298621224192961e-06,
+      "loss": 0.0999,
+      "step": 9997
+    },
+    {
+      "epoch": 1.6198963058976021,
+      "grad_norm": 0.7380390763282776,
+      "learning_rate": 2.298185344218868e-06,
+      "loss": 0.0948,
+      "step": 9998
+    },
+    {
+      "epoch": 1.6200583279325989,
+      "grad_norm": 0.8573938012123108,
+      "learning_rate": 2.297749470419815e-06,
+      "loss": 0.0986,
+      "step": 9999
+    },
+    {
+      "epoch": 1.6202203499675956,
+      "grad_norm": 0.749604344367981,
+      "learning_rate": 2.29731360280914e-06,
+      "loss": 0.1046,
+      "step": 10000
+    },
+    {
+      "epoch": 1.6203823720025925,
+      "grad_norm": 0.8711916208267212,
+      "learning_rate": 2.2968777414001773e-06,
+      "loss": 0.1106,
+      "step": 10001
+    },
+    {
+      "epoch": 1.620544394037589,
+      "grad_norm": 0.9774971008300781,
+      "learning_rate": 2.2964418862062655e-06,
+      "loss": 0.1258,
+      "step": 10002
+    },
+    {
+      "epoch": 1.620706416072586,
+      "grad_norm": 0.9113691449165344,
+      "learning_rate": 2.2960060372407377e-06,
+      "loss": 0.1038,
+      "step": 10003
+    },
+    {
+      "epoch": 1.6208684381075826,
+      "grad_norm": 0.7974822521209717,
+      "learning_rate": 2.2955701945169317e-06,
+      "loss": 0.1024,
+      "step": 10004
+    },
+    {
+      "epoch": 1.6210304601425793,
+      "grad_norm": 0.7355794310569763,
+      "learning_rate": 2.295134358048184e-06,
+      "loss": 0.093,
+      "step": 10005
+    },
+    {
+      "epoch": 1.6211924821775763,
+      "grad_norm": 0.8737595677375793,
+      "learning_rate": 2.294698527847829e-06,
+      "loss": 0.1173,
+      "step": 10006
+    },
+    {
+      "epoch": 1.6213545042125728,
+      "grad_norm": 0.8428897261619568,
+      "learning_rate": 2.2942627039292016e-06,
+      "loss": 0.1065,
+      "step": 10007
+    },
+    {
+      "epoch": 1.6215165262475697,
+      "grad_norm": 0.7106808423995972,
+      "learning_rate": 2.2938268863056373e-06,
+      "loss": 0.0899,
+      "step": 10008
+    },
+    {
+      "epoch": 1.6216785482825664,
+      "grad_norm": 0.989214301109314,
+      "learning_rate": 2.2933910749904724e-06,
+      "loss": 0.1183,
+      "step": 10009
+    },
+    {
+      "epoch": 1.6218405703175631,
+      "grad_norm": 0.7430517077445984,
+      "learning_rate": 2.29295526999704e-06,
+      "loss": 0.0899,
+      "step": 10010
+    },
+    {
+      "epoch": 1.62200259235256,
+      "grad_norm": 0.9804396033287048,
+      "learning_rate": 2.292519471338676e-06,
+      "loss": 0.1184,
+      "step": 10011
+    },
+    {
+      "epoch": 1.6221646143875565,
+      "grad_norm": 0.7713508009910583,
+      "learning_rate": 2.2920836790287134e-06,
+      "loss": 0.0946,
+      "step": 10012
+    },
+    {
+      "epoch": 1.6223266364225535,
+      "grad_norm": 0.7687755227088928,
+      "learning_rate": 2.2916478930804865e-06,
+      "loss": 0.1005,
+      "step": 10013
+    },
+    {
+      "epoch": 1.6224886584575502,
+      "grad_norm": 0.7584219574928284,
+      "learning_rate": 2.291212113507331e-06,
+      "loss": 0.0925,
+      "step": 10014
+    },
+    {
+      "epoch": 1.622650680492547,
+      "grad_norm": 0.6680698394775391,
+      "learning_rate": 2.2907763403225793e-06,
+      "loss": 0.0855,
+      "step": 10015
+    },
+    {
+      "epoch": 1.6228127025275438,
+      "grad_norm": 0.7672422528266907,
+      "learning_rate": 2.290340573539565e-06,
+      "loss": 0.0994,
+      "step": 10016
+    },
+    {
+      "epoch": 1.6229747245625405,
+      "grad_norm": 0.8922812938690186,
+      "learning_rate": 2.2899048131716223e-06,
+      "loss": 0.1152,
+      "step": 10017
+    },
+    {
+      "epoch": 1.6231367465975373,
+      "grad_norm": 0.8483216166496277,
+      "learning_rate": 2.2894690592320827e-06,
+      "loss": 0.1128,
+      "step": 10018
+    },
+    {
+      "epoch": 1.623298768632534,
+      "grad_norm": 0.6468700766563416,
+      "learning_rate": 2.2890333117342813e-06,
+      "loss": 0.086,
+      "step": 10019
+    },
+    {
+      "epoch": 1.6234607906675307,
+      "grad_norm": 0.7255584001541138,
+      "learning_rate": 2.2885975706915506e-06,
+      "loss": 0.0884,
+      "step": 10020
+    },
+    {
+      "epoch": 1.6236228127025276,
+      "grad_norm": 0.8776664137840271,
+      "learning_rate": 2.288161836117222e-06,
+      "loss": 0.1049,
+      "step": 10021
+    },
+    {
+      "epoch": 1.6237848347375243,
+      "grad_norm": 0.750275731086731,
+      "learning_rate": 2.287726108024628e-06,
+      "loss": 0.0964,
+      "step": 10022
+    },
+    {
+      "epoch": 1.623946856772521,
+      "grad_norm": 0.9269810318946838,
+      "learning_rate": 2.2872903864271017e-06,
+      "loss": 0.112,
+      "step": 10023
+    },
+    {
+      "epoch": 1.624108878807518,
+      "grad_norm": 0.8907431364059448,
+      "learning_rate": 2.2868546713379755e-06,
+      "loss": 0.1046,
+      "step": 10024
+    },
+    {
+      "epoch": 1.6242709008425145,
+      "grad_norm": 0.8910091519355774,
+      "learning_rate": 2.28641896277058e-06,
+      "loss": 0.0983,
+      "step": 10025
+    },
+    {
+      "epoch": 1.6244329228775114,
+      "grad_norm": 0.8178685307502747,
+      "learning_rate": 2.285983260738248e-06,
+      "loss": 0.1033,
+      "step": 10026
+    },
+    {
+      "epoch": 1.624594944912508,
+      "grad_norm": 0.8685728907585144,
+      "learning_rate": 2.2855475652543094e-06,
+      "loss": 0.1143,
+      "step": 10027
+    },
+    {
+      "epoch": 1.6247569669475048,
+      "grad_norm": 0.996419370174408,
+      "learning_rate": 2.285111876332097e-06,
+      "loss": 0.111,
+      "step": 10028
+    },
+    {
+      "epoch": 1.6249189889825018,
+      "grad_norm": 0.9297820925712585,
+      "learning_rate": 2.284676193984941e-06,
+      "loss": 0.1146,
+      "step": 10029
+    },
+    {
+      "epoch": 1.6250810110174982,
+      "grad_norm": 0.7859435677528381,
+      "learning_rate": 2.2842405182261725e-06,
+      "loss": 0.0975,
+      "step": 10030
+    },
+    {
+      "epoch": 1.6252430330524952,
+      "grad_norm": 0.8805785775184631,
+      "learning_rate": 2.2838048490691223e-06,
+      "loss": 0.104,
+      "step": 10031
+    },
+    {
+      "epoch": 1.625405055087492,
+      "grad_norm": 0.8301118612289429,
+      "learning_rate": 2.283369186527119e-06,
+      "loss": 0.1087,
+      "step": 10032
+    },
+    {
+      "epoch": 1.6255670771224886,
+      "grad_norm": 0.7605770230293274,
+      "learning_rate": 2.282933530613496e-06,
+      "loss": 0.0949,
+      "step": 10033
+    },
+    {
+      "epoch": 1.6257290991574855,
+      "grad_norm": 0.8895880579948425,
+      "learning_rate": 2.282497881341581e-06,
+      "loss": 0.1106,
+      "step": 10034
+    },
+    {
+      "epoch": 1.625891121192482,
+      "grad_norm": 0.8832073211669922,
+      "learning_rate": 2.282062238724705e-06,
+      "loss": 0.1151,
+      "step": 10035
+    },
+    {
+      "epoch": 1.626053143227479,
+      "grad_norm": 0.6511034965515137,
+      "learning_rate": 2.2816266027761965e-06,
+      "loss": 0.0846,
+      "step": 10036
+    },
+    {
+      "epoch": 1.6262151652624757,
+      "grad_norm": 0.7454072833061218,
+      "learning_rate": 2.2811909735093853e-06,
+      "loss": 0.0911,
+      "step": 10037
+    },
+    {
+      "epoch": 1.6263771872974724,
+      "grad_norm": 0.9449716806411743,
+      "learning_rate": 2.280755350937602e-06,
+      "loss": 0.1066,
+      "step": 10038
+    },
+    {
+      "epoch": 1.6265392093324693,
+      "grad_norm": 0.7572311162948608,
+      "learning_rate": 2.280319735074173e-06,
+      "loss": 0.0937,
+      "step": 10039
+    },
+    {
+      "epoch": 1.626701231367466,
+      "grad_norm": 0.8121317028999329,
+      "learning_rate": 2.27988412593243e-06,
+      "loss": 0.1019,
+      "step": 10040
+    },
+    {
+      "epoch": 1.6268632534024627,
+      "grad_norm": 0.8382424712181091,
+      "learning_rate": 2.279448523525699e-06,
+      "loss": 0.1101,
+      "step": 10041
+    },
+    {
+      "epoch": 1.6270252754374595,
+      "grad_norm": 0.7908496856689453,
+      "learning_rate": 2.279012927867309e-06,
+      "loss": 0.0963,
+      "step": 10042
+    },
+    {
+      "epoch": 1.6271872974724562,
+      "grad_norm": 0.8468271493911743,
+      "learning_rate": 2.278577338970589e-06,
+      "loss": 0.1045,
+      "step": 10043
+    },
+    {
+      "epoch": 1.627349319507453,
+      "grad_norm": 0.8646363019943237,
+      "learning_rate": 2.2781417568488677e-06,
+      "loss": 0.1113,
+      "step": 10044
+    },
+    {
+      "epoch": 1.6275113415424498,
+      "grad_norm": 0.8169398307800293,
+      "learning_rate": 2.2777061815154705e-06,
+      "loss": 0.1056,
+      "step": 10045
+    },
+    {
+      "epoch": 1.6276733635774465,
+      "grad_norm": 0.8818169832229614,
+      "learning_rate": 2.277270612983726e-06,
+      "loss": 0.1123,
+      "step": 10046
+    },
+    {
+      "epoch": 1.6278353856124435,
+      "grad_norm": 0.8572877049446106,
+      "learning_rate": 2.276835051266963e-06,
+      "loss": 0.0976,
+      "step": 10047
+    },
+    {
+      "epoch": 1.62799740764744,
+      "grad_norm": 0.8052610754966736,
+      "learning_rate": 2.2763994963785066e-06,
+      "loss": 0.1041,
+      "step": 10048
+    },
+    {
+      "epoch": 1.6281594296824369,
+      "grad_norm": 0.7382464408874512,
+      "learning_rate": 2.275963948331685e-06,
+      "loss": 0.0905,
+      "step": 10049
+    },
+    {
+      "epoch": 1.6283214517174336,
+      "grad_norm": 0.7890332341194153,
+      "learning_rate": 2.2755284071398243e-06,
+      "loss": 0.1029,
+      "step": 10050
+    },
+    {
+      "epoch": 1.6284834737524303,
+      "grad_norm": 0.794537365436554,
+      "learning_rate": 2.27509287281625e-06,
+      "loss": 0.0951,
+      "step": 10051
+    },
+    {
+      "epoch": 1.6286454957874272,
+      "grad_norm": 0.7802528738975525,
+      "learning_rate": 2.2746573453742905e-06,
+      "loss": 0.0916,
+      "step": 10052
+    },
+    {
+      "epoch": 1.6288075178224237,
+      "grad_norm": 0.8495976328849792,
+      "learning_rate": 2.2742218248272714e-06,
+      "loss": 0.1044,
+      "step": 10053
+    },
+    {
+      "epoch": 1.6289695398574207,
+      "grad_norm": 0.8333568572998047,
+      "learning_rate": 2.2737863111885175e-06,
+      "loss": 0.1017,
+      "step": 10054
+    },
+    {
+      "epoch": 1.6291315618924174,
+      "grad_norm": 0.7988643646240234,
+      "learning_rate": 2.273350804471355e-06,
+      "loss": 0.1004,
+      "step": 10055
+    },
+    {
+      "epoch": 1.629293583927414,
+      "grad_norm": 0.8200765252113342,
+      "learning_rate": 2.2729153046891095e-06,
+      "loss": 0.1005,
+      "step": 10056
+    },
+    {
+      "epoch": 1.629455605962411,
+      "grad_norm": 0.7785143852233887,
+      "learning_rate": 2.272479811855106e-06,
+      "loss": 0.0918,
+      "step": 10057
+    },
+    {
+      "epoch": 1.6296176279974075,
+      "grad_norm": 0.7829576730728149,
+      "learning_rate": 2.2720443259826702e-06,
+      "loss": 0.0891,
+      "step": 10058
+    },
+    {
+      "epoch": 1.6297796500324044,
+      "grad_norm": 0.8104037046432495,
+      "learning_rate": 2.271608847085126e-06,
+      "loss": 0.0981,
+      "step": 10059
+    },
+    {
+      "epoch": 1.6299416720674011,
+      "grad_norm": 0.9911179542541504,
+      "learning_rate": 2.2711733751757983e-06,
+      "loss": 0.1145,
+      "step": 10060
+    },
+    {
+      "epoch": 1.6301036941023979,
+      "grad_norm": 0.7678511142730713,
+      "learning_rate": 2.270737910268011e-06,
+      "loss": 0.0952,
+      "step": 10061
+    },
+    {
+      "epoch": 1.6302657161373948,
+      "grad_norm": 0.9023809432983398,
+      "learning_rate": 2.27030245237509e-06,
+      "loss": 0.0994,
+      "step": 10062
+    },
+    {
+      "epoch": 1.6304277381723913,
+      "grad_norm": 0.8956974148750305,
+      "learning_rate": 2.2698670015103574e-06,
+      "loss": 0.1128,
+      "step": 10063
+    },
+    {
+      "epoch": 1.6305897602073882,
+      "grad_norm": 0.8878682255744934,
+      "learning_rate": 2.2694315576871384e-06,
+      "loss": 0.0983,
+      "step": 10064
+    },
+    {
+      "epoch": 1.630751782242385,
+      "grad_norm": 0.7095435857772827,
+      "learning_rate": 2.2689961209187543e-06,
+      "loss": 0.0903,
+      "step": 10065
+    },
+    {
+      "epoch": 1.6309138042773816,
+      "grad_norm": 0.8130254149436951,
+      "learning_rate": 2.268560691218531e-06,
+      "loss": 0.0988,
+      "step": 10066
+    },
+    {
+      "epoch": 1.6310758263123786,
+      "grad_norm": 0.8411709070205688,
+      "learning_rate": 2.26812526859979e-06,
+      "loss": 0.1038,
+      "step": 10067
+    },
+    {
+      "epoch": 1.6312378483473753,
+      "grad_norm": 0.9909535050392151,
+      "learning_rate": 2.2676898530758554e-06,
+      "loss": 0.12,
+      "step": 10068
+    },
+    {
+      "epoch": 1.631399870382372,
+      "grad_norm": 0.8369352221488953,
+      "learning_rate": 2.2672544446600485e-06,
+      "loss": 0.0975,
+      "step": 10069
+    },
+    {
+      "epoch": 1.6315618924173687,
+      "grad_norm": 0.8436316847801208,
+      "learning_rate": 2.266819043365692e-06,
+      "loss": 0.102,
+      "step": 10070
+    },
+    {
+      "epoch": 1.6317239144523654,
+      "grad_norm": 0.7915019392967224,
+      "learning_rate": 2.2663836492061097e-06,
+      "loss": 0.1006,
+      "step": 10071
+    },
+    {
+      "epoch": 1.6318859364873624,
+      "grad_norm": 0.6947129964828491,
+      "learning_rate": 2.265948262194621e-06,
+      "loss": 0.0943,
+      "step": 10072
+    },
+    {
+      "epoch": 1.632047958522359,
+      "grad_norm": 0.738064169883728,
+      "learning_rate": 2.2655128823445507e-06,
+      "loss": 0.0989,
+      "step": 10073
+    },
+    {
+      "epoch": 1.6322099805573558,
+      "grad_norm": 0.9031091928482056,
+      "learning_rate": 2.2650775096692176e-06,
+      "loss": 0.1185,
+      "step": 10074
+    },
+    {
+      "epoch": 1.6323720025923527,
+      "grad_norm": 0.736219048500061,
+      "learning_rate": 2.264642144181944e-06,
+      "loss": 0.0998,
+      "step": 10075
+    },
+    {
+      "epoch": 1.6325340246273492,
+      "grad_norm": 0.7197458744049072,
+      "learning_rate": 2.2642067858960514e-06,
+      "loss": 0.0991,
+      "step": 10076
+    },
+    {
+      "epoch": 1.6326960466623461,
+      "grad_norm": 0.8605630993843079,
+      "learning_rate": 2.263771434824861e-06,
+      "loss": 0.1069,
+      "step": 10077
+    },
+    {
+      "epoch": 1.6328580686973428,
+      "grad_norm": 0.7979494333267212,
+      "learning_rate": 2.263336090981693e-06,
+      "loss": 0.0967,
+      "step": 10078
+    },
+    {
+      "epoch": 1.6330200907323396,
+      "grad_norm": 0.8447449207305908,
+      "learning_rate": 2.262900754379868e-06,
+      "loss": 0.099,
+      "step": 10079
+    },
+    {
+      "epoch": 1.6331821127673365,
+      "grad_norm": 0.9383016228675842,
+      "learning_rate": 2.2624654250327054e-06,
+      "loss": 0.1159,
+      "step": 10080
+    },
+    {
+      "epoch": 1.633344134802333,
+      "grad_norm": 0.8210467100143433,
+      "learning_rate": 2.2620301029535264e-06,
+      "loss": 0.1078,
+      "step": 10081
+    },
+    {
+      "epoch": 1.63350615683733,
+      "grad_norm": 0.8023772239685059,
+      "learning_rate": 2.2615947881556506e-06,
+      "loss": 0.101,
+      "step": 10082
+    },
+    {
+      "epoch": 1.6336681788723266,
+      "grad_norm": 0.8335037231445312,
+      "learning_rate": 2.2611594806523975e-06,
+      "loss": 0.103,
+      "step": 10083
+    },
+    {
+      "epoch": 1.6338302009073233,
+      "grad_norm": 0.7512911558151245,
+      "learning_rate": 2.2607241804570864e-06,
+      "loss": 0.0998,
+      "step": 10084
+    },
+    {
+      "epoch": 1.6339922229423203,
+      "grad_norm": 0.8822583556175232,
+      "learning_rate": 2.2602888875830346e-06,
+      "loss": 0.1034,
+      "step": 10085
+    },
+    {
+      "epoch": 1.6341542449773168,
+      "grad_norm": 0.830302357673645,
+      "learning_rate": 2.2598536020435644e-06,
+      "loss": 0.1027,
+      "step": 10086
+    },
+    {
+      "epoch": 1.6343162670123137,
+      "grad_norm": 0.7789014577865601,
+      "learning_rate": 2.2594183238519923e-06,
+      "loss": 0.0965,
+      "step": 10087
+    },
+    {
+      "epoch": 1.6344782890473104,
+      "grad_norm": 0.7837060689926147,
+      "learning_rate": 2.258983053021638e-06,
+      "loss": 0.1004,
+      "step": 10088
+    },
+    {
+      "epoch": 1.6346403110823071,
+      "grad_norm": 0.8886387348175049,
+      "learning_rate": 2.258547789565818e-06,
+      "loss": 0.1169,
+      "step": 10089
+    },
+    {
+      "epoch": 1.634802333117304,
+      "grad_norm": 0.9173817038536072,
+      "learning_rate": 2.2581125334978517e-06,
+      "loss": 0.1102,
+      "step": 10090
+    },
+    {
+      "epoch": 1.6349643551523008,
+      "grad_norm": 0.8326734304428101,
+      "learning_rate": 2.2576772848310572e-06,
+      "loss": 0.1016,
+      "step": 10091
+    },
+    {
+      "epoch": 1.6351263771872975,
+      "grad_norm": 0.7596706748008728,
+      "learning_rate": 2.257242043578751e-06,
+      "loss": 0.0957,
+      "step": 10092
+    },
+    {
+      "epoch": 1.6352883992222942,
+      "grad_norm": 0.7185407876968384,
+      "learning_rate": 2.256806809754251e-06,
+      "loss": 0.084,
+      "step": 10093
+    },
+    {
+      "epoch": 1.635450421257291,
+      "grad_norm": 0.8314585089683533,
+      "learning_rate": 2.2563715833708726e-06,
+      "loss": 0.1079,
+      "step": 10094
+    },
+    {
+      "epoch": 1.6356124432922878,
+      "grad_norm": 0.804526686668396,
+      "learning_rate": 2.2559363644419357e-06,
+      "loss": 0.1031,
+      "step": 10095
+    },
+    {
+      "epoch": 1.6357744653272845,
+      "grad_norm": 0.7893649935722351,
+      "learning_rate": 2.255501152980755e-06,
+      "loss": 0.0935,
+      "step": 10096
+    },
+    {
+      "epoch": 1.6359364873622813,
+      "grad_norm": 0.9371634721755981,
+      "learning_rate": 2.255065949000648e-06,
+      "loss": 0.1163,
+      "step": 10097
+    },
+    {
+      "epoch": 1.6360985093972782,
+      "grad_norm": 0.8292739987373352,
+      "learning_rate": 2.2546307525149293e-06,
+      "loss": 0.1,
+      "step": 10098
+    },
+    {
+      "epoch": 1.6362605314322747,
+      "grad_norm": 0.7563901543617249,
+      "learning_rate": 2.2541955635369156e-06,
+      "loss": 0.0997,
+      "step": 10099
+    },
+    {
+      "epoch": 1.6364225534672716,
+      "grad_norm": 0.8337042927742004,
+      "learning_rate": 2.253760382079924e-06,
+      "loss": 0.1052,
+      "step": 10100
+    },
+    {
+      "epoch": 1.6365845755022683,
+      "grad_norm": 0.7527819275856018,
+      "learning_rate": 2.253325208157268e-06,
+      "loss": 0.0978,
+      "step": 10101
+    },
+    {
+      "epoch": 1.636746597537265,
+      "grad_norm": 0.8317774534225464,
+      "learning_rate": 2.2528900417822636e-06,
+      "loss": 0.1005,
+      "step": 10102
+    },
+    {
+      "epoch": 1.636908619572262,
+      "grad_norm": 0.8075408935546875,
+      "learning_rate": 2.252454882968227e-06,
+      "loss": 0.1078,
+      "step": 10103
+    },
+    {
+      "epoch": 1.6370706416072585,
+      "grad_norm": 0.9008114337921143,
+      "learning_rate": 2.2520197317284702e-06,
+      "loss": 0.1172,
+      "step": 10104
+    },
+    {
+      "epoch": 1.6372326636422554,
+      "grad_norm": 0.8030856251716614,
+      "learning_rate": 2.2515845880763102e-06,
+      "loss": 0.099,
+      "step": 10105
+    },
+    {
+      "epoch": 1.637394685677252,
+      "grad_norm": 0.8751403093338013,
+      "learning_rate": 2.2511494520250613e-06,
+      "loss": 0.1087,
+      "step": 10106
+    },
+    {
+      "epoch": 1.6375567077122488,
+      "grad_norm": 0.8014907836914062,
+      "learning_rate": 2.2507143235880364e-06,
+      "loss": 0.0864,
+      "step": 10107
+    },
+    {
+      "epoch": 1.6377187297472457,
+      "grad_norm": 0.7244542241096497,
+      "learning_rate": 2.2502792027785508e-06,
+      "loss": 0.0942,
+      "step": 10108
+    },
+    {
+      "epoch": 1.6378807517822422,
+      "grad_norm": 0.9491717219352722,
+      "learning_rate": 2.249844089609916e-06,
+      "loss": 0.119,
+      "step": 10109
+    },
+    {
+      "epoch": 1.6380427738172392,
+      "grad_norm": 0.730972170829773,
+      "learning_rate": 2.249408984095447e-06,
+      "loss": 0.0864,
+      "step": 10110
+    },
+    {
+      "epoch": 1.6382047958522359,
+      "grad_norm": 0.7497561573982239,
+      "learning_rate": 2.248973886248457e-06,
+      "loss": 0.1,
+      "step": 10111
+    },
+    {
+      "epoch": 1.6383668178872326,
+      "grad_norm": 0.8914108276367188,
+      "learning_rate": 2.248538796082259e-06,
+      "loss": 0.1044,
+      "step": 10112
+    },
+    {
+      "epoch": 1.6385288399222295,
+      "grad_norm": 0.8104687929153442,
+      "learning_rate": 2.248103713610164e-06,
+      "loss": 0.104,
+      "step": 10113
+    },
+    {
+      "epoch": 1.638690861957226,
+      "grad_norm": 0.851868748664856,
+      "learning_rate": 2.2476686388454867e-06,
+      "loss": 0.1021,
+      "step": 10114
+    },
+    {
+      "epoch": 1.638852883992223,
+      "grad_norm": 0.8389064073562622,
+      "learning_rate": 2.247233571801539e-06,
+      "loss": 0.1069,
+      "step": 10115
+    },
+    {
+      "epoch": 1.6390149060272197,
+      "grad_norm": 0.7837998867034912,
+      "learning_rate": 2.2467985124916314e-06,
+      "loss": 0.0959,
+      "step": 10116
+    },
+    {
+      "epoch": 1.6391769280622164,
+      "grad_norm": 0.7479491233825684,
+      "learning_rate": 2.2463634609290776e-06,
+      "loss": 0.1026,
+      "step": 10117
+    },
+    {
+      "epoch": 1.6393389500972133,
+      "grad_norm": 0.7724059820175171,
+      "learning_rate": 2.2459284171271863e-06,
+      "loss": 0.1027,
+      "step": 10118
+    },
+    {
+      "epoch": 1.63950097213221,
+      "grad_norm": 0.7824938893318176,
+      "learning_rate": 2.245493381099272e-06,
+      "loss": 0.1015,
+      "step": 10119
+    },
+    {
+      "epoch": 1.6396629941672067,
+      "grad_norm": 0.8370165824890137,
+      "learning_rate": 2.2450583528586437e-06,
+      "loss": 0.112,
+      "step": 10120
+    },
+    {
+      "epoch": 1.6398250162022034,
+      "grad_norm": 0.8152546286582947,
+      "learning_rate": 2.244623332418614e-06,
+      "loss": 0.1056,
+      "step": 10121
+    },
+    {
+      "epoch": 1.6399870382372002,
+      "grad_norm": 0.7992122173309326,
+      "learning_rate": 2.244188319792491e-06,
+      "loss": 0.0924,
+      "step": 10122
+    },
+    {
+      "epoch": 1.640149060272197,
+      "grad_norm": 0.7578029036521912,
+      "learning_rate": 2.243753314993586e-06,
+      "loss": 0.0947,
+      "step": 10123
+    },
+    {
+      "epoch": 1.6403110823071938,
+      "grad_norm": 0.8097305297851562,
+      "learning_rate": 2.243318318035211e-06,
+      "loss": 0.1012,
+      "step": 10124
+    },
+    {
+      "epoch": 1.6404731043421905,
+      "grad_norm": 0.7099080681800842,
+      "learning_rate": 2.2428833289306735e-06,
+      "loss": 0.0909,
+      "step": 10125
+    },
+    {
+      "epoch": 1.6406351263771874,
+      "grad_norm": 0.8755943179130554,
+      "learning_rate": 2.2424483476932847e-06,
+      "loss": 0.1099,
+      "step": 10126
+    },
+    {
+      "epoch": 1.640797148412184,
+      "grad_norm": 0.7338682413101196,
+      "learning_rate": 2.2420133743363524e-06,
+      "loss": 0.0978,
+      "step": 10127
+    },
+    {
+      "epoch": 1.6409591704471809,
+      "grad_norm": 0.7446590065956116,
+      "learning_rate": 2.241578408873186e-06,
+      "loss": 0.0917,
+      "step": 10128
+    },
+    {
+      "epoch": 1.6411211924821776,
+      "grad_norm": 0.8787501454353333,
+      "learning_rate": 2.2411434513170955e-06,
+      "loss": 0.1037,
+      "step": 10129
+    },
+    {
+      "epoch": 1.6412832145171743,
+      "grad_norm": 0.73140549659729,
+      "learning_rate": 2.2407085016813895e-06,
+      "loss": 0.0915,
+      "step": 10130
+    },
+    {
+      "epoch": 1.6414452365521712,
+      "grad_norm": 0.6761218905448914,
+      "learning_rate": 2.2402735599793754e-06,
+      "loss": 0.0846,
+      "step": 10131
+    },
+    {
+      "epoch": 1.6416072585871677,
+      "grad_norm": 0.9457839131355286,
+      "learning_rate": 2.239838626224361e-06,
+      "loss": 0.1162,
+      "step": 10132
+    },
+    {
+      "epoch": 1.6417692806221647,
+      "grad_norm": 0.7481329441070557,
+      "learning_rate": 2.2394037004296566e-06,
+      "loss": 0.0894,
+      "step": 10133
+    },
+    {
+      "epoch": 1.6419313026571614,
+      "grad_norm": 0.8582826852798462,
+      "learning_rate": 2.2389687826085675e-06,
+      "loss": 0.1031,
+      "step": 10134
+    },
+    {
+      "epoch": 1.642093324692158,
+      "grad_norm": 0.9034931063652039,
+      "learning_rate": 2.2385338727744027e-06,
+      "loss": 0.1094,
+      "step": 10135
+    },
+    {
+      "epoch": 1.642255346727155,
+      "grad_norm": 0.8110890984535217,
+      "learning_rate": 2.238098970940468e-06,
+      "loss": 0.0943,
+      "step": 10136
+    },
+    {
+      "epoch": 1.6424173687621515,
+      "grad_norm": 0.7057623863220215,
+      "learning_rate": 2.23766407712007e-06,
+      "loss": 0.0871,
+      "step": 10137
+    },
+    {
+      "epoch": 1.6425793907971484,
+      "grad_norm": 0.8530978560447693,
+      "learning_rate": 2.2372291913265177e-06,
+      "loss": 0.0934,
+      "step": 10138
+    },
+    {
+      "epoch": 1.6427414128321451,
+      "grad_norm": 0.8763788342475891,
+      "learning_rate": 2.2367943135731164e-06,
+      "loss": 0.1104,
+      "step": 10139
+    },
+    {
+      "epoch": 1.6429034348671419,
+      "grad_norm": 0.8446993827819824,
+      "learning_rate": 2.236359443873172e-06,
+      "loss": 0.1086,
+      "step": 10140
+    },
+    {
+      "epoch": 1.6430654569021388,
+      "grad_norm": 0.9074305891990662,
+      "learning_rate": 2.2359245822399908e-06,
+      "loss": 0.1015,
+      "step": 10141
+    },
+    {
+      "epoch": 1.6432274789371355,
+      "grad_norm": 0.8572295904159546,
+      "learning_rate": 2.2354897286868773e-06,
+      "loss": 0.1,
+      "step": 10142
+    },
+    {
+      "epoch": 1.6433895009721322,
+      "grad_norm": 0.8138932585716248,
+      "learning_rate": 2.2350548832271386e-06,
+      "loss": 0.0973,
+      "step": 10143
+    },
+    {
+      "epoch": 1.643551523007129,
+      "grad_norm": 0.9844521880149841,
+      "learning_rate": 2.23462004587408e-06,
+      "loss": 0.1023,
+      "step": 10144
+    },
+    {
+      "epoch": 1.6437135450421256,
+      "grad_norm": 0.794780433177948,
+      "learning_rate": 2.2341852166410048e-06,
+      "loss": 0.0918,
+      "step": 10145
+    },
+    {
+      "epoch": 1.6438755670771226,
+      "grad_norm": 0.8207656145095825,
+      "learning_rate": 2.233750395541219e-06,
+      "loss": 0.0913,
+      "step": 10146
+    },
+    {
+      "epoch": 1.6440375891121193,
+      "grad_norm": 0.7891299724578857,
+      "learning_rate": 2.233315582588026e-06,
+      "loss": 0.0977,
+      "step": 10147
+    },
+    {
+      "epoch": 1.644199611147116,
+      "grad_norm": 0.7756161689758301,
+      "learning_rate": 2.2328807777947323e-06,
+      "loss": 0.094,
+      "step": 10148
+    },
+    {
+      "epoch": 1.644361633182113,
+      "grad_norm": 0.9181487560272217,
+      "learning_rate": 2.23244598117464e-06,
+      "loss": 0.1011,
+      "step": 10149
+    },
+    {
+      "epoch": 1.6445236552171094,
+      "grad_norm": 0.7459124326705933,
+      "learning_rate": 2.232011192741053e-06,
+      "loss": 0.0858,
+      "step": 10150
+    },
+    {
+      "epoch": 1.6446856772521063,
+      "grad_norm": 0.8103591203689575,
+      "learning_rate": 2.231576412507275e-06,
+      "loss": 0.1069,
+      "step": 10151
+    },
+    {
+      "epoch": 1.644847699287103,
+      "grad_norm": 0.9799508452415466,
+      "learning_rate": 2.2311416404866085e-06,
+      "loss": 0.1186,
+      "step": 10152
+    },
+    {
+      "epoch": 1.6450097213220998,
+      "grad_norm": 0.7545525431632996,
+      "learning_rate": 2.2307068766923584e-06,
+      "loss": 0.0958,
+      "step": 10153
+    },
+    {
+      "epoch": 1.6451717433570967,
+      "grad_norm": 0.8151463866233826,
+      "learning_rate": 2.2302721211378254e-06,
+      "loss": 0.0976,
+      "step": 10154
+    },
+    {
+      "epoch": 1.6453337653920932,
+      "grad_norm": 0.8132693767547607,
+      "learning_rate": 2.229837373836313e-06,
+      "loss": 0.1047,
+      "step": 10155
+    },
+    {
+      "epoch": 1.6454957874270901,
+      "grad_norm": 0.6447371244430542,
+      "learning_rate": 2.2294026348011223e-06,
+      "loss": 0.0838,
+      "step": 10156
+    },
+    {
+      "epoch": 1.6456578094620868,
+      "grad_norm": 0.8918012380599976,
+      "learning_rate": 2.228967904045558e-06,
+      "loss": 0.1196,
+      "step": 10157
+    },
+    {
+      "epoch": 1.6458198314970836,
+      "grad_norm": 0.7891850471496582,
+      "learning_rate": 2.2285331815829187e-06,
+      "loss": 0.1002,
+      "step": 10158
+    },
+    {
+      "epoch": 1.6459818535320805,
+      "grad_norm": 0.7595945000648499,
+      "learning_rate": 2.2280984674265077e-06,
+      "loss": 0.0927,
+      "step": 10159
+    },
+    {
+      "epoch": 1.646143875567077,
+      "grad_norm": 0.8276386857032776,
+      "learning_rate": 2.227663761589625e-06,
+      "loss": 0.1031,
+      "step": 10160
+    },
+    {
+      "epoch": 1.646305897602074,
+      "grad_norm": 0.8383728861808777,
+      "learning_rate": 2.227229064085572e-06,
+      "loss": 0.0979,
+      "step": 10161
+    },
+    {
+      "epoch": 1.6464679196370706,
+      "grad_norm": 0.6650486588478088,
+      "learning_rate": 2.2267943749276503e-06,
+      "loss": 0.0886,
+      "step": 10162
+    },
+    {
+      "epoch": 1.6466299416720673,
+      "grad_norm": 0.7667943835258484,
+      "learning_rate": 2.2263596941291595e-06,
+      "loss": 0.0959,
+      "step": 10163
+    },
+    {
+      "epoch": 1.6467919637070643,
+      "grad_norm": 0.9279943108558655,
+      "learning_rate": 2.225925021703399e-06,
+      "loss": 0.1164,
+      "step": 10164
+    },
+    {
+      "epoch": 1.6469539857420608,
+      "grad_norm": 0.608851969242096,
+      "learning_rate": 2.2254903576636713e-06,
+      "loss": 0.076,
+      "step": 10165
+    },
+    {
+      "epoch": 1.6471160077770577,
+      "grad_norm": 0.7714102268218994,
+      "learning_rate": 2.2250557020232724e-06,
+      "loss": 0.0894,
+      "step": 10166
+    },
+    {
+      "epoch": 1.6472780298120544,
+      "grad_norm": 0.9038023352622986,
+      "learning_rate": 2.2246210547955043e-06,
+      "loss": 0.1038,
+      "step": 10167
+    },
+    {
+      "epoch": 1.6474400518470511,
+      "grad_norm": 0.8809518814086914,
+      "learning_rate": 2.2241864159936664e-06,
+      "loss": 0.1081,
+      "step": 10168
+    },
+    {
+      "epoch": 1.647602073882048,
+      "grad_norm": 0.7901051640510559,
+      "learning_rate": 2.2237517856310558e-06,
+      "loss": 0.0955,
+      "step": 10169
+    },
+    {
+      "epoch": 1.6477640959170448,
+      "grad_norm": 0.8911927938461304,
+      "learning_rate": 2.223317163720973e-06,
+      "loss": 0.1086,
+      "step": 10170
+    },
+    {
+      "epoch": 1.6479261179520415,
+      "grad_norm": 0.7963345050811768,
+      "learning_rate": 2.2228825502767133e-06,
+      "loss": 0.0976,
+      "step": 10171
+    },
+    {
+      "epoch": 1.6480881399870384,
+      "grad_norm": 0.902242124080658,
+      "learning_rate": 2.222447945311579e-06,
+      "loss": 0.1149,
+      "step": 10172
+    },
+    {
+      "epoch": 1.648250162022035,
+      "grad_norm": 0.8996853232383728,
+      "learning_rate": 2.2220133488388652e-06,
+      "loss": 0.118,
+      "step": 10173
+    },
+    {
+      "epoch": 1.6484121840570318,
+      "grad_norm": 0.9371250867843628,
+      "learning_rate": 2.2215787608718706e-06,
+      "loss": 0.1137,
+      "step": 10174
+    },
+    {
+      "epoch": 1.6485742060920285,
+      "grad_norm": 0.9435984492301941,
+      "learning_rate": 2.221144181423892e-06,
+      "loss": 0.1131,
+      "step": 10175
+    },
+    {
+      "epoch": 1.6487362281270252,
+      "grad_norm": 0.7899986505508423,
+      "learning_rate": 2.220709610508226e-06,
+      "loss": 0.095,
+      "step": 10176
+    },
+    {
+      "epoch": 1.6488982501620222,
+      "grad_norm": 0.7958756685256958,
+      "learning_rate": 2.220275048138171e-06,
+      "loss": 0.096,
+      "step": 10177
+    },
+    {
+      "epoch": 1.6490602721970187,
+      "grad_norm": 0.7749899625778198,
+      "learning_rate": 2.2198404943270217e-06,
+      "loss": 0.0984,
+      "step": 10178
+    },
+    {
+      "epoch": 1.6492222942320156,
+      "grad_norm": 0.7689141631126404,
+      "learning_rate": 2.2194059490880764e-06,
+      "loss": 0.0908,
+      "step": 10179
+    },
+    {
+      "epoch": 1.6493843162670123,
+      "grad_norm": 0.869452714920044,
+      "learning_rate": 2.218971412434628e-06,
+      "loss": 0.1048,
+      "step": 10180
+    },
+    {
+      "epoch": 1.649546338302009,
+      "grad_norm": 0.840836226940155,
+      "learning_rate": 2.2185368843799764e-06,
+      "loss": 0.1033,
+      "step": 10181
+    },
+    {
+      "epoch": 1.649708360337006,
+      "grad_norm": 0.8797898888587952,
+      "learning_rate": 2.218102364937414e-06,
+      "loss": 0.1093,
+      "step": 10182
+    },
+    {
+      "epoch": 1.6498703823720025,
+      "grad_norm": 0.7758781909942627,
+      "learning_rate": 2.217667854120238e-06,
+      "loss": 0.094,
+      "step": 10183
+    },
+    {
+      "epoch": 1.6500324044069994,
+      "grad_norm": 0.830178439617157,
+      "learning_rate": 2.2172333519417415e-06,
+      "loss": 0.1035,
+      "step": 10184
+    },
+    {
+      "epoch": 1.650194426441996,
+      "grad_norm": 0.8420971632003784,
+      "learning_rate": 2.2167988584152198e-06,
+      "loss": 0.1108,
+      "step": 10185
+    },
+    {
+      "epoch": 1.6503564484769928,
+      "grad_norm": 0.8024832606315613,
+      "learning_rate": 2.2163643735539688e-06,
+      "loss": 0.0986,
+      "step": 10186
+    },
+    {
+      "epoch": 1.6505184705119897,
+      "grad_norm": 0.8306187391281128,
+      "learning_rate": 2.215929897371281e-06,
+      "loss": 0.1036,
+      "step": 10187
+    },
+    {
+      "epoch": 1.6506804925469862,
+      "grad_norm": 0.7863216400146484,
+      "learning_rate": 2.2154954298804514e-06,
+      "loss": 0.1019,
+      "step": 10188
+    },
+    {
+      "epoch": 1.6508425145819832,
+      "grad_norm": 0.7025604248046875,
+      "learning_rate": 2.215060971094773e-06,
+      "loss": 0.0838,
+      "step": 10189
+    },
+    {
+      "epoch": 1.6510045366169799,
+      "grad_norm": 0.9769557118415833,
+      "learning_rate": 2.214626521027538e-06,
+      "loss": 0.0856,
+      "step": 10190
+    },
+    {
+      "epoch": 1.6511665586519766,
+      "grad_norm": 0.7868211269378662,
+      "learning_rate": 2.214192079692042e-06,
+      "loss": 0.0988,
+      "step": 10191
+    },
+    {
+      "epoch": 1.6513285806869735,
+      "grad_norm": 0.8536260724067688,
+      "learning_rate": 2.213757647101577e-06,
+      "loss": 0.1069,
+      "step": 10192
+    },
+    {
+      "epoch": 1.6514906027219702,
+      "grad_norm": 0.8652046322822571,
+      "learning_rate": 2.2133232232694354e-06,
+      "loss": 0.1003,
+      "step": 10193
+    },
+    {
+      "epoch": 1.651652624756967,
+      "grad_norm": 0.7852055430412292,
+      "learning_rate": 2.2128888082089093e-06,
+      "loss": 0.0992,
+      "step": 10194
+    },
+    {
+      "epoch": 1.6518146467919637,
+      "grad_norm": 0.7507728338241577,
+      "learning_rate": 2.2124544019332898e-06,
+      "loss": 0.0894,
+      "step": 10195
+    },
+    {
+      "epoch": 1.6519766688269604,
+      "grad_norm": 0.8173365592956543,
+      "learning_rate": 2.2120200044558705e-06,
+      "loss": 0.1081,
+      "step": 10196
+    },
+    {
+      "epoch": 1.6521386908619573,
+      "grad_norm": 0.9333679676055908,
+      "learning_rate": 2.211585615789943e-06,
+      "loss": 0.1155,
+      "step": 10197
+    },
+    {
+      "epoch": 1.652300712896954,
+      "grad_norm": 0.861423671245575,
+      "learning_rate": 2.2111512359487967e-06,
+      "loss": 0.0958,
+      "step": 10198
+    },
+    {
+      "epoch": 1.6524627349319507,
+      "grad_norm": 0.7481124997138977,
+      "learning_rate": 2.2107168649457233e-06,
+      "loss": 0.097,
+      "step": 10199
+    },
+    {
+      "epoch": 1.6526247569669477,
+      "grad_norm": 0.8404136896133423,
+      "learning_rate": 2.2102825027940143e-06,
+      "loss": 0.0984,
+      "step": 10200
+    },
+    {
+      "epoch": 1.6527867790019442,
+      "grad_norm": 0.7924706339836121,
+      "learning_rate": 2.20984814950696e-06,
+      "loss": 0.103,
+      "step": 10201
+    },
+    {
+      "epoch": 1.652948801036941,
+      "grad_norm": 0.8639355301856995,
+      "learning_rate": 2.2094138050978496e-06,
+      "loss": 0.1058,
+      "step": 10202
+    },
+    {
+      "epoch": 1.6531108230719378,
+      "grad_norm": 0.7134700417518616,
+      "learning_rate": 2.2089794695799744e-06,
+      "loss": 0.093,
+      "step": 10203
+    },
+    {
+      "epoch": 1.6532728451069345,
+      "grad_norm": 0.7675890922546387,
+      "learning_rate": 2.2085451429666215e-06,
+      "loss": 0.0912,
+      "step": 10204
+    },
+    {
+      "epoch": 1.6534348671419314,
+      "grad_norm": 0.9659287929534912,
+      "learning_rate": 2.208110825271083e-06,
+      "loss": 0.1052,
+      "step": 10205
+    },
+    {
+      "epoch": 1.653596889176928,
+      "grad_norm": 0.821732759475708,
+      "learning_rate": 2.207676516506647e-06,
+      "loss": 0.0996,
+      "step": 10206
+    },
+    {
+      "epoch": 1.6537589112119249,
+      "grad_norm": 0.8499715924263,
+      "learning_rate": 2.2072422166866024e-06,
+      "loss": 0.1036,
+      "step": 10207
+    },
+    {
+      "epoch": 1.6539209332469216,
+      "grad_norm": 0.8570707440376282,
+      "learning_rate": 2.206807925824237e-06,
+      "loss": 0.0909,
+      "step": 10208
+    },
+    {
+      "epoch": 1.6540829552819183,
+      "grad_norm": 0.8913640975952148,
+      "learning_rate": 2.206373643932839e-06,
+      "loss": 0.1169,
+      "step": 10209
+    },
+    {
+      "epoch": 1.6542449773169152,
+      "grad_norm": 0.7439228296279907,
+      "learning_rate": 2.205939371025698e-06,
+      "loss": 0.0964,
+      "step": 10210
+    },
+    {
+      "epoch": 1.6544069993519117,
+      "grad_norm": 0.8434868454933167,
+      "learning_rate": 2.2055051071161e-06,
+      "loss": 0.1139,
+      "step": 10211
+    },
+    {
+      "epoch": 1.6545690213869086,
+      "grad_norm": 0.8420630097389221,
+      "learning_rate": 2.205070852217334e-06,
+      "loss": 0.1057,
+      "step": 10212
+    },
+    {
+      "epoch": 1.6547310434219054,
+      "grad_norm": 1.035088062286377,
+      "learning_rate": 2.204636606342685e-06,
+      "loss": 0.112,
+      "step": 10213
+    },
+    {
+      "epoch": 1.654893065456902,
+      "grad_norm": 0.7095579504966736,
+      "learning_rate": 2.204202369505441e-06,
+      "loss": 0.0954,
+      "step": 10214
+    },
+    {
+      "epoch": 1.655055087491899,
+      "grad_norm": 0.8334465622901917,
+      "learning_rate": 2.2037681417188895e-06,
+      "loss": 0.1064,
+      "step": 10215
+    },
+    {
+      "epoch": 1.6552171095268955,
+      "grad_norm": 0.7914650440216064,
+      "learning_rate": 2.203333922996316e-06,
+      "loss": 0.1005,
+      "step": 10216
+    },
+    {
+      "epoch": 1.6553791315618924,
+      "grad_norm": 0.6868200898170471,
+      "learning_rate": 2.2028997133510065e-06,
+      "loss": 0.0868,
+      "step": 10217
+    },
+    {
+      "epoch": 1.6555411535968891,
+      "grad_norm": 0.7120256423950195,
+      "learning_rate": 2.202465512796247e-06,
+      "loss": 0.0839,
+      "step": 10218
+    },
+    {
+      "epoch": 1.6557031756318858,
+      "grad_norm": 0.8592171669006348,
+      "learning_rate": 2.2020313213453216e-06,
+      "loss": 0.1126,
+      "step": 10219
+    },
+    {
+      "epoch": 1.6558651976668828,
+      "grad_norm": 0.8605828285217285,
+      "learning_rate": 2.2015971390115172e-06,
+      "loss": 0.0982,
+      "step": 10220
+    },
+    {
+      "epoch": 1.6560272197018795,
+      "grad_norm": 0.75331050157547,
+      "learning_rate": 2.2011629658081194e-06,
+      "loss": 0.0965,
+      "step": 10221
+    },
+    {
+      "epoch": 1.6561892417368762,
+      "grad_norm": 0.8181818127632141,
+      "learning_rate": 2.2007288017484105e-06,
+      "loss": 0.0948,
+      "step": 10222
+    },
+    {
+      "epoch": 1.6563512637718731,
+      "grad_norm": 0.8317630290985107,
+      "learning_rate": 2.2002946468456758e-06,
+      "loss": 0.0963,
+      "step": 10223
+    },
+    {
+      "epoch": 1.6565132858068696,
+      "grad_norm": 0.8039714694023132,
+      "learning_rate": 2.1998605011131997e-06,
+      "loss": 0.0992,
+      "step": 10224
+    },
+    {
+      "epoch": 1.6566753078418666,
+      "grad_norm": 0.8288991451263428,
+      "learning_rate": 2.199426364564267e-06,
+      "loss": 0.1062,
+      "step": 10225
+    },
+    {
+      "epoch": 1.6568373298768633,
+      "grad_norm": 0.781379222869873,
+      "learning_rate": 2.19899223721216e-06,
+      "loss": 0.0961,
+      "step": 10226
+    },
+    {
+      "epoch": 1.65699935191186,
+      "grad_norm": 0.8361613750457764,
+      "learning_rate": 2.1985581190701617e-06,
+      "loss": 0.1011,
+      "step": 10227
+    },
+    {
+      "epoch": 1.657161373946857,
+      "grad_norm": 0.8540022373199463,
+      "learning_rate": 2.1981240101515548e-06,
+      "loss": 0.1012,
+      "step": 10228
+    },
+    {
+      "epoch": 1.6573233959818534,
+      "grad_norm": 0.7892126441001892,
+      "learning_rate": 2.197689910469623e-06,
+      "loss": 0.087,
+      "step": 10229
+    },
+    {
+      "epoch": 1.6574854180168503,
+      "grad_norm": 0.9751836061477661,
+      "learning_rate": 2.1972558200376497e-06,
+      "loss": 0.1122,
+      "step": 10230
+    },
+    {
+      "epoch": 1.657647440051847,
+      "grad_norm": 0.9051089882850647,
+      "learning_rate": 2.1968217388689145e-06,
+      "loss": 0.1043,
+      "step": 10231
+    },
+    {
+      "epoch": 1.6578094620868438,
+      "grad_norm": 0.783577024936676,
+      "learning_rate": 2.1963876669767008e-06,
+      "loss": 0.094,
+      "step": 10232
+    },
+    {
+      "epoch": 1.6579714841218407,
+      "grad_norm": 0.835684597492218,
+      "learning_rate": 2.1959536043742887e-06,
+      "loss": 0.0975,
+      "step": 10233
+    },
+    {
+      "epoch": 1.6581335061568372,
+      "grad_norm": 0.8699052929878235,
+      "learning_rate": 2.1955195510749614e-06,
+      "loss": 0.1112,
+      "step": 10234
+    },
+    {
+      "epoch": 1.6582955281918341,
+      "grad_norm": 0.8392834067344666,
+      "learning_rate": 2.1950855070919992e-06,
+      "loss": 0.1118,
+      "step": 10235
+    },
+    {
+      "epoch": 1.6584575502268308,
+      "grad_norm": 0.7662588357925415,
+      "learning_rate": 2.1946514724386827e-06,
+      "loss": 0.1046,
+      "step": 10236
+    },
+    {
+      "epoch": 1.6586195722618275,
+      "grad_norm": 0.8604794144630432,
+      "learning_rate": 2.194217447128292e-06,
+      "loss": 0.1106,
+      "step": 10237
+    },
+    {
+      "epoch": 1.6587815942968245,
+      "grad_norm": 0.7735698223114014,
+      "learning_rate": 2.1937834311741066e-06,
+      "loss": 0.091,
+      "step": 10238
+    },
+    {
+      "epoch": 1.658943616331821,
+      "grad_norm": 0.8237504959106445,
+      "learning_rate": 2.1933494245894087e-06,
+      "loss": 0.0916,
+      "step": 10239
+    },
+    {
+      "epoch": 1.659105638366818,
+      "grad_norm": 0.7927703857421875,
+      "learning_rate": 2.192915427387475e-06,
+      "loss": 0.0967,
+      "step": 10240
+    },
+    {
+      "epoch": 1.6592676604018146,
+      "grad_norm": 0.7646491527557373,
+      "learning_rate": 2.1924814395815875e-06,
+      "loss": 0.0996,
+      "step": 10241
+    },
+    {
+      "epoch": 1.6594296824368113,
+      "grad_norm": 0.7301465272903442,
+      "learning_rate": 2.1920474611850225e-06,
+      "loss": 0.0896,
+      "step": 10242
+    },
+    {
+      "epoch": 1.6595917044718083,
+      "grad_norm": 0.7497079372406006,
+      "learning_rate": 2.19161349221106e-06,
+      "loss": 0.0904,
+      "step": 10243
+    },
+    {
+      "epoch": 1.659753726506805,
+      "grad_norm": 0.8112010955810547,
+      "learning_rate": 2.1911795326729784e-06,
+      "loss": 0.0977,
+      "step": 10244
+    },
+    {
+      "epoch": 1.6599157485418017,
+      "grad_norm": 0.7948684692382812,
+      "learning_rate": 2.1907455825840568e-06,
+      "loss": 0.0968,
+      "step": 10245
+    },
+    {
+      "epoch": 1.6600777705767984,
+      "grad_norm": 0.7585175633430481,
+      "learning_rate": 2.190311641957571e-06,
+      "loss": 0.0896,
+      "step": 10246
+    },
+    {
+      "epoch": 1.660239792611795,
+      "grad_norm": 0.7399813532829285,
+      "learning_rate": 2.189877710806799e-06,
+      "loss": 0.0879,
+      "step": 10247
+    },
+    {
+      "epoch": 1.660401814646792,
+      "grad_norm": 0.7585265636444092,
+      "learning_rate": 2.18944378914502e-06,
+      "loss": 0.0918,
+      "step": 10248
+    },
+    {
+      "epoch": 1.6605638366817888,
+      "grad_norm": 0.9892318844795227,
+      "learning_rate": 2.189009876985509e-06,
+      "loss": 0.1184,
+      "step": 10249
+    },
+    {
+      "epoch": 1.6607258587167855,
+      "grad_norm": 0.8371285200119019,
+      "learning_rate": 2.188575974341543e-06,
+      "loss": 0.103,
+      "step": 10250
+    },
+    {
+      "epoch": 1.6608878807517824,
+      "grad_norm": 0.7434375882148743,
+      "learning_rate": 2.188142081226399e-06,
+      "loss": 0.0905,
+      "step": 10251
+    },
+    {
+      "epoch": 1.6610499027867789,
+      "grad_norm": 0.7424872517585754,
+      "learning_rate": 2.1877081976533515e-06,
+      "loss": 0.0832,
+      "step": 10252
+    },
+    {
+      "epoch": 1.6612119248217758,
+      "grad_norm": 0.9748948812484741,
+      "learning_rate": 2.1872743236356783e-06,
+      "loss": 0.1224,
+      "step": 10253
+    },
+    {
+      "epoch": 1.6613739468567725,
+      "grad_norm": 0.7499638199806213,
+      "learning_rate": 2.186840459186654e-06,
+      "loss": 0.0857,
+      "step": 10254
+    },
+    {
+      "epoch": 1.6615359688917692,
+      "grad_norm": 0.8377159237861633,
+      "learning_rate": 2.186406604319554e-06,
+      "loss": 0.0981,
+      "step": 10255
+    },
+    {
+      "epoch": 1.6616979909267662,
+      "grad_norm": 0.8437566161155701,
+      "learning_rate": 2.185972759047653e-06,
+      "loss": 0.0917,
+      "step": 10256
+    },
+    {
+      "epoch": 1.6618600129617627,
+      "grad_norm": 0.9447170495986938,
+      "learning_rate": 2.185538923384225e-06,
+      "loss": 0.1206,
+      "step": 10257
+    },
+    {
+      "epoch": 1.6620220349967596,
+      "grad_norm": 0.9525909423828125,
+      "learning_rate": 2.1851050973425454e-06,
+      "loss": 0.1178,
+      "step": 10258
+    },
+    {
+      "epoch": 1.6621840570317563,
+      "grad_norm": 0.8417654633522034,
+      "learning_rate": 2.1846712809358876e-06,
+      "loss": 0.0978,
+      "step": 10259
+    },
+    {
+      "epoch": 1.662346079066753,
+      "grad_norm": 0.8003092408180237,
+      "learning_rate": 2.1842374741775262e-06,
+      "loss": 0.1017,
+      "step": 10260
+    },
+    {
+      "epoch": 1.66250810110175,
+      "grad_norm": 0.7186848521232605,
+      "learning_rate": 2.183803677080733e-06,
+      "loss": 0.0841,
+      "step": 10261
+    },
+    {
+      "epoch": 1.6626701231367464,
+      "grad_norm": 0.8107516765594482,
+      "learning_rate": 2.1833698896587816e-06,
+      "loss": 0.0997,
+      "step": 10262
+    },
+    {
+      "epoch": 1.6628321451717434,
+      "grad_norm": 0.852218508720398,
+      "learning_rate": 2.182936111924947e-06,
+      "loss": 0.1055,
+      "step": 10263
+    },
+    {
+      "epoch": 1.66299416720674,
+      "grad_norm": 0.7575379610061646,
+      "learning_rate": 2.1825023438924995e-06,
+      "loss": 0.095,
+      "step": 10264
+    },
+    {
+      "epoch": 1.6631561892417368,
+      "grad_norm": 0.7673975825309753,
+      "learning_rate": 2.182068585574712e-06,
+      "loss": 0.1017,
+      "step": 10265
+    },
+    {
+      "epoch": 1.6633182112767337,
+      "grad_norm": 0.8737732172012329,
+      "learning_rate": 2.1816348369848555e-06,
+      "loss": 0.1138,
+      "step": 10266
+    },
+    {
+      "epoch": 1.6634802333117304,
+      "grad_norm": 0.7601016163825989,
+      "learning_rate": 2.1812010981362033e-06,
+      "loss": 0.0916,
+      "step": 10267
+    },
+    {
+      "epoch": 1.6636422553467272,
+      "grad_norm": 0.7540645003318787,
+      "learning_rate": 2.180767369042026e-06,
+      "loss": 0.1005,
+      "step": 10268
+    },
+    {
+      "epoch": 1.6638042773817239,
+      "grad_norm": 0.7578513622283936,
+      "learning_rate": 2.180333649715595e-06,
+      "loss": 0.0875,
+      "step": 10269
+    },
+    {
+      "epoch": 1.6639662994167206,
+      "grad_norm": 0.682778000831604,
+      "learning_rate": 2.1798999401701802e-06,
+      "loss": 0.0829,
+      "step": 10270
+    },
+    {
+      "epoch": 1.6641283214517175,
+      "grad_norm": 0.7566827535629272,
+      "learning_rate": 2.1794662404190526e-06,
+      "loss": 0.0932,
+      "step": 10271
+    },
+    {
+      "epoch": 1.6642903434867142,
+      "grad_norm": 0.7969744205474854,
+      "learning_rate": 2.1790325504754827e-06,
+      "loss": 0.099,
+      "step": 10272
+    },
+    {
+      "epoch": 1.664452365521711,
+      "grad_norm": 0.7715543508529663,
+      "learning_rate": 2.17859887035274e-06,
+      "loss": 0.0954,
+      "step": 10273
+    },
+    {
+      "epoch": 1.6646143875567079,
+      "grad_norm": 0.9869740009307861,
+      "learning_rate": 2.1781652000640947e-06,
+      "loss": 0.1277,
+      "step": 10274
+    },
+    {
+      "epoch": 1.6647764095917044,
+      "grad_norm": 0.7891120910644531,
+      "learning_rate": 2.1777315396228145e-06,
+      "loss": 0.1008,
+      "step": 10275
+    },
+    {
+      "epoch": 1.6649384316267013,
+      "grad_norm": 0.7361397743225098,
+      "learning_rate": 2.177297889042169e-06,
+      "loss": 0.0878,
+      "step": 10276
+    },
+    {
+      "epoch": 1.665100453661698,
+      "grad_norm": 0.8581258654594421,
+      "learning_rate": 2.1768642483354274e-06,
+      "loss": 0.103,
+      "step": 10277
+    },
+    {
+      "epoch": 1.6652624756966947,
+      "grad_norm": 0.8369101285934448,
+      "learning_rate": 2.1764306175158588e-06,
+      "loss": 0.0994,
+      "step": 10278
+    },
+    {
+      "epoch": 1.6654244977316917,
+      "grad_norm": 1.0208182334899902,
+      "learning_rate": 2.1759969965967293e-06,
+      "loss": 0.1056,
+      "step": 10279
+    },
+    {
+      "epoch": 1.6655865197666881,
+      "grad_norm": 0.717728316783905,
+      "learning_rate": 2.1755633855913086e-06,
+      "loss": 0.0869,
+      "step": 10280
+    },
+    {
+      "epoch": 1.665748541801685,
+      "grad_norm": 0.7904380559921265,
+      "learning_rate": 2.175129784512862e-06,
+      "loss": 0.0956,
+      "step": 10281
+    },
+    {
+      "epoch": 1.6659105638366818,
+      "grad_norm": 0.8746015429496765,
+      "learning_rate": 2.174696193374658e-06,
+      "loss": 0.1046,
+      "step": 10282
+    },
+    {
+      "epoch": 1.6660725858716785,
+      "grad_norm": 0.7690415978431702,
+      "learning_rate": 2.1742626121899645e-06,
+      "loss": 0.0932,
+      "step": 10283
+    },
+    {
+      "epoch": 1.6662346079066754,
+      "grad_norm": 0.8206772208213806,
+      "learning_rate": 2.173829040972046e-06,
+      "loss": 0.107,
+      "step": 10284
+    },
+    {
+      "epoch": 1.666396629941672,
+      "grad_norm": 0.830139696598053,
+      "learning_rate": 2.1733954797341692e-06,
+      "loss": 0.1108,
+      "step": 10285
+    },
+    {
+      "epoch": 1.6665586519766689,
+      "grad_norm": 0.7665162086486816,
+      "learning_rate": 2.1729619284896e-06,
+      "loss": 0.096,
+      "step": 10286
+    },
+    {
+      "epoch": 1.6667206740116656,
+      "grad_norm": 0.9206477403640747,
+      "learning_rate": 2.1725283872516053e-06,
+      "loss": 0.1022,
+      "step": 10287
+    },
+    {
+      "epoch": 1.6668826960466623,
+      "grad_norm": 0.7268409729003906,
+      "learning_rate": 2.1720948560334492e-06,
+      "loss": 0.0966,
+      "step": 10288
+    },
+    {
+      "epoch": 1.6670447180816592,
+      "grad_norm": 0.7334338426589966,
+      "learning_rate": 2.171661334848397e-06,
+      "loss": 0.0948,
+      "step": 10289
+    },
+    {
+      "epoch": 1.6672067401166557,
+      "grad_norm": 0.661564826965332,
+      "learning_rate": 2.171227823709713e-06,
+      "loss": 0.0766,
+      "step": 10290
+    },
+    {
+      "epoch": 1.6673687621516526,
+      "grad_norm": 1.0224709510803223,
+      "learning_rate": 2.1707943226306626e-06,
+      "loss": 0.1268,
+      "step": 10291
+    },
+    {
+      "epoch": 1.6675307841866494,
+      "grad_norm": 0.9280935525894165,
+      "learning_rate": 2.1703608316245092e-06,
+      "loss": 0.1148,
+      "step": 10292
+    },
+    {
+      "epoch": 1.667692806221646,
+      "grad_norm": 0.8443484306335449,
+      "learning_rate": 2.1699273507045163e-06,
+      "loss": 0.1064,
+      "step": 10293
+    },
+    {
+      "epoch": 1.667854828256643,
+      "grad_norm": 0.8637576103210449,
+      "learning_rate": 2.169493879883948e-06,
+      "loss": 0.1034,
+      "step": 10294
+    },
+    {
+      "epoch": 1.6680168502916397,
+      "grad_norm": 0.8016158938407898,
+      "learning_rate": 2.169060419176066e-06,
+      "loss": 0.0982,
+      "step": 10295
+    },
+    {
+      "epoch": 1.6681788723266364,
+      "grad_norm": 0.7897706627845764,
+      "learning_rate": 2.168626968594136e-06,
+      "loss": 0.0963,
+      "step": 10296
+    },
+    {
+      "epoch": 1.6683408943616331,
+      "grad_norm": 0.7999807596206665,
+      "learning_rate": 2.1681935281514182e-06,
+      "loss": 0.0962,
+      "step": 10297
+    },
+    {
+      "epoch": 1.6685029163966298,
+      "grad_norm": 0.8183321356773376,
+      "learning_rate": 2.167760097861176e-06,
+      "loss": 0.1015,
+      "step": 10298
+    },
+    {
+      "epoch": 1.6686649384316268,
+      "grad_norm": 0.7776053547859192,
+      "learning_rate": 2.16732667773667e-06,
+      "loss": 0.0998,
+      "step": 10299
+    },
+    {
+      "epoch": 1.6688269604666235,
+      "grad_norm": 0.983116626739502,
+      "learning_rate": 2.1668932677911624e-06,
+      "loss": 0.1021,
+      "step": 10300
+    },
+    {
+      "epoch": 1.6689889825016202,
+      "grad_norm": 0.8453471064567566,
+      "learning_rate": 2.1664598680379158e-06,
+      "loss": 0.0946,
+      "step": 10301
+    },
+    {
+      "epoch": 1.6691510045366171,
+      "grad_norm": 0.7629480957984924,
+      "learning_rate": 2.166026478490189e-06,
+      "loss": 0.097,
+      "step": 10302
+    },
+    {
+      "epoch": 1.6693130265716136,
+      "grad_norm": 0.7451595664024353,
+      "learning_rate": 2.1655930991612443e-06,
+      "loss": 0.0916,
+      "step": 10303
+    },
+    {
+      "epoch": 1.6694750486066106,
+      "grad_norm": 0.7532944083213806,
+      "learning_rate": 2.1651597300643418e-06,
+      "loss": 0.0916,
+      "step": 10304
+    },
+    {
+      "epoch": 1.6696370706416073,
+      "grad_norm": 0.7477149963378906,
+      "learning_rate": 2.1647263712127402e-06,
+      "loss": 0.0871,
+      "step": 10305
+    },
+    {
+      "epoch": 1.669799092676604,
+      "grad_norm": 0.7762525081634521,
+      "learning_rate": 2.1642930226197012e-06,
+      "loss": 0.0979,
+      "step": 10306
+    },
+    {
+      "epoch": 1.669961114711601,
+      "grad_norm": 0.8461878299713135,
+      "learning_rate": 2.1638596842984834e-06,
+      "loss": 0.0979,
+      "step": 10307
+    },
+    {
+      "epoch": 1.6701231367465974,
+      "grad_norm": 0.9171335697174072,
+      "learning_rate": 2.1634263562623454e-06,
+      "loss": 0.105,
+      "step": 10308
+    },
+    {
+      "epoch": 1.6702851587815943,
+      "grad_norm": 0.7819655537605286,
+      "learning_rate": 2.162993038524547e-06,
+      "loss": 0.0919,
+      "step": 10309
+    },
+    {
+      "epoch": 1.670447180816591,
+      "grad_norm": 0.8039749264717102,
+      "learning_rate": 2.162559731098345e-06,
+      "loss": 0.1012,
+      "step": 10310
+    },
+    {
+      "epoch": 1.6706092028515878,
+      "grad_norm": 0.7070164084434509,
+      "learning_rate": 2.162126433996999e-06,
+      "loss": 0.085,
+      "step": 10311
+    },
+    {
+      "epoch": 1.6707712248865847,
+      "grad_norm": 0.6756847500801086,
+      "learning_rate": 2.161693147233767e-06,
+      "loss": 0.0854,
+      "step": 10312
+    },
+    {
+      "epoch": 1.6709332469215812,
+      "grad_norm": 0.8924721479415894,
+      "learning_rate": 2.161259870821906e-06,
+      "loss": 0.1165,
+      "step": 10313
+    },
+    {
+      "epoch": 1.6710952689565781,
+      "grad_norm": 0.8302946090698242,
+      "learning_rate": 2.1608266047746723e-06,
+      "loss": 0.1093,
+      "step": 10314
+    },
+    {
+      "epoch": 1.6712572909915748,
+      "grad_norm": 0.9837422370910645,
+      "learning_rate": 2.1603933491053243e-06,
+      "loss": 0.1093,
+      "step": 10315
+    },
+    {
+      "epoch": 1.6714193130265715,
+      "grad_norm": 0.8880236744880676,
+      "learning_rate": 2.1599601038271186e-06,
+      "loss": 0.1147,
+      "step": 10316
+    },
+    {
+      "epoch": 1.6715813350615685,
+      "grad_norm": 0.7705572247505188,
+      "learning_rate": 2.1595268689533105e-06,
+      "loss": 0.0923,
+      "step": 10317
+    },
+    {
+      "epoch": 1.6717433570965652,
+      "grad_norm": 0.834516167640686,
+      "learning_rate": 2.1590936444971563e-06,
+      "loss": 0.0956,
+      "step": 10318
+    },
+    {
+      "epoch": 1.671905379131562,
+      "grad_norm": 0.8553721904754639,
+      "learning_rate": 2.15866043047191e-06,
+      "loss": 0.1118,
+      "step": 10319
+    },
+    {
+      "epoch": 1.6720674011665586,
+      "grad_norm": 0.899256706237793,
+      "learning_rate": 2.1582272268908307e-06,
+      "loss": 0.1104,
+      "step": 10320
+    },
+    {
+      "epoch": 1.6722294232015553,
+      "grad_norm": 0.8428982496261597,
+      "learning_rate": 2.1577940337671698e-06,
+      "loss": 0.108,
+      "step": 10321
+    },
+    {
+      "epoch": 1.6723914452365523,
+      "grad_norm": 0.8363228440284729,
+      "learning_rate": 2.1573608511141845e-06,
+      "loss": 0.105,
+      "step": 10322
+    },
+    {
+      "epoch": 1.672553467271549,
+      "grad_norm": 0.7880693674087524,
+      "learning_rate": 2.1569276789451273e-06,
+      "loss": 0.1044,
+      "step": 10323
+    },
+    {
+      "epoch": 1.6727154893065457,
+      "grad_norm": 0.7894632816314697,
+      "learning_rate": 2.1564945172732523e-06,
+      "loss": 0.0986,
+      "step": 10324
+    },
+    {
+      "epoch": 1.6728775113415426,
+      "grad_norm": 0.8343285918235779,
+      "learning_rate": 2.1560613661118154e-06,
+      "loss": 0.1084,
+      "step": 10325
+    },
+    {
+      "epoch": 1.673039533376539,
+      "grad_norm": 0.8286653757095337,
+      "learning_rate": 2.155628225474067e-06,
+      "loss": 0.1083,
+      "step": 10326
+    },
+    {
+      "epoch": 1.673201555411536,
+      "grad_norm": 0.7715649008750916,
+      "learning_rate": 2.1551950953732627e-06,
+      "loss": 0.1026,
+      "step": 10327
+    },
+    {
+      "epoch": 1.6733635774465327,
+      "grad_norm": 0.7973687648773193,
+      "learning_rate": 2.154761975822653e-06,
+      "loss": 0.1054,
+      "step": 10328
+    },
+    {
+      "epoch": 1.6735255994815295,
+      "grad_norm": 0.7777444124221802,
+      "learning_rate": 2.1543288668354914e-06,
+      "loss": 0.0949,
+      "step": 10329
+    },
+    {
+      "epoch": 1.6736876215165264,
+      "grad_norm": 0.8096343278884888,
+      "learning_rate": 2.1538957684250303e-06,
+      "loss": 0.1013,
+      "step": 10330
+    },
+    {
+      "epoch": 1.6738496435515229,
+      "grad_norm": 0.751850962638855,
+      "learning_rate": 2.153462680604522e-06,
+      "loss": 0.0921,
+      "step": 10331
+    },
+    {
+      "epoch": 1.6740116655865198,
+      "grad_norm": 0.8717265725135803,
+      "learning_rate": 2.1530296033872155e-06,
+      "loss": 0.1096,
+      "step": 10332
+    },
+    {
+      "epoch": 1.6741736876215165,
+      "grad_norm": 0.8423851728439331,
+      "learning_rate": 2.152596536786364e-06,
+      "loss": 0.1089,
+      "step": 10333
+    },
+    {
+      "epoch": 1.6743357096565132,
+      "grad_norm": 0.7620172500610352,
+      "learning_rate": 2.152163480815218e-06,
+      "loss": 0.095,
+      "step": 10334
+    },
+    {
+      "epoch": 1.6744977316915102,
+      "grad_norm": 0.7917490005493164,
+      "learning_rate": 2.151730435487028e-06,
+      "loss": 0.0961,
+      "step": 10335
+    },
+    {
+      "epoch": 1.6746597537265067,
+      "grad_norm": 0.8136554956436157,
+      "learning_rate": 2.151297400815044e-06,
+      "loss": 0.1113,
+      "step": 10336
+    },
+    {
+      "epoch": 1.6748217757615036,
+      "grad_norm": 0.8516972064971924,
+      "learning_rate": 2.150864376812515e-06,
+      "loss": 0.1168,
+      "step": 10337
+    },
+    {
+      "epoch": 1.6749837977965003,
+      "grad_norm": 0.7402398586273193,
+      "learning_rate": 2.150431363492691e-06,
+      "loss": 0.0879,
+      "step": 10338
+    },
+    {
+      "epoch": 1.675145819831497,
+      "grad_norm": 0.7780933976173401,
+      "learning_rate": 2.1499983608688217e-06,
+      "loss": 0.0975,
+      "step": 10339
+    },
+    {
+      "epoch": 1.675307841866494,
+      "grad_norm": 0.8871686458587646,
+      "learning_rate": 2.1495653689541562e-06,
+      "loss": 0.1125,
+      "step": 10340
+    },
+    {
+      "epoch": 1.6754698639014904,
+      "grad_norm": 0.854103147983551,
+      "learning_rate": 2.149132387761942e-06,
+      "loss": 0.113,
+      "step": 10341
+    },
+    {
+      "epoch": 1.6756318859364874,
+      "grad_norm": 0.8291221857070923,
+      "learning_rate": 2.1486994173054276e-06,
+      "loss": 0.1044,
+      "step": 10342
+    },
+    {
+      "epoch": 1.675793907971484,
+      "grad_norm": 0.838173508644104,
+      "learning_rate": 2.14826645759786e-06,
+      "loss": 0.1034,
+      "step": 10343
+    },
+    {
+      "epoch": 1.6759559300064808,
+      "grad_norm": 0.7588802576065063,
+      "learning_rate": 2.1478335086524885e-06,
+      "loss": 0.1,
+      "step": 10344
+    },
+    {
+      "epoch": 1.6761179520414777,
+      "grad_norm": 0.7420892119407654,
+      "learning_rate": 2.14740057048256e-06,
+      "loss": 0.0877,
+      "step": 10345
+    },
+    {
+      "epoch": 1.6762799740764744,
+      "grad_norm": 0.823940634727478,
+      "learning_rate": 2.14696764310132e-06,
+      "loss": 0.1046,
+      "step": 10346
+    },
+    {
+      "epoch": 1.6764419961114712,
+      "grad_norm": 0.8017604351043701,
+      "learning_rate": 2.146534726522016e-06,
+      "loss": 0.1034,
+      "step": 10347
+    },
+    {
+      "epoch": 1.6766040181464679,
+      "grad_norm": 0.8648914694786072,
+      "learning_rate": 2.1461018207578932e-06,
+      "loss": 0.1073,
+      "step": 10348
+    },
+    {
+      "epoch": 1.6767660401814646,
+      "grad_norm": 0.7010998129844666,
+      "learning_rate": 2.145668925822199e-06,
+      "loss": 0.0923,
+      "step": 10349
+    },
+    {
+      "epoch": 1.6769280622164615,
+      "grad_norm": 0.7532062530517578,
+      "learning_rate": 2.1452360417281786e-06,
+      "loss": 0.0892,
+      "step": 10350
+    },
+    {
+      "epoch": 1.6770900842514582,
+      "grad_norm": 0.7723690867424011,
+      "learning_rate": 2.1448031684890767e-06,
+      "loss": 0.0971,
+      "step": 10351
+    },
+    {
+      "epoch": 1.677252106286455,
+      "grad_norm": 0.7597529292106628,
+      "learning_rate": 2.144370306118138e-06,
+      "loss": 0.095,
+      "step": 10352
+    },
+    {
+      "epoch": 1.6774141283214519,
+      "grad_norm": 0.9111921787261963,
+      "learning_rate": 2.1439374546286065e-06,
+      "loss": 0.0962,
+      "step": 10353
+    },
+    {
+      "epoch": 1.6775761503564484,
+      "grad_norm": 0.7498681545257568,
+      "learning_rate": 2.143504614033728e-06,
+      "loss": 0.0917,
+      "step": 10354
+    },
+    {
+      "epoch": 1.6777381723914453,
+      "grad_norm": 0.7228400707244873,
+      "learning_rate": 2.143071784346746e-06,
+      "loss": 0.0927,
+      "step": 10355
+    },
+    {
+      "epoch": 1.677900194426442,
+      "grad_norm": 0.6969931721687317,
+      "learning_rate": 2.142638965580903e-06,
+      "loss": 0.0866,
+      "step": 10356
+    },
+    {
+      "epoch": 1.6780622164614387,
+      "grad_norm": 0.8031118512153625,
+      "learning_rate": 2.1422061577494427e-06,
+      "loss": 0.1015,
+      "step": 10357
+    },
+    {
+      "epoch": 1.6782242384964356,
+      "grad_norm": 0.8556954264640808,
+      "learning_rate": 2.141773360865609e-06,
+      "loss": 0.1002,
+      "step": 10358
+    },
+    {
+      "epoch": 1.6783862605314321,
+      "grad_norm": 0.8014951348304749,
+      "learning_rate": 2.1413405749426432e-06,
+      "loss": 0.0944,
+      "step": 10359
+    },
+    {
+      "epoch": 1.678548282566429,
+      "grad_norm": 1.1695481538772583,
+      "learning_rate": 2.1409077999937883e-06,
+      "loss": 0.1157,
+      "step": 10360
+    },
+    {
+      "epoch": 1.6787103046014258,
+      "grad_norm": 0.8790541887283325,
+      "learning_rate": 2.1404750360322852e-06,
+      "loss": 0.1036,
+      "step": 10361
+    },
+    {
+      "epoch": 1.6788723266364225,
+      "grad_norm": 0.8709704279899597,
+      "learning_rate": 2.1400422830713752e-06,
+      "loss": 0.1126,
+      "step": 10362
+    },
+    {
+      "epoch": 1.6790343486714194,
+      "grad_norm": 0.8556258678436279,
+      "learning_rate": 2.139609541124301e-06,
+      "loss": 0.1063,
+      "step": 10363
+    },
+    {
+      "epoch": 1.679196370706416,
+      "grad_norm": 0.779319703578949,
+      "learning_rate": 2.1391768102043032e-06,
+      "loss": 0.0961,
+      "step": 10364
+    },
+    {
+      "epoch": 1.6793583927414129,
+      "grad_norm": 0.8869314789772034,
+      "learning_rate": 2.138744090324621e-06,
+      "loss": 0.1104,
+      "step": 10365
+    },
+    {
+      "epoch": 1.6795204147764096,
+      "grad_norm": 0.9177558422088623,
+      "learning_rate": 2.1383113814984967e-06,
+      "loss": 0.1191,
+      "step": 10366
+    },
+    {
+      "epoch": 1.6796824368114063,
+      "grad_norm": 0.7949991822242737,
+      "learning_rate": 2.1378786837391673e-06,
+      "loss": 0.1051,
+      "step": 10367
+    },
+    {
+      "epoch": 1.6798444588464032,
+      "grad_norm": 0.8216454982757568,
+      "learning_rate": 2.137445997059874e-06,
+      "loss": 0.0968,
+      "step": 10368
+    },
+    {
+      "epoch": 1.6800064808814,
+      "grad_norm": 0.7865923643112183,
+      "learning_rate": 2.1370133214738573e-06,
+      "loss": 0.1002,
+      "step": 10369
+    },
+    {
+      "epoch": 1.6801685029163966,
+      "grad_norm": 0.7961848378181458,
+      "learning_rate": 2.1365806569943533e-06,
+      "loss": 0.0945,
+      "step": 10370
+    },
+    {
+      "epoch": 1.6803305249513933,
+      "grad_norm": 0.843292236328125,
+      "learning_rate": 2.1361480036346025e-06,
+      "loss": 0.1049,
+      "step": 10371
+    },
+    {
+      "epoch": 1.68049254698639,
+      "grad_norm": 0.9252605438232422,
+      "learning_rate": 2.1357153614078407e-06,
+      "loss": 0.114,
+      "step": 10372
+    },
+    {
+      "epoch": 1.680654569021387,
+      "grad_norm": 0.7975597977638245,
+      "learning_rate": 2.135282730327309e-06,
+      "loss": 0.0955,
+      "step": 10373
+    },
+    {
+      "epoch": 1.6808165910563837,
+      "grad_norm": 0.9786580204963684,
+      "learning_rate": 2.1348501104062423e-06,
+      "loss": 0.1155,
+      "step": 10374
+    },
+    {
+      "epoch": 1.6809786130913804,
+      "grad_norm": 0.9285010695457458,
+      "learning_rate": 2.1344175016578796e-06,
+      "loss": 0.1207,
+      "step": 10375
+    },
+    {
+      "epoch": 1.6811406351263773,
+      "grad_norm": 0.7795856595039368,
+      "learning_rate": 2.1339849040954556e-06,
+      "loss": 0.0978,
+      "step": 10376
+    },
+    {
+      "epoch": 1.6813026571613738,
+      "grad_norm": 1.0095703601837158,
+      "learning_rate": 2.133552317732208e-06,
+      "loss": 0.1229,
+      "step": 10377
+    },
+    {
+      "epoch": 1.6814646791963708,
+      "grad_norm": 0.7144725918769836,
+      "learning_rate": 2.133119742581373e-06,
+      "loss": 0.0896,
+      "step": 10378
+    },
+    {
+      "epoch": 1.6816267012313675,
+      "grad_norm": 0.7718963623046875,
+      "learning_rate": 2.1326871786561856e-06,
+      "loss": 0.0949,
+      "step": 10379
+    },
+    {
+      "epoch": 1.6817887232663642,
+      "grad_norm": 0.8956825733184814,
+      "learning_rate": 2.1322546259698823e-06,
+      "loss": 0.1074,
+      "step": 10380
+    },
+    {
+      "epoch": 1.6819507453013611,
+      "grad_norm": 0.8126304745674133,
+      "learning_rate": 2.131822084535696e-06,
+      "loss": 0.1,
+      "step": 10381
+    },
+    {
+      "epoch": 1.6821127673363576,
+      "grad_norm": 0.8905256986618042,
+      "learning_rate": 2.1313895543668644e-06,
+      "loss": 0.1035,
+      "step": 10382
+    },
+    {
+      "epoch": 1.6822747893713546,
+      "grad_norm": 0.7790430188179016,
+      "learning_rate": 2.13095703547662e-06,
+      "loss": 0.0894,
+      "step": 10383
+    },
+    {
+      "epoch": 1.6824368114063513,
+      "grad_norm": 0.7982741594314575,
+      "learning_rate": 2.1305245278781977e-06,
+      "loss": 0.1081,
+      "step": 10384
+    },
+    {
+      "epoch": 1.682598833441348,
+      "grad_norm": 0.790199875831604,
+      "learning_rate": 2.1300920315848307e-06,
+      "loss": 0.1041,
+      "step": 10385
+    },
+    {
+      "epoch": 1.682760855476345,
+      "grad_norm": 0.8125954866409302,
+      "learning_rate": 2.129659546609751e-06,
+      "loss": 0.0963,
+      "step": 10386
+    },
+    {
+      "epoch": 1.6829228775113414,
+      "grad_norm": 0.9038329124450684,
+      "learning_rate": 2.1292270729661946e-06,
+      "loss": 0.1123,
+      "step": 10387
+    },
+    {
+      "epoch": 1.6830848995463383,
+      "grad_norm": 0.8060694336891174,
+      "learning_rate": 2.1287946106673916e-06,
+      "loss": 0.0908,
+      "step": 10388
+    },
+    {
+      "epoch": 1.683246921581335,
+      "grad_norm": 0.7719234824180603,
+      "learning_rate": 2.128362159726576e-06,
+      "loss": 0.097,
+      "step": 10389
+    },
+    {
+      "epoch": 1.6834089436163318,
+      "grad_norm": 1.3537698984146118,
+      "learning_rate": 2.1279297201569787e-06,
+      "loss": 0.0971,
+      "step": 10390
+    },
+    {
+      "epoch": 1.6835709656513287,
+      "grad_norm": 0.8270101547241211,
+      "learning_rate": 2.1274972919718305e-06,
+      "loss": 0.1124,
+      "step": 10391
+    },
+    {
+      "epoch": 1.6837329876863252,
+      "grad_norm": 0.8727843761444092,
+      "learning_rate": 2.127064875184365e-06,
+      "loss": 0.1085,
+      "step": 10392
+    },
+    {
+      "epoch": 1.6838950097213221,
+      "grad_norm": 0.8102154731750488,
+      "learning_rate": 2.1266324698078116e-06,
+      "loss": 0.0979,
+      "step": 10393
+    },
+    {
+      "epoch": 1.6840570317563188,
+      "grad_norm": 0.7432061433792114,
+      "learning_rate": 2.126200075855401e-06,
+      "loss": 0.0951,
+      "step": 10394
+    },
+    {
+      "epoch": 1.6842190537913155,
+      "grad_norm": 0.7901977300643921,
+      "learning_rate": 2.1257676933403637e-06,
+      "loss": 0.1051,
+      "step": 10395
+    },
+    {
+      "epoch": 1.6843810758263125,
+      "grad_norm": 0.84058678150177,
+      "learning_rate": 2.125335322275928e-06,
+      "loss": 0.11,
+      "step": 10396
+    },
+    {
+      "epoch": 1.6845430978613092,
+      "grad_norm": 0.6626092195510864,
+      "learning_rate": 2.124902962675326e-06,
+      "loss": 0.0884,
+      "step": 10397
+    },
+    {
+      "epoch": 1.684705119896306,
+      "grad_norm": 0.7434554100036621,
+      "learning_rate": 2.1244706145517853e-06,
+      "loss": 0.0998,
+      "step": 10398
+    },
+    {
+      "epoch": 1.6848671419313026,
+      "grad_norm": 0.7729745507240295,
+      "learning_rate": 2.124038277918536e-06,
+      "loss": 0.0976,
+      "step": 10399
+    },
+    {
+      "epoch": 1.6850291639662993,
+      "grad_norm": 0.6964062452316284,
+      "learning_rate": 2.1236059527888044e-06,
+      "loss": 0.0883,
+      "step": 10400
+    },
+    {
+      "epoch": 1.6851911860012962,
+      "grad_norm": 0.8118695616722107,
+      "learning_rate": 2.1231736391758195e-06,
+      "loss": 0.1021,
+      "step": 10401
+    },
+    {
+      "epoch": 1.685353208036293,
+      "grad_norm": 0.7241361737251282,
+      "learning_rate": 2.1227413370928106e-06,
+      "loss": 0.0955,
+      "step": 10402
+    },
+    {
+      "epoch": 1.6855152300712897,
+      "grad_norm": 0.8646603226661682,
+      "learning_rate": 2.1223090465530032e-06,
+      "loss": 0.0978,
+      "step": 10403
+    },
+    {
+      "epoch": 1.6856772521062866,
+      "grad_norm": 0.8380978107452393,
+      "learning_rate": 2.1218767675696255e-06,
+      "loss": 0.1027,
+      "step": 10404
+    },
+    {
+      "epoch": 1.685839274141283,
+      "grad_norm": 0.8121195435523987,
+      "learning_rate": 2.1214445001559025e-06,
+      "loss": 0.0994,
+      "step": 10405
+    },
+    {
+      "epoch": 1.68600129617628,
+      "grad_norm": 0.7910208106040955,
+      "learning_rate": 2.1210122443250625e-06,
+      "loss": 0.0954,
+      "step": 10406
+    },
+    {
+      "epoch": 1.6861633182112767,
+      "grad_norm": 0.8031769394874573,
+      "learning_rate": 2.1205800000903305e-06,
+      "loss": 0.1046,
+      "step": 10407
+    },
+    {
+      "epoch": 1.6863253402462735,
+      "grad_norm": 0.7870137095451355,
+      "learning_rate": 2.1201477674649326e-06,
+      "loss": 0.1009,
+      "step": 10408
+    },
+    {
+      "epoch": 1.6864873622812704,
+      "grad_norm": 0.8784118294715881,
+      "learning_rate": 2.1197155464620934e-06,
+      "loss": 0.1071,
+      "step": 10409
+    },
+    {
+      "epoch": 1.6866493843162669,
+      "grad_norm": 0.7897698283195496,
+      "learning_rate": 2.119283337095038e-06,
+      "loss": 0.0943,
+      "step": 10410
+    },
+    {
+      "epoch": 1.6868114063512638,
+      "grad_norm": 0.770966112613678,
+      "learning_rate": 2.118851139376992e-06,
+      "loss": 0.0914,
+      "step": 10411
+    },
+    {
+      "epoch": 1.6869734283862605,
+      "grad_norm": 0.8083310127258301,
+      "learning_rate": 2.1184189533211783e-06,
+      "loss": 0.099,
+      "step": 10412
+    },
+    {
+      "epoch": 1.6871354504212572,
+      "grad_norm": 0.8235817551612854,
+      "learning_rate": 2.117986778940822e-06,
+      "loss": 0.1128,
+      "step": 10413
+    },
+    {
+      "epoch": 1.6872974724562542,
+      "grad_norm": 0.7883199453353882,
+      "learning_rate": 2.117554616249145e-06,
+      "loss": 0.0981,
+      "step": 10414
+    },
+    {
+      "epoch": 1.6874594944912507,
+      "grad_norm": 0.8687751293182373,
+      "learning_rate": 2.11712246525937e-06,
+      "loss": 0.1017,
+      "step": 10415
+    },
+    {
+      "epoch": 1.6876215165262476,
+      "grad_norm": 0.863337516784668,
+      "learning_rate": 2.1166903259847228e-06,
+      "loss": 0.111,
+      "step": 10416
+    },
+    {
+      "epoch": 1.6877835385612443,
+      "grad_norm": 0.7061415314674377,
+      "learning_rate": 2.116258198438424e-06,
+      "loss": 0.0925,
+      "step": 10417
+    },
+    {
+      "epoch": 1.687945560596241,
+      "grad_norm": 0.7574256658554077,
+      "learning_rate": 2.115826082633695e-06,
+      "loss": 0.0906,
+      "step": 10418
+    },
+    {
+      "epoch": 1.688107582631238,
+      "grad_norm": 0.7754940390586853,
+      "learning_rate": 2.115393978583759e-06,
+      "loss": 0.1083,
+      "step": 10419
+    },
+    {
+      "epoch": 1.6882696046662347,
+      "grad_norm": 0.799916684627533,
+      "learning_rate": 2.114961886301835e-06,
+      "loss": 0.0945,
+      "step": 10420
+    },
+    {
+      "epoch": 1.6884316267012314,
+      "grad_norm": 0.9458333849906921,
+      "learning_rate": 2.114529805801147e-06,
+      "loss": 0.1102,
+      "step": 10421
+    },
+    {
+      "epoch": 1.688593648736228,
+      "grad_norm": 0.7674143314361572,
+      "learning_rate": 2.114097737094914e-06,
+      "loss": 0.0925,
+      "step": 10422
+    },
+    {
+      "epoch": 1.6887556707712248,
+      "grad_norm": 1.0405738353729248,
+      "learning_rate": 2.1136656801963556e-06,
+      "loss": 0.1039,
+      "step": 10423
+    },
+    {
+      "epoch": 1.6889176928062217,
+      "grad_norm": 0.9508488178253174,
+      "learning_rate": 2.1132336351186923e-06,
+      "loss": 0.1144,
+      "step": 10424
+    },
+    {
+      "epoch": 1.6890797148412184,
+      "grad_norm": 0.8733676075935364,
+      "learning_rate": 2.1128016018751444e-06,
+      "loss": 0.1035,
+      "step": 10425
+    },
+    {
+      "epoch": 1.6892417368762151,
+      "grad_norm": 0.8814464807510376,
+      "learning_rate": 2.1123695804789307e-06,
+      "loss": 0.1017,
+      "step": 10426
+    },
+    {
+      "epoch": 1.689403758911212,
+      "grad_norm": 0.7157515287399292,
+      "learning_rate": 2.1119375709432696e-06,
+      "loss": 0.0853,
+      "step": 10427
+    },
+    {
+      "epoch": 1.6895657809462086,
+      "grad_norm": 0.8367986083030701,
+      "learning_rate": 2.11150557328138e-06,
+      "loss": 0.1023,
+      "step": 10428
+    },
+    {
+      "epoch": 1.6897278029812055,
+      "grad_norm": 0.8281162977218628,
+      "learning_rate": 2.1110735875064787e-06,
+      "loss": 0.1074,
+      "step": 10429
+    },
+    {
+      "epoch": 1.6898898250162022,
+      "grad_norm": 0.8819671273231506,
+      "learning_rate": 2.110641613631785e-06,
+      "loss": 0.1114,
+      "step": 10430
+    },
+    {
+      "epoch": 1.690051847051199,
+      "grad_norm": 0.796600341796875,
+      "learning_rate": 2.1102096516705165e-06,
+      "loss": 0.099,
+      "step": 10431
+    },
+    {
+      "epoch": 1.6902138690861959,
+      "grad_norm": 0.8326752185821533,
+      "learning_rate": 2.109777701635889e-06,
+      "loss": 0.1044,
+      "step": 10432
+    },
+    {
+      "epoch": 1.6903758911211924,
+      "grad_norm": 0.9205048084259033,
+      "learning_rate": 2.109345763541119e-06,
+      "loss": 0.1107,
+      "step": 10433
+    },
+    {
+      "epoch": 1.6905379131561893,
+      "grad_norm": 0.799943745136261,
+      "learning_rate": 2.1089138373994226e-06,
+      "loss": 0.1003,
+      "step": 10434
+    },
+    {
+      "epoch": 1.690699935191186,
+      "grad_norm": 0.8054278492927551,
+      "learning_rate": 2.1084819232240177e-06,
+      "loss": 0.0972,
+      "step": 10435
+    },
+    {
+      "epoch": 1.6908619572261827,
+      "grad_norm": 0.7532472014427185,
+      "learning_rate": 2.108050021028118e-06,
+      "loss": 0.0929,
+      "step": 10436
+    },
+    {
+      "epoch": 1.6910239792611796,
+      "grad_norm": 0.7412585020065308,
+      "learning_rate": 2.1076181308249396e-06,
+      "loss": 0.0968,
+      "step": 10437
+    },
+    {
+      "epoch": 1.6911860012961761,
+      "grad_norm": 0.8518087267875671,
+      "learning_rate": 2.1071862526276963e-06,
+      "loss": 0.0933,
+      "step": 10438
+    },
+    {
+      "epoch": 1.691348023331173,
+      "grad_norm": 0.7559210658073425,
+      "learning_rate": 2.1067543864496028e-06,
+      "loss": 0.0815,
+      "step": 10439
+    },
+    {
+      "epoch": 1.6915100453661698,
+      "grad_norm": 0.7951977849006653,
+      "learning_rate": 2.1063225323038744e-06,
+      "loss": 0.0965,
+      "step": 10440
+    },
+    {
+      "epoch": 1.6916720674011665,
+      "grad_norm": 0.7670897841453552,
+      "learning_rate": 2.1058906902037228e-06,
+      "loss": 0.0946,
+      "step": 10441
+    },
+    {
+      "epoch": 1.6918340894361634,
+      "grad_norm": 0.7800466418266296,
+      "learning_rate": 2.1054588601623634e-06,
+      "loss": 0.0971,
+      "step": 10442
+    },
+    {
+      "epoch": 1.69199611147116,
+      "grad_norm": 0.872276246547699,
+      "learning_rate": 2.1050270421930077e-06,
+      "loss": 0.0954,
+      "step": 10443
+    },
+    {
+      "epoch": 1.6921581335061568,
+      "grad_norm": 0.8790923357009888,
+      "learning_rate": 2.104595236308868e-06,
+      "loss": 0.1148,
+      "step": 10444
+    },
+    {
+      "epoch": 1.6923201555411536,
+      "grad_norm": 0.8432591557502747,
+      "learning_rate": 2.104163442523158e-06,
+      "loss": 0.1034,
+      "step": 10445
+    },
+    {
+      "epoch": 1.6924821775761503,
+      "grad_norm": 0.7778016328811646,
+      "learning_rate": 2.1037316608490886e-06,
+      "loss": 0.0907,
+      "step": 10446
+    },
+    {
+      "epoch": 1.6926441996111472,
+      "grad_norm": 0.8325784206390381,
+      "learning_rate": 2.1032998912998712e-06,
+      "loss": 0.1069,
+      "step": 10447
+    },
+    {
+      "epoch": 1.692806221646144,
+      "grad_norm": 0.789277195930481,
+      "learning_rate": 2.1028681338887164e-06,
+      "loss": 0.1051,
+      "step": 10448
+    },
+    {
+      "epoch": 1.6929682436811406,
+      "grad_norm": 0.8284058570861816,
+      "learning_rate": 2.1024363886288375e-06,
+      "loss": 0.11,
+      "step": 10449
+    },
+    {
+      "epoch": 1.6931302657161373,
+      "grad_norm": 0.8092173337936401,
+      "learning_rate": 2.102004655533442e-06,
+      "loss": 0.1038,
+      "step": 10450
+    },
+    {
+      "epoch": 1.693292287751134,
+      "grad_norm": 0.7652307152748108,
+      "learning_rate": 2.1015729346157406e-06,
+      "loss": 0.097,
+      "step": 10451
+    },
+    {
+      "epoch": 1.693454309786131,
+      "grad_norm": 0.6917365789413452,
+      "learning_rate": 2.101141225888944e-06,
+      "loss": 0.082,
+      "step": 10452
+    },
+    {
+      "epoch": 1.6936163318211277,
+      "grad_norm": 0.770603597164154,
+      "learning_rate": 2.10070952936626e-06,
+      "loss": 0.0931,
+      "step": 10453
+    },
+    {
+      "epoch": 1.6937783538561244,
+      "grad_norm": 0.8083109855651855,
+      "learning_rate": 2.100277845060898e-06,
+      "loss": 0.1012,
+      "step": 10454
+    },
+    {
+      "epoch": 1.6939403758911213,
+      "grad_norm": 0.7694876194000244,
+      "learning_rate": 2.0998461729860675e-06,
+      "loss": 0.0965,
+      "step": 10455
+    },
+    {
+      "epoch": 1.6941023979261178,
+      "grad_norm": 0.7325409054756165,
+      "learning_rate": 2.0994145131549755e-06,
+      "loss": 0.0911,
+      "step": 10456
+    },
+    {
+      "epoch": 1.6942644199611148,
+      "grad_norm": 0.8317882418632507,
+      "learning_rate": 2.09898286558083e-06,
+      "loss": 0.099,
+      "step": 10457
+    },
+    {
+      "epoch": 1.6944264419961115,
+      "grad_norm": 0.8942394852638245,
+      "learning_rate": 2.0985512302768366e-06,
+      "loss": 0.1108,
+      "step": 10458
+    },
+    {
+      "epoch": 1.6945884640311082,
+      "grad_norm": 0.850330114364624,
+      "learning_rate": 2.0981196072562067e-06,
+      "loss": 0.1008,
+      "step": 10459
+    },
+    {
+      "epoch": 1.6947504860661051,
+      "grad_norm": 0.7139987945556641,
+      "learning_rate": 2.097687996532143e-06,
+      "loss": 0.077,
+      "step": 10460
+    },
+    {
+      "epoch": 1.6949125081011016,
+      "grad_norm": 0.7986920475959778,
+      "learning_rate": 2.097256398117854e-06,
+      "loss": 0.0954,
+      "step": 10461
+    },
+    {
+      "epoch": 1.6950745301360985,
+      "grad_norm": 0.7126197218894958,
+      "learning_rate": 2.0968248120265433e-06,
+      "loss": 0.083,
+      "step": 10462
+    },
+    {
+      "epoch": 1.6952365521710953,
+      "grad_norm": 0.8080899715423584,
+      "learning_rate": 2.0963932382714175e-06,
+      "loss": 0.1012,
+      "step": 10463
+    },
+    {
+      "epoch": 1.695398574206092,
+      "grad_norm": 0.9293273687362671,
+      "learning_rate": 2.095961676865683e-06,
+      "loss": 0.1071,
+      "step": 10464
+    },
+    {
+      "epoch": 1.695560596241089,
+      "grad_norm": 0.7605911493301392,
+      "learning_rate": 2.0955301278225433e-06,
+      "loss": 0.1003,
+      "step": 10465
+    },
+    {
+      "epoch": 1.6957226182760854,
+      "grad_norm": 0.7792356014251709,
+      "learning_rate": 2.095098591155203e-06,
+      "loss": 0.0969,
+      "step": 10466
+    },
+    {
+      "epoch": 1.6958846403110823,
+      "grad_norm": 0.7824127674102783,
+      "learning_rate": 2.0946670668768652e-06,
+      "loss": 0.0951,
+      "step": 10467
+    },
+    {
+      "epoch": 1.696046662346079,
+      "grad_norm": 0.8019390106201172,
+      "learning_rate": 2.094235555000734e-06,
+      "loss": 0.0993,
+      "step": 10468
+    },
+    {
+      "epoch": 1.6962086843810757,
+      "grad_norm": 0.8202789425849915,
+      "learning_rate": 2.0938040555400137e-06,
+      "loss": 0.0955,
+      "step": 10469
+    },
+    {
+      "epoch": 1.6963707064160727,
+      "grad_norm": 0.8047801852226257,
+      "learning_rate": 2.093372568507907e-06,
+      "loss": 0.0991,
+      "step": 10470
+    },
+    {
+      "epoch": 1.6965327284510694,
+      "grad_norm": 0.7065772414207458,
+      "learning_rate": 2.0929410939176147e-06,
+      "loss": 0.0858,
+      "step": 10471
+    },
+    {
+      "epoch": 1.696694750486066,
+      "grad_norm": 0.8477860689163208,
+      "learning_rate": 2.0925096317823393e-06,
+      "loss": 0.097,
+      "step": 10472
+    },
+    {
+      "epoch": 1.6968567725210628,
+      "grad_norm": 0.7459856867790222,
+      "learning_rate": 2.0920781821152843e-06,
+      "loss": 0.0914,
+      "step": 10473
+    },
+    {
+      "epoch": 1.6970187945560595,
+      "grad_norm": 0.8197339773178101,
+      "learning_rate": 2.091646744929649e-06,
+      "loss": 0.099,
+      "step": 10474
+    },
+    {
+      "epoch": 1.6971808165910565,
+      "grad_norm": 0.7804821133613586,
+      "learning_rate": 2.091215320238636e-06,
+      "loss": 0.101,
+      "step": 10475
+    },
+    {
+      "epoch": 1.6973428386260532,
+      "grad_norm": 0.7543721795082092,
+      "learning_rate": 2.0907839080554443e-06,
+      "loss": 0.0866,
+      "step": 10476
+    },
+    {
+      "epoch": 1.6975048606610499,
+      "grad_norm": 0.8172874450683594,
+      "learning_rate": 2.090352508393274e-06,
+      "loss": 0.1058,
+      "step": 10477
+    },
+    {
+      "epoch": 1.6976668826960468,
+      "grad_norm": 0.8281826972961426,
+      "learning_rate": 2.0899211212653262e-06,
+      "loss": 0.0992,
+      "step": 10478
+    },
+    {
+      "epoch": 1.6978289047310433,
+      "grad_norm": 0.784435510635376,
+      "learning_rate": 2.0894897466848007e-06,
+      "loss": 0.098,
+      "step": 10479
+    },
+    {
+      "epoch": 1.6979909267660402,
+      "grad_norm": 0.871392548084259,
+      "learning_rate": 2.0890583846648945e-06,
+      "loss": 0.1029,
+      "step": 10480
+    },
+    {
+      "epoch": 1.698152948801037,
+      "grad_norm": 0.7556376457214355,
+      "learning_rate": 2.0886270352188082e-06,
+      "loss": 0.0959,
+      "step": 10481
+    },
+    {
+      "epoch": 1.6983149708360337,
+      "grad_norm": 0.7945913672447205,
+      "learning_rate": 2.0881956983597375e-06,
+      "loss": 0.0944,
+      "step": 10482
+    },
+    {
+      "epoch": 1.6984769928710306,
+      "grad_norm": 0.8585154414176941,
+      "learning_rate": 2.0877643741008828e-06,
+      "loss": 0.106,
+      "step": 10483
+    },
+    {
+      "epoch": 1.698639014906027,
+      "grad_norm": 0.8008867502212524,
+      "learning_rate": 2.087333062455441e-06,
+      "loss": 0.0965,
+      "step": 10484
+    },
+    {
+      "epoch": 1.698801036941024,
+      "grad_norm": 0.8277882933616638,
+      "learning_rate": 2.0869017634366087e-06,
+      "loss": 0.1087,
+      "step": 10485
+    },
+    {
+      "epoch": 1.6989630589760207,
+      "grad_norm": 0.781322717666626,
+      "learning_rate": 2.0864704770575824e-06,
+      "loss": 0.1053,
+      "step": 10486
+    },
+    {
+      "epoch": 1.6991250810110174,
+      "grad_norm": 0.8764674663543701,
+      "learning_rate": 2.0860392033315584e-06,
+      "loss": 0.1127,
+      "step": 10487
+    },
+    {
+      "epoch": 1.6992871030460144,
+      "grad_norm": 0.8520863652229309,
+      "learning_rate": 2.085607942271734e-06,
+      "loss": 0.0996,
+      "step": 10488
+    },
+    {
+      "epoch": 1.6994491250810109,
+      "grad_norm": 0.8383838534355164,
+      "learning_rate": 2.085176693891303e-06,
+      "loss": 0.1004,
+      "step": 10489
+    },
+    {
+      "epoch": 1.6996111471160078,
+      "grad_norm": 0.8455978631973267,
+      "learning_rate": 2.0847454582034625e-06,
+      "loss": 0.0961,
+      "step": 10490
+    },
+    {
+      "epoch": 1.6997731691510045,
+      "grad_norm": 0.8698278665542603,
+      "learning_rate": 2.084314235221405e-06,
+      "loss": 0.109,
+      "step": 10491
+    },
+    {
+      "epoch": 1.6999351911860012,
+      "grad_norm": 0.7847959995269775,
+      "learning_rate": 2.0838830249583254e-06,
+      "loss": 0.1002,
+      "step": 10492
+    },
+    {
+      "epoch": 1.7000972132209982,
+      "grad_norm": 0.8136616349220276,
+      "learning_rate": 2.0834518274274195e-06,
+      "loss": 0.0992,
+      "step": 10493
+    },
+    {
+      "epoch": 1.7002592352559946,
+      "grad_norm": 0.8103066086769104,
+      "learning_rate": 2.0830206426418794e-06,
+      "loss": 0.0894,
+      "step": 10494
+    },
+    {
+      "epoch": 1.7004212572909916,
+      "grad_norm": 0.7221809029579163,
+      "learning_rate": 2.0825894706148984e-06,
+      "loss": 0.0907,
+      "step": 10495
+    },
+    {
+      "epoch": 1.7005832793259883,
+      "grad_norm": 0.9179300665855408,
+      "learning_rate": 2.0821583113596686e-06,
+      "loss": 0.1152,
+      "step": 10496
+    },
+    {
+      "epoch": 1.700745301360985,
+      "grad_norm": 0.8785554766654968,
+      "learning_rate": 2.0817271648893848e-06,
+      "loss": 0.1152,
+      "step": 10497
+    },
+    {
+      "epoch": 1.700907323395982,
+      "grad_norm": 0.7163464426994324,
+      "learning_rate": 2.081296031217237e-06,
+      "loss": 0.0873,
+      "step": 10498
+    },
+    {
+      "epoch": 1.7010693454309787,
+      "grad_norm": 0.8808443546295166,
+      "learning_rate": 2.0808649103564173e-06,
+      "loss": 0.1108,
+      "step": 10499
+    },
+    {
+      "epoch": 1.7012313674659754,
+      "grad_norm": 0.7949004769325256,
+      "learning_rate": 2.080433802320117e-06,
+      "loss": 0.0997,
+      "step": 10500
+    },
+    {
+      "epoch": 1.7013933895009723,
+      "grad_norm": 0.7522899508476257,
+      "learning_rate": 2.0800027071215265e-06,
+      "loss": 0.0996,
+      "step": 10501
+    },
+    {
+      "epoch": 1.7015554115359688,
+      "grad_norm": 0.7748807072639465,
+      "learning_rate": 2.0795716247738374e-06,
+      "loss": 0.1031,
+      "step": 10502
+    },
+    {
+      "epoch": 1.7017174335709657,
+      "grad_norm": 0.7756296396255493,
+      "learning_rate": 2.0791405552902396e-06,
+      "loss": 0.0983,
+      "step": 10503
+    },
+    {
+      "epoch": 1.7018794556059624,
+      "grad_norm": 0.8680400848388672,
+      "learning_rate": 2.078709498683922e-06,
+      "loss": 0.1047,
+      "step": 10504
+    },
+    {
+      "epoch": 1.7020414776409591,
+      "grad_norm": 0.7261191606521606,
+      "learning_rate": 2.0782784549680744e-06,
+      "loss": 0.091,
+      "step": 10505
+    },
+    {
+      "epoch": 1.702203499675956,
+      "grad_norm": 0.7516106963157654,
+      "learning_rate": 2.0778474241558845e-06,
+      "loss": 0.0944,
+      "step": 10506
+    },
+    {
+      "epoch": 1.7023655217109526,
+      "grad_norm": 0.8887795805931091,
+      "learning_rate": 2.0774164062605425e-06,
+      "loss": 0.1098,
+      "step": 10507
+    },
+    {
+      "epoch": 1.7025275437459495,
+      "grad_norm": 0.7267820835113525,
+      "learning_rate": 2.0769854012952368e-06,
+      "loss": 0.0917,
+      "step": 10508
+    },
+    {
+      "epoch": 1.7026895657809462,
+      "grad_norm": 0.6994649767875671,
+      "learning_rate": 2.076554409273153e-06,
+      "loss": 0.0883,
+      "step": 10509
+    },
+    {
+      "epoch": 1.702851587815943,
+      "grad_norm": 0.8000863194465637,
+      "learning_rate": 2.0761234302074803e-06,
+      "loss": 0.0984,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7030136098509399,
+      "grad_norm": 0.8726096749305725,
+      "learning_rate": 2.075692464111403e-06,
+      "loss": 0.1051,
+      "step": 10511
+    },
+    {
+      "epoch": 1.7031756318859363,
+      "grad_norm": 0.8187968134880066,
+      "learning_rate": 2.0752615109981116e-06,
+      "loss": 0.1046,
+      "step": 10512
+    },
+    {
+      "epoch": 1.7033376539209333,
+      "grad_norm": 0.753304660320282,
+      "learning_rate": 2.074830570880789e-06,
+      "loss": 0.0933,
+      "step": 10513
+    },
+    {
+      "epoch": 1.70349967595593,
+      "grad_norm": 0.8444907665252686,
+      "learning_rate": 2.0743996437726233e-06,
+      "loss": 0.093,
+      "step": 10514
+    },
+    {
+      "epoch": 1.7036616979909267,
+      "grad_norm": 0.8190134167671204,
+      "learning_rate": 2.073968729686797e-06,
+      "loss": 0.0956,
+      "step": 10515
+    },
+    {
+      "epoch": 1.7038237200259236,
+      "grad_norm": 0.6358878016471863,
+      "learning_rate": 2.073537828636497e-06,
+      "loss": 0.0766,
+      "step": 10516
+    },
+    {
+      "epoch": 1.7039857420609201,
+      "grad_norm": 0.7422366738319397,
+      "learning_rate": 2.0731069406349087e-06,
+      "loss": 0.086,
+      "step": 10517
+    },
+    {
+      "epoch": 1.704147764095917,
+      "grad_norm": 0.775372326374054,
+      "learning_rate": 2.0726760656952137e-06,
+      "loss": 0.0876,
+      "step": 10518
+    },
+    {
+      "epoch": 1.7043097861309138,
+      "grad_norm": 0.8271961808204651,
+      "learning_rate": 2.0722452038305976e-06,
+      "loss": 0.1032,
+      "step": 10519
+    },
+    {
+      "epoch": 1.7044718081659105,
+      "grad_norm": 0.6599870920181274,
+      "learning_rate": 2.0718143550542418e-06,
+      "loss": 0.0808,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7046338302009074,
+      "grad_norm": 0.7971540689468384,
+      "learning_rate": 2.071383519379332e-06,
+      "loss": 0.1085,
+      "step": 10521
+    },
+    {
+      "epoch": 1.7047958522359041,
+      "grad_norm": 0.7135748267173767,
+      "learning_rate": 2.0709526968190483e-06,
+      "loss": 0.0867,
+      "step": 10522
+    },
+    {
+      "epoch": 1.7049578742709008,
+      "grad_norm": 0.9244253039360046,
+      "learning_rate": 2.070521887386575e-06,
+      "loss": 0.1061,
+      "step": 10523
+    },
+    {
+      "epoch": 1.7051198963058976,
+      "grad_norm": 0.9357718825340271,
+      "learning_rate": 2.070091091095092e-06,
+      "loss": 0.1023,
+      "step": 10524
+    },
+    {
+      "epoch": 1.7052819183408943,
+      "grad_norm": 0.9502805471420288,
+      "learning_rate": 2.0696603079577808e-06,
+      "loss": 0.1071,
+      "step": 10525
+    },
+    {
+      "epoch": 1.7054439403758912,
+      "grad_norm": 0.8359770178794861,
+      "learning_rate": 2.0692295379878237e-06,
+      "loss": 0.0975,
+      "step": 10526
+    },
+    {
+      "epoch": 1.705605962410888,
+      "grad_norm": 0.9007936120033264,
+      "learning_rate": 2.0687987811983994e-06,
+      "loss": 0.1114,
+      "step": 10527
+    },
+    {
+      "epoch": 1.7057679844458846,
+      "grad_norm": 0.8316621780395508,
+      "learning_rate": 2.0683680376026897e-06,
+      "loss": 0.1015,
+      "step": 10528
+    },
+    {
+      "epoch": 1.7059300064808816,
+      "grad_norm": 0.8183777332305908,
+      "learning_rate": 2.067937307213873e-06,
+      "loss": 0.108,
+      "step": 10529
+    },
+    {
+      "epoch": 1.706092028515878,
+      "grad_norm": 0.8549315333366394,
+      "learning_rate": 2.0675065900451287e-06,
+      "loss": 0.1099,
+      "step": 10530
+    },
+    {
+      "epoch": 1.706254050550875,
+      "grad_norm": 0.8095566630363464,
+      "learning_rate": 2.0670758861096366e-06,
+      "loss": 0.1013,
+      "step": 10531
+    },
+    {
+      "epoch": 1.7064160725858717,
+      "grad_norm": 0.8298711776733398,
+      "learning_rate": 2.066645195420575e-06,
+      "loss": 0.1061,
+      "step": 10532
+    },
+    {
+      "epoch": 1.7065780946208684,
+      "grad_norm": 0.8974525928497314,
+      "learning_rate": 2.0662145179911216e-06,
+      "loss": 0.0994,
+      "step": 10533
+    },
+    {
+      "epoch": 1.7067401166558653,
+      "grad_norm": 0.8173624873161316,
+      "learning_rate": 2.0657838538344545e-06,
+      "loss": 0.1013,
+      "step": 10534
+    },
+    {
+      "epoch": 1.7069021386908618,
+      "grad_norm": 0.8199102878570557,
+      "learning_rate": 2.06535320296375e-06,
+      "loss": 0.1004,
+      "step": 10535
+    },
+    {
+      "epoch": 1.7070641607258588,
+      "grad_norm": 0.8255974054336548,
+      "learning_rate": 2.0649225653921855e-06,
+      "loss": 0.1085,
+      "step": 10536
+    },
+    {
+      "epoch": 1.7072261827608555,
+      "grad_norm": 0.703346848487854,
+      "learning_rate": 2.064491941132938e-06,
+      "loss": 0.0922,
+      "step": 10537
+    },
+    {
+      "epoch": 1.7073882047958522,
+      "grad_norm": 0.7433433532714844,
+      "learning_rate": 2.064061330199184e-06,
+      "loss": 0.0975,
+      "step": 10538
+    },
+    {
+      "epoch": 1.7075502268308491,
+      "grad_norm": 0.7282018065452576,
+      "learning_rate": 2.0636307326040972e-06,
+      "loss": 0.0926,
+      "step": 10539
+    },
+    {
+      "epoch": 1.7077122488658456,
+      "grad_norm": 0.7705861330032349,
+      "learning_rate": 2.0632001483608544e-06,
+      "loss": 0.1065,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7078742709008425,
+      "grad_norm": 0.771189272403717,
+      "learning_rate": 2.0627695774826305e-06,
+      "loss": 0.0907,
+      "step": 10541
+    },
+    {
+      "epoch": 1.7080362929358393,
+      "grad_norm": 0.7684487700462341,
+      "learning_rate": 2.062339019982599e-06,
+      "loss": 0.1023,
+      "step": 10542
+    },
+    {
+      "epoch": 1.708198314970836,
+      "grad_norm": 0.8463597297668457,
+      "learning_rate": 2.0619084758739348e-06,
+      "loss": 0.1077,
+      "step": 10543
+    },
+    {
+      "epoch": 1.708360337005833,
+      "grad_norm": 0.7792380452156067,
+      "learning_rate": 2.06147794516981e-06,
+      "loss": 0.102,
+      "step": 10544
+    },
+    {
+      "epoch": 1.7085223590408296,
+      "grad_norm": 0.8330045938491821,
+      "learning_rate": 2.061047427883399e-06,
+      "loss": 0.1016,
+      "step": 10545
+    },
+    {
+      "epoch": 1.7086843810758263,
+      "grad_norm": 0.7635950446128845,
+      "learning_rate": 2.0606169240278752e-06,
+      "loss": 0.1005,
+      "step": 10546
+    },
+    {
+      "epoch": 1.708846403110823,
+      "grad_norm": 0.8526593446731567,
+      "learning_rate": 2.0601864336164104e-06,
+      "loss": 0.1061,
+      "step": 10547
+    },
+    {
+      "epoch": 1.7090084251458197,
+      "grad_norm": 0.8109990954399109,
+      "learning_rate": 2.059755956662176e-06,
+      "loss": 0.1015,
+      "step": 10548
+    },
+    {
+      "epoch": 1.7091704471808167,
+      "grad_norm": 0.7520654201507568,
+      "learning_rate": 2.0593254931783436e-06,
+      "loss": 0.0998,
+      "step": 10549
+    },
+    {
+      "epoch": 1.7093324692158134,
+      "grad_norm": 0.9032507538795471,
+      "learning_rate": 2.058895043178085e-06,
+      "loss": 0.1102,
+      "step": 10550
+    },
+    {
+      "epoch": 1.70949449125081,
+      "grad_norm": 0.7866460084915161,
+      "learning_rate": 2.0584646066745707e-06,
+      "loss": 0.0917,
+      "step": 10551
+    },
+    {
+      "epoch": 1.709656513285807,
+      "grad_norm": 0.800690770149231,
+      "learning_rate": 2.0580341836809718e-06,
+      "loss": 0.0962,
+      "step": 10552
+    },
+    {
+      "epoch": 1.7098185353208035,
+      "grad_norm": 0.7801680564880371,
+      "learning_rate": 2.0576037742104563e-06,
+      "loss": 0.0926,
+      "step": 10553
+    },
+    {
+      "epoch": 1.7099805573558005,
+      "grad_norm": 0.7548884153366089,
+      "learning_rate": 2.0571733782761943e-06,
+      "loss": 0.0968,
+      "step": 10554
+    },
+    {
+      "epoch": 1.7101425793907972,
+      "grad_norm": 0.7508038282394409,
+      "learning_rate": 2.056742995891356e-06,
+      "loss": 0.0899,
+      "step": 10555
+    },
+    {
+      "epoch": 1.7103046014257939,
+      "grad_norm": 0.8736903667449951,
+      "learning_rate": 2.0563126270691097e-06,
+      "loss": 0.1057,
+      "step": 10556
+    },
+    {
+      "epoch": 1.7104666234607908,
+      "grad_norm": 0.8422747254371643,
+      "learning_rate": 2.0558822718226226e-06,
+      "loss": 0.1087,
+      "step": 10557
+    },
+    {
+      "epoch": 1.7106286454957873,
+      "grad_norm": 0.8379523754119873,
+      "learning_rate": 2.055451930165063e-06,
+      "loss": 0.1065,
+      "step": 10558
+    },
+    {
+      "epoch": 1.7107906675307842,
+      "grad_norm": 0.7081695795059204,
+      "learning_rate": 2.0550216021095993e-06,
+      "loss": 0.0838,
+      "step": 10559
+    },
+    {
+      "epoch": 1.710952689565781,
+      "grad_norm": 0.9157091975212097,
+      "learning_rate": 2.054591287669398e-06,
+      "loss": 0.0991,
+      "step": 10560
+    },
+    {
+      "epoch": 1.7111147116007777,
+      "grad_norm": 0.8604560494422913,
+      "learning_rate": 2.054160986857625e-06,
+      "loss": 0.0997,
+      "step": 10561
+    },
+    {
+      "epoch": 1.7112767336357746,
+      "grad_norm": 0.861049473285675,
+      "learning_rate": 2.053730699687447e-06,
+      "loss": 0.1093,
+      "step": 10562
+    },
+    {
+      "epoch": 1.711438755670771,
+      "grad_norm": 0.8312821388244629,
+      "learning_rate": 2.053300426172029e-06,
+      "loss": 0.1004,
+      "step": 10563
+    },
+    {
+      "epoch": 1.711600777705768,
+      "grad_norm": 0.8654624223709106,
+      "learning_rate": 2.052870166324537e-06,
+      "loss": 0.1035,
+      "step": 10564
+    },
+    {
+      "epoch": 1.7117627997407647,
+      "grad_norm": 0.8277021646499634,
+      "learning_rate": 2.052439920158137e-06,
+      "loss": 0.0949,
+      "step": 10565
+    },
+    {
+      "epoch": 1.7119248217757614,
+      "grad_norm": 0.8619728088378906,
+      "learning_rate": 2.0520096876859918e-06,
+      "loss": 0.1026,
+      "step": 10566
+    },
+    {
+      "epoch": 1.7120868438107584,
+      "grad_norm": 0.8873988389968872,
+      "learning_rate": 2.051579468921266e-06,
+      "loss": 0.1064,
+      "step": 10567
+    },
+    {
+      "epoch": 1.7122488658457549,
+      "grad_norm": 0.8158526420593262,
+      "learning_rate": 2.051149263877123e-06,
+      "loss": 0.1018,
+      "step": 10568
+    },
+    {
+      "epoch": 1.7124108878807518,
+      "grad_norm": 0.8675525784492493,
+      "learning_rate": 2.0507190725667263e-06,
+      "loss": 0.1038,
+      "step": 10569
+    },
+    {
+      "epoch": 1.7125729099157485,
+      "grad_norm": 0.7022666931152344,
+      "learning_rate": 2.0502888950032396e-06,
+      "loss": 0.0982,
+      "step": 10570
+    },
+    {
+      "epoch": 1.7127349319507452,
+      "grad_norm": 1.0205082893371582,
+      "learning_rate": 2.0498587311998235e-06,
+      "loss": 0.1244,
+      "step": 10571
+    },
+    {
+      "epoch": 1.7128969539857422,
+      "grad_norm": 0.7745731472969055,
+      "learning_rate": 2.0494285811696417e-06,
+      "loss": 0.1027,
+      "step": 10572
+    },
+    {
+      "epoch": 1.7130589760207389,
+      "grad_norm": 0.8017035722732544,
+      "learning_rate": 2.0489984449258537e-06,
+      "loss": 0.0978,
+      "step": 10573
+    },
+    {
+      "epoch": 1.7132209980557356,
+      "grad_norm": 0.7118944525718689,
+      "learning_rate": 2.048568322481623e-06,
+      "loss": 0.0959,
+      "step": 10574
+    },
+    {
+      "epoch": 1.7133830200907323,
+      "grad_norm": 0.8438950777053833,
+      "learning_rate": 2.048138213850109e-06,
+      "loss": 0.1062,
+      "step": 10575
+    },
+    {
+      "epoch": 1.713545042125729,
+      "grad_norm": 0.7368760108947754,
+      "learning_rate": 2.0477081190444724e-06,
+      "loss": 0.0892,
+      "step": 10576
+    },
+    {
+      "epoch": 1.713707064160726,
+      "grad_norm": 0.7474925518035889,
+      "learning_rate": 2.0472780380778724e-06,
+      "loss": 0.0932,
+      "step": 10577
+    },
+    {
+      "epoch": 1.7138690861957226,
+      "grad_norm": 0.794523298740387,
+      "learning_rate": 2.046847970963468e-06,
+      "loss": 0.0957,
+      "step": 10578
+    },
+    {
+      "epoch": 1.7140311082307194,
+      "grad_norm": 0.8395006060600281,
+      "learning_rate": 2.0464179177144207e-06,
+      "loss": 0.1031,
+      "step": 10579
+    },
+    {
+      "epoch": 1.7141931302657163,
+      "grad_norm": 0.8234453797340393,
+      "learning_rate": 2.0459878783438867e-06,
+      "loss": 0.105,
+      "step": 10580
+    },
+    {
+      "epoch": 1.7143551523007128,
+      "grad_norm": 0.7391911149024963,
+      "learning_rate": 2.045557852865025e-06,
+      "loss": 0.0862,
+      "step": 10581
+    },
+    {
+      "epoch": 1.7145171743357097,
+      "grad_norm": 0.7759642601013184,
+      "learning_rate": 2.045127841290993e-06,
+      "loss": 0.1004,
+      "step": 10582
+    },
+    {
+      "epoch": 1.7146791963707064,
+      "grad_norm": 0.7772481441497803,
+      "learning_rate": 2.0446978436349486e-06,
+      "loss": 0.0965,
+      "step": 10583
+    },
+    {
+      "epoch": 1.7148412184057031,
+      "grad_norm": 0.8068507313728333,
+      "learning_rate": 2.0442678599100484e-06,
+      "loss": 0.0977,
+      "step": 10584
+    },
+    {
+      "epoch": 1.7150032404407,
+      "grad_norm": 0.8451778292655945,
+      "learning_rate": 2.0438378901294493e-06,
+      "loss": 0.1098,
+      "step": 10585
+    },
+    {
+      "epoch": 1.7151652624756966,
+      "grad_norm": 0.7878326773643494,
+      "learning_rate": 2.043407934306306e-06,
+      "loss": 0.0967,
+      "step": 10586
+    },
+    {
+      "epoch": 1.7153272845106935,
+      "grad_norm": 0.8413339853286743,
+      "learning_rate": 2.042977992453775e-06,
+      "loss": 0.0935,
+      "step": 10587
+    },
+    {
+      "epoch": 1.7154893065456902,
+      "grad_norm": 0.9668422937393188,
+      "learning_rate": 2.0425480645850124e-06,
+      "loss": 0.1142,
+      "step": 10588
+    },
+    {
+      "epoch": 1.715651328580687,
+      "grad_norm": 0.829973578453064,
+      "learning_rate": 2.042118150713171e-06,
+      "loss": 0.1115,
+      "step": 10589
+    },
+    {
+      "epoch": 1.7158133506156839,
+      "grad_norm": 0.8030633330345154,
+      "learning_rate": 2.041688250851407e-06,
+      "loss": 0.0899,
+      "step": 10590
+    },
+    {
+      "epoch": 1.7159753726506803,
+      "grad_norm": 0.9173603057861328,
+      "learning_rate": 2.041258365012873e-06,
+      "loss": 0.1067,
+      "step": 10591
+    },
+    {
+      "epoch": 1.7161373946856773,
+      "grad_norm": 0.9267182946205139,
+      "learning_rate": 2.0408284932107227e-06,
+      "loss": 0.1038,
+      "step": 10592
+    },
+    {
+      "epoch": 1.716299416720674,
+      "grad_norm": 0.7872363328933716,
+      "learning_rate": 2.040398635458109e-06,
+      "loss": 0.1063,
+      "step": 10593
+    },
+    {
+      "epoch": 1.7164614387556707,
+      "grad_norm": 0.7939862012863159,
+      "learning_rate": 2.039968791768186e-06,
+      "loss": 0.1016,
+      "step": 10594
+    },
+    {
+      "epoch": 1.7166234607906676,
+      "grad_norm": 0.7923547625541687,
+      "learning_rate": 2.039538962154104e-06,
+      "loss": 0.0921,
+      "step": 10595
+    },
+    {
+      "epoch": 1.7167854828256643,
+      "grad_norm": 0.7851249575614929,
+      "learning_rate": 2.039109146629016e-06,
+      "loss": 0.0993,
+      "step": 10596
+    },
+    {
+      "epoch": 1.716947504860661,
+      "grad_norm": 0.7416827082633972,
+      "learning_rate": 2.0386793452060717e-06,
+      "loss": 0.0926,
+      "step": 10597
+    },
+    {
+      "epoch": 1.7171095268956578,
+      "grad_norm": 0.8344098329544067,
+      "learning_rate": 2.0382495578984236e-06,
+      "loss": 0.1,
+      "step": 10598
+    },
+    {
+      "epoch": 1.7172715489306545,
+      "grad_norm": 0.81064373254776,
+      "learning_rate": 2.0378197847192216e-06,
+      "loss": 0.0975,
+      "step": 10599
+    },
+    {
+      "epoch": 1.7174335709656514,
+      "grad_norm": 0.8358397483825684,
+      "learning_rate": 2.0373900256816166e-06,
+      "loss": 0.1013,
+      "step": 10600
+    },
+    {
+      "epoch": 1.7175955930006481,
+      "grad_norm": 0.7919216752052307,
+      "learning_rate": 2.0369602807987564e-06,
+      "loss": 0.0989,
+      "step": 10601
+    },
+    {
+      "epoch": 1.7177576150356448,
+      "grad_norm": 0.7866799831390381,
+      "learning_rate": 2.0365305500837906e-06,
+      "loss": 0.0988,
+      "step": 10602
+    },
+    {
+      "epoch": 1.7179196370706418,
+      "grad_norm": 0.7129443287849426,
+      "learning_rate": 2.0361008335498695e-06,
+      "loss": 0.0871,
+      "step": 10603
+    },
+    {
+      "epoch": 1.7180816591056383,
+      "grad_norm": 0.8652666807174683,
+      "learning_rate": 2.0356711312101394e-06,
+      "loss": 0.1064,
+      "step": 10604
+    },
+    {
+      "epoch": 1.7182436811406352,
+      "grad_norm": 0.7662056088447571,
+      "learning_rate": 2.03524144307775e-06,
+      "loss": 0.097,
+      "step": 10605
+    },
+    {
+      "epoch": 1.718405703175632,
+      "grad_norm": 0.8492453098297119,
+      "learning_rate": 2.0348117691658463e-06,
+      "loss": 0.1026,
+      "step": 10606
+    },
+    {
+      "epoch": 1.7185677252106286,
+      "grad_norm": 0.7717750668525696,
+      "learning_rate": 2.0343821094875777e-06,
+      "loss": 0.0952,
+      "step": 10607
+    },
+    {
+      "epoch": 1.7187297472456255,
+      "grad_norm": 0.8496338725090027,
+      "learning_rate": 2.03395246405609e-06,
+      "loss": 0.0991,
+      "step": 10608
+    },
+    {
+      "epoch": 1.718891769280622,
+      "grad_norm": 0.7787384390830994,
+      "learning_rate": 2.0335228328845293e-06,
+      "loss": 0.0885,
+      "step": 10609
+    },
+    {
+      "epoch": 1.719053791315619,
+      "grad_norm": 0.6998451948165894,
+      "learning_rate": 2.03309321598604e-06,
+      "loss": 0.089,
+      "step": 10610
+    },
+    {
+      "epoch": 1.7192158133506157,
+      "grad_norm": 0.8012830018997192,
+      "learning_rate": 2.0326636133737686e-06,
+      "loss": 0.0946,
+      "step": 10611
+    },
+    {
+      "epoch": 1.7193778353856124,
+      "grad_norm": 0.7706624269485474,
+      "learning_rate": 2.03223402506086e-06,
+      "loss": 0.0964,
+      "step": 10612
+    },
+    {
+      "epoch": 1.7195398574206093,
+      "grad_norm": 0.8254458904266357,
+      "learning_rate": 2.0318044510604586e-06,
+      "loss": 0.1,
+      "step": 10613
+    },
+    {
+      "epoch": 1.7197018794556058,
+      "grad_norm": 0.7956833243370056,
+      "learning_rate": 2.031374891385708e-06,
+      "loss": 0.0903,
+      "step": 10614
+    },
+    {
+      "epoch": 1.7198639014906028,
+      "grad_norm": 0.8892257213592529,
+      "learning_rate": 2.030945346049751e-06,
+      "loss": 0.1094,
+      "step": 10615
+    },
+    {
+      "epoch": 1.7200259235255995,
+      "grad_norm": 0.7731269598007202,
+      "learning_rate": 2.0305158150657316e-06,
+      "loss": 0.0916,
+      "step": 10616
+    },
+    {
+      "epoch": 1.7201879455605962,
+      "grad_norm": 0.8199908137321472,
+      "learning_rate": 2.030086298446792e-06,
+      "loss": 0.0944,
+      "step": 10617
+    },
+    {
+      "epoch": 1.720349967595593,
+      "grad_norm": 0.8723816275596619,
+      "learning_rate": 2.0296567962060753e-06,
+      "loss": 0.1072,
+      "step": 10618
+    },
+    {
+      "epoch": 1.7205119896305896,
+      "grad_norm": 0.8999892473220825,
+      "learning_rate": 2.0292273083567215e-06,
+      "loss": 0.1013,
+      "step": 10619
+    },
+    {
+      "epoch": 1.7206740116655865,
+      "grad_norm": 0.7920383810997009,
+      "learning_rate": 2.0287978349118737e-06,
+      "loss": 0.0988,
+      "step": 10620
+    },
+    {
+      "epoch": 1.7208360337005832,
+      "grad_norm": 0.9639683365821838,
+      "learning_rate": 2.0283683758846705e-06,
+      "loss": 0.1159,
+      "step": 10621
+    },
+    {
+      "epoch": 1.72099805573558,
+      "grad_norm": 0.9006828665733337,
+      "learning_rate": 2.0279389312882546e-06,
+      "loss": 0.0991,
+      "step": 10622
+    },
+    {
+      "epoch": 1.721160077770577,
+      "grad_norm": 0.8471527099609375,
+      "learning_rate": 2.0275095011357655e-06,
+      "loss": 0.0966,
+      "step": 10623
+    },
+    {
+      "epoch": 1.7213220998055736,
+      "grad_norm": 0.9707726836204529,
+      "learning_rate": 2.027080085440341e-06,
+      "loss": 0.1061,
+      "step": 10624
+    },
+    {
+      "epoch": 1.7214841218405703,
+      "grad_norm": 0.7146053910255432,
+      "learning_rate": 2.0266506842151216e-06,
+      "loss": 0.0878,
+      "step": 10625
+    },
+    {
+      "epoch": 1.721646143875567,
+      "grad_norm": 0.7884452939033508,
+      "learning_rate": 2.0262212974732465e-06,
+      "loss": 0.0955,
+      "step": 10626
+    },
+    {
+      "epoch": 1.7218081659105637,
+      "grad_norm": 0.8306124210357666,
+      "learning_rate": 2.0257919252278535e-06,
+      "loss": 0.1006,
+      "step": 10627
+    },
+    {
+      "epoch": 1.7219701879455607,
+      "grad_norm": 0.8302614688873291,
+      "learning_rate": 2.0253625674920795e-06,
+      "loss": 0.0927,
+      "step": 10628
+    },
+    {
+      "epoch": 1.7221322099805574,
+      "grad_norm": 0.7767832279205322,
+      "learning_rate": 2.0249332242790627e-06,
+      "loss": 0.0956,
+      "step": 10629
+    },
+    {
+      "epoch": 1.722294232015554,
+      "grad_norm": 0.8298105001449585,
+      "learning_rate": 2.0245038956019386e-06,
+      "loss": 0.1043,
+      "step": 10630
+    },
+    {
+      "epoch": 1.722456254050551,
+      "grad_norm": 0.8026810884475708,
+      "learning_rate": 2.024074581473845e-06,
+      "loss": 0.1006,
+      "step": 10631
+    },
+    {
+      "epoch": 1.7226182760855475,
+      "grad_norm": 0.7589970231056213,
+      "learning_rate": 2.0236452819079183e-06,
+      "loss": 0.0942,
+      "step": 10632
+    },
+    {
+      "epoch": 1.7227802981205445,
+      "grad_norm": 0.7752396464347839,
+      "learning_rate": 2.0232159969172926e-06,
+      "loss": 0.095,
+      "step": 10633
+    },
+    {
+      "epoch": 1.7229423201555412,
+      "grad_norm": 0.7742292284965515,
+      "learning_rate": 2.0227867265151035e-06,
+      "loss": 0.0986,
+      "step": 10634
+    },
+    {
+      "epoch": 1.7231043421905379,
+      "grad_norm": 0.9077650904655457,
+      "learning_rate": 2.0223574707144854e-06,
+      "loss": 0.1147,
+      "step": 10635
+    },
+    {
+      "epoch": 1.7232663642255348,
+      "grad_norm": 0.7141374349594116,
+      "learning_rate": 2.0219282295285734e-06,
+      "loss": 0.0923,
+      "step": 10636
+    },
+    {
+      "epoch": 1.7234283862605313,
+      "grad_norm": 0.7725054621696472,
+      "learning_rate": 2.0214990029705007e-06,
+      "loss": 0.0992,
+      "step": 10637
+    },
+    {
+      "epoch": 1.7235904082955282,
+      "grad_norm": 0.8039776086807251,
+      "learning_rate": 2.021069791053401e-06,
+      "loss": 0.0928,
+      "step": 10638
+    },
+    {
+      "epoch": 1.723752430330525,
+      "grad_norm": 0.8312842845916748,
+      "learning_rate": 2.0206405937904058e-06,
+      "loss": 0.0994,
+      "step": 10639
+    },
+    {
+      "epoch": 1.7239144523655217,
+      "grad_norm": 0.7526655793190002,
+      "learning_rate": 2.0202114111946483e-06,
+      "loss": 0.0902,
+      "step": 10640
+    },
+    {
+      "epoch": 1.7240764744005186,
+      "grad_norm": 0.9838579297065735,
+      "learning_rate": 2.0197822432792606e-06,
+      "loss": 0.109,
+      "step": 10641
+    },
+    {
+      "epoch": 1.724238496435515,
+      "grad_norm": 0.9125879406929016,
+      "learning_rate": 2.019353090057375e-06,
+      "loss": 0.1185,
+      "step": 10642
+    },
+    {
+      "epoch": 1.724400518470512,
+      "grad_norm": 0.7206192016601562,
+      "learning_rate": 2.0189239515421214e-06,
+      "loss": 0.0869,
+      "step": 10643
+    },
+    {
+      "epoch": 1.7245625405055087,
+      "grad_norm": 0.8060803413391113,
+      "learning_rate": 2.018494827746631e-06,
+      "loss": 0.0971,
+      "step": 10644
+    },
+    {
+      "epoch": 1.7247245625405054,
+      "grad_norm": 0.804918110370636,
+      "learning_rate": 2.0180657186840326e-06,
+      "loss": 0.1081,
+      "step": 10645
+    },
+    {
+      "epoch": 1.7248865845755024,
+      "grad_norm": 0.8952481150627136,
+      "learning_rate": 2.0176366243674575e-06,
+      "loss": 0.1074,
+      "step": 10646
+    },
+    {
+      "epoch": 1.725048606610499,
+      "grad_norm": 0.7076264023780823,
+      "learning_rate": 2.0172075448100347e-06,
+      "loss": 0.0892,
+      "step": 10647
+    },
+    {
+      "epoch": 1.7252106286454958,
+      "grad_norm": 0.7837821245193481,
+      "learning_rate": 2.0167784800248924e-06,
+      "loss": 0.1005,
+      "step": 10648
+    },
+    {
+      "epoch": 1.7253726506804925,
+      "grad_norm": 0.7912270426750183,
+      "learning_rate": 2.0163494300251586e-06,
+      "loss": 0.1001,
+      "step": 10649
+    },
+    {
+      "epoch": 1.7255346727154892,
+      "grad_norm": 0.8220000863075256,
+      "learning_rate": 2.0159203948239624e-06,
+      "loss": 0.1018,
+      "step": 10650
+    },
+    {
+      "epoch": 1.7256966947504861,
+      "grad_norm": 0.8370126485824585,
+      "learning_rate": 2.0154913744344316e-06,
+      "loss": 0.1055,
+      "step": 10651
+    },
+    {
+      "epoch": 1.7258587167854829,
+      "grad_norm": 0.8922955989837646,
+      "learning_rate": 2.015062368869691e-06,
+      "loss": 0.1027,
+      "step": 10652
+    },
+    {
+      "epoch": 1.7260207388204796,
+      "grad_norm": 0.7914833426475525,
+      "learning_rate": 2.0146333781428694e-06,
+      "loss": 0.0964,
+      "step": 10653
+    },
+    {
+      "epoch": 1.7261827608554765,
+      "grad_norm": 0.8470481038093567,
+      "learning_rate": 2.0142044022670905e-06,
+      "loss": 0.1105,
+      "step": 10654
+    },
+    {
+      "epoch": 1.726344782890473,
+      "grad_norm": 0.7832110524177551,
+      "learning_rate": 2.0137754412554823e-06,
+      "loss": 0.095,
+      "step": 10655
+    },
+    {
+      "epoch": 1.72650680492547,
+      "grad_norm": 0.8746216297149658,
+      "learning_rate": 2.013346495121169e-06,
+      "loss": 0.1111,
+      "step": 10656
+    },
+    {
+      "epoch": 1.7266688269604666,
+      "grad_norm": 0.8226485848426819,
+      "learning_rate": 2.012917563877275e-06,
+      "loss": 0.0996,
+      "step": 10657
+    },
+    {
+      "epoch": 1.7268308489954634,
+      "grad_norm": 0.7802627682685852,
+      "learning_rate": 2.012488647536925e-06,
+      "loss": 0.1043,
+      "step": 10658
+    },
+    {
+      "epoch": 1.7269928710304603,
+      "grad_norm": 0.9438074231147766,
+      "learning_rate": 2.0120597461132416e-06,
+      "loss": 0.1009,
+      "step": 10659
+    },
+    {
+      "epoch": 1.7271548930654568,
+      "grad_norm": 0.810778796672821,
+      "learning_rate": 2.0116308596193502e-06,
+      "loss": 0.1009,
+      "step": 10660
+    },
+    {
+      "epoch": 1.7273169151004537,
+      "grad_norm": 0.7767440676689148,
+      "learning_rate": 2.0112019880683724e-06,
+      "loss": 0.0921,
+      "step": 10661
+    },
+    {
+      "epoch": 1.7274789371354504,
+      "grad_norm": 0.8341066241264343,
+      "learning_rate": 2.0107731314734316e-06,
+      "loss": 0.1,
+      "step": 10662
+    },
+    {
+      "epoch": 1.7276409591704471,
+      "grad_norm": 0.8395964503288269,
+      "learning_rate": 2.0103442898476484e-06,
+      "loss": 0.1091,
+      "step": 10663
+    },
+    {
+      "epoch": 1.727802981205444,
+      "grad_norm": 0.9881349802017212,
+      "learning_rate": 2.0099154632041446e-06,
+      "loss": 0.1047,
+      "step": 10664
+    },
+    {
+      "epoch": 1.7279650032404406,
+      "grad_norm": 0.7424138784408569,
+      "learning_rate": 2.0094866515560424e-06,
+      "loss": 0.0817,
+      "step": 10665
+    },
+    {
+      "epoch": 1.7281270252754375,
+      "grad_norm": 1.004927635192871,
+      "learning_rate": 2.0090578549164614e-06,
+      "loss": 0.117,
+      "step": 10666
+    },
+    {
+      "epoch": 1.7282890473104342,
+      "grad_norm": 0.8730446100234985,
+      "learning_rate": 2.0086290732985224e-06,
+      "loss": 0.0988,
+      "step": 10667
+    },
+    {
+      "epoch": 1.728451069345431,
+      "grad_norm": 0.8061848282814026,
+      "learning_rate": 2.0082003067153436e-06,
+      "loss": 0.1047,
+      "step": 10668
+    },
+    {
+      "epoch": 1.7286130913804278,
+      "grad_norm": 0.7493446469306946,
+      "learning_rate": 2.0077715551800457e-06,
+      "loss": 0.0995,
+      "step": 10669
+    },
+    {
+      "epoch": 1.7287751134154243,
+      "grad_norm": 0.8697327971458435,
+      "learning_rate": 2.007342818705747e-06,
+      "loss": 0.0994,
+      "step": 10670
+    },
+    {
+      "epoch": 1.7289371354504213,
+      "grad_norm": 0.842828631401062,
+      "learning_rate": 2.0069140973055663e-06,
+      "loss": 0.1031,
+      "step": 10671
+    },
+    {
+      "epoch": 1.729099157485418,
+      "grad_norm": 0.7752946019172668,
+      "learning_rate": 2.006485390992621e-06,
+      "loss": 0.0983,
+      "step": 10672
+    },
+    {
+      "epoch": 1.7292611795204147,
+      "grad_norm": 0.71074378490448,
+      "learning_rate": 2.0060566997800274e-06,
+      "loss": 0.0777,
+      "step": 10673
+    },
+    {
+      "epoch": 1.7294232015554116,
+      "grad_norm": 0.7402112483978271,
+      "learning_rate": 2.0056280236809044e-06,
+      "loss": 0.0999,
+      "step": 10674
+    },
+    {
+      "epoch": 1.7295852235904083,
+      "grad_norm": 0.7987510561943054,
+      "learning_rate": 2.005199362708367e-06,
+      "loss": 0.0993,
+      "step": 10675
+    },
+    {
+      "epoch": 1.729747245625405,
+      "grad_norm": 0.9736399054527283,
+      "learning_rate": 2.004770716875533e-06,
+      "loss": 0.1144,
+      "step": 10676
+    },
+    {
+      "epoch": 1.7299092676604018,
+      "grad_norm": 0.7902863621711731,
+      "learning_rate": 2.0043420861955155e-06,
+      "loss": 0.0946,
+      "step": 10677
+    },
+    {
+      "epoch": 1.7300712896953985,
+      "grad_norm": 0.8167175650596619,
+      "learning_rate": 2.0039134706814303e-06,
+      "loss": 0.1059,
+      "step": 10678
+    },
+    {
+      "epoch": 1.7302333117303954,
+      "grad_norm": 0.8277512788772583,
+      "learning_rate": 2.003484870346393e-06,
+      "loss": 0.1024,
+      "step": 10679
+    },
+    {
+      "epoch": 1.7303953337653921,
+      "grad_norm": 0.7947298884391785,
+      "learning_rate": 2.0030562852035175e-06,
+      "loss": 0.1047,
+      "step": 10680
+    },
+    {
+      "epoch": 1.7305573558003888,
+      "grad_norm": 0.7207613587379456,
+      "learning_rate": 2.0026277152659166e-06,
+      "loss": 0.0903,
+      "step": 10681
+    },
+    {
+      "epoch": 1.7307193778353858,
+      "grad_norm": 0.8288989663124084,
+      "learning_rate": 2.0021991605467043e-06,
+      "loss": 0.0966,
+      "step": 10682
+    },
+    {
+      "epoch": 1.7308813998703823,
+      "grad_norm": 0.7648212909698486,
+      "learning_rate": 2.0017706210589925e-06,
+      "loss": 0.0918,
+      "step": 10683
+    },
+    {
+      "epoch": 1.7310434219053792,
+      "grad_norm": 0.7417365312576294,
+      "learning_rate": 2.0013420968158944e-06,
+      "loss": 0.0972,
+      "step": 10684
+    },
+    {
+      "epoch": 1.731205443940376,
+      "grad_norm": 0.8182329535484314,
+      "learning_rate": 2.000913587830521e-06,
+      "loss": 0.1003,
+      "step": 10685
+    },
+    {
+      "epoch": 1.7313674659753726,
+      "grad_norm": 0.8272933959960938,
+      "learning_rate": 2.0004850941159847e-06,
+      "loss": 0.1045,
+      "step": 10686
+    },
+    {
+      "epoch": 1.7315294880103695,
+      "grad_norm": 0.8044247031211853,
+      "learning_rate": 2.0000566156853957e-06,
+      "loss": 0.0977,
+      "step": 10687
+    },
+    {
+      "epoch": 1.731691510045366,
+      "grad_norm": 0.7936481237411499,
+      "learning_rate": 1.999628152551863e-06,
+      "loss": 0.0906,
+      "step": 10688
+    },
+    {
+      "epoch": 1.731853532080363,
+      "grad_norm": 0.8275524973869324,
+      "learning_rate": 1.9991997047285e-06,
+      "loss": 0.0972,
+      "step": 10689
+    },
+    {
+      "epoch": 1.7320155541153597,
+      "grad_norm": 0.7702337503433228,
+      "learning_rate": 1.9987712722284132e-06,
+      "loss": 0.0968,
+      "step": 10690
+    },
+    {
+      "epoch": 1.7321775761503564,
+      "grad_norm": 0.6895307302474976,
+      "learning_rate": 1.9983428550647126e-06,
+      "loss": 0.0807,
+      "step": 10691
+    },
+    {
+      "epoch": 1.7323395981853533,
+      "grad_norm": 0.8178178668022156,
+      "learning_rate": 1.9979144532505064e-06,
+      "loss": 0.0974,
+      "step": 10692
+    },
+    {
+      "epoch": 1.7325016202203498,
+      "grad_norm": 0.7914819717407227,
+      "learning_rate": 1.997486066798903e-06,
+      "loss": 0.0964,
+      "step": 10693
+    },
+    {
+      "epoch": 1.7326636422553467,
+      "grad_norm": 0.8385084867477417,
+      "learning_rate": 1.9970576957230094e-06,
+      "loss": 0.0965,
+      "step": 10694
+    },
+    {
+      "epoch": 1.7328256642903435,
+      "grad_norm": 0.9208899140357971,
+      "learning_rate": 1.9966293400359343e-06,
+      "loss": 0.1082,
+      "step": 10695
+    },
+    {
+      "epoch": 1.7329876863253402,
+      "grad_norm": 0.9039163589477539,
+      "learning_rate": 1.996200999750783e-06,
+      "loss": 0.1038,
+      "step": 10696
+    },
+    {
+      "epoch": 1.733149708360337,
+      "grad_norm": 0.7650443911552429,
+      "learning_rate": 1.9957726748806608e-06,
+      "loss": 0.0898,
+      "step": 10697
+    },
+    {
+      "epoch": 1.7333117303953338,
+      "grad_norm": 0.8120889663696289,
+      "learning_rate": 1.995344365438676e-06,
+      "loss": 0.1004,
+      "step": 10698
+    },
+    {
+      "epoch": 1.7334737524303305,
+      "grad_norm": 0.9312616586685181,
+      "learning_rate": 1.9949160714379314e-06,
+      "loss": 0.1158,
+      "step": 10699
+    },
+    {
+      "epoch": 1.7336357744653272,
+      "grad_norm": 0.8903511166572571,
+      "learning_rate": 1.994487792891534e-06,
+      "loss": 0.1053,
+      "step": 10700
+    },
+    {
+      "epoch": 1.733797796500324,
+      "grad_norm": 0.8750895261764526,
+      "learning_rate": 1.9940595298125855e-06,
+      "loss": 0.1043,
+      "step": 10701
+    },
+    {
+      "epoch": 1.7339598185353209,
+      "grad_norm": 0.8411079049110413,
+      "learning_rate": 1.993631282214191e-06,
+      "loss": 0.1051,
+      "step": 10702
+    },
+    {
+      "epoch": 1.7341218405703176,
+      "grad_norm": 0.7505423426628113,
+      "learning_rate": 1.993203050109454e-06,
+      "loss": 0.1055,
+      "step": 10703
+    },
+    {
+      "epoch": 1.7342838626053143,
+      "grad_norm": 0.7114359736442566,
+      "learning_rate": 1.992774833511478e-06,
+      "loss": 0.0923,
+      "step": 10704
+    },
+    {
+      "epoch": 1.7344458846403112,
+      "grad_norm": 1.49692964553833,
+      "learning_rate": 1.9923466324333646e-06,
+      "loss": 0.1233,
+      "step": 10705
+    },
+    {
+      "epoch": 1.7346079066753077,
+      "grad_norm": 0.8229801654815674,
+      "learning_rate": 1.991918446888216e-06,
+      "loss": 0.1105,
+      "step": 10706
+    },
+    {
+      "epoch": 1.7347699287103047,
+      "grad_norm": 0.7137386202812195,
+      "learning_rate": 1.991490276889132e-06,
+      "loss": 0.0934,
+      "step": 10707
+    },
+    {
+      "epoch": 1.7349319507453014,
+      "grad_norm": 0.8458231687545776,
+      "learning_rate": 1.9910621224492154e-06,
+      "loss": 0.1179,
+      "step": 10708
+    },
+    {
+      "epoch": 1.735093972780298,
+      "grad_norm": 0.8449655771255493,
+      "learning_rate": 1.990633983581567e-06,
+      "loss": 0.1009,
+      "step": 10709
+    },
+    {
+      "epoch": 1.735255994815295,
+      "grad_norm": 0.8100825548171997,
+      "learning_rate": 1.9902058602992856e-06,
+      "loss": 0.0982,
+      "step": 10710
+    },
+    {
+      "epoch": 1.7354180168502915,
+      "grad_norm": 0.6856768131256104,
+      "learning_rate": 1.9897777526154717e-06,
+      "loss": 0.0872,
+      "step": 10711
+    },
+    {
+      "epoch": 1.7355800388852884,
+      "grad_norm": 0.8536728620529175,
+      "learning_rate": 1.989349660543222e-06,
+      "loss": 0.1053,
+      "step": 10712
+    },
+    {
+      "epoch": 1.7357420609202852,
+      "grad_norm": 0.6712336540222168,
+      "learning_rate": 1.988921584095639e-06,
+      "loss": 0.08,
+      "step": 10713
+    },
+    {
+      "epoch": 1.7359040829552819,
+      "grad_norm": 0.8158391118049622,
+      "learning_rate": 1.988493523285818e-06,
+      "loss": 0.0878,
+      "step": 10714
+    },
+    {
+      "epoch": 1.7360661049902788,
+      "grad_norm": 0.8350120782852173,
+      "learning_rate": 1.988065478126858e-06,
+      "loss": 0.0983,
+      "step": 10715
+    },
+    {
+      "epoch": 1.7362281270252753,
+      "grad_norm": 0.8612555265426636,
+      "learning_rate": 1.9876374486318545e-06,
+      "loss": 0.1114,
+      "step": 10716
+    },
+    {
+      "epoch": 1.7363901490602722,
+      "grad_norm": 0.8192741274833679,
+      "learning_rate": 1.9872094348139054e-06,
+      "loss": 0.1062,
+      "step": 10717
+    },
+    {
+      "epoch": 1.736552171095269,
+      "grad_norm": 0.8786032795906067,
+      "learning_rate": 1.9867814366861075e-06,
+      "loss": 0.1065,
+      "step": 10718
+    },
+    {
+      "epoch": 1.7367141931302656,
+      "grad_norm": 0.7366222739219666,
+      "learning_rate": 1.986353454261555e-06,
+      "loss": 0.0918,
+      "step": 10719
+    },
+    {
+      "epoch": 1.7368762151652626,
+      "grad_norm": 0.7951706051826477,
+      "learning_rate": 1.9859254875533435e-06,
+      "loss": 0.0903,
+      "step": 10720
+    },
+    {
+      "epoch": 1.737038237200259,
+      "grad_norm": 0.8006754517555237,
+      "learning_rate": 1.985497536574568e-06,
+      "loss": 0.0967,
+      "step": 10721
+    },
+    {
+      "epoch": 1.737200259235256,
+      "grad_norm": 0.8424306511878967,
+      "learning_rate": 1.9850696013383236e-06,
+      "loss": 0.103,
+      "step": 10722
+    },
+    {
+      "epoch": 1.7373622812702527,
+      "grad_norm": 0.7816236615180969,
+      "learning_rate": 1.9846416818577025e-06,
+      "loss": 0.0935,
+      "step": 10723
+    },
+    {
+      "epoch": 1.7375243033052494,
+      "grad_norm": 0.9311386942863464,
+      "learning_rate": 1.9842137781458e-06,
+      "loss": 0.1199,
+      "step": 10724
+    },
+    {
+      "epoch": 1.7376863253402464,
+      "grad_norm": 0.7541922926902771,
+      "learning_rate": 1.983785890215707e-06,
+      "loss": 0.0901,
+      "step": 10725
+    },
+    {
+      "epoch": 1.737848347375243,
+      "grad_norm": 0.842083215713501,
+      "learning_rate": 1.9833580180805155e-06,
+      "loss": 0.1022,
+      "step": 10726
+    },
+    {
+      "epoch": 1.7380103694102398,
+      "grad_norm": 0.7754312753677368,
+      "learning_rate": 1.98293016175332e-06,
+      "loss": 0.104,
+      "step": 10727
+    },
+    {
+      "epoch": 1.7381723914452365,
+      "grad_norm": 0.7790486216545105,
+      "learning_rate": 1.9825023212472095e-06,
+      "loss": 0.0973,
+      "step": 10728
+    },
+    {
+      "epoch": 1.7383344134802332,
+      "grad_norm": 0.9156441688537598,
+      "learning_rate": 1.9820744965752754e-06,
+      "loss": 0.1056,
+      "step": 10729
+    },
+    {
+      "epoch": 1.7384964355152301,
+      "grad_norm": 0.7913225293159485,
+      "learning_rate": 1.9816466877506095e-06,
+      "loss": 0.0946,
+      "step": 10730
+    },
+    {
+      "epoch": 1.7386584575502269,
+      "grad_norm": 0.7693167328834534,
+      "learning_rate": 1.981218894786299e-06,
+      "loss": 0.0908,
+      "step": 10731
+    },
+    {
+      "epoch": 1.7388204795852236,
+      "grad_norm": 0.7898248434066772,
+      "learning_rate": 1.9807911176954357e-06,
+      "loss": 0.0949,
+      "step": 10732
+    },
+    {
+      "epoch": 1.7389825016202205,
+      "grad_norm": 0.7249963283538818,
+      "learning_rate": 1.980363356491108e-06,
+      "loss": 0.0939,
+      "step": 10733
+    },
+    {
+      "epoch": 1.739144523655217,
+      "grad_norm": 0.8590476512908936,
+      "learning_rate": 1.9799356111864036e-06,
+      "loss": 0.1106,
+      "step": 10734
+    },
+    {
+      "epoch": 1.739306545690214,
+      "grad_norm": 0.8085042834281921,
+      "learning_rate": 1.979507881794412e-06,
+      "loss": 0.1028,
+      "step": 10735
+    },
+    {
+      "epoch": 1.7394685677252106,
+      "grad_norm": 0.801541805267334,
+      "learning_rate": 1.979080168328218e-06,
+      "loss": 0.0979,
+      "step": 10736
+    },
+    {
+      "epoch": 1.7396305897602073,
+      "grad_norm": 0.8037110567092896,
+      "learning_rate": 1.978652470800911e-06,
+      "loss": 0.1054,
+      "step": 10737
+    },
+    {
+      "epoch": 1.7397926117952043,
+      "grad_norm": 0.830059826374054,
+      "learning_rate": 1.9782247892255767e-06,
+      "loss": 0.0985,
+      "step": 10738
+    },
+    {
+      "epoch": 1.7399546338302008,
+      "grad_norm": 0.8223294019699097,
+      "learning_rate": 1.977797123615302e-06,
+      "loss": 0.0968,
+      "step": 10739
+    },
+    {
+      "epoch": 1.7401166558651977,
+      "grad_norm": 0.8012174367904663,
+      "learning_rate": 1.9773694739831702e-06,
+      "loss": 0.0944,
+      "step": 10740
+    },
+    {
+      "epoch": 1.7402786779001944,
+      "grad_norm": 0.9736762046813965,
+      "learning_rate": 1.9769418403422685e-06,
+      "loss": 0.1149,
+      "step": 10741
+    },
+    {
+      "epoch": 1.7404406999351911,
+      "grad_norm": 0.8288683891296387,
+      "learning_rate": 1.976514222705681e-06,
+      "loss": 0.0896,
+      "step": 10742
+    },
+    {
+      "epoch": 1.740602721970188,
+      "grad_norm": 0.908768892288208,
+      "learning_rate": 1.976086621086491e-06,
+      "loss": 0.1,
+      "step": 10743
+    },
+    {
+      "epoch": 1.7407647440051845,
+      "grad_norm": 0.8298162221908569,
+      "learning_rate": 1.975659035497783e-06,
+      "loss": 0.0956,
+      "step": 10744
+    },
+    {
+      "epoch": 1.7409267660401815,
+      "grad_norm": 0.8755427598953247,
+      "learning_rate": 1.9752314659526383e-06,
+      "loss": 0.1017,
+      "step": 10745
+    },
+    {
+      "epoch": 1.7410887880751782,
+      "grad_norm": 0.7467525005340576,
+      "learning_rate": 1.9748039124641426e-06,
+      "loss": 0.0907,
+      "step": 10746
+    },
+    {
+      "epoch": 1.741250810110175,
+      "grad_norm": 0.7956831455230713,
+      "learning_rate": 1.974376375045375e-06,
+      "loss": 0.0944,
+      "step": 10747
+    },
+    {
+      "epoch": 1.7414128321451718,
+      "grad_norm": 0.7522971034049988,
+      "learning_rate": 1.9739488537094197e-06,
+      "loss": 0.0952,
+      "step": 10748
+    },
+    {
+      "epoch": 1.7415748541801686,
+      "grad_norm": 0.8951601982116699,
+      "learning_rate": 1.973521348469355e-06,
+      "loss": 0.1079,
+      "step": 10749
+    },
+    {
+      "epoch": 1.7417368762151653,
+      "grad_norm": 0.8221662640571594,
+      "learning_rate": 1.973093859338263e-06,
+      "loss": 0.098,
+      "step": 10750
+    },
+    {
+      "epoch": 1.741898898250162,
+      "grad_norm": 0.8099002242088318,
+      "learning_rate": 1.972666386329225e-06,
+      "loss": 0.0924,
+      "step": 10751
+    },
+    {
+      "epoch": 1.7420609202851587,
+      "grad_norm": 0.8443790674209595,
+      "learning_rate": 1.9722389294553188e-06,
+      "loss": 0.1002,
+      "step": 10752
+    },
+    {
+      "epoch": 1.7422229423201556,
+      "grad_norm": 0.777952253818512,
+      "learning_rate": 1.9718114887296245e-06,
+      "loss": 0.0979,
+      "step": 10753
+    },
+    {
+      "epoch": 1.7423849643551523,
+      "grad_norm": 0.7357156276702881,
+      "learning_rate": 1.9713840641652206e-06,
+      "loss": 0.093,
+      "step": 10754
+    },
+    {
+      "epoch": 1.742546986390149,
+      "grad_norm": 0.8331161141395569,
+      "learning_rate": 1.970956655775184e-06,
+      "loss": 0.1058,
+      "step": 10755
+    },
+    {
+      "epoch": 1.742709008425146,
+      "grad_norm": 1.1018832921981812,
+      "learning_rate": 1.970529263572594e-06,
+      "loss": 0.1138,
+      "step": 10756
+    },
+    {
+      "epoch": 1.7428710304601425,
+      "grad_norm": 0.8057063221931458,
+      "learning_rate": 1.9701018875705277e-06,
+      "loss": 0.1083,
+      "step": 10757
+    },
+    {
+      "epoch": 1.7430330524951394,
+      "grad_norm": 0.8819854259490967,
+      "learning_rate": 1.9696745277820613e-06,
+      "loss": 0.1038,
+      "step": 10758
+    },
+    {
+      "epoch": 1.7431950745301361,
+      "grad_norm": 0.7874314188957214,
+      "learning_rate": 1.969247184220271e-06,
+      "loss": 0.0942,
+      "step": 10759
+    },
+    {
+      "epoch": 1.7433570965651328,
+      "grad_norm": 0.8394420742988586,
+      "learning_rate": 1.9688198568982316e-06,
+      "loss": 0.1002,
+      "step": 10760
+    },
+    {
+      "epoch": 1.7435191186001298,
+      "grad_norm": 0.8666523694992065,
+      "learning_rate": 1.9683925458290196e-06,
+      "loss": 0.1118,
+      "step": 10761
+    },
+    {
+      "epoch": 1.7436811406351262,
+      "grad_norm": 0.8129344582557678,
+      "learning_rate": 1.96796525102571e-06,
+      "loss": 0.1116,
+      "step": 10762
+    },
+    {
+      "epoch": 1.7438431626701232,
+      "grad_norm": 0.7237949967384338,
+      "learning_rate": 1.9675379725013752e-06,
+      "loss": 0.0892,
+      "step": 10763
+    },
+    {
+      "epoch": 1.74400518470512,
+      "grad_norm": 0.757645845413208,
+      "learning_rate": 1.96711071026909e-06,
+      "loss": 0.0978,
+      "step": 10764
+    },
+    {
+      "epoch": 1.7441672067401166,
+      "grad_norm": 0.7295411825180054,
+      "learning_rate": 1.9666834643419276e-06,
+      "loss": 0.087,
+      "step": 10765
+    },
+    {
+      "epoch": 1.7443292287751135,
+      "grad_norm": 0.8933332562446594,
+      "learning_rate": 1.9662562347329613e-06,
+      "loss": 0.1023,
+      "step": 10766
+    },
+    {
+      "epoch": 1.74449125081011,
+      "grad_norm": 0.8272913098335266,
+      "learning_rate": 1.9658290214552616e-06,
+      "loss": 0.1089,
+      "step": 10767
+    },
+    {
+      "epoch": 1.744653272845107,
+      "grad_norm": 0.8127201795578003,
+      "learning_rate": 1.9654018245219024e-06,
+      "loss": 0.0923,
+      "step": 10768
+    },
+    {
+      "epoch": 1.7448152948801037,
+      "grad_norm": 0.7642537951469421,
+      "learning_rate": 1.9649746439459523e-06,
+      "loss": 0.0996,
+      "step": 10769
+    },
+    {
+      "epoch": 1.7449773169151004,
+      "grad_norm": 0.7745651602745056,
+      "learning_rate": 1.9645474797404838e-06,
+      "loss": 0.0977,
+      "step": 10770
+    },
+    {
+      "epoch": 1.7451393389500973,
+      "grad_norm": 0.7775667905807495,
+      "learning_rate": 1.964120331918567e-06,
+      "loss": 0.0981,
+      "step": 10771
+    },
+    {
+      "epoch": 1.7453013609850938,
+      "grad_norm": 0.776843786239624,
+      "learning_rate": 1.963693200493271e-06,
+      "loss": 0.1047,
+      "step": 10772
+    },
+    {
+      "epoch": 1.7454633830200907,
+      "grad_norm": 0.6424158215522766,
+      "learning_rate": 1.963266085477665e-06,
+      "loss": 0.0803,
+      "step": 10773
+    },
+    {
+      "epoch": 1.7456254050550875,
+      "grad_norm": 0.8612257242202759,
+      "learning_rate": 1.962838986884818e-06,
+      "loss": 0.1058,
+      "step": 10774
+    },
+    {
+      "epoch": 1.7457874270900842,
+      "grad_norm": 0.8193663358688354,
+      "learning_rate": 1.9624119047277987e-06,
+      "loss": 0.0973,
+      "step": 10775
+    },
+    {
+      "epoch": 1.745949449125081,
+      "grad_norm": 0.7953779101371765,
+      "learning_rate": 1.9619848390196734e-06,
+      "loss": 0.1113,
+      "step": 10776
+    },
+    {
+      "epoch": 1.7461114711600778,
+      "grad_norm": 0.8136135339736938,
+      "learning_rate": 1.961557789773511e-06,
+      "loss": 0.1051,
+      "step": 10777
+    },
+    {
+      "epoch": 1.7462734931950745,
+      "grad_norm": 0.7347664833068848,
+      "learning_rate": 1.9611307570023766e-06,
+      "loss": 0.0897,
+      "step": 10778
+    },
+    {
+      "epoch": 1.7464355152300715,
+      "grad_norm": 0.6952412128448486,
+      "learning_rate": 1.9607037407193365e-06,
+      "loss": 0.0851,
+      "step": 10779
+    },
+    {
+      "epoch": 1.746597537265068,
+      "grad_norm": 0.8789510726928711,
+      "learning_rate": 1.960276740937458e-06,
+      "loss": 0.1074,
+      "step": 10780
+    },
+    {
+      "epoch": 1.7467595593000649,
+      "grad_norm": 0.8486936092376709,
+      "learning_rate": 1.9598497576698043e-06,
+      "loss": 0.0999,
+      "step": 10781
+    },
+    {
+      "epoch": 1.7469215813350616,
+      "grad_norm": 0.8231163620948792,
+      "learning_rate": 1.959422790929441e-06,
+      "loss": 0.1075,
+      "step": 10782
+    },
+    {
+      "epoch": 1.7470836033700583,
+      "grad_norm": 0.7864092588424683,
+      "learning_rate": 1.9589958407294317e-06,
+      "loss": 0.0967,
+      "step": 10783
+    },
+    {
+      "epoch": 1.7472456254050552,
+      "grad_norm": 0.7766144871711731,
+      "learning_rate": 1.9585689070828413e-06,
+      "loss": 0.0872,
+      "step": 10784
+    },
+    {
+      "epoch": 1.7474076474400517,
+      "grad_norm": 0.9048768877983093,
+      "learning_rate": 1.9581419900027317e-06,
+      "loss": 0.101,
+      "step": 10785
+    },
+    {
+      "epoch": 1.7475696694750487,
+      "grad_norm": 0.7717433571815491,
+      "learning_rate": 1.9577150895021664e-06,
+      "loss": 0.0954,
+      "step": 10786
+    },
+    {
+      "epoch": 1.7477316915100454,
+      "grad_norm": 0.8796309232711792,
+      "learning_rate": 1.9572882055942065e-06,
+      "loss": 0.101,
+      "step": 10787
+    },
+    {
+      "epoch": 1.747893713545042,
+      "grad_norm": 0.9133003354072571,
+      "learning_rate": 1.9568613382919142e-06,
+      "loss": 0.1146,
+      "step": 10788
+    },
+    {
+      "epoch": 1.748055735580039,
+      "grad_norm": 0.7639305591583252,
+      "learning_rate": 1.9564344876083504e-06,
+      "loss": 0.0979,
+      "step": 10789
+    },
+    {
+      "epoch": 1.7482177576150355,
+      "grad_norm": 0.7829166650772095,
+      "learning_rate": 1.9560076535565766e-06,
+      "loss": 0.1073,
+      "step": 10790
+    },
+    {
+      "epoch": 1.7483797796500324,
+      "grad_norm": 0.7283692359924316,
+      "learning_rate": 1.955580836149652e-06,
+      "loss": 0.0943,
+      "step": 10791
+    },
+    {
+      "epoch": 1.7485418016850292,
+      "grad_norm": 0.886214017868042,
+      "learning_rate": 1.9551540354006366e-06,
+      "loss": 0.1067,
+      "step": 10792
+    },
+    {
+      "epoch": 1.7487038237200259,
+      "grad_norm": 0.7852003574371338,
+      "learning_rate": 1.954727251322588e-06,
+      "loss": 0.0953,
+      "step": 10793
+    },
+    {
+      "epoch": 1.7488658457550228,
+      "grad_norm": 0.8498847484588623,
+      "learning_rate": 1.954300483928567e-06,
+      "loss": 0.1003,
+      "step": 10794
+    },
+    {
+      "epoch": 1.7490278677900193,
+      "grad_norm": 0.8580987453460693,
+      "learning_rate": 1.9538737332316304e-06,
+      "loss": 0.1139,
+      "step": 10795
+    },
+    {
+      "epoch": 1.7491898898250162,
+      "grad_norm": 0.800059974193573,
+      "learning_rate": 1.953446999244836e-06,
+      "loss": 0.0961,
+      "step": 10796
+    },
+    {
+      "epoch": 1.749351911860013,
+      "grad_norm": 0.7808043956756592,
+      "learning_rate": 1.953020281981241e-06,
+      "loss": 0.0931,
+      "step": 10797
+    },
+    {
+      "epoch": 1.7495139338950096,
+      "grad_norm": 0.7637292146682739,
+      "learning_rate": 1.9525935814539e-06,
+      "loss": 0.0983,
+      "step": 10798
+    },
+    {
+      "epoch": 1.7496759559300066,
+      "grad_norm": 0.7604991793632507,
+      "learning_rate": 1.952166897675873e-06,
+      "loss": 0.0898,
+      "step": 10799
+    },
+    {
+      "epoch": 1.7498379779650033,
+      "grad_norm": 0.7365829944610596,
+      "learning_rate": 1.951740230660212e-06,
+      "loss": 0.0953,
+      "step": 10800
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8294869065284729,
+      "learning_rate": 1.951313580419974e-06,
+      "loss": 0.1092,
+      "step": 10801
+    },
+    {
+      "epoch": 1.7501620220349967,
+      "grad_norm": 0.8134753704071045,
+      "learning_rate": 1.950886946968212e-06,
+      "loss": 0.1008,
+      "step": 10802
+    },
+    {
+      "epoch": 1.7503240440699934,
+      "grad_norm": 0.967177152633667,
+      "learning_rate": 1.9504603303179805e-06,
+      "loss": 0.1083,
+      "step": 10803
+    },
+    {
+      "epoch": 1.7504860661049904,
+      "grad_norm": 0.8222710490226746,
+      "learning_rate": 1.9500337304823333e-06,
+      "loss": 0.0935,
+      "step": 10804
+    },
+    {
+      "epoch": 1.750648088139987,
+      "grad_norm": 0.7220136523246765,
+      "learning_rate": 1.949607147474323e-06,
+      "loss": 0.0931,
+      "step": 10805
+    },
+    {
+      "epoch": 1.7508101101749838,
+      "grad_norm": 0.6608802676200867,
+      "learning_rate": 1.9491805813070025e-06,
+      "loss": 0.0843,
+      "step": 10806
+    },
+    {
+      "epoch": 1.7509721322099807,
+      "grad_norm": 0.816785454750061,
+      "learning_rate": 1.948754031993422e-06,
+      "loss": 0.0995,
+      "step": 10807
+    },
+    {
+      "epoch": 1.7511341542449772,
+      "grad_norm": 0.8732054829597473,
+      "learning_rate": 1.948327499546635e-06,
+      "loss": 0.1105,
+      "step": 10808
+    },
+    {
+      "epoch": 1.7512961762799741,
+      "grad_norm": 0.9379077553749084,
+      "learning_rate": 1.9479009839796913e-06,
+      "loss": 0.1121,
+      "step": 10809
+    },
+    {
+      "epoch": 1.7514581983149708,
+      "grad_norm": 0.8605905771255493,
+      "learning_rate": 1.947474485305642e-06,
+      "loss": 0.0981,
+      "step": 10810
+    },
+    {
+      "epoch": 1.7516202203499676,
+      "grad_norm": 0.8368394374847412,
+      "learning_rate": 1.947048003537536e-06,
+      "loss": 0.1033,
+      "step": 10811
+    },
+    {
+      "epoch": 1.7517822423849645,
+      "grad_norm": 0.7215535640716553,
+      "learning_rate": 1.9466215386884223e-06,
+      "loss": 0.0826,
+      "step": 10812
+    },
+    {
+      "epoch": 1.751944264419961,
+      "grad_norm": 0.8332518339157104,
+      "learning_rate": 1.9461950907713517e-06,
+      "loss": 0.1041,
+      "step": 10813
+    },
+    {
+      "epoch": 1.752106286454958,
+      "grad_norm": 0.9338610768318176,
+      "learning_rate": 1.9457686597993704e-06,
+      "loss": 0.1167,
+      "step": 10814
+    },
+    {
+      "epoch": 1.7522683084899546,
+      "grad_norm": 0.8345838785171509,
+      "learning_rate": 1.9453422457855274e-06,
+      "loss": 0.1062,
+      "step": 10815
+    },
+    {
+      "epoch": 1.7524303305249513,
+      "grad_norm": 0.7421085834503174,
+      "learning_rate": 1.9449158487428688e-06,
+      "loss": 0.0859,
+      "step": 10816
+    },
+    {
+      "epoch": 1.7525923525599483,
+      "grad_norm": 0.9578739404678345,
+      "learning_rate": 1.9444894686844417e-06,
+      "loss": 0.1091,
+      "step": 10817
+    },
+    {
+      "epoch": 1.7527543745949448,
+      "grad_norm": 0.819701075553894,
+      "learning_rate": 1.9440631056232926e-06,
+      "loss": 0.1051,
+      "step": 10818
+    },
+    {
+      "epoch": 1.7529163966299417,
+      "grad_norm": 0.8500429391860962,
+      "learning_rate": 1.943636759572468e-06,
+      "loss": 0.1019,
+      "step": 10819
+    },
+    {
+      "epoch": 1.7530784186649384,
+      "grad_norm": 0.8731544017791748,
+      "learning_rate": 1.9432104305450117e-06,
+      "loss": 0.1023,
+      "step": 10820
+    },
+    {
+      "epoch": 1.7532404406999351,
+      "grad_norm": 0.8563147783279419,
+      "learning_rate": 1.9427841185539693e-06,
+      "loss": 0.1049,
+      "step": 10821
+    },
+    {
+      "epoch": 1.753402462734932,
+      "grad_norm": 0.8676608800888062,
+      "learning_rate": 1.942357823612383e-06,
+      "loss": 0.1079,
+      "step": 10822
+    },
+    {
+      "epoch": 1.7535644847699285,
+      "grad_norm": 0.870190441608429,
+      "learning_rate": 1.941931545733299e-06,
+      "loss": 0.1082,
+      "step": 10823
+    },
+    {
+      "epoch": 1.7537265068049255,
+      "grad_norm": 0.8860259652137756,
+      "learning_rate": 1.9415052849297585e-06,
+      "loss": 0.1054,
+      "step": 10824
+    },
+    {
+      "epoch": 1.7538885288399222,
+      "grad_norm": 1.0227184295654297,
+      "learning_rate": 1.941079041214806e-06,
+      "loss": 0.1208,
+      "step": 10825
+    },
+    {
+      "epoch": 1.754050550874919,
+      "grad_norm": 0.787338137626648,
+      "learning_rate": 1.9406528146014815e-06,
+      "loss": 0.1075,
+      "step": 10826
+    },
+    {
+      "epoch": 1.7542125729099158,
+      "grad_norm": 0.9310184121131897,
+      "learning_rate": 1.9402266051028263e-06,
+      "loss": 0.1064,
+      "step": 10827
+    },
+    {
+      "epoch": 1.7543745949449125,
+      "grad_norm": 0.8110270500183105,
+      "learning_rate": 1.939800412731884e-06,
+      "loss": 0.1018,
+      "step": 10828
+    },
+    {
+      "epoch": 1.7545366169799093,
+      "grad_norm": 0.8377658128738403,
+      "learning_rate": 1.9393742375016926e-06,
+      "loss": 0.1095,
+      "step": 10829
+    },
+    {
+      "epoch": 1.7546986390149062,
+      "grad_norm": 0.8631830811500549,
+      "learning_rate": 1.9389480794252933e-06,
+      "loss": 0.1097,
+      "step": 10830
+    },
+    {
+      "epoch": 1.7548606610499027,
+      "grad_norm": 0.76842200756073,
+      "learning_rate": 1.9385219385157244e-06,
+      "loss": 0.0994,
+      "step": 10831
+    },
+    {
+      "epoch": 1.7550226830848996,
+      "grad_norm": 0.9799662828445435,
+      "learning_rate": 1.9380958147860254e-06,
+      "loss": 0.1145,
+      "step": 10832
+    },
+    {
+      "epoch": 1.7551847051198963,
+      "grad_norm": 0.7072003483772278,
+      "learning_rate": 1.937669708249235e-06,
+      "loss": 0.0899,
+      "step": 10833
+    },
+    {
+      "epoch": 1.755346727154893,
+      "grad_norm": 0.8198095560073853,
+      "learning_rate": 1.937243618918391e-06,
+      "loss": 0.1064,
+      "step": 10834
+    },
+    {
+      "epoch": 1.75550874918989,
+      "grad_norm": 0.8468987941741943,
+      "learning_rate": 1.9368175468065305e-06,
+      "loss": 0.1043,
+      "step": 10835
+    },
+    {
+      "epoch": 1.7556707712248865,
+      "grad_norm": 0.9012013077735901,
+      "learning_rate": 1.936391491926689e-06,
+      "loss": 0.107,
+      "step": 10836
+    },
+    {
+      "epoch": 1.7558327932598834,
+      "grad_norm": 0.7513500452041626,
+      "learning_rate": 1.9359654542919054e-06,
+      "loss": 0.0948,
+      "step": 10837
+    },
+    {
+      "epoch": 1.75599481529488,
+      "grad_norm": 0.7412696480751038,
+      "learning_rate": 1.9355394339152133e-06,
+      "loss": 0.0926,
+      "step": 10838
+    },
+    {
+      "epoch": 1.7561568373298768,
+      "grad_norm": 0.8077243566513062,
+      "learning_rate": 1.9351134308096493e-06,
+      "loss": 0.0994,
+      "step": 10839
+    },
+    {
+      "epoch": 1.7563188593648738,
+      "grad_norm": 0.7305166721343994,
+      "learning_rate": 1.9346874449882465e-06,
+      "loss": 0.0951,
+      "step": 10840
+    },
+    {
+      "epoch": 1.7564808813998702,
+      "grad_norm": 0.8272958993911743,
+      "learning_rate": 1.9342614764640392e-06,
+      "loss": 0.1083,
+      "step": 10841
+    },
+    {
+      "epoch": 1.7566429034348672,
+      "grad_norm": 0.7705273628234863,
+      "learning_rate": 1.9338355252500624e-06,
+      "loss": 0.095,
+      "step": 10842
+    },
+    {
+      "epoch": 1.7568049254698639,
+      "grad_norm": 0.754991352558136,
+      "learning_rate": 1.933409591359349e-06,
+      "loss": 0.1042,
+      "step": 10843
+    },
+    {
+      "epoch": 1.7569669475048606,
+      "grad_norm": 0.8502324819564819,
+      "learning_rate": 1.93298367480493e-06,
+      "loss": 0.0999,
+      "step": 10844
+    },
+    {
+      "epoch": 1.7571289695398575,
+      "grad_norm": 0.7320595383644104,
+      "learning_rate": 1.9325577755998397e-06,
+      "loss": 0.0893,
+      "step": 10845
+    },
+    {
+      "epoch": 1.757290991574854,
+      "grad_norm": 0.7977878451347351,
+      "learning_rate": 1.932131893757107e-06,
+      "loss": 0.0983,
+      "step": 10846
+    },
+    {
+      "epoch": 1.757453013609851,
+      "grad_norm": 0.7108481526374817,
+      "learning_rate": 1.9317060292897643e-06,
+      "loss": 0.0906,
+      "step": 10847
+    },
+    {
+      "epoch": 1.7576150356448477,
+      "grad_norm": 0.7112817764282227,
+      "learning_rate": 1.9312801822108425e-06,
+      "loss": 0.0883,
+      "step": 10848
+    },
+    {
+      "epoch": 1.7577770576798444,
+      "grad_norm": 0.7918272614479065,
+      "learning_rate": 1.9308543525333707e-06,
+      "loss": 0.0957,
+      "step": 10849
+    },
+    {
+      "epoch": 1.7579390797148413,
+      "grad_norm": 0.8627058267593384,
+      "learning_rate": 1.9304285402703775e-06,
+      "loss": 0.1079,
+      "step": 10850
+    },
+    {
+      "epoch": 1.758101101749838,
+      "grad_norm": 0.7404596209526062,
+      "learning_rate": 1.9300027454348932e-06,
+      "loss": 0.0908,
+      "step": 10851
+    },
+    {
+      "epoch": 1.7582631237848347,
+      "grad_norm": 0.879859447479248,
+      "learning_rate": 1.929576968039946e-06,
+      "loss": 0.1072,
+      "step": 10852
+    },
+    {
+      "epoch": 1.7584251458198314,
+      "grad_norm": 0.8356106281280518,
+      "learning_rate": 1.9291512080985626e-06,
+      "loss": 0.105,
+      "step": 10853
+    },
+    {
+      "epoch": 1.7585871678548282,
+      "grad_norm": 0.8642502427101135,
+      "learning_rate": 1.928725465623772e-06,
+      "loss": 0.1099,
+      "step": 10854
+    },
+    {
+      "epoch": 1.758749189889825,
+      "grad_norm": 0.9117339849472046,
+      "learning_rate": 1.928299740628598e-06,
+      "loss": 0.113,
+      "step": 10855
+    },
+    {
+      "epoch": 1.7589112119248218,
+      "grad_norm": 0.8020870685577393,
+      "learning_rate": 1.927874033126069e-06,
+      "loss": 0.1042,
+      "step": 10856
+    },
+    {
+      "epoch": 1.7590732339598185,
+      "grad_norm": 0.8990389704704285,
+      "learning_rate": 1.9274483431292107e-06,
+      "loss": 0.109,
+      "step": 10857
+    },
+    {
+      "epoch": 1.7592352559948155,
+      "grad_norm": 0.9079889059066772,
+      "learning_rate": 1.927022670651047e-06,
+      "loss": 0.1068,
+      "step": 10858
+    },
+    {
+      "epoch": 1.759397278029812,
+      "grad_norm": 0.928896427154541,
+      "learning_rate": 1.9265970157046037e-06,
+      "loss": 0.1076,
+      "step": 10859
+    },
+    {
+      "epoch": 1.7595593000648089,
+      "grad_norm": 0.936068594455719,
+      "learning_rate": 1.9261713783029024e-06,
+      "loss": 0.1145,
+      "step": 10860
+    },
+    {
+      "epoch": 1.7597213220998056,
+      "grad_norm": 0.791902482509613,
+      "learning_rate": 1.9257457584589697e-06,
+      "loss": 0.1006,
+      "step": 10861
+    },
+    {
+      "epoch": 1.7598833441348023,
+      "grad_norm": 0.7125447988510132,
+      "learning_rate": 1.9253201561858266e-06,
+      "loss": 0.0929,
+      "step": 10862
+    },
+    {
+      "epoch": 1.7600453661697992,
+      "grad_norm": 0.7819201946258545,
+      "learning_rate": 1.9248945714964967e-06,
+      "loss": 0.0972,
+      "step": 10863
+    },
+    {
+      "epoch": 1.7602073882047957,
+      "grad_norm": 0.6780325174331665,
+      "learning_rate": 1.924469004404001e-06,
+      "loss": 0.0861,
+      "step": 10864
+    },
+    {
+      "epoch": 1.7603694102397927,
+      "grad_norm": 0.755989670753479,
+      "learning_rate": 1.92404345492136e-06,
+      "loss": 0.0954,
+      "step": 10865
+    },
+    {
+      "epoch": 1.7605314322747894,
+      "grad_norm": 0.6931602358818054,
+      "learning_rate": 1.9236179230615967e-06,
+      "loss": 0.0822,
+      "step": 10866
+    },
+    {
+      "epoch": 1.760693454309786,
+      "grad_norm": 0.7863826155662537,
+      "learning_rate": 1.9231924088377296e-06,
+      "loss": 0.0948,
+      "step": 10867
+    },
+    {
+      "epoch": 1.760855476344783,
+      "grad_norm": 0.6952497959136963,
+      "learning_rate": 1.922766912262779e-06,
+      "loss": 0.0819,
+      "step": 10868
+    },
+    {
+      "epoch": 1.7610174983797795,
+      "grad_norm": 0.6934300065040588,
+      "learning_rate": 1.922341433349764e-06,
+      "loss": 0.0829,
+      "step": 10869
+    },
+    {
+      "epoch": 1.7611795204147764,
+      "grad_norm": 0.7480776906013489,
+      "learning_rate": 1.921915972111703e-06,
+      "loss": 0.0958,
+      "step": 10870
+    },
+    {
+      "epoch": 1.7613415424497731,
+      "grad_norm": 0.8322224020957947,
+      "learning_rate": 1.9214905285616147e-06,
+      "loss": 0.1,
+      "step": 10871
+    },
+    {
+      "epoch": 1.7615035644847699,
+      "grad_norm": 0.8011393547058105,
+      "learning_rate": 1.9210651027125164e-06,
+      "loss": 0.1024,
+      "step": 10872
+    },
+    {
+      "epoch": 1.7616655865197668,
+      "grad_norm": 0.8309357166290283,
+      "learning_rate": 1.9206396945774246e-06,
+      "loss": 0.1033,
+      "step": 10873
+    },
+    {
+      "epoch": 1.7618276085547635,
+      "grad_norm": 0.7737796306610107,
+      "learning_rate": 1.9202143041693554e-06,
+      "loss": 0.0988,
+      "step": 10874
+    },
+    {
+      "epoch": 1.7619896305897602,
+      "grad_norm": 0.7730358242988586,
+      "learning_rate": 1.919788931501327e-06,
+      "loss": 0.0998,
+      "step": 10875
+    },
+    {
+      "epoch": 1.762151652624757,
+      "grad_norm": 0.8785960078239441,
+      "learning_rate": 1.919363576586352e-06,
+      "loss": 0.1073,
+      "step": 10876
+    },
+    {
+      "epoch": 1.7623136746597536,
+      "grad_norm": 0.8502126932144165,
+      "learning_rate": 1.918938239437447e-06,
+      "loss": 0.1092,
+      "step": 10877
+    },
+    {
+      "epoch": 1.7624756966947506,
+      "grad_norm": 0.7489198446273804,
+      "learning_rate": 1.918512920067626e-06,
+      "loss": 0.0944,
+      "step": 10878
+    },
+    {
+      "epoch": 1.7626377187297473,
+      "grad_norm": 0.7126352787017822,
+      "learning_rate": 1.9180876184899015e-06,
+      "loss": 0.0952,
+      "step": 10879
+    },
+    {
+      "epoch": 1.762799740764744,
+      "grad_norm": 0.8675928711891174,
+      "learning_rate": 1.9176623347172885e-06,
+      "loss": 0.1067,
+      "step": 10880
+    },
+    {
+      "epoch": 1.762961762799741,
+      "grad_norm": 0.820986807346344,
+      "learning_rate": 1.9172370687627987e-06,
+      "loss": 0.0972,
+      "step": 10881
+    },
+    {
+      "epoch": 1.7631237848347374,
+      "grad_norm": 0.8926191329956055,
+      "learning_rate": 1.9168118206394443e-06,
+      "loss": 0.1056,
+      "step": 10882
+    },
+    {
+      "epoch": 1.7632858068697344,
+      "grad_norm": 0.8152083158493042,
+      "learning_rate": 1.9163865903602374e-06,
+      "loss": 0.1041,
+      "step": 10883
+    },
+    {
+      "epoch": 1.763447828904731,
+      "grad_norm": 0.837230920791626,
+      "learning_rate": 1.915961377938187e-06,
+      "loss": 0.1074,
+      "step": 10884
+    },
+    {
+      "epoch": 1.7636098509397278,
+      "grad_norm": 0.8547905683517456,
+      "learning_rate": 1.915536183386306e-06,
+      "loss": 0.1056,
+      "step": 10885
+    },
+    {
+      "epoch": 1.7637718729747247,
+      "grad_norm": 0.8433102965354919,
+      "learning_rate": 1.9151110067176038e-06,
+      "loss": 0.1013,
+      "step": 10886
+    },
+    {
+      "epoch": 1.7639338950097212,
+      "grad_norm": 0.8947946429252625,
+      "learning_rate": 1.9146858479450894e-06,
+      "loss": 0.1166,
+      "step": 10887
+    },
+    {
+      "epoch": 1.7640959170447181,
+      "grad_norm": 0.9684112668037415,
+      "learning_rate": 1.914260707081771e-06,
+      "loss": 0.1186,
+      "step": 10888
+    },
+    {
+      "epoch": 1.7642579390797148,
+      "grad_norm": 0.7762362360954285,
+      "learning_rate": 1.913835584140657e-06,
+      "loss": 0.0967,
+      "step": 10889
+    },
+    {
+      "epoch": 1.7644199611147116,
+      "grad_norm": 0.7626681923866272,
+      "learning_rate": 1.913410479134757e-06,
+      "loss": 0.0881,
+      "step": 10890
+    },
+    {
+      "epoch": 1.7645819831497085,
+      "grad_norm": 0.7825966477394104,
+      "learning_rate": 1.9129853920770763e-06,
+      "loss": 0.0941,
+      "step": 10891
+    },
+    {
+      "epoch": 1.764744005184705,
+      "grad_norm": 0.8394178748130798,
+      "learning_rate": 1.9125603229806223e-06,
+      "loss": 0.1112,
+      "step": 10892
+    },
+    {
+      "epoch": 1.764906027219702,
+      "grad_norm": 0.7935326099395752,
+      "learning_rate": 1.9121352718584006e-06,
+      "loss": 0.1012,
+      "step": 10893
+    },
+    {
+      "epoch": 1.7650680492546986,
+      "grad_norm": 0.8410793542861938,
+      "learning_rate": 1.9117102387234165e-06,
+      "loss": 0.1077,
+      "step": 10894
+    },
+    {
+      "epoch": 1.7652300712896953,
+      "grad_norm": 0.7813910245895386,
+      "learning_rate": 1.9112852235886757e-06,
+      "loss": 0.0947,
+      "step": 10895
+    },
+    {
+      "epoch": 1.7653920933246923,
+      "grad_norm": 0.81944340467453,
+      "learning_rate": 1.910860226467183e-06,
+      "loss": 0.0989,
+      "step": 10896
+    },
+    {
+      "epoch": 1.7655541153596888,
+      "grad_norm": 0.7879267930984497,
+      "learning_rate": 1.910435247371941e-06,
+      "loss": 0.1065,
+      "step": 10897
+    },
+    {
+      "epoch": 1.7657161373946857,
+      "grad_norm": 0.8651462197303772,
+      "learning_rate": 1.910010286315953e-06,
+      "loss": 0.1011,
+      "step": 10898
+    },
+    {
+      "epoch": 1.7658781594296824,
+      "grad_norm": 0.7657998204231262,
+      "learning_rate": 1.909585343312224e-06,
+      "loss": 0.0925,
+      "step": 10899
+    },
+    {
+      "epoch": 1.7660401814646791,
+      "grad_norm": 0.652816891670227,
+      "learning_rate": 1.9091604183737546e-06,
+      "loss": 0.0832,
+      "step": 10900
+    },
+    {
+      "epoch": 1.766202203499676,
+      "grad_norm": 0.8381374478340149,
+      "learning_rate": 1.9087355115135465e-06,
+      "loss": 0.1008,
+      "step": 10901
+    },
+    {
+      "epoch": 1.7663642255346728,
+      "grad_norm": 0.7962815761566162,
+      "learning_rate": 1.9083106227446e-06,
+      "loss": 0.0991,
+      "step": 10902
+    },
+    {
+      "epoch": 1.7665262475696695,
+      "grad_norm": 0.7816174030303955,
+      "learning_rate": 1.9078857520799167e-06,
+      "loss": 0.098,
+      "step": 10903
+    },
+    {
+      "epoch": 1.7666882696046662,
+      "grad_norm": 0.8635630011558533,
+      "learning_rate": 1.907460899532497e-06,
+      "loss": 0.0986,
+      "step": 10904
+    },
+    {
+      "epoch": 1.766850291639663,
+      "grad_norm": 0.8344462513923645,
+      "learning_rate": 1.9070360651153402e-06,
+      "loss": 0.094,
+      "step": 10905
+    },
+    {
+      "epoch": 1.7670123136746598,
+      "grad_norm": 0.6713218092918396,
+      "learning_rate": 1.9066112488414445e-06,
+      "loss": 0.0832,
+      "step": 10906
+    },
+    {
+      "epoch": 1.7671743357096565,
+      "grad_norm": 0.8062579035758972,
+      "learning_rate": 1.906186450723809e-06,
+      "loss": 0.104,
+      "step": 10907
+    },
+    {
+      "epoch": 1.7673363577446533,
+      "grad_norm": 0.8023903965950012,
+      "learning_rate": 1.90576167077543e-06,
+      "loss": 0.091,
+      "step": 10908
+    },
+    {
+      "epoch": 1.7674983797796502,
+      "grad_norm": 0.739756166934967,
+      "learning_rate": 1.9053369090093065e-06,
+      "loss": 0.0819,
+      "step": 10909
+    },
+    {
+      "epoch": 1.7676604018146467,
+      "grad_norm": 0.9430824518203735,
+      "learning_rate": 1.904912165438435e-06,
+      "loss": 0.1085,
+      "step": 10910
+    },
+    {
+      "epoch": 1.7678224238496436,
+      "grad_norm": 1.0915939807891846,
+      "learning_rate": 1.9044874400758106e-06,
+      "loss": 0.12,
+      "step": 10911
+    },
+    {
+      "epoch": 1.7679844458846403,
+      "grad_norm": 0.6942973136901855,
+      "learning_rate": 1.9040627329344296e-06,
+      "loss": 0.0896,
+      "step": 10912
+    },
+    {
+      "epoch": 1.768146467919637,
+      "grad_norm": 0.8167296648025513,
+      "learning_rate": 1.9036380440272861e-06,
+      "loss": 0.0949,
+      "step": 10913
+    },
+    {
+      "epoch": 1.768308489954634,
+      "grad_norm": 0.8740625381469727,
+      "learning_rate": 1.9032133733673764e-06,
+      "loss": 0.0991,
+      "step": 10914
+    },
+    {
+      "epoch": 1.7684705119896305,
+      "grad_norm": 0.9350792169570923,
+      "learning_rate": 1.9027887209676925e-06,
+      "loss": 0.1148,
+      "step": 10915
+    },
+    {
+      "epoch": 1.7686325340246274,
+      "grad_norm": 0.7898386716842651,
+      "learning_rate": 1.9023640868412297e-06,
+      "loss": 0.0929,
+      "step": 10916
+    },
+    {
+      "epoch": 1.768794556059624,
+      "grad_norm": 0.7293332815170288,
+      "learning_rate": 1.901939471000978e-06,
+      "loss": 0.0902,
+      "step": 10917
+    },
+    {
+      "epoch": 1.7689565780946208,
+      "grad_norm": 0.8032740950584412,
+      "learning_rate": 1.9015148734599317e-06,
+      "loss": 0.097,
+      "step": 10918
+    },
+    {
+      "epoch": 1.7691186001296177,
+      "grad_norm": 0.7515488266944885,
+      "learning_rate": 1.9010902942310827e-06,
+      "loss": 0.1018,
+      "step": 10919
+    },
+    {
+      "epoch": 1.7692806221646142,
+      "grad_norm": 0.7839230895042419,
+      "learning_rate": 1.900665733327421e-06,
+      "loss": 0.0935,
+      "step": 10920
+    },
+    {
+      "epoch": 1.7694426441996112,
+      "grad_norm": 0.8209660649299622,
+      "learning_rate": 1.9002411907619372e-06,
+      "loss": 0.1026,
+      "step": 10921
+    },
+    {
+      "epoch": 1.7696046662346079,
+      "grad_norm": 0.7910726070404053,
+      "learning_rate": 1.899816666547621e-06,
+      "loss": 0.0966,
+      "step": 10922
+    },
+    {
+      "epoch": 1.7697666882696046,
+      "grad_norm": 0.8440234065055847,
+      "learning_rate": 1.8993921606974636e-06,
+      "loss": 0.1055,
+      "step": 10923
+    },
+    {
+      "epoch": 1.7699287103046015,
+      "grad_norm": 0.9682288765907288,
+      "learning_rate": 1.8989676732244522e-06,
+      "loss": 0.1216,
+      "step": 10924
+    },
+    {
+      "epoch": 1.7700907323395982,
+      "grad_norm": 0.8367732763290405,
+      "learning_rate": 1.8985432041415758e-06,
+      "loss": 0.1072,
+      "step": 10925
+    },
+    {
+      "epoch": 1.770252754374595,
+      "grad_norm": 1.015958547592163,
+      "learning_rate": 1.8981187534618217e-06,
+      "loss": 0.1244,
+      "step": 10926
+    },
+    {
+      "epoch": 1.7704147764095917,
+      "grad_norm": 1.0870450735092163,
+      "learning_rate": 1.8976943211981764e-06,
+      "loss": 0.1068,
+      "step": 10927
+    },
+    {
+      "epoch": 1.7705767984445884,
+      "grad_norm": 0.8047299981117249,
+      "learning_rate": 1.8972699073636283e-06,
+      "loss": 0.1014,
+      "step": 10928
+    },
+    {
+      "epoch": 1.7707388204795853,
+      "grad_norm": 0.7418354153633118,
+      "learning_rate": 1.896845511971162e-06,
+      "loss": 0.0959,
+      "step": 10929
+    },
+    {
+      "epoch": 1.770900842514582,
+      "grad_norm": 0.6954073905944824,
+      "learning_rate": 1.8964211350337637e-06,
+      "loss": 0.0887,
+      "step": 10930
+    },
+    {
+      "epoch": 1.7710628645495787,
+      "grad_norm": 0.8665586113929749,
+      "learning_rate": 1.8959967765644182e-06,
+      "loss": 0.1077,
+      "step": 10931
+    },
+    {
+      "epoch": 1.7712248865845757,
+      "grad_norm": 0.6979767084121704,
+      "learning_rate": 1.895572436576109e-06,
+      "loss": 0.0904,
+      "step": 10932
+    },
+    {
+      "epoch": 1.7713869086195722,
+      "grad_norm": 0.8404514193534851,
+      "learning_rate": 1.8951481150818206e-06,
+      "loss": 0.1054,
+      "step": 10933
+    },
+    {
+      "epoch": 1.771548930654569,
+      "grad_norm": 0.7205252647399902,
+      "learning_rate": 1.8947238120945372e-06,
+      "loss": 0.0875,
+      "step": 10934
+    },
+    {
+      "epoch": 1.7717109526895658,
+      "grad_norm": 0.8082873821258545,
+      "learning_rate": 1.8942995276272396e-06,
+      "loss": 0.0989,
+      "step": 10935
+    },
+    {
+      "epoch": 1.7718729747245625,
+      "grad_norm": 0.9559950828552246,
+      "learning_rate": 1.8938752616929112e-06,
+      "loss": 0.1251,
+      "step": 10936
+    },
+    {
+      "epoch": 1.7720349967595594,
+      "grad_norm": 0.7924472093582153,
+      "learning_rate": 1.8934510143045316e-06,
+      "loss": 0.097,
+      "step": 10937
+    },
+    {
+      "epoch": 1.772197018794556,
+      "grad_norm": 0.8172940015792847,
+      "learning_rate": 1.8930267854750845e-06,
+      "loss": 0.1057,
+      "step": 10938
+    },
+    {
+      "epoch": 1.7723590408295529,
+      "grad_norm": 0.8141445517539978,
+      "learning_rate": 1.8926025752175486e-06,
+      "loss": 0.1122,
+      "step": 10939
+    },
+    {
+      "epoch": 1.7725210628645496,
+      "grad_norm": 0.8678747415542603,
+      "learning_rate": 1.8921783835449042e-06,
+      "loss": 0.1103,
+      "step": 10940
+    },
+    {
+      "epoch": 1.7726830848995463,
+      "grad_norm": 0.6963508725166321,
+      "learning_rate": 1.8917542104701297e-06,
+      "loss": 0.0906,
+      "step": 10941
+    },
+    {
+      "epoch": 1.7728451069345432,
+      "grad_norm": 0.7853334546089172,
+      "learning_rate": 1.8913300560062047e-06,
+      "loss": 0.0972,
+      "step": 10942
+    },
+    {
+      "epoch": 1.7730071289695397,
+      "grad_norm": 0.8051917552947998,
+      "learning_rate": 1.8909059201661079e-06,
+      "loss": 0.0975,
+      "step": 10943
+    },
+    {
+      "epoch": 1.7731691510045366,
+      "grad_norm": 0.7873483896255493,
+      "learning_rate": 1.890481802962815e-06,
+      "loss": 0.1039,
+      "step": 10944
+    },
+    {
+      "epoch": 1.7733311730395334,
+      "grad_norm": 0.7235662937164307,
+      "learning_rate": 1.8900577044093045e-06,
+      "loss": 0.0817,
+      "step": 10945
+    },
+    {
+      "epoch": 1.77349319507453,
+      "grad_norm": 0.793947696685791,
+      "learning_rate": 1.889633624518551e-06,
+      "loss": 0.0988,
+      "step": 10946
+    },
+    {
+      "epoch": 1.773655217109527,
+      "grad_norm": 0.8600531816482544,
+      "learning_rate": 1.889209563303533e-06,
+      "loss": 0.1138,
+      "step": 10947
+    },
+    {
+      "epoch": 1.7738172391445235,
+      "grad_norm": 0.7487325072288513,
+      "learning_rate": 1.8887855207772235e-06,
+      "loss": 0.0858,
+      "step": 10948
+    },
+    {
+      "epoch": 1.7739792611795204,
+      "grad_norm": 0.8943690061569214,
+      "learning_rate": 1.8883614969525987e-06,
+      "loss": 0.1172,
+      "step": 10949
+    },
+    {
+      "epoch": 1.7741412832145171,
+      "grad_norm": 0.7959859371185303,
+      "learning_rate": 1.8879374918426312e-06,
+      "loss": 0.1,
+      "step": 10950
+    },
+    {
+      "epoch": 1.7743033052495139,
+      "grad_norm": 0.8319025635719299,
+      "learning_rate": 1.887513505460295e-06,
+      "loss": 0.1021,
+      "step": 10951
+    },
+    {
+      "epoch": 1.7744653272845108,
+      "grad_norm": 0.7502698302268982,
+      "learning_rate": 1.8870895378185643e-06,
+      "loss": 0.092,
+      "step": 10952
+    },
+    {
+      "epoch": 1.7746273493195075,
+      "grad_norm": 0.8591358661651611,
+      "learning_rate": 1.88666558893041e-06,
+      "loss": 0.1105,
+      "step": 10953
+    },
+    {
+      "epoch": 1.7747893713545042,
+      "grad_norm": 0.7255021333694458,
+      "learning_rate": 1.886241658808805e-06,
+      "loss": 0.0905,
+      "step": 10954
+    },
+    {
+      "epoch": 1.774951393389501,
+      "grad_norm": 0.7662566304206848,
+      "learning_rate": 1.8858177474667195e-06,
+      "loss": 0.0917,
+      "step": 10955
+    },
+    {
+      "epoch": 1.7751134154244976,
+      "grad_norm": 0.7199138402938843,
+      "learning_rate": 1.8853938549171242e-06,
+      "loss": 0.0868,
+      "step": 10956
+    },
+    {
+      "epoch": 1.7752754374594946,
+      "grad_norm": 0.8491278886795044,
+      "learning_rate": 1.88496998117299e-06,
+      "loss": 0.1025,
+      "step": 10957
+    },
+    {
+      "epoch": 1.7754374594944913,
+      "grad_norm": 0.9104598164558411,
+      "learning_rate": 1.8845461262472863e-06,
+      "loss": 0.108,
+      "step": 10958
+    },
+    {
+      "epoch": 1.775599481529488,
+      "grad_norm": 0.7580164670944214,
+      "learning_rate": 1.8841222901529816e-06,
+      "loss": 0.0952,
+      "step": 10959
+    },
+    {
+      "epoch": 1.775761503564485,
+      "grad_norm": 0.8028944730758667,
+      "learning_rate": 1.883698472903045e-06,
+      "loss": 0.1113,
+      "step": 10960
+    },
+    {
+      "epoch": 1.7759235255994814,
+      "grad_norm": 0.7645183801651001,
+      "learning_rate": 1.8832746745104425e-06,
+      "loss": 0.1035,
+      "step": 10961
+    },
+    {
+      "epoch": 1.7760855476344783,
+      "grad_norm": 0.7198721170425415,
+      "learning_rate": 1.882850894988143e-06,
+      "loss": 0.0931,
+      "step": 10962
+    },
+    {
+      "epoch": 1.776247569669475,
+      "grad_norm": 0.7722495794296265,
+      "learning_rate": 1.882427134349113e-06,
+      "loss": 0.0999,
+      "step": 10963
+    },
+    {
+      "epoch": 1.7764095917044718,
+      "grad_norm": 0.8241525292396545,
+      "learning_rate": 1.882003392606318e-06,
+      "loss": 0.1011,
+      "step": 10964
+    },
+    {
+      "epoch": 1.7765716137394687,
+      "grad_norm": 0.7520769834518433,
+      "learning_rate": 1.881579669772723e-06,
+      "loss": 0.0956,
+      "step": 10965
+    },
+    {
+      "epoch": 1.7767336357744652,
+      "grad_norm": 0.7766212224960327,
+      "learning_rate": 1.8811559658612941e-06,
+      "loss": 0.1007,
+      "step": 10966
+    },
+    {
+      "epoch": 1.7768956578094621,
+      "grad_norm": 0.7454885244369507,
+      "learning_rate": 1.8807322808849953e-06,
+      "loss": 0.0895,
+      "step": 10967
+    },
+    {
+      "epoch": 1.7770576798444588,
+      "grad_norm": 0.7523648142814636,
+      "learning_rate": 1.88030861485679e-06,
+      "loss": 0.0972,
+      "step": 10968
+    },
+    {
+      "epoch": 1.7772197018794555,
+      "grad_norm": 0.902167022228241,
+      "learning_rate": 1.879884967789642e-06,
+      "loss": 0.1076,
+      "step": 10969
+    },
+    {
+      "epoch": 1.7773817239144525,
+      "grad_norm": 0.9376851320266724,
+      "learning_rate": 1.879461339696512e-06,
+      "loss": 0.1125,
+      "step": 10970
+    },
+    {
+      "epoch": 1.777543745949449,
+      "grad_norm": 0.7125734090805054,
+      "learning_rate": 1.879037730590364e-06,
+      "loss": 0.0869,
+      "step": 10971
+    },
+    {
+      "epoch": 1.777705767984446,
+      "grad_norm": 0.9217870831489563,
+      "learning_rate": 1.8786141404841587e-06,
+      "loss": 0.1053,
+      "step": 10972
+    },
+    {
+      "epoch": 1.7778677900194426,
+      "grad_norm": 0.7570266723632812,
+      "learning_rate": 1.8781905693908575e-06,
+      "loss": 0.0966,
+      "step": 10973
+    },
+    {
+      "epoch": 1.7780298120544393,
+      "grad_norm": 0.8740879893302917,
+      "learning_rate": 1.8777670173234198e-06,
+      "loss": 0.1015,
+      "step": 10974
+    },
+    {
+      "epoch": 1.7781918340894363,
+      "grad_norm": 0.7980040907859802,
+      "learning_rate": 1.877343484294805e-06,
+      "loss": 0.0967,
+      "step": 10975
+    },
+    {
+      "epoch": 1.778353856124433,
+      "grad_norm": 0.8354352116584778,
+      "learning_rate": 1.8769199703179736e-06,
+      "loss": 0.1052,
+      "step": 10976
+    },
+    {
+      "epoch": 1.7785158781594297,
+      "grad_norm": 0.7894411683082581,
+      "learning_rate": 1.876496475405883e-06,
+      "loss": 0.0979,
+      "step": 10977
+    },
+    {
+      "epoch": 1.7786779001944264,
+      "grad_norm": 0.7108971476554871,
+      "learning_rate": 1.8760729995714916e-06,
+      "loss": 0.0876,
+      "step": 10978
+    },
+    {
+      "epoch": 1.778839922229423,
+      "grad_norm": 0.8221232891082764,
+      "learning_rate": 1.8756495428277562e-06,
+      "loss": 0.1034,
+      "step": 10979
+    },
+    {
+      "epoch": 1.77900194426442,
+      "grad_norm": 0.7995948195457458,
+      "learning_rate": 1.8752261051876337e-06,
+      "loss": 0.0979,
+      "step": 10980
+    },
+    {
+      "epoch": 1.7791639662994168,
+      "grad_norm": 0.8195783495903015,
+      "learning_rate": 1.8748026866640806e-06,
+      "loss": 0.104,
+      "step": 10981
+    },
+    {
+      "epoch": 1.7793259883344135,
+      "grad_norm": 0.8994091153144836,
+      "learning_rate": 1.8743792872700529e-06,
+      "loss": 0.1015,
+      "step": 10982
+    },
+    {
+      "epoch": 1.7794880103694104,
+      "grad_norm": 0.9404149055480957,
+      "learning_rate": 1.8739559070185045e-06,
+      "loss": 0.1078,
+      "step": 10983
+    },
+    {
+      "epoch": 1.779650032404407,
+      "grad_norm": 0.7274839878082275,
+      "learning_rate": 1.873532545922391e-06,
+      "loss": 0.0896,
+      "step": 10984
+    },
+    {
+      "epoch": 1.7798120544394038,
+      "grad_norm": 0.7796977758407593,
+      "learning_rate": 1.8731092039946646e-06,
+      "loss": 0.098,
+      "step": 10985
+    },
+    {
+      "epoch": 1.7799740764744005,
+      "grad_norm": 0.7981799840927124,
+      "learning_rate": 1.8726858812482798e-06,
+      "loss": 0.0993,
+      "step": 10986
+    },
+    {
+      "epoch": 1.7801360985093972,
+      "grad_norm": 0.8132720589637756,
+      "learning_rate": 1.8722625776961894e-06,
+      "loss": 0.1009,
+      "step": 10987
+    },
+    {
+      "epoch": 1.7802981205443942,
+      "grad_norm": 0.769990086555481,
+      "learning_rate": 1.871839293351345e-06,
+      "loss": 0.0948,
+      "step": 10988
+    },
+    {
+      "epoch": 1.7804601425793907,
+      "grad_norm": 0.7847530245780945,
+      "learning_rate": 1.8714160282266973e-06,
+      "loss": 0.0981,
+      "step": 10989
+    },
+    {
+      "epoch": 1.7806221646143876,
+      "grad_norm": 0.8644593954086304,
+      "learning_rate": 1.870992782335198e-06,
+      "loss": 0.0972,
+      "step": 10990
+    },
+    {
+      "epoch": 1.7807841866493843,
+      "grad_norm": 0.777860701084137,
+      "learning_rate": 1.8705695556897986e-06,
+      "loss": 0.0979,
+      "step": 10991
+    },
+    {
+      "epoch": 1.780946208684381,
+      "grad_norm": 0.9635716676712036,
+      "learning_rate": 1.8701463483034471e-06,
+      "loss": 0.1061,
+      "step": 10992
+    },
+    {
+      "epoch": 1.781108230719378,
+      "grad_norm": 0.7884843945503235,
+      "learning_rate": 1.8697231601890933e-06,
+      "loss": 0.0937,
+      "step": 10993
+    },
+    {
+      "epoch": 1.7812702527543745,
+      "grad_norm": 0.8972017168998718,
+      "learning_rate": 1.8692999913596846e-06,
+      "loss": 0.1078,
+      "step": 10994
+    },
+    {
+      "epoch": 1.7814322747893714,
+      "grad_norm": 0.7947750091552734,
+      "learning_rate": 1.8688768418281705e-06,
+      "loss": 0.0987,
+      "step": 10995
+    },
+    {
+      "epoch": 1.781594296824368,
+      "grad_norm": 1.048323154449463,
+      "learning_rate": 1.8684537116074983e-06,
+      "loss": 0.126,
+      "step": 10996
+    },
+    {
+      "epoch": 1.7817563188593648,
+      "grad_norm": 0.8322978615760803,
+      "learning_rate": 1.8680306007106136e-06,
+      "loss": 0.1015,
+      "step": 10997
+    },
+    {
+      "epoch": 1.7819183408943617,
+      "grad_norm": 0.8976532816886902,
+      "learning_rate": 1.8676075091504637e-06,
+      "loss": 0.1016,
+      "step": 10998
+    },
+    {
+      "epoch": 1.7820803629293582,
+      "grad_norm": 0.6236394643783569,
+      "learning_rate": 1.8671844369399922e-06,
+      "loss": 0.0794,
+      "step": 10999
+    },
+    {
+      "epoch": 1.7822423849643552,
+      "grad_norm": 0.8208524584770203,
+      "learning_rate": 1.866761384092147e-06,
+      "loss": 0.1006,
+      "step": 11000
+    },
+    {
+      "epoch": 1.7824044069993519,
+      "grad_norm": 0.8767183423042297,
+      "learning_rate": 1.8663383506198706e-06,
+      "loss": 0.0964,
+      "step": 11001
+    },
+    {
+      "epoch": 1.7825664290343486,
+      "grad_norm": 0.7991448044776917,
+      "learning_rate": 1.8659153365361076e-06,
+      "loss": 0.1013,
+      "step": 11002
+    },
+    {
+      "epoch": 1.7827284510693455,
+      "grad_norm": 0.8050661683082581,
+      "learning_rate": 1.8654923418538003e-06,
+      "loss": 0.0933,
+      "step": 11003
+    },
+    {
+      "epoch": 1.7828904731043422,
+      "grad_norm": 0.8594214916229248,
+      "learning_rate": 1.8650693665858916e-06,
+      "loss": 0.1113,
+      "step": 11004
+    },
+    {
+      "epoch": 1.783052495139339,
+      "grad_norm": 0.8001613616943359,
+      "learning_rate": 1.8646464107453247e-06,
+      "loss": 0.1044,
+      "step": 11005
+    },
+    {
+      "epoch": 1.7832145171743357,
+      "grad_norm": 0.865673840045929,
+      "learning_rate": 1.8642234743450394e-06,
+      "loss": 0.1033,
+      "step": 11006
+    },
+    {
+      "epoch": 1.7833765392093324,
+      "grad_norm": 0.720534086227417,
+      "learning_rate": 1.8638005573979776e-06,
+      "loss": 0.0874,
+      "step": 11007
+    },
+    {
+      "epoch": 1.7835385612443293,
+      "grad_norm": 0.7906041145324707,
+      "learning_rate": 1.8633776599170783e-06,
+      "loss": 0.1006,
+      "step": 11008
+    },
+    {
+      "epoch": 1.783700583279326,
+      "grad_norm": 0.8074170351028442,
+      "learning_rate": 1.8629547819152832e-06,
+      "loss": 0.1024,
+      "step": 11009
+    },
+    {
+      "epoch": 1.7838626053143227,
+      "grad_norm": 0.7554319500923157,
+      "learning_rate": 1.86253192340553e-06,
+      "loss": 0.0986,
+      "step": 11010
+    },
+    {
+      "epoch": 1.7840246273493197,
+      "grad_norm": 0.7377730011940002,
+      "learning_rate": 1.8621090844007572e-06,
+      "loss": 0.0984,
+      "step": 11011
+    },
+    {
+      "epoch": 1.7841866493843161,
+      "grad_norm": 0.9073734283447266,
+      "learning_rate": 1.8616862649139024e-06,
+      "loss": 0.1075,
+      "step": 11012
+    },
+    {
+      "epoch": 1.784348671419313,
+      "grad_norm": 0.8492515683174133,
+      "learning_rate": 1.861263464957903e-06,
+      "loss": 0.1033,
+      "step": 11013
+    },
+    {
+      "epoch": 1.7845106934543098,
+      "grad_norm": 0.7091047763824463,
+      "learning_rate": 1.8608406845456968e-06,
+      "loss": 0.0868,
+      "step": 11014
+    },
+    {
+      "epoch": 1.7846727154893065,
+      "grad_norm": 0.7672532796859741,
+      "learning_rate": 1.860417923690218e-06,
+      "loss": 0.0974,
+      "step": 11015
+    },
+    {
+      "epoch": 1.7848347375243034,
+      "grad_norm": 0.7299427390098572,
+      "learning_rate": 1.8599951824044033e-06,
+      "loss": 0.0926,
+      "step": 11016
+    },
+    {
+      "epoch": 1.7849967595593,
+      "grad_norm": 0.8017887473106384,
+      "learning_rate": 1.8595724607011878e-06,
+      "loss": 0.0925,
+      "step": 11017
+    },
+    {
+      "epoch": 1.7851587815942969,
+      "grad_norm": 0.7088046669960022,
+      "learning_rate": 1.8591497585935041e-06,
+      "loss": 0.0911,
+      "step": 11018
+    },
+    {
+      "epoch": 1.7853208036292936,
+      "grad_norm": 0.9126763939857483,
+      "learning_rate": 1.8587270760942875e-06,
+      "loss": 0.1143,
+      "step": 11019
+    },
+    {
+      "epoch": 1.7854828256642903,
+      "grad_norm": 0.9007970094680786,
+      "learning_rate": 1.858304413216471e-06,
+      "loss": 0.1132,
+      "step": 11020
+    },
+    {
+      "epoch": 1.7856448476992872,
+      "grad_norm": 0.7455145120620728,
+      "learning_rate": 1.8578817699729862e-06,
+      "loss": 0.0945,
+      "step": 11021
+    },
+    {
+      "epoch": 1.7858068697342837,
+      "grad_norm": 0.8875641226768494,
+      "learning_rate": 1.8574591463767656e-06,
+      "loss": 0.1057,
+      "step": 11022
+    },
+    {
+      "epoch": 1.7859688917692806,
+      "grad_norm": 0.8011730909347534,
+      "learning_rate": 1.8570365424407394e-06,
+      "loss": 0.1005,
+      "step": 11023
+    },
+    {
+      "epoch": 1.7861309138042774,
+      "grad_norm": 0.843669593334198,
+      "learning_rate": 1.8566139581778392e-06,
+      "loss": 0.0926,
+      "step": 11024
+    },
+    {
+      "epoch": 1.786292935839274,
+      "grad_norm": 0.7854774594306946,
+      "learning_rate": 1.856191393600995e-06,
+      "loss": 0.0967,
+      "step": 11025
+    },
+    {
+      "epoch": 1.786454957874271,
+      "grad_norm": 0.9077264666557312,
+      "learning_rate": 1.855768848723137e-06,
+      "loss": 0.1092,
+      "step": 11026
+    },
+    {
+      "epoch": 1.7866169799092677,
+      "grad_norm": 0.8441132307052612,
+      "learning_rate": 1.8553463235571927e-06,
+      "loss": 0.1036,
+      "step": 11027
+    },
+    {
+      "epoch": 1.7867790019442644,
+      "grad_norm": 1.07819664478302,
+      "learning_rate": 1.85492381811609e-06,
+      "loss": 0.0925,
+      "step": 11028
+    },
+    {
+      "epoch": 1.7869410239792611,
+      "grad_norm": 0.8448095917701721,
+      "learning_rate": 1.8545013324127587e-06,
+      "loss": 0.1002,
+      "step": 11029
+    },
+    {
+      "epoch": 1.7871030460142578,
+      "grad_norm": 0.8025045394897461,
+      "learning_rate": 1.854078866460124e-06,
+      "loss": 0.1058,
+      "step": 11030
+    },
+    {
+      "epoch": 1.7872650680492548,
+      "grad_norm": 0.7943110466003418,
+      "learning_rate": 1.8536564202711135e-06,
+      "loss": 0.1004,
+      "step": 11031
+    },
+    {
+      "epoch": 1.7874270900842515,
+      "grad_norm": 0.8177294731140137,
+      "learning_rate": 1.8532339938586513e-06,
+      "loss": 0.0892,
+      "step": 11032
+    },
+    {
+      "epoch": 1.7875891121192482,
+      "grad_norm": 0.9388481378555298,
+      "learning_rate": 1.8528115872356641e-06,
+      "loss": 0.111,
+      "step": 11033
+    },
+    {
+      "epoch": 1.7877511341542451,
+      "grad_norm": 0.8275322914123535,
+      "learning_rate": 1.8523892004150765e-06,
+      "loss": 0.1,
+      "step": 11034
+    },
+    {
+      "epoch": 1.7879131561892416,
+      "grad_norm": 0.777814507484436,
+      "learning_rate": 1.8519668334098124e-06,
+      "loss": 0.1026,
+      "step": 11035
+    },
+    {
+      "epoch": 1.7880751782242386,
+      "grad_norm": 1.0052603483200073,
+      "learning_rate": 1.8515444862327947e-06,
+      "loss": 0.1034,
+      "step": 11036
+    },
+    {
+      "epoch": 1.7882372002592353,
+      "grad_norm": 0.8155965805053711,
+      "learning_rate": 1.8511221588969457e-06,
+      "loss": 0.1,
+      "step": 11037
+    },
+    {
+      "epoch": 1.788399222294232,
+      "grad_norm": 0.7355263829231262,
+      "learning_rate": 1.8506998514151896e-06,
+      "loss": 0.091,
+      "step": 11038
+    },
+    {
+      "epoch": 1.788561244329229,
+      "grad_norm": 0.7432408332824707,
+      "learning_rate": 1.850277563800446e-06,
+      "loss": 0.092,
+      "step": 11039
+    },
+    {
+      "epoch": 1.7887232663642254,
+      "grad_norm": 0.7793557643890381,
+      "learning_rate": 1.8498552960656378e-06,
+      "loss": 0.1009,
+      "step": 11040
+    },
+    {
+      "epoch": 1.7888852883992223,
+      "grad_norm": 0.8657031655311584,
+      "learning_rate": 1.8494330482236832e-06,
+      "loss": 0.0957,
+      "step": 11041
+    },
+    {
+      "epoch": 1.789047310434219,
+      "grad_norm": 0.752953052520752,
+      "learning_rate": 1.8490108202875023e-06,
+      "loss": 0.0939,
+      "step": 11042
+    },
+    {
+      "epoch": 1.7892093324692158,
+      "grad_norm": 0.9895876049995422,
+      "learning_rate": 1.8485886122700158e-06,
+      "loss": 0.1191,
+      "step": 11043
+    },
+    {
+      "epoch": 1.7893713545042127,
+      "grad_norm": 0.7776055335998535,
+      "learning_rate": 1.848166424184142e-06,
+      "loss": 0.0981,
+      "step": 11044
+    },
+    {
+      "epoch": 1.7895333765392092,
+      "grad_norm": 0.7290987968444824,
+      "learning_rate": 1.8477442560427975e-06,
+      "loss": 0.0947,
+      "step": 11045
+    },
+    {
+      "epoch": 1.7896953985742061,
+      "grad_norm": 0.7357193231582642,
+      "learning_rate": 1.8473221078589006e-06,
+      "loss": 0.0906,
+      "step": 11046
+    },
+    {
+      "epoch": 1.7898574206092028,
+      "grad_norm": 0.8186837434768677,
+      "learning_rate": 1.8468999796453672e-06,
+      "loss": 0.1071,
+      "step": 11047
+    },
+    {
+      "epoch": 1.7900194426441995,
+      "grad_norm": 0.8652331233024597,
+      "learning_rate": 1.846477871415114e-06,
+      "loss": 0.1072,
+      "step": 11048
+    },
+    {
+      "epoch": 1.7901814646791965,
+      "grad_norm": 0.8034560084342957,
+      "learning_rate": 1.8460557831810571e-06,
+      "loss": 0.0951,
+      "step": 11049
+    },
+    {
+      "epoch": 1.790343486714193,
+      "grad_norm": 0.7994994521141052,
+      "learning_rate": 1.8456337149561105e-06,
+      "loss": 0.0932,
+      "step": 11050
+    },
+    {
+      "epoch": 1.79050550874919,
+      "grad_norm": 0.8264526724815369,
+      "learning_rate": 1.8452116667531886e-06,
+      "loss": 0.0989,
+      "step": 11051
+    },
+    {
+      "epoch": 1.7906675307841866,
+      "grad_norm": 0.8260409235954285,
+      "learning_rate": 1.8447896385852043e-06,
+      "loss": 0.104,
+      "step": 11052
+    },
+    {
+      "epoch": 1.7908295528191833,
+      "grad_norm": 0.8477978706359863,
+      "learning_rate": 1.844367630465073e-06,
+      "loss": 0.1045,
+      "step": 11053
+    },
+    {
+      "epoch": 1.7909915748541803,
+      "grad_norm": 0.790149986743927,
+      "learning_rate": 1.8439456424057044e-06,
+      "loss": 0.0985,
+      "step": 11054
+    },
+    {
+      "epoch": 1.791153596889177,
+      "grad_norm": 0.7050209045410156,
+      "learning_rate": 1.8435236744200126e-06,
+      "loss": 0.0867,
+      "step": 11055
+    },
+    {
+      "epoch": 1.7913156189241737,
+      "grad_norm": 0.9373990297317505,
+      "learning_rate": 1.8431017265209067e-06,
+      "loss": 0.1088,
+      "step": 11056
+    },
+    {
+      "epoch": 1.7914776409591704,
+      "grad_norm": 0.8682579398155212,
+      "learning_rate": 1.8426797987212985e-06,
+      "loss": 0.1047,
+      "step": 11057
+    },
+    {
+      "epoch": 1.791639662994167,
+      "grad_norm": 0.7753797769546509,
+      "learning_rate": 1.8422578910340985e-06,
+      "loss": 0.0957,
+      "step": 11058
+    },
+    {
+      "epoch": 1.791801685029164,
+      "grad_norm": 0.9039705395698547,
+      "learning_rate": 1.8418360034722149e-06,
+      "loss": 0.1037,
+      "step": 11059
+    },
+    {
+      "epoch": 1.7919637070641607,
+      "grad_norm": 0.7357102632522583,
+      "learning_rate": 1.8414141360485565e-06,
+      "loss": 0.0893,
+      "step": 11060
+    },
+    {
+      "epoch": 1.7921257290991575,
+      "grad_norm": 0.7715552449226379,
+      "learning_rate": 1.8409922887760317e-06,
+      "loss": 0.0987,
+      "step": 11061
+    },
+    {
+      "epoch": 1.7922877511341544,
+      "grad_norm": 0.8303348422050476,
+      "learning_rate": 1.840570461667549e-06,
+      "loss": 0.1043,
+      "step": 11062
+    },
+    {
+      "epoch": 1.7924497731691509,
+      "grad_norm": 0.7379674315452576,
+      "learning_rate": 1.8401486547360137e-06,
+      "loss": 0.0908,
+      "step": 11063
+    },
+    {
+      "epoch": 1.7926117952041478,
+      "grad_norm": 0.8676923513412476,
+      "learning_rate": 1.8397268679943333e-06,
+      "loss": 0.0983,
+      "step": 11064
+    },
+    {
+      "epoch": 1.7927738172391445,
+      "grad_norm": 0.8327484130859375,
+      "learning_rate": 1.8393051014554124e-06,
+      "loss": 0.1041,
+      "step": 11065
+    },
+    {
+      "epoch": 1.7929358392741412,
+      "grad_norm": 0.8785510659217834,
+      "learning_rate": 1.8388833551321562e-06,
+      "loss": 0.0972,
+      "step": 11066
+    },
+    {
+      "epoch": 1.7930978613091382,
+      "grad_norm": 0.8931607007980347,
+      "learning_rate": 1.8384616290374705e-06,
+      "loss": 0.1099,
+      "step": 11067
+    },
+    {
+      "epoch": 1.7932598833441347,
+      "grad_norm": 0.7808845639228821,
+      "learning_rate": 1.838039923184257e-06,
+      "loss": 0.0937,
+      "step": 11068
+    },
+    {
+      "epoch": 1.7934219053791316,
+      "grad_norm": 0.8837364912033081,
+      "learning_rate": 1.8376182375854207e-06,
+      "loss": 0.1046,
+      "step": 11069
+    },
+    {
+      "epoch": 1.7935839274141283,
+      "grad_norm": 0.7206534743309021,
+      "learning_rate": 1.8371965722538636e-06,
+      "loss": 0.0915,
+      "step": 11070
+    },
+    {
+      "epoch": 1.793745949449125,
+      "grad_norm": 0.7323435544967651,
+      "learning_rate": 1.8367749272024865e-06,
+      "loss": 0.0969,
+      "step": 11071
+    },
+    {
+      "epoch": 1.793907971484122,
+      "grad_norm": 0.7837962508201599,
+      "learning_rate": 1.836353302444192e-06,
+      "loss": 0.1013,
+      "step": 11072
+    },
+    {
+      "epoch": 1.7940699935191184,
+      "grad_norm": 0.7067288160324097,
+      "learning_rate": 1.8359316979918808e-06,
+      "loss": 0.0903,
+      "step": 11073
+    },
+    {
+      "epoch": 1.7942320155541154,
+      "grad_norm": 0.7567185163497925,
+      "learning_rate": 1.8355101138584524e-06,
+      "loss": 0.0876,
+      "step": 11074
+    },
+    {
+      "epoch": 1.794394037589112,
+      "grad_norm": 0.9895825982093811,
+      "learning_rate": 1.835088550056806e-06,
+      "loss": 0.1207,
+      "step": 11075
+    },
+    {
+      "epoch": 1.7945560596241088,
+      "grad_norm": 0.8607103824615479,
+      "learning_rate": 1.8346670065998411e-06,
+      "loss": 0.0999,
+      "step": 11076
+    },
+    {
+      "epoch": 1.7947180816591057,
+      "grad_norm": 0.7561851143836975,
+      "learning_rate": 1.8342454835004566e-06,
+      "loss": 0.0952,
+      "step": 11077
+    },
+    {
+      "epoch": 1.7948801036941024,
+      "grad_norm": 0.756078839302063,
+      "learning_rate": 1.8338239807715486e-06,
+      "loss": 0.0955,
+      "step": 11078
+    },
+    {
+      "epoch": 1.7950421257290992,
+      "grad_norm": 0.948203444480896,
+      "learning_rate": 1.833402498426015e-06,
+      "loss": 0.1169,
+      "step": 11079
+    },
+    {
+      "epoch": 1.7952041477640959,
+      "grad_norm": 0.809870183467865,
+      "learning_rate": 1.8329810364767511e-06,
+      "loss": 0.0992,
+      "step": 11080
+    },
+    {
+      "epoch": 1.7953661697990926,
+      "grad_norm": 0.7944110631942749,
+      "learning_rate": 1.8325595949366537e-06,
+      "loss": 0.0974,
+      "step": 11081
+    },
+    {
+      "epoch": 1.7955281918340895,
+      "grad_norm": 0.7319938540458679,
+      "learning_rate": 1.8321381738186178e-06,
+      "loss": 0.0953,
+      "step": 11082
+    },
+    {
+      "epoch": 1.7956902138690862,
+      "grad_norm": 0.7145232558250427,
+      "learning_rate": 1.8317167731355373e-06,
+      "loss": 0.0909,
+      "step": 11083
+    },
+    {
+      "epoch": 1.795852235904083,
+      "grad_norm": 0.7656394243240356,
+      "learning_rate": 1.8312953929003068e-06,
+      "loss": 0.0937,
+      "step": 11084
+    },
+    {
+      "epoch": 1.7960142579390799,
+      "grad_norm": 0.8850581645965576,
+      "learning_rate": 1.8308740331258177e-06,
+      "loss": 0.1003,
+      "step": 11085
+    },
+    {
+      "epoch": 1.7961762799740764,
+      "grad_norm": 0.7528477311134338,
+      "learning_rate": 1.8304526938249653e-06,
+      "loss": 0.0894,
+      "step": 11086
+    },
+    {
+      "epoch": 1.7963383020090733,
+      "grad_norm": 0.7964629530906677,
+      "learning_rate": 1.8300313750106396e-06,
+      "loss": 0.0934,
+      "step": 11087
+    },
+    {
+      "epoch": 1.79650032404407,
+      "grad_norm": 0.6928067207336426,
+      "learning_rate": 1.8296100766957331e-06,
+      "loss": 0.0834,
+      "step": 11088
+    },
+    {
+      "epoch": 1.7966623460790667,
+      "grad_norm": 0.9328933954238892,
+      "learning_rate": 1.8291887988931357e-06,
+      "loss": 0.1128,
+      "step": 11089
+    },
+    {
+      "epoch": 1.7968243681140637,
+      "grad_norm": 0.7624589800834656,
+      "learning_rate": 1.828767541615737e-06,
+      "loss": 0.0925,
+      "step": 11090
+    },
+    {
+      "epoch": 1.7969863901490601,
+      "grad_norm": 0.7582963109016418,
+      "learning_rate": 1.828346304876428e-06,
+      "loss": 0.0924,
+      "step": 11091
+    },
+    {
+      "epoch": 1.797148412184057,
+      "grad_norm": 0.8816431760787964,
+      "learning_rate": 1.8279250886880962e-06,
+      "loss": 0.0985,
+      "step": 11092
+    },
+    {
+      "epoch": 1.7973104342190538,
+      "grad_norm": 0.9561519026756287,
+      "learning_rate": 1.8275038930636314e-06,
+      "loss": 0.1172,
+      "step": 11093
+    },
+    {
+      "epoch": 1.7974724562540505,
+      "grad_norm": 0.845643162727356,
+      "learning_rate": 1.827082718015919e-06,
+      "loss": 0.1133,
+      "step": 11094
+    },
+    {
+      "epoch": 1.7976344782890474,
+      "grad_norm": 0.8886216282844543,
+      "learning_rate": 1.8266615635578464e-06,
+      "loss": 0.1089,
+      "step": 11095
+    },
+    {
+      "epoch": 1.797796500324044,
+      "grad_norm": 0.8109029531478882,
+      "learning_rate": 1.8262404297023013e-06,
+      "loss": 0.0978,
+      "step": 11096
+    },
+    {
+      "epoch": 1.7979585223590409,
+      "grad_norm": 0.7375802993774414,
+      "learning_rate": 1.825819316462169e-06,
+      "loss": 0.0906,
+      "step": 11097
+    },
+    {
+      "epoch": 1.7981205443940376,
+      "grad_norm": 0.8273477554321289,
+      "learning_rate": 1.8253982238503338e-06,
+      "loss": 0.1088,
+      "step": 11098
+    },
+    {
+      "epoch": 1.7982825664290343,
+      "grad_norm": 0.6932176947593689,
+      "learning_rate": 1.8249771518796794e-06,
+      "loss": 0.0845,
+      "step": 11099
+    },
+    {
+      "epoch": 1.7984445884640312,
+      "grad_norm": 0.7964252233505249,
+      "learning_rate": 1.8245561005630921e-06,
+      "loss": 0.0918,
+      "step": 11100
+    },
+    {
+      "epoch": 1.7986066104990277,
+      "grad_norm": 0.708993136882782,
+      "learning_rate": 1.824135069913453e-06,
+      "loss": 0.0911,
+      "step": 11101
+    },
+    {
+      "epoch": 1.7987686325340246,
+      "grad_norm": 0.8443676233291626,
+      "learning_rate": 1.823714059943646e-06,
+      "loss": 0.1014,
+      "step": 11102
+    },
+    {
+      "epoch": 1.7989306545690213,
+      "grad_norm": 0.8077982068061829,
+      "learning_rate": 1.823293070666551e-06,
+      "loss": 0.0996,
+      "step": 11103
+    },
+    {
+      "epoch": 1.799092676604018,
+      "grad_norm": 0.8241330981254578,
+      "learning_rate": 1.8228721020950504e-06,
+      "loss": 0.0983,
+      "step": 11104
+    },
+    {
+      "epoch": 1.799254698639015,
+      "grad_norm": 0.8449083566665649,
+      "learning_rate": 1.8224511542420254e-06,
+      "loss": 0.1019,
+      "step": 11105
+    },
+    {
+      "epoch": 1.7994167206740117,
+      "grad_norm": 0.6822419166564941,
+      "learning_rate": 1.8220302271203557e-06,
+      "loss": 0.0793,
+      "step": 11106
+    },
+    {
+      "epoch": 1.7995787427090084,
+      "grad_norm": 0.8199465870857239,
+      "learning_rate": 1.82160932074292e-06,
+      "loss": 0.0983,
+      "step": 11107
+    },
+    {
+      "epoch": 1.7997407647440054,
+      "grad_norm": 0.9466975927352905,
+      "learning_rate": 1.8211884351225978e-06,
+      "loss": 0.1151,
+      "step": 11108
+    },
+    {
+      "epoch": 1.7999027867790018,
+      "grad_norm": 0.7171126008033752,
+      "learning_rate": 1.8207675702722661e-06,
+      "loss": 0.0904,
+      "step": 11109
+    },
+    {
+      "epoch": 1.8000648088139988,
+      "grad_norm": 0.8177133202552795,
+      "learning_rate": 1.8203467262048033e-06,
+      "loss": 0.096,
+      "step": 11110
+    },
+    {
+      "epoch": 1.8002268308489955,
+      "grad_norm": 0.9946437478065491,
+      "learning_rate": 1.8199259029330865e-06,
+      "loss": 0.1111,
+      "step": 11111
+    },
+    {
+      "epoch": 1.8003888528839922,
+      "grad_norm": 0.8421857357025146,
+      "learning_rate": 1.819505100469991e-06,
+      "loss": 0.1072,
+      "step": 11112
+    },
+    {
+      "epoch": 1.8005508749189891,
+      "grad_norm": 0.7888447642326355,
+      "learning_rate": 1.8190843188283925e-06,
+      "loss": 0.0884,
+      "step": 11113
+    },
+    {
+      "epoch": 1.8007128969539856,
+      "grad_norm": 0.7778288722038269,
+      "learning_rate": 1.8186635580211654e-06,
+      "loss": 0.0878,
+      "step": 11114
+    },
+    {
+      "epoch": 1.8008749189889826,
+      "grad_norm": 0.7060684561729431,
+      "learning_rate": 1.8182428180611855e-06,
+      "loss": 0.0799,
+      "step": 11115
+    },
+    {
+      "epoch": 1.8010369410239793,
+      "grad_norm": 0.8897756934165955,
+      "learning_rate": 1.8178220989613255e-06,
+      "loss": 0.1116,
+      "step": 11116
+    },
+    {
+      "epoch": 1.801198963058976,
+      "grad_norm": 0.7480401992797852,
+      "learning_rate": 1.8174014007344586e-06,
+      "loss": 0.0954,
+      "step": 11117
+    },
+    {
+      "epoch": 1.801360985093973,
+      "grad_norm": 0.7541413307189941,
+      "learning_rate": 1.8169807233934567e-06,
+      "loss": 0.0945,
+      "step": 11118
+    },
+    {
+      "epoch": 1.8015230071289694,
+      "grad_norm": 0.7814289927482605,
+      "learning_rate": 1.8165600669511912e-06,
+      "loss": 0.0951,
+      "step": 11119
+    },
+    {
+      "epoch": 1.8016850291639663,
+      "grad_norm": 0.8429241180419922,
+      "learning_rate": 1.8161394314205343e-06,
+      "loss": 0.1045,
+      "step": 11120
+    },
+    {
+      "epoch": 1.801847051198963,
+      "grad_norm": 0.8381679654121399,
+      "learning_rate": 1.8157188168143564e-06,
+      "loss": 0.1059,
+      "step": 11121
+    },
+    {
+      "epoch": 1.8020090732339598,
+      "grad_norm": 0.8170045018196106,
+      "learning_rate": 1.8152982231455262e-06,
+      "loss": 0.0991,
+      "step": 11122
+    },
+    {
+      "epoch": 1.8021710952689567,
+      "grad_norm": 0.8006271719932556,
+      "learning_rate": 1.8148776504269129e-06,
+      "loss": 0.0989,
+      "step": 11123
+    },
+    {
+      "epoch": 1.8023331173039532,
+      "grad_norm": 0.7720314264297485,
+      "learning_rate": 1.8144570986713867e-06,
+      "loss": 0.0928,
+      "step": 11124
+    },
+    {
+      "epoch": 1.8024951393389501,
+      "grad_norm": 0.7991904020309448,
+      "learning_rate": 1.8140365678918138e-06,
+      "loss": 0.0986,
+      "step": 11125
+    },
+    {
+      "epoch": 1.8026571613739468,
+      "grad_norm": 0.9030035138130188,
+      "learning_rate": 1.8136160581010624e-06,
+      "loss": 0.1187,
+      "step": 11126
+    },
+    {
+      "epoch": 1.8028191834089435,
+      "grad_norm": 0.8388223648071289,
+      "learning_rate": 1.813195569311998e-06,
+      "loss": 0.1017,
+      "step": 11127
+    },
+    {
+      "epoch": 1.8029812054439405,
+      "grad_norm": 0.9252680540084839,
+      "learning_rate": 1.8127751015374865e-06,
+      "loss": 0.1104,
+      "step": 11128
+    },
+    {
+      "epoch": 1.8031432274789372,
+      "grad_norm": 0.8280481696128845,
+      "learning_rate": 1.8123546547903944e-06,
+      "loss": 0.1046,
+      "step": 11129
+    },
+    {
+      "epoch": 1.803305249513934,
+      "grad_norm": 0.8752947449684143,
+      "learning_rate": 1.8119342290835864e-06,
+      "loss": 0.1067,
+      "step": 11130
+    },
+    {
+      "epoch": 1.8034672715489306,
+      "grad_norm": 0.753736138343811,
+      "learning_rate": 1.8115138244299254e-06,
+      "loss": 0.0953,
+      "step": 11131
+    },
+    {
+      "epoch": 1.8036292935839273,
+      "grad_norm": 0.8405237197875977,
+      "learning_rate": 1.8110934408422758e-06,
+      "loss": 0.0967,
+      "step": 11132
+    },
+    {
+      "epoch": 1.8037913156189243,
+      "grad_norm": 0.9108558893203735,
+      "learning_rate": 1.8106730783334985e-06,
+      "loss": 0.1006,
+      "step": 11133
+    },
+    {
+      "epoch": 1.803953337653921,
+      "grad_norm": 0.7528913021087646,
+      "learning_rate": 1.810252736916458e-06,
+      "loss": 0.0983,
+      "step": 11134
+    },
+    {
+      "epoch": 1.8041153596889177,
+      "grad_norm": 0.8099750280380249,
+      "learning_rate": 1.8098324166040146e-06,
+      "loss": 0.1013,
+      "step": 11135
+    },
+    {
+      "epoch": 1.8042773817239146,
+      "grad_norm": 0.9023746848106384,
+      "learning_rate": 1.8094121174090288e-06,
+      "loss": 0.1148,
+      "step": 11136
+    },
+    {
+      "epoch": 1.804439403758911,
+      "grad_norm": 0.8526237607002258,
+      "learning_rate": 1.8089918393443611e-06,
+      "loss": 0.1046,
+      "step": 11137
+    },
+    {
+      "epoch": 1.804601425793908,
+      "grad_norm": 0.7785555124282837,
+      "learning_rate": 1.80857158242287e-06,
+      "loss": 0.0932,
+      "step": 11138
+    },
+    {
+      "epoch": 1.8047634478289047,
+      "grad_norm": 0.7802260518074036,
+      "learning_rate": 1.8081513466574164e-06,
+      "loss": 0.1013,
+      "step": 11139
+    },
+    {
+      "epoch": 1.8049254698639015,
+      "grad_norm": 0.8971158862113953,
+      "learning_rate": 1.8077311320608571e-06,
+      "loss": 0.1221,
+      "step": 11140
+    },
+    {
+      "epoch": 1.8050874918988984,
+      "grad_norm": 0.833736002445221,
+      "learning_rate": 1.8073109386460502e-06,
+      "loss": 0.1064,
+      "step": 11141
+    },
+    {
+      "epoch": 1.8052495139338949,
+      "grad_norm": 0.8659934401512146,
+      "learning_rate": 1.806890766425851e-06,
+      "loss": 0.1017,
+      "step": 11142
+    },
+    {
+      "epoch": 1.8054115359688918,
+      "grad_norm": 0.7669563889503479,
+      "learning_rate": 1.8064706154131179e-06,
+      "loss": 0.0973,
+      "step": 11143
+    },
+    {
+      "epoch": 1.8055735580038885,
+      "grad_norm": 0.9565890431404114,
+      "learning_rate": 1.8060504856207062e-06,
+      "loss": 0.1079,
+      "step": 11144
+    },
+    {
+      "epoch": 1.8057355800388852,
+      "grad_norm": 0.8553591370582581,
+      "learning_rate": 1.8056303770614697e-06,
+      "loss": 0.1065,
+      "step": 11145
+    },
+    {
+      "epoch": 1.8058976020738822,
+      "grad_norm": 0.9055613279342651,
+      "learning_rate": 1.8052102897482643e-06,
+      "loss": 0.1114,
+      "step": 11146
+    },
+    {
+      "epoch": 1.8060596241088787,
+      "grad_norm": 0.8043663501739502,
+      "learning_rate": 1.8047902236939405e-06,
+      "loss": 0.1029,
+      "step": 11147
+    },
+    {
+      "epoch": 1.8062216461438756,
+      "grad_norm": 0.8190831542015076,
+      "learning_rate": 1.8043701789113552e-06,
+      "loss": 0.098,
+      "step": 11148
+    },
+    {
+      "epoch": 1.8063836681788723,
+      "grad_norm": 0.8615659475326538,
+      "learning_rate": 1.8039501554133588e-06,
+      "loss": 0.0942,
+      "step": 11149
+    },
+    {
+      "epoch": 1.806545690213869,
+      "grad_norm": 0.9080948829650879,
+      "learning_rate": 1.8035301532128032e-06,
+      "loss": 0.1049,
+      "step": 11150
+    },
+    {
+      "epoch": 1.806707712248866,
+      "grad_norm": 0.7901988625526428,
+      "learning_rate": 1.8031101723225393e-06,
+      "loss": 0.0924,
+      "step": 11151
+    },
+    {
+      "epoch": 1.8068697342838627,
+      "grad_norm": 0.8067225813865662,
+      "learning_rate": 1.8026902127554172e-06,
+      "loss": 0.1016,
+      "step": 11152
+    },
+    {
+      "epoch": 1.8070317563188594,
+      "grad_norm": 0.7822276949882507,
+      "learning_rate": 1.8022702745242882e-06,
+      "loss": 0.0953,
+      "step": 11153
+    },
+    {
+      "epoch": 1.807193778353856,
+      "grad_norm": 0.8535957336425781,
+      "learning_rate": 1.8018503576419996e-06,
+      "loss": 0.1074,
+      "step": 11154
+    },
+    {
+      "epoch": 1.8073558003888528,
+      "grad_norm": 0.7549822926521301,
+      "learning_rate": 1.8014304621214008e-06,
+      "loss": 0.0964,
+      "step": 11155
+    },
+    {
+      "epoch": 1.8075178224238497,
+      "grad_norm": 0.816744863986969,
+      "learning_rate": 1.8010105879753398e-06,
+      "loss": 0.105,
+      "step": 11156
+    },
+    {
+      "epoch": 1.8076798444588464,
+      "grad_norm": 0.7305224537849426,
+      "learning_rate": 1.800590735216662e-06,
+      "loss": 0.0952,
+      "step": 11157
+    },
+    {
+      "epoch": 1.8078418664938432,
+      "grad_norm": 0.8567788600921631,
+      "learning_rate": 1.800170903858216e-06,
+      "loss": 0.1097,
+      "step": 11158
+    },
+    {
+      "epoch": 1.80800388852884,
+      "grad_norm": 0.8007251024246216,
+      "learning_rate": 1.799751093912847e-06,
+      "loss": 0.0898,
+      "step": 11159
+    },
+    {
+      "epoch": 1.8081659105638366,
+      "grad_norm": 0.7454932928085327,
+      "learning_rate": 1.7993313053933998e-06,
+      "loss": 0.0942,
+      "step": 11160
+    },
+    {
+      "epoch": 1.8083279325988335,
+      "grad_norm": 0.8023138046264648,
+      "learning_rate": 1.7989115383127195e-06,
+      "loss": 0.102,
+      "step": 11161
+    },
+    {
+      "epoch": 1.8084899546338302,
+      "grad_norm": 0.7055501937866211,
+      "learning_rate": 1.7984917926836484e-06,
+      "loss": 0.0865,
+      "step": 11162
+    },
+    {
+      "epoch": 1.808651976668827,
+      "grad_norm": 0.7876911759376526,
+      "learning_rate": 1.7980720685190314e-06,
+      "loss": 0.0841,
+      "step": 11163
+    },
+    {
+      "epoch": 1.8088139987038239,
+      "grad_norm": 0.8725854158401489,
+      "learning_rate": 1.7976523658317104e-06,
+      "loss": 0.1082,
+      "step": 11164
+    },
+    {
+      "epoch": 1.8089760207388204,
+      "grad_norm": 0.8327855467796326,
+      "learning_rate": 1.7972326846345277e-06,
+      "loss": 0.1103,
+      "step": 11165
+    },
+    {
+      "epoch": 1.8091380427738173,
+      "grad_norm": 0.9465359449386597,
+      "learning_rate": 1.7968130249403238e-06,
+      "loss": 0.114,
+      "step": 11166
+    },
+    {
+      "epoch": 1.809300064808814,
+      "grad_norm": 0.8367246389389038,
+      "learning_rate": 1.7963933867619396e-06,
+      "loss": 0.1114,
+      "step": 11167
+    },
+    {
+      "epoch": 1.8094620868438107,
+      "grad_norm": 0.821707546710968,
+      "learning_rate": 1.7959737701122157e-06,
+      "loss": 0.0933,
+      "step": 11168
+    },
+    {
+      "epoch": 1.8096241088788076,
+      "grad_norm": 0.951481282711029,
+      "learning_rate": 1.79555417500399e-06,
+      "loss": 0.1188,
+      "step": 11169
+    },
+    {
+      "epoch": 1.8097861309138041,
+      "grad_norm": 0.7933133244514465,
+      "learning_rate": 1.7951346014501027e-06,
+      "loss": 0.1043,
+      "step": 11170
+    },
+    {
+      "epoch": 1.809948152948801,
+      "grad_norm": 0.8114445209503174,
+      "learning_rate": 1.7947150494633897e-06,
+      "loss": 0.0908,
+      "step": 11171
+    },
+    {
+      "epoch": 1.8101101749837978,
+      "grad_norm": 0.7849157452583313,
+      "learning_rate": 1.7942955190566899e-06,
+      "loss": 0.1009,
+      "step": 11172
+    },
+    {
+      "epoch": 1.8102721970187945,
+      "grad_norm": 0.7490556836128235,
+      "learning_rate": 1.7938760102428396e-06,
+      "loss": 0.0816,
+      "step": 11173
+    },
+    {
+      "epoch": 1.8104342190537914,
+      "grad_norm": 0.8201428651809692,
+      "learning_rate": 1.7934565230346752e-06,
+      "loss": 0.1019,
+      "step": 11174
+    },
+    {
+      "epoch": 1.810596241088788,
+      "grad_norm": 0.7032244801521301,
+      "learning_rate": 1.7930370574450304e-06,
+      "loss": 0.0911,
+      "step": 11175
+    },
+    {
+      "epoch": 1.8107582631237849,
+      "grad_norm": 0.8158414959907532,
+      "learning_rate": 1.7926176134867408e-06,
+      "loss": 0.1025,
+      "step": 11176
+    },
+    {
+      "epoch": 1.8109202851587816,
+      "grad_norm": 0.8338707685470581,
+      "learning_rate": 1.792198191172641e-06,
+      "loss": 0.1059,
+      "step": 11177
+    },
+    {
+      "epoch": 1.8110823071937783,
+      "grad_norm": 0.7679630517959595,
+      "learning_rate": 1.7917787905155634e-06,
+      "loss": 0.0831,
+      "step": 11178
+    },
+    {
+      "epoch": 1.8112443292287752,
+      "grad_norm": 0.8508248329162598,
+      "learning_rate": 1.7913594115283414e-06,
+      "loss": 0.1109,
+      "step": 11179
+    },
+    {
+      "epoch": 1.811406351263772,
+      "grad_norm": 0.8345566391944885,
+      "learning_rate": 1.790940054223806e-06,
+      "loss": 0.1056,
+      "step": 11180
+    },
+    {
+      "epoch": 1.8115683732987686,
+      "grad_norm": 0.8556733131408691,
+      "learning_rate": 1.7905207186147888e-06,
+      "loss": 0.1059,
+      "step": 11181
+    },
+    {
+      "epoch": 1.8117303953337653,
+      "grad_norm": 0.7630212903022766,
+      "learning_rate": 1.7901014047141208e-06,
+      "loss": 0.0991,
+      "step": 11182
+    },
+    {
+      "epoch": 1.811892417368762,
+      "grad_norm": 0.9459897875785828,
+      "learning_rate": 1.7896821125346325e-06,
+      "loss": 0.11,
+      "step": 11183
+    },
+    {
+      "epoch": 1.812054439403759,
+      "grad_norm": 0.7780237793922424,
+      "learning_rate": 1.7892628420891526e-06,
+      "loss": 0.1011,
+      "step": 11184
+    },
+    {
+      "epoch": 1.8122164614387557,
+      "grad_norm": 0.7486929297447205,
+      "learning_rate": 1.7888435933905097e-06,
+      "loss": 0.0906,
+      "step": 11185
+    },
+    {
+      "epoch": 1.8123784834737524,
+      "grad_norm": 0.8000259399414062,
+      "learning_rate": 1.788424366451531e-06,
+      "loss": 0.0972,
+      "step": 11186
+    },
+    {
+      "epoch": 1.8125405055087493,
+      "grad_norm": 0.837604284286499,
+      "learning_rate": 1.7880051612850455e-06,
+      "loss": 0.1022,
+      "step": 11187
+    },
+    {
+      "epoch": 1.8127025275437458,
+      "grad_norm": 0.7862725853919983,
+      "learning_rate": 1.7875859779038796e-06,
+      "loss": 0.0986,
+      "step": 11188
+    },
+    {
+      "epoch": 1.8128645495787428,
+      "grad_norm": 0.7725867629051208,
+      "learning_rate": 1.7871668163208577e-06,
+      "loss": 0.0968,
+      "step": 11189
+    },
+    {
+      "epoch": 1.8130265716137395,
+      "grad_norm": 0.7349421977996826,
+      "learning_rate": 1.7867476765488061e-06,
+      "loss": 0.0899,
+      "step": 11190
+    },
+    {
+      "epoch": 1.8131885936487362,
+      "grad_norm": 0.7476187944412231,
+      "learning_rate": 1.78632855860055e-06,
+      "loss": 0.0975,
+      "step": 11191
+    },
+    {
+      "epoch": 1.8133506156837331,
+      "grad_norm": 0.8673322200775146,
+      "learning_rate": 1.7859094624889135e-06,
+      "loss": 0.0985,
+      "step": 11192
+    },
+    {
+      "epoch": 1.8135126377187296,
+      "grad_norm": 0.77607262134552,
+      "learning_rate": 1.785490388226719e-06,
+      "loss": 0.0923,
+      "step": 11193
+    },
+    {
+      "epoch": 1.8136746597537265,
+      "grad_norm": 0.8594310283660889,
+      "learning_rate": 1.7850713358267897e-06,
+      "loss": 0.1137,
+      "step": 11194
+    },
+    {
+      "epoch": 1.8138366817887233,
+      "grad_norm": 0.7893880009651184,
+      "learning_rate": 1.7846523053019466e-06,
+      "loss": 0.096,
+      "step": 11195
+    },
+    {
+      "epoch": 1.81399870382372,
+      "grad_norm": 0.8915963768959045,
+      "learning_rate": 1.7842332966650122e-06,
+      "loss": 0.1095,
+      "step": 11196
+    },
+    {
+      "epoch": 1.814160725858717,
+      "grad_norm": 0.7770095467567444,
+      "learning_rate": 1.7838143099288075e-06,
+      "loss": 0.0921,
+      "step": 11197
+    },
+    {
+      "epoch": 1.8143227478937134,
+      "grad_norm": 0.8442981243133545,
+      "learning_rate": 1.7833953451061513e-06,
+      "loss": 0.1042,
+      "step": 11198
+    },
+    {
+      "epoch": 1.8144847699287103,
+      "grad_norm": 0.8559162020683289,
+      "learning_rate": 1.7829764022098633e-06,
+      "loss": 0.1041,
+      "step": 11199
+    },
+    {
+      "epoch": 1.814646791963707,
+      "grad_norm": 0.7358631491661072,
+      "learning_rate": 1.7825574812527617e-06,
+      "loss": 0.0866,
+      "step": 11200
+    },
+    {
+      "epoch": 1.8148088139987038,
+      "grad_norm": 0.8252244591712952,
+      "learning_rate": 1.7821385822476661e-06,
+      "loss": 0.1032,
+      "step": 11201
+    },
+    {
+      "epoch": 1.8149708360337007,
+      "grad_norm": 0.725712239742279,
+      "learning_rate": 1.781719705207392e-06,
+      "loss": 0.0829,
+      "step": 11202
+    },
+    {
+      "epoch": 1.8151328580686974,
+      "grad_norm": 0.7640675902366638,
+      "learning_rate": 1.7813008501447576e-06,
+      "loss": 0.0867,
+      "step": 11203
+    },
+    {
+      "epoch": 1.815294880103694,
+      "grad_norm": 0.8494378924369812,
+      "learning_rate": 1.7808820170725772e-06,
+      "loss": 0.1012,
+      "step": 11204
+    },
+    {
+      "epoch": 1.8154569021386908,
+      "grad_norm": 0.7609313130378723,
+      "learning_rate": 1.7804632060036665e-06,
+      "loss": 0.096,
+      "step": 11205
+    },
+    {
+      "epoch": 1.8156189241736875,
+      "grad_norm": 0.7626781463623047,
+      "learning_rate": 1.7800444169508414e-06,
+      "loss": 0.0921,
+      "step": 11206
+    },
+    {
+      "epoch": 1.8157809462086845,
+      "grad_norm": 0.9517927169799805,
+      "learning_rate": 1.7796256499269141e-06,
+      "loss": 0.1178,
+      "step": 11207
+    },
+    {
+      "epoch": 1.8159429682436812,
+      "grad_norm": 0.8355745673179626,
+      "learning_rate": 1.7792069049446987e-06,
+      "loss": 0.1029,
+      "step": 11208
+    },
+    {
+      "epoch": 1.8161049902786779,
+      "grad_norm": 0.8401842713356018,
+      "learning_rate": 1.7787881820170073e-06,
+      "loss": 0.1033,
+      "step": 11209
+    },
+    {
+      "epoch": 1.8162670123136748,
+      "grad_norm": 0.8746349215507507,
+      "learning_rate": 1.7783694811566534e-06,
+      "loss": 0.104,
+      "step": 11210
+    },
+    {
+      "epoch": 1.8164290343486713,
+      "grad_norm": 0.8786240816116333,
+      "learning_rate": 1.7779508023764464e-06,
+      "loss": 0.1073,
+      "step": 11211
+    },
+    {
+      "epoch": 1.8165910563836682,
+      "grad_norm": 0.700187087059021,
+      "learning_rate": 1.777532145689198e-06,
+      "loss": 0.0841,
+      "step": 11212
+    },
+    {
+      "epoch": 1.816753078418665,
+      "grad_norm": 0.7752106785774231,
+      "learning_rate": 1.7771135111077173e-06,
+      "loss": 0.0987,
+      "step": 11213
+    },
+    {
+      "epoch": 1.8169151004536617,
+      "grad_norm": 0.8160271644592285,
+      "learning_rate": 1.7766948986448131e-06,
+      "loss": 0.096,
+      "step": 11214
+    },
+    {
+      "epoch": 1.8170771224886586,
+      "grad_norm": 0.8085820078849792,
+      "learning_rate": 1.7762763083132958e-06,
+      "loss": 0.0965,
+      "step": 11215
+    },
+    {
+      "epoch": 1.817239144523655,
+      "grad_norm": 0.8033862113952637,
+      "learning_rate": 1.7758577401259716e-06,
+      "loss": 0.0961,
+      "step": 11216
+    },
+    {
+      "epoch": 1.817401166558652,
+      "grad_norm": 0.7184333205223083,
+      "learning_rate": 1.775439194095648e-06,
+      "loss": 0.0934,
+      "step": 11217
+    },
+    {
+      "epoch": 1.8175631885936487,
+      "grad_norm": 0.8042249083518982,
+      "learning_rate": 1.7750206702351325e-06,
+      "loss": 0.1002,
+      "step": 11218
+    },
+    {
+      "epoch": 1.8177252106286454,
+      "grad_norm": 0.779135525226593,
+      "learning_rate": 1.7746021685572284e-06,
+      "loss": 0.0907,
+      "step": 11219
+    },
+    {
+      "epoch": 1.8178872326636424,
+      "grad_norm": 0.8143093585968018,
+      "learning_rate": 1.7741836890747438e-06,
+      "loss": 0.1011,
+      "step": 11220
+    },
+    {
+      "epoch": 1.8180492546986389,
+      "grad_norm": 0.8787755370140076,
+      "learning_rate": 1.7737652318004818e-06,
+      "loss": 0.1014,
+      "step": 11221
+    },
+    {
+      "epoch": 1.8182112767336358,
+      "grad_norm": 0.8515397310256958,
+      "learning_rate": 1.7733467967472459e-06,
+      "loss": 0.1032,
+      "step": 11222
+    },
+    {
+      "epoch": 1.8183732987686325,
+      "grad_norm": 0.8510406017303467,
+      "learning_rate": 1.7729283839278403e-06,
+      "loss": 0.0972,
+      "step": 11223
+    },
+    {
+      "epoch": 1.8185353208036292,
+      "grad_norm": 0.8091854453086853,
+      "learning_rate": 1.7725099933550649e-06,
+      "loss": 0.0957,
+      "step": 11224
+    },
+    {
+      "epoch": 1.8186973428386262,
+      "grad_norm": 0.7440215349197388,
+      "learning_rate": 1.7720916250417248e-06,
+      "loss": 0.0914,
+      "step": 11225
+    },
+    {
+      "epoch": 1.8188593648736227,
+      "grad_norm": 0.8673259615898132,
+      "learning_rate": 1.7716732790006188e-06,
+      "loss": 0.101,
+      "step": 11226
+    },
+    {
+      "epoch": 1.8190213869086196,
+      "grad_norm": 0.8858534097671509,
+      "learning_rate": 1.7712549552445484e-06,
+      "loss": 0.1069,
+      "step": 11227
+    },
+    {
+      "epoch": 1.8191834089436163,
+      "grad_norm": 0.7519997954368591,
+      "learning_rate": 1.7708366537863129e-06,
+      "loss": 0.0896,
+      "step": 11228
+    },
+    {
+      "epoch": 1.819345430978613,
+      "grad_norm": 0.7740821242332458,
+      "learning_rate": 1.7704183746387105e-06,
+      "loss": 0.0916,
+      "step": 11229
+    },
+    {
+      "epoch": 1.81950745301361,
+      "grad_norm": 0.898308277130127,
+      "learning_rate": 1.7700001178145409e-06,
+      "loss": 0.1155,
+      "step": 11230
+    },
+    {
+      "epoch": 1.8196694750486067,
+      "grad_norm": 0.7707512378692627,
+      "learning_rate": 1.7695818833266009e-06,
+      "loss": 0.0876,
+      "step": 11231
+    },
+    {
+      "epoch": 1.8198314970836034,
+      "grad_norm": 0.7827706336975098,
+      "learning_rate": 1.7691636711876883e-06,
+      "loss": 0.1,
+      "step": 11232
+    },
+    {
+      "epoch": 1.8199935191186,
+      "grad_norm": 0.8508397340774536,
+      "learning_rate": 1.768745481410597e-06,
+      "loss": 0.0975,
+      "step": 11233
+    },
+    {
+      "epoch": 1.8201555411535968,
+      "grad_norm": 0.8238134384155273,
+      "learning_rate": 1.768327314008126e-06,
+      "loss": 0.0968,
+      "step": 11234
+    },
+    {
+      "epoch": 1.8203175631885937,
+      "grad_norm": 0.7700952291488647,
+      "learning_rate": 1.7679091689930683e-06,
+      "loss": 0.0831,
+      "step": 11235
+    },
+    {
+      "epoch": 1.8204795852235904,
+      "grad_norm": 0.7718937993049622,
+      "learning_rate": 1.7674910463782186e-06,
+      "loss": 0.1014,
+      "step": 11236
+    },
+    {
+      "epoch": 1.8206416072585871,
+      "grad_norm": 0.9156939387321472,
+      "learning_rate": 1.76707294617637e-06,
+      "loss": 0.1105,
+      "step": 11237
+    },
+    {
+      "epoch": 1.820803629293584,
+      "grad_norm": 0.85805743932724,
+      "learning_rate": 1.766654868400315e-06,
+      "loss": 0.105,
+      "step": 11238
+    },
+    {
+      "epoch": 1.8209656513285806,
+      "grad_norm": 0.7723628878593445,
+      "learning_rate": 1.766236813062847e-06,
+      "loss": 0.0991,
+      "step": 11239
+    },
+    {
+      "epoch": 1.8211276733635775,
+      "grad_norm": 0.7656691670417786,
+      "learning_rate": 1.7658187801767568e-06,
+      "loss": 0.0986,
+      "step": 11240
+    },
+    {
+      "epoch": 1.8212896953985742,
+      "grad_norm": 0.8363527655601501,
+      "learning_rate": 1.765400769754836e-06,
+      "loss": 0.0918,
+      "step": 11241
+    },
+    {
+      "epoch": 1.821451717433571,
+      "grad_norm": 0.6773914694786072,
+      "learning_rate": 1.7649827818098727e-06,
+      "loss": 0.0836,
+      "step": 11242
+    },
+    {
+      "epoch": 1.8216137394685679,
+      "grad_norm": 0.751105546951294,
+      "learning_rate": 1.7645648163546574e-06,
+      "loss": 0.0893,
+      "step": 11243
+    },
+    {
+      "epoch": 1.8217757615035644,
+      "grad_norm": 0.8291444182395935,
+      "learning_rate": 1.7641468734019795e-06,
+      "loss": 0.0998,
+      "step": 11244
+    },
+    {
+      "epoch": 1.8219377835385613,
+      "grad_norm": 0.735002875328064,
+      "learning_rate": 1.7637289529646273e-06,
+      "loss": 0.0898,
+      "step": 11245
+    },
+    {
+      "epoch": 1.822099805573558,
+      "grad_norm": 0.7483386993408203,
+      "learning_rate": 1.7633110550553867e-06,
+      "loss": 0.0863,
+      "step": 11246
+    },
+    {
+      "epoch": 1.8222618276085547,
+      "grad_norm": 0.7296152114868164,
+      "learning_rate": 1.7628931796870454e-06,
+      "loss": 0.0961,
+      "step": 11247
+    },
+    {
+      "epoch": 1.8224238496435516,
+      "grad_norm": 0.7359948754310608,
+      "learning_rate": 1.7624753268723882e-06,
+      "loss": 0.0927,
+      "step": 11248
+    },
+    {
+      "epoch": 1.8225858716785481,
+      "grad_norm": 0.8318410515785217,
+      "learning_rate": 1.7620574966242015e-06,
+      "loss": 0.1004,
+      "step": 11249
+    },
+    {
+      "epoch": 1.822747893713545,
+      "grad_norm": 0.9866887331008911,
+      "learning_rate": 1.7616396889552706e-06,
+      "loss": 0.1179,
+      "step": 11250
+    },
+    {
+      "epoch": 1.8229099157485418,
+      "grad_norm": 0.913759708404541,
+      "learning_rate": 1.7612219038783775e-06,
+      "loss": 0.105,
+      "step": 11251
+    },
+    {
+      "epoch": 1.8230719377835385,
+      "grad_norm": 0.7952281832695007,
+      "learning_rate": 1.7608041414063065e-06,
+      "loss": 0.0996,
+      "step": 11252
+    },
+    {
+      "epoch": 1.8232339598185354,
+      "grad_norm": 0.9138891100883484,
+      "learning_rate": 1.7603864015518392e-06,
+      "loss": 0.1103,
+      "step": 11253
+    },
+    {
+      "epoch": 1.8233959818535321,
+      "grad_norm": 0.9621608853340149,
+      "learning_rate": 1.7599686843277596e-06,
+      "loss": 0.1116,
+      "step": 11254
+    },
+    {
+      "epoch": 1.8235580038885288,
+      "grad_norm": 0.9279875755310059,
+      "learning_rate": 1.7595509897468466e-06,
+      "loss": 0.1177,
+      "step": 11255
+    },
+    {
+      "epoch": 1.8237200259235256,
+      "grad_norm": 0.7723210453987122,
+      "learning_rate": 1.7591333178218823e-06,
+      "loss": 0.1005,
+      "step": 11256
+    },
+    {
+      "epoch": 1.8238820479585223,
+      "grad_norm": 0.7563832998275757,
+      "learning_rate": 1.7587156685656442e-06,
+      "loss": 0.0964,
+      "step": 11257
+    },
+    {
+      "epoch": 1.8240440699935192,
+      "grad_norm": 0.7410457134246826,
+      "learning_rate": 1.7582980419909135e-06,
+      "loss": 0.0911,
+      "step": 11258
+    },
+    {
+      "epoch": 1.824206092028516,
+      "grad_norm": 0.8052392601966858,
+      "learning_rate": 1.7578804381104678e-06,
+      "loss": 0.1009,
+      "step": 11259
+    },
+    {
+      "epoch": 1.8243681140635126,
+      "grad_norm": 0.8688940405845642,
+      "learning_rate": 1.7574628569370855e-06,
+      "loss": 0.1155,
+      "step": 11260
+    },
+    {
+      "epoch": 1.8245301360985096,
+      "grad_norm": 0.8440880179405212,
+      "learning_rate": 1.757045298483542e-06,
+      "loss": 0.1034,
+      "step": 11261
+    },
+    {
+      "epoch": 1.824692158133506,
+      "grad_norm": 0.764716386795044,
+      "learning_rate": 1.756627762762614e-06,
+      "loss": 0.0984,
+      "step": 11262
+    },
+    {
+      "epoch": 1.824854180168503,
+      "grad_norm": 0.7497915029525757,
+      "learning_rate": 1.7562102497870787e-06,
+      "loss": 0.0923,
+      "step": 11263
+    },
+    {
+      "epoch": 1.8250162022034997,
+      "grad_norm": 0.8336397409439087,
+      "learning_rate": 1.7557927595697094e-06,
+      "loss": 0.1071,
+      "step": 11264
+    },
+    {
+      "epoch": 1.8251782242384964,
+      "grad_norm": 0.8785907626152039,
+      "learning_rate": 1.7553752921232809e-06,
+      "loss": 0.1079,
+      "step": 11265
+    },
+    {
+      "epoch": 1.8253402462734933,
+      "grad_norm": 0.8184788823127747,
+      "learning_rate": 1.7549578474605661e-06,
+      "loss": 0.0985,
+      "step": 11266
+    },
+    {
+      "epoch": 1.8255022683084898,
+      "grad_norm": 0.7712414860725403,
+      "learning_rate": 1.754540425594338e-06,
+      "loss": 0.095,
+      "step": 11267
+    },
+    {
+      "epoch": 1.8256642903434868,
+      "grad_norm": 0.8796233534812927,
+      "learning_rate": 1.754123026537369e-06,
+      "loss": 0.1065,
+      "step": 11268
+    },
+    {
+      "epoch": 1.8258263123784835,
+      "grad_norm": 0.9053205251693726,
+      "learning_rate": 1.7537056503024314e-06,
+      "loss": 0.0946,
+      "step": 11269
+    },
+    {
+      "epoch": 1.8259883344134802,
+      "grad_norm": 0.7080380916595459,
+      "learning_rate": 1.7532882969022941e-06,
+      "loss": 0.0892,
+      "step": 11270
+    },
+    {
+      "epoch": 1.8261503564484771,
+      "grad_norm": 0.8562909364700317,
+      "learning_rate": 1.7528709663497282e-06,
+      "loss": 0.1094,
+      "step": 11271
+    },
+    {
+      "epoch": 1.8263123784834736,
+      "grad_norm": 0.7841035723686218,
+      "learning_rate": 1.752453658657502e-06,
+      "loss": 0.0991,
+      "step": 11272
+    },
+    {
+      "epoch": 1.8264744005184705,
+      "grad_norm": 0.7752327919006348,
+      "learning_rate": 1.752036373838385e-06,
+      "loss": 0.0916,
+      "step": 11273
+    },
+    {
+      "epoch": 1.8266364225534673,
+      "grad_norm": 0.7416565418243408,
+      "learning_rate": 1.7516191119051456e-06,
+      "loss": 0.0904,
+      "step": 11274
+    },
+    {
+      "epoch": 1.826798444588464,
+      "grad_norm": 0.8184176683425903,
+      "learning_rate": 1.7512018728705498e-06,
+      "loss": 0.1011,
+      "step": 11275
+    },
+    {
+      "epoch": 1.826960466623461,
+      "grad_norm": 0.7727778553962708,
+      "learning_rate": 1.7507846567473643e-06,
+      "loss": 0.0946,
+      "step": 11276
+    },
+    {
+      "epoch": 1.8271224886584574,
+      "grad_norm": 0.7771718502044678,
+      "learning_rate": 1.7503674635483558e-06,
+      "loss": 0.0977,
+      "step": 11277
+    },
+    {
+      "epoch": 1.8272845106934543,
+      "grad_norm": 0.958016037940979,
+      "learning_rate": 1.749950293286289e-06,
+      "loss": 0.1002,
+      "step": 11278
+    },
+    {
+      "epoch": 1.827446532728451,
+      "grad_norm": 0.7841407060623169,
+      "learning_rate": 1.7495331459739278e-06,
+      "loss": 0.1045,
+      "step": 11279
+    },
+    {
+      "epoch": 1.8276085547634477,
+      "grad_norm": 0.7898461818695068,
+      "learning_rate": 1.7491160216240368e-06,
+      "loss": 0.0953,
+      "step": 11280
+    },
+    {
+      "epoch": 1.8277705767984447,
+      "grad_norm": 0.8376603126525879,
+      "learning_rate": 1.7486989202493775e-06,
+      "loss": 0.1077,
+      "step": 11281
+    },
+    {
+      "epoch": 1.8279325988334414,
+      "grad_norm": 0.7808676958084106,
+      "learning_rate": 1.7482818418627134e-06,
+      "loss": 0.0983,
+      "step": 11282
+    },
+    {
+      "epoch": 1.828094620868438,
+      "grad_norm": 0.8467119932174683,
+      "learning_rate": 1.7478647864768067e-06,
+      "loss": 0.1026,
+      "step": 11283
+    },
+    {
+      "epoch": 1.8282566429034348,
+      "grad_norm": 0.7723277807235718,
+      "learning_rate": 1.7474477541044165e-06,
+      "loss": 0.0954,
+      "step": 11284
+    },
+    {
+      "epoch": 1.8284186649384315,
+      "grad_norm": 0.8611959218978882,
+      "learning_rate": 1.7470307447583047e-06,
+      "loss": 0.0971,
+      "step": 11285
+    },
+    {
+      "epoch": 1.8285806869734285,
+      "grad_norm": 0.7999573945999146,
+      "learning_rate": 1.746613758451228e-06,
+      "loss": 0.0988,
+      "step": 11286
+    },
+    {
+      "epoch": 1.8287427090084252,
+      "grad_norm": 0.8502542972564697,
+      "learning_rate": 1.746196795195949e-06,
+      "loss": 0.0988,
+      "step": 11287
+    },
+    {
+      "epoch": 1.8289047310434219,
+      "grad_norm": 0.6951650977134705,
+      "learning_rate": 1.7457798550052232e-06,
+      "loss": 0.0888,
+      "step": 11288
+    },
+    {
+      "epoch": 1.8290667530784188,
+      "grad_norm": 0.7761609554290771,
+      "learning_rate": 1.7453629378918094e-06,
+      "loss": 0.097,
+      "step": 11289
+    },
+    {
+      "epoch": 1.8292287751134153,
+      "grad_norm": 0.882725715637207,
+      "learning_rate": 1.744946043868463e-06,
+      "loss": 0.1073,
+      "step": 11290
+    },
+    {
+      "epoch": 1.8293907971484122,
+      "grad_norm": 0.8751239776611328,
+      "learning_rate": 1.7445291729479397e-06,
+      "loss": 0.1095,
+      "step": 11291
+    },
+    {
+      "epoch": 1.829552819183409,
+      "grad_norm": 0.7810752987861633,
+      "learning_rate": 1.7441123251429968e-06,
+      "loss": 0.0916,
+      "step": 11292
+    },
+    {
+      "epoch": 1.8297148412184057,
+      "grad_norm": 0.8557392358779907,
+      "learning_rate": 1.7436955004663868e-06,
+      "loss": 0.0986,
+      "step": 11293
+    },
+    {
+      "epoch": 1.8298768632534026,
+      "grad_norm": 0.88213050365448,
+      "learning_rate": 1.7432786989308648e-06,
+      "loss": 0.1046,
+      "step": 11294
+    },
+    {
+      "epoch": 1.830038885288399,
+      "grad_norm": 0.8699725270271301,
+      "learning_rate": 1.7428619205491831e-06,
+      "loss": 0.1018,
+      "step": 11295
+    },
+    {
+      "epoch": 1.830200907323396,
+      "grad_norm": 0.7817003726959229,
+      "learning_rate": 1.7424451653340934e-06,
+      "loss": 0.0926,
+      "step": 11296
+    },
+    {
+      "epoch": 1.8303629293583927,
+      "grad_norm": 0.8012980818748474,
+      "learning_rate": 1.7420284332983495e-06,
+      "loss": 0.1063,
+      "step": 11297
+    },
+    {
+      "epoch": 1.8305249513933894,
+      "grad_norm": 0.7765569686889648,
+      "learning_rate": 1.7416117244547014e-06,
+      "loss": 0.0899,
+      "step": 11298
+    },
+    {
+      "epoch": 1.8306869734283864,
+      "grad_norm": 0.8100259900093079,
+      "learning_rate": 1.7411950388158987e-06,
+      "loss": 0.1055,
+      "step": 11299
+    },
+    {
+      "epoch": 1.8308489954633829,
+      "grad_norm": 0.996043860912323,
+      "learning_rate": 1.7407783763946911e-06,
+      "loss": 0.1143,
+      "step": 11300
+    },
+    {
+      "epoch": 1.8310110174983798,
+      "grad_norm": 0.9620465636253357,
+      "learning_rate": 1.7403617372038293e-06,
+      "loss": 0.1167,
+      "step": 11301
+    },
+    {
+      "epoch": 1.8311730395333765,
+      "grad_norm": 0.7914530038833618,
+      "learning_rate": 1.7399451212560593e-06,
+      "loss": 0.1029,
+      "step": 11302
+    },
+    {
+      "epoch": 1.8313350615683732,
+      "grad_norm": 0.8515611886978149,
+      "learning_rate": 1.7395285285641292e-06,
+      "loss": 0.1059,
+      "step": 11303
+    },
+    {
+      "epoch": 1.8314970836033702,
+      "grad_norm": 0.8178187608718872,
+      "learning_rate": 1.7391119591407863e-06,
+      "loss": 0.0974,
+      "step": 11304
+    },
+    {
+      "epoch": 1.8316591056383669,
+      "grad_norm": 0.8800912499427795,
+      "learning_rate": 1.7386954129987754e-06,
+      "loss": 0.1066,
+      "step": 11305
+    },
+    {
+      "epoch": 1.8318211276733636,
+      "grad_norm": 0.8402878642082214,
+      "learning_rate": 1.7382788901508426e-06,
+      "loss": 0.1038,
+      "step": 11306
+    },
+    {
+      "epoch": 1.8319831497083603,
+      "grad_norm": 0.8607720732688904,
+      "learning_rate": 1.7378623906097333e-06,
+      "loss": 0.1116,
+      "step": 11307
+    },
+    {
+      "epoch": 1.832145171743357,
+      "grad_norm": 0.8464848399162292,
+      "learning_rate": 1.7374459143881899e-06,
+      "loss": 0.1,
+      "step": 11308
+    },
+    {
+      "epoch": 1.832307193778354,
+      "grad_norm": 0.7194477319717407,
+      "learning_rate": 1.737029461498957e-06,
+      "loss": 0.0803,
+      "step": 11309
+    },
+    {
+      "epoch": 1.8324692158133506,
+      "grad_norm": 0.9045358300209045,
+      "learning_rate": 1.7366130319547747e-06,
+      "loss": 0.1053,
+      "step": 11310
+    },
+    {
+      "epoch": 1.8326312378483474,
+      "grad_norm": 0.8817188143730164,
+      "learning_rate": 1.736196625768387e-06,
+      "loss": 0.1033,
+      "step": 11311
+    },
+    {
+      "epoch": 1.8327932598833443,
+      "grad_norm": 0.7992186546325684,
+      "learning_rate": 1.735780242952534e-06,
+      "loss": 0.0999,
+      "step": 11312
+    },
+    {
+      "epoch": 1.8329552819183408,
+      "grad_norm": 0.711328387260437,
+      "learning_rate": 1.7353638835199568e-06,
+      "loss": 0.0875,
+      "step": 11313
+    },
+    {
+      "epoch": 1.8331173039533377,
+      "grad_norm": 0.8536904454231262,
+      "learning_rate": 1.7349475474833938e-06,
+      "loss": 0.1047,
+      "step": 11314
+    },
+    {
+      "epoch": 1.8332793259883344,
+      "grad_norm": 0.7888259887695312,
+      "learning_rate": 1.7345312348555843e-06,
+      "loss": 0.1089,
+      "step": 11315
+    },
+    {
+      "epoch": 1.8334413480233311,
+      "grad_norm": 0.8059415221214294,
+      "learning_rate": 1.7341149456492672e-06,
+      "loss": 0.0995,
+      "step": 11316
+    },
+    {
+      "epoch": 1.833603370058328,
+      "grad_norm": 0.8481978178024292,
+      "learning_rate": 1.733698679877179e-06,
+      "loss": 0.1035,
+      "step": 11317
+    },
+    {
+      "epoch": 1.8337653920933246,
+      "grad_norm": 0.72404545545578,
+      "learning_rate": 1.7332824375520574e-06,
+      "loss": 0.0971,
+      "step": 11318
+    },
+    {
+      "epoch": 1.8339274141283215,
+      "grad_norm": 0.8225178718566895,
+      "learning_rate": 1.7328662186866373e-06,
+      "loss": 0.0951,
+      "step": 11319
+    },
+    {
+      "epoch": 1.8340894361633182,
+      "grad_norm": 0.7030364871025085,
+      "learning_rate": 1.7324500232936536e-06,
+      "loss": 0.0844,
+      "step": 11320
+    },
+    {
+      "epoch": 1.834251458198315,
+      "grad_norm": 0.9015457630157471,
+      "learning_rate": 1.7320338513858425e-06,
+      "loss": 0.1048,
+      "step": 11321
+    },
+    {
+      "epoch": 1.8344134802333119,
+      "grad_norm": 0.8932374715805054,
+      "learning_rate": 1.731617702975938e-06,
+      "loss": 0.1184,
+      "step": 11322
+    },
+    {
+      "epoch": 1.8345755022683083,
+      "grad_norm": 1.008551836013794,
+      "learning_rate": 1.7312015780766714e-06,
+      "loss": 0.1336,
+      "step": 11323
+    },
+    {
+      "epoch": 1.8347375243033053,
+      "grad_norm": 0.7870475053787231,
+      "learning_rate": 1.7307854767007756e-06,
+      "loss": 0.1007,
+      "step": 11324
+    },
+    {
+      "epoch": 1.834899546338302,
+      "grad_norm": 0.7753018736839294,
+      "learning_rate": 1.7303693988609837e-06,
+      "loss": 0.0904,
+      "step": 11325
+    },
+    {
+      "epoch": 1.8350615683732987,
+      "grad_norm": 0.9246848225593567,
+      "learning_rate": 1.7299533445700253e-06,
+      "loss": 0.1144,
+      "step": 11326
+    },
+    {
+      "epoch": 1.8352235904082956,
+      "grad_norm": 0.7962560057640076,
+      "learning_rate": 1.7295373138406318e-06,
+      "loss": 0.1006,
+      "step": 11327
+    },
+    {
+      "epoch": 1.8353856124432921,
+      "grad_norm": 0.8699910044670105,
+      "learning_rate": 1.7291213066855312e-06,
+      "loss": 0.1003,
+      "step": 11328
+    },
+    {
+      "epoch": 1.835547634478289,
+      "grad_norm": 0.8488613367080688,
+      "learning_rate": 1.7287053231174528e-06,
+      "loss": 0.1104,
+      "step": 11329
+    },
+    {
+      "epoch": 1.8357096565132858,
+      "grad_norm": 0.8561351895332336,
+      "learning_rate": 1.7282893631491253e-06,
+      "loss": 0.1013,
+      "step": 11330
+    },
+    {
+      "epoch": 1.8358716785482825,
+      "grad_norm": 0.7828042507171631,
+      "learning_rate": 1.7278734267932764e-06,
+      "loss": 0.0941,
+      "step": 11331
+    },
+    {
+      "epoch": 1.8360337005832794,
+      "grad_norm": 0.916395366191864,
+      "learning_rate": 1.7274575140626318e-06,
+      "loss": 0.1124,
+      "step": 11332
+    },
+    {
+      "epoch": 1.8361957226182761,
+      "grad_norm": 0.8568091988563538,
+      "learning_rate": 1.7270416249699179e-06,
+      "loss": 0.0995,
+      "step": 11333
+    },
+    {
+      "epoch": 1.8363577446532728,
+      "grad_norm": 0.7647148966789246,
+      "learning_rate": 1.7266257595278591e-06,
+      "loss": 0.0951,
+      "step": 11334
+    },
+    {
+      "epoch": 1.8365197666882696,
+      "grad_norm": 0.8157481551170349,
+      "learning_rate": 1.726209917749181e-06,
+      "loss": 0.0979,
+      "step": 11335
+    },
+    {
+      "epoch": 1.8366817887232663,
+      "grad_norm": 0.8966127634048462,
+      "learning_rate": 1.725794099646607e-06,
+      "loss": 0.1024,
+      "step": 11336
+    },
+    {
+      "epoch": 1.8368438107582632,
+      "grad_norm": 0.8767910003662109,
+      "learning_rate": 1.72537830523286e-06,
+      "loss": 0.1108,
+      "step": 11337
+    },
+    {
+      "epoch": 1.83700583279326,
+      "grad_norm": 0.8303874731063843,
+      "learning_rate": 1.7249625345206623e-06,
+      "loss": 0.0929,
+      "step": 11338
+    },
+    {
+      "epoch": 1.8371678548282566,
+      "grad_norm": 0.7946867346763611,
+      "learning_rate": 1.7245467875227345e-06,
+      "loss": 0.1038,
+      "step": 11339
+    },
+    {
+      "epoch": 1.8373298768632536,
+      "grad_norm": 0.7958531379699707,
+      "learning_rate": 1.7241310642517998e-06,
+      "loss": 0.0946,
+      "step": 11340
+    },
+    {
+      "epoch": 1.83749189889825,
+      "grad_norm": 0.7723376750946045,
+      "learning_rate": 1.7237153647205762e-06,
+      "loss": 0.1002,
+      "step": 11341
+    },
+    {
+      "epoch": 1.837653920933247,
+      "grad_norm": 1.005839228630066,
+      "learning_rate": 1.7232996889417846e-06,
+      "loss": 0.109,
+      "step": 11342
+    },
+    {
+      "epoch": 1.8378159429682437,
+      "grad_norm": 0.7882110476493835,
+      "learning_rate": 1.7228840369281424e-06,
+      "loss": 0.0978,
+      "step": 11343
+    },
+    {
+      "epoch": 1.8379779650032404,
+      "grad_norm": 0.6778146028518677,
+      "learning_rate": 1.7224684086923677e-06,
+      "loss": 0.0867,
+      "step": 11344
+    },
+    {
+      "epoch": 1.8381399870382373,
+      "grad_norm": 0.8371578454971313,
+      "learning_rate": 1.722052804247179e-06,
+      "loss": 0.104,
+      "step": 11345
+    },
+    {
+      "epoch": 1.8383020090732338,
+      "grad_norm": 0.684536874294281,
+      "learning_rate": 1.7216372236052914e-06,
+      "loss": 0.0873,
+      "step": 11346
+    },
+    {
+      "epoch": 1.8384640311082308,
+      "grad_norm": 0.777190089225769,
+      "learning_rate": 1.7212216667794213e-06,
+      "loss": 0.1056,
+      "step": 11347
+    },
+    {
+      "epoch": 1.8386260531432275,
+      "grad_norm": 0.7780445218086243,
+      "learning_rate": 1.7208061337822828e-06,
+      "loss": 0.0936,
+      "step": 11348
+    },
+    {
+      "epoch": 1.8387880751782242,
+      "grad_norm": 0.8263536095619202,
+      "learning_rate": 1.7203906246265921e-06,
+      "loss": 0.095,
+      "step": 11349
+    },
+    {
+      "epoch": 1.8389500972132211,
+      "grad_norm": 0.7346975803375244,
+      "learning_rate": 1.7199751393250614e-06,
+      "loss": 0.0939,
+      "step": 11350
+    },
+    {
+      "epoch": 1.8391121192482176,
+      "grad_norm": 0.8008426427841187,
+      "learning_rate": 1.719559677890404e-06,
+      "loss": 0.0932,
+      "step": 11351
+    },
+    {
+      "epoch": 1.8392741412832145,
+      "grad_norm": 0.8390111923217773,
+      "learning_rate": 1.7191442403353314e-06,
+      "loss": 0.0961,
+      "step": 11352
+    },
+    {
+      "epoch": 1.8394361633182112,
+      "grad_norm": 0.8071438074111938,
+      "learning_rate": 1.7187288266725549e-06,
+      "loss": 0.094,
+      "step": 11353
+    },
+    {
+      "epoch": 1.839598185353208,
+      "grad_norm": 0.8307203054428101,
+      "learning_rate": 1.7183134369147866e-06,
+      "loss": 0.1013,
+      "step": 11354
+    },
+    {
+      "epoch": 1.839760207388205,
+      "grad_norm": 0.8359047174453735,
+      "learning_rate": 1.717898071074735e-06,
+      "loss": 0.1013,
+      "step": 11355
+    },
+    {
+      "epoch": 1.8399222294232016,
+      "grad_norm": 0.925969660282135,
+      "learning_rate": 1.71748272916511e-06,
+      "loss": 0.1027,
+      "step": 11356
+    },
+    {
+      "epoch": 1.8400842514581983,
+      "grad_norm": 0.7569020986557007,
+      "learning_rate": 1.7170674111986202e-06,
+      "loss": 0.0964,
+      "step": 11357
+    },
+    {
+      "epoch": 1.840246273493195,
+      "grad_norm": 0.8798254728317261,
+      "learning_rate": 1.716652117187972e-06,
+      "loss": 0.105,
+      "step": 11358
+    },
+    {
+      "epoch": 1.8404082955281917,
+      "grad_norm": 0.757815957069397,
+      "learning_rate": 1.7162368471458738e-06,
+      "loss": 0.0955,
+      "step": 11359
+    },
+    {
+      "epoch": 1.8405703175631887,
+      "grad_norm": 0.8111533522605896,
+      "learning_rate": 1.7158216010850318e-06,
+      "loss": 0.0971,
+      "step": 11360
+    },
+    {
+      "epoch": 1.8407323395981854,
+      "grad_norm": 0.793456494808197,
+      "learning_rate": 1.7154063790181507e-06,
+      "loss": 0.1047,
+      "step": 11361
+    },
+    {
+      "epoch": 1.840894361633182,
+      "grad_norm": 0.7648953199386597,
+      "learning_rate": 1.7149911809579361e-06,
+      "loss": 0.0951,
+      "step": 11362
+    },
+    {
+      "epoch": 1.841056383668179,
+      "grad_norm": 0.7821177244186401,
+      "learning_rate": 1.7145760069170905e-06,
+      "loss": 0.0973,
+      "step": 11363
+    },
+    {
+      "epoch": 1.8412184057031755,
+      "grad_norm": 0.8774672150611877,
+      "learning_rate": 1.7141608569083195e-06,
+      "loss": 0.1019,
+      "step": 11364
+    },
+    {
+      "epoch": 1.8413804277381725,
+      "grad_norm": 0.815719485282898,
+      "learning_rate": 1.7137457309443245e-06,
+      "loss": 0.0915,
+      "step": 11365
+    },
+    {
+      "epoch": 1.8415424497731692,
+      "grad_norm": 0.8151124715805054,
+      "learning_rate": 1.7133306290378077e-06,
+      "loss": 0.094,
+      "step": 11366
+    },
+    {
+      "epoch": 1.8417044718081659,
+      "grad_norm": 0.6767386794090271,
+      "learning_rate": 1.7129155512014692e-06,
+      "loss": 0.0797,
+      "step": 11367
+    },
+    {
+      "epoch": 1.8418664938431628,
+      "grad_norm": 0.8355568051338196,
+      "learning_rate": 1.7125004974480102e-06,
+      "loss": 0.0956,
+      "step": 11368
+    },
+    {
+      "epoch": 1.8420285158781593,
+      "grad_norm": 0.9013076424598694,
+      "learning_rate": 1.7120854677901312e-06,
+      "loss": 0.1056,
+      "step": 11369
+    },
+    {
+      "epoch": 1.8421905379131562,
+      "grad_norm": 0.7785685062408447,
+      "learning_rate": 1.7116704622405295e-06,
+      "loss": 0.0867,
+      "step": 11370
+    },
+    {
+      "epoch": 1.842352559948153,
+      "grad_norm": 0.7849324345588684,
+      "learning_rate": 1.7112554808119043e-06,
+      "loss": 0.0983,
+      "step": 11371
+    },
+    {
+      "epoch": 1.8425145819831497,
+      "grad_norm": 0.9096629023551941,
+      "learning_rate": 1.7108405235169511e-06,
+      "loss": 0.1079,
+      "step": 11372
+    },
+    {
+      "epoch": 1.8426766040181466,
+      "grad_norm": 0.7472807765007019,
+      "learning_rate": 1.71042559036837e-06,
+      "loss": 0.0946,
+      "step": 11373
+    },
+    {
+      "epoch": 1.842838626053143,
+      "grad_norm": 0.6788498163223267,
+      "learning_rate": 1.7100106813788544e-06,
+      "loss": 0.0785,
+      "step": 11374
+    },
+    {
+      "epoch": 1.84300064808814,
+      "grad_norm": 0.8583519458770752,
+      "learning_rate": 1.7095957965611008e-06,
+      "loss": 0.1001,
+      "step": 11375
+    },
+    {
+      "epoch": 1.8431626701231367,
+      "grad_norm": 0.9140331149101257,
+      "learning_rate": 1.7091809359278025e-06,
+      "loss": 0.1117,
+      "step": 11376
+    },
+    {
+      "epoch": 1.8433246921581334,
+      "grad_norm": 0.8352115154266357,
+      "learning_rate": 1.7087660994916533e-06,
+      "loss": 0.1087,
+      "step": 11377
+    },
+    {
+      "epoch": 1.8434867141931304,
+      "grad_norm": 0.8460627198219299,
+      "learning_rate": 1.7083512872653477e-06,
+      "loss": 0.1032,
+      "step": 11378
+    },
+    {
+      "epoch": 1.8436487362281269,
+      "grad_norm": 0.8629457950592041,
+      "learning_rate": 1.7079364992615766e-06,
+      "loss": 0.1015,
+      "step": 11379
+    },
+    {
+      "epoch": 1.8438107582631238,
+      "grad_norm": 0.8817080855369568,
+      "learning_rate": 1.7075217354930324e-06,
+      "loss": 0.1085,
+      "step": 11380
+    },
+    {
+      "epoch": 1.8439727802981205,
+      "grad_norm": 0.9188164472579956,
+      "learning_rate": 1.7071069959724046e-06,
+      "loss": 0.0987,
+      "step": 11381
+    },
+    {
+      "epoch": 1.8441348023331172,
+      "grad_norm": 0.7526326179504395,
+      "learning_rate": 1.7066922807123834e-06,
+      "loss": 0.0948,
+      "step": 11382
+    },
+    {
+      "epoch": 1.8442968243681142,
+      "grad_norm": 0.7177049517631531,
+      "learning_rate": 1.7062775897256593e-06,
+      "loss": 0.0892,
+      "step": 11383
+    },
+    {
+      "epoch": 1.8444588464031109,
+      "grad_norm": 0.8121328353881836,
+      "learning_rate": 1.7058629230249207e-06,
+      "loss": 0.0961,
+      "step": 11384
+    },
+    {
+      "epoch": 1.8446208684381076,
+      "grad_norm": 0.7545210123062134,
+      "learning_rate": 1.7054482806228543e-06,
+      "loss": 0.0927,
+      "step": 11385
+    },
+    {
+      "epoch": 1.8447828904731045,
+      "grad_norm": 0.8250703811645508,
+      "learning_rate": 1.7050336625321484e-06,
+      "loss": 0.095,
+      "step": 11386
+    },
+    {
+      "epoch": 1.844944912508101,
+      "grad_norm": 0.772592306137085,
+      "learning_rate": 1.7046190687654873e-06,
+      "loss": 0.0977,
+      "step": 11387
+    },
+    {
+      "epoch": 1.845106934543098,
+      "grad_norm": 0.7452101111412048,
+      "learning_rate": 1.704204499335559e-06,
+      "loss": 0.0897,
+      "step": 11388
+    },
+    {
+      "epoch": 1.8452689565780946,
+      "grad_norm": 0.7722578048706055,
+      "learning_rate": 1.703789954255047e-06,
+      "loss": 0.0977,
+      "step": 11389
+    },
+    {
+      "epoch": 1.8454309786130914,
+      "grad_norm": 0.8865172863006592,
+      "learning_rate": 1.7033754335366356e-06,
+      "loss": 0.1031,
+      "step": 11390
+    },
+    {
+      "epoch": 1.8455930006480883,
+      "grad_norm": 0.8430839776992798,
+      "learning_rate": 1.7029609371930076e-06,
+      "loss": 0.1073,
+      "step": 11391
+    },
+    {
+      "epoch": 1.8457550226830848,
+      "grad_norm": 0.8359355330467224,
+      "learning_rate": 1.7025464652368464e-06,
+      "loss": 0.101,
+      "step": 11392
+    },
+    {
+      "epoch": 1.8459170447180817,
+      "grad_norm": 0.8882021307945251,
+      "learning_rate": 1.7021320176808343e-06,
+      "loss": 0.1072,
+      "step": 11393
+    },
+    {
+      "epoch": 1.8460790667530784,
+      "grad_norm": 0.7827454209327698,
+      "learning_rate": 1.701717594537651e-06,
+      "loss": 0.0927,
+      "step": 11394
+    },
+    {
+      "epoch": 1.8462410887880751,
+      "grad_norm": 0.8166926503181458,
+      "learning_rate": 1.7013031958199783e-06,
+      "loss": 0.1026,
+      "step": 11395
+    },
+    {
+      "epoch": 1.846403110823072,
+      "grad_norm": 0.7511534690856934,
+      "learning_rate": 1.7008888215404933e-06,
+      "loss": 0.0942,
+      "step": 11396
+    },
+    {
+      "epoch": 1.8465651328580686,
+      "grad_norm": 0.8353126645088196,
+      "learning_rate": 1.7004744717118777e-06,
+      "loss": 0.1063,
+      "step": 11397
+    },
+    {
+      "epoch": 1.8467271548930655,
+      "grad_norm": 0.6988396644592285,
+      "learning_rate": 1.7000601463468088e-06,
+      "loss": 0.0877,
+      "step": 11398
+    },
+    {
+      "epoch": 1.8468891769280622,
+      "grad_norm": 0.8091744184494019,
+      "learning_rate": 1.6996458454579632e-06,
+      "loss": 0.0984,
+      "step": 11399
+    },
+    {
+      "epoch": 1.847051198963059,
+      "grad_norm": 0.7793344855308533,
+      "learning_rate": 1.6992315690580178e-06,
+      "loss": 0.093,
+      "step": 11400
+    },
+    {
+      "epoch": 1.8472132209980558,
+      "grad_norm": 0.7006025910377502,
+      "learning_rate": 1.6988173171596479e-06,
+      "loss": 0.0805,
+      "step": 11401
+    },
+    {
+      "epoch": 1.8473752430330523,
+      "grad_norm": 0.7687976956367493,
+      "learning_rate": 1.6984030897755304e-06,
+      "loss": 0.0942,
+      "step": 11402
+    },
+    {
+      "epoch": 1.8475372650680493,
+      "grad_norm": 0.8496354222297668,
+      "learning_rate": 1.697988886918338e-06,
+      "loss": 0.1107,
+      "step": 11403
+    },
+    {
+      "epoch": 1.847699287103046,
+      "grad_norm": 0.8020095229148865,
+      "learning_rate": 1.6975747086007454e-06,
+      "loss": 0.0906,
+      "step": 11404
+    },
+    {
+      "epoch": 1.8478613091380427,
+      "grad_norm": 0.8590508699417114,
+      "learning_rate": 1.6971605548354244e-06,
+      "loss": 0.0976,
+      "step": 11405
+    },
+    {
+      "epoch": 1.8480233311730396,
+      "grad_norm": 0.8174715042114258,
+      "learning_rate": 1.6967464256350468e-06,
+      "loss": 0.0967,
+      "step": 11406
+    },
+    {
+      "epoch": 1.8481853532080363,
+      "grad_norm": 0.8774664998054504,
+      "learning_rate": 1.6963323210122856e-06,
+      "loss": 0.0971,
+      "step": 11407
+    },
+    {
+      "epoch": 1.848347375243033,
+      "grad_norm": 0.8936482667922974,
+      "learning_rate": 1.6959182409798111e-06,
+      "loss": 0.1196,
+      "step": 11408
+    },
+    {
+      "epoch": 1.8485093972780298,
+      "grad_norm": 0.7813578248023987,
+      "learning_rate": 1.6955041855502918e-06,
+      "loss": 0.1045,
+      "step": 11409
+    },
+    {
+      "epoch": 1.8486714193130265,
+      "grad_norm": 0.8374517560005188,
+      "learning_rate": 1.695090154736398e-06,
+      "loss": 0.1012,
+      "step": 11410
+    },
+    {
+      "epoch": 1.8488334413480234,
+      "grad_norm": 0.7886912822723389,
+      "learning_rate": 1.694676148550797e-06,
+      "loss": 0.0903,
+      "step": 11411
+    },
+    {
+      "epoch": 1.8489954633830201,
+      "grad_norm": 0.8474147915840149,
+      "learning_rate": 1.6942621670061574e-06,
+      "loss": 0.1001,
+      "step": 11412
+    },
+    {
+      "epoch": 1.8491574854180168,
+      "grad_norm": 0.7622694373130798,
+      "learning_rate": 1.693848210115146e-06,
+      "loss": 0.0996,
+      "step": 11413
+    },
+    {
+      "epoch": 1.8493195074530138,
+      "grad_norm": 0.7860474586486816,
+      "learning_rate": 1.693434277890428e-06,
+      "loss": 0.0952,
+      "step": 11414
+    },
+    {
+      "epoch": 1.8494815294880103,
+      "grad_norm": 0.8053168058395386,
+      "learning_rate": 1.693020370344669e-06,
+      "loss": 0.0991,
+      "step": 11415
+    },
+    {
+      "epoch": 1.8496435515230072,
+      "grad_norm": 0.769270658493042,
+      "learning_rate": 1.692606487490534e-06,
+      "loss": 0.1014,
+      "step": 11416
+    },
+    {
+      "epoch": 1.849805573558004,
+      "grad_norm": 0.7316917181015015,
+      "learning_rate": 1.6921926293406874e-06,
+      "loss": 0.0897,
+      "step": 11417
+    },
+    {
+      "epoch": 1.8499675955930006,
+      "grad_norm": 0.7723559737205505,
+      "learning_rate": 1.6917787959077907e-06,
+      "loss": 0.0973,
+      "step": 11418
+    },
+    {
+      "epoch": 1.8501296176279975,
+      "grad_norm": 0.8190063238143921,
+      "learning_rate": 1.6913649872045076e-06,
+      "loss": 0.1082,
+      "step": 11419
+    },
+    {
+      "epoch": 1.850291639662994,
+      "grad_norm": 0.8806280493736267,
+      "learning_rate": 1.6909512032434984e-06,
+      "loss": 0.1103,
+      "step": 11420
+    },
+    {
+      "epoch": 1.850453661697991,
+      "grad_norm": 0.7780027985572815,
+      "learning_rate": 1.6905374440374245e-06,
+      "loss": 0.1017,
+      "step": 11421
+    },
+    {
+      "epoch": 1.8506156837329877,
+      "grad_norm": 0.9774267077445984,
+      "learning_rate": 1.6901237095989464e-06,
+      "loss": 0.1157,
+      "step": 11422
+    },
+    {
+      "epoch": 1.8507777057679844,
+      "grad_norm": 0.7308449745178223,
+      "learning_rate": 1.689709999940723e-06,
+      "loss": 0.0916,
+      "step": 11423
+    },
+    {
+      "epoch": 1.8509397278029813,
+      "grad_norm": 0.7397428154945374,
+      "learning_rate": 1.6892963150754128e-06,
+      "loss": 0.0873,
+      "step": 11424
+    },
+    {
+      "epoch": 1.8511017498379778,
+      "grad_norm": 0.8941960334777832,
+      "learning_rate": 1.688882655015672e-06,
+      "loss": 0.102,
+      "step": 11425
+    },
+    {
+      "epoch": 1.8512637718729748,
+      "grad_norm": 0.77532958984375,
+      "learning_rate": 1.6884690197741608e-06,
+      "loss": 0.0966,
+      "step": 11426
+    },
+    {
+      "epoch": 1.8514257939079715,
+      "grad_norm": 0.834740936756134,
+      "learning_rate": 1.6880554093635331e-06,
+      "loss": 0.1077,
+      "step": 11427
+    },
+    {
+      "epoch": 1.8515878159429682,
+      "grad_norm": 0.8465557098388672,
+      "learning_rate": 1.6876418237964453e-06,
+      "loss": 0.1014,
+      "step": 11428
+    },
+    {
+      "epoch": 1.851749837977965,
+      "grad_norm": 0.8378772139549255,
+      "learning_rate": 1.6872282630855519e-06,
+      "loss": 0.0937,
+      "step": 11429
+    },
+    {
+      "epoch": 1.8519118600129616,
+      "grad_norm": 0.9061734676361084,
+      "learning_rate": 1.6868147272435057e-06,
+      "loss": 0.1131,
+      "step": 11430
+    },
+    {
+      "epoch": 1.8520738820479585,
+      "grad_norm": 0.8114215135574341,
+      "learning_rate": 1.6864012162829624e-06,
+      "loss": 0.0961,
+      "step": 11431
+    },
+    {
+      "epoch": 1.8522359040829552,
+      "grad_norm": 0.8266593217849731,
+      "learning_rate": 1.6859877302165723e-06,
+      "loss": 0.0923,
+      "step": 11432
+    },
+    {
+      "epoch": 1.852397926117952,
+      "grad_norm": 0.7605752944946289,
+      "learning_rate": 1.6855742690569881e-06,
+      "loss": 0.098,
+      "step": 11433
+    },
+    {
+      "epoch": 1.8525599481529489,
+      "grad_norm": 0.77652508020401,
+      "learning_rate": 1.6851608328168589e-06,
+      "loss": 0.0995,
+      "step": 11434
+    },
+    {
+      "epoch": 1.8527219701879456,
+      "grad_norm": 0.7098603248596191,
+      "learning_rate": 1.6847474215088382e-06,
+      "loss": 0.0844,
+      "step": 11435
+    },
+    {
+      "epoch": 1.8528839922229423,
+      "grad_norm": 0.9365832209587097,
+      "learning_rate": 1.6843340351455728e-06,
+      "loss": 0.1145,
+      "step": 11436
+    },
+    {
+      "epoch": 1.8530460142579392,
+      "grad_norm": 0.8150047659873962,
+      "learning_rate": 1.6839206737397126e-06,
+      "loss": 0.0992,
+      "step": 11437
+    },
+    {
+      "epoch": 1.8532080362929357,
+      "grad_norm": 0.7815777659416199,
+      "learning_rate": 1.6835073373039045e-06,
+      "loss": 0.092,
+      "step": 11438
+    },
+    {
+      "epoch": 1.8533700583279327,
+      "grad_norm": 0.8175069689750671,
+      "learning_rate": 1.6830940258507955e-06,
+      "loss": 0.0988,
+      "step": 11439
+    },
+    {
+      "epoch": 1.8535320803629294,
+      "grad_norm": 0.8436422944068909,
+      "learning_rate": 1.6826807393930334e-06,
+      "loss": 0.0944,
+      "step": 11440
+    },
+    {
+      "epoch": 1.853694102397926,
+      "grad_norm": 0.6752289533615112,
+      "learning_rate": 1.682267477943262e-06,
+      "loss": 0.0803,
+      "step": 11441
+    },
+    {
+      "epoch": 1.853856124432923,
+      "grad_norm": 0.7708499431610107,
+      "learning_rate": 1.6818542415141273e-06,
+      "loss": 0.0933,
+      "step": 11442
+    },
+    {
+      "epoch": 1.8540181464679195,
+      "grad_norm": 0.8612736463546753,
+      "learning_rate": 1.6814410301182732e-06,
+      "loss": 0.0977,
+      "step": 11443
+    },
+    {
+      "epoch": 1.8541801685029164,
+      "grad_norm": 0.8443825840950012,
+      "learning_rate": 1.6810278437683419e-06,
+      "loss": 0.1044,
+      "step": 11444
+    },
+    {
+      "epoch": 1.8543421905379132,
+      "grad_norm": 0.8466036319732666,
+      "learning_rate": 1.680614682476977e-06,
+      "loss": 0.0982,
+      "step": 11445
+    },
+    {
+      "epoch": 1.8545042125729099,
+      "grad_norm": 0.9107270240783691,
+      "learning_rate": 1.6802015462568205e-06,
+      "loss": 0.1083,
+      "step": 11446
+    },
+    {
+      "epoch": 1.8546662346079068,
+      "grad_norm": 0.7344313859939575,
+      "learning_rate": 1.6797884351205123e-06,
+      "loss": 0.0926,
+      "step": 11447
+    },
+    {
+      "epoch": 1.8548282566429033,
+      "grad_norm": 0.8583900332450867,
+      "learning_rate": 1.6793753490806939e-06,
+      "loss": 0.1143,
+      "step": 11448
+    },
+    {
+      "epoch": 1.8549902786779002,
+      "grad_norm": 0.7386242151260376,
+      "learning_rate": 1.678962288150003e-06,
+      "loss": 0.0855,
+      "step": 11449
+    },
+    {
+      "epoch": 1.855152300712897,
+      "grad_norm": 0.7551776766777039,
+      "learning_rate": 1.678549252341079e-06,
+      "loss": 0.0893,
+      "step": 11450
+    },
+    {
+      "epoch": 1.8553143227478937,
+      "grad_norm": 0.8036359548568726,
+      "learning_rate": 1.6781362416665602e-06,
+      "loss": 0.0918,
+      "step": 11451
+    },
+    {
+      "epoch": 1.8554763447828906,
+      "grad_norm": 0.7629950046539307,
+      "learning_rate": 1.6777232561390844e-06,
+      "loss": 0.1067,
+      "step": 11452
+    },
+    {
+      "epoch": 1.855638366817887,
+      "grad_norm": 0.7417677044868469,
+      "learning_rate": 1.6773102957712866e-06,
+      "loss": 0.087,
+      "step": 11453
+    },
+    {
+      "epoch": 1.855800388852884,
+      "grad_norm": 0.7128724455833435,
+      "learning_rate": 1.6768973605758021e-06,
+      "loss": 0.091,
+      "step": 11454
+    },
+    {
+      "epoch": 1.8559624108878807,
+      "grad_norm": 0.721200704574585,
+      "learning_rate": 1.6764844505652677e-06,
+      "loss": 0.0895,
+      "step": 11455
+    },
+    {
+      "epoch": 1.8561244329228774,
+      "grad_norm": 0.7439916729927063,
+      "learning_rate": 1.6760715657523158e-06,
+      "loss": 0.0916,
+      "step": 11456
+    },
+    {
+      "epoch": 1.8562864549578744,
+      "grad_norm": 0.8428785800933838,
+      "learning_rate": 1.6756587061495805e-06,
+      "loss": 0.1077,
+      "step": 11457
+    },
+    {
+      "epoch": 1.856448476992871,
+      "grad_norm": 0.7453662753105164,
+      "learning_rate": 1.6752458717696928e-06,
+      "loss": 0.089,
+      "step": 11458
+    },
+    {
+      "epoch": 1.8566104990278678,
+      "grad_norm": 0.8296511173248291,
+      "learning_rate": 1.6748330626252862e-06,
+      "loss": 0.1011,
+      "step": 11459
+    },
+    {
+      "epoch": 1.8567725210628645,
+      "grad_norm": 0.8146061301231384,
+      "learning_rate": 1.674420278728991e-06,
+      "loss": 0.0924,
+      "step": 11460
+    },
+    {
+      "epoch": 1.8569345430978612,
+      "grad_norm": 0.8605503439903259,
+      "learning_rate": 1.674007520093438e-06,
+      "loss": 0.0933,
+      "step": 11461
+    },
+    {
+      "epoch": 1.8570965651328581,
+      "grad_norm": 0.820659875869751,
+      "learning_rate": 1.6735947867312553e-06,
+      "loss": 0.0988,
+      "step": 11462
+    },
+    {
+      "epoch": 1.8572585871678549,
+      "grad_norm": 0.8566899299621582,
+      "learning_rate": 1.6731820786550717e-06,
+      "loss": 0.1045,
+      "step": 11463
+    },
+    {
+      "epoch": 1.8574206092028516,
+      "grad_norm": 0.8467245101928711,
+      "learning_rate": 1.6727693958775172e-06,
+      "loss": 0.1007,
+      "step": 11464
+    },
+    {
+      "epoch": 1.8575826312378485,
+      "grad_norm": 0.8376954197883606,
+      "learning_rate": 1.6723567384112161e-06,
+      "loss": 0.1074,
+      "step": 11465
+    },
+    {
+      "epoch": 1.857744653272845,
+      "grad_norm": 0.7786714434623718,
+      "learning_rate": 1.671944106268797e-06,
+      "loss": 0.0937,
+      "step": 11466
+    },
+    {
+      "epoch": 1.857906675307842,
+      "grad_norm": 0.9477052688598633,
+      "learning_rate": 1.6715314994628834e-06,
+      "loss": 0.1135,
+      "step": 11467
+    },
+    {
+      "epoch": 1.8580686973428386,
+      "grad_norm": 0.9861614108085632,
+      "learning_rate": 1.671118918006101e-06,
+      "loss": 0.1181,
+      "step": 11468
+    },
+    {
+      "epoch": 1.8582307193778353,
+      "grad_norm": 0.8557963371276855,
+      "learning_rate": 1.6707063619110742e-06,
+      "loss": 0.1045,
+      "step": 11469
+    },
+    {
+      "epoch": 1.8583927414128323,
+      "grad_norm": 0.7471309900283813,
+      "learning_rate": 1.6702938311904262e-06,
+      "loss": 0.0944,
+      "step": 11470
+    },
+    {
+      "epoch": 1.8585547634478288,
+      "grad_norm": 0.7850518822669983,
+      "learning_rate": 1.669881325856779e-06,
+      "loss": 0.0951,
+      "step": 11471
+    },
+    {
+      "epoch": 1.8587167854828257,
+      "grad_norm": 0.8640868067741394,
+      "learning_rate": 1.6694688459227545e-06,
+      "loss": 0.1049,
+      "step": 11472
+    },
+    {
+      "epoch": 1.8588788075178224,
+      "grad_norm": 0.909514844417572,
+      "learning_rate": 1.6690563914009728e-06,
+      "loss": 0.0996,
+      "step": 11473
+    },
+    {
+      "epoch": 1.8590408295528191,
+      "grad_norm": 0.8319307565689087,
+      "learning_rate": 1.6686439623040548e-06,
+      "loss": 0.1022,
+      "step": 11474
+    },
+    {
+      "epoch": 1.859202851587816,
+      "grad_norm": 0.797558605670929,
+      "learning_rate": 1.6682315586446205e-06,
+      "loss": 0.0896,
+      "step": 11475
+    },
+    {
+      "epoch": 1.8593648736228126,
+      "grad_norm": 0.7331224083900452,
+      "learning_rate": 1.6678191804352873e-06,
+      "loss": 0.084,
+      "step": 11476
+    },
+    {
+      "epoch": 1.8595268956578095,
+      "grad_norm": 0.8644435405731201,
+      "learning_rate": 1.6674068276886734e-06,
+      "loss": 0.1117,
+      "step": 11477
+    },
+    {
+      "epoch": 1.8596889176928062,
+      "grad_norm": 0.7379981279373169,
+      "learning_rate": 1.6669945004173944e-06,
+      "loss": 0.0877,
+      "step": 11478
+    },
+    {
+      "epoch": 1.859850939727803,
+      "grad_norm": 0.7866598963737488,
+      "learning_rate": 1.6665821986340695e-06,
+      "loss": 0.0984,
+      "step": 11479
+    },
+    {
+      "epoch": 1.8600129617627998,
+      "grad_norm": 0.8214098215103149,
+      "learning_rate": 1.6661699223513118e-06,
+      "loss": 0.1077,
+      "step": 11480
+    },
+    {
+      "epoch": 1.8601749837977966,
+      "grad_norm": 0.7403610348701477,
+      "learning_rate": 1.6657576715817372e-06,
+      "loss": 0.0934,
+      "step": 11481
+    },
+    {
+      "epoch": 1.8603370058327933,
+      "grad_norm": 0.8874310255050659,
+      "learning_rate": 1.6653454463379582e-06,
+      "loss": 0.1102,
+      "step": 11482
+    },
+    {
+      "epoch": 1.86049902786779,
+      "grad_norm": 1.0349223613739014,
+      "learning_rate": 1.664933246632589e-06,
+      "loss": 0.115,
+      "step": 11483
+    },
+    {
+      "epoch": 1.8606610499027867,
+      "grad_norm": 0.7765905857086182,
+      "learning_rate": 1.6645210724782423e-06,
+      "loss": 0.1052,
+      "step": 11484
+    },
+    {
+      "epoch": 1.8608230719377836,
+      "grad_norm": 0.8592008948326111,
+      "learning_rate": 1.6641089238875283e-06,
+      "loss": 0.1013,
+      "step": 11485
+    },
+    {
+      "epoch": 1.8609850939727803,
+      "grad_norm": 0.8249764442443848,
+      "learning_rate": 1.6636968008730586e-06,
+      "loss": 0.1116,
+      "step": 11486
+    },
+    {
+      "epoch": 1.861147116007777,
+      "grad_norm": 0.7741264700889587,
+      "learning_rate": 1.6632847034474423e-06,
+      "loss": 0.0926,
+      "step": 11487
+    },
+    {
+      "epoch": 1.861309138042774,
+      "grad_norm": 0.6054975390434265,
+      "learning_rate": 1.6628726316232902e-06,
+      "loss": 0.0725,
+      "step": 11488
+    },
+    {
+      "epoch": 1.8614711600777705,
+      "grad_norm": 0.7995757460594177,
+      "learning_rate": 1.6624605854132094e-06,
+      "loss": 0.0928,
+      "step": 11489
+    },
+    {
+      "epoch": 1.8616331821127674,
+      "grad_norm": 0.6884124279022217,
+      "learning_rate": 1.6620485648298084e-06,
+      "loss": 0.0815,
+      "step": 11490
+    },
+    {
+      "epoch": 1.8617952041477641,
+      "grad_norm": 0.8780040740966797,
+      "learning_rate": 1.661636569885693e-06,
+      "loss": 0.1026,
+      "step": 11491
+    },
+    {
+      "epoch": 1.8619572261827608,
+      "grad_norm": 0.8116469383239746,
+      "learning_rate": 1.6612246005934694e-06,
+      "loss": 0.0942,
+      "step": 11492
+    },
+    {
+      "epoch": 1.8621192482177578,
+      "grad_norm": 0.8133668899536133,
+      "learning_rate": 1.660812656965744e-06,
+      "loss": 0.0935,
+      "step": 11493
+    },
+    {
+      "epoch": 1.8622812702527543,
+      "grad_norm": 0.6727808713912964,
+      "learning_rate": 1.66040073901512e-06,
+      "loss": 0.0841,
+      "step": 11494
+    },
+    {
+      "epoch": 1.8624432922877512,
+      "grad_norm": 0.8516779541969299,
+      "learning_rate": 1.6599888467542017e-06,
+      "loss": 0.114,
+      "step": 11495
+    },
+    {
+      "epoch": 1.862605314322748,
+      "grad_norm": 0.7665314078330994,
+      "learning_rate": 1.6595769801955925e-06,
+      "loss": 0.0951,
+      "step": 11496
+    },
+    {
+      "epoch": 1.8627673363577446,
+      "grad_norm": 0.7693126797676086,
+      "learning_rate": 1.6591651393518926e-06,
+      "loss": 0.0922,
+      "step": 11497
+    },
+    {
+      "epoch": 1.8629293583927415,
+      "grad_norm": 0.7686997652053833,
+      "learning_rate": 1.6587533242357053e-06,
+      "loss": 0.0905,
+      "step": 11498
+    },
+    {
+      "epoch": 1.863091380427738,
+      "grad_norm": 0.7887589335441589,
+      "learning_rate": 1.658341534859631e-06,
+      "loss": 0.0993,
+      "step": 11499
+    },
+    {
+      "epoch": 1.863253402462735,
+      "grad_norm": 0.8120632171630859,
+      "learning_rate": 1.6579297712362686e-06,
+      "loss": 0.0902,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8634154244977317,
+      "grad_norm": 0.9442248940467834,
+      "learning_rate": 1.657518033378217e-06,
+      "loss": 0.1086,
+      "step": 11501
+    },
+    {
+      "epoch": 1.8635774465327284,
+      "grad_norm": 0.9302598237991333,
+      "learning_rate": 1.6571063212980753e-06,
+      "loss": 0.1051,
+      "step": 11502
+    },
+    {
+      "epoch": 1.8637394685677253,
+      "grad_norm": 0.7944306135177612,
+      "learning_rate": 1.6566946350084405e-06,
+      "loss": 0.0994,
+      "step": 11503
+    },
+    {
+      "epoch": 1.8639014906027218,
+      "grad_norm": 0.8751698732376099,
+      "learning_rate": 1.6562829745219089e-06,
+      "loss": 0.1055,
+      "step": 11504
+    },
+    {
+      "epoch": 1.8640635126377187,
+      "grad_norm": 0.8934447169303894,
+      "learning_rate": 1.6558713398510767e-06,
+      "loss": 0.1028,
+      "step": 11505
+    },
+    {
+      "epoch": 1.8642255346727155,
+      "grad_norm": 0.8261483907699585,
+      "learning_rate": 1.6554597310085383e-06,
+      "loss": 0.0983,
+      "step": 11506
+    },
+    {
+      "epoch": 1.8643875567077122,
+      "grad_norm": 0.7163456082344055,
+      "learning_rate": 1.6550481480068887e-06,
+      "loss": 0.0841,
+      "step": 11507
+    },
+    {
+      "epoch": 1.864549578742709,
+      "grad_norm": 0.7598121166229248,
+      "learning_rate": 1.6546365908587213e-06,
+      "loss": 0.0926,
+      "step": 11508
+    },
+    {
+      "epoch": 1.8647116007777058,
+      "grad_norm": 0.7900456786155701,
+      "learning_rate": 1.654225059576628e-06,
+      "loss": 0.1034,
+      "step": 11509
+    },
+    {
+      "epoch": 1.8648736228127025,
+      "grad_norm": 0.792512834072113,
+      "learning_rate": 1.653813554173202e-06,
+      "loss": 0.092,
+      "step": 11510
+    },
+    {
+      "epoch": 1.8650356448476992,
+      "grad_norm": 0.8468700051307678,
+      "learning_rate": 1.6534020746610315e-06,
+      "loss": 0.1061,
+      "step": 11511
+    },
+    {
+      "epoch": 1.865197666882696,
+      "grad_norm": 0.883598268032074,
+      "learning_rate": 1.6529906210527107e-06,
+      "loss": 0.1075,
+      "step": 11512
+    },
+    {
+      "epoch": 1.8653596889176929,
+      "grad_norm": 0.8703530430793762,
+      "learning_rate": 1.6525791933608266e-06,
+      "loss": 0.1078,
+      "step": 11513
+    },
+    {
+      "epoch": 1.8655217109526896,
+      "grad_norm": 0.7216241955757141,
+      "learning_rate": 1.6521677915979688e-06,
+      "loss": 0.0921,
+      "step": 11514
+    },
+    {
+      "epoch": 1.8656837329876863,
+      "grad_norm": 0.7918997406959534,
+      "learning_rate": 1.6517564157767245e-06,
+      "loss": 0.0948,
+      "step": 11515
+    },
+    {
+      "epoch": 1.8658457550226832,
+      "grad_norm": 0.7993650436401367,
+      "learning_rate": 1.6513450659096804e-06,
+      "loss": 0.1015,
+      "step": 11516
+    },
+    {
+      "epoch": 1.8660077770576797,
+      "grad_norm": 0.9507842659950256,
+      "learning_rate": 1.650933742009425e-06,
+      "loss": 0.096,
+      "step": 11517
+    },
+    {
+      "epoch": 1.8661697990926767,
+      "grad_norm": 0.7698103189468384,
+      "learning_rate": 1.6505224440885414e-06,
+      "loss": 0.0967,
+      "step": 11518
+    },
+    {
+      "epoch": 1.8663318211276734,
+      "grad_norm": 0.9620030522346497,
+      "learning_rate": 1.6501111721596163e-06,
+      "loss": 0.1113,
+      "step": 11519
+    },
+    {
+      "epoch": 1.86649384316267,
+      "grad_norm": 0.7965664863586426,
+      "learning_rate": 1.649699926235232e-06,
+      "loss": 0.101,
+      "step": 11520
+    },
+    {
+      "epoch": 1.866655865197667,
+      "grad_norm": 0.7318791747093201,
+      "learning_rate": 1.6492887063279717e-06,
+      "loss": 0.0812,
+      "step": 11521
+    },
+    {
+      "epoch": 1.8668178872326635,
+      "grad_norm": 0.7944748401641846,
+      "learning_rate": 1.6488775124504188e-06,
+      "loss": 0.1008,
+      "step": 11522
+    },
+    {
+      "epoch": 1.8669799092676604,
+      "grad_norm": 0.8979195356369019,
+      "learning_rate": 1.648466344615155e-06,
+      "loss": 0.1114,
+      "step": 11523
+    },
+    {
+      "epoch": 1.8671419313026572,
+      "grad_norm": 0.8624005317687988,
+      "learning_rate": 1.6480552028347597e-06,
+      "loss": 0.1084,
+      "step": 11524
+    },
+    {
+      "epoch": 1.8673039533376539,
+      "grad_norm": 0.9684587717056274,
+      "learning_rate": 1.647644087121813e-06,
+      "loss": 0.1113,
+      "step": 11525
+    },
+    {
+      "epoch": 1.8674659753726508,
+      "grad_norm": 0.7721800804138184,
+      "learning_rate": 1.6472329974888956e-06,
+      "loss": 0.0923,
+      "step": 11526
+    },
+    {
+      "epoch": 1.8676279974076473,
+      "grad_norm": 0.7591527700424194,
+      "learning_rate": 1.6468219339485845e-06,
+      "loss": 0.0978,
+      "step": 11527
+    },
+    {
+      "epoch": 1.8677900194426442,
+      "grad_norm": 0.7007966637611389,
+      "learning_rate": 1.6464108965134578e-06,
+      "loss": 0.0874,
+      "step": 11528
+    },
+    {
+      "epoch": 1.867952041477641,
+      "grad_norm": 0.8641650080680847,
+      "learning_rate": 1.6459998851960918e-06,
+      "loss": 0.1124,
+      "step": 11529
+    },
+    {
+      "epoch": 1.8681140635126376,
+      "grad_norm": 0.8576594591140747,
+      "learning_rate": 1.645588900009062e-06,
+      "loss": 0.1001,
+      "step": 11530
+    },
+    {
+      "epoch": 1.8682760855476346,
+      "grad_norm": 0.8817536234855652,
+      "learning_rate": 1.645177940964945e-06,
+      "loss": 0.1046,
+      "step": 11531
+    },
+    {
+      "epoch": 1.8684381075826313,
+      "grad_norm": 0.7928271889686584,
+      "learning_rate": 1.6447670080763146e-06,
+      "loss": 0.0985,
+      "step": 11532
+    },
+    {
+      "epoch": 1.868600129617628,
+      "grad_norm": 0.8168180584907532,
+      "learning_rate": 1.6443561013557434e-06,
+      "loss": 0.1076,
+      "step": 11533
+    },
+    {
+      "epoch": 1.8687621516526247,
+      "grad_norm": 0.845324695110321,
+      "learning_rate": 1.6439452208158058e-06,
+      "loss": 0.1029,
+      "step": 11534
+    },
+    {
+      "epoch": 1.8689241736876214,
+      "grad_norm": 0.9935814142227173,
+      "learning_rate": 1.6435343664690718e-06,
+      "loss": 0.1097,
+      "step": 11535
+    },
+    {
+      "epoch": 1.8690861957226184,
+      "grad_norm": 0.796512246131897,
+      "learning_rate": 1.6431235383281135e-06,
+      "loss": 0.0944,
+      "step": 11536
+    },
+    {
+      "epoch": 1.869248217757615,
+      "grad_norm": 0.8797528147697449,
+      "learning_rate": 1.6427127364055024e-06,
+      "loss": 0.1073,
+      "step": 11537
+    },
+    {
+      "epoch": 1.8694102397926118,
+      "grad_norm": 0.7383131384849548,
+      "learning_rate": 1.6423019607138064e-06,
+      "loss": 0.0979,
+      "step": 11538
+    },
+    {
+      "epoch": 1.8695722618276087,
+      "grad_norm": 0.794525146484375,
+      "learning_rate": 1.641891211265595e-06,
+      "loss": 0.1006,
+      "step": 11539
+    },
+    {
+      "epoch": 1.8697342838626052,
+      "grad_norm": 0.7232054471969604,
+      "learning_rate": 1.641480488073435e-06,
+      "loss": 0.0898,
+      "step": 11540
+    },
+    {
+      "epoch": 1.8698963058976021,
+      "grad_norm": 0.7711024284362793,
+      "learning_rate": 1.6410697911498957e-06,
+      "loss": 0.099,
+      "step": 11541
+    },
+    {
+      "epoch": 1.8700583279325989,
+      "grad_norm": 0.7795062065124512,
+      "learning_rate": 1.6406591205075417e-06,
+      "loss": 0.1017,
+      "step": 11542
+    },
+    {
+      "epoch": 1.8702203499675956,
+      "grad_norm": 0.7736190557479858,
+      "learning_rate": 1.6402484761589397e-06,
+      "loss": 0.0874,
+      "step": 11543
+    },
+    {
+      "epoch": 1.8703823720025925,
+      "grad_norm": 0.7875672578811646,
+      "learning_rate": 1.639837858116653e-06,
+      "loss": 0.1025,
+      "step": 11544
+    },
+    {
+      "epoch": 1.870544394037589,
+      "grad_norm": 0.7884228825569153,
+      "learning_rate": 1.639427266393246e-06,
+      "loss": 0.0903,
+      "step": 11545
+    },
+    {
+      "epoch": 1.870706416072586,
+      "grad_norm": 0.6791999340057373,
+      "learning_rate": 1.6390167010012824e-06,
+      "loss": 0.0852,
+      "step": 11546
+    },
+    {
+      "epoch": 1.8708684381075826,
+      "grad_norm": 0.9688091278076172,
+      "learning_rate": 1.638606161953325e-06,
+      "loss": 0.1071,
+      "step": 11547
+    },
+    {
+      "epoch": 1.8710304601425793,
+      "grad_norm": 0.7724692225456238,
+      "learning_rate": 1.638195649261934e-06,
+      "loss": 0.0987,
+      "step": 11548
+    },
+    {
+      "epoch": 1.8711924821775763,
+      "grad_norm": 0.685142457485199,
+      "learning_rate": 1.6377851629396695e-06,
+      "loss": 0.0857,
+      "step": 11549
+    },
+    {
+      "epoch": 1.8713545042125728,
+      "grad_norm": 0.8404377698898315,
+      "learning_rate": 1.6373747029990943e-06,
+      "loss": 0.1011,
+      "step": 11550
+    },
+    {
+      "epoch": 1.8715165262475697,
+      "grad_norm": 0.8327875137329102,
+      "learning_rate": 1.6369642694527648e-06,
+      "loss": 0.1033,
+      "step": 11551
+    },
+    {
+      "epoch": 1.8716785482825664,
+      "grad_norm": 0.777384877204895,
+      "learning_rate": 1.6365538623132405e-06,
+      "loss": 0.1067,
+      "step": 11552
+    },
+    {
+      "epoch": 1.8718405703175631,
+      "grad_norm": 0.806379497051239,
+      "learning_rate": 1.6361434815930782e-06,
+      "loss": 0.0952,
+      "step": 11553
+    },
+    {
+      "epoch": 1.87200259235256,
+      "grad_norm": 0.7385959029197693,
+      "learning_rate": 1.6357331273048343e-06,
+      "loss": 0.0857,
+      "step": 11554
+    },
+    {
+      "epoch": 1.8721646143875565,
+      "grad_norm": 0.8200345635414124,
+      "learning_rate": 1.635322799461066e-06,
+      "loss": 0.1041,
+      "step": 11555
+    },
+    {
+      "epoch": 1.8723266364225535,
+      "grad_norm": 0.8210601806640625,
+      "learning_rate": 1.6349124980743278e-06,
+      "loss": 0.101,
+      "step": 11556
+    },
+    {
+      "epoch": 1.8724886584575502,
+      "grad_norm": 0.7943647503852844,
+      "learning_rate": 1.6345022231571734e-06,
+      "loss": 0.0912,
+      "step": 11557
+    },
+    {
+      "epoch": 1.872650680492547,
+      "grad_norm": 0.8460164666175842,
+      "learning_rate": 1.6340919747221568e-06,
+      "loss": 0.0955,
+      "step": 11558
+    },
+    {
+      "epoch": 1.8728127025275438,
+      "grad_norm": 0.7457050085067749,
+      "learning_rate": 1.6336817527818292e-06,
+      "loss": 0.0884,
+      "step": 11559
+    },
+    {
+      "epoch": 1.8729747245625405,
+      "grad_norm": 1.021488070487976,
+      "learning_rate": 1.633271557348744e-06,
+      "loss": 0.1097,
+      "step": 11560
+    },
+    {
+      "epoch": 1.8731367465975373,
+      "grad_norm": 0.8276289701461792,
+      "learning_rate": 1.6328613884354524e-06,
+      "loss": 0.0963,
+      "step": 11561
+    },
+    {
+      "epoch": 1.873298768632534,
+      "grad_norm": 0.6760364770889282,
+      "learning_rate": 1.6324512460545034e-06,
+      "loss": 0.0817,
+      "step": 11562
+    },
+    {
+      "epoch": 1.8734607906675307,
+      "grad_norm": 0.7393625378608704,
+      "learning_rate": 1.6320411302184474e-06,
+      "loss": 0.0919,
+      "step": 11563
+    },
+    {
+      "epoch": 1.8736228127025276,
+      "grad_norm": 0.7952508926391602,
+      "learning_rate": 1.6316310409398306e-06,
+      "loss": 0.0932,
+      "step": 11564
+    },
+    {
+      "epoch": 1.8737848347375243,
+      "grad_norm": 0.8211168050765991,
+      "learning_rate": 1.6312209782312044e-06,
+      "loss": 0.1028,
+      "step": 11565
+    },
+    {
+      "epoch": 1.873946856772521,
+      "grad_norm": 0.8613480925559998,
+      "learning_rate": 1.6308109421051132e-06,
+      "loss": 0.106,
+      "step": 11566
+    },
+    {
+      "epoch": 1.874108878807518,
+      "grad_norm": 0.6946225166320801,
+      "learning_rate": 1.6304009325741044e-06,
+      "loss": 0.0865,
+      "step": 11567
+    },
+    {
+      "epoch": 1.8742709008425145,
+      "grad_norm": 0.8030040264129639,
+      "learning_rate": 1.6299909496507214e-06,
+      "loss": 0.1029,
+      "step": 11568
+    },
+    {
+      "epoch": 1.8744329228775114,
+      "grad_norm": 0.840975284576416,
+      "learning_rate": 1.6295809933475103e-06,
+      "loss": 0.1061,
+      "step": 11569
+    },
+    {
+      "epoch": 1.874594944912508,
+      "grad_norm": 0.893821120262146,
+      "learning_rate": 1.6291710636770152e-06,
+      "loss": 0.1042,
+      "step": 11570
+    },
+    {
+      "epoch": 1.8747569669475048,
+      "grad_norm": 0.8448208570480347,
+      "learning_rate": 1.6287611606517778e-06,
+      "loss": 0.0991,
+      "step": 11571
+    },
+    {
+      "epoch": 1.8749189889825018,
+      "grad_norm": 0.7172380089759827,
+      "learning_rate": 1.628351284284341e-06,
+      "loss": 0.0892,
+      "step": 11572
+    },
+    {
+      "epoch": 1.8750810110174982,
+      "grad_norm": 0.8485018610954285,
+      "learning_rate": 1.627941434587244e-06,
+      "loss": 0.0992,
+      "step": 11573
+    },
+    {
+      "epoch": 1.8752430330524952,
+      "grad_norm": 0.8565303087234497,
+      "learning_rate": 1.6275316115730302e-06,
+      "loss": 0.0989,
+      "step": 11574
+    },
+    {
+      "epoch": 1.875405055087492,
+      "grad_norm": 0.8015487194061279,
+      "learning_rate": 1.6271218152542373e-06,
+      "loss": 0.0978,
+      "step": 11575
+    },
+    {
+      "epoch": 1.8755670771224886,
+      "grad_norm": 0.962988555431366,
+      "learning_rate": 1.626712045643405e-06,
+      "loss": 0.1113,
+      "step": 11576
+    },
+    {
+      "epoch": 1.8757290991574855,
+      "grad_norm": 0.8654093742370605,
+      "learning_rate": 1.6263023027530706e-06,
+      "loss": 0.0986,
+      "step": 11577
+    },
+    {
+      "epoch": 1.875891121192482,
+      "grad_norm": 0.8079492449760437,
+      "learning_rate": 1.6258925865957703e-06,
+      "loss": 0.1023,
+      "step": 11578
+    },
+    {
+      "epoch": 1.876053143227479,
+      "grad_norm": 0.7464910745620728,
+      "learning_rate": 1.6254828971840432e-06,
+      "loss": 0.0936,
+      "step": 11579
+    },
+    {
+      "epoch": 1.8762151652624757,
+      "grad_norm": 0.8631336688995361,
+      "learning_rate": 1.625073234530422e-06,
+      "loss": 0.1021,
+      "step": 11580
+    },
+    {
+      "epoch": 1.8763771872974724,
+      "grad_norm": 0.7467739582061768,
+      "learning_rate": 1.6246635986474436e-06,
+      "loss": 0.0909,
+      "step": 11581
+    },
+    {
+      "epoch": 1.8765392093324693,
+      "grad_norm": 0.8743352293968201,
+      "learning_rate": 1.62425398954764e-06,
+      "loss": 0.1001,
+      "step": 11582
+    },
+    {
+      "epoch": 1.876701231367466,
+      "grad_norm": 0.70003342628479,
+      "learning_rate": 1.6238444072435447e-06,
+      "loss": 0.0814,
+      "step": 11583
+    },
+    {
+      "epoch": 1.8768632534024627,
+      "grad_norm": 0.7987949848175049,
+      "learning_rate": 1.6234348517476905e-06,
+      "loss": 0.1061,
+      "step": 11584
+    },
+    {
+      "epoch": 1.8770252754374595,
+      "grad_norm": 0.8156781196594238,
+      "learning_rate": 1.6230253230726096e-06,
+      "loss": 0.1068,
+      "step": 11585
+    },
+    {
+      "epoch": 1.8771872974724562,
+      "grad_norm": 0.8381296396255493,
+      "learning_rate": 1.6226158212308307e-06,
+      "loss": 0.1087,
+      "step": 11586
+    },
+    {
+      "epoch": 1.877349319507453,
+      "grad_norm": 0.794044017791748,
+      "learning_rate": 1.622206346234885e-06,
+      "loss": 0.0952,
+      "step": 11587
+    },
+    {
+      "epoch": 1.8775113415424498,
+      "grad_norm": 0.829454243183136,
+      "learning_rate": 1.6217968980972998e-06,
+      "loss": 0.1039,
+      "step": 11588
+    },
+    {
+      "epoch": 1.8776733635774465,
+      "grad_norm": 0.7840944528579712,
+      "learning_rate": 1.621387476830605e-06,
+      "loss": 0.0937,
+      "step": 11589
+    },
+    {
+      "epoch": 1.8778353856124435,
+      "grad_norm": 0.6949406862258911,
+      "learning_rate": 1.620978082447327e-06,
+      "loss": 0.0796,
+      "step": 11590
+    },
+    {
+      "epoch": 1.87799740764744,
+      "grad_norm": 0.8238313794136047,
+      "learning_rate": 1.6205687149599933e-06,
+      "loss": 0.1038,
+      "step": 11591
+    },
+    {
+      "epoch": 1.8781594296824369,
+      "grad_norm": 0.7915367484092712,
+      "learning_rate": 1.6201593743811275e-06,
+      "loss": 0.1021,
+      "step": 11592
+    },
+    {
+      "epoch": 1.8783214517174336,
+      "grad_norm": 0.8443301320075989,
+      "learning_rate": 1.6197500607232563e-06,
+      "loss": 0.0991,
+      "step": 11593
+    },
+    {
+      "epoch": 1.8784834737524303,
+      "grad_norm": 0.7770126461982727,
+      "learning_rate": 1.6193407739989037e-06,
+      "loss": 0.0969,
+      "step": 11594
+    },
+    {
+      "epoch": 1.8786454957874272,
+      "grad_norm": 0.8288414478302002,
+      "learning_rate": 1.6189315142205914e-06,
+      "loss": 0.1031,
+      "step": 11595
+    },
+    {
+      "epoch": 1.8788075178224237,
+      "grad_norm": 0.7647958993911743,
+      "learning_rate": 1.6185222814008434e-06,
+      "loss": 0.0886,
+      "step": 11596
+    },
+    {
+      "epoch": 1.8789695398574207,
+      "grad_norm": 0.7341493368148804,
+      "learning_rate": 1.6181130755521792e-06,
+      "loss": 0.0968,
+      "step": 11597
+    },
+    {
+      "epoch": 1.8791315618924174,
+      "grad_norm": 0.7972742319107056,
+      "learning_rate": 1.6177038966871213e-06,
+      "loss": 0.1008,
+      "step": 11598
+    },
+    {
+      "epoch": 1.879293583927414,
+      "grad_norm": 0.8043832182884216,
+      "learning_rate": 1.617294744818189e-06,
+      "loss": 0.0959,
+      "step": 11599
+    },
+    {
+      "epoch": 1.879455605962411,
+      "grad_norm": 0.8683483600616455,
+      "learning_rate": 1.6168856199579025e-06,
+      "loss": 0.1057,
+      "step": 11600
+    },
+    {
+      "epoch": 1.8796176279974075,
+      "grad_norm": 0.8487406373023987,
+      "learning_rate": 1.6164765221187778e-06,
+      "loss": 0.0997,
+      "step": 11601
+    },
+    {
+      "epoch": 1.8797796500324044,
+      "grad_norm": 0.7058298587799072,
+      "learning_rate": 1.6160674513133332e-06,
+      "loss": 0.0892,
+      "step": 11602
+    },
+    {
+      "epoch": 1.8799416720674011,
+      "grad_norm": 1.010647177696228,
+      "learning_rate": 1.6156584075540864e-06,
+      "loss": 0.1148,
+      "step": 11603
+    },
+    {
+      "epoch": 1.8801036941023979,
+      "grad_norm": 0.839988648891449,
+      "learning_rate": 1.615249390853552e-06,
+      "loss": 0.0978,
+      "step": 11604
+    },
+    {
+      "epoch": 1.8802657161373948,
+      "grad_norm": 0.8383321762084961,
+      "learning_rate": 1.6148404012242453e-06,
+      "loss": 0.0976,
+      "step": 11605
+    },
+    {
+      "epoch": 1.8804277381723913,
+      "grad_norm": 0.7959474325180054,
+      "learning_rate": 1.61443143867868e-06,
+      "loss": 0.0901,
+      "step": 11606
+    },
+    {
+      "epoch": 1.8805897602073882,
+      "grad_norm": 0.854000985622406,
+      "learning_rate": 1.614022503229369e-06,
+      "loss": 0.1036,
+      "step": 11607
+    },
+    {
+      "epoch": 1.880751782242385,
+      "grad_norm": 0.8434250950813293,
+      "learning_rate": 1.613613594888826e-06,
+      "loss": 0.0996,
+      "step": 11608
+    },
+    {
+      "epoch": 1.8809138042773816,
+      "grad_norm": 0.7653728127479553,
+      "learning_rate": 1.6132047136695625e-06,
+      "loss": 0.101,
+      "step": 11609
+    },
+    {
+      "epoch": 1.8810758263123786,
+      "grad_norm": 0.7736960649490356,
+      "learning_rate": 1.612795859584088e-06,
+      "loss": 0.0956,
+      "step": 11610
+    },
+    {
+      "epoch": 1.8812378483473753,
+      "grad_norm": 0.7493591904640198,
+      "learning_rate": 1.6123870326449144e-06,
+      "loss": 0.0897,
+      "step": 11611
+    },
+    {
+      "epoch": 1.881399870382372,
+      "grad_norm": 0.7444273829460144,
+      "learning_rate": 1.611978232864548e-06,
+      "loss": 0.0943,
+      "step": 11612
+    },
+    {
+      "epoch": 1.8815618924173687,
+      "grad_norm": 0.7813089489936829,
+      "learning_rate": 1.6115694602554994e-06,
+      "loss": 0.0952,
+      "step": 11613
+    },
+    {
+      "epoch": 1.8817239144523654,
+      "grad_norm": 0.8079869151115417,
+      "learning_rate": 1.6111607148302758e-06,
+      "loss": 0.1069,
+      "step": 11614
+    },
+    {
+      "epoch": 1.8818859364873624,
+      "grad_norm": 0.8186015486717224,
+      "learning_rate": 1.6107519966013828e-06,
+      "loss": 0.1025,
+      "step": 11615
+    },
+    {
+      "epoch": 1.882047958522359,
+      "grad_norm": 0.7281806468963623,
+      "learning_rate": 1.6103433055813265e-06,
+      "loss": 0.0893,
+      "step": 11616
+    },
+    {
+      "epoch": 1.8822099805573558,
+      "grad_norm": 0.7015278935432434,
+      "learning_rate": 1.6099346417826123e-06,
+      "loss": 0.0844,
+      "step": 11617
+    },
+    {
+      "epoch": 1.8823720025923527,
+      "grad_norm": 0.825904905796051,
+      "learning_rate": 1.6095260052177446e-06,
+      "loss": 0.1027,
+      "step": 11618
+    },
+    {
+      "epoch": 1.8825340246273492,
+      "grad_norm": 0.8017039895057678,
+      "learning_rate": 1.6091173958992261e-06,
+      "loss": 0.0955,
+      "step": 11619
+    },
+    {
+      "epoch": 1.8826960466623461,
+      "grad_norm": 0.7551977038383484,
+      "learning_rate": 1.6087088138395598e-06,
+      "loss": 0.0907,
+      "step": 11620
+    },
+    {
+      "epoch": 1.8828580686973428,
+      "grad_norm": 0.9527973532676697,
+      "learning_rate": 1.6083002590512458e-06,
+      "loss": 0.1074,
+      "step": 11621
+    },
+    {
+      "epoch": 1.8830200907323396,
+      "grad_norm": 0.8547660112380981,
+      "learning_rate": 1.6078917315467867e-06,
+      "loss": 0.106,
+      "step": 11622
+    },
+    {
+      "epoch": 1.8831821127673365,
+      "grad_norm": 0.7676581144332886,
+      "learning_rate": 1.607483231338682e-06,
+      "loss": 0.0914,
+      "step": 11623
+    },
+    {
+      "epoch": 1.883344134802333,
+      "grad_norm": 0.7763146162033081,
+      "learning_rate": 1.6070747584394303e-06,
+      "loss": 0.0927,
+      "step": 11624
+    },
+    {
+      "epoch": 1.88350615683733,
+      "grad_norm": 0.8205904364585876,
+      "learning_rate": 1.6066663128615301e-06,
+      "loss": 0.1028,
+      "step": 11625
+    },
+    {
+      "epoch": 1.8836681788723266,
+      "grad_norm": 0.7976201176643372,
+      "learning_rate": 1.6062578946174785e-06,
+      "loss": 0.0999,
+      "step": 11626
+    },
+    {
+      "epoch": 1.8838302009073233,
+      "grad_norm": 0.8827847242355347,
+      "learning_rate": 1.605849503719773e-06,
+      "loss": 0.109,
+      "step": 11627
+    },
+    {
+      "epoch": 1.8839922229423203,
+      "grad_norm": 0.8832690715789795,
+      "learning_rate": 1.605441140180909e-06,
+      "loss": 0.11,
+      "step": 11628
+    },
+    {
+      "epoch": 1.8841542449773168,
+      "grad_norm": 0.7509612441062927,
+      "learning_rate": 1.605032804013381e-06,
+      "loss": 0.091,
+      "step": 11629
+    },
+    {
+      "epoch": 1.8843162670123137,
+      "grad_norm": 0.7290849685668945,
+      "learning_rate": 1.6046244952296839e-06,
+      "loss": 0.0928,
+      "step": 11630
+    },
+    {
+      "epoch": 1.8844782890473104,
+      "grad_norm": 0.7547253966331482,
+      "learning_rate": 1.6042162138423095e-06,
+      "loss": 0.0907,
+      "step": 11631
+    },
+    {
+      "epoch": 1.8846403110823071,
+      "grad_norm": 0.7190500497817993,
+      "learning_rate": 1.6038079598637523e-06,
+      "loss": 0.0858,
+      "step": 11632
+    },
+    {
+      "epoch": 1.884802333117304,
+      "grad_norm": 0.9019315242767334,
+      "learning_rate": 1.6033997333065022e-06,
+      "loss": 0.0975,
+      "step": 11633
+    },
+    {
+      "epoch": 1.8849643551523008,
+      "grad_norm": 0.8361161947250366,
+      "learning_rate": 1.6029915341830503e-06,
+      "loss": 0.102,
+      "step": 11634
+    },
+    {
+      "epoch": 1.8851263771872975,
+      "grad_norm": 1.2841830253601074,
+      "learning_rate": 1.6025833625058878e-06,
+      "loss": 0.1219,
+      "step": 11635
+    },
+    {
+      "epoch": 1.8852883992222942,
+      "grad_norm": 0.7469258308410645,
+      "learning_rate": 1.6021752182875012e-06,
+      "loss": 0.0904,
+      "step": 11636
+    },
+    {
+      "epoch": 1.885450421257291,
+      "grad_norm": 0.8182607293128967,
+      "learning_rate": 1.601767101540381e-06,
+      "loss": 0.1008,
+      "step": 11637
+    },
+    {
+      "epoch": 1.8856124432922878,
+      "grad_norm": 0.6999529004096985,
+      "learning_rate": 1.6013590122770143e-06,
+      "loss": 0.0842,
+      "step": 11638
+    },
+    {
+      "epoch": 1.8857744653272845,
+      "grad_norm": 0.7251670956611633,
+      "learning_rate": 1.6009509505098863e-06,
+      "loss": 0.0931,
+      "step": 11639
+    },
+    {
+      "epoch": 1.8859364873622813,
+      "grad_norm": 0.825188934803009,
+      "learning_rate": 1.6005429162514834e-06,
+      "loss": 0.1004,
+      "step": 11640
+    },
+    {
+      "epoch": 1.8860985093972782,
+      "grad_norm": 0.72586590051651,
+      "learning_rate": 1.6001349095142918e-06,
+      "loss": 0.0852,
+      "step": 11641
+    },
+    {
+      "epoch": 1.8862605314322747,
+      "grad_norm": 0.8619117140769958,
+      "learning_rate": 1.5997269303107937e-06,
+      "loss": 0.0906,
+      "step": 11642
+    },
+    {
+      "epoch": 1.8864225534672716,
+      "grad_norm": 0.8306232690811157,
+      "learning_rate": 1.5993189786534727e-06,
+      "loss": 0.1013,
+      "step": 11643
+    },
+    {
+      "epoch": 1.8865845755022683,
+      "grad_norm": 0.7801927924156189,
+      "learning_rate": 1.598911054554812e-06,
+      "loss": 0.0915,
+      "step": 11644
+    },
+    {
+      "epoch": 1.886746597537265,
+      "grad_norm": 0.7791352272033691,
+      "learning_rate": 1.5985031580272914e-06,
+      "loss": 0.0875,
+      "step": 11645
+    },
+    {
+      "epoch": 1.886908619572262,
+      "grad_norm": 0.8080638647079468,
+      "learning_rate": 1.5980952890833929e-06,
+      "loss": 0.0956,
+      "step": 11646
+    },
+    {
+      "epoch": 1.8870706416072585,
+      "grad_norm": 0.7200146913528442,
+      "learning_rate": 1.597687447735597e-06,
+      "loss": 0.0807,
+      "step": 11647
+    },
+    {
+      "epoch": 1.8872326636422554,
+      "grad_norm": 0.8803225755691528,
+      "learning_rate": 1.5972796339963806e-06,
+      "loss": 0.1038,
+      "step": 11648
+    },
+    {
+      "epoch": 1.887394685677252,
+      "grad_norm": 0.8330278992652893,
+      "learning_rate": 1.5968718478782236e-06,
+      "loss": 0.1015,
+      "step": 11649
+    },
+    {
+      "epoch": 1.8875567077122488,
+      "grad_norm": 0.7998868823051453,
+      "learning_rate": 1.5964640893936015e-06,
+      "loss": 0.0993,
+      "step": 11650
+    },
+    {
+      "epoch": 1.8877187297472457,
+      "grad_norm": 0.9711896777153015,
+      "learning_rate": 1.596056358554992e-06,
+      "loss": 0.1076,
+      "step": 11651
+    },
+    {
+      "epoch": 1.8878807517822422,
+      "grad_norm": 0.8102350831031799,
+      "learning_rate": 1.595648655374871e-06,
+      "loss": 0.1031,
+      "step": 11652
+    },
+    {
+      "epoch": 1.8880427738172392,
+      "grad_norm": 0.879172146320343,
+      "learning_rate": 1.5952409798657127e-06,
+      "loss": 0.1006,
+      "step": 11653
+    },
+    {
+      "epoch": 1.8882047958522359,
+      "grad_norm": 0.8931054472923279,
+      "learning_rate": 1.5948333320399905e-06,
+      "loss": 0.1153,
+      "step": 11654
+    },
+    {
+      "epoch": 1.8883668178872326,
+      "grad_norm": 0.7753148078918457,
+      "learning_rate": 1.5944257119101775e-06,
+      "loss": 0.0959,
+      "step": 11655
+    },
+    {
+      "epoch": 1.8885288399222295,
+      "grad_norm": 0.8476399183273315,
+      "learning_rate": 1.5940181194887472e-06,
+      "loss": 0.1019,
+      "step": 11656
+    },
+    {
+      "epoch": 1.888690861957226,
+      "grad_norm": 0.7567358613014221,
+      "learning_rate": 1.5936105547881697e-06,
+      "loss": 0.0865,
+      "step": 11657
+    },
+    {
+      "epoch": 1.888852883992223,
+      "grad_norm": 0.7247362732887268,
+      "learning_rate": 1.5932030178209163e-06,
+      "loss": 0.0863,
+      "step": 11658
+    },
+    {
+      "epoch": 1.8890149060272197,
+      "grad_norm": 0.7682359218597412,
+      "learning_rate": 1.5927955085994544e-06,
+      "loss": 0.0938,
+      "step": 11659
+    },
+    {
+      "epoch": 1.8891769280622164,
+      "grad_norm": 0.750938355922699,
+      "learning_rate": 1.592388027136256e-06,
+      "loss": 0.0947,
+      "step": 11660
+    },
+    {
+      "epoch": 1.8893389500972133,
+      "grad_norm": 0.8155219554901123,
+      "learning_rate": 1.5919805734437871e-06,
+      "loss": 0.1027,
+      "step": 11661
+    },
+    {
+      "epoch": 1.88950097213221,
+      "grad_norm": 0.8531356453895569,
+      "learning_rate": 1.591573147534516e-06,
+      "loss": 0.1002,
+      "step": 11662
+    },
+    {
+      "epoch": 1.8896629941672067,
+      "grad_norm": 0.8241977095603943,
+      "learning_rate": 1.5911657494209077e-06,
+      "loss": 0.0958,
+      "step": 11663
+    },
+    {
+      "epoch": 1.8898250162022034,
+      "grad_norm": 0.8130655884742737,
+      "learning_rate": 1.5907583791154275e-06,
+      "loss": 0.1017,
+      "step": 11664
+    },
+    {
+      "epoch": 1.8899870382372002,
+      "grad_norm": 0.7908192873001099,
+      "learning_rate": 1.5903510366305416e-06,
+      "loss": 0.1012,
+      "step": 11665
+    },
+    {
+      "epoch": 1.890149060272197,
+      "grad_norm": 0.7589678168296814,
+      "learning_rate": 1.5899437219787124e-06,
+      "loss": 0.0919,
+      "step": 11666
+    },
+    {
+      "epoch": 1.8903110823071938,
+      "grad_norm": 0.7555517554283142,
+      "learning_rate": 1.5895364351724033e-06,
+      "loss": 0.0946,
+      "step": 11667
+    },
+    {
+      "epoch": 1.8904731043421905,
+      "grad_norm": 0.7375280857086182,
+      "learning_rate": 1.5891291762240757e-06,
+      "loss": 0.09,
+      "step": 11668
+    },
+    {
+      "epoch": 1.8906351263771874,
+      "grad_norm": 0.8817306756973267,
+      "learning_rate": 1.5887219451461903e-06,
+      "loss": 0.1076,
+      "step": 11669
+    },
+    {
+      "epoch": 1.890797148412184,
+      "grad_norm": 0.7232506275177002,
+      "learning_rate": 1.5883147419512086e-06,
+      "loss": 0.0868,
+      "step": 11670
+    },
+    {
+      "epoch": 1.8909591704471809,
+      "grad_norm": 0.8023999333381653,
+      "learning_rate": 1.5879075666515903e-06,
+      "loss": 0.1019,
+      "step": 11671
+    },
+    {
+      "epoch": 1.8911211924821776,
+      "grad_norm": 0.8410249948501587,
+      "learning_rate": 1.5875004192597926e-06,
+      "loss": 0.0945,
+      "step": 11672
+    },
+    {
+      "epoch": 1.8912832145171743,
+      "grad_norm": 0.8737829327583313,
+      "learning_rate": 1.5870932997882742e-06,
+      "loss": 0.1045,
+      "step": 11673
+    },
+    {
+      "epoch": 1.8914452365521712,
+      "grad_norm": 0.8467708230018616,
+      "learning_rate": 1.5866862082494907e-06,
+      "loss": 0.094,
+      "step": 11674
+    },
+    {
+      "epoch": 1.8916072585871677,
+      "grad_norm": 0.6961696743965149,
+      "learning_rate": 1.5862791446558999e-06,
+      "loss": 0.0826,
+      "step": 11675
+    },
+    {
+      "epoch": 1.8917692806221647,
+      "grad_norm": 0.8878881931304932,
+      "learning_rate": 1.5858721090199564e-06,
+      "loss": 0.097,
+      "step": 11676
+    },
+    {
+      "epoch": 1.8919313026571614,
+      "grad_norm": 0.9326772093772888,
+      "learning_rate": 1.5854651013541134e-06,
+      "loss": 0.1038,
+      "step": 11677
+    },
+    {
+      "epoch": 1.892093324692158,
+      "grad_norm": 0.769069492816925,
+      "learning_rate": 1.5850581216708254e-06,
+      "loss": 0.0946,
+      "step": 11678
+    },
+    {
+      "epoch": 1.892255346727155,
+      "grad_norm": 0.8138481378555298,
+      "learning_rate": 1.5846511699825445e-06,
+      "loss": 0.099,
+      "step": 11679
+    },
+    {
+      "epoch": 1.8924173687621515,
+      "grad_norm": 0.9026851058006287,
+      "learning_rate": 1.5842442463017235e-06,
+      "loss": 0.113,
+      "step": 11680
+    },
+    {
+      "epoch": 1.8925793907971484,
+      "grad_norm": 0.796542763710022,
+      "learning_rate": 1.583837350640812e-06,
+      "loss": 0.0943,
+      "step": 11681
+    },
+    {
+      "epoch": 1.8927414128321451,
+      "grad_norm": 0.7942173480987549,
+      "learning_rate": 1.583430483012261e-06,
+      "loss": 0.0911,
+      "step": 11682
+    },
+    {
+      "epoch": 1.8929034348671419,
+      "grad_norm": 0.748563289642334,
+      "learning_rate": 1.5830236434285182e-06,
+      "loss": 0.0934,
+      "step": 11683
+    },
+    {
+      "epoch": 1.8930654569021388,
+      "grad_norm": 0.8524467349052429,
+      "learning_rate": 1.5826168319020332e-06,
+      "loss": 0.105,
+      "step": 11684
+    },
+    {
+      "epoch": 1.8932274789371355,
+      "grad_norm": 0.8268295526504517,
+      "learning_rate": 1.5822100484452538e-06,
+      "loss": 0.0841,
+      "step": 11685
+    },
+    {
+      "epoch": 1.8933895009721322,
+      "grad_norm": 0.7947415709495544,
+      "learning_rate": 1.5818032930706254e-06,
+      "loss": 0.1011,
+      "step": 11686
+    },
+    {
+      "epoch": 1.893551523007129,
+      "grad_norm": 0.8413145542144775,
+      "learning_rate": 1.5813965657905942e-06,
+      "loss": 0.1021,
+      "step": 11687
+    },
+    {
+      "epoch": 1.8937135450421256,
+      "grad_norm": 0.7836431860923767,
+      "learning_rate": 1.5809898666176044e-06,
+      "loss": 0.0948,
+      "step": 11688
+    },
+    {
+      "epoch": 1.8938755670771226,
+      "grad_norm": 0.8638646602630615,
+      "learning_rate": 1.5805831955641022e-06,
+      "loss": 0.1104,
+      "step": 11689
+    },
+    {
+      "epoch": 1.8940375891121193,
+      "grad_norm": 0.7815718650817871,
+      "learning_rate": 1.5801765526425283e-06,
+      "loss": 0.0857,
+      "step": 11690
+    },
+    {
+      "epoch": 1.894199611147116,
+      "grad_norm": 0.8989560008049011,
+      "learning_rate": 1.5797699378653267e-06,
+      "loss": 0.1111,
+      "step": 11691
+    },
+    {
+      "epoch": 1.894361633182113,
+      "grad_norm": 0.8104327321052551,
+      "learning_rate": 1.5793633512449374e-06,
+      "loss": 0.0974,
+      "step": 11692
+    },
+    {
+      "epoch": 1.8945236552171094,
+      "grad_norm": 0.7694452404975891,
+      "learning_rate": 1.5789567927938016e-06,
+      "loss": 0.0981,
+      "step": 11693
+    },
+    {
+      "epoch": 1.8946856772521063,
+      "grad_norm": 0.8841857314109802,
+      "learning_rate": 1.578550262524359e-06,
+      "loss": 0.101,
+      "step": 11694
+    },
+    {
+      "epoch": 1.894847699287103,
+      "grad_norm": 0.7945596575737,
+      "learning_rate": 1.5781437604490493e-06,
+      "loss": 0.0931,
+      "step": 11695
+    },
+    {
+      "epoch": 1.8950097213220998,
+      "grad_norm": 0.8388197422027588,
+      "learning_rate": 1.5777372865803091e-06,
+      "loss": 0.0982,
+      "step": 11696
+    },
+    {
+      "epoch": 1.8951717433570967,
+      "grad_norm": 0.8069587349891663,
+      "learning_rate": 1.5773308409305766e-06,
+      "loss": 0.0995,
+      "step": 11697
+    },
+    {
+      "epoch": 1.8953337653920932,
+      "grad_norm": 0.7525715231895447,
+      "learning_rate": 1.5769244235122867e-06,
+      "loss": 0.0856,
+      "step": 11698
+    },
+    {
+      "epoch": 1.8954957874270901,
+      "grad_norm": 0.9167400598526001,
+      "learning_rate": 1.576518034337876e-06,
+      "loss": 0.1076,
+      "step": 11699
+    },
+    {
+      "epoch": 1.8956578094620868,
+      "grad_norm": 0.7757900357246399,
+      "learning_rate": 1.576111673419779e-06,
+      "loss": 0.0995,
+      "step": 11700
+    },
+    {
+      "epoch": 1.8958198314970836,
+      "grad_norm": 0.8285884857177734,
+      "learning_rate": 1.5757053407704287e-06,
+      "loss": 0.106,
+      "step": 11701
+    },
+    {
+      "epoch": 1.8959818535320805,
+      "grad_norm": 0.7494697570800781,
+      "learning_rate": 1.5752990364022588e-06,
+      "loss": 0.0959,
+      "step": 11702
+    },
+    {
+      "epoch": 1.896143875567077,
+      "grad_norm": 0.7144713401794434,
+      "learning_rate": 1.5748927603276992e-06,
+      "loss": 0.0885,
+      "step": 11703
+    },
+    {
+      "epoch": 1.896305897602074,
+      "grad_norm": 0.7766308784484863,
+      "learning_rate": 1.5744865125591837e-06,
+      "loss": 0.0969,
+      "step": 11704
+    },
+    {
+      "epoch": 1.8964679196370706,
+      "grad_norm": 0.818439245223999,
+      "learning_rate": 1.574080293109141e-06,
+      "loss": 0.1071,
+      "step": 11705
+    },
+    {
+      "epoch": 1.8966299416720673,
+      "grad_norm": 0.8813479542732239,
+      "learning_rate": 1.573674101990001e-06,
+      "loss": 0.1012,
+      "step": 11706
+    },
+    {
+      "epoch": 1.8967919637070643,
+      "grad_norm": 0.8053517937660217,
+      "learning_rate": 1.5732679392141906e-06,
+      "loss": 0.1002,
+      "step": 11707
+    },
+    {
+      "epoch": 1.8969539857420608,
+      "grad_norm": 0.8213833570480347,
+      "learning_rate": 1.5728618047941393e-06,
+      "loss": 0.1046,
+      "step": 11708
+    },
+    {
+      "epoch": 1.8971160077770577,
+      "grad_norm": 0.9207565784454346,
+      "learning_rate": 1.5724556987422738e-06,
+      "loss": 0.1037,
+      "step": 11709
+    },
+    {
+      "epoch": 1.8972780298120544,
+      "grad_norm": 0.7941535115242004,
+      "learning_rate": 1.5720496210710185e-06,
+      "loss": 0.0841,
+      "step": 11710
+    },
+    {
+      "epoch": 1.8974400518470511,
+      "grad_norm": 0.9520290493965149,
+      "learning_rate": 1.5716435717927996e-06,
+      "loss": 0.1096,
+      "step": 11711
+    },
+    {
+      "epoch": 1.897602073882048,
+      "grad_norm": 0.720905065536499,
+      "learning_rate": 1.5712375509200397e-06,
+      "loss": 0.0864,
+      "step": 11712
+    },
+    {
+      "epoch": 1.8977640959170448,
+      "grad_norm": 0.8394162654876709,
+      "learning_rate": 1.5708315584651646e-06,
+      "loss": 0.1007,
+      "step": 11713
+    },
+    {
+      "epoch": 1.8979261179520415,
+      "grad_norm": 0.9244471192359924,
+      "learning_rate": 1.5704255944405947e-06,
+      "loss": 0.1049,
+      "step": 11714
+    },
+    {
+      "epoch": 1.8980881399870384,
+      "grad_norm": 0.9457299709320068,
+      "learning_rate": 1.5700196588587527e-06,
+      "loss": 0.1148,
+      "step": 11715
+    },
+    {
+      "epoch": 1.898250162022035,
+      "grad_norm": 0.8000479340553284,
+      "learning_rate": 1.5696137517320582e-06,
+      "loss": 0.0928,
+      "step": 11716
+    },
+    {
+      "epoch": 1.8984121840570318,
+      "grad_norm": 0.8423092365264893,
+      "learning_rate": 1.5692078730729304e-06,
+      "loss": 0.1003,
+      "step": 11717
+    },
+    {
+      "epoch": 1.8985742060920285,
+      "grad_norm": 0.8390052914619446,
+      "learning_rate": 1.5688020228937905e-06,
+      "loss": 0.1021,
+      "step": 11718
+    },
+    {
+      "epoch": 1.8987362281270252,
+      "grad_norm": 0.8341346979141235,
+      "learning_rate": 1.5683962012070546e-06,
+      "loss": 0.1038,
+      "step": 11719
+    },
+    {
+      "epoch": 1.8988982501620222,
+      "grad_norm": 0.7826020121574402,
+      "learning_rate": 1.5679904080251414e-06,
+      "loss": 0.1,
+      "step": 11720
+    },
+    {
+      "epoch": 1.8990602721970187,
+      "grad_norm": 0.9335909485816956,
+      "learning_rate": 1.5675846433604658e-06,
+      "loss": 0.1173,
+      "step": 11721
+    },
+    {
+      "epoch": 1.8992222942320156,
+      "grad_norm": 0.8499523401260376,
+      "learning_rate": 1.567178907225443e-06,
+      "loss": 0.1009,
+      "step": 11722
+    },
+    {
+      "epoch": 1.8993843162670123,
+      "grad_norm": 0.8055524826049805,
+      "learning_rate": 1.5667731996324887e-06,
+      "loss": 0.104,
+      "step": 11723
+    },
+    {
+      "epoch": 1.899546338302009,
+      "grad_norm": 0.8389691710472107,
+      "learning_rate": 1.5663675205940164e-06,
+      "loss": 0.1059,
+      "step": 11724
+    },
+    {
+      "epoch": 1.899708360337006,
+      "grad_norm": 0.7978635430335999,
+      "learning_rate": 1.5659618701224385e-06,
+      "loss": 0.0991,
+      "step": 11725
+    },
+    {
+      "epoch": 1.8998703823720025,
+      "grad_norm": 0.8267003297805786,
+      "learning_rate": 1.5655562482301664e-06,
+      "loss": 0.1018,
+      "step": 11726
+    },
+    {
+      "epoch": 1.9000324044069994,
+      "grad_norm": 0.8983197808265686,
+      "learning_rate": 1.565150654929613e-06,
+      "loss": 0.1094,
+      "step": 11727
+    },
+    {
+      "epoch": 1.900194426441996,
+      "grad_norm": 0.7554136514663696,
+      "learning_rate": 1.5647450902331866e-06,
+      "loss": 0.0945,
+      "step": 11728
+    },
+    {
+      "epoch": 1.9003564484769928,
+      "grad_norm": 0.8314508199691772,
+      "learning_rate": 1.5643395541532972e-06,
+      "loss": 0.1002,
+      "step": 11729
+    },
+    {
+      "epoch": 1.9005184705119897,
+      "grad_norm": 0.8851383924484253,
+      "learning_rate": 1.5639340467023534e-06,
+      "loss": 0.1073,
+      "step": 11730
+    },
+    {
+      "epoch": 1.9006804925469862,
+      "grad_norm": 0.7740817070007324,
+      "learning_rate": 1.563528567892762e-06,
+      "loss": 0.0942,
+      "step": 11731
+    },
+    {
+      "epoch": 1.9008425145819832,
+      "grad_norm": 0.7232766151428223,
+      "learning_rate": 1.5631231177369305e-06,
+      "loss": 0.0925,
+      "step": 11732
+    },
+    {
+      "epoch": 1.9010045366169799,
+      "grad_norm": 0.7759274840354919,
+      "learning_rate": 1.562717696247265e-06,
+      "loss": 0.1035,
+      "step": 11733
+    },
+    {
+      "epoch": 1.9011665586519766,
+      "grad_norm": 0.7025986313819885,
+      "learning_rate": 1.562312303436169e-06,
+      "loss": 0.0873,
+      "step": 11734
+    },
+    {
+      "epoch": 1.9013285806869735,
+      "grad_norm": 0.8542094230651855,
+      "learning_rate": 1.561906939316048e-06,
+      "loss": 0.106,
+      "step": 11735
+    },
+    {
+      "epoch": 1.9014906027219702,
+      "grad_norm": 0.7692488431930542,
+      "learning_rate": 1.5615016038993036e-06,
+      "loss": 0.0989,
+      "step": 11736
+    },
+    {
+      "epoch": 1.901652624756967,
+      "grad_norm": 0.7880541682243347,
+      "learning_rate": 1.5610962971983395e-06,
+      "loss": 0.1038,
+      "step": 11737
+    },
+    {
+      "epoch": 1.9018146467919637,
+      "grad_norm": 0.8704632520675659,
+      "learning_rate": 1.5606910192255565e-06,
+      "loss": 0.0971,
+      "step": 11738
+    },
+    {
+      "epoch": 1.9019766688269604,
+      "grad_norm": 0.8276471495628357,
+      "learning_rate": 1.560285769993356e-06,
+      "loss": 0.1017,
+      "step": 11739
+    },
+    {
+      "epoch": 1.9021386908619573,
+      "grad_norm": 0.9419677257537842,
+      "learning_rate": 1.5598805495141362e-06,
+      "loss": 0.1009,
+      "step": 11740
+    },
+    {
+      "epoch": 1.902300712896954,
+      "grad_norm": 0.8188934326171875,
+      "learning_rate": 1.5594753578002957e-06,
+      "loss": 0.1005,
+      "step": 11741
+    },
+    {
+      "epoch": 1.9024627349319507,
+      "grad_norm": 0.7875339984893799,
+      "learning_rate": 1.5590701948642348e-06,
+      "loss": 0.0983,
+      "step": 11742
+    },
+    {
+      "epoch": 1.9026247569669477,
+      "grad_norm": 0.756129264831543,
+      "learning_rate": 1.5586650607183482e-06,
+      "loss": 0.0916,
+      "step": 11743
+    },
+    {
+      "epoch": 1.9027867790019442,
+      "grad_norm": 0.8351160287857056,
+      "learning_rate": 1.5582599553750332e-06,
+      "loss": 0.106,
+      "step": 11744
+    },
+    {
+      "epoch": 1.902948801036941,
+      "grad_norm": 0.8231625556945801,
+      "learning_rate": 1.5578548788466841e-06,
+      "loss": 0.1069,
+      "step": 11745
+    },
+    {
+      "epoch": 1.9031108230719378,
+      "grad_norm": 0.7880319952964783,
+      "learning_rate": 1.5574498311456953e-06,
+      "loss": 0.0998,
+      "step": 11746
+    },
+    {
+      "epoch": 1.9032728451069345,
+      "grad_norm": 0.8101006150245667,
+      "learning_rate": 1.5570448122844612e-06,
+      "loss": 0.0968,
+      "step": 11747
+    },
+    {
+      "epoch": 1.9034348671419314,
+      "grad_norm": 0.7617364525794983,
+      "learning_rate": 1.5566398222753745e-06,
+      "loss": 0.0917,
+      "step": 11748
+    },
+    {
+      "epoch": 1.903596889176928,
+      "grad_norm": 0.8054034113883972,
+      "learning_rate": 1.5562348611308259e-06,
+      "loss": 0.0939,
+      "step": 11749
+    },
+    {
+      "epoch": 1.9037589112119249,
+      "grad_norm": 0.8272371292114258,
+      "learning_rate": 1.5558299288632061e-06,
+      "loss": 0.0987,
+      "step": 11750
+    },
+    {
+      "epoch": 1.9039209332469216,
+      "grad_norm": 0.8034398555755615,
+      "learning_rate": 1.5554250254849065e-06,
+      "loss": 0.0987,
+      "step": 11751
+    },
+    {
+      "epoch": 1.9040829552819183,
+      "grad_norm": 0.681370198726654,
+      "learning_rate": 1.555020151008315e-06,
+      "loss": 0.0833,
+      "step": 11752
+    },
+    {
+      "epoch": 1.9042449773169152,
+      "grad_norm": 0.9123519062995911,
+      "learning_rate": 1.5546153054458204e-06,
+      "loss": 0.1037,
+      "step": 11753
+    },
+    {
+      "epoch": 1.9044069993519117,
+      "grad_norm": 0.8379228115081787,
+      "learning_rate": 1.5542104888098093e-06,
+      "loss": 0.1071,
+      "step": 11754
+    },
+    {
+      "epoch": 1.9045690213869086,
+      "grad_norm": 0.6765666007995605,
+      "learning_rate": 1.5538057011126678e-06,
+      "loss": 0.0874,
+      "step": 11755
+    },
+    {
+      "epoch": 1.9047310434219054,
+      "grad_norm": 0.7915697693824768,
+      "learning_rate": 1.553400942366783e-06,
+      "loss": 0.0974,
+      "step": 11756
+    },
+    {
+      "epoch": 1.904893065456902,
+      "grad_norm": 0.7531107664108276,
+      "learning_rate": 1.5529962125845387e-06,
+      "loss": 0.0915,
+      "step": 11757
+    },
+    {
+      "epoch": 1.905055087491899,
+      "grad_norm": 0.8496636152267456,
+      "learning_rate": 1.5525915117783182e-06,
+      "loss": 0.1105,
+      "step": 11758
+    },
+    {
+      "epoch": 1.9052171095268955,
+      "grad_norm": 0.9320754408836365,
+      "learning_rate": 1.5521868399605057e-06,
+      "loss": 0.1043,
+      "step": 11759
+    },
+    {
+      "epoch": 1.9053791315618924,
+      "grad_norm": 0.80690598487854,
+      "learning_rate": 1.5517821971434804e-06,
+      "loss": 0.105,
+      "step": 11760
+    },
+    {
+      "epoch": 1.9055411535968891,
+      "grad_norm": 0.8309117555618286,
+      "learning_rate": 1.5513775833396263e-06,
+      "loss": 0.0969,
+      "step": 11761
+    },
+    {
+      "epoch": 1.9057031756318858,
+      "grad_norm": 0.7836124300956726,
+      "learning_rate": 1.5509729985613232e-06,
+      "loss": 0.0968,
+      "step": 11762
+    },
+    {
+      "epoch": 1.9058651976668828,
+      "grad_norm": 0.7297149300575256,
+      "learning_rate": 1.5505684428209487e-06,
+      "loss": 0.0931,
+      "step": 11763
+    },
+    {
+      "epoch": 1.9060272197018795,
+      "grad_norm": 0.8328400254249573,
+      "learning_rate": 1.5501639161308829e-06,
+      "loss": 0.104,
+      "step": 11764
+    },
+    {
+      "epoch": 1.9061892417368762,
+      "grad_norm": 0.8439556956291199,
+      "learning_rate": 1.5497594185035015e-06,
+      "loss": 0.1035,
+      "step": 11765
+    },
+    {
+      "epoch": 1.9063512637718731,
+      "grad_norm": 0.7502602934837341,
+      "learning_rate": 1.5493549499511834e-06,
+      "loss": 0.0911,
+      "step": 11766
+    },
+    {
+      "epoch": 1.9065132858068696,
+      "grad_norm": 0.9024834632873535,
+      "learning_rate": 1.5489505104863034e-06,
+      "loss": 0.1085,
+      "step": 11767
+    },
+    {
+      "epoch": 1.9066753078418666,
+      "grad_norm": 0.8212124705314636,
+      "learning_rate": 1.5485461001212365e-06,
+      "loss": 0.1028,
+      "step": 11768
+    },
+    {
+      "epoch": 1.9068373298768633,
+      "grad_norm": 0.8439403176307678,
+      "learning_rate": 1.5481417188683557e-06,
+      "loss": 0.1067,
+      "step": 11769
+    },
+    {
+      "epoch": 1.90699935191186,
+      "grad_norm": 0.811410665512085,
+      "learning_rate": 1.5477373667400347e-06,
+      "loss": 0.0983,
+      "step": 11770
+    },
+    {
+      "epoch": 1.907161373946857,
+      "grad_norm": 0.8574284315109253,
+      "learning_rate": 1.5473330437486466e-06,
+      "loss": 0.1048,
+      "step": 11771
+    },
+    {
+      "epoch": 1.9073233959818534,
+      "grad_norm": 0.7073046565055847,
+      "learning_rate": 1.5469287499065615e-06,
+      "loss": 0.0905,
+      "step": 11772
+    },
+    {
+      "epoch": 1.9074854180168503,
+      "grad_norm": 0.7401828169822693,
+      "learning_rate": 1.5465244852261505e-06,
+      "loss": 0.0906,
+      "step": 11773
+    },
+    {
+      "epoch": 1.907647440051847,
+      "grad_norm": 0.7841954827308655,
+      "learning_rate": 1.5461202497197821e-06,
+      "loss": 0.0979,
+      "step": 11774
+    },
+    {
+      "epoch": 1.9078094620868438,
+      "grad_norm": 0.7952117919921875,
+      "learning_rate": 1.545716043399827e-06,
+      "loss": 0.0935,
+      "step": 11775
+    },
+    {
+      "epoch": 1.9079714841218407,
+      "grad_norm": 0.7880851626396179,
+      "learning_rate": 1.5453118662786509e-06,
+      "loss": 0.1009,
+      "step": 11776
+    },
+    {
+      "epoch": 1.9081335061568372,
+      "grad_norm": 0.849980354309082,
+      "learning_rate": 1.544907718368622e-06,
+      "loss": 0.1105,
+      "step": 11777
+    },
+    {
+      "epoch": 1.9082955281918341,
+      "grad_norm": 0.858780026435852,
+      "learning_rate": 1.544503599682105e-06,
+      "loss": 0.1007,
+      "step": 11778
+    },
+    {
+      "epoch": 1.9084575502268308,
+      "grad_norm": 0.7746371030807495,
+      "learning_rate": 1.5440995102314654e-06,
+      "loss": 0.0995,
+      "step": 11779
+    },
+    {
+      "epoch": 1.9086195722618275,
+      "grad_norm": 0.8200014233589172,
+      "learning_rate": 1.5436954500290684e-06,
+      "loss": 0.0951,
+      "step": 11780
+    },
+    {
+      "epoch": 1.9087815942968245,
+      "grad_norm": 0.9434183835983276,
+      "learning_rate": 1.5432914190872757e-06,
+      "loss": 0.1081,
+      "step": 11781
+    },
+    {
+      "epoch": 1.908943616331821,
+      "grad_norm": 0.7809907793998718,
+      "learning_rate": 1.5428874174184509e-06,
+      "loss": 0.0902,
+      "step": 11782
+    },
+    {
+      "epoch": 1.909105638366818,
+      "grad_norm": 0.8580632209777832,
+      "learning_rate": 1.5424834450349552e-06,
+      "loss": 0.103,
+      "step": 11783
+    },
+    {
+      "epoch": 1.9092676604018146,
+      "grad_norm": 0.8701772689819336,
+      "learning_rate": 1.5420795019491475e-06,
+      "loss": 0.1071,
+      "step": 11784
+    },
+    {
+      "epoch": 1.9094296824368113,
+      "grad_norm": 0.784704327583313,
+      "learning_rate": 1.54167558817339e-06,
+      "loss": 0.0969,
+      "step": 11785
+    },
+    {
+      "epoch": 1.9095917044718083,
+      "grad_norm": 0.7510059475898743,
+      "learning_rate": 1.5412717037200406e-06,
+      "loss": 0.0939,
+      "step": 11786
+    },
+    {
+      "epoch": 1.909753726506805,
+      "grad_norm": 0.7119048237800598,
+      "learning_rate": 1.5408678486014567e-06,
+      "loss": 0.0859,
+      "step": 11787
+    },
+    {
+      "epoch": 1.9099157485418017,
+      "grad_norm": 0.7860952615737915,
+      "learning_rate": 1.540464022829996e-06,
+      "loss": 0.095,
+      "step": 11788
+    },
+    {
+      "epoch": 1.9100777705767984,
+      "grad_norm": 0.9239879846572876,
+      "learning_rate": 1.540060226418013e-06,
+      "loss": 0.1001,
+      "step": 11789
+    },
+    {
+      "epoch": 1.910239792611795,
+      "grad_norm": 0.7606763243675232,
+      "learning_rate": 1.5396564593778646e-06,
+      "loss": 0.0849,
+      "step": 11790
+    },
+    {
+      "epoch": 1.910401814646792,
+      "grad_norm": 0.830071747303009,
+      "learning_rate": 1.5392527217219047e-06,
+      "loss": 0.0984,
+      "step": 11791
+    },
+    {
+      "epoch": 1.9105638366817888,
+      "grad_norm": 0.8571628332138062,
+      "learning_rate": 1.538849013462487e-06,
+      "loss": 0.103,
+      "step": 11792
+    },
+    {
+      "epoch": 1.9107258587167855,
+      "grad_norm": 0.8239210844039917,
+      "learning_rate": 1.5384453346119628e-06,
+      "loss": 0.0916,
+      "step": 11793
+    },
+    {
+      "epoch": 1.9108878807517824,
+      "grad_norm": 0.7969984412193298,
+      "learning_rate": 1.5380416851826845e-06,
+      "loss": 0.0952,
+      "step": 11794
+    },
+    {
+      "epoch": 1.9110499027867789,
+      "grad_norm": 0.8779683113098145,
+      "learning_rate": 1.5376380651870033e-06,
+      "loss": 0.1102,
+      "step": 11795
+    },
+    {
+      "epoch": 1.9112119248217758,
+      "grad_norm": 0.8378877639770508,
+      "learning_rate": 1.537234474637268e-06,
+      "loss": 0.1004,
+      "step": 11796
+    },
+    {
+      "epoch": 1.9113739468567725,
+      "grad_norm": 0.8366298675537109,
+      "learning_rate": 1.5368309135458287e-06,
+      "loss": 0.1061,
+      "step": 11797
+    },
+    {
+      "epoch": 1.9115359688917692,
+      "grad_norm": 0.8169813752174377,
+      "learning_rate": 1.5364273819250308e-06,
+      "loss": 0.0967,
+      "step": 11798
+    },
+    {
+      "epoch": 1.9116979909267662,
+      "grad_norm": 0.880442202091217,
+      "learning_rate": 1.536023879787225e-06,
+      "loss": 0.1037,
+      "step": 11799
+    },
+    {
+      "epoch": 1.9118600129617627,
+      "grad_norm": 0.8345661759376526,
+      "learning_rate": 1.535620407144755e-06,
+      "loss": 0.0997,
+      "step": 11800
+    },
+    {
+      "epoch": 1.9120220349967596,
+      "grad_norm": 0.9036890268325806,
+      "learning_rate": 1.5352169640099673e-06,
+      "loss": 0.1058,
+      "step": 11801
+    },
+    {
+      "epoch": 1.9121840570317563,
+      "grad_norm": 0.7998917102813721,
+      "learning_rate": 1.534813550395205e-06,
+      "loss": 0.1014,
+      "step": 11802
+    },
+    {
+      "epoch": 1.912346079066753,
+      "grad_norm": 0.8374770879745483,
+      "learning_rate": 1.5344101663128121e-06,
+      "loss": 0.1047,
+      "step": 11803
+    },
+    {
+      "epoch": 1.91250810110175,
+      "grad_norm": 0.8265429735183716,
+      "learning_rate": 1.5340068117751329e-06,
+      "loss": 0.098,
+      "step": 11804
+    },
+    {
+      "epoch": 1.9126701231367464,
+      "grad_norm": 0.7321347594261169,
+      "learning_rate": 1.5336034867945065e-06,
+      "loss": 0.0839,
+      "step": 11805
+    },
+    {
+      "epoch": 1.9128321451717434,
+      "grad_norm": 0.7928110361099243,
+      "learning_rate": 1.5332001913832754e-06,
+      "loss": 0.0964,
+      "step": 11806
+    },
+    {
+      "epoch": 1.91299416720674,
+      "grad_norm": 0.7539445161819458,
+      "learning_rate": 1.5327969255537784e-06,
+      "loss": 0.0951,
+      "step": 11807
+    },
+    {
+      "epoch": 1.9131561892417368,
+      "grad_norm": 0.8430777192115784,
+      "learning_rate": 1.5323936893183542e-06,
+      "loss": 0.1087,
+      "step": 11808
+    },
+    {
+      "epoch": 1.9133182112767337,
+      "grad_norm": 0.8476277589797974,
+      "learning_rate": 1.5319904826893422e-06,
+      "loss": 0.1013,
+      "step": 11809
+    },
+    {
+      "epoch": 1.9134802333117304,
+      "grad_norm": 0.8722755908966064,
+      "learning_rate": 1.5315873056790791e-06,
+      "loss": 0.0985,
+      "step": 11810
+    },
+    {
+      "epoch": 1.9136422553467272,
+      "grad_norm": 0.708776593208313,
+      "learning_rate": 1.5311841582999009e-06,
+      "loss": 0.0864,
+      "step": 11811
+    },
+    {
+      "epoch": 1.9138042773817239,
+      "grad_norm": 0.8320741057395935,
+      "learning_rate": 1.5307810405641433e-06,
+      "loss": 0.0971,
+      "step": 11812
+    },
+    {
+      "epoch": 1.9139662994167206,
+      "grad_norm": 0.79482501745224,
+      "learning_rate": 1.5303779524841395e-06,
+      "loss": 0.1019,
+      "step": 11813
+    },
+    {
+      "epoch": 1.9141283214517175,
+      "grad_norm": 0.8025671243667603,
+      "learning_rate": 1.5299748940722241e-06,
+      "loss": 0.0934,
+      "step": 11814
+    },
+    {
+      "epoch": 1.9142903434867142,
+      "grad_norm": 0.9434292316436768,
+      "learning_rate": 1.5295718653407305e-06,
+      "loss": 0.1017,
+      "step": 11815
+    },
+    {
+      "epoch": 1.914452365521711,
+      "grad_norm": 0.734461784362793,
+      "learning_rate": 1.5291688663019885e-06,
+      "loss": 0.0876,
+      "step": 11816
+    },
+    {
+      "epoch": 1.9146143875567079,
+      "grad_norm": 0.858860194683075,
+      "learning_rate": 1.5287658969683294e-06,
+      "loss": 0.105,
+      "step": 11817
+    },
+    {
+      "epoch": 1.9147764095917044,
+      "grad_norm": 0.7746593356132507,
+      "learning_rate": 1.5283629573520841e-06,
+      "loss": 0.0921,
+      "step": 11818
+    },
+    {
+      "epoch": 1.9149384316267013,
+      "grad_norm": 0.7499451637268066,
+      "learning_rate": 1.5279600474655814e-06,
+      "loss": 0.0905,
+      "step": 11819
+    },
+    {
+      "epoch": 1.915100453661698,
+      "grad_norm": 0.7706909775733948,
+      "learning_rate": 1.5275571673211487e-06,
+      "loss": 0.0979,
+      "step": 11820
+    },
+    {
+      "epoch": 1.9152624756966947,
+      "grad_norm": 0.9270902872085571,
+      "learning_rate": 1.5271543169311137e-06,
+      "loss": 0.1123,
+      "step": 11821
+    },
+    {
+      "epoch": 1.9154244977316917,
+      "grad_norm": 0.9533674120903015,
+      "learning_rate": 1.5267514963078014e-06,
+      "loss": 0.1095,
+      "step": 11822
+    },
+    {
+      "epoch": 1.9155865197666881,
+      "grad_norm": 0.6820551156997681,
+      "learning_rate": 1.5263487054635386e-06,
+      "loss": 0.0837,
+      "step": 11823
+    },
+    {
+      "epoch": 1.915748541801685,
+      "grad_norm": 0.7506440281867981,
+      "learning_rate": 1.5259459444106497e-06,
+      "loss": 0.0904,
+      "step": 11824
+    },
+    {
+      "epoch": 1.9159105638366818,
+      "grad_norm": 0.7518582344055176,
+      "learning_rate": 1.5255432131614572e-06,
+      "loss": 0.0924,
+      "step": 11825
+    },
+    {
+      "epoch": 1.9160725858716785,
+      "grad_norm": 0.7959820032119751,
+      "learning_rate": 1.5251405117282843e-06,
+      "loss": 0.0992,
+      "step": 11826
+    },
+    {
+      "epoch": 1.9162346079066754,
+      "grad_norm": 0.8564347624778748,
+      "learning_rate": 1.5247378401234521e-06,
+      "loss": 0.1062,
+      "step": 11827
+    },
+    {
+      "epoch": 1.916396629941672,
+      "grad_norm": 0.8178135752677917,
+      "learning_rate": 1.524335198359283e-06,
+      "loss": 0.0948,
+      "step": 11828
+    },
+    {
+      "epoch": 1.9165586519766689,
+      "grad_norm": 0.8334271907806396,
+      "learning_rate": 1.5239325864480952e-06,
+      "loss": 0.1052,
+      "step": 11829
+    },
+    {
+      "epoch": 1.9167206740116656,
+      "grad_norm": 0.7920629978179932,
+      "learning_rate": 1.5235300044022088e-06,
+      "loss": 0.0976,
+      "step": 11830
+    },
+    {
+      "epoch": 1.9168826960466623,
+      "grad_norm": 0.7598364353179932,
+      "learning_rate": 1.5231274522339407e-06,
+      "loss": 0.091,
+      "step": 11831
+    },
+    {
+      "epoch": 1.9170447180816592,
+      "grad_norm": 0.8865482211112976,
+      "learning_rate": 1.522724929955608e-06,
+      "loss": 0.1087,
+      "step": 11832
+    },
+    {
+      "epoch": 1.9172067401166557,
+      "grad_norm": 0.7840036153793335,
+      "learning_rate": 1.5223224375795287e-06,
+      "loss": 0.0972,
+      "step": 11833
+    },
+    {
+      "epoch": 1.9173687621516526,
+      "grad_norm": 0.7893587350845337,
+      "learning_rate": 1.5219199751180162e-06,
+      "loss": 0.0954,
+      "step": 11834
+    },
+    {
+      "epoch": 1.9175307841866494,
+      "grad_norm": 0.7772044539451599,
+      "learning_rate": 1.5215175425833855e-06,
+      "loss": 0.0893,
+      "step": 11835
+    },
+    {
+      "epoch": 1.917692806221646,
+      "grad_norm": 0.8612580299377441,
+      "learning_rate": 1.5211151399879505e-06,
+      "loss": 0.0945,
+      "step": 11836
+    },
+    {
+      "epoch": 1.917854828256643,
+      "grad_norm": 0.8095962405204773,
+      "learning_rate": 1.5207127673440226e-06,
+      "loss": 0.1019,
+      "step": 11837
+    },
+    {
+      "epoch": 1.9180168502916397,
+      "grad_norm": 0.794070839881897,
+      "learning_rate": 1.5203104246639144e-06,
+      "loss": 0.0995,
+      "step": 11838
+    },
+    {
+      "epoch": 1.9181788723266364,
+      "grad_norm": 0.8148672580718994,
+      "learning_rate": 1.5199081119599373e-06,
+      "loss": 0.1067,
+      "step": 11839
+    },
+    {
+      "epoch": 1.9183408943616331,
+      "grad_norm": 0.8468709588050842,
+      "learning_rate": 1.5195058292443996e-06,
+      "loss": 0.1072,
+      "step": 11840
+    },
+    {
+      "epoch": 1.9185029163966298,
+      "grad_norm": 0.8336006999015808,
+      "learning_rate": 1.5191035765296104e-06,
+      "loss": 0.1073,
+      "step": 11841
+    },
+    {
+      "epoch": 1.9186649384316268,
+      "grad_norm": 0.7484931945800781,
+      "learning_rate": 1.518701353827878e-06,
+      "loss": 0.0902,
+      "step": 11842
+    },
+    {
+      "epoch": 1.9188269604666235,
+      "grad_norm": 0.8069817423820496,
+      "learning_rate": 1.5182991611515102e-06,
+      "loss": 0.0964,
+      "step": 11843
+    },
+    {
+      "epoch": 1.9189889825016202,
+      "grad_norm": 0.7606700658798218,
+      "learning_rate": 1.5178969985128122e-06,
+      "loss": 0.0955,
+      "step": 11844
+    },
+    {
+      "epoch": 1.9191510045366171,
+      "grad_norm": 0.936642050743103,
+      "learning_rate": 1.5174948659240896e-06,
+      "loss": 0.1097,
+      "step": 11845
+    },
+    {
+      "epoch": 1.9193130265716136,
+      "grad_norm": 0.827983021736145,
+      "learning_rate": 1.5170927633976457e-06,
+      "loss": 0.1045,
+      "step": 11846
+    },
+    {
+      "epoch": 1.9194750486066106,
+      "grad_norm": 0.8345808386802673,
+      "learning_rate": 1.516690690945785e-06,
+      "loss": 0.0996,
+      "step": 11847
+    },
+    {
+      "epoch": 1.9196370706416073,
+      "grad_norm": 0.8033564686775208,
+      "learning_rate": 1.5162886485808102e-06,
+      "loss": 0.1027,
+      "step": 11848
+    },
+    {
+      "epoch": 1.919799092676604,
+      "grad_norm": 0.8006458878517151,
+      "learning_rate": 1.5158866363150215e-06,
+      "loss": 0.1001,
+      "step": 11849
+    },
+    {
+      "epoch": 1.919961114711601,
+      "grad_norm": 0.8043130040168762,
+      "learning_rate": 1.515484654160721e-06,
+      "loss": 0.1028,
+      "step": 11850
+    },
+    {
+      "epoch": 1.9201231367465974,
+      "grad_norm": 0.8533475995063782,
+      "learning_rate": 1.515082702130206e-06,
+      "loss": 0.1065,
+      "step": 11851
+    },
+    {
+      "epoch": 1.9202851587815943,
+      "grad_norm": 0.9136732816696167,
+      "learning_rate": 1.5146807802357782e-06,
+      "loss": 0.1156,
+      "step": 11852
+    },
+    {
+      "epoch": 1.920447180816591,
+      "grad_norm": 0.8210397958755493,
+      "learning_rate": 1.5142788884897336e-06,
+      "loss": 0.1002,
+      "step": 11853
+    },
+    {
+      "epoch": 1.9206092028515878,
+      "grad_norm": 0.906101405620575,
+      "learning_rate": 1.5138770269043704e-06,
+      "loss": 0.1062,
+      "step": 11854
+    },
+    {
+      "epoch": 1.9207712248865847,
+      "grad_norm": 0.8309211730957031,
+      "learning_rate": 1.5134751954919833e-06,
+      "loss": 0.0985,
+      "step": 11855
+    },
+    {
+      "epoch": 1.9209332469215812,
+      "grad_norm": 0.9222484230995178,
+      "learning_rate": 1.513073394264867e-06,
+      "loss": 0.1017,
+      "step": 11856
+    },
+    {
+      "epoch": 1.9210952689565781,
+      "grad_norm": 0.7615657448768616,
+      "learning_rate": 1.5126716232353178e-06,
+      "loss": 0.0931,
+      "step": 11857
+    },
+    {
+      "epoch": 1.9212572909915748,
+      "grad_norm": 0.823805570602417,
+      "learning_rate": 1.5122698824156271e-06,
+      "loss": 0.0979,
+      "step": 11858
+    },
+    {
+      "epoch": 1.9214193130265715,
+      "grad_norm": 0.801001787185669,
+      "learning_rate": 1.5118681718180883e-06,
+      "loss": 0.1006,
+      "step": 11859
+    },
+    {
+      "epoch": 1.9215813350615685,
+      "grad_norm": 0.7479082345962524,
+      "learning_rate": 1.5114664914549903e-06,
+      "loss": 0.0909,
+      "step": 11860
+    },
+    {
+      "epoch": 1.9217433570965652,
+      "grad_norm": 0.662217915058136,
+      "learning_rate": 1.5110648413386275e-06,
+      "loss": 0.0826,
+      "step": 11861
+    },
+    {
+      "epoch": 1.921905379131562,
+      "grad_norm": 0.837660014629364,
+      "learning_rate": 1.5106632214812865e-06,
+      "loss": 0.1019,
+      "step": 11862
+    },
+    {
+      "epoch": 1.9220674011665586,
+      "grad_norm": 0.8495209813117981,
+      "learning_rate": 1.5102616318952575e-06,
+      "loss": 0.105,
+      "step": 11863
+    },
+    {
+      "epoch": 1.9222294232015553,
+      "grad_norm": 0.723486065864563,
+      "learning_rate": 1.5098600725928269e-06,
+      "loss": 0.0883,
+      "step": 11864
+    },
+    {
+      "epoch": 1.9223914452365523,
+      "grad_norm": 0.7528011798858643,
+      "learning_rate": 1.5094585435862817e-06,
+      "loss": 0.0919,
+      "step": 11865
+    },
+    {
+      "epoch": 1.922553467271549,
+      "grad_norm": 0.9446701407432556,
+      "learning_rate": 1.5090570448879088e-06,
+      "loss": 0.115,
+      "step": 11866
+    },
+    {
+      "epoch": 1.9227154893065457,
+      "grad_norm": 0.8201817274093628,
+      "learning_rate": 1.5086555765099916e-06,
+      "loss": 0.1131,
+      "step": 11867
+    },
+    {
+      "epoch": 1.9228775113415426,
+      "grad_norm": 0.6670837998390198,
+      "learning_rate": 1.5082541384648154e-06,
+      "loss": 0.0816,
+      "step": 11868
+    },
+    {
+      "epoch": 1.923039533376539,
+      "grad_norm": 0.6985659003257751,
+      "learning_rate": 1.5078527307646622e-06,
+      "loss": 0.0847,
+      "step": 11869
+    },
+    {
+      "epoch": 1.923201555411536,
+      "grad_norm": 0.8145870566368103,
+      "learning_rate": 1.5074513534218137e-06,
+      "loss": 0.1065,
+      "step": 11870
+    },
+    {
+      "epoch": 1.9233635774465327,
+      "grad_norm": 0.7868598699569702,
+      "learning_rate": 1.5070500064485527e-06,
+      "loss": 0.0941,
+      "step": 11871
+    },
+    {
+      "epoch": 1.9235255994815295,
+      "grad_norm": 0.8277310729026794,
+      "learning_rate": 1.5066486898571588e-06,
+      "loss": 0.0937,
+      "step": 11872
+    },
+    {
+      "epoch": 1.9236876215165264,
+      "grad_norm": 0.752480685710907,
+      "learning_rate": 1.5062474036599109e-06,
+      "loss": 0.1008,
+      "step": 11873
+    },
+    {
+      "epoch": 1.9238496435515229,
+      "grad_norm": 0.7744075655937195,
+      "learning_rate": 1.5058461478690878e-06,
+      "loss": 0.0879,
+      "step": 11874
+    },
+    {
+      "epoch": 1.9240116655865198,
+      "grad_norm": 0.8470804691314697,
+      "learning_rate": 1.505444922496966e-06,
+      "loss": 0.0971,
+      "step": 11875
+    },
+    {
+      "epoch": 1.9241736876215165,
+      "grad_norm": 0.6907162666320801,
+      "learning_rate": 1.5050437275558233e-06,
+      "loss": 0.0813,
+      "step": 11876
+    },
+    {
+      "epoch": 1.9243357096565132,
+      "grad_norm": 0.8127276301383972,
+      "learning_rate": 1.5046425630579348e-06,
+      "loss": 0.1044,
+      "step": 11877
+    },
+    {
+      "epoch": 1.9244977316915102,
+      "grad_norm": 0.8600353598594666,
+      "learning_rate": 1.5042414290155754e-06,
+      "loss": 0.1021,
+      "step": 11878
+    },
+    {
+      "epoch": 1.9246597537265067,
+      "grad_norm": 0.7428255081176758,
+      "learning_rate": 1.5038403254410183e-06,
+      "loss": 0.0966,
+      "step": 11879
+    },
+    {
+      "epoch": 1.9248217757615036,
+      "grad_norm": 0.8952332139015198,
+      "learning_rate": 1.5034392523465364e-06,
+      "loss": 0.097,
+      "step": 11880
+    },
+    {
+      "epoch": 1.9249837977965003,
+      "grad_norm": 0.7860379815101624,
+      "learning_rate": 1.5030382097444025e-06,
+      "loss": 0.0909,
+      "step": 11881
+    },
+    {
+      "epoch": 1.925145819831497,
+      "grad_norm": 0.8410727977752686,
+      "learning_rate": 1.502637197646886e-06,
+      "loss": 0.0759,
+      "step": 11882
+    },
+    {
+      "epoch": 1.925307841866494,
+      "grad_norm": 0.8401789665222168,
+      "learning_rate": 1.5022362160662584e-06,
+      "loss": 0.1057,
+      "step": 11883
+    },
+    {
+      "epoch": 1.9254698639014904,
+      "grad_norm": 0.7923102378845215,
+      "learning_rate": 1.5018352650147872e-06,
+      "loss": 0.0931,
+      "step": 11884
+    },
+    {
+      "epoch": 1.9256318859364874,
+      "grad_norm": 0.8386868834495544,
+      "learning_rate": 1.5014343445047414e-06,
+      "loss": 0.1025,
+      "step": 11885
+    },
+    {
+      "epoch": 1.925793907971484,
+      "grad_norm": 0.723055362701416,
+      "learning_rate": 1.5010334545483885e-06,
+      "loss": 0.0889,
+      "step": 11886
+    },
+    {
+      "epoch": 1.9259559300064808,
+      "grad_norm": 0.8790740966796875,
+      "learning_rate": 1.5006325951579948e-06,
+      "loss": 0.1046,
+      "step": 11887
+    },
+    {
+      "epoch": 1.9261179520414777,
+      "grad_norm": 0.8958154320716858,
+      "learning_rate": 1.500231766345825e-06,
+      "loss": 0.1142,
+      "step": 11888
+    },
+    {
+      "epoch": 1.9262799740764744,
+      "grad_norm": 0.7713485360145569,
+      "learning_rate": 1.499830968124143e-06,
+      "loss": 0.1023,
+      "step": 11889
+    },
+    {
+      "epoch": 1.9264419961114712,
+      "grad_norm": 0.8863093852996826,
+      "learning_rate": 1.4994302005052141e-06,
+      "loss": 0.1149,
+      "step": 11890
+    },
+    {
+      "epoch": 1.9266040181464679,
+      "grad_norm": 0.7300902009010315,
+      "learning_rate": 1.4990294635012995e-06,
+      "loss": 0.0852,
+      "step": 11891
+    },
+    {
+      "epoch": 1.9267660401814646,
+      "grad_norm": 0.7891831398010254,
+      "learning_rate": 1.4986287571246614e-06,
+      "loss": 0.1016,
+      "step": 11892
+    },
+    {
+      "epoch": 1.9269280622164615,
+      "grad_norm": 0.8113868236541748,
+      "learning_rate": 1.4982280813875593e-06,
+      "loss": 0.1071,
+      "step": 11893
+    },
+    {
+      "epoch": 1.9270900842514582,
+      "grad_norm": 0.8107513189315796,
+      "learning_rate": 1.4978274363022532e-06,
+      "loss": 0.0996,
+      "step": 11894
+    },
+    {
+      "epoch": 1.927252106286455,
+      "grad_norm": 0.7865254878997803,
+      "learning_rate": 1.497426821881003e-06,
+      "loss": 0.0956,
+      "step": 11895
+    },
+    {
+      "epoch": 1.9274141283214519,
+      "grad_norm": 0.8345635533332825,
+      "learning_rate": 1.4970262381360664e-06,
+      "loss": 0.1072,
+      "step": 11896
+    },
+    {
+      "epoch": 1.9275761503564484,
+      "grad_norm": 0.7102372050285339,
+      "learning_rate": 1.4966256850796993e-06,
+      "loss": 0.0904,
+      "step": 11897
+    },
+    {
+      "epoch": 1.9277381723914453,
+      "grad_norm": 0.8325737118721008,
+      "learning_rate": 1.4962251627241583e-06,
+      "loss": 0.0958,
+      "step": 11898
+    },
+    {
+      "epoch": 1.927900194426442,
+      "grad_norm": 0.8160897493362427,
+      "learning_rate": 1.4958246710816976e-06,
+      "loss": 0.101,
+      "step": 11899
+    },
+    {
+      "epoch": 1.9280622164614387,
+      "grad_norm": 0.794330358505249,
+      "learning_rate": 1.4954242101645722e-06,
+      "loss": 0.0976,
+      "step": 11900
+    },
+    {
+      "epoch": 1.9282242384964356,
+      "grad_norm": 0.8225795030593872,
+      "learning_rate": 1.4950237799850354e-06,
+      "loss": 0.104,
+      "step": 11901
+    },
+    {
+      "epoch": 1.9283862605314321,
+      "grad_norm": 0.7986380457878113,
+      "learning_rate": 1.4946233805553387e-06,
+      "loss": 0.0807,
+      "step": 11902
+    },
+    {
+      "epoch": 1.928548282566429,
+      "grad_norm": 0.7666733860969543,
+      "learning_rate": 1.4942230118877337e-06,
+      "loss": 0.0926,
+      "step": 11903
+    },
+    {
+      "epoch": 1.9287103046014258,
+      "grad_norm": 0.8928430676460266,
+      "learning_rate": 1.4938226739944694e-06,
+      "loss": 0.1056,
+      "step": 11904
+    },
+    {
+      "epoch": 1.9288723266364225,
+      "grad_norm": 0.7845842838287354,
+      "learning_rate": 1.4934223668877979e-06,
+      "loss": 0.0987,
+      "step": 11905
+    },
+    {
+      "epoch": 1.9290343486714194,
+      "grad_norm": 0.8057361245155334,
+      "learning_rate": 1.4930220905799652e-06,
+      "loss": 0.0966,
+      "step": 11906
+    },
+    {
+      "epoch": 1.929196370706416,
+      "grad_norm": 0.7778323888778687,
+      "learning_rate": 1.4926218450832208e-06,
+      "loss": 0.0946,
+      "step": 11907
+    },
+    {
+      "epoch": 1.9293583927414129,
+      "grad_norm": 0.884856641292572,
+      "learning_rate": 1.4922216304098085e-06,
+      "loss": 0.1084,
+      "step": 11908
+    },
+    {
+      "epoch": 1.9295204147764096,
+      "grad_norm": 0.9439958333969116,
+      "learning_rate": 1.4918214465719765e-06,
+      "loss": 0.1146,
+      "step": 11909
+    },
+    {
+      "epoch": 1.9296824368114063,
+      "grad_norm": 0.7851779460906982,
+      "learning_rate": 1.4914212935819689e-06,
+      "loss": 0.0973,
+      "step": 11910
+    },
+    {
+      "epoch": 1.9298444588464032,
+      "grad_norm": 0.8911312818527222,
+      "learning_rate": 1.4910211714520285e-06,
+      "loss": 0.1109,
+      "step": 11911
+    },
+    {
+      "epoch": 1.9300064808814,
+      "grad_norm": 0.8925055265426636,
+      "learning_rate": 1.4906210801943985e-06,
+      "loss": 0.1051,
+      "step": 11912
+    },
+    {
+      "epoch": 1.9301685029163966,
+      "grad_norm": 0.8144211769104004,
+      "learning_rate": 1.4902210198213203e-06,
+      "loss": 0.1023,
+      "step": 11913
+    },
+    {
+      "epoch": 1.9303305249513933,
+      "grad_norm": 0.8194781541824341,
+      "learning_rate": 1.4898209903450361e-06,
+      "loss": 0.0997,
+      "step": 11914
+    },
+    {
+      "epoch": 1.93049254698639,
+      "grad_norm": 0.8512380123138428,
+      "learning_rate": 1.489420991777785e-06,
+      "loss": 0.1035,
+      "step": 11915
+    },
+    {
+      "epoch": 1.930654569021387,
+      "grad_norm": 0.8246690630912781,
+      "learning_rate": 1.489021024131806e-06,
+      "loss": 0.107,
+      "step": 11916
+    },
+    {
+      "epoch": 1.9308165910563837,
+      "grad_norm": 0.8465759754180908,
+      "learning_rate": 1.4886210874193368e-06,
+      "loss": 0.0924,
+      "step": 11917
+    },
+    {
+      "epoch": 1.9309786130913804,
+      "grad_norm": 0.757154643535614,
+      "learning_rate": 1.4882211816526144e-06,
+      "loss": 0.0958,
+      "step": 11918
+    },
+    {
+      "epoch": 1.9311406351263773,
+      "grad_norm": 0.8065274357795715,
+      "learning_rate": 1.4878213068438762e-06,
+      "loss": 0.0939,
+      "step": 11919
+    },
+    {
+      "epoch": 1.9313026571613738,
+      "grad_norm": 0.8495281934738159,
+      "learning_rate": 1.4874214630053562e-06,
+      "loss": 0.0914,
+      "step": 11920
+    },
+    {
+      "epoch": 1.9314646791963708,
+      "grad_norm": 0.8558754920959473,
+      "learning_rate": 1.4870216501492892e-06,
+      "loss": 0.1046,
+      "step": 11921
+    },
+    {
+      "epoch": 1.9316267012313675,
+      "grad_norm": 0.7717517018318176,
+      "learning_rate": 1.4866218682879088e-06,
+      "loss": 0.0917,
+      "step": 11922
+    },
+    {
+      "epoch": 1.9317887232663642,
+      "grad_norm": 0.7911754250526428,
+      "learning_rate": 1.4862221174334457e-06,
+      "loss": 0.0999,
+      "step": 11923
+    },
+    {
+      "epoch": 1.9319507453013611,
+      "grad_norm": 0.7910676598548889,
+      "learning_rate": 1.4858223975981334e-06,
+      "loss": 0.1002,
+      "step": 11924
+    },
+    {
+      "epoch": 1.9321127673363576,
+      "grad_norm": 0.816967785358429,
+      "learning_rate": 1.4854227087942016e-06,
+      "loss": 0.0959,
+      "step": 11925
+    },
+    {
+      "epoch": 1.9322747893713546,
+      "grad_norm": 0.828427255153656,
+      "learning_rate": 1.4850230510338792e-06,
+      "loss": 0.1038,
+      "step": 11926
+    },
+    {
+      "epoch": 1.9324368114063513,
+      "grad_norm": 0.7696671485900879,
+      "learning_rate": 1.484623424329395e-06,
+      "loss": 0.0931,
+      "step": 11927
+    },
+    {
+      "epoch": 1.932598833441348,
+      "grad_norm": 0.8365792036056519,
+      "learning_rate": 1.4842238286929777e-06,
+      "loss": 0.1076,
+      "step": 11928
+    },
+    {
+      "epoch": 1.932760855476345,
+      "grad_norm": 0.7698932886123657,
+      "learning_rate": 1.4838242641368526e-06,
+      "loss": 0.089,
+      "step": 11929
+    },
+    {
+      "epoch": 1.9329228775113414,
+      "grad_norm": 0.8664256930351257,
+      "learning_rate": 1.4834247306732457e-06,
+      "loss": 0.1108,
+      "step": 11930
+    },
+    {
+      "epoch": 1.9330848995463383,
+      "grad_norm": 0.9179291129112244,
+      "learning_rate": 1.4830252283143825e-06,
+      "loss": 0.107,
+      "step": 11931
+    },
+    {
+      "epoch": 1.933246921581335,
+      "grad_norm": 0.798081636428833,
+      "learning_rate": 1.4826257570724856e-06,
+      "loss": 0.0952,
+      "step": 11932
+    },
+    {
+      "epoch": 1.9334089436163318,
+      "grad_norm": 0.7657841444015503,
+      "learning_rate": 1.4822263169597789e-06,
+      "loss": 0.0892,
+      "step": 11933
+    },
+    {
+      "epoch": 1.9335709656513287,
+      "grad_norm": 0.7410942316055298,
+      "learning_rate": 1.4818269079884845e-06,
+      "loss": 0.091,
+      "step": 11934
+    },
+    {
+      "epoch": 1.9337329876863252,
+      "grad_norm": 0.8873445987701416,
+      "learning_rate": 1.4814275301708222e-06,
+      "loss": 0.1068,
+      "step": 11935
+    },
+    {
+      "epoch": 1.9338950097213221,
+      "grad_norm": 0.835503101348877,
+      "learning_rate": 1.4810281835190132e-06,
+      "loss": 0.0989,
+      "step": 11936
+    },
+    {
+      "epoch": 1.9340570317563188,
+      "grad_norm": 0.831396758556366,
+      "learning_rate": 1.4806288680452747e-06,
+      "loss": 0.1001,
+      "step": 11937
+    },
+    {
+      "epoch": 1.9342190537913155,
+      "grad_norm": 0.7926315665245056,
+      "learning_rate": 1.4802295837618268e-06,
+      "loss": 0.0978,
+      "step": 11938
+    },
+    {
+      "epoch": 1.9343810758263125,
+      "grad_norm": 0.9609958529472351,
+      "learning_rate": 1.4798303306808857e-06,
+      "loss": 0.1113,
+      "step": 11939
+    },
+    {
+      "epoch": 1.9345430978613092,
+      "grad_norm": 0.8151983618736267,
+      "learning_rate": 1.479431108814668e-06,
+      "loss": 0.0924,
+      "step": 11940
+    },
+    {
+      "epoch": 1.934705119896306,
+      "grad_norm": 0.763329803943634,
+      "learning_rate": 1.4790319181753884e-06,
+      "loss": 0.0871,
+      "step": 11941
+    },
+    {
+      "epoch": 1.9348671419313026,
+      "grad_norm": 0.827495813369751,
+      "learning_rate": 1.4786327587752608e-06,
+      "loss": 0.0946,
+      "step": 11942
+    },
+    {
+      "epoch": 1.9350291639662993,
+      "grad_norm": 0.7745943665504456,
+      "learning_rate": 1.4782336306265002e-06,
+      "loss": 0.0914,
+      "step": 11943
+    },
+    {
+      "epoch": 1.9351911860012962,
+      "grad_norm": 0.8348422646522522,
+      "learning_rate": 1.4778345337413174e-06,
+      "loss": 0.1025,
+      "step": 11944
+    },
+    {
+      "epoch": 1.935353208036293,
+      "grad_norm": 0.9461818933486938,
+      "learning_rate": 1.4774354681319247e-06,
+      "loss": 0.1125,
+      "step": 11945
+    },
+    {
+      "epoch": 1.9355152300712897,
+      "grad_norm": 8.034222602844238,
+      "learning_rate": 1.4770364338105315e-06,
+      "loss": 0.0958,
+      "step": 11946
+    },
+    {
+      "epoch": 1.9356772521062866,
+      "grad_norm": 0.747870147228241,
+      "learning_rate": 1.4766374307893477e-06,
+      "loss": 0.0903,
+      "step": 11947
+    },
+    {
+      "epoch": 1.935839274141283,
+      "grad_norm": 0.8536267876625061,
+      "learning_rate": 1.4762384590805823e-06,
+      "loss": 0.1048,
+      "step": 11948
+    },
+    {
+      "epoch": 1.93600129617628,
+      "grad_norm": 0.8837407827377319,
+      "learning_rate": 1.475839518696443e-06,
+      "loss": 0.1007,
+      "step": 11949
+    },
+    {
+      "epoch": 1.9361633182112767,
+      "grad_norm": 0.8272178173065186,
+      "learning_rate": 1.475440609649136e-06,
+      "loss": 0.1017,
+      "step": 11950
+    },
+    {
+      "epoch": 1.9363253402462735,
+      "grad_norm": 0.8612387180328369,
+      "learning_rate": 1.4750417319508658e-06,
+      "loss": 0.1037,
+      "step": 11951
+    },
+    {
+      "epoch": 1.9364873622812704,
+      "grad_norm": 0.8727052211761475,
+      "learning_rate": 1.4746428856138395e-06,
+      "loss": 0.1062,
+      "step": 11952
+    },
+    {
+      "epoch": 1.9366493843162669,
+      "grad_norm": 0.7352006435394287,
+      "learning_rate": 1.4742440706502591e-06,
+      "loss": 0.0934,
+      "step": 11953
+    },
+    {
+      "epoch": 1.9368114063512638,
+      "grad_norm": 0.8387872576713562,
+      "learning_rate": 1.4738452870723286e-06,
+      "loss": 0.1012,
+      "step": 11954
+    },
+    {
+      "epoch": 1.9369734283862605,
+      "grad_norm": 0.8393522500991821,
+      "learning_rate": 1.4734465348922484e-06,
+      "loss": 0.0972,
+      "step": 11955
+    },
+    {
+      "epoch": 1.9371354504212572,
+      "grad_norm": 0.8170238733291626,
+      "learning_rate": 1.4730478141222194e-06,
+      "loss": 0.0957,
+      "step": 11956
+    },
+    {
+      "epoch": 1.9372974724562542,
+      "grad_norm": 0.908352792263031,
+      "learning_rate": 1.4726491247744429e-06,
+      "loss": 0.1151,
+      "step": 11957
+    },
+    {
+      "epoch": 1.9374594944912507,
+      "grad_norm": 0.8904739022254944,
+      "learning_rate": 1.4722504668611172e-06,
+      "loss": 0.1008,
+      "step": 11958
+    },
+    {
+      "epoch": 1.9376215165262476,
+      "grad_norm": 0.7439394593238831,
+      "learning_rate": 1.4718518403944398e-06,
+      "loss": 0.0903,
+      "step": 11959
+    },
+    {
+      "epoch": 1.9377835385612443,
+      "grad_norm": 0.8450915217399597,
+      "learning_rate": 1.4714532453866084e-06,
+      "loss": 0.0941,
+      "step": 11960
+    },
+    {
+      "epoch": 1.937945560596241,
+      "grad_norm": 0.8702685236930847,
+      "learning_rate": 1.4710546818498178e-06,
+      "loss": 0.095,
+      "step": 11961
+    },
+    {
+      "epoch": 1.938107582631238,
+      "grad_norm": 0.807630717754364,
+      "learning_rate": 1.4706561497962644e-06,
+      "loss": 0.1006,
+      "step": 11962
+    },
+    {
+      "epoch": 1.9382696046662347,
+      "grad_norm": 0.8445601463317871,
+      "learning_rate": 1.470257649238142e-06,
+      "loss": 0.0985,
+      "step": 11963
+    },
+    {
+      "epoch": 1.9384316267012314,
+      "grad_norm": 0.7522417902946472,
+      "learning_rate": 1.4698591801876435e-06,
+      "loss": 0.0966,
+      "step": 11964
+    },
+    {
+      "epoch": 1.938593648736228,
+      "grad_norm": 0.9122772812843323,
+      "learning_rate": 1.4694607426569613e-06,
+      "loss": 0.106,
+      "step": 11965
+    },
+    {
+      "epoch": 1.9387556707712248,
+      "grad_norm": 0.8337773084640503,
+      "learning_rate": 1.4690623366582856e-06,
+      "loss": 0.1084,
+      "step": 11966
+    },
+    {
+      "epoch": 1.9389176928062217,
+      "grad_norm": 0.8264670968055725,
+      "learning_rate": 1.468663962203809e-06,
+      "loss": 0.108,
+      "step": 11967
+    },
+    {
+      "epoch": 1.9390797148412184,
+      "grad_norm": 0.7818570733070374,
+      "learning_rate": 1.4682656193057189e-06,
+      "loss": 0.0925,
+      "step": 11968
+    },
+    {
+      "epoch": 1.9392417368762151,
+      "grad_norm": 0.7806037664413452,
+      "learning_rate": 1.467867307976204e-06,
+      "loss": 0.0911,
+      "step": 11969
+    },
+    {
+      "epoch": 1.939403758911212,
+      "grad_norm": 0.8423177599906921,
+      "learning_rate": 1.4674690282274517e-06,
+      "loss": 0.0982,
+      "step": 11970
+    },
+    {
+      "epoch": 1.9395657809462086,
+      "grad_norm": 0.963801920413971,
+      "learning_rate": 1.4670707800716478e-06,
+      "loss": 0.1058,
+      "step": 11971
+    },
+    {
+      "epoch": 1.9397278029812055,
+      "grad_norm": 0.8713017106056213,
+      "learning_rate": 1.4666725635209794e-06,
+      "loss": 0.1015,
+      "step": 11972
+    },
+    {
+      "epoch": 1.9398898250162022,
+      "grad_norm": 0.7701468467712402,
+      "learning_rate": 1.4662743785876298e-06,
+      "loss": 0.0996,
+      "step": 11973
+    },
+    {
+      "epoch": 1.940051847051199,
+      "grad_norm": 0.7816067934036255,
+      "learning_rate": 1.4658762252837821e-06,
+      "loss": 0.0938,
+      "step": 11974
+    },
+    {
+      "epoch": 1.9402138690861959,
+      "grad_norm": 1.0052039623260498,
+      "learning_rate": 1.4654781036216193e-06,
+      "loss": 0.1145,
+      "step": 11975
+    },
+    {
+      "epoch": 1.9403758911211924,
+      "grad_norm": 0.8187944889068604,
+      "learning_rate": 1.4650800136133238e-06,
+      "loss": 0.094,
+      "step": 11976
+    },
+    {
+      "epoch": 1.9405379131561893,
+      "grad_norm": 0.7152525186538696,
+      "learning_rate": 1.4646819552710751e-06,
+      "loss": 0.0915,
+      "step": 11977
+    },
+    {
+      "epoch": 1.940699935191186,
+      "grad_norm": 0.8357136845588684,
+      "learning_rate": 1.4642839286070537e-06,
+      "loss": 0.0951,
+      "step": 11978
+    },
+    {
+      "epoch": 1.9408619572261827,
+      "grad_norm": 0.8096917271614075,
+      "learning_rate": 1.463885933633437e-06,
+      "loss": 0.1029,
+      "step": 11979
+    },
+    {
+      "epoch": 1.9410239792611796,
+      "grad_norm": 0.8772410154342651,
+      "learning_rate": 1.4634879703624027e-06,
+      "loss": 0.108,
+      "step": 11980
+    },
+    {
+      "epoch": 1.9411860012961761,
+      "grad_norm": 0.7658387422561646,
+      "learning_rate": 1.4630900388061292e-06,
+      "loss": 0.0968,
+      "step": 11981
+    },
+    {
+      "epoch": 1.941348023331173,
+      "grad_norm": 0.7496578693389893,
+      "learning_rate": 1.4626921389767915e-06,
+      "loss": 0.0883,
+      "step": 11982
+    },
+    {
+      "epoch": 1.9415100453661698,
+      "grad_norm": 0.852181613445282,
+      "learning_rate": 1.4622942708865635e-06,
+      "loss": 0.0962,
+      "step": 11983
+    },
+    {
+      "epoch": 1.9416720674011665,
+      "grad_norm": 0.8182169198989868,
+      "learning_rate": 1.4618964345476203e-06,
+      "loss": 0.1018,
+      "step": 11984
+    },
+    {
+      "epoch": 1.9418340894361634,
+      "grad_norm": 0.781848669052124,
+      "learning_rate": 1.4614986299721328e-06,
+      "loss": 0.0939,
+      "step": 11985
+    },
+    {
+      "epoch": 1.94199611147116,
+      "grad_norm": 0.7615278363227844,
+      "learning_rate": 1.4611008571722748e-06,
+      "loss": 0.0952,
+      "step": 11986
+    },
+    {
+      "epoch": 1.9421581335061568,
+      "grad_norm": 0.8403747081756592,
+      "learning_rate": 1.460703116160217e-06,
+      "loss": 0.1036,
+      "step": 11987
+    },
+    {
+      "epoch": 1.9423201555411536,
+      "grad_norm": 0.7197666168212891,
+      "learning_rate": 1.4603054069481282e-06,
+      "loss": 0.0873,
+      "step": 11988
+    },
+    {
+      "epoch": 1.9424821775761503,
+      "grad_norm": 0.7320033311843872,
+      "learning_rate": 1.4599077295481783e-06,
+      "loss": 0.0831,
+      "step": 11989
+    },
+    {
+      "epoch": 1.9426441996111472,
+      "grad_norm": 0.7511314749717712,
+      "learning_rate": 1.4595100839725338e-06,
+      "loss": 0.0869,
+      "step": 11990
+    },
+    {
+      "epoch": 1.942806221646144,
+      "grad_norm": 0.809007465839386,
+      "learning_rate": 1.4591124702333636e-06,
+      "loss": 0.0954,
+      "step": 11991
+    },
+    {
+      "epoch": 1.9429682436811406,
+      "grad_norm": 0.8283510804176331,
+      "learning_rate": 1.4587148883428337e-06,
+      "loss": 0.1011,
+      "step": 11992
+    },
+    {
+      "epoch": 1.9431302657161373,
+      "grad_norm": 0.7291625738143921,
+      "learning_rate": 1.4583173383131077e-06,
+      "loss": 0.0918,
+      "step": 11993
+    },
+    {
+      "epoch": 1.943292287751134,
+      "grad_norm": 0.7869513034820557,
+      "learning_rate": 1.45791982015635e-06,
+      "loss": 0.0993,
+      "step": 11994
+    },
+    {
+      "epoch": 1.943454309786131,
+      "grad_norm": 0.7561447620391846,
+      "learning_rate": 1.457522333884724e-06,
+      "loss": 0.0897,
+      "step": 11995
+    },
+    {
+      "epoch": 1.9436163318211277,
+      "grad_norm": 0.8446448445320129,
+      "learning_rate": 1.4571248795103921e-06,
+      "loss": 0.1032,
+      "step": 11996
+    },
+    {
+      "epoch": 1.9437783538561244,
+      "grad_norm": 0.8000361919403076,
+      "learning_rate": 1.456727457045515e-06,
+      "loss": 0.096,
+      "step": 11997
+    },
+    {
+      "epoch": 1.9439403758911213,
+      "grad_norm": 0.7143806219100952,
+      "learning_rate": 1.4563300665022534e-06,
+      "loss": 0.0851,
+      "step": 11998
+    },
+    {
+      "epoch": 1.9441023979261178,
+      "grad_norm": 0.7861617207527161,
+      "learning_rate": 1.4559327078927656e-06,
+      "loss": 0.098,
+      "step": 11999
+    },
+    {
+      "epoch": 1.9442644199611148,
+      "grad_norm": 0.8464603424072266,
+      "learning_rate": 1.4555353812292105e-06,
+      "loss": 0.0929,
+      "step": 12000
+    },
+    {
+      "epoch": 1.9444264419961115,
+      "grad_norm": 0.817179262638092,
+      "learning_rate": 1.4551380865237456e-06,
+      "loss": 0.09,
+      "step": 12001
+    },
+    {
+      "epoch": 1.9445884640311082,
+      "grad_norm": 0.9822414517402649,
+      "learning_rate": 1.4547408237885262e-06,
+      "loss": 0.1172,
+      "step": 12002
+    },
+    {
+      "epoch": 1.9447504860661051,
+      "grad_norm": 0.6776533722877502,
+      "learning_rate": 1.454343593035709e-06,
+      "loss": 0.078,
+      "step": 12003
+    },
+    {
+      "epoch": 1.9449125081011016,
+      "grad_norm": 0.7004638314247131,
+      "learning_rate": 1.4539463942774462e-06,
+      "loss": 0.0857,
+      "step": 12004
+    },
+    {
+      "epoch": 1.9450745301360985,
+      "grad_norm": 0.8583202362060547,
+      "learning_rate": 1.4535492275258928e-06,
+      "loss": 0.0998,
+      "step": 12005
+    },
+    {
+      "epoch": 1.9452365521710953,
+      "grad_norm": 0.9123525619506836,
+      "learning_rate": 1.4531520927932017e-06,
+      "loss": 0.107,
+      "step": 12006
+    },
+    {
+      "epoch": 1.945398574206092,
+      "grad_norm": 0.8352869153022766,
+      "learning_rate": 1.4527549900915222e-06,
+      "loss": 0.1025,
+      "step": 12007
+    },
+    {
+      "epoch": 1.945560596241089,
+      "grad_norm": 0.7704386115074158,
+      "learning_rate": 1.452357919433006e-06,
+      "loss": 0.0921,
+      "step": 12008
+    },
+    {
+      "epoch": 1.9457226182760854,
+      "grad_norm": 0.7887938618659973,
+      "learning_rate": 1.4519608808298007e-06,
+      "loss": 0.1008,
+      "step": 12009
+    },
+    {
+      "epoch": 1.9458846403110823,
+      "grad_norm": 0.7636969685554504,
+      "learning_rate": 1.4515638742940585e-06,
+      "loss": 0.0949,
+      "step": 12010
+    },
+    {
+      "epoch": 1.946046662346079,
+      "grad_norm": 0.7282156348228455,
+      "learning_rate": 1.4511668998379238e-06,
+      "loss": 0.082,
+      "step": 12011
+    },
+    {
+      "epoch": 1.9462086843810757,
+      "grad_norm": 0.8383048176765442,
+      "learning_rate": 1.4507699574735436e-06,
+      "loss": 0.1019,
+      "step": 12012
+    },
+    {
+      "epoch": 1.9463707064160727,
+      "grad_norm": 0.8490771055221558,
+      "learning_rate": 1.450373047213064e-06,
+      "loss": 0.0947,
+      "step": 12013
+    },
+    {
+      "epoch": 1.9465327284510694,
+      "grad_norm": 0.7578926086425781,
+      "learning_rate": 1.4499761690686287e-06,
+      "loss": 0.0871,
+      "step": 12014
+    },
+    {
+      "epoch": 1.946694750486066,
+      "grad_norm": 0.847022533416748,
+      "learning_rate": 1.4495793230523817e-06,
+      "loss": 0.1013,
+      "step": 12015
+    },
+    {
+      "epoch": 1.9468567725210628,
+      "grad_norm": 0.7949681282043457,
+      "learning_rate": 1.4491825091764656e-06,
+      "loss": 0.094,
+      "step": 12016
+    },
+    {
+      "epoch": 1.9470187945560595,
+      "grad_norm": 0.8004774451255798,
+      "learning_rate": 1.4487857274530214e-06,
+      "loss": 0.0959,
+      "step": 12017
+    },
+    {
+      "epoch": 1.9471808165910565,
+      "grad_norm": 0.8467872142791748,
+      "learning_rate": 1.4483889778941904e-06,
+      "loss": 0.1049,
+      "step": 12018
+    },
+    {
+      "epoch": 1.9473428386260532,
+      "grad_norm": 0.767632246017456,
+      "learning_rate": 1.4479922605121117e-06,
+      "loss": 0.0965,
+      "step": 12019
+    },
+    {
+      "epoch": 1.9475048606610499,
+      "grad_norm": 0.7307461500167847,
+      "learning_rate": 1.447595575318924e-06,
+      "loss": 0.0858,
+      "step": 12020
+    },
+    {
+      "epoch": 1.9476668826960468,
+      "grad_norm": 0.7664514780044556,
+      "learning_rate": 1.447198922326766e-06,
+      "loss": 0.1005,
+      "step": 12021
+    },
+    {
+      "epoch": 1.9478289047310433,
+      "grad_norm": 0.9641528129577637,
+      "learning_rate": 1.4468023015477722e-06,
+      "loss": 0.1177,
+      "step": 12022
+    },
+    {
+      "epoch": 1.9479909267660402,
+      "grad_norm": 0.8184154629707336,
+      "learning_rate": 1.4464057129940783e-06,
+      "loss": 0.0997,
+      "step": 12023
+    },
+    {
+      "epoch": 1.948152948801037,
+      "grad_norm": 0.7754473090171814,
+      "learning_rate": 1.446009156677822e-06,
+      "loss": 0.0928,
+      "step": 12024
+    },
+    {
+      "epoch": 1.9483149708360337,
+      "grad_norm": 0.8045803308486938,
+      "learning_rate": 1.4456126326111337e-06,
+      "loss": 0.0991,
+      "step": 12025
+    },
+    {
+      "epoch": 1.9484769928710306,
+      "grad_norm": 0.7364550828933716,
+      "learning_rate": 1.4452161408061478e-06,
+      "loss": 0.0943,
+      "step": 12026
+    },
+    {
+      "epoch": 1.948639014906027,
+      "grad_norm": 0.8381508588790894,
+      "learning_rate": 1.4448196812749948e-06,
+      "loss": 0.1039,
+      "step": 12027
+    },
+    {
+      "epoch": 1.948801036941024,
+      "grad_norm": 0.7855244278907776,
+      "learning_rate": 1.4444232540298064e-06,
+      "loss": 0.106,
+      "step": 12028
+    },
+    {
+      "epoch": 1.9489630589760207,
+      "grad_norm": 0.8092867136001587,
+      "learning_rate": 1.4440268590827117e-06,
+      "loss": 0.1001,
+      "step": 12029
+    },
+    {
+      "epoch": 1.9491250810110174,
+      "grad_norm": 0.8405110239982605,
+      "learning_rate": 1.44363049644584e-06,
+      "loss": 0.1049,
+      "step": 12030
+    },
+    {
+      "epoch": 1.9492871030460144,
+      "grad_norm": 0.821543276309967,
+      "learning_rate": 1.4432341661313188e-06,
+      "loss": 0.1045,
+      "step": 12031
+    },
+    {
+      "epoch": 1.9494491250810109,
+      "grad_norm": 0.7550950646400452,
+      "learning_rate": 1.4428378681512755e-06,
+      "loss": 0.0893,
+      "step": 12032
+    },
+    {
+      "epoch": 1.9496111471160078,
+      "grad_norm": 0.733284056186676,
+      "learning_rate": 1.4424416025178335e-06,
+      "loss": 0.0866,
+      "step": 12033
+    },
+    {
+      "epoch": 1.9497731691510045,
+      "grad_norm": 0.8085827231407166,
+      "learning_rate": 1.4420453692431197e-06,
+      "loss": 0.0921,
+      "step": 12034
+    },
+    {
+      "epoch": 1.9499351911860012,
+      "grad_norm": 0.7645505666732788,
+      "learning_rate": 1.441649168339258e-06,
+      "loss": 0.0936,
+      "step": 12035
+    },
+    {
+      "epoch": 1.9500972132209982,
+      "grad_norm": 0.781045138835907,
+      "learning_rate": 1.441252999818371e-06,
+      "loss": 0.0983,
+      "step": 12036
+    },
+    {
+      "epoch": 1.9502592352559946,
+      "grad_norm": 0.7690817713737488,
+      "learning_rate": 1.4408568636925796e-06,
+      "loss": 0.0885,
+      "step": 12037
+    },
+    {
+      "epoch": 1.9504212572909916,
+      "grad_norm": 0.8504323363304138,
+      "learning_rate": 1.440460759974004e-06,
+      "loss": 0.099,
+      "step": 12038
+    },
+    {
+      "epoch": 1.9505832793259883,
+      "grad_norm": 0.788281261920929,
+      "learning_rate": 1.4400646886747672e-06,
+      "loss": 0.0929,
+      "step": 12039
+    },
+    {
+      "epoch": 1.950745301360985,
+      "grad_norm": 0.7282304167747498,
+      "learning_rate": 1.4396686498069844e-06,
+      "loss": 0.0861,
+      "step": 12040
+    },
+    {
+      "epoch": 1.950907323395982,
+      "grad_norm": 0.8458184003829956,
+      "learning_rate": 1.4392726433827754e-06,
+      "loss": 0.1047,
+      "step": 12041
+    },
+    {
+      "epoch": 1.9510693454309787,
+      "grad_norm": 0.9088663458824158,
+      "learning_rate": 1.4388766694142553e-06,
+      "loss": 0.1065,
+      "step": 12042
+    },
+    {
+      "epoch": 1.9512313674659754,
+      "grad_norm": 0.827163815498352,
+      "learning_rate": 1.4384807279135438e-06,
+      "loss": 0.095,
+      "step": 12043
+    },
+    {
+      "epoch": 1.9513933895009723,
+      "grad_norm": 0.7869982123374939,
+      "learning_rate": 1.4380848188927516e-06,
+      "loss": 0.0893,
+      "step": 12044
+    },
+    {
+      "epoch": 1.9515554115359688,
+      "grad_norm": 0.940324604511261,
+      "learning_rate": 1.4376889423639945e-06,
+      "loss": 0.1045,
+      "step": 12045
+    },
+    {
+      "epoch": 1.9517174335709657,
+      "grad_norm": 0.8107567429542542,
+      "learning_rate": 1.4372930983393849e-06,
+      "loss": 0.0969,
+      "step": 12046
+    },
+    {
+      "epoch": 1.9518794556059624,
+      "grad_norm": 0.8711578845977783,
+      "learning_rate": 1.4368972868310349e-06,
+      "loss": 0.0977,
+      "step": 12047
+    },
+    {
+      "epoch": 1.9520414776409591,
+      "grad_norm": 0.7904216647148132,
+      "learning_rate": 1.4365015078510553e-06,
+      "loss": 0.102,
+      "step": 12048
+    },
+    {
+      "epoch": 1.952203499675956,
+      "grad_norm": 0.7687509655952454,
+      "learning_rate": 1.4361057614115557e-06,
+      "loss": 0.0939,
+      "step": 12049
+    },
+    {
+      "epoch": 1.9523655217109526,
+      "grad_norm": 0.7089811563491821,
+      "learning_rate": 1.4357100475246463e-06,
+      "loss": 0.0847,
+      "step": 12050
+    },
+    {
+      "epoch": 1.9525275437459495,
+      "grad_norm": 0.8005346059799194,
+      "learning_rate": 1.435314366202433e-06,
+      "loss": 0.0934,
+      "step": 12051
+    },
+    {
+      "epoch": 1.9526895657809462,
+      "grad_norm": 0.8521600365638733,
+      "learning_rate": 1.4349187174570226e-06,
+      "loss": 0.0972,
+      "step": 12052
+    },
+    {
+      "epoch": 1.952851587815943,
+      "grad_norm": 0.7018685936927795,
+      "learning_rate": 1.4345231013005229e-06,
+      "loss": 0.0859,
+      "step": 12053
+    },
+    {
+      "epoch": 1.9530136098509399,
+      "grad_norm": 0.8380179405212402,
+      "learning_rate": 1.4341275177450389e-06,
+      "loss": 0.0971,
+      "step": 12054
+    },
+    {
+      "epoch": 1.9531756318859363,
+      "grad_norm": 0.8714627623558044,
+      "learning_rate": 1.4337319668026726e-06,
+      "loss": 0.1073,
+      "step": 12055
+    },
+    {
+      "epoch": 1.9533376539209333,
+      "grad_norm": 0.7577527165412903,
+      "learning_rate": 1.4333364484855277e-06,
+      "loss": 0.0881,
+      "step": 12056
+    },
+    {
+      "epoch": 1.95349967595593,
+      "grad_norm": 0.8456433415412903,
+      "learning_rate": 1.4329409628057062e-06,
+      "loss": 0.0975,
+      "step": 12057
+    },
+    {
+      "epoch": 1.9536616979909267,
+      "grad_norm": 0.7880712151527405,
+      "learning_rate": 1.432545509775309e-06,
+      "loss": 0.0929,
+      "step": 12058
+    },
+    {
+      "epoch": 1.9538237200259236,
+      "grad_norm": 0.7699628472328186,
+      "learning_rate": 1.432150089406436e-06,
+      "loss": 0.0881,
+      "step": 12059
+    },
+    {
+      "epoch": 1.9539857420609201,
+      "grad_norm": 0.7075265049934387,
+      "learning_rate": 1.4317547017111865e-06,
+      "loss": 0.0902,
+      "step": 12060
+    },
+    {
+      "epoch": 1.954147764095917,
+      "grad_norm": 0.8275308609008789,
+      "learning_rate": 1.4313593467016576e-06,
+      "loss": 0.1006,
+      "step": 12061
+    },
+    {
+      "epoch": 1.9543097861309138,
+      "grad_norm": 0.876686155796051,
+      "learning_rate": 1.4309640243899467e-06,
+      "loss": 0.0997,
+      "step": 12062
+    },
+    {
+      "epoch": 1.9544718081659105,
+      "grad_norm": 0.9776948690414429,
+      "learning_rate": 1.4305687347881497e-06,
+      "loss": 0.102,
+      "step": 12063
+    },
+    {
+      "epoch": 1.9546338302009074,
+      "grad_norm": 0.8320091962814331,
+      "learning_rate": 1.4301734779083614e-06,
+      "loss": 0.0979,
+      "step": 12064
+    },
+    {
+      "epoch": 1.9547958522359041,
+      "grad_norm": 0.9126972556114197,
+      "learning_rate": 1.429778253762677e-06,
+      "loss": 0.1078,
+      "step": 12065
+    },
+    {
+      "epoch": 1.9549578742709008,
+      "grad_norm": 0.7437605857849121,
+      "learning_rate": 1.4293830623631857e-06,
+      "loss": 0.0844,
+      "step": 12066
+    },
+    {
+      "epoch": 1.9551198963058976,
+      "grad_norm": 0.8160969018936157,
+      "learning_rate": 1.4289879037219832e-06,
+      "loss": 0.0981,
+      "step": 12067
+    },
+    {
+      "epoch": 1.9552819183408943,
+      "grad_norm": 0.7898657917976379,
+      "learning_rate": 1.4285927778511598e-06,
+      "loss": 0.0951,
+      "step": 12068
+    },
+    {
+      "epoch": 1.9554439403758912,
+      "grad_norm": 0.87503582239151,
+      "learning_rate": 1.4281976847628038e-06,
+      "loss": 0.1059,
+      "step": 12069
+    },
+    {
+      "epoch": 1.955605962410888,
+      "grad_norm": 0.8932657241821289,
+      "learning_rate": 1.4278026244690046e-06,
+      "loss": 0.1041,
+      "step": 12070
+    },
+    {
+      "epoch": 1.9557679844458846,
+      "grad_norm": 0.878448486328125,
+      "learning_rate": 1.4274075969818498e-06,
+      "loss": 0.092,
+      "step": 12071
+    },
+    {
+      "epoch": 1.9559300064808816,
+      "grad_norm": 0.7802169322967529,
+      "learning_rate": 1.427012602313429e-06,
+      "loss": 0.0938,
+      "step": 12072
+    },
+    {
+      "epoch": 1.956092028515878,
+      "grad_norm": 0.9439387321472168,
+      "learning_rate": 1.4266176404758246e-06,
+      "loss": 0.1151,
+      "step": 12073
+    },
+    {
+      "epoch": 1.956254050550875,
+      "grad_norm": 0.7736334800720215,
+      "learning_rate": 1.4262227114811233e-06,
+      "loss": 0.0914,
+      "step": 12074
+    },
+    {
+      "epoch": 1.9564160725858717,
+      "grad_norm": 0.8040494322776794,
+      "learning_rate": 1.4258278153414082e-06,
+      "loss": 0.0988,
+      "step": 12075
+    },
+    {
+      "epoch": 1.9565780946208684,
+      "grad_norm": 0.8778871297836304,
+      "learning_rate": 1.4254329520687626e-06,
+      "loss": 0.1024,
+      "step": 12076
+    },
+    {
+      "epoch": 1.9567401166558653,
+      "grad_norm": 0.7685941457748413,
+      "learning_rate": 1.4250381216752685e-06,
+      "loss": 0.0928,
+      "step": 12077
+    },
+    {
+      "epoch": 1.9569021386908618,
+      "grad_norm": 0.8489927053451538,
+      "learning_rate": 1.4246433241730062e-06,
+      "loss": 0.1049,
+      "step": 12078
+    },
+    {
+      "epoch": 1.9570641607258588,
+      "grad_norm": 0.7054093480110168,
+      "learning_rate": 1.4242485595740557e-06,
+      "loss": 0.0874,
+      "step": 12079
+    },
+    {
+      "epoch": 1.9572261827608555,
+      "grad_norm": 0.8379080295562744,
+      "learning_rate": 1.4238538278904973e-06,
+      "loss": 0.1119,
+      "step": 12080
+    },
+    {
+      "epoch": 1.9573882047958522,
+      "grad_norm": 0.7417575716972351,
+      "learning_rate": 1.4234591291344058e-06,
+      "loss": 0.0906,
+      "step": 12081
+    },
+    {
+      "epoch": 1.9575502268308491,
+      "grad_norm": 0.7702432870864868,
+      "learning_rate": 1.4230644633178603e-06,
+      "loss": 0.0917,
+      "step": 12082
+    },
+    {
+      "epoch": 1.9577122488658456,
+      "grad_norm": 0.7377634644508362,
+      "learning_rate": 1.4226698304529373e-06,
+      "loss": 0.0904,
+      "step": 12083
+    },
+    {
+      "epoch": 1.9578742709008425,
+      "grad_norm": 0.7292881608009338,
+      "learning_rate": 1.4222752305517093e-06,
+      "loss": 0.0947,
+      "step": 12084
+    },
+    {
+      "epoch": 1.9580362929358393,
+      "grad_norm": 0.8291403651237488,
+      "learning_rate": 1.4218806636262504e-06,
+      "loss": 0.0972,
+      "step": 12085
+    },
+    {
+      "epoch": 1.958198314970836,
+      "grad_norm": 0.8070185780525208,
+      "learning_rate": 1.421486129688635e-06,
+      "loss": 0.0977,
+      "step": 12086
+    },
+    {
+      "epoch": 1.958360337005833,
+      "grad_norm": 0.8218309283256531,
+      "learning_rate": 1.421091628750935e-06,
+      "loss": 0.1092,
+      "step": 12087
+    },
+    {
+      "epoch": 1.9585223590408296,
+      "grad_norm": 0.911953330039978,
+      "learning_rate": 1.4206971608252196e-06,
+      "loss": 0.1041,
+      "step": 12088
+    },
+    {
+      "epoch": 1.9586843810758263,
+      "grad_norm": 0.8383828997612,
+      "learning_rate": 1.4203027259235592e-06,
+      "loss": 0.0998,
+      "step": 12089
+    },
+    {
+      "epoch": 1.958846403110823,
+      "grad_norm": 0.7073849439620972,
+      "learning_rate": 1.4199083240580218e-06,
+      "loss": 0.0869,
+      "step": 12090
+    },
+    {
+      "epoch": 1.9590084251458197,
+      "grad_norm": 0.8118810653686523,
+      "learning_rate": 1.4195139552406766e-06,
+      "loss": 0.095,
+      "step": 12091
+    },
+    {
+      "epoch": 1.9591704471808167,
+      "grad_norm": 0.8553704619407654,
+      "learning_rate": 1.41911961948359e-06,
+      "loss": 0.1069,
+      "step": 12092
+    },
+    {
+      "epoch": 1.9593324692158134,
+      "grad_norm": 0.8816598653793335,
+      "learning_rate": 1.4187253167988266e-06,
+      "loss": 0.1,
+      "step": 12093
+    },
+    {
+      "epoch": 1.95949449125081,
+      "grad_norm": 0.7908227443695068,
+      "learning_rate": 1.4183310471984532e-06,
+      "loss": 0.0944,
+      "step": 12094
+    },
+    {
+      "epoch": 1.959656513285807,
+      "grad_norm": 0.8013996481895447,
+      "learning_rate": 1.41793681069453e-06,
+      "loss": 0.1013,
+      "step": 12095
+    },
+    {
+      "epoch": 1.9598185353208035,
+      "grad_norm": 0.8322909474372864,
+      "learning_rate": 1.4175426072991234e-06,
+      "loss": 0.0962,
+      "step": 12096
+    },
+    {
+      "epoch": 1.9599805573558005,
+      "grad_norm": 0.8282430768013,
+      "learning_rate": 1.4171484370242927e-06,
+      "loss": 0.1035,
+      "step": 12097
+    },
+    {
+      "epoch": 1.9601425793907972,
+      "grad_norm": 0.8286914229393005,
+      "learning_rate": 1.416754299882101e-06,
+      "loss": 0.0942,
+      "step": 12098
+    },
+    {
+      "epoch": 1.9603046014257939,
+      "grad_norm": 0.7469590902328491,
+      "learning_rate": 1.4163601958846052e-06,
+      "loss": 0.0882,
+      "step": 12099
+    },
+    {
+      "epoch": 1.9604666234607908,
+      "grad_norm": 0.8161764144897461,
+      "learning_rate": 1.415966125043864e-06,
+      "loss": 0.1035,
+      "step": 12100
+    },
+    {
+      "epoch": 1.9606286454957873,
+      "grad_norm": 0.8531409502029419,
+      "learning_rate": 1.4155720873719378e-06,
+      "loss": 0.0952,
+      "step": 12101
+    },
+    {
+      "epoch": 1.9607906675307842,
+      "grad_norm": 0.8322476148605347,
+      "learning_rate": 1.415178082880881e-06,
+      "loss": 0.0941,
+      "step": 12102
+    },
+    {
+      "epoch": 1.960952689565781,
+      "grad_norm": 0.7497605681419373,
+      "learning_rate": 1.414784111582749e-06,
+      "loss": 0.087,
+      "step": 12103
+    },
+    {
+      "epoch": 1.9611147116007777,
+      "grad_norm": 0.7348429560661316,
+      "learning_rate": 1.4143901734895973e-06,
+      "loss": 0.0873,
+      "step": 12104
+    },
+    {
+      "epoch": 1.9612767336357746,
+      "grad_norm": 0.8899442553520203,
+      "learning_rate": 1.4139962686134792e-06,
+      "loss": 0.1037,
+      "step": 12105
+    },
+    {
+      "epoch": 1.961438755670771,
+      "grad_norm": 0.7474508881568909,
+      "learning_rate": 1.4136023969664471e-06,
+      "loss": 0.0878,
+      "step": 12106
+    },
+    {
+      "epoch": 1.961600777705768,
+      "grad_norm": 0.8204247951507568,
+      "learning_rate": 1.4132085585605528e-06,
+      "loss": 0.1014,
+      "step": 12107
+    },
+    {
+      "epoch": 1.9617627997407647,
+      "grad_norm": 0.906720757484436,
+      "learning_rate": 1.4128147534078469e-06,
+      "loss": 0.0965,
+      "step": 12108
+    },
+    {
+      "epoch": 1.9619248217757614,
+      "grad_norm": 0.8094850778579712,
+      "learning_rate": 1.4124209815203779e-06,
+      "loss": 0.0905,
+      "step": 12109
+    },
+    {
+      "epoch": 1.9620868438107584,
+      "grad_norm": 0.8408055901527405,
+      "learning_rate": 1.4120272429101955e-06,
+      "loss": 0.1028,
+      "step": 12110
+    },
+    {
+      "epoch": 1.9622488658457549,
+      "grad_norm": 0.852473258972168,
+      "learning_rate": 1.4116335375893464e-06,
+      "loss": 0.0992,
+      "step": 12111
+    },
+    {
+      "epoch": 1.9624108878807518,
+      "grad_norm": 0.7675461173057556,
+      "learning_rate": 1.4112398655698772e-06,
+      "loss": 0.0989,
+      "step": 12112
+    },
+    {
+      "epoch": 1.9625729099157485,
+      "grad_norm": 0.7929125428199768,
+      "learning_rate": 1.4108462268638346e-06,
+      "loss": 0.0971,
+      "step": 12113
+    },
+    {
+      "epoch": 1.9627349319507452,
+      "grad_norm": 0.7752059102058411,
+      "learning_rate": 1.4104526214832595e-06,
+      "loss": 0.0955,
+      "step": 12114
+    },
+    {
+      "epoch": 1.9628969539857422,
+      "grad_norm": 0.7863343358039856,
+      "learning_rate": 1.4100590494401988e-06,
+      "loss": 0.0925,
+      "step": 12115
+    },
+    {
+      "epoch": 1.9630589760207389,
+      "grad_norm": 0.8089168071746826,
+      "learning_rate": 1.4096655107466943e-06,
+      "loss": 0.0934,
+      "step": 12116
+    },
+    {
+      "epoch": 1.9632209980557356,
+      "grad_norm": 0.8375380635261536,
+      "learning_rate": 1.4092720054147857e-06,
+      "loss": 0.0978,
+      "step": 12117
+    },
+    {
+      "epoch": 1.9633830200907323,
+      "grad_norm": 0.7996324896812439,
+      "learning_rate": 1.4088785334565145e-06,
+      "loss": 0.0971,
+      "step": 12118
+    },
+    {
+      "epoch": 1.963545042125729,
+      "grad_norm": 0.7718796730041504,
+      "learning_rate": 1.4084850948839194e-06,
+      "loss": 0.1003,
+      "step": 12119
+    },
+    {
+      "epoch": 1.963707064160726,
+      "grad_norm": 0.8828782439231873,
+      "learning_rate": 1.4080916897090391e-06,
+      "loss": 0.1085,
+      "step": 12120
+    },
+    {
+      "epoch": 1.9638690861957226,
+      "grad_norm": 0.8628697395324707,
+      "learning_rate": 1.4076983179439107e-06,
+      "loss": 0.1042,
+      "step": 12121
+    },
+    {
+      "epoch": 1.9640311082307194,
+      "grad_norm": 0.802811861038208,
+      "learning_rate": 1.4073049796005705e-06,
+      "loss": 0.1005,
+      "step": 12122
+    },
+    {
+      "epoch": 1.9641931302657163,
+      "grad_norm": 0.7491890788078308,
+      "learning_rate": 1.4069116746910536e-06,
+      "loss": 0.0984,
+      "step": 12123
+    },
+    {
+      "epoch": 1.9643551523007128,
+      "grad_norm": 0.9189904928207397,
+      "learning_rate": 1.4065184032273942e-06,
+      "loss": 0.1173,
+      "step": 12124
+    },
+    {
+      "epoch": 1.9645171743357097,
+      "grad_norm": 0.6986983418464661,
+      "learning_rate": 1.4061251652216254e-06,
+      "loss": 0.0886,
+      "step": 12125
+    },
+    {
+      "epoch": 1.9646791963707064,
+      "grad_norm": 0.9891625642776489,
+      "learning_rate": 1.4057319606857795e-06,
+      "loss": 0.1135,
+      "step": 12126
+    },
+    {
+      "epoch": 1.9648412184057031,
+      "grad_norm": 0.8146089911460876,
+      "learning_rate": 1.4053387896318888e-06,
+      "loss": 0.1027,
+      "step": 12127
+    },
+    {
+      "epoch": 1.9650032404407,
+      "grad_norm": 0.8153820633888245,
+      "learning_rate": 1.4049456520719805e-06,
+      "loss": 0.0987,
+      "step": 12128
+    },
+    {
+      "epoch": 1.9651652624756966,
+      "grad_norm": 0.6867256164550781,
+      "learning_rate": 1.4045525480180849e-06,
+      "loss": 0.0889,
+      "step": 12129
+    },
+    {
+      "epoch": 1.9653272845106935,
+      "grad_norm": 0.8280233144760132,
+      "learning_rate": 1.404159477482231e-06,
+      "loss": 0.1033,
+      "step": 12130
+    },
+    {
+      "epoch": 1.9654893065456902,
+      "grad_norm": 0.7022069096565247,
+      "learning_rate": 1.4037664404764465e-06,
+      "loss": 0.083,
+      "step": 12131
+    },
+    {
+      "epoch": 1.965651328580687,
+      "grad_norm": 0.8097946643829346,
+      "learning_rate": 1.403373437012755e-06,
+      "loss": 0.0949,
+      "step": 12132
+    },
+    {
+      "epoch": 1.9658133506156839,
+      "grad_norm": 0.7029479742050171,
+      "learning_rate": 1.4029804671031812e-06,
+      "loss": 0.0877,
+      "step": 12133
+    },
+    {
+      "epoch": 1.9659753726506803,
+      "grad_norm": 0.7599433064460754,
+      "learning_rate": 1.4025875307597528e-06,
+      "loss": 0.0892,
+      "step": 12134
+    },
+    {
+      "epoch": 1.9661373946856773,
+      "grad_norm": 0.754822313785553,
+      "learning_rate": 1.4021946279944893e-06,
+      "loss": 0.0842,
+      "step": 12135
+    },
+    {
+      "epoch": 1.966299416720674,
+      "grad_norm": 0.8841560482978821,
+      "learning_rate": 1.4018017588194132e-06,
+      "loss": 0.1086,
+      "step": 12136
+    },
+    {
+      "epoch": 1.9664614387556707,
+      "grad_norm": 0.7273808121681213,
+      "learning_rate": 1.4014089232465458e-06,
+      "loss": 0.0901,
+      "step": 12137
+    },
+    {
+      "epoch": 1.9666234607906676,
+      "grad_norm": 0.7973136305809021,
+      "learning_rate": 1.401016121287907e-06,
+      "loss": 0.085,
+      "step": 12138
+    },
+    {
+      "epoch": 1.9667854828256643,
+      "grad_norm": 0.7728840708732605,
+      "learning_rate": 1.4006233529555152e-06,
+      "loss": 0.0985,
+      "step": 12139
+    },
+    {
+      "epoch": 1.966947504860661,
+      "grad_norm": 0.807805597782135,
+      "learning_rate": 1.4002306182613885e-06,
+      "loss": 0.1028,
+      "step": 12140
+    },
+    {
+      "epoch": 1.9671095268956578,
+      "grad_norm": 0.8040981888771057,
+      "learning_rate": 1.399837917217543e-06,
+      "loss": 0.105,
+      "step": 12141
+    },
+    {
+      "epoch": 1.9672715489306545,
+      "grad_norm": 0.8386695384979248,
+      "learning_rate": 1.3994452498359963e-06,
+      "loss": 0.1036,
+      "step": 12142
+    },
+    {
+      "epoch": 1.9674335709656514,
+      "grad_norm": 0.7889523506164551,
+      "learning_rate": 1.39905261612876e-06,
+      "loss": 0.091,
+      "step": 12143
+    },
+    {
+      "epoch": 1.9675955930006481,
+      "grad_norm": 0.792294979095459,
+      "learning_rate": 1.39866001610785e-06,
+      "loss": 0.0958,
+      "step": 12144
+    },
+    {
+      "epoch": 1.9677576150356448,
+      "grad_norm": 0.7585206627845764,
+      "learning_rate": 1.3982674497852794e-06,
+      "loss": 0.0992,
+      "step": 12145
+    },
+    {
+      "epoch": 1.9679196370706418,
+      "grad_norm": 0.8587349057197571,
+      "learning_rate": 1.3978749171730577e-06,
+      "loss": 0.1085,
+      "step": 12146
+    },
+    {
+      "epoch": 1.9680816591056383,
+      "grad_norm": 0.742311418056488,
+      "learning_rate": 1.3974824182831965e-06,
+      "loss": 0.0892,
+      "step": 12147
+    },
+    {
+      "epoch": 1.9682436811406352,
+      "grad_norm": 0.805219829082489,
+      "learning_rate": 1.397089953127704e-06,
+      "loss": 0.0963,
+      "step": 12148
+    },
+    {
+      "epoch": 1.968405703175632,
+      "grad_norm": 0.8392459154129028,
+      "learning_rate": 1.3966975217185922e-06,
+      "loss": 0.0924,
+      "step": 12149
+    },
+    {
+      "epoch": 1.9685677252106286,
+      "grad_norm": 0.7114721536636353,
+      "learning_rate": 1.3963051240678652e-06,
+      "loss": 0.0909,
+      "step": 12150
+    },
+    {
+      "epoch": 1.9687297472456255,
+      "grad_norm": 0.8719701766967773,
+      "learning_rate": 1.3959127601875305e-06,
+      "loss": 0.1132,
+      "step": 12151
+    },
+    {
+      "epoch": 1.968891769280622,
+      "grad_norm": 0.7276889085769653,
+      "learning_rate": 1.3955204300895937e-06,
+      "loss": 0.0862,
+      "step": 12152
+    },
+    {
+      "epoch": 1.969053791315619,
+      "grad_norm": 0.7744901180267334,
+      "learning_rate": 1.3951281337860583e-06,
+      "loss": 0.0942,
+      "step": 12153
+    },
+    {
+      "epoch": 1.9692158133506157,
+      "grad_norm": 0.8694067001342773,
+      "learning_rate": 1.3947358712889292e-06,
+      "loss": 0.0981,
+      "step": 12154
+    },
+    {
+      "epoch": 1.9693778353856124,
+      "grad_norm": 0.8138403296470642,
+      "learning_rate": 1.394343642610207e-06,
+      "loss": 0.1028,
+      "step": 12155
+    },
+    {
+      "epoch": 1.9695398574206093,
+      "grad_norm": 0.8411658406257629,
+      "learning_rate": 1.3939514477618944e-06,
+      "loss": 0.0972,
+      "step": 12156
+    },
+    {
+      "epoch": 1.9697018794556058,
+      "grad_norm": 0.7511509656906128,
+      "learning_rate": 1.3935592867559902e-06,
+      "loss": 0.0868,
+      "step": 12157
+    },
+    {
+      "epoch": 1.9698639014906028,
+      "grad_norm": 0.7799199819564819,
+      "learning_rate": 1.3931671596044946e-06,
+      "loss": 0.103,
+      "step": 12158
+    },
+    {
+      "epoch": 1.9700259235255995,
+      "grad_norm": 0.7285469770431519,
+      "learning_rate": 1.3927750663194055e-06,
+      "loss": 0.0917,
+      "step": 12159
+    },
+    {
+      "epoch": 1.9701879455605962,
+      "grad_norm": 0.8250842690467834,
+      "learning_rate": 1.392383006912721e-06,
+      "loss": 0.0929,
+      "step": 12160
+    },
+    {
+      "epoch": 1.970349967595593,
+      "grad_norm": 0.7213485836982727,
+      "learning_rate": 1.391990981396435e-06,
+      "loss": 0.0876,
+      "step": 12161
+    },
+    {
+      "epoch": 1.9705119896305896,
+      "grad_norm": 0.7967683672904968,
+      "learning_rate": 1.3915989897825424e-06,
+      "loss": 0.0975,
+      "step": 12162
+    },
+    {
+      "epoch": 1.9706740116655865,
+      "grad_norm": 0.7662811875343323,
+      "learning_rate": 1.3912070320830406e-06,
+      "loss": 0.0975,
+      "step": 12163
+    },
+    {
+      "epoch": 1.9708360337005832,
+      "grad_norm": 0.7548596262931824,
+      "learning_rate": 1.3908151083099195e-06,
+      "loss": 0.0958,
+      "step": 12164
+    },
+    {
+      "epoch": 1.97099805573558,
+      "grad_norm": 0.7877540588378906,
+      "learning_rate": 1.390423218475172e-06,
+      "loss": 0.0953,
+      "step": 12165
+    },
+    {
+      "epoch": 1.971160077770577,
+      "grad_norm": 0.7539090514183044,
+      "learning_rate": 1.3900313625907886e-06,
+      "loss": 0.0947,
+      "step": 12166
+    },
+    {
+      "epoch": 1.9713220998055736,
+      "grad_norm": 0.7667350769042969,
+      "learning_rate": 1.3896395406687597e-06,
+      "loss": 0.091,
+      "step": 12167
+    },
+    {
+      "epoch": 1.9714841218405703,
+      "grad_norm": 0.9332156777381897,
+      "learning_rate": 1.3892477527210734e-06,
+      "loss": 0.1106,
+      "step": 12168
+    },
+    {
+      "epoch": 1.971646143875567,
+      "grad_norm": 0.8445827960968018,
+      "learning_rate": 1.3888559987597182e-06,
+      "loss": 0.1082,
+      "step": 12169
+    },
+    {
+      "epoch": 1.9718081659105637,
+      "grad_norm": 0.8100192546844482,
+      "learning_rate": 1.3884642787966806e-06,
+      "loss": 0.0948,
+      "step": 12170
+    },
+    {
+      "epoch": 1.9719701879455607,
+      "grad_norm": 1.1084216833114624,
+      "learning_rate": 1.3880725928439472e-06,
+      "loss": 0.1227,
+      "step": 12171
+    },
+    {
+      "epoch": 1.9721322099805574,
+      "grad_norm": 0.7511377930641174,
+      "learning_rate": 1.3876809409134994e-06,
+      "loss": 0.0881,
+      "step": 12172
+    },
+    {
+      "epoch": 1.972294232015554,
+      "grad_norm": 0.8674477338790894,
+      "learning_rate": 1.3872893230173245e-06,
+      "loss": 0.1032,
+      "step": 12173
+    },
+    {
+      "epoch": 1.972456254050551,
+      "grad_norm": 0.8338192701339722,
+      "learning_rate": 1.3868977391674033e-06,
+      "loss": 0.1097,
+      "step": 12174
+    },
+    {
+      "epoch": 1.9726182760855475,
+      "grad_norm": 0.7292324304580688,
+      "learning_rate": 1.3865061893757187e-06,
+      "loss": 0.0908,
+      "step": 12175
+    },
+    {
+      "epoch": 1.9727802981205445,
+      "grad_norm": 0.8517992496490479,
+      "learning_rate": 1.386114673654248e-06,
+      "loss": 0.0971,
+      "step": 12176
+    },
+    {
+      "epoch": 1.9729423201555412,
+      "grad_norm": 0.6976124048233032,
+      "learning_rate": 1.3857231920149738e-06,
+      "loss": 0.0792,
+      "step": 12177
+    },
+    {
+      "epoch": 1.9731043421905379,
+      "grad_norm": 0.8370488286018372,
+      "learning_rate": 1.3853317444698744e-06,
+      "loss": 0.1007,
+      "step": 12178
+    },
+    {
+      "epoch": 1.9732663642255348,
+      "grad_norm": 1.0718648433685303,
+      "learning_rate": 1.3849403310309251e-06,
+      "loss": 0.1228,
+      "step": 12179
+    },
+    {
+      "epoch": 1.9734283862605313,
+      "grad_norm": 0.7935207486152649,
+      "learning_rate": 1.3845489517101036e-06,
+      "loss": 0.1041,
+      "step": 12180
+    },
+    {
+      "epoch": 1.9735904082955282,
+      "grad_norm": 0.8802226185798645,
+      "learning_rate": 1.3841576065193834e-06,
+      "loss": 0.1127,
+      "step": 12181
+    },
+    {
+      "epoch": 1.973752430330525,
+      "grad_norm": 0.8075469136238098,
+      "learning_rate": 1.3837662954707426e-06,
+      "loss": 0.0987,
+      "step": 12182
+    },
+    {
+      "epoch": 1.9739144523655217,
+      "grad_norm": 0.7968436479568481,
+      "learning_rate": 1.3833750185761507e-06,
+      "loss": 0.0968,
+      "step": 12183
+    },
+    {
+      "epoch": 1.9740764744005186,
+      "grad_norm": 0.8242005109786987,
+      "learning_rate": 1.3829837758475808e-06,
+      "loss": 0.0978,
+      "step": 12184
+    },
+    {
+      "epoch": 1.974238496435515,
+      "grad_norm": 0.7901813983917236,
+      "learning_rate": 1.3825925672970048e-06,
+      "loss": 0.1014,
+      "step": 12185
+    },
+    {
+      "epoch": 1.974400518470512,
+      "grad_norm": 0.8069804310798645,
+      "learning_rate": 1.3822013929363914e-06,
+      "loss": 0.0971,
+      "step": 12186
+    },
+    {
+      "epoch": 1.9745625405055087,
+      "grad_norm": 0.8082523345947266,
+      "learning_rate": 1.3818102527777111e-06,
+      "loss": 0.0949,
+      "step": 12187
+    },
+    {
+      "epoch": 1.9747245625405054,
+      "grad_norm": 0.7194974422454834,
+      "learning_rate": 1.3814191468329307e-06,
+      "loss": 0.0868,
+      "step": 12188
+    },
+    {
+      "epoch": 1.9748865845755024,
+      "grad_norm": 0.8970668911933899,
+      "learning_rate": 1.3810280751140188e-06,
+      "loss": 0.1057,
+      "step": 12189
+    },
+    {
+      "epoch": 1.975048606610499,
+      "grad_norm": 0.7686210870742798,
+      "learning_rate": 1.3806370376329388e-06,
+      "loss": 0.0913,
+      "step": 12190
+    },
+    {
+      "epoch": 1.9752106286454958,
+      "grad_norm": 0.8302438259124756,
+      "learning_rate": 1.3802460344016552e-06,
+      "loss": 0.1052,
+      "step": 12191
+    },
+    {
+      "epoch": 1.9753726506804925,
+      "grad_norm": 0.7827860116958618,
+      "learning_rate": 1.3798550654321347e-06,
+      "loss": 0.091,
+      "step": 12192
+    },
+    {
+      "epoch": 1.9755346727154892,
+      "grad_norm": 0.8492727875709534,
+      "learning_rate": 1.3794641307363393e-06,
+      "loss": 0.1065,
+      "step": 12193
+    },
+    {
+      "epoch": 1.9756966947504861,
+      "grad_norm": 0.8518586158752441,
+      "learning_rate": 1.379073230326229e-06,
+      "loss": 0.1037,
+      "step": 12194
+    },
+    {
+      "epoch": 1.9758587167854829,
+      "grad_norm": 0.8506153225898743,
+      "learning_rate": 1.378682364213765e-06,
+      "loss": 0.1091,
+      "step": 12195
+    },
+    {
+      "epoch": 1.9760207388204796,
+      "grad_norm": 0.895915687084198,
+      "learning_rate": 1.3782915324109075e-06,
+      "loss": 0.1101,
+      "step": 12196
+    },
+    {
+      "epoch": 1.9761827608554765,
+      "grad_norm": 0.8467504978179932,
+      "learning_rate": 1.377900734929614e-06,
+      "loss": 0.0992,
+      "step": 12197
+    },
+    {
+      "epoch": 1.976344782890473,
+      "grad_norm": 0.7702296376228333,
+      "learning_rate": 1.3775099717818432e-06,
+      "loss": 0.0924,
+      "step": 12198
+    },
+    {
+      "epoch": 1.97650680492547,
+      "grad_norm": 0.8202542066574097,
+      "learning_rate": 1.377119242979551e-06,
+      "loss": 0.0916,
+      "step": 12199
+    },
+    {
+      "epoch": 1.9766688269604666,
+      "grad_norm": 0.9093190431594849,
+      "learning_rate": 1.376728548534692e-06,
+      "loss": 0.1095,
+      "step": 12200
+    },
+    {
+      "epoch": 1.9768308489954634,
+      "grad_norm": 0.8310642242431641,
+      "learning_rate": 1.3763378884592215e-06,
+      "loss": 0.0967,
+      "step": 12201
+    },
+    {
+      "epoch": 1.9769928710304603,
+      "grad_norm": 0.774259090423584,
+      "learning_rate": 1.3759472627650926e-06,
+      "loss": 0.0962,
+      "step": 12202
+    },
+    {
+      "epoch": 1.9771548930654568,
+      "grad_norm": 0.7820743918418884,
+      "learning_rate": 1.3755566714642571e-06,
+      "loss": 0.0974,
+      "step": 12203
+    },
+    {
+      "epoch": 1.9773169151004537,
+      "grad_norm": 0.8799723386764526,
+      "learning_rate": 1.3751661145686673e-06,
+      "loss": 0.1064,
+      "step": 12204
+    },
+    {
+      "epoch": 1.9774789371354504,
+      "grad_norm": 0.7707881927490234,
+      "learning_rate": 1.3747755920902706e-06,
+      "loss": 0.094,
+      "step": 12205
+    },
+    {
+      "epoch": 1.9776409591704471,
+      "grad_norm": 0.9051207900047302,
+      "learning_rate": 1.3743851040410183e-06,
+      "loss": 0.1111,
+      "step": 12206
+    },
+    {
+      "epoch": 1.977802981205444,
+      "grad_norm": 0.8284469246864319,
+      "learning_rate": 1.3739946504328594e-06,
+      "loss": 0.1007,
+      "step": 12207
+    },
+    {
+      "epoch": 1.9779650032404406,
+      "grad_norm": 0.7556292414665222,
+      "learning_rate": 1.3736042312777381e-06,
+      "loss": 0.0925,
+      "step": 12208
+    },
+    {
+      "epoch": 1.9781270252754375,
+      "grad_norm": 0.798641562461853,
+      "learning_rate": 1.3732138465876012e-06,
+      "loss": 0.0906,
+      "step": 12209
+    },
+    {
+      "epoch": 1.9782890473104342,
+      "grad_norm": 0.7734487652778625,
+      "learning_rate": 1.3728234963743931e-06,
+      "loss": 0.0985,
+      "step": 12210
+    },
+    {
+      "epoch": 1.978451069345431,
+      "grad_norm": 0.7930302619934082,
+      "learning_rate": 1.3724331806500604e-06,
+      "loss": 0.0921,
+      "step": 12211
+    },
+    {
+      "epoch": 1.9786130913804278,
+      "grad_norm": 0.8522673845291138,
+      "learning_rate": 1.3720428994265427e-06,
+      "loss": 0.0993,
+      "step": 12212
+    },
+    {
+      "epoch": 1.9787751134154243,
+      "grad_norm": 0.7778444290161133,
+      "learning_rate": 1.3716526527157826e-06,
+      "loss": 0.0865,
+      "step": 12213
+    },
+    {
+      "epoch": 1.9789371354504213,
+      "grad_norm": 0.8435918092727661,
+      "learning_rate": 1.3712624405297209e-06,
+      "loss": 0.0991,
+      "step": 12214
+    },
+    {
+      "epoch": 1.979099157485418,
+      "grad_norm": 0.9310854077339172,
+      "learning_rate": 1.3708722628802968e-06,
+      "loss": 0.1127,
+      "step": 12215
+    },
+    {
+      "epoch": 1.9792611795204147,
+      "grad_norm": 0.7594826221466064,
+      "learning_rate": 1.3704821197794491e-06,
+      "loss": 0.093,
+      "step": 12216
+    },
+    {
+      "epoch": 1.9794232015554116,
+      "grad_norm": 0.6724945902824402,
+      "learning_rate": 1.3700920112391152e-06,
+      "loss": 0.0864,
+      "step": 12217
+    },
+    {
+      "epoch": 1.9795852235904083,
+      "grad_norm": 0.7259680032730103,
+      "learning_rate": 1.369701937271231e-06,
+      "loss": 0.0876,
+      "step": 12218
+    },
+    {
+      "epoch": 1.979747245625405,
+      "grad_norm": 1.109673261642456,
+      "learning_rate": 1.369311897887733e-06,
+      "loss": 0.107,
+      "step": 12219
+    },
+    {
+      "epoch": 1.9799092676604018,
+      "grad_norm": 0.7597100138664246,
+      "learning_rate": 1.3689218931005543e-06,
+      "loss": 0.09,
+      "step": 12220
+    },
+    {
+      "epoch": 1.9800712896953985,
+      "grad_norm": 0.8943421244621277,
+      "learning_rate": 1.3685319229216287e-06,
+      "loss": 0.1104,
+      "step": 12221
+    },
+    {
+      "epoch": 1.9802333117303954,
+      "grad_norm": 0.8570875525474548,
+      "learning_rate": 1.368141987362889e-06,
+      "loss": 0.0965,
+      "step": 12222
+    },
+    {
+      "epoch": 1.9803953337653921,
+      "grad_norm": 0.8116633296012878,
+      "learning_rate": 1.3677520864362644e-06,
+      "loss": 0.0989,
+      "step": 12223
+    },
+    {
+      "epoch": 1.9805573558003888,
+      "grad_norm": 0.8231261968612671,
+      "learning_rate": 1.3673622201536852e-06,
+      "loss": 0.0923,
+      "step": 12224
+    },
+    {
+      "epoch": 1.9807193778353858,
+      "grad_norm": 0.7679911851882935,
+      "learning_rate": 1.366972388527082e-06,
+      "loss": 0.0988,
+      "step": 12225
+    },
+    {
+      "epoch": 1.9808813998703823,
+      "grad_norm": 0.7934950590133667,
+      "learning_rate": 1.3665825915683829e-06,
+      "loss": 0.0949,
+      "step": 12226
+    },
+    {
+      "epoch": 1.9810434219053792,
+      "grad_norm": 0.8774157762527466,
+      "learning_rate": 1.3661928292895123e-06,
+      "loss": 0.1021,
+      "step": 12227
+    },
+    {
+      "epoch": 1.981205443940376,
+      "grad_norm": 0.8431322574615479,
+      "learning_rate": 1.3658031017023977e-06,
+      "loss": 0.1071,
+      "step": 12228
+    },
+    {
+      "epoch": 1.9813674659753726,
+      "grad_norm": 0.7929509878158569,
+      "learning_rate": 1.3654134088189636e-06,
+      "loss": 0.0955,
+      "step": 12229
+    },
+    {
+      "epoch": 1.9815294880103695,
+      "grad_norm": 0.8263244032859802,
+      "learning_rate": 1.3650237506511333e-06,
+      "loss": 0.0889,
+      "step": 12230
+    },
+    {
+      "epoch": 1.981691510045366,
+      "grad_norm": 0.8262062668800354,
+      "learning_rate": 1.36463412721083e-06,
+      "loss": 0.1029,
+      "step": 12231
+    },
+    {
+      "epoch": 1.981853532080363,
+      "grad_norm": 0.8104032874107361,
+      "learning_rate": 1.3642445385099746e-06,
+      "loss": 0.093,
+      "step": 12232
+    },
+    {
+      "epoch": 1.9820155541153597,
+      "grad_norm": 0.7863628268241882,
+      "learning_rate": 1.3638549845604886e-06,
+      "loss": 0.0944,
+      "step": 12233
+    },
+    {
+      "epoch": 1.9821775761503564,
+      "grad_norm": 0.7456783652305603,
+      "learning_rate": 1.363465465374289e-06,
+      "loss": 0.093,
+      "step": 12234
+    },
+    {
+      "epoch": 1.9823395981853533,
+      "grad_norm": 0.7725854516029358,
+      "learning_rate": 1.3630759809632965e-06,
+      "loss": 0.0886,
+      "step": 12235
+    },
+    {
+      "epoch": 1.9825016202203498,
+      "grad_norm": 0.7280664443969727,
+      "learning_rate": 1.362686531339428e-06,
+      "loss": 0.0885,
+      "step": 12236
+    },
+    {
+      "epoch": 1.9826636422553467,
+      "grad_norm": 0.9137564301490784,
+      "learning_rate": 1.3622971165146005e-06,
+      "loss": 0.1132,
+      "step": 12237
+    },
+    {
+      "epoch": 1.9828256642903435,
+      "grad_norm": 0.8117855787277222,
+      "learning_rate": 1.3619077365007266e-06,
+      "loss": 0.0978,
+      "step": 12238
+    },
+    {
+      "epoch": 1.9829876863253402,
+      "grad_norm": 0.7241969108581543,
+      "learning_rate": 1.3615183913097211e-06,
+      "loss": 0.0911,
+      "step": 12239
+    },
+    {
+      "epoch": 1.983149708360337,
+      "grad_norm": 0.9119266867637634,
+      "learning_rate": 1.3611290809534997e-06,
+      "loss": 0.1046,
+      "step": 12240
+    },
+    {
+      "epoch": 1.9833117303953338,
+      "grad_norm": 0.7932629585266113,
+      "learning_rate": 1.3607398054439713e-06,
+      "loss": 0.1004,
+      "step": 12241
+    },
+    {
+      "epoch": 1.9834737524303305,
+      "grad_norm": 0.79780513048172,
+      "learning_rate": 1.3603505647930481e-06,
+      "loss": 0.0932,
+      "step": 12242
+    },
+    {
+      "epoch": 1.9836357744653272,
+      "grad_norm": 0.7685292959213257,
+      "learning_rate": 1.3599613590126388e-06,
+      "loss": 0.0969,
+      "step": 12243
+    },
+    {
+      "epoch": 1.983797796500324,
+      "grad_norm": 0.7224975228309631,
+      "learning_rate": 1.3595721881146548e-06,
+      "loss": 0.0876,
+      "step": 12244
+    },
+    {
+      "epoch": 1.9839598185353209,
+      "grad_norm": 0.7350090742111206,
+      "learning_rate": 1.359183052111001e-06,
+      "loss": 0.0943,
+      "step": 12245
+    },
+    {
+      "epoch": 1.9841218405703176,
+      "grad_norm": 0.7922118306159973,
+      "learning_rate": 1.3587939510135856e-06,
+      "loss": 0.0939,
+      "step": 12246
+    },
+    {
+      "epoch": 1.9842838626053143,
+      "grad_norm": 0.7330470085144043,
+      "learning_rate": 1.358404884834313e-06,
+      "loss": 0.0913,
+      "step": 12247
+    },
+    {
+      "epoch": 1.9844458846403112,
+      "grad_norm": 0.8037322759628296,
+      "learning_rate": 1.3580158535850884e-06,
+      "loss": 0.1021,
+      "step": 12248
+    },
+    {
+      "epoch": 1.9846079066753077,
+      "grad_norm": 0.7963745594024658,
+      "learning_rate": 1.3576268572778156e-06,
+      "loss": 0.1056,
+      "step": 12249
+    },
+    {
+      "epoch": 1.9847699287103047,
+      "grad_norm": 0.8091640472412109,
+      "learning_rate": 1.357237895924396e-06,
+      "loss": 0.0959,
+      "step": 12250
+    },
+    {
+      "epoch": 1.9849319507453014,
+      "grad_norm": 0.8512594103813171,
+      "learning_rate": 1.3568489695367325e-06,
+      "loss": 0.0948,
+      "step": 12251
+    },
+    {
+      "epoch": 1.985093972780298,
+      "grad_norm": 0.8013456463813782,
+      "learning_rate": 1.3564600781267234e-06,
+      "loss": 0.0988,
+      "step": 12252
+    },
+    {
+      "epoch": 1.985255994815295,
+      "grad_norm": 0.6600555181503296,
+      "learning_rate": 1.3560712217062676e-06,
+      "loss": 0.0808,
+      "step": 12253
+    },
+    {
+      "epoch": 1.9854180168502915,
+      "grad_norm": 0.8468796014785767,
+      "learning_rate": 1.3556824002872648e-06,
+      "loss": 0.0989,
+      "step": 12254
+    },
+    {
+      "epoch": 1.9855800388852884,
+      "grad_norm": 0.8054071664810181,
+      "learning_rate": 1.3552936138816124e-06,
+      "loss": 0.1016,
+      "step": 12255
+    },
+    {
+      "epoch": 1.9857420609202852,
+      "grad_norm": 0.7995196580886841,
+      "learning_rate": 1.3549048625012046e-06,
+      "loss": 0.0946,
+      "step": 12256
+    },
+    {
+      "epoch": 1.9859040829552819,
+      "grad_norm": 0.9262045621871948,
+      "learning_rate": 1.3545161461579367e-06,
+      "loss": 0.1046,
+      "step": 12257
+    },
+    {
+      "epoch": 1.9860661049902788,
+      "grad_norm": 0.9522650837898254,
+      "learning_rate": 1.354127464863703e-06,
+      "loss": 0.1038,
+      "step": 12258
+    },
+    {
+      "epoch": 1.9862281270252753,
+      "grad_norm": 0.8085470199584961,
+      "learning_rate": 1.3537388186303956e-06,
+      "loss": 0.0969,
+      "step": 12259
+    },
+    {
+      "epoch": 1.9863901490602722,
+      "grad_norm": 0.7567479014396667,
+      "learning_rate": 1.3533502074699065e-06,
+      "loss": 0.0935,
+      "step": 12260
+    },
+    {
+      "epoch": 1.986552171095269,
+      "grad_norm": 0.7655736804008484,
+      "learning_rate": 1.3529616313941264e-06,
+      "loss": 0.0944,
+      "step": 12261
+    },
+    {
+      "epoch": 1.9867141931302656,
+      "grad_norm": 0.8020748496055603,
+      "learning_rate": 1.3525730904149443e-06,
+      "loss": 0.0902,
+      "step": 12262
+    },
+    {
+      "epoch": 1.9868762151652626,
+      "grad_norm": 0.8509409427642822,
+      "learning_rate": 1.3521845845442489e-06,
+      "loss": 0.1011,
+      "step": 12263
+    },
+    {
+      "epoch": 1.987038237200259,
+      "grad_norm": 0.8280978798866272,
+      "learning_rate": 1.351796113793928e-06,
+      "loss": 0.0972,
+      "step": 12264
+    },
+    {
+      "epoch": 1.987200259235256,
+      "grad_norm": 0.8213714957237244,
+      "learning_rate": 1.351407678175867e-06,
+      "loss": 0.0884,
+      "step": 12265
+    },
+    {
+      "epoch": 1.9873622812702527,
+      "grad_norm": 0.8149861097335815,
+      "learning_rate": 1.3510192777019527e-06,
+      "loss": 0.0976,
+      "step": 12266
+    },
+    {
+      "epoch": 1.9875243033052494,
+      "grad_norm": 0.7219667434692383,
+      "learning_rate": 1.3506309123840659e-06,
+      "loss": 0.0856,
+      "step": 12267
+    },
+    {
+      "epoch": 1.9876863253402464,
+      "grad_norm": 0.7291889786720276,
+      "learning_rate": 1.3502425822340925e-06,
+      "loss": 0.0941,
+      "step": 12268
+    },
+    {
+      "epoch": 1.987848347375243,
+      "grad_norm": 0.7790191769599915,
+      "learning_rate": 1.3498542872639142e-06,
+      "loss": 0.0993,
+      "step": 12269
+    },
+    {
+      "epoch": 1.9880103694102398,
+      "grad_norm": 0.7414619326591492,
+      "learning_rate": 1.3494660274854122e-06,
+      "loss": 0.0905,
+      "step": 12270
+    },
+    {
+      "epoch": 1.9881723914452365,
+      "grad_norm": 0.7423388957977295,
+      "learning_rate": 1.3490778029104646e-06,
+      "loss": 0.0886,
+      "step": 12271
+    },
+    {
+      "epoch": 1.9883344134802332,
+      "grad_norm": 0.8346954584121704,
+      "learning_rate": 1.3486896135509503e-06,
+      "loss": 0.1073,
+      "step": 12272
+    },
+    {
+      "epoch": 1.9884964355152301,
+      "grad_norm": 0.7810722589492798,
+      "learning_rate": 1.3483014594187493e-06,
+      "loss": 0.0918,
+      "step": 12273
+    },
+    {
+      "epoch": 1.9886584575502269,
+      "grad_norm": 0.7663103938102722,
+      "learning_rate": 1.3479133405257355e-06,
+      "loss": 0.0959,
+      "step": 12274
+    },
+    {
+      "epoch": 1.9888204795852236,
+      "grad_norm": 0.7423315048217773,
+      "learning_rate": 1.347525256883786e-06,
+      "loss": 0.0895,
+      "step": 12275
+    },
+    {
+      "epoch": 1.9889825016202205,
+      "grad_norm": 0.7934430241584778,
+      "learning_rate": 1.3471372085047743e-06,
+      "loss": 0.0964,
+      "step": 12276
+    },
+    {
+      "epoch": 1.989144523655217,
+      "grad_norm": 0.8418073654174805,
+      "learning_rate": 1.346749195400574e-06,
+      "loss": 0.1047,
+      "step": 12277
+    },
+    {
+      "epoch": 1.989306545690214,
+      "grad_norm": 0.7353390455245972,
+      "learning_rate": 1.3463612175830578e-06,
+      "loss": 0.0857,
+      "step": 12278
+    },
+    {
+      "epoch": 1.9894685677252106,
+      "grad_norm": 0.8936602473258972,
+      "learning_rate": 1.3459732750640967e-06,
+      "loss": 0.1083,
+      "step": 12279
+    },
+    {
+      "epoch": 1.9896305897602073,
+      "grad_norm": 0.7523042559623718,
+      "learning_rate": 1.3455853678555605e-06,
+      "loss": 0.086,
+      "step": 12280
+    },
+    {
+      "epoch": 1.9897926117952043,
+      "grad_norm": 0.7889019846916199,
+      "learning_rate": 1.3451974959693193e-06,
+      "loss": 0.0887,
+      "step": 12281
+    },
+    {
+      "epoch": 1.9899546338302008,
+      "grad_norm": 0.8025287389755249,
+      "learning_rate": 1.3448096594172383e-06,
+      "loss": 0.0981,
+      "step": 12282
+    },
+    {
+      "epoch": 1.9901166558651977,
+      "grad_norm": 0.7614328861236572,
+      "learning_rate": 1.3444218582111872e-06,
+      "loss": 0.0908,
+      "step": 12283
+    },
+    {
+      "epoch": 1.9902786779001944,
+      "grad_norm": 0.7384836673736572,
+      "learning_rate": 1.344034092363032e-06,
+      "loss": 0.0878,
+      "step": 12284
+    },
+    {
+      "epoch": 1.9904406999351911,
+      "grad_norm": 0.7912430167198181,
+      "learning_rate": 1.3436463618846351e-06,
+      "loss": 0.0902,
+      "step": 12285
+    },
+    {
+      "epoch": 1.990602721970188,
+      "grad_norm": 0.8135248422622681,
+      "learning_rate": 1.343258666787861e-06,
+      "loss": 0.0986,
+      "step": 12286
+    },
+    {
+      "epoch": 1.9907647440051845,
+      "grad_norm": 0.7323644757270813,
+      "learning_rate": 1.3428710070845716e-06,
+      "loss": 0.0917,
+      "step": 12287
+    },
+    {
+      "epoch": 1.9909267660401815,
+      "grad_norm": 0.7355465292930603,
+      "learning_rate": 1.3424833827866312e-06,
+      "loss": 0.0901,
+      "step": 12288
+    },
+    {
+      "epoch": 1.9910887880751782,
+      "grad_norm": 0.7884531021118164,
+      "learning_rate": 1.342095793905897e-06,
+      "loss": 0.0937,
+      "step": 12289
+    },
+    {
+      "epoch": 1.991250810110175,
+      "grad_norm": 0.7865089178085327,
+      "learning_rate": 1.3417082404542295e-06,
+      "loss": 0.0997,
+      "step": 12290
+    },
+    {
+      "epoch": 1.9914128321451718,
+      "grad_norm": 0.711365818977356,
+      "learning_rate": 1.3413207224434867e-06,
+      "loss": 0.0833,
+      "step": 12291
+    },
+    {
+      "epoch": 1.9915748541801686,
+      "grad_norm": 0.9250528812408447,
+      "learning_rate": 1.3409332398855263e-06,
+      "loss": 0.1068,
+      "step": 12292
+    },
+    {
+      "epoch": 1.9917368762151653,
+      "grad_norm": 0.9538305401802063,
+      "learning_rate": 1.3405457927922032e-06,
+      "loss": 0.0967,
+      "step": 12293
+    },
+    {
+      "epoch": 1.991898898250162,
+      "grad_norm": 0.7699099183082581,
+      "learning_rate": 1.3401583811753735e-06,
+      "loss": 0.0936,
+      "step": 12294
+    },
+    {
+      "epoch": 1.9920609202851587,
+      "grad_norm": 0.7329535484313965,
+      "learning_rate": 1.3397710050468903e-06,
+      "loss": 0.0914,
+      "step": 12295
+    },
+    {
+      "epoch": 1.9922229423201556,
+      "grad_norm": 0.7969571352005005,
+      "learning_rate": 1.339383664418607e-06,
+      "loss": 0.0971,
+      "step": 12296
+    },
+    {
+      "epoch": 1.9923849643551523,
+      "grad_norm": 0.7721100449562073,
+      "learning_rate": 1.3389963593023747e-06,
+      "loss": 0.0894,
+      "step": 12297
+    },
+    {
+      "epoch": 1.992546986390149,
+      "grad_norm": 0.8299007415771484,
+      "learning_rate": 1.3386090897100442e-06,
+      "loss": 0.0902,
+      "step": 12298
+    },
+    {
+      "epoch": 1.992709008425146,
+      "grad_norm": 0.8381181955337524,
+      "learning_rate": 1.338221855653466e-06,
+      "loss": 0.0899,
+      "step": 12299
+    },
+    {
+      "epoch": 1.9928710304601425,
+      "grad_norm": 0.7096841931343079,
+      "learning_rate": 1.3378346571444866e-06,
+      "loss": 0.0798,
+      "step": 12300
+    },
+    {
+      "epoch": 1.9930330524951394,
+      "grad_norm": 0.7077094912528992,
+      "learning_rate": 1.3374474941949535e-06,
+      "loss": 0.0878,
+      "step": 12301
+    },
+    {
+      "epoch": 1.9931950745301361,
+      "grad_norm": 0.8370009660720825,
+      "learning_rate": 1.3370603668167156e-06,
+      "loss": 0.1005,
+      "step": 12302
+    },
+    {
+      "epoch": 1.9933570965651328,
+      "grad_norm": 0.7673652768135071,
+      "learning_rate": 1.3366732750216154e-06,
+      "loss": 0.0939,
+      "step": 12303
+    },
+    {
+      "epoch": 1.9935191186001298,
+      "grad_norm": 0.7367184162139893,
+      "learning_rate": 1.3362862188214977e-06,
+      "loss": 0.0913,
+      "step": 12304
+    },
+    {
+      "epoch": 1.9936811406351262,
+      "grad_norm": 0.8323283195495605,
+      "learning_rate": 1.3358991982282055e-06,
+      "loss": 0.1035,
+      "step": 12305
+    },
+    {
+      "epoch": 1.9938431626701232,
+      "grad_norm": 0.7226582765579224,
+      "learning_rate": 1.3355122132535806e-06,
+      "loss": 0.0931,
+      "step": 12306
+    },
+    {
+      "epoch": 1.99400518470512,
+      "grad_norm": 0.8198063969612122,
+      "learning_rate": 1.3351252639094641e-06,
+      "loss": 0.1003,
+      "step": 12307
+    },
+    {
+      "epoch": 1.9941672067401166,
+      "grad_norm": 0.6655625104904175,
+      "learning_rate": 1.3347383502076955e-06,
+      "loss": 0.0806,
+      "step": 12308
+    },
+    {
+      "epoch": 1.9943292287751135,
+      "grad_norm": 0.8288505673408508,
+      "learning_rate": 1.3343514721601136e-06,
+      "loss": 0.0963,
+      "step": 12309
+    },
+    {
+      "epoch": 1.99449125081011,
+      "grad_norm": 0.7732831835746765,
+      "learning_rate": 1.333964629778556e-06,
+      "loss": 0.1015,
+      "step": 12310
+    },
+    {
+      "epoch": 1.994653272845107,
+      "grad_norm": 0.821479320526123,
+      "learning_rate": 1.3335778230748588e-06,
+      "loss": 0.1061,
+      "step": 12311
+    },
+    {
+      "epoch": 1.9948152948801037,
+      "grad_norm": 0.8334558010101318,
+      "learning_rate": 1.3331910520608576e-06,
+      "loss": 0.095,
+      "step": 12312
+    },
+    {
+      "epoch": 1.9949773169151004,
+      "grad_norm": 0.8023782968521118,
+      "learning_rate": 1.3328043167483868e-06,
+      "loss": 0.1001,
+      "step": 12313
+    },
+    {
+      "epoch": 1.9951393389500973,
+      "grad_norm": 0.8799611330032349,
+      "learning_rate": 1.3324176171492798e-06,
+      "loss": 0.1116,
+      "step": 12314
+    },
+    {
+      "epoch": 1.9953013609850938,
+      "grad_norm": 0.8682718276977539,
+      "learning_rate": 1.3320309532753667e-06,
+      "loss": 0.1067,
+      "step": 12315
+    },
+    {
+      "epoch": 1.9954633830200907,
+      "grad_norm": 0.7998306751251221,
+      "learning_rate": 1.3316443251384808e-06,
+      "loss": 0.0978,
+      "step": 12316
+    },
+    {
+      "epoch": 1.9956254050550875,
+      "grad_norm": 1.0022040605545044,
+      "learning_rate": 1.3312577327504522e-06,
+      "loss": 0.1222,
+      "step": 12317
+    },
+    {
+      "epoch": 1.9957874270900842,
+      "grad_norm": 0.7971022129058838,
+      "learning_rate": 1.3308711761231074e-06,
+      "loss": 0.0886,
+      "step": 12318
+    },
+    {
+      "epoch": 1.995949449125081,
+      "grad_norm": 0.7757930755615234,
+      "learning_rate": 1.3304846552682756e-06,
+      "loss": 0.0984,
+      "step": 12319
+    },
+    {
+      "epoch": 1.9961114711600778,
+      "grad_norm": 0.8637643456459045,
+      "learning_rate": 1.3300981701977834e-06,
+      "loss": 0.1074,
+      "step": 12320
+    },
+    {
+      "epoch": 1.9962734931950745,
+      "grad_norm": 0.7365697622299194,
+      "learning_rate": 1.3297117209234558e-06,
+      "loss": 0.0892,
+      "step": 12321
+    },
+    {
+      "epoch": 1.9964355152300715,
+      "grad_norm": 0.861108660697937,
+      "learning_rate": 1.3293253074571178e-06,
+      "loss": 0.0967,
+      "step": 12322
+    },
+    {
+      "epoch": 1.996597537265068,
+      "grad_norm": 0.7773643732070923,
+      "learning_rate": 1.328938929810592e-06,
+      "loss": 0.0964,
+      "step": 12323
+    },
+    {
+      "epoch": 1.9967595593000649,
+      "grad_norm": 0.7831589579582214,
+      "learning_rate": 1.3285525879957011e-06,
+      "loss": 0.0884,
+      "step": 12324
+    },
+    {
+      "epoch": 1.9969215813350616,
+      "grad_norm": 0.7906063795089722,
+      "learning_rate": 1.3281662820242664e-06,
+      "loss": 0.0979,
+      "step": 12325
+    },
+    {
+      "epoch": 1.9970836033700583,
+      "grad_norm": 0.9140375256538391,
+      "learning_rate": 1.3277800119081077e-06,
+      "loss": 0.1065,
+      "step": 12326
+    },
+    {
+      "epoch": 1.9972456254050552,
+      "grad_norm": 0.8144938349723816,
+      "learning_rate": 1.327393777659044e-06,
+      "loss": 0.1046,
+      "step": 12327
+    },
+    {
+      "epoch": 1.9974076474400517,
+      "grad_norm": 0.795957624912262,
+      "learning_rate": 1.3270075792888937e-06,
+      "loss": 0.0933,
+      "step": 12328
+    },
+    {
+      "epoch": 1.9975696694750487,
+      "grad_norm": 0.7858502268791199,
+      "learning_rate": 1.326621416809472e-06,
+      "loss": 0.093,
+      "step": 12329
+    },
+    {
+      "epoch": 1.9977316915100454,
+      "grad_norm": 0.7812694907188416,
+      "learning_rate": 1.3262352902325944e-06,
+      "loss": 0.1023,
+      "step": 12330
+    },
+    {
+      "epoch": 1.997893713545042,
+      "grad_norm": 0.8662413954734802,
+      "learning_rate": 1.3258491995700777e-06,
+      "loss": 0.1051,
+      "step": 12331
+    },
+    {
+      "epoch": 1.998055735580039,
+      "grad_norm": 0.8278566598892212,
+      "learning_rate": 1.325463144833735e-06,
+      "loss": 0.0963,
+      "step": 12332
+    },
+    {
+      "epoch": 1.9982177576150355,
+      "grad_norm": 0.7443529963493347,
+      "learning_rate": 1.3250771260353764e-06,
+      "loss": 0.0903,
+      "step": 12333
+    },
+    {
+      "epoch": 1.9983797796500324,
+      "grad_norm": 0.7548669576644897,
+      "learning_rate": 1.324691143186814e-06,
+      "loss": 0.0901,
+      "step": 12334
+    },
+    {
+      "epoch": 1.9985418016850292,
+      "grad_norm": 0.839152455329895,
+      "learning_rate": 1.3243051962998598e-06,
+      "loss": 0.0943,
+      "step": 12335
+    },
+    {
+      "epoch": 1.9987038237200259,
+      "grad_norm": 0.7400187849998474,
+      "learning_rate": 1.323919285386321e-06,
+      "loss": 0.0864,
+      "step": 12336
+    },
+    {
+      "epoch": 1.9988658457550228,
+      "grad_norm": 0.7706547379493713,
+      "learning_rate": 1.3235334104580061e-06,
+      "loss": 0.0983,
+      "step": 12337
+    },
+    {
+      "epoch": 1.9990278677900193,
+      "grad_norm": 0.8732326626777649,
+      "learning_rate": 1.3231475715267217e-06,
+      "loss": 0.1046,
+      "step": 12338
+    },
+    {
+      "epoch": 1.9991898898250162,
+      "grad_norm": 0.7628592848777771,
+      "learning_rate": 1.3227617686042734e-06,
+      "loss": 0.0925,
+      "step": 12339
+    },
+    {
+      "epoch": 1.999351911860013,
+      "grad_norm": 0.7879310846328735,
+      "learning_rate": 1.3223760017024661e-06,
+      "loss": 0.1,
+      "step": 12340
+    },
+    {
+      "epoch": 1.9995139338950096,
+      "grad_norm": 0.7694489359855652,
+      "learning_rate": 1.321990270833104e-06,
+      "loss": 0.0851,
+      "step": 12341
+    },
+    {
+      "epoch": 1.9996759559300066,
+      "grad_norm": 0.7563639283180237,
+      "learning_rate": 1.3216045760079882e-06,
+      "loss": 0.0893,
+      "step": 12342
+    },
+    {
+      "epoch": 1.9998379779650033,
+      "grad_norm": 0.8943942189216614,
+      "learning_rate": 1.321218917238922e-06,
+      "loss": 0.1126,
+      "step": 12343
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.822969377040863,
+      "learning_rate": 1.3208332945377022e-06,
+      "loss": 0.1075,
+      "step": 12344
+    },
+    {
+      "epoch": 2.000162022034997,
+      "grad_norm": 0.6357743740081787,
+      "learning_rate": 1.3204477079161312e-06,
+      "loss": 0.0723,
+      "step": 12345
+    },
+    {
+      "epoch": 2.0003240440699934,
+      "grad_norm": 0.5724396109580994,
+      "learning_rate": 1.3200621573860068e-06,
+      "loss": 0.0654,
+      "step": 12346
+    },
+    {
+      "epoch": 2.0004860661049904,
+      "grad_norm": 0.6248465776443481,
+      "learning_rate": 1.319676642959124e-06,
+      "loss": 0.0719,
+      "step": 12347
+    },
+    {
+      "epoch": 2.000648088139987,
+      "grad_norm": 0.5666736960411072,
+      "learning_rate": 1.3192911646472796e-06,
+      "loss": 0.0628,
+      "step": 12348
+    },
+    {
+      "epoch": 2.000810110174984,
+      "grad_norm": 0.5950966477394104,
+      "learning_rate": 1.3189057224622676e-06,
+      "loss": 0.0634,
+      "step": 12349
+    },
+    {
+      "epoch": 2.0009721322099807,
+      "grad_norm": 0.624751091003418,
+      "learning_rate": 1.3185203164158838e-06,
+      "loss": 0.0675,
+      "step": 12350
+    },
+    {
+      "epoch": 2.001134154244977,
+      "grad_norm": 0.7292012572288513,
+      "learning_rate": 1.3181349465199184e-06,
+      "loss": 0.0717,
+      "step": 12351
+    },
+    {
+      "epoch": 2.001296176279974,
+      "grad_norm": 0.5715608596801758,
+      "learning_rate": 1.3177496127861635e-06,
+      "loss": 0.0591,
+      "step": 12352
+    },
+    {
+      "epoch": 2.0014581983149706,
+      "grad_norm": 0.6922410130500793,
+      "learning_rate": 1.31736431522641e-06,
+      "loss": 0.0729,
+      "step": 12353
+    },
+    {
+      "epoch": 2.0016202203499676,
+      "grad_norm": 0.5971651077270508,
+      "learning_rate": 1.3169790538524457e-06,
+      "loss": 0.0648,
+      "step": 12354
+    },
+    {
+      "epoch": 2.0017822423849645,
+      "grad_norm": 0.6563034057617188,
+      "learning_rate": 1.3165938286760599e-06,
+      "loss": 0.0687,
+      "step": 12355
+    },
+    {
+      "epoch": 2.001944264419961,
+      "grad_norm": 0.6092911958694458,
+      "learning_rate": 1.316208639709039e-06,
+      "loss": 0.0641,
+      "step": 12356
+    },
+    {
+      "epoch": 2.002106286454958,
+      "grad_norm": 0.5998396873474121,
+      "learning_rate": 1.3158234869631692e-06,
+      "loss": 0.0602,
+      "step": 12357
+    },
+    {
+      "epoch": 2.002268308489955,
+      "grad_norm": 0.6295647025108337,
+      "learning_rate": 1.3154383704502349e-06,
+      "loss": 0.0667,
+      "step": 12358
+    },
+    {
+      "epoch": 2.0024303305249513,
+      "grad_norm": 0.6843347549438477,
+      "learning_rate": 1.31505329018202e-06,
+      "loss": 0.0644,
+      "step": 12359
+    },
+    {
+      "epoch": 2.0025923525599483,
+      "grad_norm": 0.8760754466056824,
+      "learning_rate": 1.3146682461703069e-06,
+      "loss": 0.0674,
+      "step": 12360
+    },
+    {
+      "epoch": 2.0027543745949448,
+      "grad_norm": 0.617087721824646,
+      "learning_rate": 1.314283238426878e-06,
+      "loss": 0.0603,
+      "step": 12361
+    },
+    {
+      "epoch": 2.0029163966299417,
+      "grad_norm": 0.7366885542869568,
+      "learning_rate": 1.3138982669635117e-06,
+      "loss": 0.0619,
+      "step": 12362
+    },
+    {
+      "epoch": 2.0030784186649386,
+      "grad_norm": 0.6823921799659729,
+      "learning_rate": 1.3135133317919868e-06,
+      "loss": 0.0606,
+      "step": 12363
+    },
+    {
+      "epoch": 2.003240440699935,
+      "grad_norm": 0.6802733540534973,
+      "learning_rate": 1.313128432924084e-06,
+      "loss": 0.0667,
+      "step": 12364
+    },
+    {
+      "epoch": 2.003402462734932,
+      "grad_norm": 0.6357323527336121,
+      "learning_rate": 1.3127435703715802e-06,
+      "loss": 0.0608,
+      "step": 12365
+    },
+    {
+      "epoch": 2.0035644847699285,
+      "grad_norm": 0.6992286443710327,
+      "learning_rate": 1.3123587441462487e-06,
+      "loss": 0.0601,
+      "step": 12366
+    },
+    {
+      "epoch": 2.0037265068049255,
+      "grad_norm": 0.7034692168235779,
+      "learning_rate": 1.3119739542598655e-06,
+      "loss": 0.0617,
+      "step": 12367
+    },
+    {
+      "epoch": 2.0038885288399224,
+      "grad_norm": 0.7643810510635376,
+      "learning_rate": 1.3115892007242046e-06,
+      "loss": 0.0653,
+      "step": 12368
+    },
+    {
+      "epoch": 2.004050550874919,
+      "grad_norm": 0.6599617004394531,
+      "learning_rate": 1.3112044835510378e-06,
+      "loss": 0.0582,
+      "step": 12369
+    },
+    {
+      "epoch": 2.004212572909916,
+      "grad_norm": 0.7879641652107239,
+      "learning_rate": 1.3108198027521374e-06,
+      "loss": 0.0667,
+      "step": 12370
+    },
+    {
+      "epoch": 2.0043745949449123,
+      "grad_norm": 0.7727499008178711,
+      "learning_rate": 1.3104351583392732e-06,
+      "loss": 0.0622,
+      "step": 12371
+    },
+    {
+      "epoch": 2.0045366169799093,
+      "grad_norm": 0.7340289354324341,
+      "learning_rate": 1.3100505503242156e-06,
+      "loss": 0.0605,
+      "step": 12372
+    },
+    {
+      "epoch": 2.004698639014906,
+      "grad_norm": 0.7155749797821045,
+      "learning_rate": 1.3096659787187294e-06,
+      "loss": 0.062,
+      "step": 12373
+    },
+    {
+      "epoch": 2.0048606610499027,
+      "grad_norm": 0.7992011308670044,
+      "learning_rate": 1.3092814435345845e-06,
+      "loss": 0.0655,
+      "step": 12374
+    },
+    {
+      "epoch": 2.0050226830848996,
+      "grad_norm": 0.6771997809410095,
+      "learning_rate": 1.3088969447835464e-06,
+      "loss": 0.0567,
+      "step": 12375
+    },
+    {
+      "epoch": 2.005184705119896,
+      "grad_norm": 0.7328808307647705,
+      "learning_rate": 1.3085124824773797e-06,
+      "loss": 0.0597,
+      "step": 12376
+    },
+    {
+      "epoch": 2.005346727154893,
+      "grad_norm": 0.7776128649711609,
+      "learning_rate": 1.3081280566278464e-06,
+      "loss": 0.0617,
+      "step": 12377
+    },
+    {
+      "epoch": 2.00550874918989,
+      "grad_norm": 0.8074184656143188,
+      "learning_rate": 1.307743667246711e-06,
+      "loss": 0.0738,
+      "step": 12378
+    },
+    {
+      "epoch": 2.0056707712248865,
+      "grad_norm": 0.7445822358131409,
+      "learning_rate": 1.3073593143457353e-06,
+      "loss": 0.0609,
+      "step": 12379
+    },
+    {
+      "epoch": 2.0058327932598834,
+      "grad_norm": 0.811712920665741,
+      "learning_rate": 1.306974997936677e-06,
+      "loss": 0.0675,
+      "step": 12380
+    },
+    {
+      "epoch": 2.00599481529488,
+      "grad_norm": 0.8170737624168396,
+      "learning_rate": 1.306590718031297e-06,
+      "loss": 0.0713,
+      "step": 12381
+    },
+    {
+      "epoch": 2.006156837329877,
+      "grad_norm": 0.6781127452850342,
+      "learning_rate": 1.3062064746413522e-06,
+      "loss": 0.0607,
+      "step": 12382
+    },
+    {
+      "epoch": 2.0063188593648738,
+      "grad_norm": 0.8460241556167603,
+      "learning_rate": 1.305822267778602e-06,
+      "loss": 0.0651,
+      "step": 12383
+    },
+    {
+      "epoch": 2.0064808813998702,
+      "grad_norm": 0.7995967864990234,
+      "learning_rate": 1.3054380974547998e-06,
+      "loss": 0.0606,
+      "step": 12384
+    },
+    {
+      "epoch": 2.006642903434867,
+      "grad_norm": 0.8841012120246887,
+      "learning_rate": 1.3050539636817012e-06,
+      "loss": 0.0734,
+      "step": 12385
+    },
+    {
+      "epoch": 2.006804925469864,
+      "grad_norm": 0.8261283040046692,
+      "learning_rate": 1.3046698664710595e-06,
+      "loss": 0.0605,
+      "step": 12386
+    },
+    {
+      "epoch": 2.0069669475048606,
+      "grad_norm": 0.6940838098526001,
+      "learning_rate": 1.304285805834627e-06,
+      "loss": 0.0601,
+      "step": 12387
+    },
+    {
+      "epoch": 2.0071289695398575,
+      "grad_norm": 0.8337867856025696,
+      "learning_rate": 1.3039017817841553e-06,
+      "loss": 0.0617,
+      "step": 12388
+    },
+    {
+      "epoch": 2.007290991574854,
+      "grad_norm": 0.7418117523193359,
+      "learning_rate": 1.3035177943313947e-06,
+      "loss": 0.0594,
+      "step": 12389
+    },
+    {
+      "epoch": 2.007453013609851,
+      "grad_norm": 0.853374183177948,
+      "learning_rate": 1.3031338434880952e-06,
+      "loss": 0.0681,
+      "step": 12390
+    },
+    {
+      "epoch": 2.007615035644848,
+      "grad_norm": 0.8610039949417114,
+      "learning_rate": 1.3027499292660022e-06,
+      "loss": 0.0707,
+      "step": 12391
+    },
+    {
+      "epoch": 2.0077770576798444,
+      "grad_norm": 0.8662420511245728,
+      "learning_rate": 1.3023660516768638e-06,
+      "loss": 0.0693,
+      "step": 12392
+    },
+    {
+      "epoch": 2.0079390797148413,
+      "grad_norm": 0.8422592282295227,
+      "learning_rate": 1.3019822107324267e-06,
+      "loss": 0.0759,
+      "step": 12393
+    },
+    {
+      "epoch": 2.008101101749838,
+      "grad_norm": 0.8292291760444641,
+      "learning_rate": 1.301598406444436e-06,
+      "loss": 0.0707,
+      "step": 12394
+    },
+    {
+      "epoch": 2.0082631237848347,
+      "grad_norm": 0.7821356058120728,
+      "learning_rate": 1.3012146388246328e-06,
+      "loss": 0.0616,
+      "step": 12395
+    },
+    {
+      "epoch": 2.0084251458198317,
+      "grad_norm": 0.8509025573730469,
+      "learning_rate": 1.3008309078847605e-06,
+      "loss": 0.0696,
+      "step": 12396
+    },
+    {
+      "epoch": 2.008587167854828,
+      "grad_norm": 0.8189096450805664,
+      "learning_rate": 1.3004472136365609e-06,
+      "loss": 0.0658,
+      "step": 12397
+    },
+    {
+      "epoch": 2.008749189889825,
+      "grad_norm": 0.7005947828292847,
+      "learning_rate": 1.3000635560917735e-06,
+      "loss": 0.0585,
+      "step": 12398
+    },
+    {
+      "epoch": 2.0089112119248216,
+      "grad_norm": 0.749654233455658,
+      "learning_rate": 1.2996799352621372e-06,
+      "loss": 0.0564,
+      "step": 12399
+    },
+    {
+      "epoch": 2.0090732339598185,
+      "grad_norm": 0.8583326935768127,
+      "learning_rate": 1.2992963511593904e-06,
+      "loss": 0.0669,
+      "step": 12400
+    },
+    {
+      "epoch": 2.0092352559948155,
+      "grad_norm": 0.834697425365448,
+      "learning_rate": 1.2989128037952698e-06,
+      "loss": 0.0709,
+      "step": 12401
+    },
+    {
+      "epoch": 2.009397278029812,
+      "grad_norm": 0.7957351207733154,
+      "learning_rate": 1.2985292931815105e-06,
+      "loss": 0.063,
+      "step": 12402
+    },
+    {
+      "epoch": 2.009559300064809,
+      "grad_norm": 0.7907761335372925,
+      "learning_rate": 1.2981458193298473e-06,
+      "loss": 0.0607,
+      "step": 12403
+    },
+    {
+      "epoch": 2.0097213220998054,
+      "grad_norm": 0.813335120677948,
+      "learning_rate": 1.2977623822520141e-06,
+      "loss": 0.0622,
+      "step": 12404
+    },
+    {
+      "epoch": 2.0098833441348023,
+      "grad_norm": 0.9492120146751404,
+      "learning_rate": 1.2973789819597431e-06,
+      "loss": 0.0669,
+      "step": 12405
+    },
+    {
+      "epoch": 2.0100453661697992,
+      "grad_norm": 0.8153406381607056,
+      "learning_rate": 1.296995618464763e-06,
+      "loss": 0.0651,
+      "step": 12406
+    },
+    {
+      "epoch": 2.0102073882047957,
+      "grad_norm": 0.7369259595870972,
+      "learning_rate": 1.296612291778807e-06,
+      "loss": 0.058,
+      "step": 12407
+    },
+    {
+      "epoch": 2.0103694102397927,
+      "grad_norm": 0.7126782536506653,
+      "learning_rate": 1.2962290019136028e-06,
+      "loss": 0.0606,
+      "step": 12408
+    },
+    {
+      "epoch": 2.0105314322747896,
+      "grad_norm": 0.7159781455993652,
+      "learning_rate": 1.295845748880879e-06,
+      "loss": 0.0635,
+      "step": 12409
+    },
+    {
+      "epoch": 2.010693454309786,
+      "grad_norm": 0.7070374488830566,
+      "learning_rate": 1.2954625326923602e-06,
+      "loss": 0.0552,
+      "step": 12410
+    },
+    {
+      "epoch": 2.010855476344783,
+      "grad_norm": 0.7887537479400635,
+      "learning_rate": 1.2950793533597722e-06,
+      "loss": 0.0673,
+      "step": 12411
+    },
+    {
+      "epoch": 2.0110174983797795,
+      "grad_norm": 0.7388550639152527,
+      "learning_rate": 1.294696210894842e-06,
+      "loss": 0.0542,
+      "step": 12412
+    },
+    {
+      "epoch": 2.0111795204147764,
+      "grad_norm": 0.8547154068946838,
+      "learning_rate": 1.2943131053092895e-06,
+      "loss": 0.0693,
+      "step": 12413
+    },
+    {
+      "epoch": 2.0113415424497734,
+      "grad_norm": 0.7168740034103394,
+      "learning_rate": 1.2939300366148389e-06,
+      "loss": 0.0591,
+      "step": 12414
+    },
+    {
+      "epoch": 2.01150356448477,
+      "grad_norm": 0.717652440071106,
+      "learning_rate": 1.2935470048232102e-06,
+      "loss": 0.0568,
+      "step": 12415
+    },
+    {
+      "epoch": 2.011665586519767,
+      "grad_norm": 0.8575769066810608,
+      "learning_rate": 1.2931640099461237e-06,
+      "loss": 0.0617,
+      "step": 12416
+    },
+    {
+      "epoch": 2.0118276085547633,
+      "grad_norm": 0.862146258354187,
+      "learning_rate": 1.292781051995298e-06,
+      "loss": 0.0645,
+      "step": 12417
+    },
+    {
+      "epoch": 2.01198963058976,
+      "grad_norm": 0.8514499068260193,
+      "learning_rate": 1.2923981309824507e-06,
+      "loss": 0.0685,
+      "step": 12418
+    },
+    {
+      "epoch": 2.012151652624757,
+      "grad_norm": 0.8388655781745911,
+      "learning_rate": 1.292015246919298e-06,
+      "loss": 0.0724,
+      "step": 12419
+    },
+    {
+      "epoch": 2.0123136746597536,
+      "grad_norm": 0.8496232032775879,
+      "learning_rate": 1.291632399817557e-06,
+      "loss": 0.0593,
+      "step": 12420
+    },
+    {
+      "epoch": 2.0124756966947506,
+      "grad_norm": 0.7816452383995056,
+      "learning_rate": 1.2912495896889383e-06,
+      "loss": 0.0612,
+      "step": 12421
+    },
+    {
+      "epoch": 2.012637718729747,
+      "grad_norm": 0.808066189289093,
+      "learning_rate": 1.2908668165451577e-06,
+      "loss": 0.0599,
+      "step": 12422
+    },
+    {
+      "epoch": 2.012799740764744,
+      "grad_norm": 0.8434270024299622,
+      "learning_rate": 1.2904840803979276e-06,
+      "loss": 0.0663,
+      "step": 12423
+    },
+    {
+      "epoch": 2.012961762799741,
+      "grad_norm": 0.8326642513275146,
+      "learning_rate": 1.290101381258957e-06,
+      "loss": 0.0616,
+      "step": 12424
+    },
+    {
+      "epoch": 2.0131237848347374,
+      "grad_norm": 0.971601128578186,
+      "learning_rate": 1.2897187191399546e-06,
+      "loss": 0.0676,
+      "step": 12425
+    },
+    {
+      "epoch": 2.0132858068697344,
+      "grad_norm": 0.8341547250747681,
+      "learning_rate": 1.289336094052632e-06,
+      "loss": 0.0643,
+      "step": 12426
+    },
+    {
+      "epoch": 2.013447828904731,
+      "grad_norm": 0.917320728302002,
+      "learning_rate": 1.288953506008696e-06,
+      "loss": 0.0688,
+      "step": 12427
+    },
+    {
+      "epoch": 2.0136098509397278,
+      "grad_norm": 0.8390260934829712,
+      "learning_rate": 1.288570955019851e-06,
+      "loss": 0.0646,
+      "step": 12428
+    },
+    {
+      "epoch": 2.0137718729747247,
+      "grad_norm": 0.9914124608039856,
+      "learning_rate": 1.2881884410978034e-06,
+      "loss": 0.0705,
+      "step": 12429
+    },
+    {
+      "epoch": 2.013933895009721,
+      "grad_norm": 0.8766891956329346,
+      "learning_rate": 1.2878059642542566e-06,
+      "loss": 0.0725,
+      "step": 12430
+    },
+    {
+      "epoch": 2.014095917044718,
+      "grad_norm": 0.7540597915649414,
+      "learning_rate": 1.2874235245009143e-06,
+      "loss": 0.0608,
+      "step": 12431
+    },
+    {
+      "epoch": 2.0142579390797146,
+      "grad_norm": 0.8450373411178589,
+      "learning_rate": 1.2870411218494778e-06,
+      "loss": 0.0668,
+      "step": 12432
+    },
+    {
+      "epoch": 2.0144199611147116,
+      "grad_norm": 0.743740975856781,
+      "learning_rate": 1.2866587563116473e-06,
+      "loss": 0.058,
+      "step": 12433
+    },
+    {
+      "epoch": 2.0145819831497085,
+      "grad_norm": 0.7940791845321655,
+      "learning_rate": 1.2862764278991236e-06,
+      "loss": 0.0618,
+      "step": 12434
+    },
+    {
+      "epoch": 2.014744005184705,
+      "grad_norm": 0.8542442321777344,
+      "learning_rate": 1.2858941366236021e-06,
+      "loss": 0.0641,
+      "step": 12435
+    },
+    {
+      "epoch": 2.014906027219702,
+      "grad_norm": 0.8491990566253662,
+      "learning_rate": 1.2855118824967833e-06,
+      "loss": 0.0603,
+      "step": 12436
+    },
+    {
+      "epoch": 2.015068049254699,
+      "grad_norm": 1.0003999471664429,
+      "learning_rate": 1.2851296655303616e-06,
+      "loss": 0.064,
+      "step": 12437
+    },
+    {
+      "epoch": 2.0152300712896953,
+      "grad_norm": 0.8259167671203613,
+      "learning_rate": 1.2847474857360332e-06,
+      "loss": 0.0616,
+      "step": 12438
+    },
+    {
+      "epoch": 2.0153920933246923,
+      "grad_norm": 0.9689138531684875,
+      "learning_rate": 1.28436534312549e-06,
+      "loss": 0.0662,
+      "step": 12439
+    },
+    {
+      "epoch": 2.0155541153596888,
+      "grad_norm": 0.8320235013961792,
+      "learning_rate": 1.2839832377104245e-06,
+      "loss": 0.0579,
+      "step": 12440
+    },
+    {
+      "epoch": 2.0157161373946857,
+      "grad_norm": 0.8748543858528137,
+      "learning_rate": 1.283601169502531e-06,
+      "loss": 0.0762,
+      "step": 12441
+    },
+    {
+      "epoch": 2.0158781594296826,
+      "grad_norm": 0.920680820941925,
+      "learning_rate": 1.2832191385134972e-06,
+      "loss": 0.0721,
+      "step": 12442
+    },
+    {
+      "epoch": 2.016040181464679,
+      "grad_norm": 0.6789097785949707,
+      "learning_rate": 1.2828371447550133e-06,
+      "loss": 0.0546,
+      "step": 12443
+    },
+    {
+      "epoch": 2.016202203499676,
+      "grad_norm": 0.8041635155677795,
+      "learning_rate": 1.2824551882387664e-06,
+      "loss": 0.0615,
+      "step": 12444
+    },
+    {
+      "epoch": 2.0163642255346725,
+      "grad_norm": 0.8807735443115234,
+      "learning_rate": 1.2820732689764462e-06,
+      "loss": 0.0688,
+      "step": 12445
+    },
+    {
+      "epoch": 2.0165262475696695,
+      "grad_norm": 0.7847744822502136,
+      "learning_rate": 1.2816913869797353e-06,
+      "loss": 0.0639,
+      "step": 12446
+    },
+    {
+      "epoch": 2.0166882696046664,
+      "grad_norm": 0.9476509094238281,
+      "learning_rate": 1.2813095422603203e-06,
+      "loss": 0.0652,
+      "step": 12447
+    },
+    {
+      "epoch": 2.016850291639663,
+      "grad_norm": 0.7772141695022583,
+      "learning_rate": 1.2809277348298838e-06,
+      "loss": 0.0583,
+      "step": 12448
+    },
+    {
+      "epoch": 2.01701231367466,
+      "grad_norm": 0.8144988417625427,
+      "learning_rate": 1.2805459647001087e-06,
+      "loss": 0.0634,
+      "step": 12449
+    },
+    {
+      "epoch": 2.0171743357096563,
+      "grad_norm": 0.9111658334732056,
+      "learning_rate": 1.2801642318826759e-06,
+      "loss": 0.0599,
+      "step": 12450
+    },
+    {
+      "epoch": 2.0173363577446533,
+      "grad_norm": 0.9419159889221191,
+      "learning_rate": 1.2797825363892658e-06,
+      "loss": 0.0759,
+      "step": 12451
+    },
+    {
+      "epoch": 2.01749837977965,
+      "grad_norm": 0.9087123870849609,
+      "learning_rate": 1.279400878231557e-06,
+      "loss": 0.0676,
+      "step": 12452
+    },
+    {
+      "epoch": 2.0176604018146467,
+      "grad_norm": 0.7883154153823853,
+      "learning_rate": 1.2790192574212287e-06,
+      "loss": 0.0585,
+      "step": 12453
+    },
+    {
+      "epoch": 2.0178224238496436,
+      "grad_norm": 0.7743737101554871,
+      "learning_rate": 1.2786376739699547e-06,
+      "loss": 0.0643,
+      "step": 12454
+    },
+    {
+      "epoch": 2.01798444588464,
+      "grad_norm": 0.8277117013931274,
+      "learning_rate": 1.2782561278894126e-06,
+      "loss": 0.069,
+      "step": 12455
+    },
+    {
+      "epoch": 2.018146467919637,
+      "grad_norm": 0.7758954763412476,
+      "learning_rate": 1.2778746191912778e-06,
+      "loss": 0.0629,
+      "step": 12456
+    },
+    {
+      "epoch": 2.018308489954634,
+      "grad_norm": 0.7998802065849304,
+      "learning_rate": 1.277493147887221e-06,
+      "loss": 0.0572,
+      "step": 12457
+    },
+    {
+      "epoch": 2.0184705119896305,
+      "grad_norm": 0.7002156972885132,
+      "learning_rate": 1.2771117139889155e-06,
+      "loss": 0.0612,
+      "step": 12458
+    },
+    {
+      "epoch": 2.0186325340246274,
+      "grad_norm": 0.7348694801330566,
+      "learning_rate": 1.2767303175080325e-06,
+      "loss": 0.0593,
+      "step": 12459
+    },
+    {
+      "epoch": 2.0187945560596243,
+      "grad_norm": 0.7863420248031616,
+      "learning_rate": 1.276348958456241e-06,
+      "loss": 0.0638,
+      "step": 12460
+    },
+    {
+      "epoch": 2.018956578094621,
+      "grad_norm": 0.8472959995269775,
+      "learning_rate": 1.2759676368452106e-06,
+      "loss": 0.0664,
+      "step": 12461
+    },
+    {
+      "epoch": 2.0191186001296177,
+      "grad_norm": 0.7812498807907104,
+      "learning_rate": 1.2755863526866087e-06,
+      "loss": 0.0568,
+      "step": 12462
+    },
+    {
+      "epoch": 2.0192806221646142,
+      "grad_norm": 0.85575932264328,
+      "learning_rate": 1.2752051059921005e-06,
+      "loss": 0.0662,
+      "step": 12463
+    },
+    {
+      "epoch": 2.019442644199611,
+      "grad_norm": 0.9001664519309998,
+      "learning_rate": 1.2748238967733529e-06,
+      "loss": 0.0682,
+      "step": 12464
+    },
+    {
+      "epoch": 2.019604666234608,
+      "grad_norm": 0.948502779006958,
+      "learning_rate": 1.2744427250420288e-06,
+      "loss": 0.0695,
+      "step": 12465
+    },
+    {
+      "epoch": 2.0197666882696046,
+      "grad_norm": 0.8993748426437378,
+      "learning_rate": 1.2740615908097915e-06,
+      "loss": 0.0641,
+      "step": 12466
+    },
+    {
+      "epoch": 2.0199287103046015,
+      "grad_norm": 0.8371912837028503,
+      "learning_rate": 1.273680494088304e-06,
+      "loss": 0.064,
+      "step": 12467
+    },
+    {
+      "epoch": 2.020090732339598,
+      "grad_norm": 1.5723109245300293,
+      "learning_rate": 1.2732994348892237e-06,
+      "loss": 0.0644,
+      "step": 12468
+    },
+    {
+      "epoch": 2.020252754374595,
+      "grad_norm": 0.9138685464859009,
+      "learning_rate": 1.2729184132242131e-06,
+      "loss": 0.0678,
+      "step": 12469
+    },
+    {
+      "epoch": 2.020414776409592,
+      "grad_norm": 0.8680717349052429,
+      "learning_rate": 1.2725374291049296e-06,
+      "loss": 0.0615,
+      "step": 12470
+    },
+    {
+      "epoch": 2.0205767984445884,
+      "grad_norm": 1.0072784423828125,
+      "learning_rate": 1.2721564825430313e-06,
+      "loss": 0.0675,
+      "step": 12471
+    },
+    {
+      "epoch": 2.0207388204795853,
+      "grad_norm": 0.8381701707839966,
+      "learning_rate": 1.2717755735501725e-06,
+      "loss": 0.0593,
+      "step": 12472
+    },
+    {
+      "epoch": 2.020900842514582,
+      "grad_norm": 0.8638261556625366,
+      "learning_rate": 1.2713947021380078e-06,
+      "loss": 0.0674,
+      "step": 12473
+    },
+    {
+      "epoch": 2.0210628645495787,
+      "grad_norm": 0.8350870609283447,
+      "learning_rate": 1.2710138683181937e-06,
+      "loss": 0.0686,
+      "step": 12474
+    },
+    {
+      "epoch": 2.0212248865845757,
+      "grad_norm": 0.8235458731651306,
+      "learning_rate": 1.2706330721023807e-06,
+      "loss": 0.0662,
+      "step": 12475
+    },
+    {
+      "epoch": 2.021386908619572,
+      "grad_norm": 0.7439950704574585,
+      "learning_rate": 1.2702523135022205e-06,
+      "loss": 0.0597,
+      "step": 12476
+    },
+    {
+      "epoch": 2.021548930654569,
+      "grad_norm": 0.7728898525238037,
+      "learning_rate": 1.2698715925293634e-06,
+      "loss": 0.0563,
+      "step": 12477
+    },
+    {
+      "epoch": 2.0217109526895656,
+      "grad_norm": 0.7948071956634521,
+      "learning_rate": 1.2694909091954588e-06,
+      "loss": 0.0595,
+      "step": 12478
+    },
+    {
+      "epoch": 2.0218729747245625,
+      "grad_norm": 0.8696475625038147,
+      "learning_rate": 1.2691102635121544e-06,
+      "loss": 0.0629,
+      "step": 12479
+    },
+    {
+      "epoch": 2.0220349967595594,
+      "grad_norm": 0.869878888130188,
+      "learning_rate": 1.2687296554910978e-06,
+      "loss": 0.0731,
+      "step": 12480
+    },
+    {
+      "epoch": 2.022197018794556,
+      "grad_norm": 0.812859833240509,
+      "learning_rate": 1.2683490851439334e-06,
+      "loss": 0.0624,
+      "step": 12481
+    },
+    {
+      "epoch": 2.022359040829553,
+      "grad_norm": 0.8346046209335327,
+      "learning_rate": 1.2679685524823082e-06,
+      "loss": 0.0628,
+      "step": 12482
+    },
+    {
+      "epoch": 2.0225210628645494,
+      "grad_norm": 0.808914065361023,
+      "learning_rate": 1.2675880575178613e-06,
+      "loss": 0.0587,
+      "step": 12483
+    },
+    {
+      "epoch": 2.0226830848995463,
+      "grad_norm": 0.7890689969062805,
+      "learning_rate": 1.2672076002622386e-06,
+      "loss": 0.0623,
+      "step": 12484
+    },
+    {
+      "epoch": 2.0228451069345432,
+      "grad_norm": 0.7109611630439758,
+      "learning_rate": 1.266827180727081e-06,
+      "loss": 0.053,
+      "step": 12485
+    },
+    {
+      "epoch": 2.0230071289695397,
+      "grad_norm": 0.9208827018737793,
+      "learning_rate": 1.2664467989240265e-06,
+      "loss": 0.0608,
+      "step": 12486
+    },
+    {
+      "epoch": 2.0231691510045366,
+      "grad_norm": 1.064787745475769,
+      "learning_rate": 1.266066454864715e-06,
+      "loss": 0.0668,
+      "step": 12487
+    },
+    {
+      "epoch": 2.0233311730395336,
+      "grad_norm": 0.8241201639175415,
+      "learning_rate": 1.2656861485607828e-06,
+      "loss": 0.06,
+      "step": 12488
+    },
+    {
+      "epoch": 2.02349319507453,
+      "grad_norm": 0.8332204222679138,
+      "learning_rate": 1.2653058800238693e-06,
+      "loss": 0.0669,
+      "step": 12489
+    },
+    {
+      "epoch": 2.023655217109527,
+      "grad_norm": 0.8376948833465576,
+      "learning_rate": 1.264925649265607e-06,
+      "loss": 0.0604,
+      "step": 12490
+    },
+    {
+      "epoch": 2.0238172391445235,
+      "grad_norm": 0.971579372882843,
+      "learning_rate": 1.2645454562976311e-06,
+      "loss": 0.0593,
+      "step": 12491
+    },
+    {
+      "epoch": 2.0239792611795204,
+      "grad_norm": 1.004150390625,
+      "learning_rate": 1.2641653011315746e-06,
+      "loss": 0.0687,
+      "step": 12492
+    },
+    {
+      "epoch": 2.0241412832145174,
+      "grad_norm": 0.7998213768005371,
+      "learning_rate": 1.2637851837790694e-06,
+      "loss": 0.0622,
+      "step": 12493
+    },
+    {
+      "epoch": 2.024303305249514,
+      "grad_norm": 1.0076199769973755,
+      "learning_rate": 1.2634051042517453e-06,
+      "loss": 0.064,
+      "step": 12494
+    },
+    {
+      "epoch": 2.024465327284511,
+      "grad_norm": 0.8546391725540161,
+      "learning_rate": 1.2630250625612331e-06,
+      "loss": 0.0648,
+      "step": 12495
+    },
+    {
+      "epoch": 2.0246273493195073,
+      "grad_norm": 0.9467459917068481,
+      "learning_rate": 1.2626450587191602e-06,
+      "loss": 0.0631,
+      "step": 12496
+    },
+    {
+      "epoch": 2.024789371354504,
+      "grad_norm": 0.8554684519767761,
+      "learning_rate": 1.2622650927371543e-06,
+      "loss": 0.0655,
+      "step": 12497
+    },
+    {
+      "epoch": 2.024951393389501,
+      "grad_norm": 0.8624327182769775,
+      "learning_rate": 1.2618851646268416e-06,
+      "loss": 0.0617,
+      "step": 12498
+    },
+    {
+      "epoch": 2.0251134154244976,
+      "grad_norm": 0.9098758101463318,
+      "learning_rate": 1.2615052743998463e-06,
+      "loss": 0.066,
+      "step": 12499
+    },
+    {
+      "epoch": 2.0252754374594946,
+      "grad_norm": 0.8350497484207153,
+      "learning_rate": 1.2611254220677937e-06,
+      "loss": 0.064,
+      "step": 12500
+    },
+    {
+      "epoch": 2.025437459494491,
+      "grad_norm": 0.877099871635437,
+      "learning_rate": 1.260745607642304e-06,
+      "loss": 0.0547,
+      "step": 12501
+    },
+    {
+      "epoch": 2.025599481529488,
+      "grad_norm": 0.7435240149497986,
+      "learning_rate": 1.260365831134999e-06,
+      "loss": 0.052,
+      "step": 12502
+    },
+    {
+      "epoch": 2.025761503564485,
+      "grad_norm": 0.7691801190376282,
+      "learning_rate": 1.2599860925575014e-06,
+      "loss": 0.0535,
+      "step": 12503
+    },
+    {
+      "epoch": 2.0259235255994814,
+      "grad_norm": 0.7755981087684631,
+      "learning_rate": 1.259606391921428e-06,
+      "loss": 0.0574,
+      "step": 12504
+    },
+    {
+      "epoch": 2.0260855476344783,
+      "grad_norm": 1.0379624366760254,
+      "learning_rate": 1.259226729238397e-06,
+      "loss": 0.0631,
+      "step": 12505
+    },
+    {
+      "epoch": 2.026247569669475,
+      "grad_norm": 0.8337475657463074,
+      "learning_rate": 1.2588471045200256e-06,
+      "loss": 0.0557,
+      "step": 12506
+    },
+    {
+      "epoch": 2.0264095917044718,
+      "grad_norm": 0.8900526165962219,
+      "learning_rate": 1.2584675177779294e-06,
+      "loss": 0.0599,
+      "step": 12507
+    },
+    {
+      "epoch": 2.0265716137394687,
+      "grad_norm": 0.7923566102981567,
+      "learning_rate": 1.2580879690237224e-06,
+      "loss": 0.0552,
+      "step": 12508
+    },
+    {
+      "epoch": 2.026733635774465,
+      "grad_norm": 0.8455385565757751,
+      "learning_rate": 1.257708458269018e-06,
+      "loss": 0.0609,
+      "step": 12509
+    },
+    {
+      "epoch": 2.026895657809462,
+      "grad_norm": 0.8646577000617981,
+      "learning_rate": 1.257328985525429e-06,
+      "loss": 0.068,
+      "step": 12510
+    },
+    {
+      "epoch": 2.027057679844459,
+      "grad_norm": 0.8488696217536926,
+      "learning_rate": 1.2569495508045656e-06,
+      "loss": 0.0622,
+      "step": 12511
+    },
+    {
+      "epoch": 2.0272197018794555,
+      "grad_norm": 0.9858669638633728,
+      "learning_rate": 1.256570154118038e-06,
+      "loss": 0.0687,
+      "step": 12512
+    },
+    {
+      "epoch": 2.0273817239144525,
+      "grad_norm": 0.9625844359397888,
+      "learning_rate": 1.2561907954774544e-06,
+      "loss": 0.0651,
+      "step": 12513
+    },
+    {
+      "epoch": 2.027543745949449,
+      "grad_norm": 0.9064575433731079,
+      "learning_rate": 1.2558114748944226e-06,
+      "loss": 0.0618,
+      "step": 12514
+    },
+    {
+      "epoch": 2.027705767984446,
+      "grad_norm": 0.8604201674461365,
+      "learning_rate": 1.2554321923805496e-06,
+      "loss": 0.0648,
+      "step": 12515
+    },
+    {
+      "epoch": 2.027867790019443,
+      "grad_norm": 1.1641322374343872,
+      "learning_rate": 1.2550529479474383e-06,
+      "loss": 0.0645,
+      "step": 12516
+    },
+    {
+      "epoch": 2.0280298120544393,
+      "grad_norm": 0.8054379224777222,
+      "learning_rate": 1.2546737416066945e-06,
+      "loss": 0.0605,
+      "step": 12517
+    },
+    {
+      "epoch": 2.0281918340894363,
+      "grad_norm": 0.9893051385879517,
+      "learning_rate": 1.2542945733699216e-06,
+      "loss": 0.0739,
+      "step": 12518
+    },
+    {
+      "epoch": 2.0283538561244328,
+      "grad_norm": 0.8926251530647278,
+      "learning_rate": 1.2539154432487193e-06,
+      "loss": 0.0588,
+      "step": 12519
+    },
+    {
+      "epoch": 2.0285158781594297,
+      "grad_norm": 0.8101766705513,
+      "learning_rate": 1.2535363512546892e-06,
+      "loss": 0.0602,
+      "step": 12520
+    },
+    {
+      "epoch": 2.0286779001944266,
+      "grad_norm": 0.9098559021949768,
+      "learning_rate": 1.2531572973994293e-06,
+      "loss": 0.0676,
+      "step": 12521
+    },
+    {
+      "epoch": 2.028839922229423,
+      "grad_norm": 1.1210435628890991,
+      "learning_rate": 1.2527782816945405e-06,
+      "loss": 0.0665,
+      "step": 12522
+    },
+    {
+      "epoch": 2.02900194426442,
+      "grad_norm": 0.73442542552948,
+      "learning_rate": 1.2523993041516175e-06,
+      "loss": 0.045,
+      "step": 12523
+    },
+    {
+      "epoch": 2.0291639662994165,
+      "grad_norm": 0.8347865343093872,
+      "learning_rate": 1.2520203647822563e-06,
+      "loss": 0.0643,
+      "step": 12524
+    },
+    {
+      "epoch": 2.0293259883344135,
+      "grad_norm": 0.9335924983024597,
+      "learning_rate": 1.2516414635980518e-06,
+      "loss": 0.0657,
+      "step": 12525
+    },
+    {
+      "epoch": 2.0294880103694104,
+      "grad_norm": 0.845848023891449,
+      "learning_rate": 1.2512626006105977e-06,
+      "loss": 0.0618,
+      "step": 12526
+    },
+    {
+      "epoch": 2.029650032404407,
+      "grad_norm": 0.8612715601921082,
+      "learning_rate": 1.2508837758314862e-06,
+      "loss": 0.0725,
+      "step": 12527
+    },
+    {
+      "epoch": 2.029812054439404,
+      "grad_norm": 0.936465322971344,
+      "learning_rate": 1.2505049892723083e-06,
+      "loss": 0.0711,
+      "step": 12528
+    },
+    {
+      "epoch": 2.0299740764744003,
+      "grad_norm": 0.8308136463165283,
+      "learning_rate": 1.2501262409446552e-06,
+      "loss": 0.0601,
+      "step": 12529
+    },
+    {
+      "epoch": 2.0301360985093972,
+      "grad_norm": 0.8098605871200562,
+      "learning_rate": 1.2497475308601134e-06,
+      "loss": 0.0643,
+      "step": 12530
+    },
+    {
+      "epoch": 2.030298120544394,
+      "grad_norm": 0.932350754737854,
+      "learning_rate": 1.2493688590302705e-06,
+      "loss": 0.0697,
+      "step": 12531
+    },
+    {
+      "epoch": 2.0304601425793907,
+      "grad_norm": 0.8417340517044067,
+      "learning_rate": 1.248990225466715e-06,
+      "loss": 0.0661,
+      "step": 12532
+    },
+    {
+      "epoch": 2.0306221646143876,
+      "grad_norm": 0.7914611101150513,
+      "learning_rate": 1.2486116301810322e-06,
+      "loss": 0.0647,
+      "step": 12533
+    },
+    {
+      "epoch": 2.0307841866493845,
+      "grad_norm": 1.3596220016479492,
+      "learning_rate": 1.2482330731848044e-06,
+      "loss": 0.0616,
+      "step": 12534
+    },
+    {
+      "epoch": 2.030946208684381,
+      "grad_norm": 1.0075905323028564,
+      "learning_rate": 1.2478545544896143e-06,
+      "loss": 0.0648,
+      "step": 12535
+    },
+    {
+      "epoch": 2.031108230719378,
+      "grad_norm": 0.8666545748710632,
+      "learning_rate": 1.2474760741070465e-06,
+      "loss": 0.0621,
+      "step": 12536
+    },
+    {
+      "epoch": 2.0312702527543745,
+      "grad_norm": 0.8424669504165649,
+      "learning_rate": 1.2470976320486792e-06,
+      "loss": 0.0616,
+      "step": 12537
+    },
+    {
+      "epoch": 2.0314322747893714,
+      "grad_norm": 0.858910083770752,
+      "learning_rate": 1.246719228326092e-06,
+      "loss": 0.0631,
+      "step": 12538
+    },
+    {
+      "epoch": 2.0315942968243683,
+      "grad_norm": 0.9575031995773315,
+      "learning_rate": 1.2463408629508635e-06,
+      "loss": 0.0641,
+      "step": 12539
+    },
+    {
+      "epoch": 2.031756318859365,
+      "grad_norm": 1.0462077856063843,
+      "learning_rate": 1.2459625359345712e-06,
+      "loss": 0.0718,
+      "step": 12540
+    },
+    {
+      "epoch": 2.0319183408943617,
+      "grad_norm": 0.9476367831230164,
+      "learning_rate": 1.2455842472887903e-06,
+      "loss": 0.0599,
+      "step": 12541
+    },
+    {
+      "epoch": 2.0320803629293582,
+      "grad_norm": 1.0100579261779785,
+      "learning_rate": 1.2452059970250957e-06,
+      "loss": 0.0597,
+      "step": 12542
+    },
+    {
+      "epoch": 2.032242384964355,
+      "grad_norm": 0.8134729266166687,
+      "learning_rate": 1.2448277851550613e-06,
+      "loss": 0.0586,
+      "step": 12543
+    },
+    {
+      "epoch": 2.032404406999352,
+      "grad_norm": 0.9285221695899963,
+      "learning_rate": 1.2444496116902602e-06,
+      "loss": 0.0693,
+      "step": 12544
+    },
+    {
+      "epoch": 2.0325664290343486,
+      "grad_norm": 1.1232776641845703,
+      "learning_rate": 1.2440714766422604e-06,
+      "loss": 0.0791,
+      "step": 12545
+    },
+    {
+      "epoch": 2.0327284510693455,
+      "grad_norm": 0.7684057950973511,
+      "learning_rate": 1.2436933800226352e-06,
+      "loss": 0.0565,
+      "step": 12546
+    },
+    {
+      "epoch": 2.032890473104342,
+      "grad_norm": 0.7600722908973694,
+      "learning_rate": 1.2433153218429526e-06,
+      "loss": 0.0612,
+      "step": 12547
+    },
+    {
+      "epoch": 2.033052495139339,
+      "grad_norm": 0.8198429942131042,
+      "learning_rate": 1.2429373021147808e-06,
+      "loss": 0.0616,
+      "step": 12548
+    },
+    {
+      "epoch": 2.033214517174336,
+      "grad_norm": 0.9691295027732849,
+      "learning_rate": 1.2425593208496844e-06,
+      "loss": 0.0707,
+      "step": 12549
+    },
+    {
+      "epoch": 2.0333765392093324,
+      "grad_norm": 0.7660728693008423,
+      "learning_rate": 1.2421813780592294e-06,
+      "loss": 0.0643,
+      "step": 12550
+    },
+    {
+      "epoch": 2.0335385612443293,
+      "grad_norm": 0.9578198194503784,
+      "learning_rate": 1.2418034737549818e-06,
+      "loss": 0.0708,
+      "step": 12551
+    },
+    {
+      "epoch": 2.033700583279326,
+      "grad_norm": 0.7433199882507324,
+      "learning_rate": 1.2414256079485021e-06,
+      "loss": 0.0557,
+      "step": 12552
+    },
+    {
+      "epoch": 2.0338626053143227,
+      "grad_norm": 0.862591028213501,
+      "learning_rate": 1.2410477806513535e-06,
+      "loss": 0.0666,
+      "step": 12553
+    },
+    {
+      "epoch": 2.0340246273493197,
+      "grad_norm": 0.8760819435119629,
+      "learning_rate": 1.240669991875096e-06,
+      "loss": 0.0661,
+      "step": 12554
+    },
+    {
+      "epoch": 2.034186649384316,
+      "grad_norm": 0.8833930492401123,
+      "learning_rate": 1.2402922416312891e-06,
+      "loss": 0.0626,
+      "step": 12555
+    },
+    {
+      "epoch": 2.034348671419313,
+      "grad_norm": 0.8621806502342224,
+      "learning_rate": 1.2399145299314913e-06,
+      "loss": 0.0646,
+      "step": 12556
+    },
+    {
+      "epoch": 2.0345106934543096,
+      "grad_norm": 0.8710420727729797,
+      "learning_rate": 1.2395368567872596e-06,
+      "loss": 0.0686,
+      "step": 12557
+    },
+    {
+      "epoch": 2.0346727154893065,
+      "grad_norm": 0.935194194316864,
+      "learning_rate": 1.2391592222101497e-06,
+      "loss": 0.0659,
+      "step": 12558
+    },
+    {
+      "epoch": 2.0348347375243034,
+      "grad_norm": 0.9124763011932373,
+      "learning_rate": 1.2387816262117167e-06,
+      "loss": 0.0611,
+      "step": 12559
+    },
+    {
+      "epoch": 2.0349967595593,
+      "grad_norm": 0.861625075340271,
+      "learning_rate": 1.2384040688035135e-06,
+      "loss": 0.0576,
+      "step": 12560
+    },
+    {
+      "epoch": 2.035158781594297,
+      "grad_norm": 0.9553015232086182,
+      "learning_rate": 1.2380265499970932e-06,
+      "loss": 0.0698,
+      "step": 12561
+    },
+    {
+      "epoch": 2.035320803629294,
+      "grad_norm": 0.7892826795578003,
+      "learning_rate": 1.2376490698040069e-06,
+      "loss": 0.0561,
+      "step": 12562
+    },
+    {
+      "epoch": 2.0354828256642903,
+      "grad_norm": 0.7279547452926636,
+      "learning_rate": 1.2372716282358038e-06,
+      "loss": 0.0527,
+      "step": 12563
+    },
+    {
+      "epoch": 2.035644847699287,
+      "grad_norm": 0.7964885234832764,
+      "learning_rate": 1.236894225304032e-06,
+      "loss": 0.0593,
+      "step": 12564
+    },
+    {
+      "epoch": 2.0358068697342837,
+      "grad_norm": 0.7441739439964294,
+      "learning_rate": 1.2365168610202411e-06,
+      "loss": 0.0557,
+      "step": 12565
+    },
+    {
+      "epoch": 2.0359688917692806,
+      "grad_norm": 1.0166600942611694,
+      "learning_rate": 1.2361395353959776e-06,
+      "loss": 0.064,
+      "step": 12566
+    },
+    {
+      "epoch": 2.0361309138042776,
+      "grad_norm": 0.8261235356330872,
+      "learning_rate": 1.2357622484427854e-06,
+      "loss": 0.0591,
+      "step": 12567
+    },
+    {
+      "epoch": 2.036292935839274,
+      "grad_norm": 0.8217520117759705,
+      "learning_rate": 1.2353850001722084e-06,
+      "loss": 0.06,
+      "step": 12568
+    },
+    {
+      "epoch": 2.036454957874271,
+      "grad_norm": 0.887876570224762,
+      "learning_rate": 1.2350077905957902e-06,
+      "loss": 0.0678,
+      "step": 12569
+    },
+    {
+      "epoch": 2.0366169799092675,
+      "grad_norm": 0.8636623024940491,
+      "learning_rate": 1.2346306197250727e-06,
+      "loss": 0.0623,
+      "step": 12570
+    },
+    {
+      "epoch": 2.0367790019442644,
+      "grad_norm": 0.9496206045150757,
+      "learning_rate": 1.2342534875715958e-06,
+      "loss": 0.0559,
+      "step": 12571
+    },
+    {
+      "epoch": 2.0369410239792614,
+      "grad_norm": 0.827489972114563,
+      "learning_rate": 1.2338763941468993e-06,
+      "loss": 0.0609,
+      "step": 12572
+    },
+    {
+      "epoch": 2.037103046014258,
+      "grad_norm": 0.8658286929130554,
+      "learning_rate": 1.2334993394625219e-06,
+      "loss": 0.0712,
+      "step": 12573
+    },
+    {
+      "epoch": 2.037265068049255,
+      "grad_norm": 0.8474152088165283,
+      "learning_rate": 1.2331223235299983e-06,
+      "loss": 0.0627,
+      "step": 12574
+    },
+    {
+      "epoch": 2.0374270900842513,
+      "grad_norm": 0.8493207693099976,
+      "learning_rate": 1.2327453463608663e-06,
+      "loss": 0.0655,
+      "step": 12575
+    },
+    {
+      "epoch": 2.037589112119248,
+      "grad_norm": 0.9206482172012329,
+      "learning_rate": 1.2323684079666604e-06,
+      "loss": 0.0604,
+      "step": 12576
+    },
+    {
+      "epoch": 2.037751134154245,
+      "grad_norm": 1.0142006874084473,
+      "learning_rate": 1.2319915083589143e-06,
+      "loss": 0.0613,
+      "step": 12577
+    },
+    {
+      "epoch": 2.0379131561892416,
+      "grad_norm": 0.7615647315979004,
+      "learning_rate": 1.2316146475491578e-06,
+      "loss": 0.059,
+      "step": 12578
+    },
+    {
+      "epoch": 2.0380751782242386,
+      "grad_norm": 0.7730655670166016,
+      "learning_rate": 1.2312378255489246e-06,
+      "loss": 0.0621,
+      "step": 12579
+    },
+    {
+      "epoch": 2.038237200259235,
+      "grad_norm": 0.788704514503479,
+      "learning_rate": 1.2308610423697446e-06,
+      "loss": 0.0553,
+      "step": 12580
+    },
+    {
+      "epoch": 2.038399222294232,
+      "grad_norm": 0.8466129302978516,
+      "learning_rate": 1.2304842980231442e-06,
+      "loss": 0.0635,
+      "step": 12581
+    },
+    {
+      "epoch": 2.038561244329229,
+      "grad_norm": 0.8458765745162964,
+      "learning_rate": 1.2301075925206524e-06,
+      "loss": 0.0605,
+      "step": 12582
+    },
+    {
+      "epoch": 2.0387232663642254,
+      "grad_norm": 0.8162645101547241,
+      "learning_rate": 1.229730925873794e-06,
+      "loss": 0.0538,
+      "step": 12583
+    },
+    {
+      "epoch": 2.0388852883992223,
+      "grad_norm": 0.9227370023727417,
+      "learning_rate": 1.2293542980940974e-06,
+      "loss": 0.0655,
+      "step": 12584
+    },
+    {
+      "epoch": 2.039047310434219,
+      "grad_norm": 0.759735643863678,
+      "learning_rate": 1.2289777091930832e-06,
+      "loss": 0.0537,
+      "step": 12585
+    },
+    {
+      "epoch": 2.0392093324692158,
+      "grad_norm": 0.8518356680870056,
+      "learning_rate": 1.2286011591822756e-06,
+      "loss": 0.0646,
+      "step": 12586
+    },
+    {
+      "epoch": 2.0393713545042127,
+      "grad_norm": 0.9240495562553406,
+      "learning_rate": 1.2282246480731955e-06,
+      "loss": 0.0714,
+      "step": 12587
+    },
+    {
+      "epoch": 2.039533376539209,
+      "grad_norm": 0.8255005478858948,
+      "learning_rate": 1.2278481758773636e-06,
+      "loss": 0.0656,
+      "step": 12588
+    },
+    {
+      "epoch": 2.039695398574206,
+      "grad_norm": 0.8890777826309204,
+      "learning_rate": 1.227471742606299e-06,
+      "loss": 0.0655,
+      "step": 12589
+    },
+    {
+      "epoch": 2.039857420609203,
+      "grad_norm": 0.9725179076194763,
+      "learning_rate": 1.2270953482715197e-06,
+      "loss": 0.0681,
+      "step": 12590
+    },
+    {
+      "epoch": 2.0400194426441995,
+      "grad_norm": 0.9976947903633118,
+      "learning_rate": 1.2267189928845424e-06,
+      "loss": 0.0639,
+      "step": 12591
+    },
+    {
+      "epoch": 2.0401814646791965,
+      "grad_norm": 0.8435698747634888,
+      "learning_rate": 1.2263426764568835e-06,
+      "loss": 0.0615,
+      "step": 12592
+    },
+    {
+      "epoch": 2.040343486714193,
+      "grad_norm": 1.729702353477478,
+      "learning_rate": 1.2259663990000544e-06,
+      "loss": 0.0608,
+      "step": 12593
+    },
+    {
+      "epoch": 2.04050550874919,
+      "grad_norm": 0.8571469783782959,
+      "learning_rate": 1.2255901605255715e-06,
+      "loss": 0.0673,
+      "step": 12594
+    },
+    {
+      "epoch": 2.040667530784187,
+      "grad_norm": 1.0115898847579956,
+      "learning_rate": 1.2252139610449468e-06,
+      "loss": 0.0618,
+      "step": 12595
+    },
+    {
+      "epoch": 2.0408295528191833,
+      "grad_norm": 0.7657039165496826,
+      "learning_rate": 1.224837800569689e-06,
+      "loss": 0.0583,
+      "step": 12596
+    },
+    {
+      "epoch": 2.0409915748541803,
+      "grad_norm": 0.8274670243263245,
+      "learning_rate": 1.2244616791113085e-06,
+      "loss": 0.0617,
+      "step": 12597
+    },
+    {
+      "epoch": 2.0411535968891767,
+      "grad_norm": 0.8327546119689941,
+      "learning_rate": 1.224085596681314e-06,
+      "loss": 0.0532,
+      "step": 12598
+    },
+    {
+      "epoch": 2.0413156189241737,
+      "grad_norm": 0.8720483779907227,
+      "learning_rate": 1.2237095532912125e-06,
+      "loss": 0.0622,
+      "step": 12599
+    },
+    {
+      "epoch": 2.0414776409591706,
+      "grad_norm": 0.9215652942657471,
+      "learning_rate": 1.22333354895251e-06,
+      "loss": 0.0658,
+      "step": 12600
+    },
+    {
+      "epoch": 2.041639662994167,
+      "grad_norm": 0.792291522026062,
+      "learning_rate": 1.2229575836767115e-06,
+      "loss": 0.0579,
+      "step": 12601
+    },
+    {
+      "epoch": 2.041801685029164,
+      "grad_norm": 0.8133555054664612,
+      "learning_rate": 1.2225816574753208e-06,
+      "loss": 0.0622,
+      "step": 12602
+    },
+    {
+      "epoch": 2.0419637070641605,
+      "grad_norm": 0.7037177681922913,
+      "learning_rate": 1.2222057703598398e-06,
+      "loss": 0.0573,
+      "step": 12603
+    },
+    {
+      "epoch": 2.0421257290991575,
+      "grad_norm": 0.9011344909667969,
+      "learning_rate": 1.2218299223417702e-06,
+      "loss": 0.0612,
+      "step": 12604
+    },
+    {
+      "epoch": 2.0422877511341544,
+      "grad_norm": 0.8436145782470703,
+      "learning_rate": 1.2214541134326117e-06,
+      "loss": 0.0604,
+      "step": 12605
+    },
+    {
+      "epoch": 2.042449773169151,
+      "grad_norm": 0.7856259346008301,
+      "learning_rate": 1.2210783436438644e-06,
+      "loss": 0.0538,
+      "step": 12606
+    },
+    {
+      "epoch": 2.042611795204148,
+      "grad_norm": 1.0986276865005493,
+      "learning_rate": 1.2207026129870229e-06,
+      "loss": 0.0682,
+      "step": 12607
+    },
+    {
+      "epoch": 2.0427738172391443,
+      "grad_norm": 0.8682820796966553,
+      "learning_rate": 1.2203269214735866e-06,
+      "loss": 0.0579,
+      "step": 12608
+    },
+    {
+      "epoch": 2.0429358392741412,
+      "grad_norm": 0.8327391147613525,
+      "learning_rate": 1.2199512691150496e-06,
+      "loss": 0.0624,
+      "step": 12609
+    },
+    {
+      "epoch": 2.043097861309138,
+      "grad_norm": 0.8508796691894531,
+      "learning_rate": 1.2195756559229072e-06,
+      "loss": 0.0579,
+      "step": 12610
+    },
+    {
+      "epoch": 2.0432598833441347,
+      "grad_norm": 0.8715740442276001,
+      "learning_rate": 1.2192000819086502e-06,
+      "loss": 0.0582,
+      "step": 12611
+    },
+    {
+      "epoch": 2.0434219053791316,
+      "grad_norm": 0.8794846534729004,
+      "learning_rate": 1.2188245470837702e-06,
+      "loss": 0.0638,
+      "step": 12612
+    },
+    {
+      "epoch": 2.0435839274141285,
+      "grad_norm": 0.8919373750686646,
+      "learning_rate": 1.2184490514597606e-06,
+      "loss": 0.068,
+      "step": 12613
+    },
+    {
+      "epoch": 2.043745949449125,
+      "grad_norm": 1.1181578636169434,
+      "learning_rate": 1.218073595048108e-06,
+      "loss": 0.074,
+      "step": 12614
+    },
+    {
+      "epoch": 2.043907971484122,
+      "grad_norm": 0.9722860455513,
+      "learning_rate": 1.217698177860301e-06,
+      "loss": 0.0629,
+      "step": 12615
+    },
+    {
+      "epoch": 2.0440699935191184,
+      "grad_norm": 0.8910360932350159,
+      "learning_rate": 1.2173227999078264e-06,
+      "loss": 0.0586,
+      "step": 12616
+    },
+    {
+      "epoch": 2.0442320155541154,
+      "grad_norm": 1.0762535333633423,
+      "learning_rate": 1.2169474612021703e-06,
+      "loss": 0.066,
+      "step": 12617
+    },
+    {
+      "epoch": 2.0443940375891123,
+      "grad_norm": 0.9633141756057739,
+      "learning_rate": 1.2165721617548172e-06,
+      "loss": 0.0638,
+      "step": 12618
+    },
+    {
+      "epoch": 2.044556059624109,
+      "grad_norm": 0.9522542357444763,
+      "learning_rate": 1.2161969015772498e-06,
+      "loss": 0.0645,
+      "step": 12619
+    },
+    {
+      "epoch": 2.0447180816591057,
+      "grad_norm": 0.907698392868042,
+      "learning_rate": 1.2158216806809505e-06,
+      "loss": 0.0629,
+      "step": 12620
+    },
+    {
+      "epoch": 2.0448801036941022,
+      "grad_norm": 0.9610335230827332,
+      "learning_rate": 1.2154464990774013e-06,
+      "loss": 0.0621,
+      "step": 12621
+    },
+    {
+      "epoch": 2.045042125729099,
+      "grad_norm": 0.8890466690063477,
+      "learning_rate": 1.2150713567780786e-06,
+      "loss": 0.0585,
+      "step": 12622
+    },
+    {
+      "epoch": 2.045204147764096,
+      "grad_norm": 0.8295314908027649,
+      "learning_rate": 1.2146962537944638e-06,
+      "loss": 0.0649,
+      "step": 12623
+    },
+    {
+      "epoch": 2.0453661697990926,
+      "grad_norm": 0.883902907371521,
+      "learning_rate": 1.2143211901380341e-06,
+      "loss": 0.055,
+      "step": 12624
+    },
+    {
+      "epoch": 2.0455281918340895,
+      "grad_norm": 0.8097559213638306,
+      "learning_rate": 1.2139461658202642e-06,
+      "loss": 0.0591,
+      "step": 12625
+    },
+    {
+      "epoch": 2.045690213869086,
+      "grad_norm": 0.8150593638420105,
+      "learning_rate": 1.2135711808526282e-06,
+      "loss": 0.0614,
+      "step": 12626
+    },
+    {
+      "epoch": 2.045852235904083,
+      "grad_norm": 0.8336517810821533,
+      "learning_rate": 1.213196235246602e-06,
+      "loss": 0.0626,
+      "step": 12627
+    },
+    {
+      "epoch": 2.04601425793908,
+      "grad_norm": 0.7647026777267456,
+      "learning_rate": 1.2128213290136578e-06,
+      "loss": 0.0603,
+      "step": 12628
+    },
+    {
+      "epoch": 2.0461762799740764,
+      "grad_norm": 0.7869901657104492,
+      "learning_rate": 1.212446462165265e-06,
+      "loss": 0.0578,
+      "step": 12629
+    },
+    {
+      "epoch": 2.0463383020090733,
+      "grad_norm": 0.957288920879364,
+      "learning_rate": 1.212071634712895e-06,
+      "loss": 0.0735,
+      "step": 12630
+    },
+    {
+      "epoch": 2.04650032404407,
+      "grad_norm": 0.9335697889328003,
+      "learning_rate": 1.2116968466680159e-06,
+      "loss": 0.062,
+      "step": 12631
+    },
+    {
+      "epoch": 2.0466623460790667,
+      "grad_norm": 0.85823655128479,
+      "learning_rate": 1.211322098042096e-06,
+      "loss": 0.0543,
+      "step": 12632
+    },
+    {
+      "epoch": 2.0468243681140637,
+      "grad_norm": 0.9083124995231628,
+      "learning_rate": 1.210947388846601e-06,
+      "loss": 0.0638,
+      "step": 12633
+    },
+    {
+      "epoch": 2.04698639014906,
+      "grad_norm": 0.773314893245697,
+      "learning_rate": 1.2105727190929967e-06,
+      "loss": 0.0619,
+      "step": 12634
+    },
+    {
+      "epoch": 2.047148412184057,
+      "grad_norm": 0.8823757767677307,
+      "learning_rate": 1.2101980887927467e-06,
+      "loss": 0.0663,
+      "step": 12635
+    },
+    {
+      "epoch": 2.047310434219054,
+      "grad_norm": 1.0058071613311768,
+      "learning_rate": 1.209823497957314e-06,
+      "loss": 0.0649,
+      "step": 12636
+    },
+    {
+      "epoch": 2.0474724562540505,
+      "grad_norm": 0.9731652736663818,
+      "learning_rate": 1.2094489465981602e-06,
+      "loss": 0.0596,
+      "step": 12637
+    },
+    {
+      "epoch": 2.0476344782890474,
+      "grad_norm": 0.9225329160690308,
+      "learning_rate": 1.2090744347267452e-06,
+      "loss": 0.074,
+      "step": 12638
+    },
+    {
+      "epoch": 2.047796500324044,
+      "grad_norm": 0.7908154129981995,
+      "learning_rate": 1.2086999623545297e-06,
+      "loss": 0.0611,
+      "step": 12639
+    },
+    {
+      "epoch": 2.047958522359041,
+      "grad_norm": 0.7675386667251587,
+      "learning_rate": 1.2083255294929697e-06,
+      "loss": 0.0613,
+      "step": 12640
+    },
+    {
+      "epoch": 2.048120544394038,
+      "grad_norm": 0.7376556396484375,
+      "learning_rate": 1.2079511361535214e-06,
+      "loss": 0.0549,
+      "step": 12641
+    },
+    {
+      "epoch": 2.0482825664290343,
+      "grad_norm": 1.0447229146957397,
+      "learning_rate": 1.2075767823476439e-06,
+      "loss": 0.0698,
+      "step": 12642
+    },
+    {
+      "epoch": 2.048444588464031,
+      "grad_norm": 0.8112842440605164,
+      "learning_rate": 1.207202468086788e-06,
+      "loss": 0.0597,
+      "step": 12643
+    },
+    {
+      "epoch": 2.0486066104990277,
+      "grad_norm": 0.9694043397903442,
+      "learning_rate": 1.2068281933824084e-06,
+      "loss": 0.0697,
+      "step": 12644
+    },
+    {
+      "epoch": 2.0487686325340246,
+      "grad_norm": 1.0804142951965332,
+      "learning_rate": 1.2064539582459564e-06,
+      "loss": 0.0691,
+      "step": 12645
+    },
+    {
+      "epoch": 2.0489306545690216,
+      "grad_norm": 0.8556890487670898,
+      "learning_rate": 1.2060797626888828e-06,
+      "loss": 0.0708,
+      "step": 12646
+    },
+    {
+      "epoch": 2.049092676604018,
+      "grad_norm": 0.839736819267273,
+      "learning_rate": 1.2057056067226374e-06,
+      "loss": 0.0549,
+      "step": 12647
+    },
+    {
+      "epoch": 2.049254698639015,
+      "grad_norm": 0.7325344681739807,
+      "learning_rate": 1.2053314903586685e-06,
+      "loss": 0.0569,
+      "step": 12648
+    },
+    {
+      "epoch": 2.0494167206740115,
+      "grad_norm": 0.9087137579917908,
+      "learning_rate": 1.2049574136084228e-06,
+      "loss": 0.066,
+      "step": 12649
+    },
+    {
+      "epoch": 2.0495787427090084,
+      "grad_norm": 0.8081238865852356,
+      "learning_rate": 1.2045833764833461e-06,
+      "loss": 0.064,
+      "step": 12650
+    },
+    {
+      "epoch": 2.0497407647440054,
+      "grad_norm": 1.0425145626068115,
+      "learning_rate": 1.2042093789948836e-06,
+      "loss": 0.0651,
+      "step": 12651
+    },
+    {
+      "epoch": 2.049902786779002,
+      "grad_norm": 0.8078098893165588,
+      "learning_rate": 1.2038354211544781e-06,
+      "loss": 0.0593,
+      "step": 12652
+    },
+    {
+      "epoch": 2.0500648088139988,
+      "grad_norm": 0.8982751369476318,
+      "learning_rate": 1.2034615029735722e-06,
+      "loss": 0.0703,
+      "step": 12653
+    },
+    {
+      "epoch": 2.0502268308489953,
+      "grad_norm": 1.2443002462387085,
+      "learning_rate": 1.2030876244636078e-06,
+      "loss": 0.0627,
+      "step": 12654
+    },
+    {
+      "epoch": 2.050388852883992,
+      "grad_norm": 0.8545292019844055,
+      "learning_rate": 1.2027137856360212e-06,
+      "loss": 0.0665,
+      "step": 12655
+    },
+    {
+      "epoch": 2.050550874918989,
+      "grad_norm": 0.7731471061706543,
+      "learning_rate": 1.202339986502255e-06,
+      "loss": 0.058,
+      "step": 12656
+    },
+    {
+      "epoch": 2.0507128969539856,
+      "grad_norm": 0.8670762777328491,
+      "learning_rate": 1.2019662270737455e-06,
+      "loss": 0.0616,
+      "step": 12657
+    },
+    {
+      "epoch": 2.0508749189889826,
+      "grad_norm": 0.7677872776985168,
+      "learning_rate": 1.2015925073619275e-06,
+      "loss": 0.0561,
+      "step": 12658
+    },
+    {
+      "epoch": 2.051036941023979,
+      "grad_norm": 0.8861536979675293,
+      "learning_rate": 1.2012188273782367e-06,
+      "loss": 0.0666,
+      "step": 12659
+    },
+    {
+      "epoch": 2.051198963058976,
+      "grad_norm": 0.8551197052001953,
+      "learning_rate": 1.2008451871341056e-06,
+      "loss": 0.0547,
+      "step": 12660
+    },
+    {
+      "epoch": 2.051360985093973,
+      "grad_norm": 0.7268028855323792,
+      "learning_rate": 1.20047158664097e-06,
+      "loss": 0.0543,
+      "step": 12661
+    },
+    {
+      "epoch": 2.0515230071289694,
+      "grad_norm": 1.0221055746078491,
+      "learning_rate": 1.200098025910258e-06,
+      "loss": 0.0664,
+      "step": 12662
+    },
+    {
+      "epoch": 2.0516850291639663,
+      "grad_norm": 0.7850842475891113,
+      "learning_rate": 1.1997245049534007e-06,
+      "loss": 0.0598,
+      "step": 12663
+    },
+    {
+      "epoch": 2.0518470511989633,
+      "grad_norm": 0.9158909320831299,
+      "learning_rate": 1.1993510237818269e-06,
+      "loss": 0.0676,
+      "step": 12664
+    },
+    {
+      "epoch": 2.0520090732339598,
+      "grad_norm": 0.9576240181922913,
+      "learning_rate": 1.1989775824069645e-06,
+      "loss": 0.0613,
+      "step": 12665
+    },
+    {
+      "epoch": 2.0521710952689567,
+      "grad_norm": 0.7837166786193848,
+      "learning_rate": 1.1986041808402393e-06,
+      "loss": 0.0554,
+      "step": 12666
+    },
+    {
+      "epoch": 2.052333117303953,
+      "grad_norm": 0.9599198698997498,
+      "learning_rate": 1.198230819093077e-06,
+      "loss": 0.0716,
+      "step": 12667
+    },
+    {
+      "epoch": 2.05249513933895,
+      "grad_norm": 1.0300556421279907,
+      "learning_rate": 1.1978574971769025e-06,
+      "loss": 0.0639,
+      "step": 12668
+    },
+    {
+      "epoch": 2.052657161373947,
+      "grad_norm": 0.9385240077972412,
+      "learning_rate": 1.1974842151031354e-06,
+      "loss": 0.0651,
+      "step": 12669
+    },
+    {
+      "epoch": 2.0528191834089435,
+      "grad_norm": 0.8229407072067261,
+      "learning_rate": 1.1971109728832003e-06,
+      "loss": 0.0599,
+      "step": 12670
+    },
+    {
+      "epoch": 2.0529812054439405,
+      "grad_norm": 0.8111064434051514,
+      "learning_rate": 1.1967377705285163e-06,
+      "loss": 0.0614,
+      "step": 12671
+    },
+    {
+      "epoch": 2.053143227478937,
+      "grad_norm": 0.716200053691864,
+      "learning_rate": 1.196364608050504e-06,
+      "loss": 0.061,
+      "step": 12672
+    },
+    {
+      "epoch": 2.053305249513934,
+      "grad_norm": 0.8643745183944702,
+      "learning_rate": 1.1959914854605788e-06,
+      "loss": 0.0606,
+      "step": 12673
+    },
+    {
+      "epoch": 2.053467271548931,
+      "grad_norm": 0.9790539741516113,
+      "learning_rate": 1.1956184027701576e-06,
+      "loss": 0.0629,
+      "step": 12674
+    },
+    {
+      "epoch": 2.0536292935839273,
+      "grad_norm": 0.9859734177589417,
+      "learning_rate": 1.1952453599906585e-06,
+      "loss": 0.0655,
+      "step": 12675
+    },
+    {
+      "epoch": 2.0537913156189243,
+      "grad_norm": 0.784511148929596,
+      "learning_rate": 1.1948723571334932e-06,
+      "loss": 0.0544,
+      "step": 12676
+    },
+    {
+      "epoch": 2.0539533376539207,
+      "grad_norm": 0.7424229383468628,
+      "learning_rate": 1.1944993942100755e-06,
+      "loss": 0.0563,
+      "step": 12677
+    },
+    {
+      "epoch": 2.0541153596889177,
+      "grad_norm": 0.9328092932701111,
+      "learning_rate": 1.1941264712318167e-06,
+      "loss": 0.0677,
+      "step": 12678
+    },
+    {
+      "epoch": 2.0542773817239146,
+      "grad_norm": 0.8559496998786926,
+      "learning_rate": 1.193753588210128e-06,
+      "loss": 0.068,
+      "step": 12679
+    },
+    {
+      "epoch": 2.054439403758911,
+      "grad_norm": 1.0441864728927612,
+      "learning_rate": 1.1933807451564186e-06,
+      "loss": 0.0597,
+      "step": 12680
+    },
+    {
+      "epoch": 2.054601425793908,
+      "grad_norm": 0.8556796908378601,
+      "learning_rate": 1.1930079420820962e-06,
+      "loss": 0.0575,
+      "step": 12681
+    },
+    {
+      "epoch": 2.0547634478289045,
+      "grad_norm": 0.9726545214653015,
+      "learning_rate": 1.192635178998568e-06,
+      "loss": 0.0698,
+      "step": 12682
+    },
+    {
+      "epoch": 2.0549254698639015,
+      "grad_norm": 0.7726395726203918,
+      "learning_rate": 1.1922624559172404e-06,
+      "loss": 0.053,
+      "step": 12683
+    },
+    {
+      "epoch": 2.0550874918988984,
+      "grad_norm": 1.3257524967193604,
+      "learning_rate": 1.191889772849515e-06,
+      "loss": 0.057,
+      "step": 12684
+    },
+    {
+      "epoch": 2.055249513933895,
+      "grad_norm": 0.8716280460357666,
+      "learning_rate": 1.1915171298067982e-06,
+      "loss": 0.0626,
+      "step": 12685
+    },
+    {
+      "epoch": 2.055411535968892,
+      "grad_norm": 0.7775639891624451,
+      "learning_rate": 1.1911445268004917e-06,
+      "loss": 0.0586,
+      "step": 12686
+    },
+    {
+      "epoch": 2.0555735580038887,
+      "grad_norm": 0.7742239236831665,
+      "learning_rate": 1.1907719638419943e-06,
+      "loss": 0.0577,
+      "step": 12687
+    },
+    {
+      "epoch": 2.0557355800388852,
+      "grad_norm": 0.9106540679931641,
+      "learning_rate": 1.1903994409427063e-06,
+      "loss": 0.0641,
+      "step": 12688
+    },
+    {
+      "epoch": 2.055897602073882,
+      "grad_norm": 0.8494895100593567,
+      "learning_rate": 1.1900269581140257e-06,
+      "loss": 0.0622,
+      "step": 12689
+    },
+    {
+      "epoch": 2.0560596241088787,
+      "grad_norm": 0.9407444596290588,
+      "learning_rate": 1.1896545153673517e-06,
+      "loss": 0.067,
+      "step": 12690
+    },
+    {
+      "epoch": 2.0562216461438756,
+      "grad_norm": 0.734492301940918,
+      "learning_rate": 1.1892821127140777e-06,
+      "loss": 0.0553,
+      "step": 12691
+    },
+    {
+      "epoch": 2.0563836681788725,
+      "grad_norm": 0.8618312478065491,
+      "learning_rate": 1.1889097501655991e-06,
+      "loss": 0.0624,
+      "step": 12692
+    },
+    {
+      "epoch": 2.056545690213869,
+      "grad_norm": 0.8262837529182434,
+      "learning_rate": 1.1885374277333095e-06,
+      "loss": 0.0579,
+      "step": 12693
+    },
+    {
+      "epoch": 2.056707712248866,
+      "grad_norm": 0.832383930683136,
+      "learning_rate": 1.1881651454286008e-06,
+      "loss": 0.0627,
+      "step": 12694
+    },
+    {
+      "epoch": 2.0568697342838624,
+      "grad_norm": 0.8754958510398865,
+      "learning_rate": 1.187792903262864e-06,
+      "loss": 0.0592,
+      "step": 12695
+    },
+    {
+      "epoch": 2.0570317563188594,
+      "grad_norm": 0.9576855301856995,
+      "learning_rate": 1.1874207012474891e-06,
+      "loss": 0.0672,
+      "step": 12696
+    },
+    {
+      "epoch": 2.0571937783538563,
+      "grad_norm": 0.8752747774124146,
+      "learning_rate": 1.1870485393938644e-06,
+      "loss": 0.0658,
+      "step": 12697
+    },
+    {
+      "epoch": 2.057355800388853,
+      "grad_norm": 0.9489736557006836,
+      "learning_rate": 1.186676417713377e-06,
+      "loss": 0.062,
+      "step": 12698
+    },
+    {
+      "epoch": 2.0575178224238497,
+      "grad_norm": 0.791542112827301,
+      "learning_rate": 1.1863043362174129e-06,
+      "loss": 0.056,
+      "step": 12699
+    },
+    {
+      "epoch": 2.057679844458846,
+      "grad_norm": 0.8600035309791565,
+      "learning_rate": 1.1859322949173572e-06,
+      "loss": 0.0646,
+      "step": 12700
+    },
+    {
+      "epoch": 2.057841866493843,
+      "grad_norm": 0.8470669984817505,
+      "learning_rate": 1.1855602938245942e-06,
+      "loss": 0.0658,
+      "step": 12701
+    },
+    {
+      "epoch": 2.05800388852884,
+      "grad_norm": 0.8929564952850342,
+      "learning_rate": 1.1851883329505043e-06,
+      "loss": 0.0589,
+      "step": 12702
+    },
+    {
+      "epoch": 2.0581659105638366,
+      "grad_norm": 0.9186873435974121,
+      "learning_rate": 1.1848164123064687e-06,
+      "loss": 0.073,
+      "step": 12703
+    },
+    {
+      "epoch": 2.0583279325988335,
+      "grad_norm": 0.803128719329834,
+      "learning_rate": 1.1844445319038694e-06,
+      "loss": 0.059,
+      "step": 12704
+    },
+    {
+      "epoch": 2.05848995463383,
+      "grad_norm": 0.7790929675102234,
+      "learning_rate": 1.1840726917540846e-06,
+      "loss": 0.0562,
+      "step": 12705
+    },
+    {
+      "epoch": 2.058651976668827,
+      "grad_norm": 0.9201391935348511,
+      "learning_rate": 1.18370089186849e-06,
+      "loss": 0.0644,
+      "step": 12706
+    },
+    {
+      "epoch": 2.058813998703824,
+      "grad_norm": 1.1075505018234253,
+      "learning_rate": 1.1833291322584625e-06,
+      "loss": 0.0671,
+      "step": 12707
+    },
+    {
+      "epoch": 2.0589760207388204,
+      "grad_norm": 0.869609534740448,
+      "learning_rate": 1.1829574129353777e-06,
+      "loss": 0.0663,
+      "step": 12708
+    },
+    {
+      "epoch": 2.0591380427738173,
+      "grad_norm": 1.0262631177902222,
+      "learning_rate": 1.1825857339106086e-06,
+      "loss": 0.059,
+      "step": 12709
+    },
+    {
+      "epoch": 2.059300064808814,
+      "grad_norm": 0.8231914043426514,
+      "learning_rate": 1.182214095195528e-06,
+      "loss": 0.0576,
+      "step": 12710
+    },
+    {
+      "epoch": 2.0594620868438107,
+      "grad_norm": 0.8270055651664734,
+      "learning_rate": 1.181842496801507e-06,
+      "loss": 0.0574,
+      "step": 12711
+    },
+    {
+      "epoch": 2.0596241088788076,
+      "grad_norm": 1.0092082023620605,
+      "learning_rate": 1.181470938739917e-06,
+      "loss": 0.0619,
+      "step": 12712
+    },
+    {
+      "epoch": 2.059786130913804,
+      "grad_norm": 0.7911940217018127,
+      "learning_rate": 1.1810994210221234e-06,
+      "loss": 0.0587,
+      "step": 12713
+    },
+    {
+      "epoch": 2.059948152948801,
+      "grad_norm": 0.9792639017105103,
+      "learning_rate": 1.1807279436594967e-06,
+      "loss": 0.0734,
+      "step": 12714
+    },
+    {
+      "epoch": 2.060110174983798,
+      "grad_norm": 1.0325294733047485,
+      "learning_rate": 1.1803565066634027e-06,
+      "loss": 0.0708,
+      "step": 12715
+    },
+    {
+      "epoch": 2.0602721970187945,
+      "grad_norm": 0.8511001467704773,
+      "learning_rate": 1.1799851100452067e-06,
+      "loss": 0.0583,
+      "step": 12716
+    },
+    {
+      "epoch": 2.0604342190537914,
+      "grad_norm": 0.9019601345062256,
+      "learning_rate": 1.17961375381627e-06,
+      "loss": 0.0669,
+      "step": 12717
+    },
+    {
+      "epoch": 2.060596241088788,
+      "grad_norm": 0.9868301153182983,
+      "learning_rate": 1.1792424379879582e-06,
+      "loss": 0.0608,
+      "step": 12718
+    },
+    {
+      "epoch": 2.060758263123785,
+      "grad_norm": 0.954416036605835,
+      "learning_rate": 1.178871162571633e-06,
+      "loss": 0.0735,
+      "step": 12719
+    },
+    {
+      "epoch": 2.060920285158782,
+      "grad_norm": 1.1071062088012695,
+      "learning_rate": 1.1784999275786515e-06,
+      "loss": 0.0564,
+      "step": 12720
+    },
+    {
+      "epoch": 2.0610823071937783,
+      "grad_norm": 0.8846614956855774,
+      "learning_rate": 1.1781287330203747e-06,
+      "loss": 0.0631,
+      "step": 12721
+    },
+    {
+      "epoch": 2.061244329228775,
+      "grad_norm": 0.894943118095398,
+      "learning_rate": 1.177757578908159e-06,
+      "loss": 0.0604,
+      "step": 12722
+    },
+    {
+      "epoch": 2.0614063512637717,
+      "grad_norm": 1.3799413442611694,
+      "learning_rate": 1.177386465253363e-06,
+      "loss": 0.0675,
+      "step": 12723
+    },
+    {
+      "epoch": 2.0615683732987686,
+      "grad_norm": 0.8426232933998108,
+      "learning_rate": 1.17701539206734e-06,
+      "loss": 0.0613,
+      "step": 12724
+    },
+    {
+      "epoch": 2.0617303953337656,
+      "grad_norm": 0.9407797455787659,
+      "learning_rate": 1.176644359361444e-06,
+      "loss": 0.0729,
+      "step": 12725
+    },
+    {
+      "epoch": 2.061892417368762,
+      "grad_norm": 0.9105675220489502,
+      "learning_rate": 1.1762733671470285e-06,
+      "loss": 0.0679,
+      "step": 12726
+    },
+    {
+      "epoch": 2.062054439403759,
+      "grad_norm": 0.748227596282959,
+      "learning_rate": 1.1759024154354446e-06,
+      "loss": 0.0594,
+      "step": 12727
+    },
+    {
+      "epoch": 2.0622164614387555,
+      "grad_norm": 0.8079039454460144,
+      "learning_rate": 1.1755315042380425e-06,
+      "loss": 0.0629,
+      "step": 12728
+    },
+    {
+      "epoch": 2.0623784834737524,
+      "grad_norm": 0.9434335231781006,
+      "learning_rate": 1.175160633566171e-06,
+      "loss": 0.0646,
+      "step": 12729
+    },
+    {
+      "epoch": 2.0625405055087493,
+      "grad_norm": 0.8539111018180847,
+      "learning_rate": 1.1747898034311782e-06,
+      "loss": 0.0605,
+      "step": 12730
+    },
+    {
+      "epoch": 2.062702527543746,
+      "grad_norm": 0.8666152954101562,
+      "learning_rate": 1.1744190138444118e-06,
+      "loss": 0.071,
+      "step": 12731
+    },
+    {
+      "epoch": 2.0628645495787428,
+      "grad_norm": 1.0582692623138428,
+      "learning_rate": 1.1740482648172132e-06,
+      "loss": 0.0718,
+      "step": 12732
+    },
+    {
+      "epoch": 2.0630265716137393,
+      "grad_norm": 0.9556350111961365,
+      "learning_rate": 1.1736775563609305e-06,
+      "loss": 0.0716,
+      "step": 12733
+    },
+    {
+      "epoch": 2.063188593648736,
+      "grad_norm": 0.829980731010437,
+      "learning_rate": 1.1733068884869053e-06,
+      "loss": 0.0654,
+      "step": 12734
+    },
+    {
+      "epoch": 2.063350615683733,
+      "grad_norm": 0.7806819081306458,
+      "learning_rate": 1.1729362612064782e-06,
+      "loss": 0.0574,
+      "step": 12735
+    },
+    {
+      "epoch": 2.0635126377187296,
+      "grad_norm": 1.1992825269699097,
+      "learning_rate": 1.172565674530989e-06,
+      "loss": 0.0556,
+      "step": 12736
+    },
+    {
+      "epoch": 2.0636746597537265,
+      "grad_norm": 0.9343496561050415,
+      "learning_rate": 1.1721951284717797e-06,
+      "loss": 0.0697,
+      "step": 12737
+    },
+    {
+      "epoch": 2.0638366817887235,
+      "grad_norm": 0.8090648055076599,
+      "learning_rate": 1.1718246230401856e-06,
+      "loss": 0.0644,
+      "step": 12738
+    },
+    {
+      "epoch": 2.06399870382372,
+      "grad_norm": 0.9163773059844971,
+      "learning_rate": 1.1714541582475435e-06,
+      "loss": 0.0707,
+      "step": 12739
+    },
+    {
+      "epoch": 2.064160725858717,
+      "grad_norm": 0.9538021683692932,
+      "learning_rate": 1.1710837341051892e-06,
+      "loss": 0.0655,
+      "step": 12740
+    },
+    {
+      "epoch": 2.0643227478937134,
+      "grad_norm": 0.894012451171875,
+      "learning_rate": 1.170713350624457e-06,
+      "loss": 0.067,
+      "step": 12741
+    },
+    {
+      "epoch": 2.0644847699287103,
+      "grad_norm": 0.896256148815155,
+      "learning_rate": 1.1703430078166792e-06,
+      "loss": 0.0683,
+      "step": 12742
+    },
+    {
+      "epoch": 2.0646467919637073,
+      "grad_norm": 0.8130720257759094,
+      "learning_rate": 1.1699727056931878e-06,
+      "loss": 0.0666,
+      "step": 12743
+    },
+    {
+      "epoch": 2.0648088139987038,
+      "grad_norm": 0.7944749593734741,
+      "learning_rate": 1.169602444265313e-06,
+      "loss": 0.0639,
+      "step": 12744
+    },
+    {
+      "epoch": 2.0649708360337007,
+      "grad_norm": 0.9021603465080261,
+      "learning_rate": 1.1692322235443845e-06,
+      "loss": 0.069,
+      "step": 12745
+    },
+    {
+      "epoch": 2.065132858068697,
+      "grad_norm": 0.8242917060852051,
+      "learning_rate": 1.168862043541728e-06,
+      "loss": 0.0675,
+      "step": 12746
+    },
+    {
+      "epoch": 2.065294880103694,
+      "grad_norm": 0.806065559387207,
+      "learning_rate": 1.1684919042686727e-06,
+      "loss": 0.0595,
+      "step": 12747
+    },
+    {
+      "epoch": 2.065456902138691,
+      "grad_norm": 0.8553471565246582,
+      "learning_rate": 1.1681218057365429e-06,
+      "loss": 0.0588,
+      "step": 12748
+    },
+    {
+      "epoch": 2.0656189241736875,
+      "grad_norm": 0.8592274785041809,
+      "learning_rate": 1.1677517479566636e-06,
+      "loss": 0.0677,
+      "step": 12749
+    },
+    {
+      "epoch": 2.0657809462086845,
+      "grad_norm": 0.8480504155158997,
+      "learning_rate": 1.167381730940356e-06,
+      "loss": 0.0636,
+      "step": 12750
+    },
+    {
+      "epoch": 2.065942968243681,
+      "grad_norm": 0.768464207649231,
+      "learning_rate": 1.1670117546989416e-06,
+      "loss": 0.0603,
+      "step": 12751
+    },
+    {
+      "epoch": 2.066104990278678,
+      "grad_norm": 0.8079333305358887,
+      "learning_rate": 1.1666418192437434e-06,
+      "loss": 0.0615,
+      "step": 12752
+    },
+    {
+      "epoch": 2.066267012313675,
+      "grad_norm": 0.8408043384552002,
+      "learning_rate": 1.1662719245860782e-06,
+      "loss": 0.0604,
+      "step": 12753
+    },
+    {
+      "epoch": 2.0664290343486713,
+      "grad_norm": 0.8342251181602478,
+      "learning_rate": 1.1659020707372643e-06,
+      "loss": 0.061,
+      "step": 12754
+    },
+    {
+      "epoch": 2.0665910563836682,
+      "grad_norm": 0.868290364742279,
+      "learning_rate": 1.1655322577086186e-06,
+      "loss": 0.0628,
+      "step": 12755
+    },
+    {
+      "epoch": 2.0667530784186647,
+      "grad_norm": 0.7598575353622437,
+      "learning_rate": 1.1651624855114565e-06,
+      "loss": 0.0547,
+      "step": 12756
+    },
+    {
+      "epoch": 2.0669151004536617,
+      "grad_norm": 0.7536554932594299,
+      "learning_rate": 1.1647927541570922e-06,
+      "loss": 0.0583,
+      "step": 12757
+    },
+    {
+      "epoch": 2.0670771224886586,
+      "grad_norm": 0.8469033241271973,
+      "learning_rate": 1.1644230636568384e-06,
+      "loss": 0.0659,
+      "step": 12758
+    },
+    {
+      "epoch": 2.067239144523655,
+      "grad_norm": 0.796055257320404,
+      "learning_rate": 1.164053414022007e-06,
+      "loss": 0.0568,
+      "step": 12759
+    },
+    {
+      "epoch": 2.067401166558652,
+      "grad_norm": 0.922766387462616,
+      "learning_rate": 1.1636838052639081e-06,
+      "loss": 0.0627,
+      "step": 12760
+    },
+    {
+      "epoch": 2.067563188593649,
+      "grad_norm": 0.876051664352417,
+      "learning_rate": 1.163314237393851e-06,
+      "loss": 0.0594,
+      "step": 12761
+    },
+    {
+      "epoch": 2.0677252106286454,
+      "grad_norm": 0.8684565424919128,
+      "learning_rate": 1.1629447104231435e-06,
+      "loss": 0.0642,
+      "step": 12762
+    },
+    {
+      "epoch": 2.0678872326636424,
+      "grad_norm": 0.8998528122901917,
+      "learning_rate": 1.162575224363093e-06,
+      "loss": 0.0714,
+      "step": 12763
+    },
+    {
+      "epoch": 2.068049254698639,
+      "grad_norm": 0.874921977519989,
+      "learning_rate": 1.1622057792250033e-06,
+      "loss": 0.0613,
+      "step": 12764
+    },
+    {
+      "epoch": 2.068211276733636,
+      "grad_norm": 0.9340734481811523,
+      "learning_rate": 1.1618363750201784e-06,
+      "loss": 0.0604,
+      "step": 12765
+    },
+    {
+      "epoch": 2.0683732987686327,
+      "grad_norm": 1.0002615451812744,
+      "learning_rate": 1.1614670117599231e-06,
+      "loss": 0.0656,
+      "step": 12766
+    },
+    {
+      "epoch": 2.0685353208036292,
+      "grad_norm": 0.8985110521316528,
+      "learning_rate": 1.161097689455539e-06,
+      "loss": 0.0595,
+      "step": 12767
+    },
+    {
+      "epoch": 2.068697342838626,
+      "grad_norm": 0.829571008682251,
+      "learning_rate": 1.1607284081183245e-06,
+      "loss": 0.0554,
+      "step": 12768
+    },
+    {
+      "epoch": 2.0688593648736227,
+      "grad_norm": 0.9190569519996643,
+      "learning_rate": 1.16035916775958e-06,
+      "loss": 0.0607,
+      "step": 12769
+    },
+    {
+      "epoch": 2.0690213869086196,
+      "grad_norm": 0.9735108017921448,
+      "learning_rate": 1.1599899683906026e-06,
+      "loss": 0.0649,
+      "step": 12770
+    },
+    {
+      "epoch": 2.0691834089436165,
+      "grad_norm": 0.8792665600776672,
+      "learning_rate": 1.1596208100226899e-06,
+      "loss": 0.0554,
+      "step": 12771
+    },
+    {
+      "epoch": 2.069345430978613,
+      "grad_norm": 0.8092507719993591,
+      "learning_rate": 1.1592516926671367e-06,
+      "loss": 0.0611,
+      "step": 12772
+    },
+    {
+      "epoch": 2.06950745301361,
+      "grad_norm": 0.9477719068527222,
+      "learning_rate": 1.1588826163352369e-06,
+      "loss": 0.0648,
+      "step": 12773
+    },
+    {
+      "epoch": 2.0696694750486064,
+      "grad_norm": 0.806533694267273,
+      "learning_rate": 1.1585135810382836e-06,
+      "loss": 0.0603,
+      "step": 12774
+    },
+    {
+      "epoch": 2.0698314970836034,
+      "grad_norm": 0.8330726027488708,
+      "learning_rate": 1.1581445867875684e-06,
+      "loss": 0.0618,
+      "step": 12775
+    },
+    {
+      "epoch": 2.0699935191186003,
+      "grad_norm": 0.9184839129447937,
+      "learning_rate": 1.1577756335943818e-06,
+      "loss": 0.0624,
+      "step": 12776
+    },
+    {
+      "epoch": 2.070155541153597,
+      "grad_norm": 0.8566496968269348,
+      "learning_rate": 1.1574067214700127e-06,
+      "loss": 0.0613,
+      "step": 12777
+    },
+    {
+      "epoch": 2.0703175631885937,
+      "grad_norm": 0.7928066849708557,
+      "learning_rate": 1.1570378504257499e-06,
+      "loss": 0.0562,
+      "step": 12778
+    },
+    {
+      "epoch": 2.07047958522359,
+      "grad_norm": 0.8417590856552124,
+      "learning_rate": 1.1566690204728779e-06,
+      "loss": 0.0549,
+      "step": 12779
+    },
+    {
+      "epoch": 2.070641607258587,
+      "grad_norm": 0.9719398021697998,
+      "learning_rate": 1.156300231622682e-06,
+      "loss": 0.0687,
+      "step": 12780
+    },
+    {
+      "epoch": 2.070803629293584,
+      "grad_norm": 0.9284224510192871,
+      "learning_rate": 1.1559314838864494e-06,
+      "loss": 0.0637,
+      "step": 12781
+    },
+    {
+      "epoch": 2.0709656513285806,
+      "grad_norm": 1.0845189094543457,
+      "learning_rate": 1.1555627772754595e-06,
+      "loss": 0.0689,
+      "step": 12782
+    },
+    {
+      "epoch": 2.0711276733635775,
+      "grad_norm": 0.9417441487312317,
+      "learning_rate": 1.1551941118009957e-06,
+      "loss": 0.0633,
+      "step": 12783
+    },
+    {
+      "epoch": 2.071289695398574,
+      "grad_norm": 0.8231827020645142,
+      "learning_rate": 1.1548254874743365e-06,
+      "loss": 0.06,
+      "step": 12784
+    },
+    {
+      "epoch": 2.071451717433571,
+      "grad_norm": 0.8818009495735168,
+      "learning_rate": 1.154456904306764e-06,
+      "loss": 0.0611,
+      "step": 12785
+    },
+    {
+      "epoch": 2.071613739468568,
+      "grad_norm": 0.8181620240211487,
+      "learning_rate": 1.154088362309553e-06,
+      "loss": 0.0578,
+      "step": 12786
+    },
+    {
+      "epoch": 2.0717757615035644,
+      "grad_norm": 0.9999836087226868,
+      "learning_rate": 1.1537198614939812e-06,
+      "loss": 0.0684,
+      "step": 12787
+    },
+    {
+      "epoch": 2.0719377835385613,
+      "grad_norm": 0.8913933634757996,
+      "learning_rate": 1.1533514018713238e-06,
+      "loss": 0.0593,
+      "step": 12788
+    },
+    {
+      "epoch": 2.072099805573558,
+      "grad_norm": 0.8349035978317261,
+      "learning_rate": 1.1529829834528547e-06,
+      "loss": 0.0594,
+      "step": 12789
+    },
+    {
+      "epoch": 2.0722618276085547,
+      "grad_norm": 1.0070099830627441,
+      "learning_rate": 1.1526146062498464e-06,
+      "loss": 0.067,
+      "step": 12790
+    },
+    {
+      "epoch": 2.0724238496435516,
+      "grad_norm": 0.8604516386985779,
+      "learning_rate": 1.1522462702735708e-06,
+      "loss": 0.0622,
+      "step": 12791
+    },
+    {
+      "epoch": 2.072585871678548,
+      "grad_norm": 0.8990740776062012,
+      "learning_rate": 1.1518779755352977e-06,
+      "loss": 0.0691,
+      "step": 12792
+    },
+    {
+      "epoch": 2.072747893713545,
+      "grad_norm": 0.9744791388511658,
+      "learning_rate": 1.151509722046297e-06,
+      "loss": 0.0604,
+      "step": 12793
+    },
+    {
+      "epoch": 2.072909915748542,
+      "grad_norm": 0.8340404629707336,
+      "learning_rate": 1.1511415098178336e-06,
+      "loss": 0.0611,
+      "step": 12794
+    },
+    {
+      "epoch": 2.0730719377835385,
+      "grad_norm": 0.9593438506126404,
+      "learning_rate": 1.1507733388611768e-06,
+      "loss": 0.0699,
+      "step": 12795
+    },
+    {
+      "epoch": 2.0732339598185354,
+      "grad_norm": 0.8299152255058289,
+      "learning_rate": 1.1504052091875917e-06,
+      "loss": 0.0567,
+      "step": 12796
+    },
+    {
+      "epoch": 2.073395981853532,
+      "grad_norm": 0.7919155359268188,
+      "learning_rate": 1.1500371208083405e-06,
+      "loss": 0.0604,
+      "step": 12797
+    },
+    {
+      "epoch": 2.073558003888529,
+      "grad_norm": 0.9348534345626831,
+      "learning_rate": 1.1496690737346864e-06,
+      "loss": 0.0608,
+      "step": 12798
+    },
+    {
+      "epoch": 2.073720025923526,
+      "grad_norm": 0.8300426006317139,
+      "learning_rate": 1.14930106797789e-06,
+      "loss": 0.0642,
+      "step": 12799
+    },
+    {
+      "epoch": 2.0738820479585223,
+      "grad_norm": 0.8624827861785889,
+      "learning_rate": 1.148933103549214e-06,
+      "loss": 0.0623,
+      "step": 12800
+    },
+    {
+      "epoch": 2.074044069993519,
+      "grad_norm": 0.9354908466339111,
+      "learning_rate": 1.148565180459915e-06,
+      "loss": 0.0667,
+      "step": 12801
+    },
+    {
+      "epoch": 2.0742060920285157,
+      "grad_norm": 0.9018098711967468,
+      "learning_rate": 1.1481972987212505e-06,
+      "loss": 0.0565,
+      "step": 12802
+    },
+    {
+      "epoch": 2.0743681140635126,
+      "grad_norm": 0.9622835516929626,
+      "learning_rate": 1.1478294583444779e-06,
+      "loss": 0.0651,
+      "step": 12803
+    },
+    {
+      "epoch": 2.0745301360985096,
+      "grad_norm": 0.8161885142326355,
+      "learning_rate": 1.1474616593408513e-06,
+      "loss": 0.0615,
+      "step": 12804
+    },
+    {
+      "epoch": 2.074692158133506,
+      "grad_norm": 0.909511148929596,
+      "learning_rate": 1.147093901721625e-06,
+      "loss": 0.062,
+      "step": 12805
+    },
+    {
+      "epoch": 2.074854180168503,
+      "grad_norm": 0.7995123267173767,
+      "learning_rate": 1.1467261854980513e-06,
+      "loss": 0.063,
+      "step": 12806
+    },
+    {
+      "epoch": 2.0750162022034995,
+      "grad_norm": 0.8639039993286133,
+      "learning_rate": 1.1463585106813823e-06,
+      "loss": 0.0635,
+      "step": 12807
+    },
+    {
+      "epoch": 2.0751782242384964,
+      "grad_norm": 0.8831427693367004,
+      "learning_rate": 1.1459908772828658e-06,
+      "loss": 0.0637,
+      "step": 12808
+    },
+    {
+      "epoch": 2.0753402462734933,
+      "grad_norm": 0.9369395971298218,
+      "learning_rate": 1.1456232853137522e-06,
+      "loss": 0.065,
+      "step": 12809
+    },
+    {
+      "epoch": 2.07550226830849,
+      "grad_norm": 0.8888958096504211,
+      "learning_rate": 1.1452557347852885e-06,
+      "loss": 0.062,
+      "step": 12810
+    },
+    {
+      "epoch": 2.0756642903434868,
+      "grad_norm": 0.9641885757446289,
+      "learning_rate": 1.1448882257087222e-06,
+      "loss": 0.0675,
+      "step": 12811
+    },
+    {
+      "epoch": 2.0758263123784833,
+      "grad_norm": 1.0952035188674927,
+      "learning_rate": 1.1445207580952956e-06,
+      "loss": 0.0677,
+      "step": 12812
+    },
+    {
+      "epoch": 2.07598833441348,
+      "grad_norm": 1.0690641403198242,
+      "learning_rate": 1.1441533319562528e-06,
+      "loss": 0.0685,
+      "step": 12813
+    },
+    {
+      "epoch": 2.076150356448477,
+      "grad_norm": 0.9642135500907898,
+      "learning_rate": 1.143785947302839e-06,
+      "loss": 0.0711,
+      "step": 12814
+    },
+    {
+      "epoch": 2.0763123784834736,
+      "grad_norm": 1.0191136598587036,
+      "learning_rate": 1.143418604146292e-06,
+      "loss": 0.0728,
+      "step": 12815
+    },
+    {
+      "epoch": 2.0764744005184705,
+      "grad_norm": 0.8897978663444519,
+      "learning_rate": 1.143051302497853e-06,
+      "loss": 0.0671,
+      "step": 12816
+    },
+    {
+      "epoch": 2.0766364225534675,
+      "grad_norm": 1.1707680225372314,
+      "learning_rate": 1.1426840423687605e-06,
+      "loss": 0.0633,
+      "step": 12817
+    },
+    {
+      "epoch": 2.076798444588464,
+      "grad_norm": 0.7090242505073547,
+      "learning_rate": 1.1423168237702515e-06,
+      "loss": 0.0524,
+      "step": 12818
+    },
+    {
+      "epoch": 2.076960466623461,
+      "grad_norm": 0.8160051703453064,
+      "learning_rate": 1.141949646713562e-06,
+      "loss": 0.0589,
+      "step": 12819
+    },
+    {
+      "epoch": 2.0771224886584574,
+      "grad_norm": 0.8117794394493103,
+      "learning_rate": 1.1415825112099274e-06,
+      "loss": 0.0533,
+      "step": 12820
+    },
+    {
+      "epoch": 2.0772845106934543,
+      "grad_norm": 0.9710679054260254,
+      "learning_rate": 1.1412154172705803e-06,
+      "loss": 0.0625,
+      "step": 12821
+    },
+    {
+      "epoch": 2.0774465327284513,
+      "grad_norm": 0.8708137273788452,
+      "learning_rate": 1.1408483649067541e-06,
+      "loss": 0.0557,
+      "step": 12822
+    },
+    {
+      "epoch": 2.0776085547634477,
+      "grad_norm": 0.8636717200279236,
+      "learning_rate": 1.1404813541296772e-06,
+      "loss": 0.0605,
+      "step": 12823
+    },
+    {
+      "epoch": 2.0777705767984447,
+      "grad_norm": 0.763903021812439,
+      "learning_rate": 1.1401143849505816e-06,
+      "loss": 0.0597,
+      "step": 12824
+    },
+    {
+      "epoch": 2.077932598833441,
+      "grad_norm": 0.9749400019645691,
+      "learning_rate": 1.139747457380696e-06,
+      "loss": 0.0633,
+      "step": 12825
+    },
+    {
+      "epoch": 2.078094620868438,
+      "grad_norm": 0.9545486569404602,
+      "learning_rate": 1.1393805714312456e-06,
+      "loss": 0.0734,
+      "step": 12826
+    },
+    {
+      "epoch": 2.078256642903435,
+      "grad_norm": 0.8538177013397217,
+      "learning_rate": 1.1390137271134564e-06,
+      "loss": 0.064,
+      "step": 12827
+    },
+    {
+      "epoch": 2.0784186649384315,
+      "grad_norm": 1.0204607248306274,
+      "learning_rate": 1.138646924438554e-06,
+      "loss": 0.0705,
+      "step": 12828
+    },
+    {
+      "epoch": 2.0785806869734285,
+      "grad_norm": 0.9345843195915222,
+      "learning_rate": 1.1382801634177627e-06,
+      "loss": 0.0648,
+      "step": 12829
+    },
+    {
+      "epoch": 2.078742709008425,
+      "grad_norm": 0.9654140472412109,
+      "learning_rate": 1.1379134440623018e-06,
+      "loss": 0.0674,
+      "step": 12830
+    },
+    {
+      "epoch": 2.078904731043422,
+      "grad_norm": 0.8916672468185425,
+      "learning_rate": 1.1375467663833935e-06,
+      "loss": 0.0664,
+      "step": 12831
+    },
+    {
+      "epoch": 2.079066753078419,
+      "grad_norm": 0.9074366092681885,
+      "learning_rate": 1.137180130392257e-06,
+      "loss": 0.0525,
+      "step": 12832
+    },
+    {
+      "epoch": 2.0792287751134153,
+      "grad_norm": 0.9792858362197876,
+      "learning_rate": 1.1368135361001107e-06,
+      "loss": 0.0692,
+      "step": 12833
+    },
+    {
+      "epoch": 2.0793907971484122,
+      "grad_norm": 0.9119909405708313,
+      "learning_rate": 1.1364469835181712e-06,
+      "loss": 0.0681,
+      "step": 12834
+    },
+    {
+      "epoch": 2.0795528191834087,
+      "grad_norm": 0.876586377620697,
+      "learning_rate": 1.1360804726576543e-06,
+      "loss": 0.0603,
+      "step": 12835
+    },
+    {
+      "epoch": 2.0797148412184057,
+      "grad_norm": 0.8376288414001465,
+      "learning_rate": 1.1357140035297745e-06,
+      "loss": 0.0638,
+      "step": 12836
+    },
+    {
+      "epoch": 2.0798768632534026,
+      "grad_norm": 0.8386695981025696,
+      "learning_rate": 1.1353475761457445e-06,
+      "loss": 0.057,
+      "step": 12837
+    },
+    {
+      "epoch": 2.080038885288399,
+      "grad_norm": 0.8405042886734009,
+      "learning_rate": 1.1349811905167762e-06,
+      "loss": 0.066,
+      "step": 12838
+    },
+    {
+      "epoch": 2.080200907323396,
+      "grad_norm": 0.9115575551986694,
+      "learning_rate": 1.13461484665408e-06,
+      "loss": 0.0623,
+      "step": 12839
+    },
+    {
+      "epoch": 2.080362929358393,
+      "grad_norm": 0.7548916339874268,
+      "learning_rate": 1.134248544568867e-06,
+      "loss": 0.0554,
+      "step": 12840
+    },
+    {
+      "epoch": 2.0805249513933894,
+      "grad_norm": 0.873468816280365,
+      "learning_rate": 1.1338822842723418e-06,
+      "loss": 0.0674,
+      "step": 12841
+    },
+    {
+      "epoch": 2.0806869734283864,
+      "grad_norm": 0.9145466685295105,
+      "learning_rate": 1.1335160657757121e-06,
+      "loss": 0.0608,
+      "step": 12842
+    },
+    {
+      "epoch": 2.080848995463383,
+      "grad_norm": 0.8636155724525452,
+      "learning_rate": 1.1331498890901851e-06,
+      "loss": 0.064,
+      "step": 12843
+    },
+    {
+      "epoch": 2.08101101749838,
+      "grad_norm": 0.9404274821281433,
+      "learning_rate": 1.1327837542269645e-06,
+      "loss": 0.0603,
+      "step": 12844
+    },
+    {
+      "epoch": 2.0811730395333767,
+      "grad_norm": 0.867577850818634,
+      "learning_rate": 1.1324176611972515e-06,
+      "loss": 0.0659,
+      "step": 12845
+    },
+    {
+      "epoch": 2.0813350615683732,
+      "grad_norm": 0.7574745416641235,
+      "learning_rate": 1.1320516100122487e-06,
+      "loss": 0.0572,
+      "step": 12846
+    },
+    {
+      "epoch": 2.08149708360337,
+      "grad_norm": 0.8237707614898682,
+      "learning_rate": 1.1316856006831562e-06,
+      "loss": 0.0643,
+      "step": 12847
+    },
+    {
+      "epoch": 2.0816591056383666,
+      "grad_norm": 0.950802743434906,
+      "learning_rate": 1.1313196332211728e-06,
+      "loss": 0.0632,
+      "step": 12848
+    },
+    {
+      "epoch": 2.0818211276733636,
+      "grad_norm": 0.9620135426521301,
+      "learning_rate": 1.1309537076374968e-06,
+      "loss": 0.0682,
+      "step": 12849
+    },
+    {
+      "epoch": 2.0819831497083605,
+      "grad_norm": 0.9418061375617981,
+      "learning_rate": 1.130587823943324e-06,
+      "loss": 0.0629,
+      "step": 12850
+    },
+    {
+      "epoch": 2.082145171743357,
+      "grad_norm": 0.8530777096748352,
+      "learning_rate": 1.1302219821498502e-06,
+      "loss": 0.0565,
+      "step": 12851
+    },
+    {
+      "epoch": 2.082307193778354,
+      "grad_norm": 0.8891876339912415,
+      "learning_rate": 1.1298561822682687e-06,
+      "loss": 0.0709,
+      "step": 12852
+    },
+    {
+      "epoch": 2.0824692158133504,
+      "grad_norm": 0.9837715029716492,
+      "learning_rate": 1.1294904243097726e-06,
+      "loss": 0.0619,
+      "step": 12853
+    },
+    {
+      "epoch": 2.0826312378483474,
+      "grad_norm": 0.8900513648986816,
+      "learning_rate": 1.1291247082855528e-06,
+      "loss": 0.0661,
+      "step": 12854
+    },
+    {
+      "epoch": 2.0827932598833443,
+      "grad_norm": 0.8813618421554565,
+      "learning_rate": 1.1287590342068005e-06,
+      "loss": 0.0588,
+      "step": 12855
+    },
+    {
+      "epoch": 2.082955281918341,
+      "grad_norm": 0.7814889550209045,
+      "learning_rate": 1.1283934020847015e-06,
+      "loss": 0.0516,
+      "step": 12856
+    },
+    {
+      "epoch": 2.0831173039533377,
+      "grad_norm": 0.7788052558898926,
+      "learning_rate": 1.128027811930446e-06,
+      "loss": 0.0581,
+      "step": 12857
+    },
+    {
+      "epoch": 2.083279325988334,
+      "grad_norm": 0.8594955205917358,
+      "learning_rate": 1.1276622637552203e-06,
+      "loss": 0.0575,
+      "step": 12858
+    },
+    {
+      "epoch": 2.083441348023331,
+      "grad_norm": 0.8879890441894531,
+      "learning_rate": 1.1272967575702075e-06,
+      "loss": 0.0587,
+      "step": 12859
+    },
+    {
+      "epoch": 2.083603370058328,
+      "grad_norm": 1.0055869817733765,
+      "learning_rate": 1.126931293386592e-06,
+      "loss": 0.0642,
+      "step": 12860
+    },
+    {
+      "epoch": 2.0837653920933246,
+      "grad_norm": 1.0345205068588257,
+      "learning_rate": 1.1265658712155552e-06,
+      "loss": 0.0691,
+      "step": 12861
+    },
+    {
+      "epoch": 2.0839274141283215,
+      "grad_norm": 0.8441528081893921,
+      "learning_rate": 1.1262004910682811e-06,
+      "loss": 0.0597,
+      "step": 12862
+    },
+    {
+      "epoch": 2.0840894361633184,
+      "grad_norm": 0.9225409030914307,
+      "learning_rate": 1.1258351529559463e-06,
+      "loss": 0.0598,
+      "step": 12863
+    },
+    {
+      "epoch": 2.084251458198315,
+      "grad_norm": 0.9410536885261536,
+      "learning_rate": 1.1254698568897308e-06,
+      "loss": 0.0637,
+      "step": 12864
+    },
+    {
+      "epoch": 2.084413480233312,
+      "grad_norm": 0.9590262174606323,
+      "learning_rate": 1.1251046028808107e-06,
+      "loss": 0.0622,
+      "step": 12865
+    },
+    {
+      "epoch": 2.0845755022683083,
+      "grad_norm": 0.944026529788971,
+      "learning_rate": 1.124739390940363e-06,
+      "loss": 0.0729,
+      "step": 12866
+    },
+    {
+      "epoch": 2.0847375243033053,
+      "grad_norm": 0.8297381401062012,
+      "learning_rate": 1.124374221079562e-06,
+      "loss": 0.0582,
+      "step": 12867
+    },
+    {
+      "epoch": 2.084899546338302,
+      "grad_norm": 0.8802173137664795,
+      "learning_rate": 1.1240090933095806e-06,
+      "loss": 0.0617,
+      "step": 12868
+    },
+    {
+      "epoch": 2.0850615683732987,
+      "grad_norm": 0.8409082889556885,
+      "learning_rate": 1.1236440076415923e-06,
+      "loss": 0.0575,
+      "step": 12869
+    },
+    {
+      "epoch": 2.0852235904082956,
+      "grad_norm": 0.8129790425300598,
+      "learning_rate": 1.1232789640867644e-06,
+      "loss": 0.0586,
+      "step": 12870
+    },
+    {
+      "epoch": 2.085385612443292,
+      "grad_norm": 0.9125465750694275,
+      "learning_rate": 1.1229139626562698e-06,
+      "loss": 0.0634,
+      "step": 12871
+    },
+    {
+      "epoch": 2.085547634478289,
+      "grad_norm": 0.8047336935997009,
+      "learning_rate": 1.1225490033612755e-06,
+      "loss": 0.0589,
+      "step": 12872
+    },
+    {
+      "epoch": 2.085709656513286,
+      "grad_norm": 0.9558115005493164,
+      "learning_rate": 1.1221840862129493e-06,
+      "loss": 0.0637,
+      "step": 12873
+    },
+    {
+      "epoch": 2.0858716785482825,
+      "grad_norm": 0.873754620552063,
+      "learning_rate": 1.1218192112224547e-06,
+      "loss": 0.0675,
+      "step": 12874
+    },
+    {
+      "epoch": 2.0860337005832794,
+      "grad_norm": 0.857243001461029,
+      "learning_rate": 1.1214543784009563e-06,
+      "loss": 0.065,
+      "step": 12875
+    },
+    {
+      "epoch": 2.086195722618276,
+      "grad_norm": 0.8386802673339844,
+      "learning_rate": 1.1210895877596195e-06,
+      "loss": 0.0564,
+      "step": 12876
+    },
+    {
+      "epoch": 2.086357744653273,
+      "grad_norm": 0.958071231842041,
+      "learning_rate": 1.1207248393096038e-06,
+      "loss": 0.0732,
+      "step": 12877
+    },
+    {
+      "epoch": 2.0865197666882698,
+      "grad_norm": 0.8746595978736877,
+      "learning_rate": 1.12036013306207e-06,
+      "loss": 0.0591,
+      "step": 12878
+    },
+    {
+      "epoch": 2.0866817887232663,
+      "grad_norm": 0.854904294013977,
+      "learning_rate": 1.1199954690281779e-06,
+      "loss": 0.0653,
+      "step": 12879
+    },
+    {
+      "epoch": 2.086843810758263,
+      "grad_norm": 0.8066260814666748,
+      "learning_rate": 1.1196308472190845e-06,
+      "loss": 0.0576,
+      "step": 12880
+    },
+    {
+      "epoch": 2.0870058327932597,
+      "grad_norm": 0.894486129283905,
+      "learning_rate": 1.1192662676459468e-06,
+      "loss": 0.0684,
+      "step": 12881
+    },
+    {
+      "epoch": 2.0871678548282566,
+      "grad_norm": 0.8854237198829651,
+      "learning_rate": 1.1189017303199198e-06,
+      "loss": 0.0656,
+      "step": 12882
+    },
+    {
+      "epoch": 2.0873298768632536,
+      "grad_norm": 0.8725504875183105,
+      "learning_rate": 1.1185372352521581e-06,
+      "loss": 0.0578,
+      "step": 12883
+    },
+    {
+      "epoch": 2.08749189889825,
+      "grad_norm": 0.8695107102394104,
+      "learning_rate": 1.1181727824538147e-06,
+      "loss": 0.0676,
+      "step": 12884
+    },
+    {
+      "epoch": 2.087653920933247,
+      "grad_norm": 0.9441255331039429,
+      "learning_rate": 1.117808371936038e-06,
+      "loss": 0.065,
+      "step": 12885
+    },
+    {
+      "epoch": 2.087815942968244,
+      "grad_norm": 0.9579770565032959,
+      "learning_rate": 1.1174440037099815e-06,
+      "loss": 0.0623,
+      "step": 12886
+    },
+    {
+      "epoch": 2.0879779650032404,
+      "grad_norm": 0.9786186218261719,
+      "learning_rate": 1.117079677786793e-06,
+      "loss": 0.0702,
+      "step": 12887
+    },
+    {
+      "epoch": 2.0881399870382373,
+      "grad_norm": 0.8708941340446472,
+      "learning_rate": 1.1167153941776205e-06,
+      "loss": 0.056,
+      "step": 12888
+    },
+    {
+      "epoch": 2.088302009073234,
+      "grad_norm": 0.8261350989341736,
+      "learning_rate": 1.1163511528936084e-06,
+      "loss": 0.0617,
+      "step": 12889
+    },
+    {
+      "epoch": 2.0884640311082308,
+      "grad_norm": 0.9200928211212158,
+      "learning_rate": 1.1159869539459018e-06,
+      "loss": 0.0622,
+      "step": 12890
+    },
+    {
+      "epoch": 2.0886260531432277,
+      "grad_norm": 1.0103826522827148,
+      "learning_rate": 1.1156227973456468e-06,
+      "loss": 0.0651,
+      "step": 12891
+    },
+    {
+      "epoch": 2.088788075178224,
+      "grad_norm": 1.0453431606292725,
+      "learning_rate": 1.1152586831039835e-06,
+      "loss": 0.0617,
+      "step": 12892
+    },
+    {
+      "epoch": 2.088950097213221,
+      "grad_norm": 0.8877750039100647,
+      "learning_rate": 1.1148946112320533e-06,
+      "loss": 0.06,
+      "step": 12893
+    },
+    {
+      "epoch": 2.0891121192482176,
+      "grad_norm": 0.9256067872047424,
+      "learning_rate": 1.1145305817409962e-06,
+      "loss": 0.0679,
+      "step": 12894
+    },
+    {
+      "epoch": 2.0892741412832145,
+      "grad_norm": 0.7883033752441406,
+      "learning_rate": 1.1141665946419506e-06,
+      "loss": 0.0532,
+      "step": 12895
+    },
+    {
+      "epoch": 2.0894361633182115,
+      "grad_norm": 0.9102799892425537,
+      "learning_rate": 1.1138026499460532e-06,
+      "loss": 0.0666,
+      "step": 12896
+    },
+    {
+      "epoch": 2.089598185353208,
+      "grad_norm": 0.9670751690864563,
+      "learning_rate": 1.1134387476644407e-06,
+      "loss": 0.0625,
+      "step": 12897
+    },
+    {
+      "epoch": 2.089760207388205,
+      "grad_norm": 0.8531269431114197,
+      "learning_rate": 1.1130748878082467e-06,
+      "loss": 0.0646,
+      "step": 12898
+    },
+    {
+      "epoch": 2.0899222294232014,
+      "grad_norm": 0.749323844909668,
+      "learning_rate": 1.1127110703886048e-06,
+      "loss": 0.0549,
+      "step": 12899
+    },
+    {
+      "epoch": 2.0900842514581983,
+      "grad_norm": 0.8951242566108704,
+      "learning_rate": 1.1123472954166473e-06,
+      "loss": 0.0562,
+      "step": 12900
+    },
+    {
+      "epoch": 2.0902462734931953,
+      "grad_norm": 0.8564949035644531,
+      "learning_rate": 1.1119835629035042e-06,
+      "loss": 0.0605,
+      "step": 12901
+    },
+    {
+      "epoch": 2.0904082955281917,
+      "grad_norm": 0.7760194540023804,
+      "learning_rate": 1.1116198728603061e-06,
+      "loss": 0.06,
+      "step": 12902
+    },
+    {
+      "epoch": 2.0905703175631887,
+      "grad_norm": 0.9799525141716003,
+      "learning_rate": 1.1112562252981793e-06,
+      "loss": 0.0622,
+      "step": 12903
+    },
+    {
+      "epoch": 2.090732339598185,
+      "grad_norm": 0.8512018918991089,
+      "learning_rate": 1.1108926202282505e-06,
+      "loss": 0.0646,
+      "step": 12904
+    },
+    {
+      "epoch": 2.090894361633182,
+      "grad_norm": 0.9717202186584473,
+      "learning_rate": 1.1105290576616467e-06,
+      "loss": 0.0642,
+      "step": 12905
+    },
+    {
+      "epoch": 2.091056383668179,
+      "grad_norm": 0.8135266304016113,
+      "learning_rate": 1.110165537609492e-06,
+      "loss": 0.0635,
+      "step": 12906
+    },
+    {
+      "epoch": 2.0912184057031755,
+      "grad_norm": 0.8754247426986694,
+      "learning_rate": 1.109802060082908e-06,
+      "loss": 0.0609,
+      "step": 12907
+    },
+    {
+      "epoch": 2.0913804277381725,
+      "grad_norm": 0.9074878692626953,
+      "learning_rate": 1.1094386250930164e-06,
+      "loss": 0.0601,
+      "step": 12908
+    },
+    {
+      "epoch": 2.091542449773169,
+      "grad_norm": 0.8210632801055908,
+      "learning_rate": 1.1090752326509379e-06,
+      "loss": 0.059,
+      "step": 12909
+    },
+    {
+      "epoch": 2.091704471808166,
+      "grad_norm": 0.9320157766342163,
+      "learning_rate": 1.1087118827677915e-06,
+      "loss": 0.0734,
+      "step": 12910
+    },
+    {
+      "epoch": 2.091866493843163,
+      "grad_norm": 0.8074702024459839,
+      "learning_rate": 1.1083485754546944e-06,
+      "loss": 0.0599,
+      "step": 12911
+    },
+    {
+      "epoch": 2.0920285158781593,
+      "grad_norm": 0.81490558385849,
+      "learning_rate": 1.1079853107227634e-06,
+      "loss": 0.0622,
+      "step": 12912
+    },
+    {
+      "epoch": 2.0921905379131562,
+      "grad_norm": 0.9869770407676697,
+      "learning_rate": 1.107622088583113e-06,
+      "loss": 0.0634,
+      "step": 12913
+    },
+    {
+      "epoch": 2.0923525599481527,
+      "grad_norm": 1.0530625581741333,
+      "learning_rate": 1.1072589090468571e-06,
+      "loss": 0.067,
+      "step": 12914
+    },
+    {
+      "epoch": 2.0925145819831497,
+      "grad_norm": 0.9498258233070374,
+      "learning_rate": 1.1068957721251085e-06,
+      "loss": 0.0615,
+      "step": 12915
+    },
+    {
+      "epoch": 2.0926766040181466,
+      "grad_norm": 0.9322092533111572,
+      "learning_rate": 1.1065326778289782e-06,
+      "loss": 0.0678,
+      "step": 12916
+    },
+    {
+      "epoch": 2.092838626053143,
+      "grad_norm": 1.2187130451202393,
+      "learning_rate": 1.1061696261695765e-06,
+      "loss": 0.0707,
+      "step": 12917
+    },
+    {
+      "epoch": 2.09300064808814,
+      "grad_norm": 0.8521288633346558,
+      "learning_rate": 1.1058066171580092e-06,
+      "loss": 0.0667,
+      "step": 12918
+    },
+    {
+      "epoch": 2.093162670123137,
+      "grad_norm": 0.9618637561798096,
+      "learning_rate": 1.1054436508053866e-06,
+      "loss": 0.0642,
+      "step": 12919
+    },
+    {
+      "epoch": 2.0933246921581334,
+      "grad_norm": 0.7557185888290405,
+      "learning_rate": 1.1050807271228146e-06,
+      "loss": 0.0558,
+      "step": 12920
+    },
+    {
+      "epoch": 2.0934867141931304,
+      "grad_norm": 0.9032410383224487,
+      "learning_rate": 1.1047178461213956e-06,
+      "loss": 0.0654,
+      "step": 12921
+    },
+    {
+      "epoch": 2.093648736228127,
+      "grad_norm": 0.813040018081665,
+      "learning_rate": 1.1043550078122342e-06,
+      "loss": 0.0575,
+      "step": 12922
+    },
+    {
+      "epoch": 2.093810758263124,
+      "grad_norm": 0.8633285164833069,
+      "learning_rate": 1.103992212206431e-06,
+      "loss": 0.061,
+      "step": 12923
+    },
+    {
+      "epoch": 2.0939727802981207,
+      "grad_norm": 0.9251546263694763,
+      "learning_rate": 1.1036294593150898e-06,
+      "loss": 0.0627,
+      "step": 12924
+    },
+    {
+      "epoch": 2.094134802333117,
+      "grad_norm": 0.7767398953437805,
+      "learning_rate": 1.103266749149307e-06,
+      "loss": 0.0583,
+      "step": 12925
+    },
+    {
+      "epoch": 2.094296824368114,
+      "grad_norm": 0.823419988155365,
+      "learning_rate": 1.1029040817201819e-06,
+      "loss": 0.0631,
+      "step": 12926
+    },
+    {
+      "epoch": 2.0944588464031106,
+      "grad_norm": 0.9466148018836975,
+      "learning_rate": 1.1025414570388108e-06,
+      "loss": 0.0655,
+      "step": 12927
+    },
+    {
+      "epoch": 2.0946208684381076,
+      "grad_norm": 0.8676579594612122,
+      "learning_rate": 1.1021788751162893e-06,
+      "loss": 0.0575,
+      "step": 12928
+    },
+    {
+      "epoch": 2.0947828904731045,
+      "grad_norm": 0.8906098008155823,
+      "learning_rate": 1.1018163359637116e-06,
+      "loss": 0.0659,
+      "step": 12929
+    },
+    {
+      "epoch": 2.094944912508101,
+      "grad_norm": 0.9030002951622009,
+      "learning_rate": 1.1014538395921704e-06,
+      "loss": 0.0685,
+      "step": 12930
+    },
+    {
+      "epoch": 2.095106934543098,
+      "grad_norm": 0.8031485676765442,
+      "learning_rate": 1.1010913860127572e-06,
+      "loss": 0.0627,
+      "step": 12931
+    },
+    {
+      "epoch": 2.0952689565780944,
+      "grad_norm": 0.8047374486923218,
+      "learning_rate": 1.1007289752365635e-06,
+      "loss": 0.0581,
+      "step": 12932
+    },
+    {
+      "epoch": 2.0954309786130914,
+      "grad_norm": 0.9250509738922119,
+      "learning_rate": 1.100366607274675e-06,
+      "loss": 0.0613,
+      "step": 12933
+    },
+    {
+      "epoch": 2.0955930006480883,
+      "grad_norm": 0.7186943292617798,
+      "learning_rate": 1.1000042821381823e-06,
+      "loss": 0.0544,
+      "step": 12934
+    },
+    {
+      "epoch": 2.095755022683085,
+      "grad_norm": 0.7666640281677246,
+      "learning_rate": 1.0996419998381713e-06,
+      "loss": 0.0594,
+      "step": 12935
+    },
+    {
+      "epoch": 2.0959170447180817,
+      "grad_norm": 0.913008987903595,
+      "learning_rate": 1.0992797603857257e-06,
+      "loss": 0.0621,
+      "step": 12936
+    },
+    {
+      "epoch": 2.096079066753078,
+      "grad_norm": 0.9413930177688599,
+      "learning_rate": 1.0989175637919297e-06,
+      "loss": 0.067,
+      "step": 12937
+    },
+    {
+      "epoch": 2.096241088788075,
+      "grad_norm": 0.8046421408653259,
+      "learning_rate": 1.0985554100678647e-06,
+      "loss": 0.057,
+      "step": 12938
+    },
+    {
+      "epoch": 2.096403110823072,
+      "grad_norm": 0.99111407995224,
+      "learning_rate": 1.0981932992246144e-06,
+      "loss": 0.0691,
+      "step": 12939
+    },
+    {
+      "epoch": 2.0965651328580686,
+      "grad_norm": 0.8477975130081177,
+      "learning_rate": 1.0978312312732562e-06,
+      "loss": 0.0584,
+      "step": 12940
+    },
+    {
+      "epoch": 2.0967271548930655,
+      "grad_norm": 0.8106995820999146,
+      "learning_rate": 1.097469206224869e-06,
+      "loss": 0.0588,
+      "step": 12941
+    },
+    {
+      "epoch": 2.0968891769280624,
+      "grad_norm": 0.8158612847328186,
+      "learning_rate": 1.09710722409053e-06,
+      "loss": 0.0622,
+      "step": 12942
+    },
+    {
+      "epoch": 2.097051198963059,
+      "grad_norm": 0.8949916362762451,
+      "learning_rate": 1.096745284881315e-06,
+      "loss": 0.0689,
+      "step": 12943
+    },
+    {
+      "epoch": 2.097213220998056,
+      "grad_norm": 0.8782987594604492,
+      "learning_rate": 1.0963833886082987e-06,
+      "loss": 0.0649,
+      "step": 12944
+    },
+    {
+      "epoch": 2.0973752430330523,
+      "grad_norm": 0.9232782125473022,
+      "learning_rate": 1.0960215352825537e-06,
+      "loss": 0.0711,
+      "step": 12945
+    },
+    {
+      "epoch": 2.0975372650680493,
+      "grad_norm": 0.8627330660820007,
+      "learning_rate": 1.0956597249151532e-06,
+      "loss": 0.0611,
+      "step": 12946
+    },
+    {
+      "epoch": 2.097699287103046,
+      "grad_norm": 0.779122531414032,
+      "learning_rate": 1.0952979575171649e-06,
+      "loss": 0.0556,
+      "step": 12947
+    },
+    {
+      "epoch": 2.0978613091380427,
+      "grad_norm": 1.0145422220230103,
+      "learning_rate": 1.0949362330996605e-06,
+      "loss": 0.0594,
+      "step": 12948
+    },
+    {
+      "epoch": 2.0980233311730396,
+      "grad_norm": 0.8492004871368408,
+      "learning_rate": 1.0945745516737075e-06,
+      "loss": 0.0625,
+      "step": 12949
+    },
+    {
+      "epoch": 2.098185353208036,
+      "grad_norm": 0.8178960680961609,
+      "learning_rate": 1.094212913250373e-06,
+      "loss": 0.053,
+      "step": 12950
+    },
+    {
+      "epoch": 2.098347375243033,
+      "grad_norm": 0.9389135837554932,
+      "learning_rate": 1.0938513178407201e-06,
+      "loss": 0.0631,
+      "step": 12951
+    },
+    {
+      "epoch": 2.09850939727803,
+      "grad_norm": 0.9091375470161438,
+      "learning_rate": 1.0934897654558134e-06,
+      "loss": 0.0636,
+      "step": 12952
+    },
+    {
+      "epoch": 2.0986714193130265,
+      "grad_norm": 0.8559199571609497,
+      "learning_rate": 1.0931282561067183e-06,
+      "loss": 0.0569,
+      "step": 12953
+    },
+    {
+      "epoch": 2.0988334413480234,
+      "grad_norm": 0.8254791498184204,
+      "learning_rate": 1.0927667898044927e-06,
+      "loss": 0.0629,
+      "step": 12954
+    },
+    {
+      "epoch": 2.09899546338302,
+      "grad_norm": 0.7848954796791077,
+      "learning_rate": 1.092405366560198e-06,
+      "loss": 0.0606,
+      "step": 12955
+    },
+    {
+      "epoch": 2.099157485418017,
+      "grad_norm": 0.9098137021064758,
+      "learning_rate": 1.092043986384893e-06,
+      "loss": 0.0613,
+      "step": 12956
+    },
+    {
+      "epoch": 2.0993195074530138,
+      "grad_norm": 0.7770287990570068,
+      "learning_rate": 1.0916826492896346e-06,
+      "loss": 0.0554,
+      "step": 12957
+    },
+    {
+      "epoch": 2.0994815294880103,
+      "grad_norm": 0.8795650005340576,
+      "learning_rate": 1.091321355285479e-06,
+      "loss": 0.0574,
+      "step": 12958
+    },
+    {
+      "epoch": 2.099643551523007,
+      "grad_norm": 1.0702177286148071,
+      "learning_rate": 1.0909601043834812e-06,
+      "loss": 0.0708,
+      "step": 12959
+    },
+    {
+      "epoch": 2.0998055735580037,
+      "grad_norm": 0.8428245782852173,
+      "learning_rate": 1.0905988965946942e-06,
+      "loss": 0.0629,
+      "step": 12960
+    },
+    {
+      "epoch": 2.0999675955930006,
+      "grad_norm": 0.8084932565689087,
+      "learning_rate": 1.0902377319301704e-06,
+      "loss": 0.0526,
+      "step": 12961
+    },
+    {
+      "epoch": 2.1001296176279975,
+      "grad_norm": 0.9167864918708801,
+      "learning_rate": 1.0898766104009606e-06,
+      "loss": 0.0617,
+      "step": 12962
+    },
+    {
+      "epoch": 2.100291639662994,
+      "grad_norm": 0.9969832301139832,
+      "learning_rate": 1.089515532018114e-06,
+      "loss": 0.0737,
+      "step": 12963
+    },
+    {
+      "epoch": 2.100453661697991,
+      "grad_norm": 1.0059758424758911,
+      "learning_rate": 1.0891544967926795e-06,
+      "loss": 0.068,
+      "step": 12964
+    },
+    {
+      "epoch": 2.100615683732988,
+      "grad_norm": 0.7887366414070129,
+      "learning_rate": 1.0887935047357023e-06,
+      "loss": 0.0574,
+      "step": 12965
+    },
+    {
+      "epoch": 2.1007777057679844,
+      "grad_norm": 0.8655834794044495,
+      "learning_rate": 1.0884325558582283e-06,
+      "loss": 0.0611,
+      "step": 12966
+    },
+    {
+      "epoch": 2.1009397278029813,
+      "grad_norm": 0.8879144191741943,
+      "learning_rate": 1.0880716501713025e-06,
+      "loss": 0.0605,
+      "step": 12967
+    },
+    {
+      "epoch": 2.101101749837978,
+      "grad_norm": 0.7682301998138428,
+      "learning_rate": 1.0877107876859688e-06,
+      "loss": 0.0619,
+      "step": 12968
+    },
+    {
+      "epoch": 2.1012637718729748,
+      "grad_norm": 1.0318995714187622,
+      "learning_rate": 1.0873499684132663e-06,
+      "loss": 0.0671,
+      "step": 12969
+    },
+    {
+      "epoch": 2.1014257939079717,
+      "grad_norm": 0.8464838862419128,
+      "learning_rate": 1.086989192364236e-06,
+      "loss": 0.0651,
+      "step": 12970
+    },
+    {
+      "epoch": 2.101587815942968,
+      "grad_norm": 1.0440428256988525,
+      "learning_rate": 1.0866284595499172e-06,
+      "loss": 0.068,
+      "step": 12971
+    },
+    {
+      "epoch": 2.101749837977965,
+      "grad_norm": 0.8565576076507568,
+      "learning_rate": 1.0862677699813471e-06,
+      "loss": 0.0539,
+      "step": 12972
+    },
+    {
+      "epoch": 2.1019118600129616,
+      "grad_norm": 0.86493980884552,
+      "learning_rate": 1.0859071236695623e-06,
+      "loss": 0.0623,
+      "step": 12973
+    },
+    {
+      "epoch": 2.1020738820479585,
+      "grad_norm": 0.8555430173873901,
+      "learning_rate": 1.0855465206255972e-06,
+      "loss": 0.0614,
+      "step": 12974
+    },
+    {
+      "epoch": 2.1022359040829555,
+      "grad_norm": 1.089228630065918,
+      "learning_rate": 1.0851859608604858e-06,
+      "loss": 0.0658,
+      "step": 12975
+    },
+    {
+      "epoch": 2.102397926117952,
+      "grad_norm": 0.8358057141304016,
+      "learning_rate": 1.0848254443852602e-06,
+      "loss": 0.0618,
+      "step": 12976
+    },
+    {
+      "epoch": 2.102559948152949,
+      "grad_norm": 0.9118403792381287,
+      "learning_rate": 1.0844649712109515e-06,
+      "loss": 0.0661,
+      "step": 12977
+    },
+    {
+      "epoch": 2.1027219701879454,
+      "grad_norm": 0.7774009704589844,
+      "learning_rate": 1.084104541348589e-06,
+      "loss": 0.06,
+      "step": 12978
+    },
+    {
+      "epoch": 2.1028839922229423,
+      "grad_norm": 0.921910285949707,
+      "learning_rate": 1.083744154809202e-06,
+      "loss": 0.0647,
+      "step": 12979
+    },
+    {
+      "epoch": 2.1030460142579392,
+      "grad_norm": 0.8106131553649902,
+      "learning_rate": 1.0833838116038156e-06,
+      "loss": 0.0586,
+      "step": 12980
+    },
+    {
+      "epoch": 2.1032080362929357,
+      "grad_norm": 0.8339465856552124,
+      "learning_rate": 1.0830235117434557e-06,
+      "loss": 0.062,
+      "step": 12981
+    },
+    {
+      "epoch": 2.1033700583279327,
+      "grad_norm": 0.9177747368812561,
+      "learning_rate": 1.0826632552391484e-06,
+      "loss": 0.0655,
+      "step": 12982
+    },
+    {
+      "epoch": 2.103532080362929,
+      "grad_norm": 0.8722216486930847,
+      "learning_rate": 1.0823030421019163e-06,
+      "loss": 0.062,
+      "step": 12983
+    },
+    {
+      "epoch": 2.103694102397926,
+      "grad_norm": 0.8818717002868652,
+      "learning_rate": 1.081942872342779e-06,
+      "loss": 0.0585,
+      "step": 12984
+    },
+    {
+      "epoch": 2.103856124432923,
+      "grad_norm": 0.8914388418197632,
+      "learning_rate": 1.081582745972758e-06,
+      "loss": 0.066,
+      "step": 12985
+    },
+    {
+      "epoch": 2.1040181464679195,
+      "grad_norm": 0.9045519828796387,
+      "learning_rate": 1.0812226630028738e-06,
+      "loss": 0.0622,
+      "step": 12986
+    },
+    {
+      "epoch": 2.1041801685029164,
+      "grad_norm": 1.0446083545684814,
+      "learning_rate": 1.080862623444142e-06,
+      "loss": 0.0668,
+      "step": 12987
+    },
+    {
+      "epoch": 2.1043421905379134,
+      "grad_norm": 0.7879253029823303,
+      "learning_rate": 1.0805026273075797e-06,
+      "loss": 0.0604,
+      "step": 12988
+    },
+    {
+      "epoch": 2.10450421257291,
+      "grad_norm": 0.8485068082809448,
+      "learning_rate": 1.0801426746042018e-06,
+      "loss": 0.0651,
+      "step": 12989
+    },
+    {
+      "epoch": 2.104666234607907,
+      "grad_norm": 0.8506676554679871,
+      "learning_rate": 1.0797827653450222e-06,
+      "loss": 0.0647,
+      "step": 12990
+    },
+    {
+      "epoch": 2.1048282566429033,
+      "grad_norm": 0.8726806640625,
+      "learning_rate": 1.079422899541053e-06,
+      "loss": 0.0626,
+      "step": 12991
+    },
+    {
+      "epoch": 2.1049902786779002,
+      "grad_norm": 0.9030711650848389,
+      "learning_rate": 1.0790630772033057e-06,
+      "loss": 0.0698,
+      "step": 12992
+    },
+    {
+      "epoch": 2.105152300712897,
+      "grad_norm": 0.8381808400154114,
+      "learning_rate": 1.0787032983427892e-06,
+      "loss": 0.0691,
+      "step": 12993
+    },
+    {
+      "epoch": 2.1053143227478937,
+      "grad_norm": 0.9773826003074646,
+      "learning_rate": 1.0783435629705134e-06,
+      "loss": 0.0644,
+      "step": 12994
+    },
+    {
+      "epoch": 2.1054763447828906,
+      "grad_norm": 0.8287825584411621,
+      "learning_rate": 1.0779838710974822e-06,
+      "loss": 0.0559,
+      "step": 12995
+    },
+    {
+      "epoch": 2.105638366817887,
+      "grad_norm": 0.862917423248291,
+      "learning_rate": 1.0776242227347044e-06,
+      "loss": 0.0587,
+      "step": 12996
+    },
+    {
+      "epoch": 2.105800388852884,
+      "grad_norm": 0.8506877422332764,
+      "learning_rate": 1.0772646178931843e-06,
+      "loss": 0.0565,
+      "step": 12997
+    },
+    {
+      "epoch": 2.105962410887881,
+      "grad_norm": 0.8896209001541138,
+      "learning_rate": 1.0769050565839228e-06,
+      "loss": 0.0658,
+      "step": 12998
+    },
+    {
+      "epoch": 2.1061244329228774,
+      "grad_norm": 0.9213668704032898,
+      "learning_rate": 1.076545538817923e-06,
+      "loss": 0.0604,
+      "step": 12999
+    },
+    {
+      "epoch": 2.1062864549578744,
+      "grad_norm": 0.9148227572441101,
+      "learning_rate": 1.0761860646061838e-06,
+      "loss": 0.0712,
+      "step": 13000
+    },
+    {
+      "epoch": 2.106448476992871,
+      "grad_norm": 0.8725471496582031,
+      "learning_rate": 1.0758266339597077e-06,
+      "loss": 0.0651,
+      "step": 13001
+    },
+    {
+      "epoch": 2.106610499027868,
+      "grad_norm": 0.9468567967414856,
+      "learning_rate": 1.0754672468894889e-06,
+      "loss": 0.0638,
+      "step": 13002
+    },
+    {
+      "epoch": 2.1067725210628647,
+      "grad_norm": 0.9501492977142334,
+      "learning_rate": 1.075107903406525e-06,
+      "loss": 0.0691,
+      "step": 13003
+    },
+    {
+      "epoch": 2.106934543097861,
+      "grad_norm": 0.8555724024772644,
+      "learning_rate": 1.0747486035218116e-06,
+      "loss": 0.0641,
+      "step": 13004
+    },
+    {
+      "epoch": 2.107096565132858,
+      "grad_norm": 1.0084507465362549,
+      "learning_rate": 1.0743893472463416e-06,
+      "loss": 0.0585,
+      "step": 13005
+    },
+    {
+      "epoch": 2.1072585871678546,
+      "grad_norm": 0.8593863844871521,
+      "learning_rate": 1.0740301345911075e-06,
+      "loss": 0.0604,
+      "step": 13006
+    },
+    {
+      "epoch": 2.1074206092028516,
+      "grad_norm": 0.9241552948951721,
+      "learning_rate": 1.0736709655671004e-06,
+      "loss": 0.0619,
+      "step": 13007
+    },
+    {
+      "epoch": 2.1075826312378485,
+      "grad_norm": 0.9510383009910583,
+      "learning_rate": 1.0733118401853112e-06,
+      "loss": 0.0679,
+      "step": 13008
+    },
+    {
+      "epoch": 2.107744653272845,
+      "grad_norm": 0.8677833676338196,
+      "learning_rate": 1.0729527584567255e-06,
+      "loss": 0.0654,
+      "step": 13009
+    },
+    {
+      "epoch": 2.107906675307842,
+      "grad_norm": 0.8571850061416626,
+      "learning_rate": 1.0725937203923327e-06,
+      "loss": 0.0565,
+      "step": 13010
+    },
+    {
+      "epoch": 2.1080686973428384,
+      "grad_norm": 0.7107164263725281,
+      "learning_rate": 1.072234726003118e-06,
+      "loss": 0.053,
+      "step": 13011
+    },
+    {
+      "epoch": 2.1082307193778353,
+      "grad_norm": 0.9328707456588745,
+      "learning_rate": 1.0718757753000665e-06,
+      "loss": 0.0591,
+      "step": 13012
+    },
+    {
+      "epoch": 2.1083927414128323,
+      "grad_norm": 0.8621041178703308,
+      "learning_rate": 1.0715168682941591e-06,
+      "loss": 0.0615,
+      "step": 13013
+    },
+    {
+      "epoch": 2.1085547634478288,
+      "grad_norm": 0.8346017599105835,
+      "learning_rate": 1.071158004996378e-06,
+      "loss": 0.063,
+      "step": 13014
+    },
+    {
+      "epoch": 2.1087167854828257,
+      "grad_norm": 0.8426311016082764,
+      "learning_rate": 1.0707991854177057e-06,
+      "loss": 0.06,
+      "step": 13015
+    },
+    {
+      "epoch": 2.108878807517822,
+      "grad_norm": 0.9094135761260986,
+      "learning_rate": 1.070440409569119e-06,
+      "loss": 0.0647,
+      "step": 13016
+    },
+    {
+      "epoch": 2.109040829552819,
+      "grad_norm": 1.048732042312622,
+      "learning_rate": 1.0700816774615964e-06,
+      "loss": 0.0622,
+      "step": 13017
+    },
+    {
+      "epoch": 2.109202851587816,
+      "grad_norm": 0.9794455170631409,
+      "learning_rate": 1.0697229891061141e-06,
+      "loss": 0.0608,
+      "step": 13018
+    },
+    {
+      "epoch": 2.1093648736228126,
+      "grad_norm": 0.8779764771461487,
+      "learning_rate": 1.069364344513647e-06,
+      "loss": 0.0625,
+      "step": 13019
+    },
+    {
+      "epoch": 2.1095268956578095,
+      "grad_norm": 0.8880550861358643,
+      "learning_rate": 1.0690057436951689e-06,
+      "loss": 0.0581,
+      "step": 13020
+    },
+    {
+      "epoch": 2.1096889176928064,
+      "grad_norm": 0.8012394309043884,
+      "learning_rate": 1.0686471866616525e-06,
+      "loss": 0.0571,
+      "step": 13021
+    },
+    {
+      "epoch": 2.109850939727803,
+      "grad_norm": 0.9028552770614624,
+      "learning_rate": 1.068288673424068e-06,
+      "loss": 0.0609,
+      "step": 13022
+    },
+    {
+      "epoch": 2.1100129617628,
+      "grad_norm": 0.8535613417625427,
+      "learning_rate": 1.0679302039933864e-06,
+      "loss": 0.0554,
+      "step": 13023
+    },
+    {
+      "epoch": 2.1101749837977963,
+      "grad_norm": 0.8417321443557739,
+      "learning_rate": 1.067571778380573e-06,
+      "loss": 0.0549,
+      "step": 13024
+    },
+    {
+      "epoch": 2.1103370058327933,
+      "grad_norm": 0.7988325357437134,
+      "learning_rate": 1.0672133965965981e-06,
+      "loss": 0.0567,
+      "step": 13025
+    },
+    {
+      "epoch": 2.11049902786779,
+      "grad_norm": 1.0051651000976562,
+      "learning_rate": 1.0668550586524256e-06,
+      "loss": 0.0633,
+      "step": 13026
+    },
+    {
+      "epoch": 2.1106610499027867,
+      "grad_norm": 0.9682597517967224,
+      "learning_rate": 1.0664967645590212e-06,
+      "loss": 0.0544,
+      "step": 13027
+    },
+    {
+      "epoch": 2.1108230719377836,
+      "grad_norm": 0.841816782951355,
+      "learning_rate": 1.066138514327345e-06,
+      "loss": 0.0603,
+      "step": 13028
+    },
+    {
+      "epoch": 2.11098509397278,
+      "grad_norm": 0.9751487970352173,
+      "learning_rate": 1.0657803079683616e-06,
+      "loss": 0.0612,
+      "step": 13029
+    },
+    {
+      "epoch": 2.111147116007777,
+      "grad_norm": 0.8450110554695129,
+      "learning_rate": 1.0654221454930305e-06,
+      "loss": 0.0551,
+      "step": 13030
+    },
+    {
+      "epoch": 2.111309138042774,
+      "grad_norm": 0.9710997343063354,
+      "learning_rate": 1.0650640269123095e-06,
+      "loss": 0.0593,
+      "step": 13031
+    },
+    {
+      "epoch": 2.1114711600777705,
+      "grad_norm": 0.8456966280937195,
+      "learning_rate": 1.0647059522371565e-06,
+      "loss": 0.0608,
+      "step": 13032
+    },
+    {
+      "epoch": 2.1116331821127674,
+      "grad_norm": 1.07785964012146,
+      "learning_rate": 1.0643479214785283e-06,
+      "loss": 0.0738,
+      "step": 13033
+    },
+    {
+      "epoch": 2.111795204147764,
+      "grad_norm": 1.0498229265213013,
+      "learning_rate": 1.0639899346473792e-06,
+      "loss": 0.0668,
+      "step": 13034
+    },
+    {
+      "epoch": 2.111957226182761,
+      "grad_norm": 0.8354172706604004,
+      "learning_rate": 1.0636319917546631e-06,
+      "loss": 0.0569,
+      "step": 13035
+    },
+    {
+      "epoch": 2.1121192482177578,
+      "grad_norm": 0.8369887471199036,
+      "learning_rate": 1.0632740928113323e-06,
+      "loss": 0.0584,
+      "step": 13036
+    },
+    {
+      "epoch": 2.1122812702527543,
+      "grad_norm": 0.9104105830192566,
+      "learning_rate": 1.0629162378283372e-06,
+      "loss": 0.0655,
+      "step": 13037
+    },
+    {
+      "epoch": 2.112443292287751,
+      "grad_norm": 0.890618622303009,
+      "learning_rate": 1.0625584268166278e-06,
+      "loss": 0.0622,
+      "step": 13038
+    },
+    {
+      "epoch": 2.1126053143227477,
+      "grad_norm": 0.9149554967880249,
+      "learning_rate": 1.062200659787152e-06,
+      "loss": 0.067,
+      "step": 13039
+    },
+    {
+      "epoch": 2.1127673363577446,
+      "grad_norm": 0.8385174870491028,
+      "learning_rate": 1.0618429367508564e-06,
+      "loss": 0.0558,
+      "step": 13040
+    },
+    {
+      "epoch": 2.1129293583927415,
+      "grad_norm": 0.7837886214256287,
+      "learning_rate": 1.0614852577186877e-06,
+      "loss": 0.0578,
+      "step": 13041
+    },
+    {
+      "epoch": 2.113091380427738,
+      "grad_norm": 0.9414006471633911,
+      "learning_rate": 1.061127622701588e-06,
+      "loss": 0.0681,
+      "step": 13042
+    },
+    {
+      "epoch": 2.113253402462735,
+      "grad_norm": 0.8170459270477295,
+      "learning_rate": 1.0607700317105002e-06,
+      "loss": 0.0576,
+      "step": 13043
+    },
+    {
+      "epoch": 2.113415424497732,
+      "grad_norm": 0.7887612581253052,
+      "learning_rate": 1.0604124847563674e-06,
+      "loss": 0.0565,
+      "step": 13044
+    },
+    {
+      "epoch": 2.1135774465327284,
+      "grad_norm": 0.9711698293685913,
+      "learning_rate": 1.0600549818501298e-06,
+      "loss": 0.0656,
+      "step": 13045
+    },
+    {
+      "epoch": 2.1137394685677253,
+      "grad_norm": 1.077919363975525,
+      "learning_rate": 1.0596975230027243e-06,
+      "loss": 0.0674,
+      "step": 13046
+    },
+    {
+      "epoch": 2.113901490602722,
+      "grad_norm": 0.9038639068603516,
+      "learning_rate": 1.059340108225089e-06,
+      "loss": 0.0589,
+      "step": 13047
+    },
+    {
+      "epoch": 2.1140635126377187,
+      "grad_norm": 0.9091435074806213,
+      "learning_rate": 1.05898273752816e-06,
+      "loss": 0.068,
+      "step": 13048
+    },
+    {
+      "epoch": 2.1142255346727157,
+      "grad_norm": 0.8895291686058044,
+      "learning_rate": 1.0586254109228722e-06,
+      "loss": 0.0593,
+      "step": 13049
+    },
+    {
+      "epoch": 2.114387556707712,
+      "grad_norm": 0.8643745183944702,
+      "learning_rate": 1.0582681284201587e-06,
+      "loss": 0.0627,
+      "step": 13050
+    },
+    {
+      "epoch": 2.114549578742709,
+      "grad_norm": 0.9382444024085999,
+      "learning_rate": 1.0579108900309518e-06,
+      "loss": 0.0609,
+      "step": 13051
+    },
+    {
+      "epoch": 2.1147116007777056,
+      "grad_norm": 0.8059027791023254,
+      "learning_rate": 1.0575536957661814e-06,
+      "loss": 0.061,
+      "step": 13052
+    },
+    {
+      "epoch": 2.1148736228127025,
+      "grad_norm": 0.8786547780036926,
+      "learning_rate": 1.0571965456367774e-06,
+      "loss": 0.0579,
+      "step": 13053
+    },
+    {
+      "epoch": 2.1150356448476995,
+      "grad_norm": 0.7798833847045898,
+      "learning_rate": 1.056839439653668e-06,
+      "loss": 0.0528,
+      "step": 13054
+    },
+    {
+      "epoch": 2.115197666882696,
+      "grad_norm": 0.9558907151222229,
+      "learning_rate": 1.056482377827779e-06,
+      "loss": 0.0664,
+      "step": 13055
+    },
+    {
+      "epoch": 2.115359688917693,
+      "grad_norm": 0.9504494667053223,
+      "learning_rate": 1.056125360170037e-06,
+      "loss": 0.0659,
+      "step": 13056
+    },
+    {
+      "epoch": 2.1155217109526894,
+      "grad_norm": 1.01918363571167,
+      "learning_rate": 1.055768386691363e-06,
+      "loss": 0.0646,
+      "step": 13057
+    },
+    {
+      "epoch": 2.1156837329876863,
+      "grad_norm": 0.8410986065864563,
+      "learning_rate": 1.0554114574026823e-06,
+      "loss": 0.0581,
+      "step": 13058
+    },
+    {
+      "epoch": 2.1158457550226832,
+      "grad_norm": 1.0521918535232544,
+      "learning_rate": 1.055054572314916e-06,
+      "loss": 0.0752,
+      "step": 13059
+    },
+    {
+      "epoch": 2.1160077770576797,
+      "grad_norm": 0.847416877746582,
+      "learning_rate": 1.0546977314389822e-06,
+      "loss": 0.062,
+      "step": 13060
+    },
+    {
+      "epoch": 2.1161697990926767,
+      "grad_norm": 1.0600688457489014,
+      "learning_rate": 1.0543409347858001e-06,
+      "loss": 0.069,
+      "step": 13061
+    },
+    {
+      "epoch": 2.116331821127673,
+      "grad_norm": 0.8803973197937012,
+      "learning_rate": 1.0539841823662867e-06,
+      "loss": 0.0616,
+      "step": 13062
+    },
+    {
+      "epoch": 2.11649384316267,
+      "grad_norm": 0.8875924348831177,
+      "learning_rate": 1.053627474191359e-06,
+      "loss": 0.0583,
+      "step": 13063
+    },
+    {
+      "epoch": 2.116655865197667,
+      "grad_norm": 0.8933229446411133,
+      "learning_rate": 1.0532708102719303e-06,
+      "loss": 0.07,
+      "step": 13064
+    },
+    {
+      "epoch": 2.1168178872326635,
+      "grad_norm": 0.9175959825515747,
+      "learning_rate": 1.0529141906189133e-06,
+      "loss": 0.0598,
+      "step": 13065
+    },
+    {
+      "epoch": 2.1169799092676604,
+      "grad_norm": 0.8931822180747986,
+      "learning_rate": 1.0525576152432204e-06,
+      "loss": 0.0656,
+      "step": 13066
+    },
+    {
+      "epoch": 2.1171419313026574,
+      "grad_norm": 0.9046509265899658,
+      "learning_rate": 1.0522010841557615e-06,
+      "loss": 0.0662,
+      "step": 13067
+    },
+    {
+      "epoch": 2.117303953337654,
+      "grad_norm": 1.0218912363052368,
+      "learning_rate": 1.051844597367446e-06,
+      "loss": 0.0644,
+      "step": 13068
+    },
+    {
+      "epoch": 2.117465975372651,
+      "grad_norm": 0.8441329598426819,
+      "learning_rate": 1.0514881548891813e-06,
+      "loss": 0.062,
+      "step": 13069
+    },
+    {
+      "epoch": 2.1176279974076473,
+      "grad_norm": 0.9138961434364319,
+      "learning_rate": 1.0511317567318737e-06,
+      "loss": 0.0687,
+      "step": 13070
+    },
+    {
+      "epoch": 2.1177900194426442,
+      "grad_norm": 0.9639347791671753,
+      "learning_rate": 1.0507754029064293e-06,
+      "loss": 0.0589,
+      "step": 13071
+    },
+    {
+      "epoch": 2.117952041477641,
+      "grad_norm": 0.9006848931312561,
+      "learning_rate": 1.0504190934237484e-06,
+      "loss": 0.0641,
+      "step": 13072
+    },
+    {
+      "epoch": 2.1181140635126376,
+      "grad_norm": 1.6156253814697266,
+      "learning_rate": 1.0500628282947365e-06,
+      "loss": 0.0539,
+      "step": 13073
+    },
+    {
+      "epoch": 2.1182760855476346,
+      "grad_norm": 0.8980904817581177,
+      "learning_rate": 1.0497066075302939e-06,
+      "loss": 0.0691,
+      "step": 13074
+    },
+    {
+      "epoch": 2.118438107582631,
+      "grad_norm": 1.003623604774475,
+      "learning_rate": 1.0493504311413185e-06,
+      "loss": 0.0757,
+      "step": 13075
+    },
+    {
+      "epoch": 2.118600129617628,
+      "grad_norm": 0.8838923573493958,
+      "learning_rate": 1.0489942991387088e-06,
+      "loss": 0.0564,
+      "step": 13076
+    },
+    {
+      "epoch": 2.118762151652625,
+      "grad_norm": 0.8242804408073425,
+      "learning_rate": 1.0486382115333638e-06,
+      "loss": 0.0575,
+      "step": 13077
+    },
+    {
+      "epoch": 2.1189241736876214,
+      "grad_norm": 0.7853438258171082,
+      "learning_rate": 1.0482821683361767e-06,
+      "loss": 0.0624,
+      "step": 13078
+    },
+    {
+      "epoch": 2.1190861957226184,
+      "grad_norm": 0.8636956810951233,
+      "learning_rate": 1.0479261695580417e-06,
+      "loss": 0.062,
+      "step": 13079
+    },
+    {
+      "epoch": 2.119248217757615,
+      "grad_norm": 0.9305495023727417,
+      "learning_rate": 1.0475702152098522e-06,
+      "loss": 0.0606,
+      "step": 13080
+    },
+    {
+      "epoch": 2.119410239792612,
+      "grad_norm": 0.8130801320075989,
+      "learning_rate": 1.0472143053024994e-06,
+      "loss": 0.0592,
+      "step": 13081
+    },
+    {
+      "epoch": 2.1195722618276087,
+      "grad_norm": 0.9285720586776733,
+      "learning_rate": 1.0468584398468729e-06,
+      "loss": 0.0645,
+      "step": 13082
+    },
+    {
+      "epoch": 2.119734283862605,
+      "grad_norm": 0.7963539958000183,
+      "learning_rate": 1.0465026188538618e-06,
+      "loss": 0.0576,
+      "step": 13083
+    },
+    {
+      "epoch": 2.119896305897602,
+      "grad_norm": 0.8605098724365234,
+      "learning_rate": 1.0461468423343532e-06,
+      "loss": 0.0569,
+      "step": 13084
+    },
+    {
+      "epoch": 2.1200583279325986,
+      "grad_norm": 0.8885794878005981,
+      "learning_rate": 1.0457911102992338e-06,
+      "loss": 0.0636,
+      "step": 13085
+    },
+    {
+      "epoch": 2.1202203499675956,
+      "grad_norm": 0.8622367978096008,
+      "learning_rate": 1.0454354227593855e-06,
+      "loss": 0.0614,
+      "step": 13086
+    },
+    {
+      "epoch": 2.1203823720025925,
+      "grad_norm": 0.8698747754096985,
+      "learning_rate": 1.0450797797256942e-06,
+      "loss": 0.063,
+      "step": 13087
+    },
+    {
+      "epoch": 2.120544394037589,
+      "grad_norm": 0.8544244766235352,
+      "learning_rate": 1.0447241812090408e-06,
+      "loss": 0.0548,
+      "step": 13088
+    },
+    {
+      "epoch": 2.120706416072586,
+      "grad_norm": 0.7378759980201721,
+      "learning_rate": 1.0443686272203066e-06,
+      "loss": 0.056,
+      "step": 13089
+    },
+    {
+      "epoch": 2.120868438107583,
+      "grad_norm": 0.8694587349891663,
+      "learning_rate": 1.0440131177703692e-06,
+      "loss": 0.068,
+      "step": 13090
+    },
+    {
+      "epoch": 2.1210304601425793,
+      "grad_norm": 0.8376104831695557,
+      "learning_rate": 1.0436576528701057e-06,
+      "loss": 0.0592,
+      "step": 13091
+    },
+    {
+      "epoch": 2.1211924821775763,
+      "grad_norm": 0.9313507080078125,
+      "learning_rate": 1.0433022325303956e-06,
+      "loss": 0.0627,
+      "step": 13092
+    },
+    {
+      "epoch": 2.1213545042125728,
+      "grad_norm": 0.9477971792221069,
+      "learning_rate": 1.0429468567621115e-06,
+      "loss": 0.0663,
+      "step": 13093
+    },
+    {
+      "epoch": 2.1215165262475697,
+      "grad_norm": 0.8192235827445984,
+      "learning_rate": 1.042591525576127e-06,
+      "loss": 0.0603,
+      "step": 13094
+    },
+    {
+      "epoch": 2.1216785482825666,
+      "grad_norm": 0.8660534620285034,
+      "learning_rate": 1.0422362389833145e-06,
+      "loss": 0.0631,
+      "step": 13095
+    },
+    {
+      "epoch": 2.121840570317563,
+      "grad_norm": 0.8360748887062073,
+      "learning_rate": 1.041880996994547e-06,
+      "loss": 0.0584,
+      "step": 13096
+    },
+    {
+      "epoch": 2.12200259235256,
+      "grad_norm": 0.8174402713775635,
+      "learning_rate": 1.0415257996206918e-06,
+      "loss": 0.0593,
+      "step": 13097
+    },
+    {
+      "epoch": 2.1221646143875565,
+      "grad_norm": 0.9432098269462585,
+      "learning_rate": 1.0411706468726173e-06,
+      "loss": 0.0633,
+      "step": 13098
+    },
+    {
+      "epoch": 2.1223266364225535,
+      "grad_norm": 1.0108842849731445,
+      "learning_rate": 1.040815538761191e-06,
+      "loss": 0.0607,
+      "step": 13099
+    },
+    {
+      "epoch": 2.1224886584575504,
+      "grad_norm": 0.98541259765625,
+      "learning_rate": 1.040460475297278e-06,
+      "loss": 0.069,
+      "step": 13100
+    },
+    {
+      "epoch": 2.122650680492547,
+      "grad_norm": 0.8681195974349976,
+      "learning_rate": 1.0401054564917423e-06,
+      "loss": 0.0654,
+      "step": 13101
+    },
+    {
+      "epoch": 2.122812702527544,
+      "grad_norm": 0.9383473992347717,
+      "learning_rate": 1.039750482355447e-06,
+      "loss": 0.0651,
+      "step": 13102
+    },
+    {
+      "epoch": 2.1229747245625403,
+      "grad_norm": 1.2092524766921997,
+      "learning_rate": 1.0393955528992535e-06,
+      "loss": 0.0658,
+      "step": 13103
+    },
+    {
+      "epoch": 2.1231367465975373,
+      "grad_norm": 0.8957673907279968,
+      "learning_rate": 1.0390406681340212e-06,
+      "loss": 0.0572,
+      "step": 13104
+    },
+    {
+      "epoch": 2.123298768632534,
+      "grad_norm": 0.9200911521911621,
+      "learning_rate": 1.038685828070608e-06,
+      "loss": 0.0673,
+      "step": 13105
+    },
+    {
+      "epoch": 2.1234607906675307,
+      "grad_norm": 0.8420102000236511,
+      "learning_rate": 1.0383310327198728e-06,
+      "loss": 0.0562,
+      "step": 13106
+    },
+    {
+      "epoch": 2.1236228127025276,
+      "grad_norm": 0.9516891837120056,
+      "learning_rate": 1.037976282092672e-06,
+      "loss": 0.0657,
+      "step": 13107
+    },
+    {
+      "epoch": 2.123784834737524,
+      "grad_norm": 0.9453445076942444,
+      "learning_rate": 1.0376215761998578e-06,
+      "loss": 0.0617,
+      "step": 13108
+    },
+    {
+      "epoch": 2.123946856772521,
+      "grad_norm": 0.8707835674285889,
+      "learning_rate": 1.0372669150522845e-06,
+      "loss": 0.0593,
+      "step": 13109
+    },
+    {
+      "epoch": 2.124108878807518,
+      "grad_norm": 0.9308366775512695,
+      "learning_rate": 1.0369122986608044e-06,
+      "loss": 0.0588,
+      "step": 13110
+    },
+    {
+      "epoch": 2.1242709008425145,
+      "grad_norm": 0.8552700877189636,
+      "learning_rate": 1.0365577270362668e-06,
+      "loss": 0.0617,
+      "step": 13111
+    },
+    {
+      "epoch": 2.1244329228775114,
+      "grad_norm": 0.8702750205993652,
+      "learning_rate": 1.0362032001895214e-06,
+      "loss": 0.0667,
+      "step": 13112
+    },
+    {
+      "epoch": 2.124594944912508,
+      "grad_norm": 0.8597316145896912,
+      "learning_rate": 1.0358487181314155e-06,
+      "loss": 0.0642,
+      "step": 13113
+    },
+    {
+      "epoch": 2.124756966947505,
+      "grad_norm": 0.7736619114875793,
+      "learning_rate": 1.0354942808727962e-06,
+      "loss": 0.0589,
+      "step": 13114
+    },
+    {
+      "epoch": 2.1249189889825018,
+      "grad_norm": 1.05555260181427,
+      "learning_rate": 1.0351398884245076e-06,
+      "loss": 0.0707,
+      "step": 13115
+    },
+    {
+      "epoch": 2.1250810110174982,
+      "grad_norm": 0.8743281364440918,
+      "learning_rate": 1.0347855407973933e-06,
+      "loss": 0.0598,
+      "step": 13116
+    },
+    {
+      "epoch": 2.125243033052495,
+      "grad_norm": 0.9425612092018127,
+      "learning_rate": 1.0344312380022961e-06,
+      "loss": 0.0646,
+      "step": 13117
+    },
+    {
+      "epoch": 2.1254050550874917,
+      "grad_norm": 0.8367472290992737,
+      "learning_rate": 1.034076980050057e-06,
+      "loss": 0.0601,
+      "step": 13118
+    },
+    {
+      "epoch": 2.1255670771224886,
+      "grad_norm": 0.9993242621421814,
+      "learning_rate": 1.0337227669515131e-06,
+      "loss": 0.0639,
+      "step": 13119
+    },
+    {
+      "epoch": 2.1257290991574855,
+      "grad_norm": 0.7848953008651733,
+      "learning_rate": 1.0333685987175052e-06,
+      "loss": 0.0534,
+      "step": 13120
+    },
+    {
+      "epoch": 2.125891121192482,
+      "grad_norm": 0.973812460899353,
+      "learning_rate": 1.0330144753588688e-06,
+      "loss": 0.0685,
+      "step": 13121
+    },
+    {
+      "epoch": 2.126053143227479,
+      "grad_norm": 0.8284931182861328,
+      "learning_rate": 1.0326603968864407e-06,
+      "loss": 0.0557,
+      "step": 13122
+    },
+    {
+      "epoch": 2.126215165262476,
+      "grad_norm": 0.995773196220398,
+      "learning_rate": 1.0323063633110525e-06,
+      "loss": 0.0593,
+      "step": 13123
+    },
+    {
+      "epoch": 2.1263771872974724,
+      "grad_norm": 0.8277838230133057,
+      "learning_rate": 1.0319523746435367e-06,
+      "loss": 0.063,
+      "step": 13124
+    },
+    {
+      "epoch": 2.1265392093324693,
+      "grad_norm": 0.8277966380119324,
+      "learning_rate": 1.0315984308947277e-06,
+      "loss": 0.0618,
+      "step": 13125
+    },
+    {
+      "epoch": 2.126701231367466,
+      "grad_norm": 1.1131237745285034,
+      "learning_rate": 1.0312445320754522e-06,
+      "loss": 0.0729,
+      "step": 13126
+    },
+    {
+      "epoch": 2.1268632534024627,
+      "grad_norm": 0.846409797668457,
+      "learning_rate": 1.0308906781965396e-06,
+      "loss": 0.0642,
+      "step": 13127
+    },
+    {
+      "epoch": 2.1270252754374597,
+      "grad_norm": 0.8704565763473511,
+      "learning_rate": 1.0305368692688175e-06,
+      "loss": 0.07,
+      "step": 13128
+    },
+    {
+      "epoch": 2.127187297472456,
+      "grad_norm": 0.8594701886177063,
+      "learning_rate": 1.0301831053031109e-06,
+      "loss": 0.0583,
+      "step": 13129
+    },
+    {
+      "epoch": 2.127349319507453,
+      "grad_norm": 0.9101399183273315,
+      "learning_rate": 1.0298293863102444e-06,
+      "loss": 0.064,
+      "step": 13130
+    },
+    {
+      "epoch": 2.1275113415424496,
+      "grad_norm": 0.896522045135498,
+      "learning_rate": 1.0294757123010406e-06,
+      "loss": 0.0645,
+      "step": 13131
+    },
+    {
+      "epoch": 2.1276733635774465,
+      "grad_norm": 0.8539774417877197,
+      "learning_rate": 1.0291220832863219e-06,
+      "loss": 0.0617,
+      "step": 13132
+    },
+    {
+      "epoch": 2.1278353856124435,
+      "grad_norm": 0.8655493259429932,
+      "learning_rate": 1.0287684992769086e-06,
+      "loss": 0.065,
+      "step": 13133
+    },
+    {
+      "epoch": 2.12799740764744,
+      "grad_norm": 0.8757855296134949,
+      "learning_rate": 1.0284149602836174e-06,
+      "loss": 0.0602,
+      "step": 13134
+    },
+    {
+      "epoch": 2.128159429682437,
+      "grad_norm": 0.9795240759849548,
+      "learning_rate": 1.028061466317268e-06,
+      "loss": 0.0667,
+      "step": 13135
+    },
+    {
+      "epoch": 2.1283214517174334,
+      "grad_norm": 0.8074472546577454,
+      "learning_rate": 1.0277080173886766e-06,
+      "loss": 0.0633,
+      "step": 13136
+    },
+    {
+      "epoch": 2.1284834737524303,
+      "grad_norm": 0.8354179263114929,
+      "learning_rate": 1.0273546135086559e-06,
+      "loss": 0.0551,
+      "step": 13137
+    },
+    {
+      "epoch": 2.1286454957874272,
+      "grad_norm": 0.8012445569038391,
+      "learning_rate": 1.0270012546880207e-06,
+      "loss": 0.0567,
+      "step": 13138
+    },
+    {
+      "epoch": 2.1288075178224237,
+      "grad_norm": 0.8529260158538818,
+      "learning_rate": 1.0266479409375813e-06,
+      "loss": 0.0646,
+      "step": 13139
+    },
+    {
+      "epoch": 2.1289695398574207,
+      "grad_norm": 0.8222674131393433,
+      "learning_rate": 1.0262946722681513e-06,
+      "loss": 0.0624,
+      "step": 13140
+    },
+    {
+      "epoch": 2.129131561892417,
+      "grad_norm": 0.757540762424469,
+      "learning_rate": 1.0259414486905373e-06,
+      "loss": 0.0591,
+      "step": 13141
+    },
+    {
+      "epoch": 2.129293583927414,
+      "grad_norm": 0.9325169920921326,
+      "learning_rate": 1.0255882702155476e-06,
+      "loss": 0.0637,
+      "step": 13142
+    },
+    {
+      "epoch": 2.129455605962411,
+      "grad_norm": 0.7963911294937134,
+      "learning_rate": 1.025235136853989e-06,
+      "loss": 0.0561,
+      "step": 13143
+    },
+    {
+      "epoch": 2.1296176279974075,
+      "grad_norm": 0.8913578391075134,
+      "learning_rate": 1.024882048616666e-06,
+      "loss": 0.058,
+      "step": 13144
+    },
+    {
+      "epoch": 2.1297796500324044,
+      "grad_norm": 0.9068898558616638,
+      "learning_rate": 1.024529005514383e-06,
+      "loss": 0.0654,
+      "step": 13145
+    },
+    {
+      "epoch": 2.1299416720674014,
+      "grad_norm": 0.9450251460075378,
+      "learning_rate": 1.0241760075579418e-06,
+      "loss": 0.0663,
+      "step": 13146
+    },
+    {
+      "epoch": 2.130103694102398,
+      "grad_norm": 0.8904002904891968,
+      "learning_rate": 1.023823054758144e-06,
+      "loss": 0.0605,
+      "step": 13147
+    },
+    {
+      "epoch": 2.130265716137395,
+      "grad_norm": 0.7318344712257385,
+      "learning_rate": 1.0234701471257868e-06,
+      "loss": 0.0577,
+      "step": 13148
+    },
+    {
+      "epoch": 2.1304277381723913,
+      "grad_norm": 0.8698272705078125,
+      "learning_rate": 1.023117284671671e-06,
+      "loss": 0.062,
+      "step": 13149
+    },
+    {
+      "epoch": 2.130589760207388,
+      "grad_norm": 0.868567168712616,
+      "learning_rate": 1.0227644674065923e-06,
+      "loss": 0.0588,
+      "step": 13150
+    },
+    {
+      "epoch": 2.130751782242385,
+      "grad_norm": 0.8194864988327026,
+      "learning_rate": 1.0224116953413468e-06,
+      "loss": 0.0544,
+      "step": 13151
+    },
+    {
+      "epoch": 2.1309138042773816,
+      "grad_norm": 0.8874611258506775,
+      "learning_rate": 1.0220589684867269e-06,
+      "loss": 0.0614,
+      "step": 13152
+    },
+    {
+      "epoch": 2.1310758263123786,
+      "grad_norm": 0.9767943620681763,
+      "learning_rate": 1.0217062868535249e-06,
+      "loss": 0.0582,
+      "step": 13153
+    },
+    {
+      "epoch": 2.131237848347375,
+      "grad_norm": 0.8773373365402222,
+      "learning_rate": 1.021353650452535e-06,
+      "loss": 0.0601,
+      "step": 13154
+    },
+    {
+      "epoch": 2.131399870382372,
+      "grad_norm": 0.8418240547180176,
+      "learning_rate": 1.0210010592945442e-06,
+      "loss": 0.0585,
+      "step": 13155
+    },
+    {
+      "epoch": 2.131561892417369,
+      "grad_norm": 0.8298531174659729,
+      "learning_rate": 1.0206485133903424e-06,
+      "loss": 0.0618,
+      "step": 13156
+    },
+    {
+      "epoch": 2.1317239144523654,
+      "grad_norm": 0.81301349401474,
+      "learning_rate": 1.0202960127507155e-06,
+      "loss": 0.0536,
+      "step": 13157
+    },
+    {
+      "epoch": 2.1318859364873624,
+      "grad_norm": 0.8824573755264282,
+      "learning_rate": 1.0199435573864502e-06,
+      "loss": 0.0617,
+      "step": 13158
+    },
+    {
+      "epoch": 2.132047958522359,
+      "grad_norm": 0.9553408622741699,
+      "learning_rate": 1.01959114730833e-06,
+      "loss": 0.0646,
+      "step": 13159
+    },
+    {
+      "epoch": 2.1322099805573558,
+      "grad_norm": 0.8561825156211853,
+      "learning_rate": 1.0192387825271384e-06,
+      "loss": 0.0622,
+      "step": 13160
+    },
+    {
+      "epoch": 2.1323720025923527,
+      "grad_norm": 0.9502358436584473,
+      "learning_rate": 1.0188864630536568e-06,
+      "loss": 0.0666,
+      "step": 13161
+    },
+    {
+      "epoch": 2.132534024627349,
+      "grad_norm": 0.973298966884613,
+      "learning_rate": 1.018534188898665e-06,
+      "loss": 0.0671,
+      "step": 13162
+    },
+    {
+      "epoch": 2.132696046662346,
+      "grad_norm": 0.9729242920875549,
+      "learning_rate": 1.018181960072942e-06,
+      "loss": 0.0594,
+      "step": 13163
+    },
+    {
+      "epoch": 2.1328580686973426,
+      "grad_norm": 0.9466162919998169,
+      "learning_rate": 1.0178297765872651e-06,
+      "loss": 0.0709,
+      "step": 13164
+    },
+    {
+      "epoch": 2.1330200907323396,
+      "grad_norm": 0.8360689878463745,
+      "learning_rate": 1.0174776384524104e-06,
+      "loss": 0.0593,
+      "step": 13165
+    },
+    {
+      "epoch": 2.1331821127673365,
+      "grad_norm": 0.8890894055366516,
+      "learning_rate": 1.0171255456791531e-06,
+      "loss": 0.0664,
+      "step": 13166
+    },
+    {
+      "epoch": 2.133344134802333,
+      "grad_norm": 0.9430520534515381,
+      "learning_rate": 1.0167734982782636e-06,
+      "loss": 0.0605,
+      "step": 13167
+    },
+    {
+      "epoch": 2.13350615683733,
+      "grad_norm": 1.001865029335022,
+      "learning_rate": 1.016421496260517e-06,
+      "loss": 0.0626,
+      "step": 13168
+    },
+    {
+      "epoch": 2.133668178872327,
+      "grad_norm": 1.0136559009552002,
+      "learning_rate": 1.0160695396366832e-06,
+      "loss": 0.0638,
+      "step": 13169
+    },
+    {
+      "epoch": 2.1338302009073233,
+      "grad_norm": 1.062779188156128,
+      "learning_rate": 1.0157176284175293e-06,
+      "loss": 0.0654,
+      "step": 13170
+    },
+    {
+      "epoch": 2.1339922229423203,
+      "grad_norm": 0.7638817429542542,
+      "learning_rate": 1.0153657626138244e-06,
+      "loss": 0.0505,
+      "step": 13171
+    },
+    {
+      "epoch": 2.1341542449773168,
+      "grad_norm": 0.8198952078819275,
+      "learning_rate": 1.0150139422363342e-06,
+      "loss": 0.0552,
+      "step": 13172
+    },
+    {
+      "epoch": 2.1343162670123137,
+      "grad_norm": 0.9854885935783386,
+      "learning_rate": 1.0146621672958238e-06,
+      "loss": 0.061,
+      "step": 13173
+    },
+    {
+      "epoch": 2.1344782890473106,
+      "grad_norm": 0.9766724705696106,
+      "learning_rate": 1.0143104378030565e-06,
+      "loss": 0.0619,
+      "step": 13174
+    },
+    {
+      "epoch": 2.134640311082307,
+      "grad_norm": 1.0882524251937866,
+      "learning_rate": 1.013958753768795e-06,
+      "loss": 0.0698,
+      "step": 13175
+    },
+    {
+      "epoch": 2.134802333117304,
+      "grad_norm": 0.8586061596870422,
+      "learning_rate": 1.013607115203799e-06,
+      "loss": 0.0645,
+      "step": 13176
+    },
+    {
+      "epoch": 2.1349643551523005,
+      "grad_norm": 0.8522896766662598,
+      "learning_rate": 1.0132555221188285e-06,
+      "loss": 0.0626,
+      "step": 13177
+    },
+    {
+      "epoch": 2.1351263771872975,
+      "grad_norm": 0.9053431153297424,
+      "learning_rate": 1.012903974524641e-06,
+      "loss": 0.0658,
+      "step": 13178
+    },
+    {
+      "epoch": 2.1352883992222944,
+      "grad_norm": 0.8362330794334412,
+      "learning_rate": 1.0125524724319936e-06,
+      "loss": 0.0619,
+      "step": 13179
+    },
+    {
+      "epoch": 2.135450421257291,
+      "grad_norm": 0.8866271376609802,
+      "learning_rate": 1.0122010158516412e-06,
+      "loss": 0.0626,
+      "step": 13180
+    },
+    {
+      "epoch": 2.135612443292288,
+      "grad_norm": 0.9750482439994812,
+      "learning_rate": 1.0118496047943368e-06,
+      "loss": 0.0746,
+      "step": 13181
+    },
+    {
+      "epoch": 2.1357744653272843,
+      "grad_norm": 0.9252564311027527,
+      "learning_rate": 1.0114982392708325e-06,
+      "loss": 0.0663,
+      "step": 13182
+    },
+    {
+      "epoch": 2.1359364873622813,
+      "grad_norm": 1.186126947402954,
+      "learning_rate": 1.0111469192918808e-06,
+      "loss": 0.0598,
+      "step": 13183
+    },
+    {
+      "epoch": 2.136098509397278,
+      "grad_norm": 0.952793300151825,
+      "learning_rate": 1.010795644868231e-06,
+      "loss": 0.0554,
+      "step": 13184
+    },
+    {
+      "epoch": 2.1362605314322747,
+      "grad_norm": 1.0087121725082397,
+      "learning_rate": 1.01044441601063e-06,
+      "loss": 0.0665,
+      "step": 13185
+    },
+    {
+      "epoch": 2.1364225534672716,
+      "grad_norm": 0.8914116024971008,
+      "learning_rate": 1.0100932327298244e-06,
+      "loss": 0.064,
+      "step": 13186
+    },
+    {
+      "epoch": 2.136584575502268,
+      "grad_norm": 0.8840557336807251,
+      "learning_rate": 1.009742095036562e-06,
+      "loss": 0.0569,
+      "step": 13187
+    },
+    {
+      "epoch": 2.136746597537265,
+      "grad_norm": 0.9341459274291992,
+      "learning_rate": 1.0093910029415843e-06,
+      "loss": 0.0543,
+      "step": 13188
+    },
+    {
+      "epoch": 2.136908619572262,
+      "grad_norm": 0.9242621064186096,
+      "learning_rate": 1.0090399564556348e-06,
+      "loss": 0.0626,
+      "step": 13189
+    },
+    {
+      "epoch": 2.1370706416072585,
+      "grad_norm": 0.9278979897499084,
+      "learning_rate": 1.0086889555894545e-06,
+      "loss": 0.0592,
+      "step": 13190
+    },
+    {
+      "epoch": 2.1372326636422554,
+      "grad_norm": 0.8471470475196838,
+      "learning_rate": 1.0083380003537831e-06,
+      "loss": 0.0644,
+      "step": 13191
+    },
+    {
+      "epoch": 2.1373946856772523,
+      "grad_norm": 0.8700275421142578,
+      "learning_rate": 1.0079870907593592e-06,
+      "loss": 0.063,
+      "step": 13192
+    },
+    {
+      "epoch": 2.137556707712249,
+      "grad_norm": 0.7276986241340637,
+      "learning_rate": 1.0076362268169199e-06,
+      "loss": 0.0525,
+      "step": 13193
+    },
+    {
+      "epoch": 2.1377187297472457,
+      "grad_norm": 1.0328106880187988,
+      "learning_rate": 1.0072854085372005e-06,
+      "loss": 0.0698,
+      "step": 13194
+    },
+    {
+      "epoch": 2.1378807517822422,
+      "grad_norm": 0.8020308017730713,
+      "learning_rate": 1.0069346359309362e-06,
+      "loss": 0.0555,
+      "step": 13195
+    },
+    {
+      "epoch": 2.138042773817239,
+      "grad_norm": 0.9141387343406677,
+      "learning_rate": 1.0065839090088572e-06,
+      "loss": 0.0637,
+      "step": 13196
+    },
+    {
+      "epoch": 2.138204795852236,
+      "grad_norm": 0.8611671328544617,
+      "learning_rate": 1.0062332277816972e-06,
+      "loss": 0.0624,
+      "step": 13197
+    },
+    {
+      "epoch": 2.1383668178872326,
+      "grad_norm": 0.8084477186203003,
+      "learning_rate": 1.0058825922601866e-06,
+      "loss": 0.0594,
+      "step": 13198
+    },
+    {
+      "epoch": 2.1385288399222295,
+      "grad_norm": 0.9061813950538635,
+      "learning_rate": 1.0055320024550521e-06,
+      "loss": 0.0678,
+      "step": 13199
+    },
+    {
+      "epoch": 2.138690861957226,
+      "grad_norm": 0.9596869349479675,
+      "learning_rate": 1.005181458377022e-06,
+      "loss": 0.0662,
+      "step": 13200
+    },
+    {
+      "epoch": 2.138852883992223,
+      "grad_norm": 0.8806290626525879,
+      "learning_rate": 1.004830960036821e-06,
+      "loss": 0.0613,
+      "step": 13201
+    },
+    {
+      "epoch": 2.13901490602722,
+      "grad_norm": 0.7227916121482849,
+      "learning_rate": 1.0044805074451757e-06,
+      "loss": 0.0509,
+      "step": 13202
+    },
+    {
+      "epoch": 2.1391769280622164,
+      "grad_norm": 0.7963190674781799,
+      "learning_rate": 1.0041301006128073e-06,
+      "loss": 0.0637,
+      "step": 13203
+    },
+    {
+      "epoch": 2.1393389500972133,
+      "grad_norm": 0.936798632144928,
+      "learning_rate": 1.003779739550438e-06,
+      "loss": 0.0689,
+      "step": 13204
+    },
+    {
+      "epoch": 2.13950097213221,
+      "grad_norm": 0.857064962387085,
+      "learning_rate": 1.0034294242687875e-06,
+      "loss": 0.0585,
+      "step": 13205
+    },
+    {
+      "epoch": 2.1396629941672067,
+      "grad_norm": 0.9607579112052917,
+      "learning_rate": 1.003079154778575e-06,
+      "loss": 0.0691,
+      "step": 13206
+    },
+    {
+      "epoch": 2.1398250162022037,
+      "grad_norm": 0.9430453181266785,
+      "learning_rate": 1.0027289310905181e-06,
+      "loss": 0.059,
+      "step": 13207
+    },
+    {
+      "epoch": 2.1399870382372,
+      "grad_norm": 0.8209136724472046,
+      "learning_rate": 1.0023787532153325e-06,
+      "loss": 0.0648,
+      "step": 13208
+    },
+    {
+      "epoch": 2.140149060272197,
+      "grad_norm": 0.9626993536949158,
+      "learning_rate": 1.0020286211637328e-06,
+      "loss": 0.0641,
+      "step": 13209
+    },
+    {
+      "epoch": 2.1403110823071936,
+      "grad_norm": 0.9732939004898071,
+      "learning_rate": 1.0016785349464326e-06,
+      "loss": 0.0647,
+      "step": 13210
+    },
+    {
+      "epoch": 2.1404731043421905,
+      "grad_norm": 0.8415422439575195,
+      "learning_rate": 1.0013284945741431e-06,
+      "loss": 0.0608,
+      "step": 13211
+    },
+    {
+      "epoch": 2.1406351263771874,
+      "grad_norm": 0.9403479099273682,
+      "learning_rate": 1.0009785000575747e-06,
+      "loss": 0.0601,
+      "step": 13212
+    },
+    {
+      "epoch": 2.140797148412184,
+      "grad_norm": 0.9788222312927246,
+      "learning_rate": 1.000628551407438e-06,
+      "loss": 0.0685,
+      "step": 13213
+    },
+    {
+      "epoch": 2.140959170447181,
+      "grad_norm": 0.8836336135864258,
+      "learning_rate": 1.0002786486344379e-06,
+      "loss": 0.0625,
+      "step": 13214
+    },
+    {
+      "epoch": 2.141121192482178,
+      "grad_norm": 0.8284887671470642,
+      "learning_rate": 9.999287917492814e-07,
+      "loss": 0.057,
+      "step": 13215
+    },
+    {
+      "epoch": 2.1412832145171743,
+      "grad_norm": 0.9652777314186096,
+      "learning_rate": 9.995789807626754e-07,
+      "loss": 0.0552,
+      "step": 13216
+    },
+    {
+      "epoch": 2.1414452365521712,
+      "grad_norm": 0.9502488970756531,
+      "learning_rate": 9.992292156853207e-07,
+      "loss": 0.0636,
+      "step": 13217
+    },
+    {
+      "epoch": 2.1416072585871677,
+      "grad_norm": 0.9706515073776245,
+      "learning_rate": 9.988794965279203e-07,
+      "loss": 0.0636,
+      "step": 13218
+    },
+    {
+      "epoch": 2.1417692806221647,
+      "grad_norm": 0.9208532571792603,
+      "learning_rate": 9.985298233011746e-07,
+      "loss": 0.072,
+      "step": 13219
+    },
+    {
+      "epoch": 2.141931302657161,
+      "grad_norm": 0.8984748125076294,
+      "learning_rate": 9.981801960157827e-07,
+      "loss": 0.0574,
+      "step": 13220
+    },
+    {
+      "epoch": 2.142093324692158,
+      "grad_norm": 0.8832286596298218,
+      "learning_rate": 9.978306146824427e-07,
+      "loss": 0.0625,
+      "step": 13221
+    },
+    {
+      "epoch": 2.142255346727155,
+      "grad_norm": 1.0187299251556396,
+      "learning_rate": 9.974810793118505e-07,
+      "loss": 0.0616,
+      "step": 13222
+    },
+    {
+      "epoch": 2.1424173687621515,
+      "grad_norm": 0.8346063494682312,
+      "learning_rate": 9.971315899147012e-07,
+      "loss": 0.0624,
+      "step": 13223
+    },
+    {
+      "epoch": 2.1425793907971484,
+      "grad_norm": 1.0231930017471313,
+      "learning_rate": 9.967821465016893e-07,
+      "loss": 0.0675,
+      "step": 13224
+    },
+    {
+      "epoch": 2.1427414128321454,
+      "grad_norm": 0.936922550201416,
+      "learning_rate": 9.964327490835045e-07,
+      "loss": 0.0647,
+      "step": 13225
+    },
+    {
+      "epoch": 2.142903434867142,
+      "grad_norm": 0.8798637390136719,
+      "learning_rate": 9.960833976708398e-07,
+      "loss": 0.052,
+      "step": 13226
+    },
+    {
+      "epoch": 2.143065456902139,
+      "grad_norm": 0.7966178059577942,
+      "learning_rate": 9.957340922743835e-07,
+      "loss": 0.0607,
+      "step": 13227
+    },
+    {
+      "epoch": 2.1432274789371353,
+      "grad_norm": 0.7437213659286499,
+      "learning_rate": 9.953848329048248e-07,
+      "loss": 0.0558,
+      "step": 13228
+    },
+    {
+      "epoch": 2.143389500972132,
+      "grad_norm": 1.0451050996780396,
+      "learning_rate": 9.950356195728483e-07,
+      "loss": 0.0761,
+      "step": 13229
+    },
+    {
+      "epoch": 2.143551523007129,
+      "grad_norm": 0.8481780290603638,
+      "learning_rate": 9.94686452289139e-07,
+      "loss": 0.0546,
+      "step": 13230
+    },
+    {
+      "epoch": 2.1437135450421256,
+      "grad_norm": 0.8842169046401978,
+      "learning_rate": 9.943373310643831e-07,
+      "loss": 0.0643,
+      "step": 13231
+    },
+    {
+      "epoch": 2.1438755670771226,
+      "grad_norm": 0.9944992065429688,
+      "learning_rate": 9.939882559092604e-07,
+      "loss": 0.0705,
+      "step": 13232
+    },
+    {
+      "epoch": 2.144037589112119,
+      "grad_norm": 0.9027183651924133,
+      "learning_rate": 9.936392268344525e-07,
+      "loss": 0.0682,
+      "step": 13233
+    },
+    {
+      "epoch": 2.144199611147116,
+      "grad_norm": 0.7881889939308167,
+      "learning_rate": 9.93290243850638e-07,
+      "loss": 0.0604,
+      "step": 13234
+    },
+    {
+      "epoch": 2.144361633182113,
+      "grad_norm": 0.8811602592468262,
+      "learning_rate": 9.929413069684979e-07,
+      "loss": 0.0649,
+      "step": 13235
+    },
+    {
+      "epoch": 2.1445236552171094,
+      "grad_norm": 0.8543482422828674,
+      "learning_rate": 9.925924161987057e-07,
+      "loss": 0.0558,
+      "step": 13236
+    },
+    {
+      "epoch": 2.1446856772521063,
+      "grad_norm": 1.0113856792449951,
+      "learning_rate": 9.922435715519379e-07,
+      "loss": 0.0649,
+      "step": 13237
+    },
+    {
+      "epoch": 2.144847699287103,
+      "grad_norm": 0.8757432103157043,
+      "learning_rate": 9.918947730388682e-07,
+      "loss": 0.0651,
+      "step": 13238
+    },
+    {
+      "epoch": 2.1450097213220998,
+      "grad_norm": 1.0214018821716309,
+      "learning_rate": 9.915460206701685e-07,
+      "loss": 0.0677,
+      "step": 13239
+    },
+    {
+      "epoch": 2.1451717433570967,
+      "grad_norm": 0.9439998865127563,
+      "learning_rate": 9.911973144565105e-07,
+      "loss": 0.066,
+      "step": 13240
+    },
+    {
+      "epoch": 2.145333765392093,
+      "grad_norm": 1.0487086772918701,
+      "learning_rate": 9.908486544085632e-07,
+      "loss": 0.0734,
+      "step": 13241
+    },
+    {
+      "epoch": 2.14549578742709,
+      "grad_norm": 0.9988869428634644,
+      "learning_rate": 9.90500040536996e-07,
+      "loss": 0.0692,
+      "step": 13242
+    },
+    {
+      "epoch": 2.1456578094620866,
+      "grad_norm": 0.87641841173172,
+      "learning_rate": 9.901514728524739e-07,
+      "loss": 0.056,
+      "step": 13243
+    },
+    {
+      "epoch": 2.1458198314970836,
+      "grad_norm": 0.9582691192626953,
+      "learning_rate": 9.898029513656618e-07,
+      "loss": 0.0598,
+      "step": 13244
+    },
+    {
+      "epoch": 2.1459818535320805,
+      "grad_norm": 1.0479621887207031,
+      "learning_rate": 9.89454476087226e-07,
+      "loss": 0.0682,
+      "step": 13245
+    },
+    {
+      "epoch": 2.146143875567077,
+      "grad_norm": 0.893338143825531,
+      "learning_rate": 9.891060470278286e-07,
+      "loss": 0.0585,
+      "step": 13246
+    },
+    {
+      "epoch": 2.146305897602074,
+      "grad_norm": 0.880436360836029,
+      "learning_rate": 9.887576641981285e-07,
+      "loss": 0.0668,
+      "step": 13247
+    },
+    {
+      "epoch": 2.146467919637071,
+      "grad_norm": 0.9374086856842041,
+      "learning_rate": 9.884093276087871e-07,
+      "loss": 0.0614,
+      "step": 13248
+    },
+    {
+      "epoch": 2.1466299416720673,
+      "grad_norm": 0.9596349596977234,
+      "learning_rate": 9.880610372704624e-07,
+      "loss": 0.0688,
+      "step": 13249
+    },
+    {
+      "epoch": 2.1467919637070643,
+      "grad_norm": 1.77931809425354,
+      "learning_rate": 9.877127931938111e-07,
+      "loss": 0.0633,
+      "step": 13250
+    },
+    {
+      "epoch": 2.1469539857420608,
+      "grad_norm": 0.8775843977928162,
+      "learning_rate": 9.873645953894887e-07,
+      "loss": 0.0577,
+      "step": 13251
+    },
+    {
+      "epoch": 2.1471160077770577,
+      "grad_norm": 0.9120129942893982,
+      "learning_rate": 9.87016443868149e-07,
+      "loss": 0.0629,
+      "step": 13252
+    },
+    {
+      "epoch": 2.1472780298120546,
+      "grad_norm": 0.838457465171814,
+      "learning_rate": 9.86668338640445e-07,
+      "loss": 0.0577,
+      "step": 13253
+    },
+    {
+      "epoch": 2.147440051847051,
+      "grad_norm": 0.864747941493988,
+      "learning_rate": 9.863202797170273e-07,
+      "loss": 0.0634,
+      "step": 13254
+    },
+    {
+      "epoch": 2.147602073882048,
+      "grad_norm": 0.7873530387878418,
+      "learning_rate": 9.85972267108546e-07,
+      "loss": 0.0582,
+      "step": 13255
+    },
+    {
+      "epoch": 2.1477640959170445,
+      "grad_norm": 0.9956313371658325,
+      "learning_rate": 9.8562430082565e-07,
+      "loss": 0.0662,
+      "step": 13256
+    },
+    {
+      "epoch": 2.1479261179520415,
+      "grad_norm": 0.8575721383094788,
+      "learning_rate": 9.852763808789862e-07,
+      "loss": 0.06,
+      "step": 13257
+    },
+    {
+      "epoch": 2.1480881399870384,
+      "grad_norm": 0.9134724140167236,
+      "learning_rate": 9.849285072791978e-07,
+      "loss": 0.0657,
+      "step": 13258
+    },
+    {
+      "epoch": 2.148250162022035,
+      "grad_norm": 0.8377612829208374,
+      "learning_rate": 9.845806800369316e-07,
+      "loss": 0.0613,
+      "step": 13259
+    },
+    {
+      "epoch": 2.148412184057032,
+      "grad_norm": 0.7430257201194763,
+      "learning_rate": 9.8423289916283e-07,
+      "loss": 0.0535,
+      "step": 13260
+    },
+    {
+      "epoch": 2.1485742060920283,
+      "grad_norm": 0.9967155456542969,
+      "learning_rate": 9.838851646675329e-07,
+      "loss": 0.0691,
+      "step": 13261
+    },
+    {
+      "epoch": 2.1487362281270252,
+      "grad_norm": 0.809754490852356,
+      "learning_rate": 9.835374765616809e-07,
+      "loss": 0.052,
+      "step": 13262
+    },
+    {
+      "epoch": 2.148898250162022,
+      "grad_norm": 0.816158652305603,
+      "learning_rate": 9.831898348559115e-07,
+      "loss": 0.0594,
+      "step": 13263
+    },
+    {
+      "epoch": 2.1490602721970187,
+      "grad_norm": 0.8675599098205566,
+      "learning_rate": 9.82842239560864e-07,
+      "loss": 0.0603,
+      "step": 13264
+    },
+    {
+      "epoch": 2.1492222942320156,
+      "grad_norm": 0.8979247808456421,
+      "learning_rate": 9.824946906871721e-07,
+      "loss": 0.0641,
+      "step": 13265
+    },
+    {
+      "epoch": 2.149384316267012,
+      "grad_norm": 0.7833936214447021,
+      "learning_rate": 9.821471882454703e-07,
+      "loss": 0.0568,
+      "step": 13266
+    },
+    {
+      "epoch": 2.149546338302009,
+      "grad_norm": 0.8592799305915833,
+      "learning_rate": 9.817997322463912e-07,
+      "loss": 0.0624,
+      "step": 13267
+    },
+    {
+      "epoch": 2.149708360337006,
+      "grad_norm": 0.9396870732307434,
+      "learning_rate": 9.814523227005662e-07,
+      "loss": 0.0621,
+      "step": 13268
+    },
+    {
+      "epoch": 2.1498703823720025,
+      "grad_norm": 0.876347005367279,
+      "learning_rate": 9.811049596186255e-07,
+      "loss": 0.0709,
+      "step": 13269
+    },
+    {
+      "epoch": 2.1500324044069994,
+      "grad_norm": 0.9403771162033081,
+      "learning_rate": 9.807576430111975e-07,
+      "loss": 0.0612,
+      "step": 13270
+    },
+    {
+      "epoch": 2.1501944264419963,
+      "grad_norm": 0.8575648665428162,
+      "learning_rate": 9.804103728889089e-07,
+      "loss": 0.0569,
+      "step": 13271
+    },
+    {
+      "epoch": 2.150356448476993,
+      "grad_norm": 0.7617552876472473,
+      "learning_rate": 9.800631492623867e-07,
+      "loss": 0.0533,
+      "step": 13272
+    },
+    {
+      "epoch": 2.1505184705119897,
+      "grad_norm": 0.7716460824012756,
+      "learning_rate": 9.79715972142252e-07,
+      "loss": 0.0536,
+      "step": 13273
+    },
+    {
+      "epoch": 2.1506804925469862,
+      "grad_norm": 0.911914050579071,
+      "learning_rate": 9.793688415391304e-07,
+      "loss": 0.0623,
+      "step": 13274
+    },
+    {
+      "epoch": 2.150842514581983,
+      "grad_norm": 0.832737386226654,
+      "learning_rate": 9.790217574636433e-07,
+      "loss": 0.0606,
+      "step": 13275
+    },
+    {
+      "epoch": 2.15100453661698,
+      "grad_norm": 1.0580108165740967,
+      "learning_rate": 9.786747199264088e-07,
+      "loss": 0.0632,
+      "step": 13276
+    },
+    {
+      "epoch": 2.1511665586519766,
+      "grad_norm": 0.8876791000366211,
+      "learning_rate": 9.783277289380456e-07,
+      "loss": 0.0628,
+      "step": 13277
+    },
+    {
+      "epoch": 2.1513285806869735,
+      "grad_norm": 0.9212145209312439,
+      "learning_rate": 9.779807845091722e-07,
+      "loss": 0.0622,
+      "step": 13278
+    },
+    {
+      "epoch": 2.15149060272197,
+      "grad_norm": 0.7946116924285889,
+      "learning_rate": 9.776338866504045e-07,
+      "loss": 0.0588,
+      "step": 13279
+    },
+    {
+      "epoch": 2.151652624756967,
+      "grad_norm": 0.796489417552948,
+      "learning_rate": 9.77287035372355e-07,
+      "loss": 0.0572,
+      "step": 13280
+    },
+    {
+      "epoch": 2.151814646791964,
+      "grad_norm": 0.8203312754631042,
+      "learning_rate": 9.769402306856373e-07,
+      "loss": 0.0578,
+      "step": 13281
+    },
+    {
+      "epoch": 2.1519766688269604,
+      "grad_norm": 1.2622346878051758,
+      "learning_rate": 9.76593472600863e-07,
+      "loss": 0.0665,
+      "step": 13282
+    },
+    {
+      "epoch": 2.1521386908619573,
+      "grad_norm": 0.9482854604721069,
+      "learning_rate": 9.762467611286416e-07,
+      "loss": 0.0666,
+      "step": 13283
+    },
+    {
+      "epoch": 2.152300712896954,
+      "grad_norm": 0.8730493783950806,
+      "learning_rate": 9.75900096279582e-07,
+      "loss": 0.0604,
+      "step": 13284
+    },
+    {
+      "epoch": 2.1524627349319507,
+      "grad_norm": 0.9414070248603821,
+      "learning_rate": 9.755534780642911e-07,
+      "loss": 0.0669,
+      "step": 13285
+    },
+    {
+      "epoch": 2.1526247569669477,
+      "grad_norm": 0.9391159415245056,
+      "learning_rate": 9.752069064933758e-07,
+      "loss": 0.0634,
+      "step": 13286
+    },
+    {
+      "epoch": 2.152786779001944,
+      "grad_norm": 0.8309518098831177,
+      "learning_rate": 9.748603815774371e-07,
+      "loss": 0.0635,
+      "step": 13287
+    },
+    {
+      "epoch": 2.152948801036941,
+      "grad_norm": 0.7785058617591858,
+      "learning_rate": 9.745139033270812e-07,
+      "loss": 0.06,
+      "step": 13288
+    },
+    {
+      "epoch": 2.1531108230719376,
+      "grad_norm": 0.9786415696144104,
+      "learning_rate": 9.74167471752908e-07,
+      "loss": 0.0626,
+      "step": 13289
+    },
+    {
+      "epoch": 2.1532728451069345,
+      "grad_norm": 1.009415626525879,
+      "learning_rate": 9.738210868655187e-07,
+      "loss": 0.0712,
+      "step": 13290
+    },
+    {
+      "epoch": 2.1534348671419314,
+      "grad_norm": 0.8282557725906372,
+      "learning_rate": 9.7347474867551e-07,
+      "loss": 0.0582,
+      "step": 13291
+    },
+    {
+      "epoch": 2.153596889176928,
+      "grad_norm": 1.0145543813705444,
+      "learning_rate": 9.73128457193479e-07,
+      "loss": 0.0637,
+      "step": 13292
+    },
+    {
+      "epoch": 2.153758911211925,
+      "grad_norm": 0.950143039226532,
+      "learning_rate": 9.72782212430024e-07,
+      "loss": 0.0676,
+      "step": 13293
+    },
+    {
+      "epoch": 2.153920933246922,
+      "grad_norm": 0.8464543223381042,
+      "learning_rate": 9.724360143957367e-07,
+      "loss": 0.0622,
+      "step": 13294
+    },
+    {
+      "epoch": 2.1540829552819183,
+      "grad_norm": 0.9932398200035095,
+      "learning_rate": 9.720898631012106e-07,
+      "loss": 0.0596,
+      "step": 13295
+    },
+    {
+      "epoch": 2.154244977316915,
+      "grad_norm": 0.8855057954788208,
+      "learning_rate": 9.717437585570375e-07,
+      "loss": 0.0592,
+      "step": 13296
+    },
+    {
+      "epoch": 2.1544069993519117,
+      "grad_norm": 0.9049186110496521,
+      "learning_rate": 9.713977007738068e-07,
+      "loss": 0.0685,
+      "step": 13297
+    },
+    {
+      "epoch": 2.1545690213869086,
+      "grad_norm": 1.360369324684143,
+      "learning_rate": 9.710516897621072e-07,
+      "loss": 0.0636,
+      "step": 13298
+    },
+    {
+      "epoch": 2.1547310434219056,
+      "grad_norm": 0.8323555588722229,
+      "learning_rate": 9.707057255325262e-07,
+      "loss": 0.0633,
+      "step": 13299
+    },
+    {
+      "epoch": 2.154893065456902,
+      "grad_norm": 0.8513553738594055,
+      "learning_rate": 9.703598080956488e-07,
+      "loss": 0.0532,
+      "step": 13300
+    },
+    {
+      "epoch": 2.155055087491899,
+      "grad_norm": 0.8135280609130859,
+      "learning_rate": 9.700139374620602e-07,
+      "loss": 0.0648,
+      "step": 13301
+    },
+    {
+      "epoch": 2.1552171095268955,
+      "grad_norm": 0.821293830871582,
+      "learning_rate": 9.696681136423422e-07,
+      "loss": 0.0607,
+      "step": 13302
+    },
+    {
+      "epoch": 2.1553791315618924,
+      "grad_norm": 0.984480082988739,
+      "learning_rate": 9.693223366470767e-07,
+      "loss": 0.0664,
+      "step": 13303
+    },
+    {
+      "epoch": 2.1555411535968894,
+      "grad_norm": 0.8220646381378174,
+      "learning_rate": 9.689766064868434e-07,
+      "loss": 0.0597,
+      "step": 13304
+    },
+    {
+      "epoch": 2.155703175631886,
+      "grad_norm": 0.9679084420204163,
+      "learning_rate": 9.686309231722219e-07,
+      "loss": 0.0619,
+      "step": 13305
+    },
+    {
+      "epoch": 2.155865197666883,
+      "grad_norm": 1.0595329999923706,
+      "learning_rate": 9.682852867137865e-07,
+      "loss": 0.0668,
+      "step": 13306
+    },
+    {
+      "epoch": 2.1560272197018793,
+      "grad_norm": 0.989339292049408,
+      "learning_rate": 9.679396971221155e-07,
+      "loss": 0.0646,
+      "step": 13307
+    },
+    {
+      "epoch": 2.156189241736876,
+      "grad_norm": 1.0220208168029785,
+      "learning_rate": 9.675941544077833e-07,
+      "loss": 0.0686,
+      "step": 13308
+    },
+    {
+      "epoch": 2.156351263771873,
+      "grad_norm": 0.9169559478759766,
+      "learning_rate": 9.672486585813606e-07,
+      "loss": 0.0612,
+      "step": 13309
+    },
+    {
+      "epoch": 2.1565132858068696,
+      "grad_norm": 0.8436880707740784,
+      "learning_rate": 9.6690320965342e-07,
+      "loss": 0.0566,
+      "step": 13310
+    },
+    {
+      "epoch": 2.1566753078418666,
+      "grad_norm": 0.835830569267273,
+      "learning_rate": 9.665578076345307e-07,
+      "loss": 0.0622,
+      "step": 13311
+    },
+    {
+      "epoch": 2.156837329876863,
+      "grad_norm": 0.8049341440200806,
+      "learning_rate": 9.66212452535262e-07,
+      "loss": 0.0625,
+      "step": 13312
+    },
+    {
+      "epoch": 2.15699935191186,
+      "grad_norm": 0.9086779952049255,
+      "learning_rate": 9.658671443661804e-07,
+      "loss": 0.0625,
+      "step": 13313
+    },
+    {
+      "epoch": 2.157161373946857,
+      "grad_norm": 0.9593300223350525,
+      "learning_rate": 9.655218831378518e-07,
+      "loss": 0.0657,
+      "step": 13314
+    },
+    {
+      "epoch": 2.1573233959818534,
+      "grad_norm": 0.790436863899231,
+      "learning_rate": 9.651766688608402e-07,
+      "loss": 0.0626,
+      "step": 13315
+    },
+    {
+      "epoch": 2.1574854180168503,
+      "grad_norm": 0.783841073513031,
+      "learning_rate": 9.648315015457083e-07,
+      "loss": 0.0611,
+      "step": 13316
+    },
+    {
+      "epoch": 2.1576474400518473,
+      "grad_norm": 0.861705482006073,
+      "learning_rate": 9.644863812030176e-07,
+      "loss": 0.0595,
+      "step": 13317
+    },
+    {
+      "epoch": 2.1578094620868438,
+      "grad_norm": 0.8327277302742004,
+      "learning_rate": 9.641413078433274e-07,
+      "loss": 0.0562,
+      "step": 13318
+    },
+    {
+      "epoch": 2.1579714841218407,
+      "grad_norm": 0.9547343850135803,
+      "learning_rate": 9.637962814771976e-07,
+      "loss": 0.0606,
+      "step": 13319
+    },
+    {
+      "epoch": 2.158133506156837,
+      "grad_norm": 0.9020182490348816,
+      "learning_rate": 9.63451302115182e-07,
+      "loss": 0.0591,
+      "step": 13320
+    },
+    {
+      "epoch": 2.158295528191834,
+      "grad_norm": 0.896525502204895,
+      "learning_rate": 9.631063697678392e-07,
+      "loss": 0.0574,
+      "step": 13321
+    },
+    {
+      "epoch": 2.158457550226831,
+      "grad_norm": 0.833091139793396,
+      "learning_rate": 9.627614844457222e-07,
+      "loss": 0.0551,
+      "step": 13322
+    },
+    {
+      "epoch": 2.1586195722618275,
+      "grad_norm": 0.8933079838752747,
+      "learning_rate": 9.624166461593848e-07,
+      "loss": 0.0587,
+      "step": 13323
+    },
+    {
+      "epoch": 2.1587815942968245,
+      "grad_norm": 0.9775946736335754,
+      "learning_rate": 9.620718549193764e-07,
+      "loss": 0.0632,
+      "step": 13324
+    },
+    {
+      "epoch": 2.158943616331821,
+      "grad_norm": 0.898419976234436,
+      "learning_rate": 9.617271107362465e-07,
+      "loss": 0.0688,
+      "step": 13325
+    },
+    {
+      "epoch": 2.159105638366818,
+      "grad_norm": 0.8033084869384766,
+      "learning_rate": 9.61382413620546e-07,
+      "loss": 0.0575,
+      "step": 13326
+    },
+    {
+      "epoch": 2.159267660401815,
+      "grad_norm": 0.9245520830154419,
+      "learning_rate": 9.610377635828197e-07,
+      "loss": 0.0599,
+      "step": 13327
+    },
+    {
+      "epoch": 2.1594296824368113,
+      "grad_norm": 0.9904624819755554,
+      "learning_rate": 9.606931606336134e-07,
+      "loss": 0.0731,
+      "step": 13328
+    },
+    {
+      "epoch": 2.1595917044718083,
+      "grad_norm": 1.0654817819595337,
+      "learning_rate": 9.603486047834712e-07,
+      "loss": 0.0672,
+      "step": 13329
+    },
+    {
+      "epoch": 2.1597537265068047,
+      "grad_norm": 0.7820091843605042,
+      "learning_rate": 9.60004096042936e-07,
+      "loss": 0.0544,
+      "step": 13330
+    },
+    {
+      "epoch": 2.1599157485418017,
+      "grad_norm": 0.8341110348701477,
+      "learning_rate": 9.59659634422549e-07,
+      "loss": 0.054,
+      "step": 13331
+    },
+    {
+      "epoch": 2.1600777705767986,
+      "grad_norm": 0.8635512590408325,
+      "learning_rate": 9.593152199328494e-07,
+      "loss": 0.0546,
+      "step": 13332
+    },
+    {
+      "epoch": 2.160239792611795,
+      "grad_norm": 0.8399879932403564,
+      "learning_rate": 9.589708525843754e-07,
+      "loss": 0.0633,
+      "step": 13333
+    },
+    {
+      "epoch": 2.160401814646792,
+      "grad_norm": 0.8963909149169922,
+      "learning_rate": 9.586265323876653e-07,
+      "loss": 0.0589,
+      "step": 13334
+    },
+    {
+      "epoch": 2.1605638366817885,
+      "grad_norm": 0.9390884637832642,
+      "learning_rate": 9.582822593532514e-07,
+      "loss": 0.0639,
+      "step": 13335
+    },
+    {
+      "epoch": 2.1607258587167855,
+      "grad_norm": 1.0376230478286743,
+      "learning_rate": 9.579380334916704e-07,
+      "loss": 0.0714,
+      "step": 13336
+    },
+    {
+      "epoch": 2.1608878807517824,
+      "grad_norm": 0.9008492827415466,
+      "learning_rate": 9.575938548134548e-07,
+      "loss": 0.0586,
+      "step": 13337
+    },
+    {
+      "epoch": 2.161049902786779,
+      "grad_norm": 0.8083958625793457,
+      "learning_rate": 9.572497233291337e-07,
+      "loss": 0.0612,
+      "step": 13338
+    },
+    {
+      "epoch": 2.161211924821776,
+      "grad_norm": 0.9335343837738037,
+      "learning_rate": 9.56905639049238e-07,
+      "loss": 0.0678,
+      "step": 13339
+    },
+    {
+      "epoch": 2.1613739468567728,
+      "grad_norm": 0.9797310829162598,
+      "learning_rate": 9.56561601984294e-07,
+      "loss": 0.0699,
+      "step": 13340
+    },
+    {
+      "epoch": 2.1615359688917692,
+      "grad_norm": 0.8286275267601013,
+      "learning_rate": 9.562176121448322e-07,
+      "loss": 0.059,
+      "step": 13341
+    },
+    {
+      "epoch": 2.161697990926766,
+      "grad_norm": 0.887620210647583,
+      "learning_rate": 9.558736695413745e-07,
+      "loss": 0.0608,
+      "step": 13342
+    },
+    {
+      "epoch": 2.1618600129617627,
+      "grad_norm": 0.888256847858429,
+      "learning_rate": 9.55529774184446e-07,
+      "loss": 0.0678,
+      "step": 13343
+    },
+    {
+      "epoch": 2.1620220349967596,
+      "grad_norm": 0.908109724521637,
+      "learning_rate": 9.551859260845686e-07,
+      "loss": 0.0631,
+      "step": 13344
+    },
+    {
+      "epoch": 2.162184057031756,
+      "grad_norm": 1.2315661907196045,
+      "learning_rate": 9.548421252522635e-07,
+      "loss": 0.0606,
+      "step": 13345
+    },
+    {
+      "epoch": 2.162346079066753,
+      "grad_norm": 0.8499714136123657,
+      "learning_rate": 9.544983716980505e-07,
+      "loss": 0.06,
+      "step": 13346
+    },
+    {
+      "epoch": 2.16250810110175,
+      "grad_norm": 0.9646509885787964,
+      "learning_rate": 9.54154665432447e-07,
+      "loss": 0.0673,
+      "step": 13347
+    },
+    {
+      "epoch": 2.1626701231367464,
+      "grad_norm": 0.8329025506973267,
+      "learning_rate": 9.5381100646597e-07,
+      "loss": 0.0613,
+      "step": 13348
+    },
+    {
+      "epoch": 2.1628321451717434,
+      "grad_norm": 0.9725831151008606,
+      "learning_rate": 9.534673948091344e-07,
+      "loss": 0.0676,
+      "step": 13349
+    },
+    {
+      "epoch": 2.1629941672067403,
+      "grad_norm": 0.8599954843521118,
+      "learning_rate": 9.531238304724538e-07,
+      "loss": 0.06,
+      "step": 13350
+    },
+    {
+      "epoch": 2.163156189241737,
+      "grad_norm": 0.8382558822631836,
+      "learning_rate": 9.52780313466441e-07,
+      "loss": 0.0622,
+      "step": 13351
+    },
+    {
+      "epoch": 2.1633182112767337,
+      "grad_norm": 0.8793473839759827,
+      "learning_rate": 9.524368438016071e-07,
+      "loss": 0.0563,
+      "step": 13352
+    },
+    {
+      "epoch": 2.1634802333117302,
+      "grad_norm": 0.9254469275474548,
+      "learning_rate": 9.520934214884598e-07,
+      "loss": 0.0632,
+      "step": 13353
+    },
+    {
+      "epoch": 2.163642255346727,
+      "grad_norm": 0.8985252976417542,
+      "learning_rate": 9.517500465375071e-07,
+      "loss": 0.0598,
+      "step": 13354
+    },
+    {
+      "epoch": 2.163804277381724,
+      "grad_norm": 0.7945815920829773,
+      "learning_rate": 9.514067189592583e-07,
+      "loss": 0.0595,
+      "step": 13355
+    },
+    {
+      "epoch": 2.1639662994167206,
+      "grad_norm": 0.8715730309486389,
+      "learning_rate": 9.510634387642151e-07,
+      "loss": 0.0598,
+      "step": 13356
+    },
+    {
+      "epoch": 2.1641283214517175,
+      "grad_norm": 0.8382450342178345,
+      "learning_rate": 9.507202059628826e-07,
+      "loss": 0.0691,
+      "step": 13357
+    },
+    {
+      "epoch": 2.164290343486714,
+      "grad_norm": 1.0573970079421997,
+      "learning_rate": 9.503770205657625e-07,
+      "loss": 0.0636,
+      "step": 13358
+    },
+    {
+      "epoch": 2.164452365521711,
+      "grad_norm": 0.8974131345748901,
+      "learning_rate": 9.500338825833555e-07,
+      "loss": 0.0545,
+      "step": 13359
+    },
+    {
+      "epoch": 2.164614387556708,
+      "grad_norm": 0.8579263687133789,
+      "learning_rate": 9.496907920261609e-07,
+      "loss": 0.063,
+      "step": 13360
+    },
+    {
+      "epoch": 2.1647764095917044,
+      "grad_norm": 0.8979440331459045,
+      "learning_rate": 9.493477489046762e-07,
+      "loss": 0.0669,
+      "step": 13361
+    },
+    {
+      "epoch": 2.1649384316267013,
+      "grad_norm": 0.8528677225112915,
+      "learning_rate": 9.490047532293984e-07,
+      "loss": 0.0587,
+      "step": 13362
+    },
+    {
+      "epoch": 2.165100453661698,
+      "grad_norm": 0.8880028128623962,
+      "learning_rate": 9.486618050108223e-07,
+      "loss": 0.0628,
+      "step": 13363
+    },
+    {
+      "epoch": 2.1652624756966947,
+      "grad_norm": 0.9306471347808838,
+      "learning_rate": 9.48318904259439e-07,
+      "loss": 0.0645,
+      "step": 13364
+    },
+    {
+      "epoch": 2.1654244977316917,
+      "grad_norm": 0.8764260411262512,
+      "learning_rate": 9.479760509857433e-07,
+      "loss": 0.0621,
+      "step": 13365
+    },
+    {
+      "epoch": 2.165586519766688,
+      "grad_norm": 0.8734313249588013,
+      "learning_rate": 9.476332452002245e-07,
+      "loss": 0.0585,
+      "step": 13366
+    },
+    {
+      "epoch": 2.165748541801685,
+      "grad_norm": 0.9644814729690552,
+      "learning_rate": 9.472904869133726e-07,
+      "loss": 0.07,
+      "step": 13367
+    },
+    {
+      "epoch": 2.1659105638366816,
+      "grad_norm": 0.8495195508003235,
+      "learning_rate": 9.469477761356727e-07,
+      "loss": 0.0591,
+      "step": 13368
+    },
+    {
+      "epoch": 2.1660725858716785,
+      "grad_norm": 0.8992612361907959,
+      "learning_rate": 9.466051128776133e-07,
+      "loss": 0.0637,
+      "step": 13369
+    },
+    {
+      "epoch": 2.1662346079066754,
+      "grad_norm": 0.9654873013496399,
+      "learning_rate": 9.462624971496793e-07,
+      "loss": 0.0631,
+      "step": 13370
+    },
+    {
+      "epoch": 2.166396629941672,
+      "grad_norm": 1.062732219696045,
+      "learning_rate": 9.459199289623519e-07,
+      "loss": 0.0635,
+      "step": 13371
+    },
+    {
+      "epoch": 2.166558651976669,
+      "grad_norm": 0.927365243434906,
+      "learning_rate": 9.455774083261138e-07,
+      "loss": 0.0664,
+      "step": 13372
+    },
+    {
+      "epoch": 2.166720674011666,
+      "grad_norm": 0.9588886499404907,
+      "learning_rate": 9.452349352514448e-07,
+      "loss": 0.0605,
+      "step": 13373
+    },
+    {
+      "epoch": 2.1668826960466623,
+      "grad_norm": 0.8306574821472168,
+      "learning_rate": 9.448925097488257e-07,
+      "loss": 0.0587,
+      "step": 13374
+    },
+    {
+      "epoch": 2.167044718081659,
+      "grad_norm": 0.8604657649993896,
+      "learning_rate": 9.445501318287317e-07,
+      "loss": 0.0605,
+      "step": 13375
+    },
+    {
+      "epoch": 2.1672067401166557,
+      "grad_norm": 0.8701779842376709,
+      "learning_rate": 9.442078015016398e-07,
+      "loss": 0.0555,
+      "step": 13376
+    },
+    {
+      "epoch": 2.1673687621516526,
+      "grad_norm": 1.0996429920196533,
+      "learning_rate": 9.43865518778024e-07,
+      "loss": 0.068,
+      "step": 13377
+    },
+    {
+      "epoch": 2.1675307841866496,
+      "grad_norm": 0.9881284236907959,
+      "learning_rate": 9.435232836683577e-07,
+      "loss": 0.0707,
+      "step": 13378
+    },
+    {
+      "epoch": 2.167692806221646,
+      "grad_norm": 1.0332772731781006,
+      "learning_rate": 9.431810961831123e-07,
+      "loss": 0.0676,
+      "step": 13379
+    },
+    {
+      "epoch": 2.167854828256643,
+      "grad_norm": 0.9715332984924316,
+      "learning_rate": 9.42838956332758e-07,
+      "loss": 0.0713,
+      "step": 13380
+    },
+    {
+      "epoch": 2.1680168502916395,
+      "grad_norm": 1.1476157903671265,
+      "learning_rate": 9.42496864127764e-07,
+      "loss": 0.0618,
+      "step": 13381
+    },
+    {
+      "epoch": 2.1681788723266364,
+      "grad_norm": 0.9035125970840454,
+      "learning_rate": 9.421548195785962e-07,
+      "loss": 0.0641,
+      "step": 13382
+    },
+    {
+      "epoch": 2.1683408943616334,
+      "grad_norm": 1.018491268157959,
+      "learning_rate": 9.418128226957202e-07,
+      "loss": 0.0682,
+      "step": 13383
+    },
+    {
+      "epoch": 2.16850291639663,
+      "grad_norm": 0.8716436624526978,
+      "learning_rate": 9.414708734896019e-07,
+      "loss": 0.0631,
+      "step": 13384
+    },
+    {
+      "epoch": 2.1686649384316268,
+      "grad_norm": 0.9046812057495117,
+      "learning_rate": 9.411289719707039e-07,
+      "loss": 0.0658,
+      "step": 13385
+    },
+    {
+      "epoch": 2.1688269604666233,
+      "grad_norm": 0.9490453004837036,
+      "learning_rate": 9.407871181494865e-07,
+      "loss": 0.0601,
+      "step": 13386
+    },
+    {
+      "epoch": 2.16898898250162,
+      "grad_norm": 0.8839377164840698,
+      "learning_rate": 9.40445312036409e-07,
+      "loss": 0.0684,
+      "step": 13387
+    },
+    {
+      "epoch": 2.169151004536617,
+      "grad_norm": 1.0054982900619507,
+      "learning_rate": 9.401035536419326e-07,
+      "loss": 0.0635,
+      "step": 13388
+    },
+    {
+      "epoch": 2.1693130265716136,
+      "grad_norm": 0.8407875299453735,
+      "learning_rate": 9.397618429765118e-07,
+      "loss": 0.0653,
+      "step": 13389
+    },
+    {
+      "epoch": 2.1694750486066106,
+      "grad_norm": 0.8645152449607849,
+      "learning_rate": 9.394201800506028e-07,
+      "loss": 0.0612,
+      "step": 13390
+    },
+    {
+      "epoch": 2.169637070641607,
+      "grad_norm": 1.003045678138733,
+      "learning_rate": 9.390785648746598e-07,
+      "loss": 0.0665,
+      "step": 13391
+    },
+    {
+      "epoch": 2.169799092676604,
+      "grad_norm": 0.8517448306083679,
+      "learning_rate": 9.387369974591353e-07,
+      "loss": 0.065,
+      "step": 13392
+    },
+    {
+      "epoch": 2.169961114711601,
+      "grad_norm": 0.9961963891983032,
+      "learning_rate": 9.383954778144807e-07,
+      "loss": 0.0635,
+      "step": 13393
+    },
+    {
+      "epoch": 2.1701231367465974,
+      "grad_norm": 0.8124943375587463,
+      "learning_rate": 9.380540059511453e-07,
+      "loss": 0.0552,
+      "step": 13394
+    },
+    {
+      "epoch": 2.1702851587815943,
+      "grad_norm": 0.801011323928833,
+      "learning_rate": 9.377125818795777e-07,
+      "loss": 0.0616,
+      "step": 13395
+    },
+    {
+      "epoch": 2.1704471808165913,
+      "grad_norm": 0.938008725643158,
+      "learning_rate": 9.373712056102249e-07,
+      "loss": 0.0663,
+      "step": 13396
+    },
+    {
+      "epoch": 2.1706092028515878,
+      "grad_norm": 0.9394778609275818,
+      "learning_rate": 9.370298771535302e-07,
+      "loss": 0.0658,
+      "step": 13397
+    },
+    {
+      "epoch": 2.1707712248865847,
+      "grad_norm": 0.9980944991111755,
+      "learning_rate": 9.366885965199398e-07,
+      "loss": 0.0664,
+      "step": 13398
+    },
+    {
+      "epoch": 2.170933246921581,
+      "grad_norm": 0.8614987730979919,
+      "learning_rate": 9.363473637198964e-07,
+      "loss": 0.0605,
+      "step": 13399
+    },
+    {
+      "epoch": 2.171095268956578,
+      "grad_norm": 0.7758714556694031,
+      "learning_rate": 9.360061787638383e-07,
+      "loss": 0.0641,
+      "step": 13400
+    },
+    {
+      "epoch": 2.171257290991575,
+      "grad_norm": 0.9587964415550232,
+      "learning_rate": 9.356650416622065e-07,
+      "loss": 0.0654,
+      "step": 13401
+    },
+    {
+      "epoch": 2.1714193130265715,
+      "grad_norm": 0.970716118812561,
+      "learning_rate": 9.353239524254382e-07,
+      "loss": 0.0568,
+      "step": 13402
+    },
+    {
+      "epoch": 2.1715813350615685,
+      "grad_norm": 0.8484756946563721,
+      "learning_rate": 9.349829110639718e-07,
+      "loss": 0.0628,
+      "step": 13403
+    },
+    {
+      "epoch": 2.171743357096565,
+      "grad_norm": 0.8023274540901184,
+      "learning_rate": 9.346419175882407e-07,
+      "loss": 0.0593,
+      "step": 13404
+    },
+    {
+      "epoch": 2.171905379131562,
+      "grad_norm": 0.8955522179603577,
+      "learning_rate": 9.343009720086785e-07,
+      "loss": 0.0591,
+      "step": 13405
+    },
+    {
+      "epoch": 2.172067401166559,
+      "grad_norm": 0.9078856706619263,
+      "learning_rate": 9.339600743357177e-07,
+      "loss": 0.0688,
+      "step": 13406
+    },
+    {
+      "epoch": 2.1722294232015553,
+      "grad_norm": 1.0572879314422607,
+      "learning_rate": 9.33619224579789e-07,
+      "loss": 0.0575,
+      "step": 13407
+    },
+    {
+      "epoch": 2.1723914452365523,
+      "grad_norm": 0.8648132681846619,
+      "learning_rate": 9.332784227513212e-07,
+      "loss": 0.0621,
+      "step": 13408
+    },
+    {
+      "epoch": 2.1725534672715487,
+      "grad_norm": 0.8254750967025757,
+      "learning_rate": 9.329376688607425e-07,
+      "loss": 0.06,
+      "step": 13409
+    },
+    {
+      "epoch": 2.1727154893065457,
+      "grad_norm": 0.8656960725784302,
+      "learning_rate": 9.325969629184789e-07,
+      "loss": 0.0655,
+      "step": 13410
+    },
+    {
+      "epoch": 2.1728775113415426,
+      "grad_norm": 0.8903995156288147,
+      "learning_rate": 9.32256304934955e-07,
+      "loss": 0.0603,
+      "step": 13411
+    },
+    {
+      "epoch": 2.173039533376539,
+      "grad_norm": 0.9863763451576233,
+      "learning_rate": 9.319156949205943e-07,
+      "loss": 0.0615,
+      "step": 13412
+    },
+    {
+      "epoch": 2.173201555411536,
+      "grad_norm": 0.7889726161956787,
+      "learning_rate": 9.315751328858189e-07,
+      "loss": 0.0544,
+      "step": 13413
+    },
+    {
+      "epoch": 2.1733635774465325,
+      "grad_norm": 0.8542593717575073,
+      "learning_rate": 9.312346188410496e-07,
+      "loss": 0.0636,
+      "step": 13414
+    },
+    {
+      "epoch": 2.1735255994815295,
+      "grad_norm": 0.8379946947097778,
+      "learning_rate": 9.308941527967039e-07,
+      "loss": 0.0606,
+      "step": 13415
+    },
+    {
+      "epoch": 2.1736876215165264,
+      "grad_norm": 0.9521237015724182,
+      "learning_rate": 9.30553734763199e-07,
+      "loss": 0.0582,
+      "step": 13416
+    },
+    {
+      "epoch": 2.173849643551523,
+      "grad_norm": 0.8582279682159424,
+      "learning_rate": 9.302133647509526e-07,
+      "loss": 0.0613,
+      "step": 13417
+    },
+    {
+      "epoch": 2.17401166558652,
+      "grad_norm": 0.9010220766067505,
+      "learning_rate": 9.298730427703795e-07,
+      "loss": 0.0658,
+      "step": 13418
+    },
+    {
+      "epoch": 2.1741736876215167,
+      "grad_norm": 1.0892761945724487,
+      "learning_rate": 9.295327688318906e-07,
+      "loss": 0.0687,
+      "step": 13419
+    },
+    {
+      "epoch": 2.1743357096565132,
+      "grad_norm": 0.9256922006607056,
+      "learning_rate": 9.291925429458987e-07,
+      "loss": 0.0606,
+      "step": 13420
+    },
+    {
+      "epoch": 2.17449773169151,
+      "grad_norm": 1.116697907447815,
+      "learning_rate": 9.288523651228134e-07,
+      "loss": 0.0613,
+      "step": 13421
+    },
+    {
+      "epoch": 2.1746597537265067,
+      "grad_norm": 0.9603244066238403,
+      "learning_rate": 9.285122353730439e-07,
+      "loss": 0.0636,
+      "step": 13422
+    },
+    {
+      "epoch": 2.1748217757615036,
+      "grad_norm": 0.980385959148407,
+      "learning_rate": 9.281721537069971e-07,
+      "loss": 0.0623,
+      "step": 13423
+    },
+    {
+      "epoch": 2.1749837977965005,
+      "grad_norm": 0.8736245036125183,
+      "learning_rate": 9.278321201350784e-07,
+      "loss": 0.0626,
+      "step": 13424
+    },
+    {
+      "epoch": 2.175145819831497,
+      "grad_norm": 0.8763680458068848,
+      "learning_rate": 9.274921346676935e-07,
+      "loss": 0.0618,
+      "step": 13425
+    },
+    {
+      "epoch": 2.175307841866494,
+      "grad_norm": 0.9780583381652832,
+      "learning_rate": 9.271521973152418e-07,
+      "loss": 0.0612,
+      "step": 13426
+    },
+    {
+      "epoch": 2.1754698639014904,
+      "grad_norm": 1.1210222244262695,
+      "learning_rate": 9.268123080881275e-07,
+      "loss": 0.0623,
+      "step": 13427
+    },
+    {
+      "epoch": 2.1756318859364874,
+      "grad_norm": 0.9153855443000793,
+      "learning_rate": 9.264724669967498e-07,
+      "loss": 0.0649,
+      "step": 13428
+    },
+    {
+      "epoch": 2.1757939079714843,
+      "grad_norm": 0.8484050631523132,
+      "learning_rate": 9.261326740515075e-07,
+      "loss": 0.0588,
+      "step": 13429
+    },
+    {
+      "epoch": 2.175955930006481,
+      "grad_norm": 1.0825287103652954,
+      "learning_rate": 9.257929292627956e-07,
+      "loss": 0.0637,
+      "step": 13430
+    },
+    {
+      "epoch": 2.1761179520414777,
+      "grad_norm": 0.7929718494415283,
+      "learning_rate": 9.254532326410101e-07,
+      "loss": 0.0594,
+      "step": 13431
+    },
+    {
+      "epoch": 2.176279974076474,
+      "grad_norm": 0.9638476967811584,
+      "learning_rate": 9.251135841965467e-07,
+      "loss": 0.0708,
+      "step": 13432
+    },
+    {
+      "epoch": 2.176441996111471,
+      "grad_norm": 0.856880247592926,
+      "learning_rate": 9.24773983939796e-07,
+      "loss": 0.0646,
+      "step": 13433
+    },
+    {
+      "epoch": 2.176604018146468,
+      "grad_norm": 0.894572377204895,
+      "learning_rate": 9.244344318811491e-07,
+      "loss": 0.0658,
+      "step": 13434
+    },
+    {
+      "epoch": 2.1767660401814646,
+      "grad_norm": 0.8938385248184204,
+      "learning_rate": 9.240949280309949e-07,
+      "loss": 0.0606,
+      "step": 13435
+    },
+    {
+      "epoch": 2.1769280622164615,
+      "grad_norm": 0.8438916206359863,
+      "learning_rate": 9.237554723997242e-07,
+      "loss": 0.0586,
+      "step": 13436
+    },
+    {
+      "epoch": 2.177090084251458,
+      "grad_norm": 0.9637516140937805,
+      "learning_rate": 9.234160649977206e-07,
+      "loss": 0.0687,
+      "step": 13437
+    },
+    {
+      "epoch": 2.177252106286455,
+      "grad_norm": 0.8094197511672974,
+      "learning_rate": 9.230767058353701e-07,
+      "loss": 0.0559,
+      "step": 13438
+    },
+    {
+      "epoch": 2.177414128321452,
+      "grad_norm": 0.8560066223144531,
+      "learning_rate": 9.227373949230567e-07,
+      "loss": 0.0559,
+      "step": 13439
+    },
+    {
+      "epoch": 2.1775761503564484,
+      "grad_norm": 1.0146995782852173,
+      "learning_rate": 9.223981322711617e-07,
+      "loss": 0.0702,
+      "step": 13440
+    },
+    {
+      "epoch": 2.1777381723914453,
+      "grad_norm": 0.8843867182731628,
+      "learning_rate": 9.220589178900663e-07,
+      "loss": 0.0625,
+      "step": 13441
+    },
+    {
+      "epoch": 2.1779001944264422,
+      "grad_norm": 0.7497606873512268,
+      "learning_rate": 9.217197517901494e-07,
+      "loss": 0.0572,
+      "step": 13442
+    },
+    {
+      "epoch": 2.1780622164614387,
+      "grad_norm": 0.8896629810333252,
+      "learning_rate": 9.213806339817897e-07,
+      "loss": 0.0626,
+      "step": 13443
+    },
+    {
+      "epoch": 2.1782242384964356,
+      "grad_norm": 0.8931178450584412,
+      "learning_rate": 9.210415644753615e-07,
+      "loss": 0.0585,
+      "step": 13444
+    },
+    {
+      "epoch": 2.178386260531432,
+      "grad_norm": 0.8438544869422913,
+      "learning_rate": 9.207025432812397e-07,
+      "loss": 0.0594,
+      "step": 13445
+    },
+    {
+      "epoch": 2.178548282566429,
+      "grad_norm": 0.9344618320465088,
+      "learning_rate": 9.203635704097988e-07,
+      "loss": 0.0618,
+      "step": 13446
+    },
+    {
+      "epoch": 2.1787103046014256,
+      "grad_norm": 0.863814115524292,
+      "learning_rate": 9.20024645871411e-07,
+      "loss": 0.0653,
+      "step": 13447
+    },
+    {
+      "epoch": 2.1788723266364225,
+      "grad_norm": 0.8747982382774353,
+      "learning_rate": 9.196857696764446e-07,
+      "loss": 0.0585,
+      "step": 13448
+    },
+    {
+      "epoch": 2.1790343486714194,
+      "grad_norm": 0.9165652394294739,
+      "learning_rate": 9.193469418352696e-07,
+      "loss": 0.069,
+      "step": 13449
+    },
+    {
+      "epoch": 2.179196370706416,
+      "grad_norm": 0.9408144950866699,
+      "learning_rate": 9.190081623582531e-07,
+      "loss": 0.0711,
+      "step": 13450
+    },
+    {
+      "epoch": 2.179358392741413,
+      "grad_norm": 0.9184764623641968,
+      "learning_rate": 9.186694312557606e-07,
+      "loss": 0.0666,
+      "step": 13451
+    },
+    {
+      "epoch": 2.17952041477641,
+      "grad_norm": 0.8569600582122803,
+      "learning_rate": 9.183307485381571e-07,
+      "loss": 0.0613,
+      "step": 13452
+    },
+    {
+      "epoch": 2.1796824368114063,
+      "grad_norm": 0.8777344226837158,
+      "learning_rate": 9.17992114215805e-07,
+      "loss": 0.0628,
+      "step": 13453
+    },
+    {
+      "epoch": 2.179844458846403,
+      "grad_norm": 0.9402377605438232,
+      "learning_rate": 9.176535282990656e-07,
+      "loss": 0.0669,
+      "step": 13454
+    },
+    {
+      "epoch": 2.1800064808813997,
+      "grad_norm": 0.9360073208808899,
+      "learning_rate": 9.173149907982993e-07,
+      "loss": 0.0617,
+      "step": 13455
+    },
+    {
+      "epoch": 2.1801685029163966,
+      "grad_norm": 0.9287109971046448,
+      "learning_rate": 9.169765017238641e-07,
+      "loss": 0.0681,
+      "step": 13456
+    },
+    {
+      "epoch": 2.1803305249513936,
+      "grad_norm": 0.8160260915756226,
+      "learning_rate": 9.166380610861172e-07,
+      "loss": 0.0637,
+      "step": 13457
+    },
+    {
+      "epoch": 2.18049254698639,
+      "grad_norm": 0.8967964053153992,
+      "learning_rate": 9.162996688954148e-07,
+      "loss": 0.0634,
+      "step": 13458
+    },
+    {
+      "epoch": 2.180654569021387,
+      "grad_norm": 0.968971848487854,
+      "learning_rate": 9.159613251621083e-07,
+      "loss": 0.065,
+      "step": 13459
+    },
+    {
+      "epoch": 2.1808165910563835,
+      "grad_norm": 0.8118306994438171,
+      "learning_rate": 9.156230298965529e-07,
+      "loss": 0.0613,
+      "step": 13460
+    },
+    {
+      "epoch": 2.1809786130913804,
+      "grad_norm": 0.9221633672714233,
+      "learning_rate": 9.152847831090986e-07,
+      "loss": 0.0642,
+      "step": 13461
+    },
+    {
+      "epoch": 2.1811406351263773,
+      "grad_norm": 0.8746361136436462,
+      "learning_rate": 9.149465848100958e-07,
+      "loss": 0.0629,
+      "step": 13462
+    },
+    {
+      "epoch": 2.181302657161374,
+      "grad_norm": 0.9616087079048157,
+      "learning_rate": 9.14608435009891e-07,
+      "loss": 0.0727,
+      "step": 13463
+    },
+    {
+      "epoch": 2.1814646791963708,
+      "grad_norm": 0.8658609390258789,
+      "learning_rate": 9.142703337188305e-07,
+      "loss": 0.06,
+      "step": 13464
+    },
+    {
+      "epoch": 2.1816267012313673,
+      "grad_norm": 0.833065390586853,
+      "learning_rate": 9.139322809472623e-07,
+      "loss": 0.0552,
+      "step": 13465
+    },
+    {
+      "epoch": 2.181788723266364,
+      "grad_norm": 0.8363578915596008,
+      "learning_rate": 9.135942767055272e-07,
+      "loss": 0.0673,
+      "step": 13466
+    },
+    {
+      "epoch": 2.181950745301361,
+      "grad_norm": 0.9329082369804382,
+      "learning_rate": 9.132563210039683e-07,
+      "loss": 0.0571,
+      "step": 13467
+    },
+    {
+      "epoch": 2.1821127673363576,
+      "grad_norm": 0.8257620334625244,
+      "learning_rate": 9.129184138529259e-07,
+      "loss": 0.0663,
+      "step": 13468
+    },
+    {
+      "epoch": 2.1822747893713546,
+      "grad_norm": 1.1378121376037598,
+      "learning_rate": 9.125805552627395e-07,
+      "loss": 0.0648,
+      "step": 13469
+    },
+    {
+      "epoch": 2.182436811406351,
+      "grad_norm": 0.8082774877548218,
+      "learning_rate": 9.122427452437465e-07,
+      "loss": 0.0596,
+      "step": 13470
+    },
+    {
+      "epoch": 2.182598833441348,
+      "grad_norm": 1.0287437438964844,
+      "learning_rate": 9.119049838062832e-07,
+      "loss": 0.067,
+      "step": 13471
+    },
+    {
+      "epoch": 2.182760855476345,
+      "grad_norm": 0.9428204894065857,
+      "learning_rate": 9.115672709606846e-07,
+      "loss": 0.0679,
+      "step": 13472
+    },
+    {
+      "epoch": 2.1829228775113414,
+      "grad_norm": 0.9620710611343384,
+      "learning_rate": 9.11229606717284e-07,
+      "loss": 0.0726,
+      "step": 13473
+    },
+    {
+      "epoch": 2.1830848995463383,
+      "grad_norm": 0.8988024592399597,
+      "learning_rate": 9.108919910864111e-07,
+      "loss": 0.063,
+      "step": 13474
+    },
+    {
+      "epoch": 2.1832469215813353,
+      "grad_norm": 0.978762686252594,
+      "learning_rate": 9.105544240783987e-07,
+      "loss": 0.0629,
+      "step": 13475
+    },
+    {
+      "epoch": 2.1834089436163318,
+      "grad_norm": 0.9975540041923523,
+      "learning_rate": 9.102169057035753e-07,
+      "loss": 0.0659,
+      "step": 13476
+    },
+    {
+      "epoch": 2.1835709656513287,
+      "grad_norm": 0.9281214475631714,
+      "learning_rate": 9.098794359722668e-07,
+      "loss": 0.0587,
+      "step": 13477
+    },
+    {
+      "epoch": 2.183732987686325,
+      "grad_norm": 0.8466314673423767,
+      "learning_rate": 9.095420148947984e-07,
+      "loss": 0.0631,
+      "step": 13478
+    },
+    {
+      "epoch": 2.183895009721322,
+      "grad_norm": 0.8057236671447754,
+      "learning_rate": 9.092046424814962e-07,
+      "loss": 0.0631,
+      "step": 13479
+    },
+    {
+      "epoch": 2.184057031756319,
+      "grad_norm": 0.8913443088531494,
+      "learning_rate": 9.088673187426836e-07,
+      "loss": 0.0579,
+      "step": 13480
+    },
+    {
+      "epoch": 2.1842190537913155,
+      "grad_norm": 0.8665725588798523,
+      "learning_rate": 9.085300436886793e-07,
+      "loss": 0.0602,
+      "step": 13481
+    },
+    {
+      "epoch": 2.1843810758263125,
+      "grad_norm": 0.8465587496757507,
+      "learning_rate": 9.081928173298046e-07,
+      "loss": 0.0576,
+      "step": 13482
+    },
+    {
+      "epoch": 2.184543097861309,
+      "grad_norm": 0.9389297962188721,
+      "learning_rate": 9.078556396763777e-07,
+      "loss": 0.0637,
+      "step": 13483
+    },
+    {
+      "epoch": 2.184705119896306,
+      "grad_norm": 0.9527860879898071,
+      "learning_rate": 9.075185107387149e-07,
+      "loss": 0.0648,
+      "step": 13484
+    },
+    {
+      "epoch": 2.184867141931303,
+      "grad_norm": 0.8697720170021057,
+      "learning_rate": 9.071814305271323e-07,
+      "loss": 0.0654,
+      "step": 13485
+    },
+    {
+      "epoch": 2.1850291639662993,
+      "grad_norm": 0.963236391544342,
+      "learning_rate": 9.068443990519432e-07,
+      "loss": 0.066,
+      "step": 13486
+    },
+    {
+      "epoch": 2.1851911860012962,
+      "grad_norm": 0.8899818658828735,
+      "learning_rate": 9.065074163234602e-07,
+      "loss": 0.0604,
+      "step": 13487
+    },
+    {
+      "epoch": 2.1853532080362927,
+      "grad_norm": 0.8355003595352173,
+      "learning_rate": 9.061704823519943e-07,
+      "loss": 0.067,
+      "step": 13488
+    },
+    {
+      "epoch": 2.1855152300712897,
+      "grad_norm": 0.9532532691955566,
+      "learning_rate": 9.058335971478543e-07,
+      "loss": 0.0583,
+      "step": 13489
+    },
+    {
+      "epoch": 2.1856772521062866,
+      "grad_norm": 0.9321505427360535,
+      "learning_rate": 9.054967607213486e-07,
+      "loss": 0.0689,
+      "step": 13490
+    },
+    {
+      "epoch": 2.185839274141283,
+      "grad_norm": 0.9366322755813599,
+      "learning_rate": 9.051599730827842e-07,
+      "loss": 0.0692,
+      "step": 13491
+    },
+    {
+      "epoch": 2.18600129617628,
+      "grad_norm": 0.8937353491783142,
+      "learning_rate": 9.048232342424642e-07,
+      "loss": 0.0551,
+      "step": 13492
+    },
+    {
+      "epoch": 2.1861633182112765,
+      "grad_norm": 0.9362406730651855,
+      "learning_rate": 9.044865442106923e-07,
+      "loss": 0.0698,
+      "step": 13493
+    },
+    {
+      "epoch": 2.1863253402462735,
+      "grad_norm": 0.7960699200630188,
+      "learning_rate": 9.04149902997773e-07,
+      "loss": 0.0627,
+      "step": 13494
+    },
+    {
+      "epoch": 2.1864873622812704,
+      "grad_norm": 0.9175137281417847,
+      "learning_rate": 9.038133106140034e-07,
+      "loss": 0.0695,
+      "step": 13495
+    },
+    {
+      "epoch": 2.186649384316267,
+      "grad_norm": 1.0003280639648438,
+      "learning_rate": 9.034767670696842e-07,
+      "loss": 0.0628,
+      "step": 13496
+    },
+    {
+      "epoch": 2.186811406351264,
+      "grad_norm": 1.0133966207504272,
+      "learning_rate": 9.031402723751123e-07,
+      "loss": 0.068,
+      "step": 13497
+    },
+    {
+      "epoch": 2.1869734283862607,
+      "grad_norm": 1.0157874822616577,
+      "learning_rate": 9.028038265405836e-07,
+      "loss": 0.0642,
+      "step": 13498
+    },
+    {
+      "epoch": 2.1871354504212572,
+      "grad_norm": 0.7945261001586914,
+      "learning_rate": 9.02467429576393e-07,
+      "loss": 0.0581,
+      "step": 13499
+    },
+    {
+      "epoch": 2.187297472456254,
+      "grad_norm": 1.0244712829589844,
+      "learning_rate": 9.021310814928328e-07,
+      "loss": 0.0721,
+      "step": 13500
+    },
+    {
+      "epoch": 2.1874594944912507,
+      "grad_norm": 0.8607012033462524,
+      "learning_rate": 9.01794782300195e-07,
+      "loss": 0.0603,
+      "step": 13501
+    },
+    {
+      "epoch": 2.1876215165262476,
+      "grad_norm": 1.0268734693527222,
+      "learning_rate": 9.01458532008769e-07,
+      "loss": 0.0597,
+      "step": 13502
+    },
+    {
+      "epoch": 2.1877835385612445,
+      "grad_norm": 0.8796284198760986,
+      "learning_rate": 9.011223306288436e-07,
+      "loss": 0.0594,
+      "step": 13503
+    },
+    {
+      "epoch": 2.187945560596241,
+      "grad_norm": 0.8240768313407898,
+      "learning_rate": 9.007861781707056e-07,
+      "loss": 0.0583,
+      "step": 13504
+    },
+    {
+      "epoch": 2.188107582631238,
+      "grad_norm": 0.7857369780540466,
+      "learning_rate": 9.004500746446407e-07,
+      "loss": 0.0549,
+      "step": 13505
+    },
+    {
+      "epoch": 2.1882696046662344,
+      "grad_norm": 0.9636929631233215,
+      "learning_rate": 9.001140200609334e-07,
+      "loss": 0.0667,
+      "step": 13506
+    },
+    {
+      "epoch": 2.1884316267012314,
+      "grad_norm": 0.9375138878822327,
+      "learning_rate": 8.997780144298641e-07,
+      "loss": 0.0653,
+      "step": 13507
+    },
+    {
+      "epoch": 2.1885936487362283,
+      "grad_norm": 0.9468481540679932,
+      "learning_rate": 8.994420577617155e-07,
+      "loss": 0.0652,
+      "step": 13508
+    },
+    {
+      "epoch": 2.188755670771225,
+      "grad_norm": 0.946871280670166,
+      "learning_rate": 8.991061500667674e-07,
+      "loss": 0.0666,
+      "step": 13509
+    },
+    {
+      "epoch": 2.1889176928062217,
+      "grad_norm": 0.8926617503166199,
+      "learning_rate": 8.987702913552964e-07,
+      "loss": 0.066,
+      "step": 13510
+    },
+    {
+      "epoch": 2.189079714841218,
+      "grad_norm": 0.8519583940505981,
+      "learning_rate": 8.984344816375798e-07,
+      "loss": 0.0596,
+      "step": 13511
+    },
+    {
+      "epoch": 2.189241736876215,
+      "grad_norm": 0.8857722878456116,
+      "learning_rate": 8.980987209238922e-07,
+      "loss": 0.0557,
+      "step": 13512
+    },
+    {
+      "epoch": 2.189403758911212,
+      "grad_norm": 0.9426934719085693,
+      "learning_rate": 8.977630092245071e-07,
+      "loss": 0.0582,
+      "step": 13513
+    },
+    {
+      "epoch": 2.1895657809462086,
+      "grad_norm": 0.86836177110672,
+      "learning_rate": 8.974273465496966e-07,
+      "loss": 0.062,
+      "step": 13514
+    },
+    {
+      "epoch": 2.1897278029812055,
+      "grad_norm": 0.8490039110183716,
+      "learning_rate": 8.970917329097312e-07,
+      "loss": 0.0581,
+      "step": 13515
+    },
+    {
+      "epoch": 2.189889825016202,
+      "grad_norm": 0.9485715627670288,
+      "learning_rate": 8.967561683148798e-07,
+      "loss": 0.0639,
+      "step": 13516
+    },
+    {
+      "epoch": 2.190051847051199,
+      "grad_norm": 0.7836689949035645,
+      "learning_rate": 8.964206527754099e-07,
+      "loss": 0.0571,
+      "step": 13517
+    },
+    {
+      "epoch": 2.190213869086196,
+      "grad_norm": 0.9445635080337524,
+      "learning_rate": 8.960851863015874e-07,
+      "loss": 0.0558,
+      "step": 13518
+    },
+    {
+      "epoch": 2.1903758911211924,
+      "grad_norm": 0.8634076118469238,
+      "learning_rate": 8.957497689036768e-07,
+      "loss": 0.0585,
+      "step": 13519
+    },
+    {
+      "epoch": 2.1905379131561893,
+      "grad_norm": 1.2281583547592163,
+      "learning_rate": 8.954144005919422e-07,
+      "loss": 0.0657,
+      "step": 13520
+    },
+    {
+      "epoch": 2.190699935191186,
+      "grad_norm": 0.9403387308120728,
+      "learning_rate": 8.950790813766416e-07,
+      "loss": 0.0543,
+      "step": 13521
+    },
+    {
+      "epoch": 2.1908619572261827,
+      "grad_norm": 1.0141313076019287,
+      "learning_rate": 8.947438112680387e-07,
+      "loss": 0.0712,
+      "step": 13522
+    },
+    {
+      "epoch": 2.1910239792611796,
+      "grad_norm": 0.9815289378166199,
+      "learning_rate": 8.944085902763902e-07,
+      "loss": 0.0637,
+      "step": 13523
+    },
+    {
+      "epoch": 2.191186001296176,
+      "grad_norm": 0.9215153455734253,
+      "learning_rate": 8.940734184119542e-07,
+      "loss": 0.0563,
+      "step": 13524
+    },
+    {
+      "epoch": 2.191348023331173,
+      "grad_norm": 0.9290637969970703,
+      "learning_rate": 8.937382956849847e-07,
+      "loss": 0.0601,
+      "step": 13525
+    },
+    {
+      "epoch": 2.19151004536617,
+      "grad_norm": 0.8032289743423462,
+      "learning_rate": 8.934032221057354e-07,
+      "loss": 0.0625,
+      "step": 13526
+    },
+    {
+      "epoch": 2.1916720674011665,
+      "grad_norm": 0.9063378572463989,
+      "learning_rate": 8.930681976844613e-07,
+      "loss": 0.059,
+      "step": 13527
+    },
+    {
+      "epoch": 2.1918340894361634,
+      "grad_norm": 1.034857988357544,
+      "learning_rate": 8.927332224314106e-07,
+      "loss": 0.0695,
+      "step": 13528
+    },
+    {
+      "epoch": 2.19199611147116,
+      "grad_norm": 1.0457537174224854,
+      "learning_rate": 8.92398296356834e-07,
+      "loss": 0.0632,
+      "step": 13529
+    },
+    {
+      "epoch": 2.192158133506157,
+      "grad_norm": 0.8522835969924927,
+      "learning_rate": 8.92063419470979e-07,
+      "loss": 0.059,
+      "step": 13530
+    },
+    {
+      "epoch": 2.192320155541154,
+      "grad_norm": 0.9645273685455322,
+      "learning_rate": 8.917285917840926e-07,
+      "loss": 0.0624,
+      "step": 13531
+    },
+    {
+      "epoch": 2.1924821775761503,
+      "grad_norm": 1.049136996269226,
+      "learning_rate": 8.91393813306419e-07,
+      "loss": 0.0591,
+      "step": 13532
+    },
+    {
+      "epoch": 2.192644199611147,
+      "grad_norm": 0.846191942691803,
+      "learning_rate": 8.910590840482023e-07,
+      "loss": 0.0586,
+      "step": 13533
+    },
+    {
+      "epoch": 2.1928062216461437,
+      "grad_norm": 0.8312897682189941,
+      "learning_rate": 8.907244040196836e-07,
+      "loss": 0.0588,
+      "step": 13534
+    },
+    {
+      "epoch": 2.1929682436811406,
+      "grad_norm": 0.8862128257751465,
+      "learning_rate": 8.903897732311048e-07,
+      "loss": 0.0608,
+      "step": 13535
+    },
+    {
+      "epoch": 2.1931302657161376,
+      "grad_norm": 0.8992831110954285,
+      "learning_rate": 8.900551916927022e-07,
+      "loss": 0.0632,
+      "step": 13536
+    },
+    {
+      "epoch": 2.193292287751134,
+      "grad_norm": 1.0759751796722412,
+      "learning_rate": 8.897206594147156e-07,
+      "loss": 0.0659,
+      "step": 13537
+    },
+    {
+      "epoch": 2.193454309786131,
+      "grad_norm": 1.051218867301941,
+      "learning_rate": 8.893861764073808e-07,
+      "loss": 0.0672,
+      "step": 13538
+    },
+    {
+      "epoch": 2.1936163318211275,
+      "grad_norm": 0.8192399144172668,
+      "learning_rate": 8.890517426809306e-07,
+      "loss": 0.0631,
+      "step": 13539
+    },
+    {
+      "epoch": 2.1937783538561244,
+      "grad_norm": 0.7525157332420349,
+      "learning_rate": 8.887173582455985e-07,
+      "loss": 0.0506,
+      "step": 13540
+    },
+    {
+      "epoch": 2.1939403758911213,
+      "grad_norm": 0.8876081109046936,
+      "learning_rate": 8.883830231116153e-07,
+      "loss": 0.0565,
+      "step": 13541
+    },
+    {
+      "epoch": 2.194102397926118,
+      "grad_norm": 0.7892178893089294,
+      "learning_rate": 8.88048737289213e-07,
+      "loss": 0.0581,
+      "step": 13542
+    },
+    {
+      "epoch": 2.1942644199611148,
+      "grad_norm": 0.8460841178894043,
+      "learning_rate": 8.877145007886179e-07,
+      "loss": 0.0665,
+      "step": 13543
+    },
+    {
+      "epoch": 2.1944264419961117,
+      "grad_norm": 1.0782990455627441,
+      "learning_rate": 8.873803136200574e-07,
+      "loss": 0.0677,
+      "step": 13544
+    },
+    {
+      "epoch": 2.194588464031108,
+      "grad_norm": 0.9548028707504272,
+      "learning_rate": 8.870461757937568e-07,
+      "loss": 0.0645,
+      "step": 13545
+    },
+    {
+      "epoch": 2.194750486066105,
+      "grad_norm": 0.8092406392097473,
+      "learning_rate": 8.8671208731994e-07,
+      "loss": 0.0627,
+      "step": 13546
+    },
+    {
+      "epoch": 2.1949125081011016,
+      "grad_norm": 0.8001680970191956,
+      "learning_rate": 8.863780482088291e-07,
+      "loss": 0.0547,
+      "step": 13547
+    },
+    {
+      "epoch": 2.1950745301360985,
+      "grad_norm": 0.9187164306640625,
+      "learning_rate": 8.860440584706451e-07,
+      "loss": 0.0666,
+      "step": 13548
+    },
+    {
+      "epoch": 2.195236552171095,
+      "grad_norm": 0.9525156617164612,
+      "learning_rate": 8.857101181156072e-07,
+      "loss": 0.0621,
+      "step": 13549
+    },
+    {
+      "epoch": 2.195398574206092,
+      "grad_norm": 0.9552687406539917,
+      "learning_rate": 8.853762271539332e-07,
+      "loss": 0.0632,
+      "step": 13550
+    },
+    {
+      "epoch": 2.195560596241089,
+      "grad_norm": 0.9133409857749939,
+      "learning_rate": 8.850423855958393e-07,
+      "loss": 0.0629,
+      "step": 13551
+    },
+    {
+      "epoch": 2.1957226182760854,
+      "grad_norm": 0.7785288691520691,
+      "learning_rate": 8.847085934515404e-07,
+      "loss": 0.0539,
+      "step": 13552
+    },
+    {
+      "epoch": 2.1958846403110823,
+      "grad_norm": 0.856397271156311,
+      "learning_rate": 8.843748507312505e-07,
+      "loss": 0.061,
+      "step": 13553
+    },
+    {
+      "epoch": 2.1960466623460793,
+      "grad_norm": 0.9479225873947144,
+      "learning_rate": 8.840411574451793e-07,
+      "loss": 0.0606,
+      "step": 13554
+    },
+    {
+      "epoch": 2.1962086843810757,
+      "grad_norm": 0.817360520362854,
+      "learning_rate": 8.837075136035375e-07,
+      "loss": 0.0542,
+      "step": 13555
+    },
+    {
+      "epoch": 2.1963707064160727,
+      "grad_norm": 0.8960983157157898,
+      "learning_rate": 8.833739192165352e-07,
+      "loss": 0.0653,
+      "step": 13556
+    },
+    {
+      "epoch": 2.196532728451069,
+      "grad_norm": 0.8757344484329224,
+      "learning_rate": 8.830403742943797e-07,
+      "loss": 0.0521,
+      "step": 13557
+    },
+    {
+      "epoch": 2.196694750486066,
+      "grad_norm": 0.868101954460144,
+      "learning_rate": 8.827068788472751e-07,
+      "loss": 0.0589,
+      "step": 13558
+    },
+    {
+      "epoch": 2.196856772521063,
+      "grad_norm": 0.8532265424728394,
+      "learning_rate": 8.823734328854259e-07,
+      "loss": 0.0614,
+      "step": 13559
+    },
+    {
+      "epoch": 2.1970187945560595,
+      "grad_norm": 0.8555817008018494,
+      "learning_rate": 8.820400364190351e-07,
+      "loss": 0.0593,
+      "step": 13560
+    },
+    {
+      "epoch": 2.1971808165910565,
+      "grad_norm": 0.8289614915847778,
+      "learning_rate": 8.81706689458304e-07,
+      "loss": 0.0605,
+      "step": 13561
+    },
+    {
+      "epoch": 2.197342838626053,
+      "grad_norm": 0.767209529876709,
+      "learning_rate": 8.813733920134321e-07,
+      "loss": 0.0496,
+      "step": 13562
+    },
+    {
+      "epoch": 2.19750486066105,
+      "grad_norm": 0.9439572691917419,
+      "learning_rate": 8.81040144094617e-07,
+      "loss": 0.0634,
+      "step": 13563
+    },
+    {
+      "epoch": 2.197666882696047,
+      "grad_norm": 0.8966138958930969,
+      "learning_rate": 8.807069457120571e-07,
+      "loss": 0.0612,
+      "step": 13564
+    },
+    {
+      "epoch": 2.1978289047310433,
+      "grad_norm": 1.0695065259933472,
+      "learning_rate": 8.803737968759438e-07,
+      "loss": 0.0759,
+      "step": 13565
+    },
+    {
+      "epoch": 2.1979909267660402,
+      "grad_norm": 0.8874850273132324,
+      "learning_rate": 8.80040697596474e-07,
+      "loss": 0.0636,
+      "step": 13566
+    },
+    {
+      "epoch": 2.198152948801037,
+      "grad_norm": 0.8661726713180542,
+      "learning_rate": 8.797076478838388e-07,
+      "loss": 0.0588,
+      "step": 13567
+    },
+    {
+      "epoch": 2.1983149708360337,
+      "grad_norm": 0.8928310871124268,
+      "learning_rate": 8.79374647748229e-07,
+      "loss": 0.0601,
+      "step": 13568
+    },
+    {
+      "epoch": 2.1984769928710306,
+      "grad_norm": 0.8463062644004822,
+      "learning_rate": 8.790416971998317e-07,
+      "loss": 0.059,
+      "step": 13569
+    },
+    {
+      "epoch": 2.198639014906027,
+      "grad_norm": 0.9622822999954224,
+      "learning_rate": 8.787087962488367e-07,
+      "loss": 0.0595,
+      "step": 13570
+    },
+    {
+      "epoch": 2.198801036941024,
+      "grad_norm": 0.9242643117904663,
+      "learning_rate": 8.783759449054296e-07,
+      "loss": 0.0645,
+      "step": 13571
+    },
+    {
+      "epoch": 2.1989630589760205,
+      "grad_norm": 0.957097589969635,
+      "learning_rate": 8.780431431797937e-07,
+      "loss": 0.0685,
+      "step": 13572
+    },
+    {
+      "epoch": 2.1991250810110174,
+      "grad_norm": 0.9083278775215149,
+      "learning_rate": 8.777103910821127e-07,
+      "loss": 0.0633,
+      "step": 13573
+    },
+    {
+      "epoch": 2.1992871030460144,
+      "grad_norm": 0.9371110796928406,
+      "learning_rate": 8.773776886225668e-07,
+      "loss": 0.0639,
+      "step": 13574
+    },
+    {
+      "epoch": 2.199449125081011,
+      "grad_norm": 0.9061893820762634,
+      "learning_rate": 8.770450358113389e-07,
+      "loss": 0.0606,
+      "step": 13575
+    },
+    {
+      "epoch": 2.199611147116008,
+      "grad_norm": 0.9202576875686646,
+      "learning_rate": 8.767124326586043e-07,
+      "loss": 0.0632,
+      "step": 13576
+    },
+    {
+      "epoch": 2.1997731691510047,
+      "grad_norm": 0.9155436754226685,
+      "learning_rate": 8.763798791745413e-07,
+      "loss": 0.0637,
+      "step": 13577
+    },
+    {
+      "epoch": 2.1999351911860012,
+      "grad_norm": 1.0268107652664185,
+      "learning_rate": 8.760473753693243e-07,
+      "loss": 0.0673,
+      "step": 13578
+    },
+    {
+      "epoch": 2.200097213220998,
+      "grad_norm": 1.1556326150894165,
+      "learning_rate": 8.757149212531282e-07,
+      "loss": 0.0645,
+      "step": 13579
+    },
+    {
+      "epoch": 2.2002592352559946,
+      "grad_norm": 0.7752742171287537,
+      "learning_rate": 8.753825168361249e-07,
+      "loss": 0.0561,
+      "step": 13580
+    },
+    {
+      "epoch": 2.2004212572909916,
+      "grad_norm": 0.9289119839668274,
+      "learning_rate": 8.750501621284849e-07,
+      "loss": 0.0638,
+      "step": 13581
+    },
+    {
+      "epoch": 2.2005832793259885,
+      "grad_norm": 0.8528614640235901,
+      "learning_rate": 8.747178571403786e-07,
+      "loss": 0.0601,
+      "step": 13582
+    },
+    {
+      "epoch": 2.200745301360985,
+      "grad_norm": 0.8271398544311523,
+      "learning_rate": 8.743856018819719e-07,
+      "loss": 0.0596,
+      "step": 13583
+    },
+    {
+      "epoch": 2.200907323395982,
+      "grad_norm": 0.8116289377212524,
+      "learning_rate": 8.74053396363431e-07,
+      "loss": 0.0615,
+      "step": 13584
+    },
+    {
+      "epoch": 2.2010693454309784,
+      "grad_norm": 0.7971740961074829,
+      "learning_rate": 8.737212405949222e-07,
+      "loss": 0.0615,
+      "step": 13585
+    },
+    {
+      "epoch": 2.2012313674659754,
+      "grad_norm": 1.0324573516845703,
+      "learning_rate": 8.733891345866088e-07,
+      "loss": 0.0666,
+      "step": 13586
+    },
+    {
+      "epoch": 2.2013933895009723,
+      "grad_norm": 0.8397610783576965,
+      "learning_rate": 8.730570783486508e-07,
+      "loss": 0.0664,
+      "step": 13587
+    },
+    {
+      "epoch": 2.201555411535969,
+      "grad_norm": 0.9985054135322571,
+      "learning_rate": 8.727250718912089e-07,
+      "loss": 0.0602,
+      "step": 13588
+    },
+    {
+      "epoch": 2.2017174335709657,
+      "grad_norm": 1.0729403495788574,
+      "learning_rate": 8.723931152244421e-07,
+      "loss": 0.0624,
+      "step": 13589
+    },
+    {
+      "epoch": 2.201879455605962,
+      "grad_norm": 1.0045422315597534,
+      "learning_rate": 8.72061208358507e-07,
+      "loss": 0.0627,
+      "step": 13590
+    },
+    {
+      "epoch": 2.202041477640959,
+      "grad_norm": 0.8267765045166016,
+      "learning_rate": 8.717293513035596e-07,
+      "loss": 0.0644,
+      "step": 13591
+    },
+    {
+      "epoch": 2.202203499675956,
+      "grad_norm": 0.8496410250663757,
+      "learning_rate": 8.713975440697536e-07,
+      "loss": 0.0568,
+      "step": 13592
+    },
+    {
+      "epoch": 2.2023655217109526,
+      "grad_norm": 0.9486085772514343,
+      "learning_rate": 8.710657866672417e-07,
+      "loss": 0.0668,
+      "step": 13593
+    },
+    {
+      "epoch": 2.2025275437459495,
+      "grad_norm": 0.867942750453949,
+      "learning_rate": 8.707340791061747e-07,
+      "loss": 0.057,
+      "step": 13594
+    },
+    {
+      "epoch": 2.202689565780946,
+      "grad_norm": 0.8752855062484741,
+      "learning_rate": 8.704024213967021e-07,
+      "loss": 0.0651,
+      "step": 13595
+    },
+    {
+      "epoch": 2.202851587815943,
+      "grad_norm": 1.1905330419540405,
+      "learning_rate": 8.700708135489722e-07,
+      "loss": 0.0653,
+      "step": 13596
+    },
+    {
+      "epoch": 2.20301360985094,
+      "grad_norm": 0.9045699834823608,
+      "learning_rate": 8.697392555731315e-07,
+      "loss": 0.0604,
+      "step": 13597
+    },
+    {
+      "epoch": 2.2031756318859363,
+      "grad_norm": 0.8511764407157898,
+      "learning_rate": 8.694077474793227e-07,
+      "loss": 0.0662,
+      "step": 13598
+    },
+    {
+      "epoch": 2.2033376539209333,
+      "grad_norm": 0.8251369595527649,
+      "learning_rate": 8.690762892776918e-07,
+      "loss": 0.0641,
+      "step": 13599
+    },
+    {
+      "epoch": 2.20349967595593,
+      "grad_norm": 0.9330815076828003,
+      "learning_rate": 8.687448809783799e-07,
+      "loss": 0.0646,
+      "step": 13600
+    },
+    {
+      "epoch": 2.2036616979909267,
+      "grad_norm": 0.8293221592903137,
+      "learning_rate": 8.684135225915277e-07,
+      "loss": 0.0595,
+      "step": 13601
+    },
+    {
+      "epoch": 2.2038237200259236,
+      "grad_norm": 1.0434253215789795,
+      "learning_rate": 8.680822141272727e-07,
+      "loss": 0.0698,
+      "step": 13602
+    },
+    {
+      "epoch": 2.20398574206092,
+      "grad_norm": 0.8310006260871887,
+      "learning_rate": 8.677509555957517e-07,
+      "loss": 0.0568,
+      "step": 13603
+    },
+    {
+      "epoch": 2.204147764095917,
+      "grad_norm": 0.8653905391693115,
+      "learning_rate": 8.674197470071033e-07,
+      "loss": 0.0583,
+      "step": 13604
+    },
+    {
+      "epoch": 2.204309786130914,
+      "grad_norm": 0.8970719575881958,
+      "learning_rate": 8.670885883714591e-07,
+      "loss": 0.0587,
+      "step": 13605
+    },
+    {
+      "epoch": 2.2044718081659105,
+      "grad_norm": 1.2163879871368408,
+      "learning_rate": 8.667574796989526e-07,
+      "loss": 0.0714,
+      "step": 13606
+    },
+    {
+      "epoch": 2.2046338302009074,
+      "grad_norm": 0.865898609161377,
+      "learning_rate": 8.664264209997144e-07,
+      "loss": 0.0627,
+      "step": 13607
+    },
+    {
+      "epoch": 2.204795852235904,
+      "grad_norm": 0.900995671749115,
+      "learning_rate": 8.66095412283875e-07,
+      "loss": 0.0608,
+      "step": 13608
+    },
+    {
+      "epoch": 2.204957874270901,
+      "grad_norm": 0.8361657857894897,
+      "learning_rate": 8.657644535615617e-07,
+      "loss": 0.0561,
+      "step": 13609
+    },
+    {
+      "epoch": 2.2051198963058978,
+      "grad_norm": 0.9006054401397705,
+      "learning_rate": 8.654335448429016e-07,
+      "loss": 0.0639,
+      "step": 13610
+    },
+    {
+      "epoch": 2.2052819183408943,
+      "grad_norm": 0.8334879875183105,
+      "learning_rate": 8.651026861380193e-07,
+      "loss": 0.0558,
+      "step": 13611
+    },
+    {
+      "epoch": 2.205443940375891,
+      "grad_norm": 0.9533306360244751,
+      "learning_rate": 8.647718774570385e-07,
+      "loss": 0.0637,
+      "step": 13612
+    },
+    {
+      "epoch": 2.2056059624108877,
+      "grad_norm": 0.8768250346183777,
+      "learning_rate": 8.644411188100812e-07,
+      "loss": 0.0609,
+      "step": 13613
+    },
+    {
+      "epoch": 2.2057679844458846,
+      "grad_norm": 0.8379773497581482,
+      "learning_rate": 8.641104102072676e-07,
+      "loss": 0.0561,
+      "step": 13614
+    },
+    {
+      "epoch": 2.2059300064808816,
+      "grad_norm": 1.2272884845733643,
+      "learning_rate": 8.637797516587173e-07,
+      "loss": 0.0692,
+      "step": 13615
+    },
+    {
+      "epoch": 2.206092028515878,
+      "grad_norm": 0.8420472145080566,
+      "learning_rate": 8.634491431745465e-07,
+      "loss": 0.057,
+      "step": 13616
+    },
+    {
+      "epoch": 2.206254050550875,
+      "grad_norm": 0.835096001625061,
+      "learning_rate": 8.631185847648704e-07,
+      "loss": 0.0571,
+      "step": 13617
+    },
+    {
+      "epoch": 2.2064160725858715,
+      "grad_norm": 0.9500763416290283,
+      "learning_rate": 8.627880764398055e-07,
+      "loss": 0.0538,
+      "step": 13618
+    },
+    {
+      "epoch": 2.2065780946208684,
+      "grad_norm": 0.8583593368530273,
+      "learning_rate": 8.62457618209464e-07,
+      "loss": 0.0588,
+      "step": 13619
+    },
+    {
+      "epoch": 2.2067401166558653,
+      "grad_norm": 0.816092848777771,
+      "learning_rate": 8.621272100839562e-07,
+      "loss": 0.056,
+      "step": 13620
+    },
+    {
+      "epoch": 2.206902138690862,
+      "grad_norm": 0.8541872501373291,
+      "learning_rate": 8.617968520733919e-07,
+      "loss": 0.0554,
+      "step": 13621
+    },
+    {
+      "epoch": 2.2070641607258588,
+      "grad_norm": 0.8463016152381897,
+      "learning_rate": 8.614665441878798e-07,
+      "loss": 0.0608,
+      "step": 13622
+    },
+    {
+      "epoch": 2.2072261827608557,
+      "grad_norm": 0.942939281463623,
+      "learning_rate": 8.611362864375261e-07,
+      "loss": 0.0682,
+      "step": 13623
+    },
+    {
+      "epoch": 2.207388204795852,
+      "grad_norm": 1.0515944957733154,
+      "learning_rate": 8.60806078832436e-07,
+      "loss": 0.0771,
+      "step": 13624
+    },
+    {
+      "epoch": 2.207550226830849,
+      "grad_norm": 0.9806361794471741,
+      "learning_rate": 8.604759213827133e-07,
+      "loss": 0.0698,
+      "step": 13625
+    },
+    {
+      "epoch": 2.2077122488658456,
+      "grad_norm": 0.9298086762428284,
+      "learning_rate": 8.601458140984606e-07,
+      "loss": 0.0579,
+      "step": 13626
+    },
+    {
+      "epoch": 2.2078742709008425,
+      "grad_norm": 0.8107653856277466,
+      "learning_rate": 8.598157569897758e-07,
+      "loss": 0.0553,
+      "step": 13627
+    },
+    {
+      "epoch": 2.2080362929358395,
+      "grad_norm": 0.874210774898529,
+      "learning_rate": 8.594857500667606e-07,
+      "loss": 0.0583,
+      "step": 13628
+    },
+    {
+      "epoch": 2.208198314970836,
+      "grad_norm": 0.8932999968528748,
+      "learning_rate": 8.591557933395115e-07,
+      "loss": 0.0645,
+      "step": 13629
+    },
+    {
+      "epoch": 2.208360337005833,
+      "grad_norm": 0.8853711485862732,
+      "learning_rate": 8.588258868181251e-07,
+      "loss": 0.0611,
+      "step": 13630
+    },
+    {
+      "epoch": 2.2085223590408294,
+      "grad_norm": 0.9822854995727539,
+      "learning_rate": 8.584960305126943e-07,
+      "loss": 0.0742,
+      "step": 13631
+    },
+    {
+      "epoch": 2.2086843810758263,
+      "grad_norm": 0.930212676525116,
+      "learning_rate": 8.581662244333116e-07,
+      "loss": 0.0657,
+      "step": 13632
+    },
+    {
+      "epoch": 2.2088464031108233,
+      "grad_norm": 0.8969360589981079,
+      "learning_rate": 8.578364685900711e-07,
+      "loss": 0.0596,
+      "step": 13633
+    },
+    {
+      "epoch": 2.2090084251458197,
+      "grad_norm": 0.8433337211608887,
+      "learning_rate": 8.575067629930601e-07,
+      "loss": 0.0564,
+      "step": 13634
+    },
+    {
+      "epoch": 2.2091704471808167,
+      "grad_norm": 0.8767102360725403,
+      "learning_rate": 8.571771076523669e-07,
+      "loss": 0.0608,
+      "step": 13635
+    },
+    {
+      "epoch": 2.209332469215813,
+      "grad_norm": 0.8478491306304932,
+      "learning_rate": 8.568475025780781e-07,
+      "loss": 0.061,
+      "step": 13636
+    },
+    {
+      "epoch": 2.20949449125081,
+      "grad_norm": 1.1657311916351318,
+      "learning_rate": 8.56517947780281e-07,
+      "loss": 0.0705,
+      "step": 13637
+    },
+    {
+      "epoch": 2.209656513285807,
+      "grad_norm": 0.980461835861206,
+      "learning_rate": 8.561884432690568e-07,
+      "loss": 0.0583,
+      "step": 13638
+    },
+    {
+      "epoch": 2.2098185353208035,
+      "grad_norm": 0.9112688302993774,
+      "learning_rate": 8.55858989054488e-07,
+      "loss": 0.0618,
+      "step": 13639
+    },
+    {
+      "epoch": 2.2099805573558005,
+      "grad_norm": 1.0051236152648926,
+      "learning_rate": 8.555295851466556e-07,
+      "loss": 0.066,
+      "step": 13640
+    },
+    {
+      "epoch": 2.210142579390797,
+      "grad_norm": 0.8563796281814575,
+      "learning_rate": 8.552002315556382e-07,
+      "loss": 0.0613,
+      "step": 13641
+    },
+    {
+      "epoch": 2.210304601425794,
+      "grad_norm": 0.7843127846717834,
+      "learning_rate": 8.548709282915135e-07,
+      "loss": 0.0638,
+      "step": 13642
+    },
+    {
+      "epoch": 2.210466623460791,
+      "grad_norm": 0.9892216324806213,
+      "learning_rate": 8.545416753643574e-07,
+      "loss": 0.0652,
+      "step": 13643
+    },
+    {
+      "epoch": 2.2106286454957873,
+      "grad_norm": 0.9359889030456543,
+      "learning_rate": 8.542124727842438e-07,
+      "loss": 0.0596,
+      "step": 13644
+    },
+    {
+      "epoch": 2.2107906675307842,
+      "grad_norm": 0.7579911947250366,
+      "learning_rate": 8.538833205612468e-07,
+      "loss": 0.0536,
+      "step": 13645
+    },
+    {
+      "epoch": 2.210952689565781,
+      "grad_norm": 0.7758036255836487,
+      "learning_rate": 8.535542187054352e-07,
+      "loss": 0.059,
+      "step": 13646
+    },
+    {
+      "epoch": 2.2111147116007777,
+      "grad_norm": 0.8697202801704407,
+      "learning_rate": 8.532251672268807e-07,
+      "loss": 0.0619,
+      "step": 13647
+    },
+    {
+      "epoch": 2.2112767336357746,
+      "grad_norm": 0.9321410059928894,
+      "learning_rate": 8.528961661356519e-07,
+      "loss": 0.0647,
+      "step": 13648
+    },
+    {
+      "epoch": 2.211438755670771,
+      "grad_norm": 1.245924711227417,
+      "learning_rate": 8.525672154418138e-07,
+      "loss": 0.0676,
+      "step": 13649
+    },
+    {
+      "epoch": 2.211600777705768,
+      "grad_norm": 0.8289289474487305,
+      "learning_rate": 8.52238315155432e-07,
+      "loss": 0.0551,
+      "step": 13650
+    },
+    {
+      "epoch": 2.211762799740765,
+      "grad_norm": 0.8378967642784119,
+      "learning_rate": 8.519094652865703e-07,
+      "loss": 0.0534,
+      "step": 13651
+    },
+    {
+      "epoch": 2.2119248217757614,
+      "grad_norm": 0.9754065871238708,
+      "learning_rate": 8.515806658452908e-07,
+      "loss": 0.0669,
+      "step": 13652
+    },
+    {
+      "epoch": 2.2120868438107584,
+      "grad_norm": 0.8882692456245422,
+      "learning_rate": 8.512519168416536e-07,
+      "loss": 0.0619,
+      "step": 13653
+    },
+    {
+      "epoch": 2.212248865845755,
+      "grad_norm": 0.9585002660751343,
+      "learning_rate": 8.50923218285718e-07,
+      "loss": 0.0655,
+      "step": 13654
+    },
+    {
+      "epoch": 2.212410887880752,
+      "grad_norm": 0.9813425540924072,
+      "learning_rate": 8.505945701875412e-07,
+      "loss": 0.0657,
+      "step": 13655
+    },
+    {
+      "epoch": 2.2125729099157487,
+      "grad_norm": 0.9413607716560364,
+      "learning_rate": 8.502659725571791e-07,
+      "loss": 0.0606,
+      "step": 13656
+    },
+    {
+      "epoch": 2.212734931950745,
+      "grad_norm": 0.8411443829536438,
+      "learning_rate": 8.499374254046858e-07,
+      "loss": 0.0617,
+      "step": 13657
+    },
+    {
+      "epoch": 2.212896953985742,
+      "grad_norm": 0.9197219610214233,
+      "learning_rate": 8.496089287401144e-07,
+      "loss": 0.0619,
+      "step": 13658
+    },
+    {
+      "epoch": 2.2130589760207386,
+      "grad_norm": 1.0152454376220703,
+      "learning_rate": 8.492804825735166e-07,
+      "loss": 0.0579,
+      "step": 13659
+    },
+    {
+      "epoch": 2.2132209980557356,
+      "grad_norm": 0.9079082012176514,
+      "learning_rate": 8.489520869149398e-07,
+      "loss": 0.0711,
+      "step": 13660
+    },
+    {
+      "epoch": 2.2133830200907325,
+      "grad_norm": 0.9641551375389099,
+      "learning_rate": 8.486237417744344e-07,
+      "loss": 0.0587,
+      "step": 13661
+    },
+    {
+      "epoch": 2.213545042125729,
+      "grad_norm": 0.8907647132873535,
+      "learning_rate": 8.482954471620464e-07,
+      "loss": 0.0601,
+      "step": 13662
+    },
+    {
+      "epoch": 2.213707064160726,
+      "grad_norm": 0.7322314381599426,
+      "learning_rate": 8.479672030878213e-07,
+      "loss": 0.054,
+      "step": 13663
+    },
+    {
+      "epoch": 2.2138690861957224,
+      "grad_norm": 0.8832927346229553,
+      "learning_rate": 8.476390095618015e-07,
+      "loss": 0.0571,
+      "step": 13664
+    },
+    {
+      "epoch": 2.2140311082307194,
+      "grad_norm": 0.8694517016410828,
+      "learning_rate": 8.47310866594028e-07,
+      "loss": 0.06,
+      "step": 13665
+    },
+    {
+      "epoch": 2.2141931302657163,
+      "grad_norm": 0.8959830403327942,
+      "learning_rate": 8.469827741945447e-07,
+      "loss": 0.0642,
+      "step": 13666
+    },
+    {
+      "epoch": 2.214355152300713,
+      "grad_norm": 0.9911413192749023,
+      "learning_rate": 8.466547323733873e-07,
+      "loss": 0.0602,
+      "step": 13667
+    },
+    {
+      "epoch": 2.2145171743357097,
+      "grad_norm": 0.8468191027641296,
+      "learning_rate": 8.46326741140594e-07,
+      "loss": 0.0601,
+      "step": 13668
+    },
+    {
+      "epoch": 2.2146791963707066,
+      "grad_norm": 1.0290790796279907,
+      "learning_rate": 8.459988005062006e-07,
+      "loss": 0.06,
+      "step": 13669
+    },
+    {
+      "epoch": 2.214841218405703,
+      "grad_norm": 0.9934695363044739,
+      "learning_rate": 8.456709104802413e-07,
+      "loss": 0.0616,
+      "step": 13670
+    },
+    {
+      "epoch": 2.2150032404407,
+      "grad_norm": 0.8425288200378418,
+      "learning_rate": 8.453430710727486e-07,
+      "loss": 0.0595,
+      "step": 13671
+    },
+    {
+      "epoch": 2.2151652624756966,
+      "grad_norm": 0.862170398235321,
+      "learning_rate": 8.450152822937541e-07,
+      "loss": 0.0619,
+      "step": 13672
+    },
+    {
+      "epoch": 2.2153272845106935,
+      "grad_norm": 0.8608142137527466,
+      "learning_rate": 8.446875441532868e-07,
+      "loss": 0.0645,
+      "step": 13673
+    },
+    {
+      "epoch": 2.21548930654569,
+      "grad_norm": 0.8314158916473389,
+      "learning_rate": 8.443598566613756e-07,
+      "loss": 0.058,
+      "step": 13674
+    },
+    {
+      "epoch": 2.215651328580687,
+      "grad_norm": 0.9272691607475281,
+      "learning_rate": 8.440322198280446e-07,
+      "loss": 0.0604,
+      "step": 13675
+    },
+    {
+      "epoch": 2.215813350615684,
+      "grad_norm": 1.0183619260787964,
+      "learning_rate": 8.437046336633212e-07,
+      "loss": 0.0642,
+      "step": 13676
+    },
+    {
+      "epoch": 2.2159753726506803,
+      "grad_norm": 0.8743815422058105,
+      "learning_rate": 8.433770981772285e-07,
+      "loss": 0.0625,
+      "step": 13677
+    },
+    {
+      "epoch": 2.2161373946856773,
+      "grad_norm": 0.8977541327476501,
+      "learning_rate": 8.430496133797872e-07,
+      "loss": 0.067,
+      "step": 13678
+    },
+    {
+      "epoch": 2.216299416720674,
+      "grad_norm": 0.9570757150650024,
+      "learning_rate": 8.427221792810169e-07,
+      "loss": 0.0597,
+      "step": 13679
+    },
+    {
+      "epoch": 2.2164614387556707,
+      "grad_norm": 0.8252487778663635,
+      "learning_rate": 8.423947958909381e-07,
+      "loss": 0.0618,
+      "step": 13680
+    },
+    {
+      "epoch": 2.2166234607906676,
+      "grad_norm": 0.801275372505188,
+      "learning_rate": 8.420674632195683e-07,
+      "loss": 0.0581,
+      "step": 13681
+    },
+    {
+      "epoch": 2.216785482825664,
+      "grad_norm": 0.8917983770370483,
+      "learning_rate": 8.41740181276921e-07,
+      "loss": 0.0685,
+      "step": 13682
+    },
+    {
+      "epoch": 2.216947504860661,
+      "grad_norm": 0.9908973574638367,
+      "learning_rate": 8.414129500730115e-07,
+      "loss": 0.0666,
+      "step": 13683
+    },
+    {
+      "epoch": 2.217109526895658,
+      "grad_norm": 1.1001694202423096,
+      "learning_rate": 8.410857696178518e-07,
+      "loss": 0.0664,
+      "step": 13684
+    },
+    {
+      "epoch": 2.2172715489306545,
+      "grad_norm": 0.8610407710075378,
+      "learning_rate": 8.407586399214529e-07,
+      "loss": 0.0634,
+      "step": 13685
+    },
+    {
+      "epoch": 2.2174335709656514,
+      "grad_norm": 0.8666363954544067,
+      "learning_rate": 8.404315609938246e-07,
+      "loss": 0.0623,
+      "step": 13686
+    },
+    {
+      "epoch": 2.217595593000648,
+      "grad_norm": 0.8178418278694153,
+      "learning_rate": 8.401045328449742e-07,
+      "loss": 0.0607,
+      "step": 13687
+    },
+    {
+      "epoch": 2.217757615035645,
+      "grad_norm": 0.8369463086128235,
+      "learning_rate": 8.397775554849086e-07,
+      "loss": 0.0577,
+      "step": 13688
+    },
+    {
+      "epoch": 2.2179196370706418,
+      "grad_norm": 1.0143942832946777,
+      "learning_rate": 8.394506289236317e-07,
+      "loss": 0.0673,
+      "step": 13689
+    },
+    {
+      "epoch": 2.2180816591056383,
+      "grad_norm": 1.006434679031372,
+      "learning_rate": 8.391237531711474e-07,
+      "loss": 0.0585,
+      "step": 13690
+    },
+    {
+      "epoch": 2.218243681140635,
+      "grad_norm": 0.9059908390045166,
+      "learning_rate": 8.38796928237457e-07,
+      "loss": 0.064,
+      "step": 13691
+    },
+    {
+      "epoch": 2.2184057031756317,
+      "grad_norm": 0.9959692358970642,
+      "learning_rate": 8.384701541325612e-07,
+      "loss": 0.066,
+      "step": 13692
+    },
+    {
+      "epoch": 2.2185677252106286,
+      "grad_norm": 0.9817578792572021,
+      "learning_rate": 8.381434308664574e-07,
+      "loss": 0.0707,
+      "step": 13693
+    },
+    {
+      "epoch": 2.2187297472456255,
+      "grad_norm": 0.9044135808944702,
+      "learning_rate": 8.378167584491417e-07,
+      "loss": 0.0646,
+      "step": 13694
+    },
+    {
+      "epoch": 2.218891769280622,
+      "grad_norm": 0.751171886920929,
+      "learning_rate": 8.374901368906127e-07,
+      "loss": 0.0566,
+      "step": 13695
+    },
+    {
+      "epoch": 2.219053791315619,
+      "grad_norm": 0.8422271609306335,
+      "learning_rate": 8.371635662008615e-07,
+      "loss": 0.0642,
+      "step": 13696
+    },
+    {
+      "epoch": 2.2192158133506155,
+      "grad_norm": 0.8549221754074097,
+      "learning_rate": 8.368370463898812e-07,
+      "loss": 0.0518,
+      "step": 13697
+    },
+    {
+      "epoch": 2.2193778353856124,
+      "grad_norm": 0.8980638384819031,
+      "learning_rate": 8.365105774676624e-07,
+      "loss": 0.0633,
+      "step": 13698
+    },
+    {
+      "epoch": 2.2195398574206093,
+      "grad_norm": 0.9105513691902161,
+      "learning_rate": 8.361841594441944e-07,
+      "loss": 0.06,
+      "step": 13699
+    },
+    {
+      "epoch": 2.219701879455606,
+      "grad_norm": 0.8758683204650879,
+      "learning_rate": 8.358577923294647e-07,
+      "loss": 0.0646,
+      "step": 13700
+    },
+    {
+      "epoch": 2.2198639014906028,
+      "grad_norm": 1.3330185413360596,
+      "learning_rate": 8.355314761334596e-07,
+      "loss": 0.0658,
+      "step": 13701
+    },
+    {
+      "epoch": 2.2200259235255997,
+      "grad_norm": 0.9840524792671204,
+      "learning_rate": 8.352052108661634e-07,
+      "loss": 0.0674,
+      "step": 13702
+    },
+    {
+      "epoch": 2.220187945560596,
+      "grad_norm": 0.8081815242767334,
+      "learning_rate": 8.34878996537559e-07,
+      "loss": 0.06,
+      "step": 13703
+    },
+    {
+      "epoch": 2.220349967595593,
+      "grad_norm": 0.8371722102165222,
+      "learning_rate": 8.345528331576275e-07,
+      "loss": 0.0601,
+      "step": 13704
+    },
+    {
+      "epoch": 2.2205119896305896,
+      "grad_norm": 1.0199873447418213,
+      "learning_rate": 8.342267207363492e-07,
+      "loss": 0.0612,
+      "step": 13705
+    },
+    {
+      "epoch": 2.2206740116655865,
+      "grad_norm": 1.0433920621871948,
+      "learning_rate": 8.339006592837021e-07,
+      "loss": 0.0671,
+      "step": 13706
+    },
+    {
+      "epoch": 2.2208360337005835,
+      "grad_norm": 0.8671245574951172,
+      "learning_rate": 8.335746488096639e-07,
+      "loss": 0.0677,
+      "step": 13707
+    },
+    {
+      "epoch": 2.22099805573558,
+      "grad_norm": 0.9357213377952576,
+      "learning_rate": 8.33248689324207e-07,
+      "loss": 0.0716,
+      "step": 13708
+    },
+    {
+      "epoch": 2.221160077770577,
+      "grad_norm": 0.8133964538574219,
+      "learning_rate": 8.329227808373078e-07,
+      "loss": 0.0578,
+      "step": 13709
+    },
+    {
+      "epoch": 2.2213220998055734,
+      "grad_norm": 0.8895506858825684,
+      "learning_rate": 8.325969233589376e-07,
+      "loss": 0.0653,
+      "step": 13710
+    },
+    {
+      "epoch": 2.2214841218405703,
+      "grad_norm": 0.8264877200126648,
+      "learning_rate": 8.322711168990661e-07,
+      "loss": 0.057,
+      "step": 13711
+    },
+    {
+      "epoch": 2.2216461438755672,
+      "grad_norm": 0.8504375219345093,
+      "learning_rate": 8.319453614676626e-07,
+      "loss": 0.0644,
+      "step": 13712
+    },
+    {
+      "epoch": 2.2218081659105637,
+      "grad_norm": 0.824181318283081,
+      "learning_rate": 8.316196570746934e-07,
+      "loss": 0.0557,
+      "step": 13713
+    },
+    {
+      "epoch": 2.2219701879455607,
+      "grad_norm": 0.832217812538147,
+      "learning_rate": 8.31294003730127e-07,
+      "loss": 0.0613,
+      "step": 13714
+    },
+    {
+      "epoch": 2.222132209980557,
+      "grad_norm": 0.8441714644432068,
+      "learning_rate": 8.309684014439251e-07,
+      "loss": 0.0639,
+      "step": 13715
+    },
+    {
+      "epoch": 2.222294232015554,
+      "grad_norm": 0.998146653175354,
+      "learning_rate": 8.306428502260511e-07,
+      "loss": 0.0665,
+      "step": 13716
+    },
+    {
+      "epoch": 2.222456254050551,
+      "grad_norm": 0.8211577534675598,
+      "learning_rate": 8.303173500864661e-07,
+      "loss": 0.0599,
+      "step": 13717
+    },
+    {
+      "epoch": 2.2226182760855475,
+      "grad_norm": 0.9386523962020874,
+      "learning_rate": 8.299919010351296e-07,
+      "loss": 0.0625,
+      "step": 13718
+    },
+    {
+      "epoch": 2.2227802981205445,
+      "grad_norm": 0.9545656442642212,
+      "learning_rate": 8.296665030819998e-07,
+      "loss": 0.0648,
+      "step": 13719
+    },
+    {
+      "epoch": 2.222942320155541,
+      "grad_norm": 0.8156535029411316,
+      "learning_rate": 8.293411562370327e-07,
+      "loss": 0.0536,
+      "step": 13720
+    },
+    {
+      "epoch": 2.223104342190538,
+      "grad_norm": 0.9009618163108826,
+      "learning_rate": 8.290158605101842e-07,
+      "loss": 0.0621,
+      "step": 13721
+    },
+    {
+      "epoch": 2.223266364225535,
+      "grad_norm": 1.074135184288025,
+      "learning_rate": 8.286906159114058e-07,
+      "loss": 0.061,
+      "step": 13722
+    },
+    {
+      "epoch": 2.2234283862605313,
+      "grad_norm": 0.9844740033149719,
+      "learning_rate": 8.283654224506491e-07,
+      "loss": 0.0656,
+      "step": 13723
+    },
+    {
+      "epoch": 2.2235904082955282,
+      "grad_norm": 0.8761224746704102,
+      "learning_rate": 8.280402801378662e-07,
+      "loss": 0.0586,
+      "step": 13724
+    },
+    {
+      "epoch": 2.223752430330525,
+      "grad_norm": 0.88485187292099,
+      "learning_rate": 8.277151889830054e-07,
+      "loss": 0.063,
+      "step": 13725
+    },
+    {
+      "epoch": 2.2239144523655217,
+      "grad_norm": 1.0839747190475464,
+      "learning_rate": 8.27390148996012e-07,
+      "loss": 0.069,
+      "step": 13726
+    },
+    {
+      "epoch": 2.2240764744005186,
+      "grad_norm": 0.9806898832321167,
+      "learning_rate": 8.270651601868321e-07,
+      "loss": 0.0658,
+      "step": 13727
+    },
+    {
+      "epoch": 2.224238496435515,
+      "grad_norm": 0.8707731366157532,
+      "learning_rate": 8.267402225654112e-07,
+      "loss": 0.0614,
+      "step": 13728
+    },
+    {
+      "epoch": 2.224400518470512,
+      "grad_norm": 0.7389007210731506,
+      "learning_rate": 8.264153361416893e-07,
+      "loss": 0.0543,
+      "step": 13729
+    },
+    {
+      "epoch": 2.224562540505509,
+      "grad_norm": 0.907781183719635,
+      "learning_rate": 8.260905009256081e-07,
+      "loss": 0.061,
+      "step": 13730
+    },
+    {
+      "epoch": 2.2247245625405054,
+      "grad_norm": 0.976334273815155,
+      "learning_rate": 8.257657169271071e-07,
+      "loss": 0.0654,
+      "step": 13731
+    },
+    {
+      "epoch": 2.2248865845755024,
+      "grad_norm": 0.9755533337593079,
+      "learning_rate": 8.254409841561234e-07,
+      "loss": 0.0697,
+      "step": 13732
+    },
+    {
+      "epoch": 2.225048606610499,
+      "grad_norm": 0.8844923377037048,
+      "learning_rate": 8.251163026225934e-07,
+      "loss": 0.0638,
+      "step": 13733
+    },
+    {
+      "epoch": 2.225210628645496,
+      "grad_norm": 0.8760762810707092,
+      "learning_rate": 8.24791672336451e-07,
+      "loss": 0.059,
+      "step": 13734
+    },
+    {
+      "epoch": 2.2253726506804927,
+      "grad_norm": 0.8645013570785522,
+      "learning_rate": 8.244670933076298e-07,
+      "loss": 0.0717,
+      "step": 13735
+    },
+    {
+      "epoch": 2.225534672715489,
+      "grad_norm": 0.8673563003540039,
+      "learning_rate": 8.241425655460616e-07,
+      "loss": 0.0601,
+      "step": 13736
+    },
+    {
+      "epoch": 2.225696694750486,
+      "grad_norm": 0.8699643015861511,
+      "learning_rate": 8.23818089061674e-07,
+      "loss": 0.0619,
+      "step": 13737
+    },
+    {
+      "epoch": 2.2258587167854826,
+      "grad_norm": 0.8396235108375549,
+      "learning_rate": 8.23493663864397e-07,
+      "loss": 0.0601,
+      "step": 13738
+    },
+    {
+      "epoch": 2.2260207388204796,
+      "grad_norm": 0.9061728715896606,
+      "learning_rate": 8.231692899641572e-07,
+      "loss": 0.0656,
+      "step": 13739
+    },
+    {
+      "epoch": 2.2261827608554765,
+      "grad_norm": 0.9282206892967224,
+      "learning_rate": 8.228449673708797e-07,
+      "loss": 0.0617,
+      "step": 13740
+    },
+    {
+      "epoch": 2.226344782890473,
+      "grad_norm": 0.8784255385398865,
+      "learning_rate": 8.22520696094487e-07,
+      "loss": 0.0644,
+      "step": 13741
+    },
+    {
+      "epoch": 2.22650680492547,
+      "grad_norm": 1.012131690979004,
+      "learning_rate": 8.221964761449008e-07,
+      "loss": 0.0589,
+      "step": 13742
+    },
+    {
+      "epoch": 2.2266688269604664,
+      "grad_norm": 0.80937260389328,
+      "learning_rate": 8.218723075320437e-07,
+      "loss": 0.0556,
+      "step": 13743
+    },
+    {
+      "epoch": 2.2268308489954634,
+      "grad_norm": 0.9021971821784973,
+      "learning_rate": 8.215481902658323e-07,
+      "loss": 0.0644,
+      "step": 13744
+    },
+    {
+      "epoch": 2.2269928710304603,
+      "grad_norm": 0.916597306728363,
+      "learning_rate": 8.212241243561845e-07,
+      "loss": 0.0588,
+      "step": 13745
+    },
+    {
+      "epoch": 2.2271548930654568,
+      "grad_norm": 0.8220802545547485,
+      "learning_rate": 8.209001098130157e-07,
+      "loss": 0.0563,
+      "step": 13746
+    },
+    {
+      "epoch": 2.2273169151004537,
+      "grad_norm": 0.8931221961975098,
+      "learning_rate": 8.205761466462403e-07,
+      "loss": 0.0614,
+      "step": 13747
+    },
+    {
+      "epoch": 2.2274789371354506,
+      "grad_norm": 0.755279004573822,
+      "learning_rate": 8.20252234865771e-07,
+      "loss": 0.0602,
+      "step": 13748
+    },
+    {
+      "epoch": 2.227640959170447,
+      "grad_norm": 0.8481705188751221,
+      "learning_rate": 8.199283744815181e-07,
+      "loss": 0.057,
+      "step": 13749
+    },
+    {
+      "epoch": 2.227802981205444,
+      "grad_norm": 0.9167376160621643,
+      "learning_rate": 8.196045655033913e-07,
+      "loss": 0.0679,
+      "step": 13750
+    },
+    {
+      "epoch": 2.2279650032404406,
+      "grad_norm": 0.9409997463226318,
+      "learning_rate": 8.192808079412984e-07,
+      "loss": 0.0669,
+      "step": 13751
+    },
+    {
+      "epoch": 2.2281270252754375,
+      "grad_norm": 0.7802384495735168,
+      "learning_rate": 8.189571018051454e-07,
+      "loss": 0.0577,
+      "step": 13752
+    },
+    {
+      "epoch": 2.2282890473104344,
+      "grad_norm": 1.0504868030548096,
+      "learning_rate": 8.186334471048371e-07,
+      "loss": 0.0615,
+      "step": 13753
+    },
+    {
+      "epoch": 2.228451069345431,
+      "grad_norm": 0.762758731842041,
+      "learning_rate": 8.183098438502771e-07,
+      "loss": 0.0565,
+      "step": 13754
+    },
+    {
+      "epoch": 2.228613091380428,
+      "grad_norm": 0.9312151074409485,
+      "learning_rate": 8.179862920513656e-07,
+      "loss": 0.063,
+      "step": 13755
+    },
+    {
+      "epoch": 2.2287751134154243,
+      "grad_norm": 1.0615299940109253,
+      "learning_rate": 8.176627917180025e-07,
+      "loss": 0.0693,
+      "step": 13756
+    },
+    {
+      "epoch": 2.2289371354504213,
+      "grad_norm": 1.0079830884933472,
+      "learning_rate": 8.173393428600876e-07,
+      "loss": 0.0621,
+      "step": 13757
+    },
+    {
+      "epoch": 2.229099157485418,
+      "grad_norm": 1.1644383668899536,
+      "learning_rate": 8.170159454875173e-07,
+      "loss": 0.0672,
+      "step": 13758
+    },
+    {
+      "epoch": 2.2292611795204147,
+      "grad_norm": 0.910068929195404,
+      "learning_rate": 8.16692599610186e-07,
+      "loss": 0.0656,
+      "step": 13759
+    },
+    {
+      "epoch": 2.2294232015554116,
+      "grad_norm": 0.8690840005874634,
+      "learning_rate": 8.163693052379873e-07,
+      "loss": 0.0583,
+      "step": 13760
+    },
+    {
+      "epoch": 2.229585223590408,
+      "grad_norm": 0.8714174628257751,
+      "learning_rate": 8.160460623808136e-07,
+      "loss": 0.067,
+      "step": 13761
+    },
+    {
+      "epoch": 2.229747245625405,
+      "grad_norm": 0.9453072547912598,
+      "learning_rate": 8.157228710485554e-07,
+      "loss": 0.0694,
+      "step": 13762
+    },
+    {
+      "epoch": 2.229909267660402,
+      "grad_norm": 0.8522040843963623,
+      "learning_rate": 8.153997312511014e-07,
+      "loss": 0.06,
+      "step": 13763
+    },
+    {
+      "epoch": 2.2300712896953985,
+      "grad_norm": 1.055083155632019,
+      "learning_rate": 8.15076642998339e-07,
+      "loss": 0.0612,
+      "step": 13764
+    },
+    {
+      "epoch": 2.2302333117303954,
+      "grad_norm": 0.7458101511001587,
+      "learning_rate": 8.147536063001549e-07,
+      "loss": 0.0586,
+      "step": 13765
+    },
+    {
+      "epoch": 2.230395333765392,
+      "grad_norm": 0.8568379282951355,
+      "learning_rate": 8.144306211664302e-07,
+      "loss": 0.0622,
+      "step": 13766
+    },
+    {
+      "epoch": 2.230557355800389,
+      "grad_norm": 0.8845552802085876,
+      "learning_rate": 8.141076876070505e-07,
+      "loss": 0.0635,
+      "step": 13767
+    },
+    {
+      "epoch": 2.2307193778353858,
+      "grad_norm": 0.8834683299064636,
+      "learning_rate": 8.137848056318959e-07,
+      "loss": 0.0577,
+      "step": 13768
+    },
+    {
+      "epoch": 2.2308813998703823,
+      "grad_norm": 0.8642720580101013,
+      "learning_rate": 8.134619752508463e-07,
+      "loss": 0.0618,
+      "step": 13769
+    },
+    {
+      "epoch": 2.231043421905379,
+      "grad_norm": 0.991050660610199,
+      "learning_rate": 8.131391964737773e-07,
+      "loss": 0.0673,
+      "step": 13770
+    },
+    {
+      "epoch": 2.231205443940376,
+      "grad_norm": 0.8958488702774048,
+      "learning_rate": 8.128164693105678e-07,
+      "loss": 0.064,
+      "step": 13771
+    },
+    {
+      "epoch": 2.2313674659753726,
+      "grad_norm": 1.0806387662887573,
+      "learning_rate": 8.12493793771092e-07,
+      "loss": 0.0648,
+      "step": 13772
+    },
+    {
+      "epoch": 2.2315294880103695,
+      "grad_norm": 0.922431468963623,
+      "learning_rate": 8.121711698652219e-07,
+      "loss": 0.0572,
+      "step": 13773
+    },
+    {
+      "epoch": 2.231691510045366,
+      "grad_norm": 0.9419412016868591,
+      "learning_rate": 8.118485976028292e-07,
+      "loss": 0.0612,
+      "step": 13774
+    },
+    {
+      "epoch": 2.231853532080363,
+      "grad_norm": 1.0189546346664429,
+      "learning_rate": 8.115260769937835e-07,
+      "loss": 0.065,
+      "step": 13775
+    },
+    {
+      "epoch": 2.2320155541153595,
+      "grad_norm": 0.8037611246109009,
+      "learning_rate": 8.112036080479554e-07,
+      "loss": 0.0612,
+      "step": 13776
+    },
+    {
+      "epoch": 2.2321775761503564,
+      "grad_norm": 0.9175558090209961,
+      "learning_rate": 8.108811907752093e-07,
+      "loss": 0.065,
+      "step": 13777
+    },
+    {
+      "epoch": 2.2323395981853533,
+      "grad_norm": 0.9038329124450684,
+      "learning_rate": 8.10558825185411e-07,
+      "loss": 0.0585,
+      "step": 13778
+    },
+    {
+      "epoch": 2.23250162022035,
+      "grad_norm": 0.8255992531776428,
+      "learning_rate": 8.102365112884248e-07,
+      "loss": 0.0649,
+      "step": 13779
+    },
+    {
+      "epoch": 2.2326636422553467,
+      "grad_norm": 0.8868559002876282,
+      "learning_rate": 8.099142490941117e-07,
+      "loss": 0.0623,
+      "step": 13780
+    },
+    {
+      "epoch": 2.2328256642903437,
+      "grad_norm": 0.8174579739570618,
+      "learning_rate": 8.095920386123327e-07,
+      "loss": 0.059,
+      "step": 13781
+    },
+    {
+      "epoch": 2.23298768632534,
+      "grad_norm": 0.9110251069068909,
+      "learning_rate": 8.09269879852947e-07,
+      "loss": 0.064,
+      "step": 13782
+    },
+    {
+      "epoch": 2.233149708360337,
+      "grad_norm": 1.100704550743103,
+      "learning_rate": 8.08947772825811e-07,
+      "loss": 0.0637,
+      "step": 13783
+    },
+    {
+      "epoch": 2.2333117303953336,
+      "grad_norm": 0.9022136330604553,
+      "learning_rate": 8.086257175407819e-07,
+      "loss": 0.0534,
+      "step": 13784
+    },
+    {
+      "epoch": 2.2334737524303305,
+      "grad_norm": 1.0603289604187012,
+      "learning_rate": 8.083037140077113e-07,
+      "loss": 0.0619,
+      "step": 13785
+    },
+    {
+      "epoch": 2.2336357744653275,
+      "grad_norm": 0.9479257464408875,
+      "learning_rate": 8.079817622364539e-07,
+      "loss": 0.0647,
+      "step": 13786
+    },
+    {
+      "epoch": 2.233797796500324,
+      "grad_norm": 0.8262972235679626,
+      "learning_rate": 8.076598622368606e-07,
+      "loss": 0.0535,
+      "step": 13787
+    },
+    {
+      "epoch": 2.233959818535321,
+      "grad_norm": 1.2946476936340332,
+      "learning_rate": 8.073380140187795e-07,
+      "loss": 0.0728,
+      "step": 13788
+    },
+    {
+      "epoch": 2.2341218405703174,
+      "grad_norm": 0.8733221292495728,
+      "learning_rate": 8.07016217592059e-07,
+      "loss": 0.0654,
+      "step": 13789
+    },
+    {
+      "epoch": 2.2342838626053143,
+      "grad_norm": 0.9510716795921326,
+      "learning_rate": 8.066944729665455e-07,
+      "loss": 0.0665,
+      "step": 13790
+    },
+    {
+      "epoch": 2.2344458846403112,
+      "grad_norm": 0.9815870523452759,
+      "learning_rate": 8.063727801520832e-07,
+      "loss": 0.0639,
+      "step": 13791
+    },
+    {
+      "epoch": 2.2346079066753077,
+      "grad_norm": 0.7917437553405762,
+      "learning_rate": 8.060511391585152e-07,
+      "loss": 0.0597,
+      "step": 13792
+    },
+    {
+      "epoch": 2.2347699287103047,
+      "grad_norm": 1.0030494928359985,
+      "learning_rate": 8.057295499956832e-07,
+      "loss": 0.0701,
+      "step": 13793
+    },
+    {
+      "epoch": 2.234931950745301,
+      "grad_norm": 0.933049201965332,
+      "learning_rate": 8.054080126734271e-07,
+      "loss": 0.0682,
+      "step": 13794
+    },
+    {
+      "epoch": 2.235093972780298,
+      "grad_norm": 0.9111407399177551,
+      "learning_rate": 8.050865272015848e-07,
+      "loss": 0.0632,
+      "step": 13795
+    },
+    {
+      "epoch": 2.235255994815295,
+      "grad_norm": 0.9394775032997131,
+      "learning_rate": 8.047650935899931e-07,
+      "loss": 0.0664,
+      "step": 13796
+    },
+    {
+      "epoch": 2.2354180168502915,
+      "grad_norm": 0.9155844449996948,
+      "learning_rate": 8.044437118484874e-07,
+      "loss": 0.0626,
+      "step": 13797
+    },
+    {
+      "epoch": 2.2355800388852884,
+      "grad_norm": 0.9947577118873596,
+      "learning_rate": 8.041223819869015e-07,
+      "loss": 0.0598,
+      "step": 13798
+    },
+    {
+      "epoch": 2.235742060920285,
+      "grad_norm": 0.9133854508399963,
+      "learning_rate": 8.03801104015065e-07,
+      "loss": 0.0639,
+      "step": 13799
+    },
+    {
+      "epoch": 2.235904082955282,
+      "grad_norm": 0.981562077999115,
+      "learning_rate": 8.034798779428113e-07,
+      "loss": 0.0565,
+      "step": 13800
+    },
+    {
+      "epoch": 2.236066104990279,
+      "grad_norm": 0.8836338520050049,
+      "learning_rate": 8.031587037799673e-07,
+      "loss": 0.0606,
+      "step": 13801
+    },
+    {
+      "epoch": 2.2362281270252753,
+      "grad_norm": 0.9972084760665894,
+      "learning_rate": 8.02837581536362e-07,
+      "loss": 0.0707,
+      "step": 13802
+    },
+    {
+      "epoch": 2.2363901490602722,
+      "grad_norm": 0.8902301788330078,
+      "learning_rate": 8.025165112218186e-07,
+      "loss": 0.0634,
+      "step": 13803
+    },
+    {
+      "epoch": 2.236552171095269,
+      "grad_norm": 0.9194254875183105,
+      "learning_rate": 8.021954928461611e-07,
+      "loss": 0.0632,
+      "step": 13804
+    },
+    {
+      "epoch": 2.2367141931302656,
+      "grad_norm": 1.0688180923461914,
+      "learning_rate": 8.018745264192148e-07,
+      "loss": 0.0654,
+      "step": 13805
+    },
+    {
+      "epoch": 2.2368762151652626,
+      "grad_norm": 0.9109787940979004,
+      "learning_rate": 8.015536119507977e-07,
+      "loss": 0.0604,
+      "step": 13806
+    },
+    {
+      "epoch": 2.237038237200259,
+      "grad_norm": 0.9395532011985779,
+      "learning_rate": 8.012327494507302e-07,
+      "loss": 0.0617,
+      "step": 13807
+    },
+    {
+      "epoch": 2.237200259235256,
+      "grad_norm": 0.8452723622322083,
+      "learning_rate": 8.009119389288292e-07,
+      "loss": 0.0591,
+      "step": 13808
+    },
+    {
+      "epoch": 2.237362281270253,
+      "grad_norm": 0.9636961221694946,
+      "learning_rate": 8.005911803949115e-07,
+      "loss": 0.0644,
+      "step": 13809
+    },
+    {
+      "epoch": 2.2375243033052494,
+      "grad_norm": 0.8222397565841675,
+      "learning_rate": 8.002704738587911e-07,
+      "loss": 0.0579,
+      "step": 13810
+    },
+    {
+      "epoch": 2.2376863253402464,
+      "grad_norm": 0.8121089935302734,
+      "learning_rate": 7.999498193302807e-07,
+      "loss": 0.059,
+      "step": 13811
+    },
+    {
+      "epoch": 2.237848347375243,
+      "grad_norm": 0.8787469863891602,
+      "learning_rate": 7.996292168191919e-07,
+      "loss": 0.0537,
+      "step": 13812
+    },
+    {
+      "epoch": 2.23801036941024,
+      "grad_norm": 0.7887811064720154,
+      "learning_rate": 7.993086663353344e-07,
+      "loss": 0.06,
+      "step": 13813
+    },
+    {
+      "epoch": 2.2381723914452367,
+      "grad_norm": 0.9501045942306519,
+      "learning_rate": 7.989881678885158e-07,
+      "loss": 0.061,
+      "step": 13814
+    },
+    {
+      "epoch": 2.238334413480233,
+      "grad_norm": 0.8966881036758423,
+      "learning_rate": 7.986677214885433e-07,
+      "loss": 0.0625,
+      "step": 13815
+    },
+    {
+      "epoch": 2.23849643551523,
+      "grad_norm": 0.8494555354118347,
+      "learning_rate": 7.983473271452219e-07,
+      "loss": 0.0558,
+      "step": 13816
+    },
+    {
+      "epoch": 2.2386584575502266,
+      "grad_norm": 0.8177490234375,
+      "learning_rate": 7.980269848683536e-07,
+      "loss": 0.0542,
+      "step": 13817
+    },
+    {
+      "epoch": 2.2388204795852236,
+      "grad_norm": 0.9580766558647156,
+      "learning_rate": 7.977066946677404e-07,
+      "loss": 0.0619,
+      "step": 13818
+    },
+    {
+      "epoch": 2.2389825016202205,
+      "grad_norm": 1.0076230764389038,
+      "learning_rate": 7.973864565531833e-07,
+      "loss": 0.0616,
+      "step": 13819
+    },
+    {
+      "epoch": 2.239144523655217,
+      "grad_norm": 1.0264263153076172,
+      "learning_rate": 7.970662705344812e-07,
+      "loss": 0.058,
+      "step": 13820
+    },
+    {
+      "epoch": 2.239306545690214,
+      "grad_norm": 0.7906507253646851,
+      "learning_rate": 7.967461366214293e-07,
+      "loss": 0.0569,
+      "step": 13821
+    },
+    {
+      "epoch": 2.2394685677252104,
+      "grad_norm": 0.7899366617202759,
+      "learning_rate": 7.964260548238242e-07,
+      "loss": 0.0547,
+      "step": 13822
+    },
+    {
+      "epoch": 2.2396305897602073,
+      "grad_norm": 0.9811822772026062,
+      "learning_rate": 7.961060251514591e-07,
+      "loss": 0.0609,
+      "step": 13823
+    },
+    {
+      "epoch": 2.2397926117952043,
+      "grad_norm": 0.7992488145828247,
+      "learning_rate": 7.957860476141261e-07,
+      "loss": 0.0589,
+      "step": 13824
+    },
+    {
+      "epoch": 2.2399546338302008,
+      "grad_norm": 0.9526104927062988,
+      "learning_rate": 7.954661222216162e-07,
+      "loss": 0.067,
+      "step": 13825
+    },
+    {
+      "epoch": 2.2401166558651977,
+      "grad_norm": 0.8279509544372559,
+      "learning_rate": 7.951462489837178e-07,
+      "loss": 0.0608,
+      "step": 13826
+    },
+    {
+      "epoch": 2.2402786779001946,
+      "grad_norm": 0.9480863213539124,
+      "learning_rate": 7.948264279102186e-07,
+      "loss": 0.0639,
+      "step": 13827
+    },
+    {
+      "epoch": 2.240440699935191,
+      "grad_norm": 1.0198253393173218,
+      "learning_rate": 7.945066590109044e-07,
+      "loss": 0.0667,
+      "step": 13828
+    },
+    {
+      "epoch": 2.240602721970188,
+      "grad_norm": 0.9767336845397949,
+      "learning_rate": 7.941869422955592e-07,
+      "loss": 0.0663,
+      "step": 13829
+    },
+    {
+      "epoch": 2.2407647440051845,
+      "grad_norm": 0.8995802402496338,
+      "learning_rate": 7.938672777739654e-07,
+      "loss": 0.0588,
+      "step": 13830
+    },
+    {
+      "epoch": 2.2409267660401815,
+      "grad_norm": 1.019019603729248,
+      "learning_rate": 7.935476654559052e-07,
+      "loss": 0.0574,
+      "step": 13831
+    },
+    {
+      "epoch": 2.2410887880751784,
+      "grad_norm": 0.8723592758178711,
+      "learning_rate": 7.932281053511559e-07,
+      "loss": 0.0649,
+      "step": 13832
+    },
+    {
+      "epoch": 2.241250810110175,
+      "grad_norm": 1.199785590171814,
+      "learning_rate": 7.929085974694956e-07,
+      "loss": 0.0671,
+      "step": 13833
+    },
+    {
+      "epoch": 2.241412832145172,
+      "grad_norm": 0.9804271459579468,
+      "learning_rate": 7.925891418207024e-07,
+      "loss": 0.0613,
+      "step": 13834
+    },
+    {
+      "epoch": 2.2415748541801683,
+      "grad_norm": 0.7946493625640869,
+      "learning_rate": 7.922697384145492e-07,
+      "loss": 0.0595,
+      "step": 13835
+    },
+    {
+      "epoch": 2.2417368762151653,
+      "grad_norm": 0.9132854342460632,
+      "learning_rate": 7.919503872608092e-07,
+      "loss": 0.0628,
+      "step": 13836
+    },
+    {
+      "epoch": 2.241898898250162,
+      "grad_norm": 0.9299383163452148,
+      "learning_rate": 7.916310883692532e-07,
+      "loss": 0.07,
+      "step": 13837
+    },
+    {
+      "epoch": 2.2420609202851587,
+      "grad_norm": 0.9028305411338806,
+      "learning_rate": 7.913118417496532e-07,
+      "loss": 0.069,
+      "step": 13838
+    },
+    {
+      "epoch": 2.2422229423201556,
+      "grad_norm": 0.9909502267837524,
+      "learning_rate": 7.909926474117752e-07,
+      "loss": 0.0608,
+      "step": 13839
+    },
+    {
+      "epoch": 2.242384964355152,
+      "grad_norm": 1.0152370929718018,
+      "learning_rate": 7.906735053653866e-07,
+      "loss": 0.0648,
+      "step": 13840
+    },
+    {
+      "epoch": 2.242546986390149,
+      "grad_norm": 0.8076937794685364,
+      "learning_rate": 7.90354415620252e-07,
+      "loss": 0.0589,
+      "step": 13841
+    },
+    {
+      "epoch": 2.242709008425146,
+      "grad_norm": 0.9617068767547607,
+      "learning_rate": 7.900353781861353e-07,
+      "loss": 0.0603,
+      "step": 13842
+    },
+    {
+      "epoch": 2.2428710304601425,
+      "grad_norm": 1.0339735746383667,
+      "learning_rate": 7.897163930727977e-07,
+      "loss": 0.0601,
+      "step": 13843
+    },
+    {
+      "epoch": 2.2430330524951394,
+      "grad_norm": 1.0885800123214722,
+      "learning_rate": 7.893974602899998e-07,
+      "loss": 0.0664,
+      "step": 13844
+    },
+    {
+      "epoch": 2.243195074530136,
+      "grad_norm": 0.8064873218536377,
+      "learning_rate": 7.890785798475001e-07,
+      "loss": 0.0589,
+      "step": 13845
+    },
+    {
+      "epoch": 2.243357096565133,
+      "grad_norm": 0.8364428281784058,
+      "learning_rate": 7.887597517550564e-07,
+      "loss": 0.0587,
+      "step": 13846
+    },
+    {
+      "epoch": 2.2435191186001298,
+      "grad_norm": 1.0745041370391846,
+      "learning_rate": 7.884409760224212e-07,
+      "loss": 0.074,
+      "step": 13847
+    },
+    {
+      "epoch": 2.2436811406351262,
+      "grad_norm": 0.9691507816314697,
+      "learning_rate": 7.881222526593513e-07,
+      "loss": 0.0607,
+      "step": 13848
+    },
+    {
+      "epoch": 2.243843162670123,
+      "grad_norm": 0.9089518189430237,
+      "learning_rate": 7.878035816755985e-07,
+      "loss": 0.0575,
+      "step": 13849
+    },
+    {
+      "epoch": 2.24400518470512,
+      "grad_norm": 0.8140989542007446,
+      "learning_rate": 7.87484963080912e-07,
+      "loss": 0.0586,
+      "step": 13850
+    },
+    {
+      "epoch": 2.2441672067401166,
+      "grad_norm": 0.6760884523391724,
+      "learning_rate": 7.871663968850413e-07,
+      "loss": 0.0508,
+      "step": 13851
+    },
+    {
+      "epoch": 2.2443292287751135,
+      "grad_norm": 0.9919477105140686,
+      "learning_rate": 7.868478830977331e-07,
+      "loss": 0.0655,
+      "step": 13852
+    },
+    {
+      "epoch": 2.24449125081011,
+      "grad_norm": 0.8631847500801086,
+      "learning_rate": 7.865294217287356e-07,
+      "loss": 0.0672,
+      "step": 13853
+    },
+    {
+      "epoch": 2.244653272845107,
+      "grad_norm": 0.964144229888916,
+      "learning_rate": 7.862110127877903e-07,
+      "loss": 0.0652,
+      "step": 13854
+    },
+    {
+      "epoch": 2.244815294880104,
+      "grad_norm": 0.8476447463035583,
+      "learning_rate": 7.858926562846409e-07,
+      "loss": 0.0572,
+      "step": 13855
+    },
+    {
+      "epoch": 2.2449773169151004,
+      "grad_norm": 0.7753357887268066,
+      "learning_rate": 7.855743522290283e-07,
+      "loss": 0.056,
+      "step": 13856
+    },
+    {
+      "epoch": 2.2451393389500973,
+      "grad_norm": 0.9003604650497437,
+      "learning_rate": 7.852561006306913e-07,
+      "loss": 0.0638,
+      "step": 13857
+    },
+    {
+      "epoch": 2.245301360985094,
+      "grad_norm": 0.8805102109909058,
+      "learning_rate": 7.849379014993683e-07,
+      "loss": 0.0593,
+      "step": 13858
+    },
+    {
+      "epoch": 2.2454633830200907,
+      "grad_norm": 1.0749431848526,
+      "learning_rate": 7.84619754844795e-07,
+      "loss": 0.07,
+      "step": 13859
+    },
+    {
+      "epoch": 2.2456254050550877,
+      "grad_norm": 0.8165993094444275,
+      "learning_rate": 7.84301660676707e-07,
+      "loss": 0.0574,
+      "step": 13860
+    },
+    {
+      "epoch": 2.245787427090084,
+      "grad_norm": 0.9074599742889404,
+      "learning_rate": 7.839836190048344e-07,
+      "loss": 0.0617,
+      "step": 13861
+    },
+    {
+      "epoch": 2.245949449125081,
+      "grad_norm": 0.8614346981048584,
+      "learning_rate": 7.836656298389114e-07,
+      "loss": 0.0585,
+      "step": 13862
+    },
+    {
+      "epoch": 2.2461114711600776,
+      "grad_norm": 0.908950924873352,
+      "learning_rate": 7.833476931886666e-07,
+      "loss": 0.0617,
+      "step": 13863
+    },
+    {
+      "epoch": 2.2462734931950745,
+      "grad_norm": 0.9891899228096008,
+      "learning_rate": 7.830298090638291e-07,
+      "loss": 0.0613,
+      "step": 13864
+    },
+    {
+      "epoch": 2.2464355152300715,
+      "grad_norm": 0.8221472501754761,
+      "learning_rate": 7.827119774741238e-07,
+      "loss": 0.0609,
+      "step": 13865
+    },
+    {
+      "epoch": 2.246597537265068,
+      "grad_norm": 0.8985829949378967,
+      "learning_rate": 7.823941984292752e-07,
+      "loss": 0.0644,
+      "step": 13866
+    },
+    {
+      "epoch": 2.246759559300065,
+      "grad_norm": 1.03141450881958,
+      "learning_rate": 7.82076471939009e-07,
+      "loss": 0.0713,
+      "step": 13867
+    },
+    {
+      "epoch": 2.2469215813350614,
+      "grad_norm": 0.8261798620223999,
+      "learning_rate": 7.817587980130451e-07,
+      "loss": 0.0625,
+      "step": 13868
+    },
+    {
+      "epoch": 2.2470836033700583,
+      "grad_norm": 0.9878301620483398,
+      "learning_rate": 7.814411766611035e-07,
+      "loss": 0.0581,
+      "step": 13869
+    },
+    {
+      "epoch": 2.2472456254050552,
+      "grad_norm": 0.9584047794342041,
+      "learning_rate": 7.811236078929033e-07,
+      "loss": 0.0579,
+      "step": 13870
+    },
+    {
+      "epoch": 2.2474076474400517,
+      "grad_norm": 1.0649832487106323,
+      "learning_rate": 7.808060917181609e-07,
+      "loss": 0.0656,
+      "step": 13871
+    },
+    {
+      "epoch": 2.2475696694750487,
+      "grad_norm": 0.9474183917045593,
+      "learning_rate": 7.80488628146592e-07,
+      "loss": 0.0678,
+      "step": 13872
+    },
+    {
+      "epoch": 2.2477316915100456,
+      "grad_norm": 1.1049025058746338,
+      "learning_rate": 7.801712171879098e-07,
+      "loss": 0.0667,
+      "step": 13873
+    },
+    {
+      "epoch": 2.247893713545042,
+      "grad_norm": 0.8262870907783508,
+      "learning_rate": 7.798538588518265e-07,
+      "loss": 0.0596,
+      "step": 13874
+    },
+    {
+      "epoch": 2.248055735580039,
+      "grad_norm": 0.9269388914108276,
+      "learning_rate": 7.795365531480531e-07,
+      "loss": 0.0747,
+      "step": 13875
+    },
+    {
+      "epoch": 2.2482177576150355,
+      "grad_norm": 0.8992902040481567,
+      "learning_rate": 7.792193000862964e-07,
+      "loss": 0.0583,
+      "step": 13876
+    },
+    {
+      "epoch": 2.2483797796500324,
+      "grad_norm": 0.9247242212295532,
+      "learning_rate": 7.789020996762656e-07,
+      "loss": 0.0561,
+      "step": 13877
+    },
+    {
+      "epoch": 2.248541801685029,
+      "grad_norm": 0.9147598147392273,
+      "learning_rate": 7.785849519276661e-07,
+      "loss": 0.0595,
+      "step": 13878
+    },
+    {
+      "epoch": 2.248703823720026,
+      "grad_norm": 0.9816052913665771,
+      "learning_rate": 7.782678568502008e-07,
+      "loss": 0.0665,
+      "step": 13879
+    },
+    {
+      "epoch": 2.248865845755023,
+      "grad_norm": 0.8535481691360474,
+      "learning_rate": 7.779508144535725e-07,
+      "loss": 0.0621,
+      "step": 13880
+    },
+    {
+      "epoch": 2.2490278677900193,
+      "grad_norm": 0.8713977336883545,
+      "learning_rate": 7.776338247474812e-07,
+      "loss": 0.0641,
+      "step": 13881
+    },
+    {
+      "epoch": 2.249189889825016,
+      "grad_norm": 0.8746277093887329,
+      "learning_rate": 7.773168877416285e-07,
+      "loss": 0.058,
+      "step": 13882
+    },
+    {
+      "epoch": 2.249351911860013,
+      "grad_norm": 0.837213397026062,
+      "learning_rate": 7.770000034457092e-07,
+      "loss": 0.0616,
+      "step": 13883
+    },
+    {
+      "epoch": 2.2495139338950096,
+      "grad_norm": 0.8210061192512512,
+      "learning_rate": 7.766831718694204e-07,
+      "loss": 0.0613,
+      "step": 13884
+    },
+    {
+      "epoch": 2.2496759559300066,
+      "grad_norm": 0.8450611233711243,
+      "learning_rate": 7.763663930224563e-07,
+      "loss": 0.0617,
+      "step": 13885
+    },
+    {
+      "epoch": 2.249837977965003,
+      "grad_norm": 1.0194191932678223,
+      "learning_rate": 7.760496669145093e-07,
+      "loss": 0.0662,
+      "step": 13886
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.9273967146873474,
+      "learning_rate": 7.757329935552707e-07,
+      "loss": 0.0602,
+      "step": 13887
+    },
+    {
+      "epoch": 2.250162022034997,
+      "grad_norm": 0.8759946227073669,
+      "learning_rate": 7.754163729544297e-07,
+      "loss": 0.059,
+      "step": 13888
+    },
+    {
+      "epoch": 2.2503240440699934,
+      "grad_norm": 0.870991051197052,
+      "learning_rate": 7.750998051216743e-07,
+      "loss": 0.0674,
+      "step": 13889
+    },
+    {
+      "epoch": 2.2504860661049904,
+      "grad_norm": 0.8487216830253601,
+      "learning_rate": 7.747832900666907e-07,
+      "loss": 0.0642,
+      "step": 13890
+    },
+    {
+      "epoch": 2.250648088139987,
+      "grad_norm": 1.057746171951294,
+      "learning_rate": 7.744668277991635e-07,
+      "loss": 0.0703,
+      "step": 13891
+    },
+    {
+      "epoch": 2.250810110174984,
+      "grad_norm": 0.8206788301467896,
+      "learning_rate": 7.741504183287757e-07,
+      "loss": 0.0604,
+      "step": 13892
+    },
+    {
+      "epoch": 2.2509721322099807,
+      "grad_norm": 0.9289180040359497,
+      "learning_rate": 7.738340616652096e-07,
+      "loss": 0.0557,
+      "step": 13893
+    },
+    {
+      "epoch": 2.251134154244977,
+      "grad_norm": 0.8246344923973083,
+      "learning_rate": 7.73517757818143e-07,
+      "loss": 0.0589,
+      "step": 13894
+    },
+    {
+      "epoch": 2.251296176279974,
+      "grad_norm": 0.9617959856987,
+      "learning_rate": 7.73201506797254e-07,
+      "loss": 0.0698,
+      "step": 13895
+    },
+    {
+      "epoch": 2.251458198314971,
+      "grad_norm": 0.9463308453559875,
+      "learning_rate": 7.728853086122212e-07,
+      "loss": 0.062,
+      "step": 13896
+    },
+    {
+      "epoch": 2.2516202203499676,
+      "grad_norm": 0.9526486396789551,
+      "learning_rate": 7.725691632727192e-07,
+      "loss": 0.0671,
+      "step": 13897
+    },
+    {
+      "epoch": 2.2517822423849645,
+      "grad_norm": 1.0356853008270264,
+      "learning_rate": 7.722530707884196e-07,
+      "loss": 0.0744,
+      "step": 13898
+    },
+    {
+      "epoch": 2.251944264419961,
+      "grad_norm": 0.8983619213104248,
+      "learning_rate": 7.71937031168995e-07,
+      "loss": 0.0556,
+      "step": 13899
+    },
+    {
+      "epoch": 2.252106286454958,
+      "grad_norm": 0.8707557916641235,
+      "learning_rate": 7.716210444241154e-07,
+      "loss": 0.0623,
+      "step": 13900
+    },
+    {
+      "epoch": 2.2522683084899544,
+      "grad_norm": 0.9758105874061584,
+      "learning_rate": 7.713051105634492e-07,
+      "loss": 0.0633,
+      "step": 13901
+    },
+    {
+      "epoch": 2.2524303305249513,
+      "grad_norm": 0.8898715972900391,
+      "learning_rate": 7.709892295966634e-07,
+      "loss": 0.056,
+      "step": 13902
+    },
+    {
+      "epoch": 2.2525923525599483,
+      "grad_norm": 0.957287609577179,
+      "learning_rate": 7.706734015334228e-07,
+      "loss": 0.0673,
+      "step": 13903
+    },
+    {
+      "epoch": 2.2527543745949448,
+      "grad_norm": 0.8354018330574036,
+      "learning_rate": 7.703576263833915e-07,
+      "loss": 0.0635,
+      "step": 13904
+    },
+    {
+      "epoch": 2.2529163966299417,
+      "grad_norm": 0.9206665754318237,
+      "learning_rate": 7.70041904156231e-07,
+      "loss": 0.058,
+      "step": 13905
+    },
+    {
+      "epoch": 2.2530784186649386,
+      "grad_norm": 0.9412673711776733,
+      "learning_rate": 7.697262348616019e-07,
+      "loss": 0.0661,
+      "step": 13906
+    },
+    {
+      "epoch": 2.253240440699935,
+      "grad_norm": 0.8687724471092224,
+      "learning_rate": 7.694106185091627e-07,
+      "loss": 0.062,
+      "step": 13907
+    },
+    {
+      "epoch": 2.253402462734932,
+      "grad_norm": 1.0838840007781982,
+      "learning_rate": 7.690950551085716e-07,
+      "loss": 0.0682,
+      "step": 13908
+    },
+    {
+      "epoch": 2.2535644847699285,
+      "grad_norm": 0.8716097474098206,
+      "learning_rate": 7.687795446694815e-07,
+      "loss": 0.0592,
+      "step": 13909
+    },
+    {
+      "epoch": 2.2537265068049255,
+      "grad_norm": 0.7702820897102356,
+      "learning_rate": 7.684640872015484e-07,
+      "loss": 0.0537,
+      "step": 13910
+    },
+    {
+      "epoch": 2.2538885288399224,
+      "grad_norm": 0.8733231425285339,
+      "learning_rate": 7.68148682714425e-07,
+      "loss": 0.0612,
+      "step": 13911
+    },
+    {
+      "epoch": 2.254050550874919,
+      "grad_norm": 0.8055692315101624,
+      "learning_rate": 7.678333312177602e-07,
+      "loss": 0.0562,
+      "step": 13912
+    },
+    {
+      "epoch": 2.254212572909916,
+      "grad_norm": 0.948461651802063,
+      "learning_rate": 7.675180327212037e-07,
+      "loss": 0.0644,
+      "step": 13913
+    },
+    {
+      "epoch": 2.2543745949449123,
+      "grad_norm": 0.8782212138175964,
+      "learning_rate": 7.672027872344017e-07,
+      "loss": 0.063,
+      "step": 13914
+    },
+    {
+      "epoch": 2.2545366169799093,
+      "grad_norm": 0.9695743322372437,
+      "learning_rate": 7.668875947670032e-07,
+      "loss": 0.0607,
+      "step": 13915
+    },
+    {
+      "epoch": 2.254698639014906,
+      "grad_norm": 0.8532809615135193,
+      "learning_rate": 7.665724553286491e-07,
+      "loss": 0.0584,
+      "step": 13916
+    },
+    {
+      "epoch": 2.2548606610499027,
+      "grad_norm": 0.8807068467140198,
+      "learning_rate": 7.662573689289832e-07,
+      "loss": 0.0658,
+      "step": 13917
+    },
+    {
+      "epoch": 2.2550226830848996,
+      "grad_norm": 0.8487615585327148,
+      "learning_rate": 7.659423355776463e-07,
+      "loss": 0.0602,
+      "step": 13918
+    },
+    {
+      "epoch": 2.2551847051198965,
+      "grad_norm": 0.9274862408638,
+      "learning_rate": 7.656273552842774e-07,
+      "loss": 0.0589,
+      "step": 13919
+    },
+    {
+      "epoch": 2.255346727154893,
+      "grad_norm": 0.8932498693466187,
+      "learning_rate": 7.653124280585145e-07,
+      "loss": 0.0658,
+      "step": 13920
+    },
+    {
+      "epoch": 2.25550874918989,
+      "grad_norm": 0.9409620761871338,
+      "learning_rate": 7.649975539099935e-07,
+      "loss": 0.0634,
+      "step": 13921
+    },
+    {
+      "epoch": 2.2556707712248865,
+      "grad_norm": 0.9468281269073486,
+      "learning_rate": 7.646827328483486e-07,
+      "loss": 0.0615,
+      "step": 13922
+    },
+    {
+      "epoch": 2.2558327932598834,
+      "grad_norm": 1.0393003225326538,
+      "learning_rate": 7.643679648832133e-07,
+      "loss": 0.0657,
+      "step": 13923
+    },
+    {
+      "epoch": 2.25599481529488,
+      "grad_norm": 0.9196584820747375,
+      "learning_rate": 7.64053250024217e-07,
+      "loss": 0.0633,
+      "step": 13924
+    },
+    {
+      "epoch": 2.256156837329877,
+      "grad_norm": 1.1040884256362915,
+      "learning_rate": 7.637385882809909e-07,
+      "loss": 0.0673,
+      "step": 13925
+    },
+    {
+      "epoch": 2.2563188593648738,
+      "grad_norm": 1.06027352809906,
+      "learning_rate": 7.634239796631629e-07,
+      "loss": 0.0616,
+      "step": 13926
+    },
+    {
+      "epoch": 2.2564808813998702,
+      "grad_norm": 1.0992965698242188,
+      "learning_rate": 7.631094241803582e-07,
+      "loss": 0.078,
+      "step": 13927
+    },
+    {
+      "epoch": 2.256642903434867,
+      "grad_norm": 0.9714069366455078,
+      "learning_rate": 7.62794921842201e-07,
+      "loss": 0.0667,
+      "step": 13928
+    },
+    {
+      "epoch": 2.256804925469864,
+      "grad_norm": 1.0589038133621216,
+      "learning_rate": 7.624804726583169e-07,
+      "loss": 0.0668,
+      "step": 13929
+    },
+    {
+      "epoch": 2.2569669475048606,
+      "grad_norm": 0.8686789870262146,
+      "learning_rate": 7.621660766383246e-07,
+      "loss": 0.0573,
+      "step": 13930
+    },
+    {
+      "epoch": 2.2571289695398575,
+      "grad_norm": 0.8950777649879456,
+      "learning_rate": 7.618517337918451e-07,
+      "loss": 0.0524,
+      "step": 13931
+    },
+    {
+      "epoch": 2.257290991574854,
+      "grad_norm": 0.8135221600532532,
+      "learning_rate": 7.615374441284962e-07,
+      "loss": 0.0596,
+      "step": 13932
+    },
+    {
+      "epoch": 2.257453013609851,
+      "grad_norm": 0.8295992612838745,
+      "learning_rate": 7.612232076578946e-07,
+      "loss": 0.0618,
+      "step": 13933
+    },
+    {
+      "epoch": 2.257615035644848,
+      "grad_norm": 0.9673225283622742,
+      "learning_rate": 7.60909024389655e-07,
+      "loss": 0.0575,
+      "step": 13934
+    },
+    {
+      "epoch": 2.2577770576798444,
+      "grad_norm": 0.8375781178474426,
+      "learning_rate": 7.605948943333908e-07,
+      "loss": 0.0577,
+      "step": 13935
+    },
+    {
+      "epoch": 2.2579390797148413,
+      "grad_norm": 1.1369965076446533,
+      "learning_rate": 7.602808174987137e-07,
+      "loss": 0.0696,
+      "step": 13936
+    },
+    {
+      "epoch": 2.258101101749838,
+      "grad_norm": 0.774905800819397,
+      "learning_rate": 7.599667938952341e-07,
+      "loss": 0.0567,
+      "step": 13937
+    },
+    {
+      "epoch": 2.2582631237848347,
+      "grad_norm": 0.8955923914909363,
+      "learning_rate": 7.596528235325582e-07,
+      "loss": 0.0651,
+      "step": 13938
+    },
+    {
+      "epoch": 2.2584251458198317,
+      "grad_norm": 0.830094039440155,
+      "learning_rate": 7.593389064202952e-07,
+      "loss": 0.0658,
+      "step": 13939
+    },
+    {
+      "epoch": 2.258587167854828,
+      "grad_norm": 0.7732976675033569,
+      "learning_rate": 7.590250425680496e-07,
+      "loss": 0.0535,
+      "step": 13940
+    },
+    {
+      "epoch": 2.258749189889825,
+      "grad_norm": 1.0507856607437134,
+      "learning_rate": 7.58711231985425e-07,
+      "loss": 0.0673,
+      "step": 13941
+    },
+    {
+      "epoch": 2.2589112119248216,
+      "grad_norm": 0.9185804128646851,
+      "learning_rate": 7.583974746820222e-07,
+      "loss": 0.0662,
+      "step": 13942
+    },
+    {
+      "epoch": 2.2590732339598185,
+      "grad_norm": 0.9794391393661499,
+      "learning_rate": 7.580837706674415e-07,
+      "loss": 0.0692,
+      "step": 13943
+    },
+    {
+      "epoch": 2.2592352559948155,
+      "grad_norm": 0.8743561506271362,
+      "learning_rate": 7.577701199512835e-07,
+      "loss": 0.0611,
+      "step": 13944
+    },
+    {
+      "epoch": 2.259397278029812,
+      "grad_norm": 0.845887303352356,
+      "learning_rate": 7.574565225431427e-07,
+      "loss": 0.0609,
+      "step": 13945
+    },
+    {
+      "epoch": 2.259559300064809,
+      "grad_norm": 1.0307512283325195,
+      "learning_rate": 7.571429784526157e-07,
+      "loss": 0.0634,
+      "step": 13946
+    },
+    {
+      "epoch": 2.2597213220998054,
+      "grad_norm": 0.8212060332298279,
+      "learning_rate": 7.56829487689296e-07,
+      "loss": 0.0584,
+      "step": 13947
+    },
+    {
+      "epoch": 2.2598833441348023,
+      "grad_norm": 0.892123281955719,
+      "learning_rate": 7.565160502627752e-07,
+      "loss": 0.0619,
+      "step": 13948
+    },
+    {
+      "epoch": 2.2600453661697992,
+      "grad_norm": 0.8927651643753052,
+      "learning_rate": 7.562026661826446e-07,
+      "loss": 0.063,
+      "step": 13949
+    },
+    {
+      "epoch": 2.2602073882047957,
+      "grad_norm": 0.908979594707489,
+      "learning_rate": 7.558893354584923e-07,
+      "loss": 0.06,
+      "step": 13950
+    },
+    {
+      "epoch": 2.2603694102397927,
+      "grad_norm": 0.8717660903930664,
+      "learning_rate": 7.555760580999055e-07,
+      "loss": 0.0631,
+      "step": 13951
+    },
+    {
+      "epoch": 2.2605314322747896,
+      "grad_norm": 1.0289645195007324,
+      "learning_rate": 7.5526283411647e-07,
+      "loss": 0.0634,
+      "step": 13952
+    },
+    {
+      "epoch": 2.260693454309786,
+      "grad_norm": 0.8531572222709656,
+      "learning_rate": 7.549496635177698e-07,
+      "loss": 0.0628,
+      "step": 13953
+    },
+    {
+      "epoch": 2.260855476344783,
+      "grad_norm": 0.8821384310722351,
+      "learning_rate": 7.546365463133867e-07,
+      "loss": 0.0641,
+      "step": 13954
+    },
+    {
+      "epoch": 2.2610174983797795,
+      "grad_norm": 0.923827052116394,
+      "learning_rate": 7.543234825129026e-07,
+      "loss": 0.0629,
+      "step": 13955
+    },
+    {
+      "epoch": 2.2611795204147764,
+      "grad_norm": 0.8377568125724792,
+      "learning_rate": 7.540104721258945e-07,
+      "loss": 0.0635,
+      "step": 13956
+    },
+    {
+      "epoch": 2.261341542449773,
+      "grad_norm": 0.8890910148620605,
+      "learning_rate": 7.5369751516194e-07,
+      "loss": 0.0614,
+      "step": 13957
+    },
+    {
+      "epoch": 2.26150356448477,
+      "grad_norm": 1.1075289249420166,
+      "learning_rate": 7.533846116306162e-07,
+      "loss": 0.0598,
+      "step": 13958
+    },
+    {
+      "epoch": 2.261665586519767,
+      "grad_norm": 0.8395311236381531,
+      "learning_rate": 7.530717615414976e-07,
+      "loss": 0.0583,
+      "step": 13959
+    },
+    {
+      "epoch": 2.2618276085547633,
+      "grad_norm": 0.872234046459198,
+      "learning_rate": 7.527589649041548e-07,
+      "loss": 0.0643,
+      "step": 13960
+    },
+    {
+      "epoch": 2.26198963058976,
+      "grad_norm": 0.8673476576805115,
+      "learning_rate": 7.524462217281592e-07,
+      "loss": 0.0624,
+      "step": 13961
+    },
+    {
+      "epoch": 2.262151652624757,
+      "grad_norm": 0.9594100117683411,
+      "learning_rate": 7.521335320230804e-07,
+      "loss": 0.0614,
+      "step": 13962
+    },
+    {
+      "epoch": 2.2623136746597536,
+      "grad_norm": 0.8554080128669739,
+      "learning_rate": 7.518208957984857e-07,
+      "loss": 0.0596,
+      "step": 13963
+    },
+    {
+      "epoch": 2.2624756966947506,
+      "grad_norm": 0.8185822367668152,
+      "learning_rate": 7.515083130639411e-07,
+      "loss": 0.0617,
+      "step": 13964
+    },
+    {
+      "epoch": 2.262637718729747,
+      "grad_norm": 0.7946313619613647,
+      "learning_rate": 7.51195783829011e-07,
+      "loss": 0.0602,
+      "step": 13965
+    },
+    {
+      "epoch": 2.262799740764744,
+      "grad_norm": 0.8937302827835083,
+      "learning_rate": 7.508833081032577e-07,
+      "loss": 0.065,
+      "step": 13966
+    },
+    {
+      "epoch": 2.262961762799741,
+      "grad_norm": 0.9195818901062012,
+      "learning_rate": 7.505708858962424e-07,
+      "loss": 0.0651,
+      "step": 13967
+    },
+    {
+      "epoch": 2.2631237848347374,
+      "grad_norm": 0.9528251886367798,
+      "learning_rate": 7.502585172175244e-07,
+      "loss": 0.0698,
+      "step": 13968
+    },
+    {
+      "epoch": 2.2632858068697344,
+      "grad_norm": 0.9592039585113525,
+      "learning_rate": 7.499462020766618e-07,
+      "loss": 0.0562,
+      "step": 13969
+    },
+    {
+      "epoch": 2.263447828904731,
+      "grad_norm": 0.8582433462142944,
+      "learning_rate": 7.496339404832109e-07,
+      "loss": 0.0619,
+      "step": 13970
+    },
+    {
+      "epoch": 2.2636098509397278,
+      "grad_norm": 0.9389162063598633,
+      "learning_rate": 7.493217324467239e-07,
+      "loss": 0.071,
+      "step": 13971
+    },
+    {
+      "epoch": 2.2637718729747247,
+      "grad_norm": 0.9363165497779846,
+      "learning_rate": 7.490095779767564e-07,
+      "loss": 0.0612,
+      "step": 13972
+    },
+    {
+      "epoch": 2.263933895009721,
+      "grad_norm": 0.8165293335914612,
+      "learning_rate": 7.486974770828592e-07,
+      "loss": 0.0586,
+      "step": 13973
+    },
+    {
+      "epoch": 2.264095917044718,
+      "grad_norm": 0.822784423828125,
+      "learning_rate": 7.483854297745805e-07,
+      "loss": 0.0595,
+      "step": 13974
+    },
+    {
+      "epoch": 2.264257939079715,
+      "grad_norm": 1.0148231983184814,
+      "learning_rate": 7.480734360614686e-07,
+      "loss": 0.0658,
+      "step": 13975
+    },
+    {
+      "epoch": 2.2644199611147116,
+      "grad_norm": 0.8952381610870361,
+      "learning_rate": 7.47761495953069e-07,
+      "loss": 0.0551,
+      "step": 13976
+    },
+    {
+      "epoch": 2.2645819831497085,
+      "grad_norm": 1.0946059226989746,
+      "learning_rate": 7.474496094589292e-07,
+      "loss": 0.069,
+      "step": 13977
+    },
+    {
+      "epoch": 2.264744005184705,
+      "grad_norm": 0.7693538069725037,
+      "learning_rate": 7.471377765885893e-07,
+      "loss": 0.0551,
+      "step": 13978
+    },
+    {
+      "epoch": 2.264906027219702,
+      "grad_norm": 0.9450730681419373,
+      "learning_rate": 7.468259973515918e-07,
+      "loss": 0.0671,
+      "step": 13979
+    },
+    {
+      "epoch": 2.2650680492546984,
+      "grad_norm": 0.9570711255073547,
+      "learning_rate": 7.465142717574761e-07,
+      "loss": 0.0681,
+      "step": 13980
+    },
+    {
+      "epoch": 2.2652300712896953,
+      "grad_norm": 1.0384302139282227,
+      "learning_rate": 7.462025998157801e-07,
+      "loss": 0.0637,
+      "step": 13981
+    },
+    {
+      "epoch": 2.2653920933246923,
+      "grad_norm": 0.9101808071136475,
+      "learning_rate": 7.458909815360407e-07,
+      "loss": 0.0672,
+      "step": 13982
+    },
+    {
+      "epoch": 2.2655541153596888,
+      "grad_norm": 0.8050899505615234,
+      "learning_rate": 7.455794169277922e-07,
+      "loss": 0.0546,
+      "step": 13983
+    },
+    {
+      "epoch": 2.2657161373946857,
+      "grad_norm": 0.863530695438385,
+      "learning_rate": 7.45267906000568e-07,
+      "loss": 0.0597,
+      "step": 13984
+    },
+    {
+      "epoch": 2.2658781594296826,
+      "grad_norm": 0.8293231129646301,
+      "learning_rate": 7.449564487639005e-07,
+      "loss": 0.0586,
+      "step": 13985
+    },
+    {
+      "epoch": 2.266040181464679,
+      "grad_norm": 0.9019443988800049,
+      "learning_rate": 7.446450452273168e-07,
+      "loss": 0.0652,
+      "step": 13986
+    },
+    {
+      "epoch": 2.266202203499676,
+      "grad_norm": 1.059104084968567,
+      "learning_rate": 7.443336954003474e-07,
+      "loss": 0.0654,
+      "step": 13987
+    },
+    {
+      "epoch": 2.2663642255346725,
+      "grad_norm": 0.8297203779220581,
+      "learning_rate": 7.440223992925194e-07,
+      "loss": 0.054,
+      "step": 13988
+    },
+    {
+      "epoch": 2.2665262475696695,
+      "grad_norm": 1.012291669845581,
+      "learning_rate": 7.437111569133556e-07,
+      "loss": 0.0667,
+      "step": 13989
+    },
+    {
+      "epoch": 2.2666882696046664,
+      "grad_norm": 0.8839371800422668,
+      "learning_rate": 7.433999682723805e-07,
+      "loss": 0.0632,
+      "step": 13990
+    },
+    {
+      "epoch": 2.266850291639663,
+      "grad_norm": 0.8371898531913757,
+      "learning_rate": 7.430888333791144e-07,
+      "loss": 0.0566,
+      "step": 13991
+    },
+    {
+      "epoch": 2.26701231367466,
+      "grad_norm": 0.8507769703865051,
+      "learning_rate": 7.427777522430804e-07,
+      "loss": 0.0567,
+      "step": 13992
+    },
+    {
+      "epoch": 2.2671743357096563,
+      "grad_norm": 0.8428038358688354,
+      "learning_rate": 7.424667248737936e-07,
+      "loss": 0.0611,
+      "step": 13993
+    },
+    {
+      "epoch": 2.2673363577446533,
+      "grad_norm": 0.8883320093154907,
+      "learning_rate": 7.42155751280772e-07,
+      "loss": 0.0616,
+      "step": 13994
+    },
+    {
+      "epoch": 2.26749837977965,
+      "grad_norm": 0.839263916015625,
+      "learning_rate": 7.418448314735305e-07,
+      "loss": 0.0583,
+      "step": 13995
+    },
+    {
+      "epoch": 2.2676604018146467,
+      "grad_norm": 0.9638897180557251,
+      "learning_rate": 7.415339654615824e-07,
+      "loss": 0.0617,
+      "step": 13996
+    },
+    {
+      "epoch": 2.2678224238496436,
+      "grad_norm": 0.8128506541252136,
+      "learning_rate": 7.412231532544398e-07,
+      "loss": 0.0549,
+      "step": 13997
+    },
+    {
+      "epoch": 2.2679844458846405,
+      "grad_norm": 0.9353517889976501,
+      "learning_rate": 7.409123948616123e-07,
+      "loss": 0.0626,
+      "step": 13998
+    },
+    {
+      "epoch": 2.268146467919637,
+      "grad_norm": 0.7773940563201904,
+      "learning_rate": 7.406016902926094e-07,
+      "loss": 0.0569,
+      "step": 13999
+    },
+    {
+      "epoch": 2.268308489954634,
+      "grad_norm": 0.9288927912712097,
+      "learning_rate": 7.402910395569357e-07,
+      "loss": 0.0608,
+      "step": 14000
+    },
+    {
+      "epoch": 2.2684705119896305,
+      "grad_norm": 0.931483268737793,
+      "learning_rate": 7.399804426640983e-07,
+      "loss": 0.0536,
+      "step": 14001
+    },
+    {
+      "epoch": 2.2686325340246274,
+      "grad_norm": 0.8563708662986755,
+      "learning_rate": 7.396698996236004e-07,
+      "loss": 0.0656,
+      "step": 14002
+    },
+    {
+      "epoch": 2.268794556059624,
+      "grad_norm": 0.9786216020584106,
+      "learning_rate": 7.393594104449445e-07,
+      "loss": 0.0657,
+      "step": 14003
+    },
+    {
+      "epoch": 2.268956578094621,
+      "grad_norm": 0.9174492955207825,
+      "learning_rate": 7.39048975137629e-07,
+      "loss": 0.0639,
+      "step": 14004
+    },
+    {
+      "epoch": 2.2691186001296177,
+      "grad_norm": 0.8567657470703125,
+      "learning_rate": 7.387385937111527e-07,
+      "loss": 0.0573,
+      "step": 14005
+    },
+    {
+      "epoch": 2.2692806221646142,
+      "grad_norm": 0.9229654669761658,
+      "learning_rate": 7.38428266175015e-07,
+      "loss": 0.0627,
+      "step": 14006
+    },
+    {
+      "epoch": 2.269442644199611,
+      "grad_norm": 0.8565087914466858,
+      "learning_rate": 7.381179925387086e-07,
+      "loss": 0.0587,
+      "step": 14007
+    },
+    {
+      "epoch": 2.269604666234608,
+      "grad_norm": 0.9805523753166199,
+      "learning_rate": 7.378077728117277e-07,
+      "loss": 0.0606,
+      "step": 14008
+    },
+    {
+      "epoch": 2.2697666882696046,
+      "grad_norm": 0.7847617268562317,
+      "learning_rate": 7.374976070035647e-07,
+      "loss": 0.0519,
+      "step": 14009
+    },
+    {
+      "epoch": 2.2699287103046015,
+      "grad_norm": 0.8503203392028809,
+      "learning_rate": 7.371874951237099e-07,
+      "loss": 0.058,
+      "step": 14010
+    },
+    {
+      "epoch": 2.270090732339598,
+      "grad_norm": 0.955046534538269,
+      "learning_rate": 7.368774371816517e-07,
+      "loss": 0.0654,
+      "step": 14011
+    },
+    {
+      "epoch": 2.270252754374595,
+      "grad_norm": 0.9594247937202454,
+      "learning_rate": 7.365674331868772e-07,
+      "loss": 0.0609,
+      "step": 14012
+    },
+    {
+      "epoch": 2.270414776409592,
+      "grad_norm": 0.8345913887023926,
+      "learning_rate": 7.36257483148872e-07,
+      "loss": 0.0617,
+      "step": 14013
+    },
+    {
+      "epoch": 2.2705767984445884,
+      "grad_norm": 0.8707557916641235,
+      "learning_rate": 7.359475870771202e-07,
+      "loss": 0.0613,
+      "step": 14014
+    },
+    {
+      "epoch": 2.2707388204795853,
+      "grad_norm": 0.9228283166885376,
+      "learning_rate": 7.356377449811017e-07,
+      "loss": 0.0598,
+      "step": 14015
+    },
+    {
+      "epoch": 2.270900842514582,
+      "grad_norm": 1.0070610046386719,
+      "learning_rate": 7.353279568702995e-07,
+      "loss": 0.0629,
+      "step": 14016
+    },
+    {
+      "epoch": 2.2710628645495787,
+      "grad_norm": 0.901984453201294,
+      "learning_rate": 7.350182227541922e-07,
+      "loss": 0.0562,
+      "step": 14017
+    },
+    {
+      "epoch": 2.2712248865845757,
+      "grad_norm": 0.9663466215133667,
+      "learning_rate": 7.347085426422551e-07,
+      "loss": 0.0641,
+      "step": 14018
+    },
+    {
+      "epoch": 2.271386908619572,
+      "grad_norm": 0.9153159856796265,
+      "learning_rate": 7.343989165439641e-07,
+      "loss": 0.0643,
+      "step": 14019
+    },
+    {
+      "epoch": 2.271548930654569,
+      "grad_norm": 0.9758960008621216,
+      "learning_rate": 7.340893444687944e-07,
+      "loss": 0.0615,
+      "step": 14020
+    },
+    {
+      "epoch": 2.271710952689566,
+      "grad_norm": 0.8344132900238037,
+      "learning_rate": 7.33779826426218e-07,
+      "loss": 0.0622,
+      "step": 14021
+    },
+    {
+      "epoch": 2.2718729747245625,
+      "grad_norm": 0.9549560546875,
+      "learning_rate": 7.334703624257039e-07,
+      "loss": 0.0642,
+      "step": 14022
+    },
+    {
+      "epoch": 2.2720349967595594,
+      "grad_norm": 0.8046680688858032,
+      "learning_rate": 7.331609524767219e-07,
+      "loss": 0.0568,
+      "step": 14023
+    },
+    {
+      "epoch": 2.272197018794556,
+      "grad_norm": 0.9434444904327393,
+      "learning_rate": 7.328515965887389e-07,
+      "loss": 0.0654,
+      "step": 14024
+    },
+    {
+      "epoch": 2.272359040829553,
+      "grad_norm": 0.8848881721496582,
+      "learning_rate": 7.32542294771221e-07,
+      "loss": 0.0582,
+      "step": 14025
+    },
+    {
+      "epoch": 2.2725210628645494,
+      "grad_norm": 0.8448486328125,
+      "learning_rate": 7.322330470336314e-07,
+      "loss": 0.0562,
+      "step": 14026
+    },
+    {
+      "epoch": 2.2726830848995463,
+      "grad_norm": 0.896988034248352,
+      "learning_rate": 7.319238533854328e-07,
+      "loss": 0.0596,
+      "step": 14027
+    },
+    {
+      "epoch": 2.2728451069345432,
+      "grad_norm": 0.843999981880188,
+      "learning_rate": 7.316147138360855e-07,
+      "loss": 0.0583,
+      "step": 14028
+    },
+    {
+      "epoch": 2.2730071289695397,
+      "grad_norm": 0.8346846699714661,
+      "learning_rate": 7.313056283950487e-07,
+      "loss": 0.0557,
+      "step": 14029
+    },
+    {
+      "epoch": 2.2731691510045366,
+      "grad_norm": 0.9750611186027527,
+      "learning_rate": 7.309965970717795e-07,
+      "loss": 0.0644,
+      "step": 14030
+    },
+    {
+      "epoch": 2.2733311730395336,
+      "grad_norm": 0.939724862575531,
+      "learning_rate": 7.306876198757332e-07,
+      "loss": 0.0625,
+      "step": 14031
+    },
+    {
+      "epoch": 2.27349319507453,
+      "grad_norm": 1.0688453912734985,
+      "learning_rate": 7.303786968163651e-07,
+      "loss": 0.073,
+      "step": 14032
+    },
+    {
+      "epoch": 2.273655217109527,
+      "grad_norm": 0.954890251159668,
+      "learning_rate": 7.300698279031257e-07,
+      "loss": 0.0579,
+      "step": 14033
+    },
+    {
+      "epoch": 2.2738172391445235,
+      "grad_norm": 1.0080914497375488,
+      "learning_rate": 7.297610131454657e-07,
+      "loss": 0.0592,
+      "step": 14034
+    },
+    {
+      "epoch": 2.2739792611795204,
+      "grad_norm": 0.9062652587890625,
+      "learning_rate": 7.294522525528355e-07,
+      "loss": 0.0622,
+      "step": 14035
+    },
+    {
+      "epoch": 2.2741412832145174,
+      "grad_norm": 1.0314079523086548,
+      "learning_rate": 7.291435461346827e-07,
+      "loss": 0.0665,
+      "step": 14036
+    },
+    {
+      "epoch": 2.274303305249514,
+      "grad_norm": 0.8445022702217102,
+      "learning_rate": 7.288348939004508e-07,
+      "loss": 0.063,
+      "step": 14037
+    },
+    {
+      "epoch": 2.274465327284511,
+      "grad_norm": 0.9952003955841064,
+      "learning_rate": 7.285262958595846e-07,
+      "loss": 0.0579,
+      "step": 14038
+    },
+    {
+      "epoch": 2.2746273493195073,
+      "grad_norm": 0.773461639881134,
+      "learning_rate": 7.282177520215283e-07,
+      "loss": 0.0542,
+      "step": 14039
+    },
+    {
+      "epoch": 2.274789371354504,
+      "grad_norm": 0.8747705221176147,
+      "learning_rate": 7.279092623957204e-07,
+      "loss": 0.06,
+      "step": 14040
+    },
+    {
+      "epoch": 2.274951393389501,
+      "grad_norm": 0.9074735045433044,
+      "learning_rate": 7.276008269916008e-07,
+      "loss": 0.0685,
+      "step": 14041
+    },
+    {
+      "epoch": 2.2751134154244976,
+      "grad_norm": 0.9272582530975342,
+      "learning_rate": 7.272924458186064e-07,
+      "loss": 0.0637,
+      "step": 14042
+    },
+    {
+      "epoch": 2.2752754374594946,
+      "grad_norm": 0.9657379984855652,
+      "learning_rate": 7.269841188861737e-07,
+      "loss": 0.0644,
+      "step": 14043
+    },
+    {
+      "epoch": 2.2754374594944915,
+      "grad_norm": 1.0053846836090088,
+      "learning_rate": 7.26675846203736e-07,
+      "loss": 0.0702,
+      "step": 14044
+    },
+    {
+      "epoch": 2.275599481529488,
+      "grad_norm": 0.7461309432983398,
+      "learning_rate": 7.263676277807263e-07,
+      "loss": 0.0544,
+      "step": 14045
+    },
+    {
+      "epoch": 2.275761503564485,
+      "grad_norm": 0.8807432055473328,
+      "learning_rate": 7.26059463626575e-07,
+      "loss": 0.063,
+      "step": 14046
+    },
+    {
+      "epoch": 2.2759235255994814,
+      "grad_norm": 0.9560192823410034,
+      "learning_rate": 7.257513537507121e-07,
+      "loss": 0.0615,
+      "step": 14047
+    },
+    {
+      "epoch": 2.2760855476344783,
+      "grad_norm": 0.7775177359580994,
+      "learning_rate": 7.254432981625626e-07,
+      "loss": 0.0573,
+      "step": 14048
+    },
+    {
+      "epoch": 2.276247569669475,
+      "grad_norm": 0.8764267563819885,
+      "learning_rate": 7.251352968715544e-07,
+      "loss": 0.0547,
+      "step": 14049
+    },
+    {
+      "epoch": 2.2764095917044718,
+      "grad_norm": 0.8146836161613464,
+      "learning_rate": 7.248273498871119e-07,
+      "loss": 0.0563,
+      "step": 14050
+    },
+    {
+      "epoch": 2.2765716137394687,
+      "grad_norm": 0.8144499659538269,
+      "learning_rate": 7.245194572186562e-07,
+      "loss": 0.0548,
+      "step": 14051
+    },
+    {
+      "epoch": 2.276733635774465,
+      "grad_norm": 0.8982124328613281,
+      "learning_rate": 7.242116188756082e-07,
+      "loss": 0.0632,
+      "step": 14052
+    },
+    {
+      "epoch": 2.276895657809462,
+      "grad_norm": 1.0810835361480713,
+      "learning_rate": 7.239038348673866e-07,
+      "loss": 0.0683,
+      "step": 14053
+    },
+    {
+      "epoch": 2.277057679844459,
+      "grad_norm": 0.9049067497253418,
+      "learning_rate": 7.235961052034113e-07,
+      "loss": 0.0659,
+      "step": 14054
+    },
+    {
+      "epoch": 2.2772197018794555,
+      "grad_norm": 0.8906067609786987,
+      "learning_rate": 7.232884298930953e-07,
+      "loss": 0.062,
+      "step": 14055
+    },
+    {
+      "epoch": 2.2773817239144525,
+      "grad_norm": 0.9106201529502869,
+      "learning_rate": 7.22980808945854e-07,
+      "loss": 0.063,
+      "step": 14056
+    },
+    {
+      "epoch": 2.277543745949449,
+      "grad_norm": 0.9221261739730835,
+      "learning_rate": 7.226732423710998e-07,
+      "loss": 0.0688,
+      "step": 14057
+    },
+    {
+      "epoch": 2.277705767984446,
+      "grad_norm": 0.8350453972816467,
+      "learning_rate": 7.22365730178243e-07,
+      "loss": 0.0599,
+      "step": 14058
+    },
+    {
+      "epoch": 2.2778677900194424,
+      "grad_norm": 0.9052740335464478,
+      "learning_rate": 7.220582723766931e-07,
+      "loss": 0.0645,
+      "step": 14059
+    },
+    {
+      "epoch": 2.2780298120544393,
+      "grad_norm": 0.8463963866233826,
+      "learning_rate": 7.217508689758576e-07,
+      "loss": 0.0665,
+      "step": 14060
+    },
+    {
+      "epoch": 2.2781918340894363,
+      "grad_norm": 0.822176992893219,
+      "learning_rate": 7.214435199851432e-07,
+      "loss": 0.0555,
+      "step": 14061
+    },
+    {
+      "epoch": 2.2783538561244328,
+      "grad_norm": 0.9079717993736267,
+      "learning_rate": 7.211362254139512e-07,
+      "loss": 0.0625,
+      "step": 14062
+    },
+    {
+      "epoch": 2.2785158781594297,
+      "grad_norm": 0.9785674214363098,
+      "learning_rate": 7.208289852716868e-07,
+      "loss": 0.0684,
+      "step": 14063
+    },
+    {
+      "epoch": 2.2786779001944266,
+      "grad_norm": 0.8325532078742981,
+      "learning_rate": 7.205217995677502e-07,
+      "loss": 0.0595,
+      "step": 14064
+    },
+    {
+      "epoch": 2.278839922229423,
+      "grad_norm": 0.9397990107536316,
+      "learning_rate": 7.202146683115408e-07,
+      "loss": 0.0605,
+      "step": 14065
+    },
+    {
+      "epoch": 2.27900194426442,
+      "grad_norm": 0.7694252729415894,
+      "learning_rate": 7.199075915124548e-07,
+      "loss": 0.0561,
+      "step": 14066
+    },
+    {
+      "epoch": 2.2791639662994165,
+      "grad_norm": 0.9818680286407471,
+      "learning_rate": 7.19600569179888e-07,
+      "loss": 0.069,
+      "step": 14067
+    },
+    {
+      "epoch": 2.2793259883344135,
+      "grad_norm": 0.8381089568138123,
+      "learning_rate": 7.192936013232368e-07,
+      "loss": 0.0555,
+      "step": 14068
+    },
+    {
+      "epoch": 2.2794880103694104,
+      "grad_norm": 1.1629765033721924,
+      "learning_rate": 7.189866879518914e-07,
+      "loss": 0.0749,
+      "step": 14069
+    },
+    {
+      "epoch": 2.279650032404407,
+      "grad_norm": 0.8009434342384338,
+      "learning_rate": 7.186798290752436e-07,
+      "loss": 0.0594,
+      "step": 14070
+    },
+    {
+      "epoch": 2.279812054439404,
+      "grad_norm": 0.85289067029953,
+      "learning_rate": 7.183730247026821e-07,
+      "loss": 0.0626,
+      "step": 14071
+    },
+    {
+      "epoch": 2.2799740764744003,
+      "grad_norm": 0.7887542247772217,
+      "learning_rate": 7.180662748435946e-07,
+      "loss": 0.0601,
+      "step": 14072
+    },
+    {
+      "epoch": 2.2801360985093972,
+      "grad_norm": 0.8790128827095032,
+      "learning_rate": 7.17759579507367e-07,
+      "loss": 0.0633,
+      "step": 14073
+    },
+    {
+      "epoch": 2.280298120544394,
+      "grad_norm": 0.8336759209632874,
+      "learning_rate": 7.174529387033832e-07,
+      "loss": 0.058,
+      "step": 14074
+    },
+    {
+      "epoch": 2.2804601425793907,
+      "grad_norm": 0.8878383040428162,
+      "learning_rate": 7.171463524410258e-07,
+      "loss": 0.0615,
+      "step": 14075
+    },
+    {
+      "epoch": 2.2806221646143876,
+      "grad_norm": 0.8392332196235657,
+      "learning_rate": 7.168398207296764e-07,
+      "loss": 0.0626,
+      "step": 14076
+    },
+    {
+      "epoch": 2.2807841866493845,
+      "grad_norm": 0.8218026161193848,
+      "learning_rate": 7.165333435787119e-07,
+      "loss": 0.0639,
+      "step": 14077
+    },
+    {
+      "epoch": 2.280946208684381,
+      "grad_norm": 1.1141018867492676,
+      "learning_rate": 7.162269209975117e-07,
+      "loss": 0.0645,
+      "step": 14078
+    },
+    {
+      "epoch": 2.281108230719378,
+      "grad_norm": 0.8250923156738281,
+      "learning_rate": 7.159205529954513e-07,
+      "loss": 0.0575,
+      "step": 14079
+    },
+    {
+      "epoch": 2.2812702527543745,
+      "grad_norm": 0.8710193037986755,
+      "learning_rate": 7.156142395819055e-07,
+      "loss": 0.0597,
+      "step": 14080
+    },
+    {
+      "epoch": 2.2814322747893714,
+      "grad_norm": 0.8891174793243408,
+      "learning_rate": 7.153079807662447e-07,
+      "loss": 0.0643,
+      "step": 14081
+    },
+    {
+      "epoch": 2.281594296824368,
+      "grad_norm": 1.0327494144439697,
+      "learning_rate": 7.150017765578401e-07,
+      "loss": 0.0676,
+      "step": 14082
+    },
+    {
+      "epoch": 2.281756318859365,
+      "grad_norm": 0.9421011805534363,
+      "learning_rate": 7.146956269660632e-07,
+      "loss": 0.0664,
+      "step": 14083
+    },
+    {
+      "epoch": 2.2819183408943617,
+      "grad_norm": 0.9506751894950867,
+      "learning_rate": 7.143895320002789e-07,
+      "loss": 0.0699,
+      "step": 14084
+    },
+    {
+      "epoch": 2.2820803629293582,
+      "grad_norm": 0.8076117038726807,
+      "learning_rate": 7.140834916698539e-07,
+      "loss": 0.0615,
+      "step": 14085
+    },
+    {
+      "epoch": 2.282242384964355,
+      "grad_norm": 0.993713915348053,
+      "learning_rate": 7.137775059841523e-07,
+      "loss": 0.0682,
+      "step": 14086
+    },
+    {
+      "epoch": 2.282404406999352,
+      "grad_norm": 0.8705652952194214,
+      "learning_rate": 7.134715749525364e-07,
+      "loss": 0.0666,
+      "step": 14087
+    },
+    {
+      "epoch": 2.2825664290343486,
+      "grad_norm": 0.9547131657600403,
+      "learning_rate": 7.131656985843669e-07,
+      "loss": 0.0683,
+      "step": 14088
+    },
+    {
+      "epoch": 2.2827284510693455,
+      "grad_norm": 1.0205283164978027,
+      "learning_rate": 7.12859876889003e-07,
+      "loss": 0.0634,
+      "step": 14089
+    },
+    {
+      "epoch": 2.282890473104342,
+      "grad_norm": 0.892135739326477,
+      "learning_rate": 7.125541098758021e-07,
+      "loss": 0.062,
+      "step": 14090
+    },
+    {
+      "epoch": 2.283052495139339,
+      "grad_norm": 0.9227086305618286,
+      "learning_rate": 7.122483975541197e-07,
+      "loss": 0.0658,
+      "step": 14091
+    },
+    {
+      "epoch": 2.283214517174336,
+      "grad_norm": 0.8980578184127808,
+      "learning_rate": 7.119427399333104e-07,
+      "loss": 0.0683,
+      "step": 14092
+    },
+    {
+      "epoch": 2.2833765392093324,
+      "grad_norm": 0.7832695245742798,
+      "learning_rate": 7.116371370227259e-07,
+      "loss": 0.054,
+      "step": 14093
+    },
+    {
+      "epoch": 2.2835385612443293,
+      "grad_norm": 1.094404935836792,
+      "learning_rate": 7.113315888317182e-07,
+      "loss": 0.0658,
+      "step": 14094
+    },
+    {
+      "epoch": 2.283700583279326,
+      "grad_norm": 1.046769380569458,
+      "learning_rate": 7.110260953696346e-07,
+      "loss": 0.0687,
+      "step": 14095
+    },
+    {
+      "epoch": 2.2838626053143227,
+      "grad_norm": 0.9280028343200684,
+      "learning_rate": 7.107206566458225e-07,
+      "loss": 0.0613,
+      "step": 14096
+    },
+    {
+      "epoch": 2.2840246273493197,
+      "grad_norm": 0.8256856203079224,
+      "learning_rate": 7.104152726696292e-07,
+      "loss": 0.0597,
+      "step": 14097
+    },
+    {
+      "epoch": 2.284186649384316,
+      "grad_norm": 0.9466334581375122,
+      "learning_rate": 7.101099434503986e-07,
+      "loss": 0.0654,
+      "step": 14098
+    },
+    {
+      "epoch": 2.284348671419313,
+      "grad_norm": 0.8240459561347961,
+      "learning_rate": 7.098046689974714e-07,
+      "loss": 0.0608,
+      "step": 14099
+    },
+    {
+      "epoch": 2.28451069345431,
+      "grad_norm": 0.8247748613357544,
+      "learning_rate": 7.09499449320189e-07,
+      "loss": 0.0593,
+      "step": 14100
+    },
+    {
+      "epoch": 2.2846727154893065,
+      "grad_norm": 0.9899136424064636,
+      "learning_rate": 7.091942844278907e-07,
+      "loss": 0.0675,
+      "step": 14101
+    },
+    {
+      "epoch": 2.2848347375243034,
+      "grad_norm": 1.1560351848602295,
+      "learning_rate": 7.088891743299136e-07,
+      "loss": 0.068,
+      "step": 14102
+    },
+    {
+      "epoch": 2.2849967595593,
+      "grad_norm": 0.8623514771461487,
+      "learning_rate": 7.085841190355932e-07,
+      "loss": 0.0599,
+      "step": 14103
+    },
+    {
+      "epoch": 2.285158781594297,
+      "grad_norm": 0.869429886341095,
+      "learning_rate": 7.08279118554264e-07,
+      "loss": 0.0593,
+      "step": 14104
+    },
+    {
+      "epoch": 2.2853208036292934,
+      "grad_norm": 1.031187891960144,
+      "learning_rate": 7.079741728952578e-07,
+      "loss": 0.0635,
+      "step": 14105
+    },
+    {
+      "epoch": 2.2854828256642903,
+      "grad_norm": 1.0560411214828491,
+      "learning_rate": 7.076692820679051e-07,
+      "loss": 0.0651,
+      "step": 14106
+    },
+    {
+      "epoch": 2.285644847699287,
+      "grad_norm": 0.8320545554161072,
+      "learning_rate": 7.073644460815348e-07,
+      "loss": 0.06,
+      "step": 14107
+    },
+    {
+      "epoch": 2.2858068697342837,
+      "grad_norm": 0.963650107383728,
+      "learning_rate": 7.070596649454748e-07,
+      "loss": 0.0651,
+      "step": 14108
+    },
+    {
+      "epoch": 2.2859688917692806,
+      "grad_norm": 0.8947854042053223,
+      "learning_rate": 7.067549386690509e-07,
+      "loss": 0.057,
+      "step": 14109
+    },
+    {
+      "epoch": 2.2861309138042776,
+      "grad_norm": 0.8852594494819641,
+      "learning_rate": 7.064502672615847e-07,
+      "loss": 0.0603,
+      "step": 14110
+    },
+    {
+      "epoch": 2.286292935839274,
+      "grad_norm": 0.9978240728378296,
+      "learning_rate": 7.061456507324008e-07,
+      "loss": 0.0671,
+      "step": 14111
+    },
+    {
+      "epoch": 2.286454957874271,
+      "grad_norm": 0.8848221302032471,
+      "learning_rate": 7.058410890908196e-07,
+      "loss": 0.0624,
+      "step": 14112
+    },
+    {
+      "epoch": 2.2866169799092675,
+      "grad_norm": 0.8844780921936035,
+      "learning_rate": 7.055365823461585e-07,
+      "loss": 0.0554,
+      "step": 14113
+    },
+    {
+      "epoch": 2.2867790019442644,
+      "grad_norm": 0.9852705001831055,
+      "learning_rate": 7.052321305077356e-07,
+      "loss": 0.0651,
+      "step": 14114
+    },
+    {
+      "epoch": 2.2869410239792614,
+      "grad_norm": 0.9480606913566589,
+      "learning_rate": 7.049277335848656e-07,
+      "loss": 0.0659,
+      "step": 14115
+    },
+    {
+      "epoch": 2.287103046014258,
+      "grad_norm": 0.9317473769187927,
+      "learning_rate": 7.046233915868642e-07,
+      "loss": 0.0626,
+      "step": 14116
+    },
+    {
+      "epoch": 2.287265068049255,
+      "grad_norm": 0.8471056222915649,
+      "learning_rate": 7.043191045230418e-07,
+      "loss": 0.0551,
+      "step": 14117
+    },
+    {
+      "epoch": 2.2874270900842513,
+      "grad_norm": 0.8562641143798828,
+      "learning_rate": 7.04014872402709e-07,
+      "loss": 0.0583,
+      "step": 14118
+    },
+    {
+      "epoch": 2.287589112119248,
+      "grad_norm": 0.9273170232772827,
+      "learning_rate": 7.037106952351752e-07,
+      "loss": 0.0621,
+      "step": 14119
+    },
+    {
+      "epoch": 2.287751134154245,
+      "grad_norm": 0.8602285981178284,
+      "learning_rate": 7.034065730297471e-07,
+      "loss": 0.0566,
+      "step": 14120
+    },
+    {
+      "epoch": 2.2879131561892416,
+      "grad_norm": 0.9283899664878845,
+      "learning_rate": 7.031025057957302e-07,
+      "loss": 0.0624,
+      "step": 14121
+    },
+    {
+      "epoch": 2.2880751782242386,
+      "grad_norm": 0.8553148508071899,
+      "learning_rate": 7.027984935424284e-07,
+      "loss": 0.064,
+      "step": 14122
+    },
+    {
+      "epoch": 2.2882372002592355,
+      "grad_norm": 0.964326798915863,
+      "learning_rate": 7.024945362791432e-07,
+      "loss": 0.0618,
+      "step": 14123
+    },
+    {
+      "epoch": 2.288399222294232,
+      "grad_norm": 0.9071498513221741,
+      "learning_rate": 7.021906340151763e-07,
+      "loss": 0.0718,
+      "step": 14124
+    },
+    {
+      "epoch": 2.288561244329229,
+      "grad_norm": 0.9609620571136475,
+      "learning_rate": 7.01886786759824e-07,
+      "loss": 0.0663,
+      "step": 14125
+    },
+    {
+      "epoch": 2.2887232663642254,
+      "grad_norm": 0.9779571890830994,
+      "learning_rate": 7.015829945223851e-07,
+      "loss": 0.065,
+      "step": 14126
+    },
+    {
+      "epoch": 2.2888852883992223,
+      "grad_norm": 1.0127842426300049,
+      "learning_rate": 7.012792573121551e-07,
+      "loss": 0.0688,
+      "step": 14127
+    },
+    {
+      "epoch": 2.289047310434219,
+      "grad_norm": 0.9270831942558289,
+      "learning_rate": 7.009755751384267e-07,
+      "loss": 0.0569,
+      "step": 14128
+    },
+    {
+      "epoch": 2.2892093324692158,
+      "grad_norm": 1.1184446811676025,
+      "learning_rate": 7.006719480104913e-07,
+      "loss": 0.0677,
+      "step": 14129
+    },
+    {
+      "epoch": 2.2893713545042127,
+      "grad_norm": 0.8415164947509766,
+      "learning_rate": 7.003683759376415e-07,
+      "loss": 0.0582,
+      "step": 14130
+    },
+    {
+      "epoch": 2.289533376539209,
+      "grad_norm": 1.0956602096557617,
+      "learning_rate": 7.000648589291634e-07,
+      "loss": 0.0626,
+      "step": 14131
+    },
+    {
+      "epoch": 2.289695398574206,
+      "grad_norm": 0.8971714973449707,
+      "learning_rate": 6.997613969943451e-07,
+      "loss": 0.0578,
+      "step": 14132
+    },
+    {
+      "epoch": 2.289857420609203,
+      "grad_norm": 0.8495174050331116,
+      "learning_rate": 6.994579901424714e-07,
+      "loss": 0.0575,
+      "step": 14133
+    },
+    {
+      "epoch": 2.2900194426441995,
+      "grad_norm": 1.226990818977356,
+      "learning_rate": 6.99154638382826e-07,
+      "loss": 0.0649,
+      "step": 14134
+    },
+    {
+      "epoch": 2.2901814646791965,
+      "grad_norm": 1.0273133516311646,
+      "learning_rate": 6.988513417246906e-07,
+      "loss": 0.0696,
+      "step": 14135
+    },
+    {
+      "epoch": 2.290343486714193,
+      "grad_norm": 0.996566116809845,
+      "learning_rate": 6.985481001773456e-07,
+      "loss": 0.0684,
+      "step": 14136
+    },
+    {
+      "epoch": 2.29050550874919,
+      "grad_norm": 0.8226109743118286,
+      "learning_rate": 6.982449137500694e-07,
+      "loss": 0.0552,
+      "step": 14137
+    },
+    {
+      "epoch": 2.290667530784187,
+      "grad_norm": 0.8552298545837402,
+      "learning_rate": 6.979417824521393e-07,
+      "loss": 0.0626,
+      "step": 14138
+    },
+    {
+      "epoch": 2.2908295528191833,
+      "grad_norm": 0.8617782592773438,
+      "learning_rate": 6.976387062928283e-07,
+      "loss": 0.0569,
+      "step": 14139
+    },
+    {
+      "epoch": 2.2909915748541803,
+      "grad_norm": 0.887869656085968,
+      "learning_rate": 6.97335685281412e-07,
+      "loss": 0.0597,
+      "step": 14140
+    },
+    {
+      "epoch": 2.2911535968891767,
+      "grad_norm": 0.8527069687843323,
+      "learning_rate": 6.970327194271612e-07,
+      "loss": 0.0682,
+      "step": 14141
+    },
+    {
+      "epoch": 2.2913156189241737,
+      "grad_norm": 0.9207058548927307,
+      "learning_rate": 6.967298087393471e-07,
+      "loss": 0.0674,
+      "step": 14142
+    },
+    {
+      "epoch": 2.2914776409591706,
+      "grad_norm": 0.8777424097061157,
+      "learning_rate": 6.964269532272361e-07,
+      "loss": 0.062,
+      "step": 14143
+    },
+    {
+      "epoch": 2.291639662994167,
+      "grad_norm": 0.9275683164596558,
+      "learning_rate": 6.96124152900095e-07,
+      "loss": 0.0591,
+      "step": 14144
+    },
+    {
+      "epoch": 2.291801685029164,
+      "grad_norm": 0.9526832699775696,
+      "learning_rate": 6.958214077671912e-07,
+      "loss": 0.0673,
+      "step": 14145
+    },
+    {
+      "epoch": 2.291963707064161,
+      "grad_norm": 1.0193464756011963,
+      "learning_rate": 6.955187178377853e-07,
+      "loss": 0.0628,
+      "step": 14146
+    },
+    {
+      "epoch": 2.2921257290991575,
+      "grad_norm": 0.9001911282539368,
+      "learning_rate": 6.952160831211401e-07,
+      "loss": 0.0649,
+      "step": 14147
+    },
+    {
+      "epoch": 2.2922877511341544,
+      "grad_norm": 0.7867611646652222,
+      "learning_rate": 6.949135036265153e-07,
+      "loss": 0.0577,
+      "step": 14148
+    },
+    {
+      "epoch": 2.292449773169151,
+      "grad_norm": 0.8737292885780334,
+      "learning_rate": 6.946109793631689e-07,
+      "loss": 0.0549,
+      "step": 14149
+    },
+    {
+      "epoch": 2.292611795204148,
+      "grad_norm": 0.9339378476142883,
+      "learning_rate": 6.943085103403577e-07,
+      "loss": 0.0673,
+      "step": 14150
+    },
+    {
+      "epoch": 2.2927738172391443,
+      "grad_norm": 0.8713966012001038,
+      "learning_rate": 6.940060965673362e-07,
+      "loss": 0.0591,
+      "step": 14151
+    },
+    {
+      "epoch": 2.2929358392741412,
+      "grad_norm": 0.8593830466270447,
+      "learning_rate": 6.937037380533579e-07,
+      "loss": 0.062,
+      "step": 14152
+    },
+    {
+      "epoch": 2.293097861309138,
+      "grad_norm": 0.8531447649002075,
+      "learning_rate": 6.93401434807674e-07,
+      "loss": 0.057,
+      "step": 14153
+    },
+    {
+      "epoch": 2.2932598833441347,
+      "grad_norm": 0.8757892847061157,
+      "learning_rate": 6.930991868395343e-07,
+      "loss": 0.0614,
+      "step": 14154
+    },
+    {
+      "epoch": 2.2934219053791316,
+      "grad_norm": 0.868497371673584,
+      "learning_rate": 6.92796994158187e-07,
+      "loss": 0.0633,
+      "step": 14155
+    },
+    {
+      "epoch": 2.2935839274141285,
+      "grad_norm": 0.9063879251480103,
+      "learning_rate": 6.924948567728787e-07,
+      "loss": 0.0596,
+      "step": 14156
+    },
+    {
+      "epoch": 2.293745949449125,
+      "grad_norm": 1.0460374355316162,
+      "learning_rate": 6.921927746928533e-07,
+      "loss": 0.0672,
+      "step": 14157
+    },
+    {
+      "epoch": 2.293907971484122,
+      "grad_norm": 0.8834719657897949,
+      "learning_rate": 6.918907479273535e-07,
+      "loss": 0.0637,
+      "step": 14158
+    },
+    {
+      "epoch": 2.2940699935191184,
+      "grad_norm": 0.8647652864456177,
+      "learning_rate": 6.915887764856216e-07,
+      "loss": 0.0618,
+      "step": 14159
+    },
+    {
+      "epoch": 2.2942320155541154,
+      "grad_norm": 0.8388234376907349,
+      "learning_rate": 6.912868603768979e-07,
+      "loss": 0.0574,
+      "step": 14160
+    },
+    {
+      "epoch": 2.2943940375891123,
+      "grad_norm": 0.8639540076255798,
+      "learning_rate": 6.909849996104187e-07,
+      "loss": 0.0632,
+      "step": 14161
+    },
+    {
+      "epoch": 2.294556059624109,
+      "grad_norm": 0.9691962599754333,
+      "learning_rate": 6.906831941954206e-07,
+      "loss": 0.0672,
+      "step": 14162
+    },
+    {
+      "epoch": 2.2947180816591057,
+      "grad_norm": 0.8707907795906067,
+      "learning_rate": 6.903814441411383e-07,
+      "loss": 0.0623,
+      "step": 14163
+    },
+    {
+      "epoch": 2.2948801036941022,
+      "grad_norm": 0.8201960325241089,
+      "learning_rate": 6.900797494568045e-07,
+      "loss": 0.0648,
+      "step": 14164
+    },
+    {
+      "epoch": 2.295042125729099,
+      "grad_norm": 0.9355173110961914,
+      "learning_rate": 6.897781101516504e-07,
+      "loss": 0.0678,
+      "step": 14165
+    },
+    {
+      "epoch": 2.295204147764096,
+      "grad_norm": 0.7753675580024719,
+      "learning_rate": 6.894765262349056e-07,
+      "loss": 0.0546,
+      "step": 14166
+    },
+    {
+      "epoch": 2.2953661697990926,
+      "grad_norm": 1.1092512607574463,
+      "learning_rate": 6.891749977157979e-07,
+      "loss": 0.0609,
+      "step": 14167
+    },
+    {
+      "epoch": 2.2955281918340895,
+      "grad_norm": 0.8517485857009888,
+      "learning_rate": 6.88873524603553e-07,
+      "loss": 0.0617,
+      "step": 14168
+    },
+    {
+      "epoch": 2.295690213869086,
+      "grad_norm": 0.8324517011642456,
+      "learning_rate": 6.885721069073953e-07,
+      "loss": 0.0565,
+      "step": 14169
+    },
+    {
+      "epoch": 2.295852235904083,
+      "grad_norm": 0.9597250819206238,
+      "learning_rate": 6.882707446365477e-07,
+      "loss": 0.0662,
+      "step": 14170
+    },
+    {
+      "epoch": 2.29601425793908,
+      "grad_norm": 0.9508599042892456,
+      "learning_rate": 6.879694378002316e-07,
+      "loss": 0.0623,
+      "step": 14171
+    },
+    {
+      "epoch": 2.2961762799740764,
+      "grad_norm": 1.0893688201904297,
+      "learning_rate": 6.876681864076646e-07,
+      "loss": 0.0757,
+      "step": 14172
+    },
+    {
+      "epoch": 2.2963383020090733,
+      "grad_norm": 0.9242887496948242,
+      "learning_rate": 6.873669904680655e-07,
+      "loss": 0.063,
+      "step": 14173
+    },
+    {
+      "epoch": 2.29650032404407,
+      "grad_norm": 0.8941336870193481,
+      "learning_rate": 6.870658499906505e-07,
+      "loss": 0.0595,
+      "step": 14174
+    },
+    {
+      "epoch": 2.2966623460790667,
+      "grad_norm": 0.8984914422035217,
+      "learning_rate": 6.867647649846338e-07,
+      "loss": 0.0615,
+      "step": 14175
+    },
+    {
+      "epoch": 2.2968243681140637,
+      "grad_norm": 0.9848359823226929,
+      "learning_rate": 6.864637354592266e-07,
+      "loss": 0.0634,
+      "step": 14176
+    },
+    {
+      "epoch": 2.29698639014906,
+      "grad_norm": 0.9562865495681763,
+      "learning_rate": 6.861627614236396e-07,
+      "loss": 0.0661,
+      "step": 14177
+    },
+    {
+      "epoch": 2.297148412184057,
+      "grad_norm": 0.8696370124816895,
+      "learning_rate": 6.858618428870842e-07,
+      "loss": 0.0591,
+      "step": 14178
+    },
+    {
+      "epoch": 2.297310434219054,
+      "grad_norm": 0.9466931819915771,
+      "learning_rate": 6.855609798587656e-07,
+      "loss": 0.0605,
+      "step": 14179
+    },
+    {
+      "epoch": 2.2974724562540505,
+      "grad_norm": 0.826383650302887,
+      "learning_rate": 6.852601723478902e-07,
+      "loss": 0.0515,
+      "step": 14180
+    },
+    {
+      "epoch": 2.2976344782890474,
+      "grad_norm": 0.9467977285385132,
+      "learning_rate": 6.849594203636619e-07,
+      "loss": 0.0621,
+      "step": 14181
+    },
+    {
+      "epoch": 2.297796500324044,
+      "grad_norm": 0.9184082746505737,
+      "learning_rate": 6.84658723915283e-07,
+      "loss": 0.0634,
+      "step": 14182
+    },
+    {
+      "epoch": 2.297958522359041,
+      "grad_norm": 0.8161764740943909,
+      "learning_rate": 6.843580830119542e-07,
+      "loss": 0.0573,
+      "step": 14183
+    },
+    {
+      "epoch": 2.2981205443940373,
+      "grad_norm": 0.8612385392189026,
+      "learning_rate": 6.840574976628741e-07,
+      "loss": 0.0674,
+      "step": 14184
+    },
+    {
+      "epoch": 2.2982825664290343,
+      "grad_norm": 0.8990957140922546,
+      "learning_rate": 6.837569678772401e-07,
+      "loss": 0.0606,
+      "step": 14185
+    },
+    {
+      "epoch": 2.298444588464031,
+      "grad_norm": 0.809965193271637,
+      "learning_rate": 6.834564936642488e-07,
+      "loss": 0.0644,
+      "step": 14186
+    },
+    {
+      "epoch": 2.2986066104990277,
+      "grad_norm": 0.9587274193763733,
+      "learning_rate": 6.831560750330909e-07,
+      "loss": 0.0612,
+      "step": 14187
+    },
+    {
+      "epoch": 2.2987686325340246,
+      "grad_norm": 0.9014949202537537,
+      "learning_rate": 6.828557119929613e-07,
+      "loss": 0.0591,
+      "step": 14188
+    },
+    {
+      "epoch": 2.2989306545690216,
+      "grad_norm": 1.0473854541778564,
+      "learning_rate": 6.825554045530502e-07,
+      "loss": 0.0749,
+      "step": 14189
+    },
+    {
+      "epoch": 2.299092676604018,
+      "grad_norm": 0.9291993379592896,
+      "learning_rate": 6.822551527225452e-07,
+      "loss": 0.0603,
+      "step": 14190
+    },
+    {
+      "epoch": 2.299254698639015,
+      "grad_norm": 0.8897577524185181,
+      "learning_rate": 6.819549565106331e-07,
+      "loss": 0.056,
+      "step": 14191
+    },
+    {
+      "epoch": 2.2994167206740115,
+      "grad_norm": 0.8365473747253418,
+      "learning_rate": 6.816548159264993e-07,
+      "loss": 0.0583,
+      "step": 14192
+    },
+    {
+      "epoch": 2.2995787427090084,
+      "grad_norm": 0.8146736025810242,
+      "learning_rate": 6.813547309793295e-07,
+      "loss": 0.0569,
+      "step": 14193
+    },
+    {
+      "epoch": 2.2997407647440054,
+      "grad_norm": 1.0734623670578003,
+      "learning_rate": 6.810547016783029e-07,
+      "loss": 0.0584,
+      "step": 14194
+    },
+    {
+      "epoch": 2.299902786779002,
+      "grad_norm": 0.953218936920166,
+      "learning_rate": 6.807547280326007e-07,
+      "loss": 0.0679,
+      "step": 14195
+    },
+    {
+      "epoch": 2.3000648088139988,
+      "grad_norm": 1.0048823356628418,
+      "learning_rate": 6.804548100514013e-07,
+      "loss": 0.0612,
+      "step": 14196
+    },
+    {
+      "epoch": 2.3002268308489953,
+      "grad_norm": 1.0320632457733154,
+      "learning_rate": 6.801549477438815e-07,
+      "loss": 0.0585,
+      "step": 14197
+    },
+    {
+      "epoch": 2.300388852883992,
+      "grad_norm": 0.9690525531768799,
+      "learning_rate": 6.798551411192165e-07,
+      "loss": 0.0676,
+      "step": 14198
+    },
+    {
+      "epoch": 2.300550874918989,
+      "grad_norm": 0.8694241046905518,
+      "learning_rate": 6.795553901865795e-07,
+      "loss": 0.0623,
+      "step": 14199
+    },
+    {
+      "epoch": 2.3007128969539856,
+      "grad_norm": 0.9456620216369629,
+      "learning_rate": 6.792556949551426e-07,
+      "loss": 0.0662,
+      "step": 14200
+    },
+    {
+      "epoch": 2.3008749189889826,
+      "grad_norm": 0.9079062342643738,
+      "learning_rate": 6.789560554340743e-07,
+      "loss": 0.0629,
+      "step": 14201
+    },
+    {
+      "epoch": 2.3010369410239795,
+      "grad_norm": 0.9178680181503296,
+      "learning_rate": 6.786564716325441e-07,
+      "loss": 0.0721,
+      "step": 14202
+    },
+    {
+      "epoch": 2.301198963058976,
+      "grad_norm": 0.7588401436805725,
+      "learning_rate": 6.783569435597188e-07,
+      "loss": 0.0523,
+      "step": 14203
+    },
+    {
+      "epoch": 2.301360985093973,
+      "grad_norm": 0.9008545875549316,
+      "learning_rate": 6.780574712247632e-07,
+      "loss": 0.0673,
+      "step": 14204
+    },
+    {
+      "epoch": 2.3015230071289694,
+      "grad_norm": 0.960690438747406,
+      "learning_rate": 6.777580546368393e-07,
+      "loss": 0.0612,
+      "step": 14205
+    },
+    {
+      "epoch": 2.3016850291639663,
+      "grad_norm": 0.9084741473197937,
+      "learning_rate": 6.774586938051084e-07,
+      "loss": 0.0618,
+      "step": 14206
+    },
+    {
+      "epoch": 2.301847051198963,
+      "grad_norm": 1.0799715518951416,
+      "learning_rate": 6.771593887387326e-07,
+      "loss": 0.0649,
+      "step": 14207
+    },
+    {
+      "epoch": 2.3020090732339598,
+      "grad_norm": 0.854361355304718,
+      "learning_rate": 6.768601394468674e-07,
+      "loss": 0.06,
+      "step": 14208
+    },
+    {
+      "epoch": 2.3021710952689567,
+      "grad_norm": 0.9102272987365723,
+      "learning_rate": 6.765609459386702e-07,
+      "loss": 0.0681,
+      "step": 14209
+    },
+    {
+      "epoch": 2.302333117303953,
+      "grad_norm": 0.9026054739952087,
+      "learning_rate": 6.762618082232952e-07,
+      "loss": 0.0594,
+      "step": 14210
+    },
+    {
+      "epoch": 2.30249513933895,
+      "grad_norm": 0.9153203964233398,
+      "learning_rate": 6.759627263098955e-07,
+      "loss": 0.0695,
+      "step": 14211
+    },
+    {
+      "epoch": 2.302657161373947,
+      "grad_norm": 0.8652878403663635,
+      "learning_rate": 6.756637002076225e-07,
+      "loss": 0.0624,
+      "step": 14212
+    },
+    {
+      "epoch": 2.3028191834089435,
+      "grad_norm": 0.8761528730392456,
+      "learning_rate": 6.753647299256255e-07,
+      "loss": 0.0601,
+      "step": 14213
+    },
+    {
+      "epoch": 2.3029812054439405,
+      "grad_norm": 0.8554459810256958,
+      "learning_rate": 6.750658154730522e-07,
+      "loss": 0.0597,
+      "step": 14214
+    },
+    {
+      "epoch": 2.303143227478937,
+      "grad_norm": 0.8388009071350098,
+      "learning_rate": 6.747669568590492e-07,
+      "loss": 0.062,
+      "step": 14215
+    },
+    {
+      "epoch": 2.303305249513934,
+      "grad_norm": 0.928914487361908,
+      "learning_rate": 6.744681540927588e-07,
+      "loss": 0.0618,
+      "step": 14216
+    },
+    {
+      "epoch": 2.303467271548931,
+      "grad_norm": 0.8588629961013794,
+      "learning_rate": 6.741694071833263e-07,
+      "loss": 0.0574,
+      "step": 14217
+    },
+    {
+      "epoch": 2.3036292935839273,
+      "grad_norm": 0.9985714554786682,
+      "learning_rate": 6.738707161398914e-07,
+      "loss": 0.0705,
+      "step": 14218
+    },
+    {
+      "epoch": 2.3037913156189243,
+      "grad_norm": 0.9504481554031372,
+      "learning_rate": 6.73572080971594e-07,
+      "loss": 0.0678,
+      "step": 14219
+    },
+    {
+      "epoch": 2.3039533376539207,
+      "grad_norm": 0.9666747450828552,
+      "learning_rate": 6.732735016875697e-07,
+      "loss": 0.0676,
+      "step": 14220
+    },
+    {
+      "epoch": 2.3041153596889177,
+      "grad_norm": 0.8083941340446472,
+      "learning_rate": 6.729749782969563e-07,
+      "loss": 0.0612,
+      "step": 14221
+    },
+    {
+      "epoch": 2.3042773817239146,
+      "grad_norm": 0.9107043743133545,
+      "learning_rate": 6.726765108088881e-07,
+      "loss": 0.0617,
+      "step": 14222
+    },
+    {
+      "epoch": 2.304439403758911,
+      "grad_norm": 0.8032419681549072,
+      "learning_rate": 6.72378099232496e-07,
+      "loss": 0.0553,
+      "step": 14223
+    },
+    {
+      "epoch": 2.304601425793908,
+      "grad_norm": 0.9696297645568848,
+      "learning_rate": 6.720797435769111e-07,
+      "loss": 0.0601,
+      "step": 14224
+    },
+    {
+      "epoch": 2.304763447828905,
+      "grad_norm": 0.9108163118362427,
+      "learning_rate": 6.717814438512626e-07,
+      "loss": 0.0624,
+      "step": 14225
+    },
+    {
+      "epoch": 2.3049254698639015,
+      "grad_norm": 0.8767720460891724,
+      "learning_rate": 6.714832000646778e-07,
+      "loss": 0.0662,
+      "step": 14226
+    },
+    {
+      "epoch": 2.3050874918988984,
+      "grad_norm": 0.9097752571105957,
+      "learning_rate": 6.711850122262823e-07,
+      "loss": 0.0645,
+      "step": 14227
+    },
+    {
+      "epoch": 2.305249513933895,
+      "grad_norm": 0.8309717178344727,
+      "learning_rate": 6.708868803451992e-07,
+      "loss": 0.052,
+      "step": 14228
+    },
+    {
+      "epoch": 2.305411535968892,
+      "grad_norm": 0.8942591547966003,
+      "learning_rate": 6.705888044305516e-07,
+      "loss": 0.0665,
+      "step": 14229
+    },
+    {
+      "epoch": 2.3055735580038883,
+      "grad_norm": 0.7907509207725525,
+      "learning_rate": 6.702907844914597e-07,
+      "loss": 0.0545,
+      "step": 14230
+    },
+    {
+      "epoch": 2.3057355800388852,
+      "grad_norm": 0.913838267326355,
+      "learning_rate": 6.699928205370418e-07,
+      "loss": 0.0648,
+      "step": 14231
+    },
+    {
+      "epoch": 2.305897602073882,
+      "grad_norm": 0.7393428087234497,
+      "learning_rate": 6.696949125764149e-07,
+      "loss": 0.0531,
+      "step": 14232
+    },
+    {
+      "epoch": 2.3060596241088787,
+      "grad_norm": 1.152111291885376,
+      "learning_rate": 6.693970606186953e-07,
+      "loss": 0.0706,
+      "step": 14233
+    },
+    {
+      "epoch": 2.3062216461438756,
+      "grad_norm": 0.8411812782287598,
+      "learning_rate": 6.690992646729949e-07,
+      "loss": 0.0603,
+      "step": 14234
+    },
+    {
+      "epoch": 2.3063836681788725,
+      "grad_norm": 0.9782255291938782,
+      "learning_rate": 6.688015247484256e-07,
+      "loss": 0.0644,
+      "step": 14235
+    },
+    {
+      "epoch": 2.306545690213869,
+      "grad_norm": 0.9612037539482117,
+      "learning_rate": 6.685038408540989e-07,
+      "loss": 0.0636,
+      "step": 14236
+    },
+    {
+      "epoch": 2.306707712248866,
+      "grad_norm": 1.0477268695831299,
+      "learning_rate": 6.682062129991232e-07,
+      "loss": 0.07,
+      "step": 14237
+    },
+    {
+      "epoch": 2.3068697342838624,
+      "grad_norm": 0.8569392561912537,
+      "learning_rate": 6.679086411926039e-07,
+      "loss": 0.0622,
+      "step": 14238
+    },
+    {
+      "epoch": 2.3070317563188594,
+      "grad_norm": 0.8216943144798279,
+      "learning_rate": 6.676111254436465e-07,
+      "loss": 0.0564,
+      "step": 14239
+    },
+    {
+      "epoch": 2.3071937783538563,
+      "grad_norm": 0.8884442448616028,
+      "learning_rate": 6.673136657613547e-07,
+      "loss": 0.0602,
+      "step": 14240
+    },
+    {
+      "epoch": 2.307355800388853,
+      "grad_norm": 0.9046189785003662,
+      "learning_rate": 6.670162621548293e-07,
+      "loss": 0.0612,
+      "step": 14241
+    },
+    {
+      "epoch": 2.3075178224238497,
+      "grad_norm": 0.9811939597129822,
+      "learning_rate": 6.667189146331707e-07,
+      "loss": 0.0644,
+      "step": 14242
+    },
+    {
+      "epoch": 2.307679844458846,
+      "grad_norm": 0.9112708568572998,
+      "learning_rate": 6.664216232054771e-07,
+      "loss": 0.0598,
+      "step": 14243
+    },
+    {
+      "epoch": 2.307841866493843,
+      "grad_norm": 1.0917285680770874,
+      "learning_rate": 6.661243878808443e-07,
+      "loss": 0.0707,
+      "step": 14244
+    },
+    {
+      "epoch": 2.30800388852884,
+      "grad_norm": 0.8850948810577393,
+      "learning_rate": 6.658272086683676e-07,
+      "loss": 0.0545,
+      "step": 14245
+    },
+    {
+      "epoch": 2.3081659105638366,
+      "grad_norm": 0.8096259832382202,
+      "learning_rate": 6.655300855771393e-07,
+      "loss": 0.0557,
+      "step": 14246
+    },
+    {
+      "epoch": 2.3083279325988335,
+      "grad_norm": 0.9947896599769592,
+      "learning_rate": 6.652330186162514e-07,
+      "loss": 0.0618,
+      "step": 14247
+    },
+    {
+      "epoch": 2.3084899546338304,
+      "grad_norm": 0.7889495491981506,
+      "learning_rate": 6.649360077947939e-07,
+      "loss": 0.0539,
+      "step": 14248
+    },
+    {
+      "epoch": 2.308651976668827,
+      "grad_norm": 0.9984772801399231,
+      "learning_rate": 6.646390531218522e-07,
+      "loss": 0.0655,
+      "step": 14249
+    },
+    {
+      "epoch": 2.308813998703824,
+      "grad_norm": 0.9736108183860779,
+      "learning_rate": 6.643421546065146e-07,
+      "loss": 0.0626,
+      "step": 14250
+    },
+    {
+      "epoch": 2.3089760207388204,
+      "grad_norm": 0.9100701808929443,
+      "learning_rate": 6.640453122578655e-07,
+      "loss": 0.0652,
+      "step": 14251
+    },
+    {
+      "epoch": 2.3091380427738173,
+      "grad_norm": 1.022388219833374,
+      "learning_rate": 6.637485260849866e-07,
+      "loss": 0.0645,
+      "step": 14252
+    },
+    {
+      "epoch": 2.309300064808814,
+      "grad_norm": 0.9333613514900208,
+      "learning_rate": 6.634517960969588e-07,
+      "loss": 0.0595,
+      "step": 14253
+    },
+    {
+      "epoch": 2.3094620868438107,
+      "grad_norm": 0.8503563404083252,
+      "learning_rate": 6.63155122302861e-07,
+      "loss": 0.053,
+      "step": 14254
+    },
+    {
+      "epoch": 2.3096241088788076,
+      "grad_norm": 1.0957640409469604,
+      "learning_rate": 6.628585047117731e-07,
+      "loss": 0.063,
+      "step": 14255
+    },
+    {
+      "epoch": 2.309786130913804,
+      "grad_norm": 1.1101269721984863,
+      "learning_rate": 6.625619433327681e-07,
+      "loss": 0.0574,
+      "step": 14256
+    },
+    {
+      "epoch": 2.309948152948801,
+      "grad_norm": 1.0195109844207764,
+      "learning_rate": 6.622654381749213e-07,
+      "loss": 0.0655,
+      "step": 14257
+    },
+    {
+      "epoch": 2.310110174983798,
+      "grad_norm": 0.9084102511405945,
+      "learning_rate": 6.619689892473046e-07,
+      "loss": 0.0594,
+      "step": 14258
+    },
+    {
+      "epoch": 2.3102721970187945,
+      "grad_norm": 0.9905080795288086,
+      "learning_rate": 6.616725965589893e-07,
+      "loss": 0.0613,
+      "step": 14259
+    },
+    {
+      "epoch": 2.3104342190537914,
+      "grad_norm": 0.8357704281806946,
+      "learning_rate": 6.613762601190435e-07,
+      "loss": 0.0619,
+      "step": 14260
+    },
+    {
+      "epoch": 2.310596241088788,
+      "grad_norm": 0.9966726899147034,
+      "learning_rate": 6.61079979936535e-07,
+      "loss": 0.0677,
+      "step": 14261
+    },
+    {
+      "epoch": 2.310758263123785,
+      "grad_norm": 1.088291049003601,
+      "learning_rate": 6.60783756020529e-07,
+      "loss": 0.0691,
+      "step": 14262
+    },
+    {
+      "epoch": 2.310920285158782,
+      "grad_norm": 0.8392730951309204,
+      "learning_rate": 6.60487588380089e-07,
+      "loss": 0.0599,
+      "step": 14263
+    },
+    {
+      "epoch": 2.3110823071937783,
+      "grad_norm": 0.935095489025116,
+      "learning_rate": 6.601914770242776e-07,
+      "loss": 0.0603,
+      "step": 14264
+    },
+    {
+      "epoch": 2.311244329228775,
+      "grad_norm": 0.8228254318237305,
+      "learning_rate": 6.598954219621545e-07,
+      "loss": 0.0618,
+      "step": 14265
+    },
+    {
+      "epoch": 2.3114063512637717,
+      "grad_norm": 0.9754266142845154,
+      "learning_rate": 6.595994232027794e-07,
+      "loss": 0.0648,
+      "step": 14266
+    },
+    {
+      "epoch": 2.3115683732987686,
+      "grad_norm": 0.8244330286979675,
+      "learning_rate": 6.593034807552076e-07,
+      "loss": 0.051,
+      "step": 14267
+    },
+    {
+      "epoch": 2.3117303953337656,
+      "grad_norm": 0.829617440700531,
+      "learning_rate": 6.590075946284941e-07,
+      "loss": 0.0569,
+      "step": 14268
+    },
+    {
+      "epoch": 2.311892417368762,
+      "grad_norm": 1.0251657962799072,
+      "learning_rate": 6.587117648316943e-07,
+      "loss": 0.0588,
+      "step": 14269
+    },
+    {
+      "epoch": 2.312054439403759,
+      "grad_norm": 0.8598863482475281,
+      "learning_rate": 6.584159913738583e-07,
+      "loss": 0.0531,
+      "step": 14270
+    },
+    {
+      "epoch": 2.3122164614387555,
+      "grad_norm": 1.17449951171875,
+      "learning_rate": 6.581202742640361e-07,
+      "loss": 0.0737,
+      "step": 14271
+    },
+    {
+      "epoch": 2.3123784834737524,
+      "grad_norm": 0.9494041800498962,
+      "learning_rate": 6.578246135112765e-07,
+      "loss": 0.0646,
+      "step": 14272
+    },
+    {
+      "epoch": 2.3125405055087493,
+      "grad_norm": 0.9004831910133362,
+      "learning_rate": 6.575290091246256e-07,
+      "loss": 0.0553,
+      "step": 14273
+    },
+    {
+      "epoch": 2.312702527543746,
+      "grad_norm": 0.8012595772743225,
+      "learning_rate": 6.572334611131284e-07,
+      "loss": 0.0556,
+      "step": 14274
+    },
+    {
+      "epoch": 2.3128645495787428,
+      "grad_norm": 0.8960158228874207,
+      "learning_rate": 6.569379694858277e-07,
+      "loss": 0.0645,
+      "step": 14275
+    },
+    {
+      "epoch": 2.3130265716137393,
+      "grad_norm": 1.01628577709198,
+      "learning_rate": 6.566425342517652e-07,
+      "loss": 0.0653,
+      "step": 14276
+    },
+    {
+      "epoch": 2.313188593648736,
+      "grad_norm": 0.8243879079818726,
+      "learning_rate": 6.56347155419981e-07,
+      "loss": 0.0575,
+      "step": 14277
+    },
+    {
+      "epoch": 2.313350615683733,
+      "grad_norm": 0.9484763741493225,
+      "learning_rate": 6.560518329995108e-07,
+      "loss": 0.0646,
+      "step": 14278
+    },
+    {
+      "epoch": 2.3135126377187296,
+      "grad_norm": 1.1005399227142334,
+      "learning_rate": 6.557565669993931e-07,
+      "loss": 0.068,
+      "step": 14279
+    },
+    {
+      "epoch": 2.3136746597537265,
+      "grad_norm": 0.9087778925895691,
+      "learning_rate": 6.554613574286614e-07,
+      "loss": 0.0637,
+      "step": 14280
+    },
+    {
+      "epoch": 2.3138366817887235,
+      "grad_norm": 0.8552685976028442,
+      "learning_rate": 6.551662042963491e-07,
+      "loss": 0.0602,
+      "step": 14281
+    },
+    {
+      "epoch": 2.31399870382372,
+      "grad_norm": 0.8670865297317505,
+      "learning_rate": 6.548711076114858e-07,
+      "loss": 0.0605,
+      "step": 14282
+    },
+    {
+      "epoch": 2.314160725858717,
+      "grad_norm": 1.124015212059021,
+      "learning_rate": 6.545760673831009e-07,
+      "loss": 0.0686,
+      "step": 14283
+    },
+    {
+      "epoch": 2.3143227478937134,
+      "grad_norm": 1.0868921279907227,
+      "learning_rate": 6.542810836202237e-07,
+      "loss": 0.0644,
+      "step": 14284
+    },
+    {
+      "epoch": 2.3144847699287103,
+      "grad_norm": 0.8501294255256653,
+      "learning_rate": 6.539861563318784e-07,
+      "loss": 0.0641,
+      "step": 14285
+    },
+    {
+      "epoch": 2.314646791963707,
+      "grad_norm": 0.9735018610954285,
+      "learning_rate": 6.536912855270894e-07,
+      "loss": 0.0572,
+      "step": 14286
+    },
+    {
+      "epoch": 2.3148088139987038,
+      "grad_norm": 1.222994089126587,
+      "learning_rate": 6.533964712148779e-07,
+      "loss": 0.0736,
+      "step": 14287
+    },
+    {
+      "epoch": 2.3149708360337007,
+      "grad_norm": 0.9061808586120605,
+      "learning_rate": 6.531017134042678e-07,
+      "loss": 0.0617,
+      "step": 14288
+    },
+    {
+      "epoch": 2.315132858068697,
+      "grad_norm": 0.8229312896728516,
+      "learning_rate": 6.528070121042746e-07,
+      "loss": 0.0584,
+      "step": 14289
+    },
+    {
+      "epoch": 2.315294880103694,
+      "grad_norm": 0.8917800784111023,
+      "learning_rate": 6.52512367323917e-07,
+      "loss": 0.0568,
+      "step": 14290
+    },
+    {
+      "epoch": 2.315456902138691,
+      "grad_norm": 0.9813393950462341,
+      "learning_rate": 6.522177790722101e-07,
+      "loss": 0.0597,
+      "step": 14291
+    },
+    {
+      "epoch": 2.3156189241736875,
+      "grad_norm": 0.951609194278717,
+      "learning_rate": 6.519232473581675e-07,
+      "loss": 0.054,
+      "step": 14292
+    },
+    {
+      "epoch": 2.3157809462086845,
+      "grad_norm": 0.9440679550170898,
+      "learning_rate": 6.516287721908013e-07,
+      "loss": 0.0684,
+      "step": 14293
+    },
+    {
+      "epoch": 2.315942968243681,
+      "grad_norm": 0.8250246644020081,
+      "learning_rate": 6.513343535791216e-07,
+      "loss": 0.0576,
+      "step": 14294
+    },
+    {
+      "epoch": 2.316104990278678,
+      "grad_norm": 0.9275084733963013,
+      "learning_rate": 6.510399915321381e-07,
+      "loss": 0.0622,
+      "step": 14295
+    },
+    {
+      "epoch": 2.316267012313675,
+      "grad_norm": 1.0191025733947754,
+      "learning_rate": 6.507456860588554e-07,
+      "loss": 0.0587,
+      "step": 14296
+    },
+    {
+      "epoch": 2.3164290343486713,
+      "grad_norm": 1.1428616046905518,
+      "learning_rate": 6.504514371682788e-07,
+      "loss": 0.0684,
+      "step": 14297
+    },
+    {
+      "epoch": 2.3165910563836682,
+      "grad_norm": 0.9954843521118164,
+      "learning_rate": 6.501572448694135e-07,
+      "loss": 0.0539,
+      "step": 14298
+    },
+    {
+      "epoch": 2.3167530784186647,
+      "grad_norm": 0.9105631113052368,
+      "learning_rate": 6.498631091712603e-07,
+      "loss": 0.0609,
+      "step": 14299
+    },
+    {
+      "epoch": 2.3169151004536617,
+      "grad_norm": 0.7692684531211853,
+      "learning_rate": 6.495690300828183e-07,
+      "loss": 0.056,
+      "step": 14300
+    },
+    {
+      "epoch": 2.3170771224886586,
+      "grad_norm": 0.9106401205062866,
+      "learning_rate": 6.492750076130858e-07,
+      "loss": 0.0609,
+      "step": 14301
+    },
+    {
+      "epoch": 2.317239144523655,
+      "grad_norm": 1.094725489616394,
+      "learning_rate": 6.489810417710596e-07,
+      "loss": 0.0642,
+      "step": 14302
+    },
+    {
+      "epoch": 2.317401166558652,
+      "grad_norm": 0.7931134104728699,
+      "learning_rate": 6.48687132565734e-07,
+      "loss": 0.0578,
+      "step": 14303
+    },
+    {
+      "epoch": 2.317563188593649,
+      "grad_norm": 0.9079036116600037,
+      "learning_rate": 6.483932800061021e-07,
+      "loss": 0.0586,
+      "step": 14304
+    },
+    {
+      "epoch": 2.3177252106286454,
+      "grad_norm": 0.8913233280181885,
+      "learning_rate": 6.480994841011551e-07,
+      "loss": 0.0562,
+      "step": 14305
+    },
+    {
+      "epoch": 2.3178872326636424,
+      "grad_norm": 0.9209250211715698,
+      "learning_rate": 6.478057448598821e-07,
+      "loss": 0.0622,
+      "step": 14306
+    },
+    {
+      "epoch": 2.318049254698639,
+      "grad_norm": 0.994839608669281,
+      "learning_rate": 6.475120622912714e-07,
+      "loss": 0.0689,
+      "step": 14307
+    },
+    {
+      "epoch": 2.318211276733636,
+      "grad_norm": 0.9868736863136292,
+      "learning_rate": 6.472184364043085e-07,
+      "loss": 0.0632,
+      "step": 14308
+    },
+    {
+      "epoch": 2.3183732987686323,
+      "grad_norm": 0.9076941013336182,
+      "learning_rate": 6.469248672079778e-07,
+      "loss": 0.0614,
+      "step": 14309
+    },
+    {
+      "epoch": 2.3185353208036292,
+      "grad_norm": 0.9764859080314636,
+      "learning_rate": 6.466313547112627e-07,
+      "loss": 0.0699,
+      "step": 14310
+    },
+    {
+      "epoch": 2.318697342838626,
+      "grad_norm": 1.0119270086288452,
+      "learning_rate": 6.463378989231414e-07,
+      "loss": 0.0575,
+      "step": 14311
+    },
+    {
+      "epoch": 2.3188593648736227,
+      "grad_norm": 0.8330714106559753,
+      "learning_rate": 6.460444998525953e-07,
+      "loss": 0.0593,
+      "step": 14312
+    },
+    {
+      "epoch": 2.3190213869086196,
+      "grad_norm": 0.9974405765533447,
+      "learning_rate": 6.45751157508602e-07,
+      "loss": 0.0687,
+      "step": 14313
+    },
+    {
+      "epoch": 2.3191834089436165,
+      "grad_norm": 0.9587564468383789,
+      "learning_rate": 6.454578719001353e-07,
+      "loss": 0.0632,
+      "step": 14314
+    },
+    {
+      "epoch": 2.319345430978613,
+      "grad_norm": 0.8796777129173279,
+      "learning_rate": 6.451646430361696e-07,
+      "loss": 0.0626,
+      "step": 14315
+    },
+    {
+      "epoch": 2.31950745301361,
+      "grad_norm": 0.9801618456840515,
+      "learning_rate": 6.448714709256768e-07,
+      "loss": 0.0687,
+      "step": 14316
+    },
+    {
+      "epoch": 2.3196694750486064,
+      "grad_norm": 1.0900496244430542,
+      "learning_rate": 6.445783555776289e-07,
+      "loss": 0.0667,
+      "step": 14317
+    },
+    {
+      "epoch": 2.3198314970836034,
+      "grad_norm": 0.8728010654449463,
+      "learning_rate": 6.442852970009925e-07,
+      "loss": 0.0648,
+      "step": 14318
+    },
+    {
+      "epoch": 2.3199935191186003,
+      "grad_norm": 1.05231511592865,
+      "learning_rate": 6.439922952047354e-07,
+      "loss": 0.0703,
+      "step": 14319
+    },
+    {
+      "epoch": 2.320155541153597,
+      "grad_norm": 1.0301355123519897,
+      "learning_rate": 6.436993501978226e-07,
+      "loss": 0.0609,
+      "step": 14320
+    },
+    {
+      "epoch": 2.3203175631885937,
+      "grad_norm": 1.086633563041687,
+      "learning_rate": 6.434064619892175e-07,
+      "loss": 0.0576,
+      "step": 14321
+    },
+    {
+      "epoch": 2.32047958522359,
+      "grad_norm": 0.8054091930389404,
+      "learning_rate": 6.431136305878819e-07,
+      "loss": 0.061,
+      "step": 14322
+    },
+    {
+      "epoch": 2.320641607258587,
+      "grad_norm": 0.8511539697647095,
+      "learning_rate": 6.428208560027755e-07,
+      "loss": 0.0574,
+      "step": 14323
+    },
+    {
+      "epoch": 2.320803629293584,
+      "grad_norm": 1.0268864631652832,
+      "learning_rate": 6.425281382428566e-07,
+      "loss": 0.0602,
+      "step": 14324
+    },
+    {
+      "epoch": 2.3209656513285806,
+      "grad_norm": 0.941626787185669,
+      "learning_rate": 6.422354773170825e-07,
+      "loss": 0.0707,
+      "step": 14325
+    },
+    {
+      "epoch": 2.3211276733635775,
+      "grad_norm": 0.8575886487960815,
+      "learning_rate": 6.419428732344055e-07,
+      "loss": 0.0642,
+      "step": 14326
+    },
+    {
+      "epoch": 2.3212896953985744,
+      "grad_norm": 0.7996649146080017,
+      "learning_rate": 6.41650326003781e-07,
+      "loss": 0.0594,
+      "step": 14327
+    },
+    {
+      "epoch": 2.321451717433571,
+      "grad_norm": 0.8261630535125732,
+      "learning_rate": 6.413578356341602e-07,
+      "loss": 0.061,
+      "step": 14328
+    },
+    {
+      "epoch": 2.321613739468568,
+      "grad_norm": 0.9302612543106079,
+      "learning_rate": 6.410654021344909e-07,
+      "loss": 0.0629,
+      "step": 14329
+    },
+    {
+      "epoch": 2.3217757615035644,
+      "grad_norm": 0.91783207654953,
+      "learning_rate": 6.407730255137212e-07,
+      "loss": 0.0618,
+      "step": 14330
+    },
+    {
+      "epoch": 2.3219377835385613,
+      "grad_norm": 0.9778722524642944,
+      "learning_rate": 6.404807057807982e-07,
+      "loss": 0.0656,
+      "step": 14331
+    },
+    {
+      "epoch": 2.3220998055735578,
+      "grad_norm": 0.9770204424858093,
+      "learning_rate": 6.401884429446667e-07,
+      "loss": 0.0747,
+      "step": 14332
+    },
+    {
+      "epoch": 2.3222618276085547,
+      "grad_norm": 0.974606990814209,
+      "learning_rate": 6.398962370142672e-07,
+      "loss": 0.0708,
+      "step": 14333
+    },
+    {
+      "epoch": 2.3224238496435516,
+      "grad_norm": 0.8390527963638306,
+      "learning_rate": 6.396040879985416e-07,
+      "loss": 0.0654,
+      "step": 14334
+    },
+    {
+      "epoch": 2.322585871678548,
+      "grad_norm": 0.8269057869911194,
+      "learning_rate": 6.393119959064287e-07,
+      "loss": 0.0601,
+      "step": 14335
+    },
+    {
+      "epoch": 2.322747893713545,
+      "grad_norm": 0.9386374354362488,
+      "learning_rate": 6.390199607468661e-07,
+      "loss": 0.0661,
+      "step": 14336
+    },
+    {
+      "epoch": 2.322909915748542,
+      "grad_norm": 0.795021116733551,
+      "learning_rate": 6.387279825287892e-07,
+      "loss": 0.0568,
+      "step": 14337
+    },
+    {
+      "epoch": 2.3230719377835385,
+      "grad_norm": 0.8094752430915833,
+      "learning_rate": 6.384360612611317e-07,
+      "loss": 0.0615,
+      "step": 14338
+    },
+    {
+      "epoch": 2.3232339598185354,
+      "grad_norm": 0.8009088039398193,
+      "learning_rate": 6.381441969528268e-07,
+      "loss": 0.0567,
+      "step": 14339
+    },
+    {
+      "epoch": 2.323395981853532,
+      "grad_norm": 1.1142011880874634,
+      "learning_rate": 6.378523896128022e-07,
+      "loss": 0.0741,
+      "step": 14340
+    },
+    {
+      "epoch": 2.323558003888529,
+      "grad_norm": 0.8346922397613525,
+      "learning_rate": 6.37560639249989e-07,
+      "loss": 0.0549,
+      "step": 14341
+    },
+    {
+      "epoch": 2.323720025923526,
+      "grad_norm": 0.9480348825454712,
+      "learning_rate": 6.37268945873313e-07,
+      "loss": 0.0684,
+      "step": 14342
+    },
+    {
+      "epoch": 2.3238820479585223,
+      "grad_norm": 0.8845011591911316,
+      "learning_rate": 6.369773094917006e-07,
+      "loss": 0.0602,
+      "step": 14343
+    },
+    {
+      "epoch": 2.324044069993519,
+      "grad_norm": 0.7797037363052368,
+      "learning_rate": 6.36685730114073e-07,
+      "loss": 0.0554,
+      "step": 14344
+    },
+    {
+      "epoch": 2.3242060920285157,
+      "grad_norm": 0.9391816854476929,
+      "learning_rate": 6.363942077493526e-07,
+      "loss": 0.0599,
+      "step": 14345
+    },
+    {
+      "epoch": 2.3243681140635126,
+      "grad_norm": 0.895332396030426,
+      "learning_rate": 6.361027424064609e-07,
+      "loss": 0.0646,
+      "step": 14346
+    },
+    {
+      "epoch": 2.3245301360985096,
+      "grad_norm": 1.2685229778289795,
+      "learning_rate": 6.35811334094314e-07,
+      "loss": 0.0661,
+      "step": 14347
+    },
+    {
+      "epoch": 2.324692158133506,
+      "grad_norm": 0.9253360033035278,
+      "learning_rate": 6.355199828218289e-07,
+      "loss": 0.0673,
+      "step": 14348
+    },
+    {
+      "epoch": 2.324854180168503,
+      "grad_norm": 0.8045247197151184,
+      "learning_rate": 6.352286885979206e-07,
+      "loss": 0.0547,
+      "step": 14349
+    },
+    {
+      "epoch": 2.3250162022035,
+      "grad_norm": 0.9229259490966797,
+      "learning_rate": 6.349374514315015e-07,
+      "loss": 0.0606,
+      "step": 14350
+    },
+    {
+      "epoch": 2.3251782242384964,
+      "grad_norm": 0.9514033794403076,
+      "learning_rate": 6.346462713314832e-07,
+      "loss": 0.0599,
+      "step": 14351
+    },
+    {
+      "epoch": 2.3253402462734933,
+      "grad_norm": 0.9198815822601318,
+      "learning_rate": 6.343551483067751e-07,
+      "loss": 0.0639,
+      "step": 14352
+    },
+    {
+      "epoch": 2.32550226830849,
+      "grad_norm": 0.8940281271934509,
+      "learning_rate": 6.340640823662842e-07,
+      "loss": 0.059,
+      "step": 14353
+    },
+    {
+      "epoch": 2.3256642903434868,
+      "grad_norm": 0.9115906953811646,
+      "learning_rate": 6.337730735189174e-07,
+      "loss": 0.0633,
+      "step": 14354
+    },
+    {
+      "epoch": 2.3258263123784833,
+      "grad_norm": 0.9600539803504944,
+      "learning_rate": 6.334821217735778e-07,
+      "loss": 0.0585,
+      "step": 14355
+    },
+    {
+      "epoch": 2.32598833441348,
+      "grad_norm": 0.9461213946342468,
+      "learning_rate": 6.331912271391688e-07,
+      "loss": 0.0607,
+      "step": 14356
+    },
+    {
+      "epoch": 2.326150356448477,
+      "grad_norm": 0.9921082258224487,
+      "learning_rate": 6.329003896245908e-07,
+      "loss": 0.0614,
+      "step": 14357
+    },
+    {
+      "epoch": 2.3263123784834736,
+      "grad_norm": 0.8889320492744446,
+      "learning_rate": 6.326096092387429e-07,
+      "loss": 0.0627,
+      "step": 14358
+    },
+    {
+      "epoch": 2.3264744005184705,
+      "grad_norm": 0.9381389021873474,
+      "learning_rate": 6.323188859905207e-07,
+      "loss": 0.0633,
+      "step": 14359
+    },
+    {
+      "epoch": 2.3266364225534675,
+      "grad_norm": 0.850811779499054,
+      "learning_rate": 6.320282198888217e-07,
+      "loss": 0.0619,
+      "step": 14360
+    },
+    {
+      "epoch": 2.326798444588464,
+      "grad_norm": 0.831915557384491,
+      "learning_rate": 6.317376109425397e-07,
+      "loss": 0.0546,
+      "step": 14361
+    },
+    {
+      "epoch": 2.326960466623461,
+      "grad_norm": 0.8602516651153564,
+      "learning_rate": 6.314470591605646e-07,
+      "loss": 0.0624,
+      "step": 14362
+    },
+    {
+      "epoch": 2.3271224886584574,
+      "grad_norm": 0.9104455709457397,
+      "learning_rate": 6.311565645517878e-07,
+      "loss": 0.0652,
+      "step": 14363
+    },
+    {
+      "epoch": 2.3272845106934543,
+      "grad_norm": 1.1242001056671143,
+      "learning_rate": 6.308661271250974e-07,
+      "loss": 0.0677,
+      "step": 14364
+    },
+    {
+      "epoch": 2.3274465327284513,
+      "grad_norm": 0.8856641054153442,
+      "learning_rate": 6.305757468893805e-07,
+      "loss": 0.0649,
+      "step": 14365
+    },
+    {
+      "epoch": 2.3276085547634477,
+      "grad_norm": 0.8960265517234802,
+      "learning_rate": 6.302854238535219e-07,
+      "loss": 0.0605,
+      "step": 14366
+    },
+    {
+      "epoch": 2.3277705767984447,
+      "grad_norm": 0.9673840999603271,
+      "learning_rate": 6.299951580264047e-07,
+      "loss": 0.0579,
+      "step": 14367
+    },
+    {
+      "epoch": 2.327932598833441,
+      "grad_norm": 1.0338740348815918,
+      "learning_rate": 6.2970494941691e-07,
+      "loss": 0.0713,
+      "step": 14368
+    },
+    {
+      "epoch": 2.328094620868438,
+      "grad_norm": 1.090754508972168,
+      "learning_rate": 6.294147980339182e-07,
+      "loss": 0.0671,
+      "step": 14369
+    },
+    {
+      "epoch": 2.328256642903435,
+      "grad_norm": 0.877884566783905,
+      "learning_rate": 6.291247038863066e-07,
+      "loss": 0.0598,
+      "step": 14370
+    },
+    {
+      "epoch": 2.3284186649384315,
+      "grad_norm": 0.8679099678993225,
+      "learning_rate": 6.288346669829518e-07,
+      "loss": 0.0611,
+      "step": 14371
+    },
+    {
+      "epoch": 2.3285806869734285,
+      "grad_norm": 0.972556471824646,
+      "learning_rate": 6.285446873327289e-07,
+      "loss": 0.0667,
+      "step": 14372
+    },
+    {
+      "epoch": 2.3287427090084254,
+      "grad_norm": 0.8393288254737854,
+      "learning_rate": 6.282547649445087e-07,
+      "loss": 0.0602,
+      "step": 14373
+    },
+    {
+      "epoch": 2.328904731043422,
+      "grad_norm": 0.8273356556892395,
+      "learning_rate": 6.279648998271626e-07,
+      "loss": 0.0599,
+      "step": 14374
+    },
+    {
+      "epoch": 2.329066753078419,
+      "grad_norm": 0.8478513360023499,
+      "learning_rate": 6.276750919895611e-07,
+      "loss": 0.0557,
+      "step": 14375
+    },
+    {
+      "epoch": 2.3292287751134153,
+      "grad_norm": 0.9557338953018188,
+      "learning_rate": 6.273853414405715e-07,
+      "loss": 0.0706,
+      "step": 14376
+    },
+    {
+      "epoch": 2.3293907971484122,
+      "grad_norm": 0.9526261687278748,
+      "learning_rate": 6.270956481890581e-07,
+      "loss": 0.0676,
+      "step": 14377
+    },
+    {
+      "epoch": 2.3295528191834087,
+      "grad_norm": 0.8927558660507202,
+      "learning_rate": 6.268060122438846e-07,
+      "loss": 0.0581,
+      "step": 14378
+    },
+    {
+      "epoch": 2.3297148412184057,
+      "grad_norm": 0.9341113567352295,
+      "learning_rate": 6.265164336139157e-07,
+      "loss": 0.0619,
+      "step": 14379
+    },
+    {
+      "epoch": 2.3298768632534026,
+      "grad_norm": 0.8526856303215027,
+      "learning_rate": 6.262269123080095e-07,
+      "loss": 0.06,
+      "step": 14380
+    },
+    {
+      "epoch": 2.330038885288399,
+      "grad_norm": 1.0245894193649292,
+      "learning_rate": 6.259374483350253e-07,
+      "loss": 0.0641,
+      "step": 14381
+    },
+    {
+      "epoch": 2.330200907323396,
+      "grad_norm": 0.8189849853515625,
+      "learning_rate": 6.256480417038202e-07,
+      "loss": 0.0572,
+      "step": 14382
+    },
+    {
+      "epoch": 2.330362929358393,
+      "grad_norm": 0.9356175661087036,
+      "learning_rate": 6.253586924232489e-07,
+      "loss": 0.0652,
+      "step": 14383
+    },
+    {
+      "epoch": 2.3305249513933894,
+      "grad_norm": 0.9660431742668152,
+      "learning_rate": 6.250694005021651e-07,
+      "loss": 0.0656,
+      "step": 14384
+    },
+    {
+      "epoch": 2.3306869734283864,
+      "grad_norm": 0.7970089316368103,
+      "learning_rate": 6.247801659494207e-07,
+      "loss": 0.0563,
+      "step": 14385
+    },
+    {
+      "epoch": 2.330848995463383,
+      "grad_norm": 0.9310674071311951,
+      "learning_rate": 6.244909887738651e-07,
+      "loss": 0.069,
+      "step": 14386
+    },
+    {
+      "epoch": 2.33101101749838,
+      "grad_norm": 1.0296223163604736,
+      "learning_rate": 6.242018689843471e-07,
+      "loss": 0.07,
+      "step": 14387
+    },
+    {
+      "epoch": 2.3311730395333763,
+      "grad_norm": 0.875980019569397,
+      "learning_rate": 6.239128065897113e-07,
+      "loss": 0.0623,
+      "step": 14388
+    },
+    {
+      "epoch": 2.3313350615683732,
+      "grad_norm": 0.9242268204689026,
+      "learning_rate": 6.236238015988044e-07,
+      "loss": 0.0655,
+      "step": 14389
+    },
+    {
+      "epoch": 2.33149708360337,
+      "grad_norm": 0.8916738629341125,
+      "learning_rate": 6.233348540204689e-07,
+      "loss": 0.0632,
+      "step": 14390
+    },
+    {
+      "epoch": 2.3316591056383666,
+      "grad_norm": 0.9092872142791748,
+      "learning_rate": 6.23045963863545e-07,
+      "loss": 0.0652,
+      "step": 14391
+    },
+    {
+      "epoch": 2.3318211276733636,
+      "grad_norm": 0.8567991852760315,
+      "learning_rate": 6.227571311368724e-07,
+      "loss": 0.0694,
+      "step": 14392
+    },
+    {
+      "epoch": 2.3319831497083605,
+      "grad_norm": 0.921323299407959,
+      "learning_rate": 6.22468355849288e-07,
+      "loss": 0.0666,
+      "step": 14393
+    },
+    {
+      "epoch": 2.332145171743357,
+      "grad_norm": 0.9615211486816406,
+      "learning_rate": 6.221796380096298e-07,
+      "loss": 0.0626,
+      "step": 14394
+    },
+    {
+      "epoch": 2.332307193778354,
+      "grad_norm": 0.9106801748275757,
+      "learning_rate": 6.218909776267295e-07,
+      "loss": 0.0597,
+      "step": 14395
+    },
+    {
+      "epoch": 2.3324692158133504,
+      "grad_norm": 0.896317183971405,
+      "learning_rate": 6.216023747094207e-07,
+      "loss": 0.06,
+      "step": 14396
+    },
+    {
+      "epoch": 2.3326312378483474,
+      "grad_norm": 0.9247400760650635,
+      "learning_rate": 6.213138292665333e-07,
+      "loss": 0.0679,
+      "step": 14397
+    },
+    {
+      "epoch": 2.3327932598833443,
+      "grad_norm": 0.9041847586631775,
+      "learning_rate": 6.210253413068964e-07,
+      "loss": 0.0586,
+      "step": 14398
+    },
+    {
+      "epoch": 2.332955281918341,
+      "grad_norm": 0.840080201625824,
+      "learning_rate": 6.207369108393374e-07,
+      "loss": 0.0549,
+      "step": 14399
+    },
+    {
+      "epoch": 2.3331173039533377,
+      "grad_norm": 0.8104625344276428,
+      "learning_rate": 6.20448537872681e-07,
+      "loss": 0.0578,
+      "step": 14400
+    },
+    {
+      "epoch": 2.333279325988334,
+      "grad_norm": 0.9511587023735046,
+      "learning_rate": 6.201602224157508e-07,
+      "loss": 0.0697,
+      "step": 14401
+    },
+    {
+      "epoch": 2.333441348023331,
+      "grad_norm": 1.060603380203247,
+      "learning_rate": 6.198719644773687e-07,
+      "loss": 0.0735,
+      "step": 14402
+    },
+    {
+      "epoch": 2.333603370058328,
+      "grad_norm": 0.9461984634399414,
+      "learning_rate": 6.195837640663546e-07,
+      "loss": 0.06,
+      "step": 14403
+    },
+    {
+      "epoch": 2.3337653920933246,
+      "grad_norm": 0.7724205851554871,
+      "learning_rate": 6.192956211915269e-07,
+      "loss": 0.0539,
+      "step": 14404
+    },
+    {
+      "epoch": 2.3339274141283215,
+      "grad_norm": 0.7702530026435852,
+      "learning_rate": 6.190075358617029e-07,
+      "loss": 0.0553,
+      "step": 14405
+    },
+    {
+      "epoch": 2.3340894361633184,
+      "grad_norm": 1.1382404565811157,
+      "learning_rate": 6.187195080856953e-07,
+      "loss": 0.0647,
+      "step": 14406
+    },
+    {
+      "epoch": 2.334251458198315,
+      "grad_norm": 0.8377068042755127,
+      "learning_rate": 6.184315378723177e-07,
+      "loss": 0.0581,
+      "step": 14407
+    },
+    {
+      "epoch": 2.334413480233312,
+      "grad_norm": 1.097277283668518,
+      "learning_rate": 6.181436252303829e-07,
+      "loss": 0.0576,
+      "step": 14408
+    },
+    {
+      "epoch": 2.3345755022683083,
+      "grad_norm": 1.051430106163025,
+      "learning_rate": 6.178557701686985e-07,
+      "loss": 0.0641,
+      "step": 14409
+    },
+    {
+      "epoch": 2.3347375243033053,
+      "grad_norm": 0.9613043665885925,
+      "learning_rate": 6.175679726960731e-07,
+      "loss": 0.0686,
+      "step": 14410
+    },
+    {
+      "epoch": 2.3348995463383018,
+      "grad_norm": 0.791201114654541,
+      "learning_rate": 6.17280232821312e-07,
+      "loss": 0.0547,
+      "step": 14411
+    },
+    {
+      "epoch": 2.3350615683732987,
+      "grad_norm": 1.0079299211502075,
+      "learning_rate": 6.169925505532201e-07,
+      "loss": 0.0659,
+      "step": 14412
+    },
+    {
+      "epoch": 2.3352235904082956,
+      "grad_norm": 1.0087671279907227,
+      "learning_rate": 6.167049259005989e-07,
+      "loss": 0.0701,
+      "step": 14413
+    },
+    {
+      "epoch": 2.335385612443292,
+      "grad_norm": 0.9747778177261353,
+      "learning_rate": 6.164173588722497e-07,
+      "loss": 0.0582,
+      "step": 14414
+    },
+    {
+      "epoch": 2.335547634478289,
+      "grad_norm": 0.862362265586853,
+      "learning_rate": 6.161298494769713e-07,
+      "loss": 0.0548,
+      "step": 14415
+    },
+    {
+      "epoch": 2.335709656513286,
+      "grad_norm": 1.0373525619506836,
+      "learning_rate": 6.158423977235611e-07,
+      "loss": 0.0639,
+      "step": 14416
+    },
+    {
+      "epoch": 2.3358716785482825,
+      "grad_norm": 0.9192182421684265,
+      "learning_rate": 6.155550036208125e-07,
+      "loss": 0.066,
+      "step": 14417
+    },
+    {
+      "epoch": 2.3360337005832794,
+      "grad_norm": 0.8472986221313477,
+      "learning_rate": 6.152676671775215e-07,
+      "loss": 0.056,
+      "step": 14418
+    },
+    {
+      "epoch": 2.336195722618276,
+      "grad_norm": 1.020595669746399,
+      "learning_rate": 6.149803884024786e-07,
+      "loss": 0.0643,
+      "step": 14419
+    },
+    {
+      "epoch": 2.336357744653273,
+      "grad_norm": 0.877204954624176,
+      "learning_rate": 6.146931673044751e-07,
+      "loss": 0.0563,
+      "step": 14420
+    },
+    {
+      "epoch": 2.3365197666882698,
+      "grad_norm": 0.8729059100151062,
+      "learning_rate": 6.144060038922967e-07,
+      "loss": 0.0586,
+      "step": 14421
+    },
+    {
+      "epoch": 2.3366817887232663,
+      "grad_norm": 1.002275824546814,
+      "learning_rate": 6.141188981747323e-07,
+      "loss": 0.0644,
+      "step": 14422
+    },
+    {
+      "epoch": 2.336843810758263,
+      "grad_norm": 0.8643121123313904,
+      "learning_rate": 6.138318501605667e-07,
+      "loss": 0.0591,
+      "step": 14423
+    },
+    {
+      "epoch": 2.3370058327932597,
+      "grad_norm": 0.9158600568771362,
+      "learning_rate": 6.135448598585814e-07,
+      "loss": 0.0647,
+      "step": 14424
+    },
+    {
+      "epoch": 2.3371678548282566,
+      "grad_norm": 0.9756930470466614,
+      "learning_rate": 6.132579272775583e-07,
+      "loss": 0.0646,
+      "step": 14425
+    },
+    {
+      "epoch": 2.3373298768632536,
+      "grad_norm": 0.9086965322494507,
+      "learning_rate": 6.129710524262758e-07,
+      "loss": 0.0622,
+      "step": 14426
+    },
+    {
+      "epoch": 2.33749189889825,
+      "grad_norm": 0.8095213174819946,
+      "learning_rate": 6.12684235313514e-07,
+      "loss": 0.0585,
+      "step": 14427
+    },
+    {
+      "epoch": 2.337653920933247,
+      "grad_norm": 0.8328830599784851,
+      "learning_rate": 6.123974759480469e-07,
+      "loss": 0.0553,
+      "step": 14428
+    },
+    {
+      "epoch": 2.337815942968244,
+      "grad_norm": 0.9773808121681213,
+      "learning_rate": 6.12110774338649e-07,
+      "loss": 0.0597,
+      "step": 14429
+    },
+    {
+      "epoch": 2.3379779650032404,
+      "grad_norm": 1.039530634880066,
+      "learning_rate": 6.118241304940928e-07,
+      "loss": 0.0648,
+      "step": 14430
+    },
+    {
+      "epoch": 2.3381399870382373,
+      "grad_norm": 0.8209445476531982,
+      "learning_rate": 6.115375444231489e-07,
+      "loss": 0.0472,
+      "step": 14431
+    },
+    {
+      "epoch": 2.338302009073234,
+      "grad_norm": 0.8805145025253296,
+      "learning_rate": 6.112510161345861e-07,
+      "loss": 0.0629,
+      "step": 14432
+    },
+    {
+      "epoch": 2.3384640311082308,
+      "grad_norm": 1.0358737707138062,
+      "learning_rate": 6.109645456371715e-07,
+      "loss": 0.0585,
+      "step": 14433
+    },
+    {
+      "epoch": 2.3386260531432272,
+      "grad_norm": 0.8988499641418457,
+      "learning_rate": 6.106781329396714e-07,
+      "loss": 0.0542,
+      "step": 14434
+    },
+    {
+      "epoch": 2.338788075178224,
+      "grad_norm": 0.8143104910850525,
+      "learning_rate": 6.103917780508475e-07,
+      "loss": 0.0535,
+      "step": 14435
+    },
+    {
+      "epoch": 2.338950097213221,
+      "grad_norm": 0.8741886019706726,
+      "learning_rate": 6.101054809794615e-07,
+      "loss": 0.0623,
+      "step": 14436
+    },
+    {
+      "epoch": 2.3391121192482176,
+      "grad_norm": 0.9815524816513062,
+      "learning_rate": 6.098192417342755e-07,
+      "loss": 0.0693,
+      "step": 14437
+    },
+    {
+      "epoch": 2.3392741412832145,
+      "grad_norm": 0.9720104336738586,
+      "learning_rate": 6.095330603240468e-07,
+      "loss": 0.0597,
+      "step": 14438
+    },
+    {
+      "epoch": 2.3394361633182115,
+      "grad_norm": 0.9975051879882812,
+      "learning_rate": 6.092469367575312e-07,
+      "loss": 0.0565,
+      "step": 14439
+    },
+    {
+      "epoch": 2.339598185353208,
+      "grad_norm": 1.0102384090423584,
+      "learning_rate": 6.089608710434836e-07,
+      "loss": 0.0639,
+      "step": 14440
+    },
+    {
+      "epoch": 2.339760207388205,
+      "grad_norm": 1.2839295864105225,
+      "learning_rate": 6.086748631906572e-07,
+      "loss": 0.0529,
+      "step": 14441
+    },
+    {
+      "epoch": 2.3399222294232014,
+      "grad_norm": 0.770176112651825,
+      "learning_rate": 6.083889132078033e-07,
+      "loss": 0.0553,
+      "step": 14442
+    },
+    {
+      "epoch": 2.3400842514581983,
+      "grad_norm": 0.8635346293449402,
+      "learning_rate": 6.081030211036707e-07,
+      "loss": 0.0545,
+      "step": 14443
+    },
+    {
+      "epoch": 2.3402462734931953,
+      "grad_norm": 0.951804518699646,
+      "learning_rate": 6.078171868870075e-07,
+      "loss": 0.0615,
+      "step": 14444
+    },
+    {
+      "epoch": 2.3404082955281917,
+      "grad_norm": 0.8703580498695374,
+      "learning_rate": 6.075314105665595e-07,
+      "loss": 0.0609,
+      "step": 14445
+    },
+    {
+      "epoch": 2.3405703175631887,
+      "grad_norm": 0.9203516840934753,
+      "learning_rate": 6.072456921510703e-07,
+      "loss": 0.0607,
+      "step": 14446
+    },
+    {
+      "epoch": 2.340732339598185,
+      "grad_norm": 0.8416757583618164,
+      "learning_rate": 6.06960031649283e-07,
+      "loss": 0.0551,
+      "step": 14447
+    },
+    {
+      "epoch": 2.340894361633182,
+      "grad_norm": 1.4582421779632568,
+      "learning_rate": 6.066744290699372e-07,
+      "loss": 0.0577,
+      "step": 14448
+    },
+    {
+      "epoch": 2.341056383668179,
+      "grad_norm": 1.0675171613693237,
+      "learning_rate": 6.06388884421773e-07,
+      "loss": 0.0636,
+      "step": 14449
+    },
+    {
+      "epoch": 2.3412184057031755,
+      "grad_norm": 1.3599189519882202,
+      "learning_rate": 6.061033977135253e-07,
+      "loss": 0.0655,
+      "step": 14450
+    },
+    {
+      "epoch": 2.3413804277381725,
+      "grad_norm": 0.8678320050239563,
+      "learning_rate": 6.058179689539309e-07,
+      "loss": 0.0632,
+      "step": 14451
+    },
+    {
+      "epoch": 2.3415424497731694,
+      "grad_norm": 0.7486401200294495,
+      "learning_rate": 6.055325981517238e-07,
+      "loss": 0.0604,
+      "step": 14452
+    },
+    {
+      "epoch": 2.341704471808166,
+      "grad_norm": 0.9437369108200073,
+      "learning_rate": 6.052472853156339e-07,
+      "loss": 0.066,
+      "step": 14453
+    },
+    {
+      "epoch": 2.341866493843163,
+      "grad_norm": 0.8997758030891418,
+      "learning_rate": 6.049620304543916e-07,
+      "loss": 0.0664,
+      "step": 14454
+    },
+    {
+      "epoch": 2.3420285158781593,
+      "grad_norm": 0.826973557472229,
+      "learning_rate": 6.046768335767248e-07,
+      "loss": 0.0587,
+      "step": 14455
+    },
+    {
+      "epoch": 2.3421905379131562,
+      "grad_norm": 0.9194096326828003,
+      "learning_rate": 6.043916946913613e-07,
+      "loss": 0.061,
+      "step": 14456
+    },
+    {
+      "epoch": 2.3423525599481527,
+      "grad_norm": 1.0393024682998657,
+      "learning_rate": 6.041066138070245e-07,
+      "loss": 0.0642,
+      "step": 14457
+    },
+    {
+      "epoch": 2.3425145819831497,
+      "grad_norm": 1.180245041847229,
+      "learning_rate": 6.038215909324372e-07,
+      "loss": 0.0696,
+      "step": 14458
+    },
+    {
+      "epoch": 2.3426766040181466,
+      "grad_norm": 0.9270623326301575,
+      "learning_rate": 6.035366260763203e-07,
+      "loss": 0.0644,
+      "step": 14459
+    },
+    {
+      "epoch": 2.342838626053143,
+      "grad_norm": 0.8007288575172424,
+      "learning_rate": 6.032517192473935e-07,
+      "loss": 0.0534,
+      "step": 14460
+    },
+    {
+      "epoch": 2.34300064808814,
+      "grad_norm": 0.85854172706604,
+      "learning_rate": 6.02966870454374e-07,
+      "loss": 0.0591,
+      "step": 14461
+    },
+    {
+      "epoch": 2.343162670123137,
+      "grad_norm": 0.8734108805656433,
+      "learning_rate": 6.026820797059777e-07,
+      "loss": 0.0645,
+      "step": 14462
+    },
+    {
+      "epoch": 2.3433246921581334,
+      "grad_norm": 0.9076907634735107,
+      "learning_rate": 6.023973470109182e-07,
+      "loss": 0.0626,
+      "step": 14463
+    },
+    {
+      "epoch": 2.3434867141931304,
+      "grad_norm": 0.7974627614021301,
+      "learning_rate": 6.021126723779075e-07,
+      "loss": 0.0583,
+      "step": 14464
+    },
+    {
+      "epoch": 2.343648736228127,
+      "grad_norm": 0.9349268078804016,
+      "learning_rate": 6.018280558156566e-07,
+      "loss": 0.0679,
+      "step": 14465
+    },
+    {
+      "epoch": 2.343810758263124,
+      "grad_norm": 0.9849607348442078,
+      "learning_rate": 6.015434973328735e-07,
+      "loss": 0.0693,
+      "step": 14466
+    },
+    {
+      "epoch": 2.3439727802981207,
+      "grad_norm": 0.8764670491218567,
+      "learning_rate": 6.012589969382659e-07,
+      "loss": 0.0605,
+      "step": 14467
+    },
+    {
+      "epoch": 2.344134802333117,
+      "grad_norm": 0.8385621905326843,
+      "learning_rate": 6.009745546405377e-07,
+      "loss": 0.0605,
+      "step": 14468
+    },
+    {
+      "epoch": 2.344296824368114,
+      "grad_norm": 0.8304133415222168,
+      "learning_rate": 6.006901704483917e-07,
+      "loss": 0.0562,
+      "step": 14469
+    },
+    {
+      "epoch": 2.3444588464031106,
+      "grad_norm": 0.8149062395095825,
+      "learning_rate": 6.00405844370531e-07,
+      "loss": 0.0544,
+      "step": 14470
+    },
+    {
+      "epoch": 2.3446208684381076,
+      "grad_norm": 0.8023223280906677,
+      "learning_rate": 6.001215764156551e-07,
+      "loss": 0.057,
+      "step": 14471
+    },
+    {
+      "epoch": 2.3447828904731045,
+      "grad_norm": 0.7750385403633118,
+      "learning_rate": 5.998373665924606e-07,
+      "loss": 0.0537,
+      "step": 14472
+    },
+    {
+      "epoch": 2.344944912508101,
+      "grad_norm": 0.9637815952301025,
+      "learning_rate": 5.995532149096447e-07,
+      "loss": 0.066,
+      "step": 14473
+    },
+    {
+      "epoch": 2.345106934543098,
+      "grad_norm": 0.8537014722824097,
+      "learning_rate": 5.992691213759011e-07,
+      "loss": 0.0598,
+      "step": 14474
+    },
+    {
+      "epoch": 2.345268956578095,
+      "grad_norm": 0.8019862174987793,
+      "learning_rate": 5.989850859999227e-07,
+      "loss": 0.0572,
+      "step": 14475
+    },
+    {
+      "epoch": 2.3454309786130914,
+      "grad_norm": 0.8479922413825989,
+      "learning_rate": 5.987011087904007e-07,
+      "loss": 0.0562,
+      "step": 14476
+    },
+    {
+      "epoch": 2.3455930006480883,
+      "grad_norm": 0.8056397438049316,
+      "learning_rate": 5.984171897560234e-07,
+      "loss": 0.0596,
+      "step": 14477
+    },
+    {
+      "epoch": 2.345755022683085,
+      "grad_norm": 0.8855328559875488,
+      "learning_rate": 5.981333289054792e-07,
+      "loss": 0.0617,
+      "step": 14478
+    },
+    {
+      "epoch": 2.3459170447180817,
+      "grad_norm": 0.8664817810058594,
+      "learning_rate": 5.978495262474509e-07,
+      "loss": 0.0606,
+      "step": 14479
+    },
+    {
+      "epoch": 2.346079066753078,
+      "grad_norm": 0.8697784543037415,
+      "learning_rate": 5.975657817906253e-07,
+      "loss": 0.0604,
+      "step": 14480
+    },
+    {
+      "epoch": 2.346241088788075,
+      "grad_norm": 0.860231339931488,
+      "learning_rate": 5.972820955436825e-07,
+      "loss": 0.062,
+      "step": 14481
+    },
+    {
+      "epoch": 2.346403110823072,
+      "grad_norm": 0.8649452328681946,
+      "learning_rate": 5.96998467515304e-07,
+      "loss": 0.0594,
+      "step": 14482
+    },
+    {
+      "epoch": 2.3465651328580686,
+      "grad_norm": 0.9872585535049438,
+      "learning_rate": 5.967148977141665e-07,
+      "loss": 0.0711,
+      "step": 14483
+    },
+    {
+      "epoch": 2.3467271548930655,
+      "grad_norm": 0.8951234817504883,
+      "learning_rate": 5.964313861489466e-07,
+      "loss": 0.0692,
+      "step": 14484
+    },
+    {
+      "epoch": 2.3468891769280624,
+      "grad_norm": 0.8037145137786865,
+      "learning_rate": 5.96147932828321e-07,
+      "loss": 0.0561,
+      "step": 14485
+    },
+    {
+      "epoch": 2.347051198963059,
+      "grad_norm": 0.9256864786148071,
+      "learning_rate": 5.958645377609606e-07,
+      "loss": 0.0604,
+      "step": 14486
+    },
+    {
+      "epoch": 2.347213220998056,
+      "grad_norm": 0.9966773390769958,
+      "learning_rate": 5.955812009555378e-07,
+      "loss": 0.0661,
+      "step": 14487
+    },
+    {
+      "epoch": 2.3473752430330523,
+      "grad_norm": 0.930876612663269,
+      "learning_rate": 5.952979224207205e-07,
+      "loss": 0.0658,
+      "step": 14488
+    },
+    {
+      "epoch": 2.3475372650680493,
+      "grad_norm": 0.9100360870361328,
+      "learning_rate": 5.950147021651792e-07,
+      "loss": 0.0675,
+      "step": 14489
+    },
+    {
+      "epoch": 2.347699287103046,
+      "grad_norm": 0.8395872116088867,
+      "learning_rate": 5.947315401975773e-07,
+      "loss": 0.0549,
+      "step": 14490
+    },
+    {
+      "epoch": 2.3478613091380427,
+      "grad_norm": 0.807367205619812,
+      "learning_rate": 5.944484365265795e-07,
+      "loss": 0.0558,
+      "step": 14491
+    },
+    {
+      "epoch": 2.3480233311730396,
+      "grad_norm": 0.902279257774353,
+      "learning_rate": 5.941653911608486e-07,
+      "loss": 0.0609,
+      "step": 14492
+    },
+    {
+      "epoch": 2.348185353208036,
+      "grad_norm": 0.7829273343086243,
+      "learning_rate": 5.938824041090443e-07,
+      "loss": 0.0568,
+      "step": 14493
+    },
+    {
+      "epoch": 2.348347375243033,
+      "grad_norm": 1.0976831912994385,
+      "learning_rate": 5.935994753798258e-07,
+      "loss": 0.0659,
+      "step": 14494
+    },
+    {
+      "epoch": 2.34850939727803,
+      "grad_norm": 0.8173691630363464,
+      "learning_rate": 5.933166049818501e-07,
+      "loss": 0.0597,
+      "step": 14495
+    },
+    {
+      "epoch": 2.3486714193130265,
+      "grad_norm": 0.7735475897789001,
+      "learning_rate": 5.930337929237726e-07,
+      "loss": 0.0606,
+      "step": 14496
+    },
+    {
+      "epoch": 2.3488334413480234,
+      "grad_norm": 0.9407158493995667,
+      "learning_rate": 5.927510392142458e-07,
+      "loss": 0.0622,
+      "step": 14497
+    },
+    {
+      "epoch": 2.34899546338302,
+      "grad_norm": 0.8489283323287964,
+      "learning_rate": 5.924683438619208e-07,
+      "loss": 0.0552,
+      "step": 14498
+    },
+    {
+      "epoch": 2.349157485418017,
+      "grad_norm": 0.9036144614219666,
+      "learning_rate": 5.921857068754494e-07,
+      "loss": 0.0573,
+      "step": 14499
+    },
+    {
+      "epoch": 2.3493195074530138,
+      "grad_norm": 0.8492865562438965,
+      "learning_rate": 5.91903128263479e-07,
+      "loss": 0.0608,
+      "step": 14500
+    },
+    {
+      "epoch": 2.3494815294880103,
+      "grad_norm": 0.8074772953987122,
+      "learning_rate": 5.916206080346549e-07,
+      "loss": 0.0574,
+      "step": 14501
+    },
+    {
+      "epoch": 2.349643551523007,
+      "grad_norm": 0.9476245641708374,
+      "learning_rate": 5.913381461976217e-07,
+      "loss": 0.0591,
+      "step": 14502
+    },
+    {
+      "epoch": 2.3498055735580037,
+      "grad_norm": 0.938563346862793,
+      "learning_rate": 5.910557427610225e-07,
+      "loss": 0.0571,
+      "step": 14503
+    },
+    {
+      "epoch": 2.3499675955930006,
+      "grad_norm": 0.9736203551292419,
+      "learning_rate": 5.907733977334978e-07,
+      "loss": 0.0621,
+      "step": 14504
+    },
+    {
+      "epoch": 2.3501296176279975,
+      "grad_norm": 0.8642393350601196,
+      "learning_rate": 5.904911111236872e-07,
+      "loss": 0.0633,
+      "step": 14505
+    },
+    {
+      "epoch": 2.350291639662994,
+      "grad_norm": 0.9992730617523193,
+      "learning_rate": 5.902088829402274e-07,
+      "loss": 0.0643,
+      "step": 14506
+    },
+    {
+      "epoch": 2.350453661697991,
+      "grad_norm": 0.8306195735931396,
+      "learning_rate": 5.899267131917547e-07,
+      "loss": 0.0538,
+      "step": 14507
+    },
+    {
+      "epoch": 2.350615683732988,
+      "grad_norm": 0.9145315885543823,
+      "learning_rate": 5.896446018869018e-07,
+      "loss": 0.0614,
+      "step": 14508
+    },
+    {
+      "epoch": 2.3507777057679844,
+      "grad_norm": 0.948464035987854,
+      "learning_rate": 5.893625490343014e-07,
+      "loss": 0.0604,
+      "step": 14509
+    },
+    {
+      "epoch": 2.3509397278029813,
+      "grad_norm": 0.7787171006202698,
+      "learning_rate": 5.890805546425832e-07,
+      "loss": 0.0572,
+      "step": 14510
+    },
+    {
+      "epoch": 2.351101749837978,
+      "grad_norm": 0.9339138269424438,
+      "learning_rate": 5.887986187203762e-07,
+      "loss": 0.0636,
+      "step": 14511
+    },
+    {
+      "epoch": 2.3512637718729748,
+      "grad_norm": 1.0364383459091187,
+      "learning_rate": 5.885167412763051e-07,
+      "loss": 0.067,
+      "step": 14512
+    },
+    {
+      "epoch": 2.3514257939079712,
+      "grad_norm": 0.8669085502624512,
+      "learning_rate": 5.88234922318997e-07,
+      "loss": 0.0552,
+      "step": 14513
+    },
+    {
+      "epoch": 2.351587815942968,
+      "grad_norm": 0.9558306932449341,
+      "learning_rate": 5.879531618570738e-07,
+      "loss": 0.0647,
+      "step": 14514
+    },
+    {
+      "epoch": 2.351749837977965,
+      "grad_norm": 0.9057703614234924,
+      "learning_rate": 5.876714598991573e-07,
+      "loss": 0.0636,
+      "step": 14515
+    },
+    {
+      "epoch": 2.3519118600129616,
+      "grad_norm": 0.9369593262672424,
+      "learning_rate": 5.873898164538658e-07,
+      "loss": 0.0662,
+      "step": 14516
+    },
+    {
+      "epoch": 2.3520738820479585,
+      "grad_norm": 0.9057307839393616,
+      "learning_rate": 5.871082315298168e-07,
+      "loss": 0.0591,
+      "step": 14517
+    },
+    {
+      "epoch": 2.3522359040829555,
+      "grad_norm": 0.8824877738952637,
+      "learning_rate": 5.868267051356283e-07,
+      "loss": 0.0576,
+      "step": 14518
+    },
+    {
+      "epoch": 2.352397926117952,
+      "grad_norm": 0.895741879940033,
+      "learning_rate": 5.865452372799121e-07,
+      "loss": 0.0594,
+      "step": 14519
+    },
+    {
+      "epoch": 2.352559948152949,
+      "grad_norm": 0.8106254935264587,
+      "learning_rate": 5.86263827971281e-07,
+      "loss": 0.0609,
+      "step": 14520
+    },
+    {
+      "epoch": 2.3527219701879454,
+      "grad_norm": 0.7932037115097046,
+      "learning_rate": 5.859824772183459e-07,
+      "loss": 0.0541,
+      "step": 14521
+    },
+    {
+      "epoch": 2.3528839922229423,
+      "grad_norm": 1.2299085855484009,
+      "learning_rate": 5.857011850297148e-07,
+      "loss": 0.0635,
+      "step": 14522
+    },
+    {
+      "epoch": 2.3530460142579392,
+      "grad_norm": 0.9761585593223572,
+      "learning_rate": 5.854199514139952e-07,
+      "loss": 0.0563,
+      "step": 14523
+    },
+    {
+      "epoch": 2.3532080362929357,
+      "grad_norm": 0.9831428527832031,
+      "learning_rate": 5.851387763797916e-07,
+      "loss": 0.0723,
+      "step": 14524
+    },
+    {
+      "epoch": 2.3533700583279327,
+      "grad_norm": 0.9172781705856323,
+      "learning_rate": 5.848576599357078e-07,
+      "loss": 0.0666,
+      "step": 14525
+    },
+    {
+      "epoch": 2.353532080362929,
+      "grad_norm": 0.9665817022323608,
+      "learning_rate": 5.845766020903459e-07,
+      "loss": 0.0599,
+      "step": 14526
+    },
+    {
+      "epoch": 2.353694102397926,
+      "grad_norm": 0.8048878908157349,
+      "learning_rate": 5.842956028523031e-07,
+      "loss": 0.0518,
+      "step": 14527
+    },
+    {
+      "epoch": 2.353856124432923,
+      "grad_norm": 1.1949489116668701,
+      "learning_rate": 5.840146622301796e-07,
+      "loss": 0.0697,
+      "step": 14528
+    },
+    {
+      "epoch": 2.3540181464679195,
+      "grad_norm": 1.0436763763427734,
+      "learning_rate": 5.837337802325718e-07,
+      "loss": 0.0735,
+      "step": 14529
+    },
+    {
+      "epoch": 2.3541801685029164,
+      "grad_norm": 0.9715738892555237,
+      "learning_rate": 5.834529568680722e-07,
+      "loss": 0.067,
+      "step": 14530
+    },
+    {
+      "epoch": 2.3543421905379134,
+      "grad_norm": 0.9725156426429749,
+      "learning_rate": 5.83172192145274e-07,
+      "loss": 0.065,
+      "step": 14531
+    },
+    {
+      "epoch": 2.35450421257291,
+      "grad_norm": 1.021435022354126,
+      "learning_rate": 5.828914860727674e-07,
+      "loss": 0.0635,
+      "step": 14532
+    },
+    {
+      "epoch": 2.354666234607907,
+      "grad_norm": 0.9970274567604065,
+      "learning_rate": 5.826108386591436e-07,
+      "loss": 0.0587,
+      "step": 14533
+    },
+    {
+      "epoch": 2.3548282566429033,
+      "grad_norm": 1.0088868141174316,
+      "learning_rate": 5.823302499129873e-07,
+      "loss": 0.0653,
+      "step": 14534
+    },
+    {
+      "epoch": 2.3549902786779002,
+      "grad_norm": 0.9348923563957214,
+      "learning_rate": 5.820497198428849e-07,
+      "loss": 0.0649,
+      "step": 14535
+    },
+    {
+      "epoch": 2.3551523007128967,
+      "grad_norm": 0.8347666263580322,
+      "learning_rate": 5.817692484574197e-07,
+      "loss": 0.0563,
+      "step": 14536
+    },
+    {
+      "epoch": 2.3553143227478937,
+      "grad_norm": 0.8557009696960449,
+      "learning_rate": 5.814888357651733e-07,
+      "loss": 0.0638,
+      "step": 14537
+    },
+    {
+      "epoch": 2.3554763447828906,
+      "grad_norm": 0.8046319484710693,
+      "learning_rate": 5.81208481774726e-07,
+      "loss": 0.0533,
+      "step": 14538
+    },
+    {
+      "epoch": 2.355638366817887,
+      "grad_norm": 0.9410936832427979,
+      "learning_rate": 5.809281864946556e-07,
+      "loss": 0.0613,
+      "step": 14539
+    },
+    {
+      "epoch": 2.355800388852884,
+      "grad_norm": 0.840984582901001,
+      "learning_rate": 5.806479499335385e-07,
+      "loss": 0.0563,
+      "step": 14540
+    },
+    {
+      "epoch": 2.355962410887881,
+      "grad_norm": 0.9175610542297363,
+      "learning_rate": 5.803677720999495e-07,
+      "loss": 0.059,
+      "step": 14541
+    },
+    {
+      "epoch": 2.3561244329228774,
+      "grad_norm": 0.8809838891029358,
+      "learning_rate": 5.800876530024615e-07,
+      "loss": 0.0598,
+      "step": 14542
+    },
+    {
+      "epoch": 2.3562864549578744,
+      "grad_norm": 0.9057488441467285,
+      "learning_rate": 5.798075926496449e-07,
+      "loss": 0.0637,
+      "step": 14543
+    },
+    {
+      "epoch": 2.356448476992871,
+      "grad_norm": 0.8945654630661011,
+      "learning_rate": 5.795275910500703e-07,
+      "loss": 0.067,
+      "step": 14544
+    },
+    {
+      "epoch": 2.356610499027868,
+      "grad_norm": 1.0310019254684448,
+      "learning_rate": 5.792476482123027e-07,
+      "loss": 0.0654,
+      "step": 14545
+    },
+    {
+      "epoch": 2.3567725210628647,
+      "grad_norm": 0.8311236500740051,
+      "learning_rate": 5.789677641449087e-07,
+      "loss": 0.0557,
+      "step": 14546
+    },
+    {
+      "epoch": 2.356934543097861,
+      "grad_norm": 0.9365234971046448,
+      "learning_rate": 5.786879388564534e-07,
+      "loss": 0.0632,
+      "step": 14547
+    },
+    {
+      "epoch": 2.357096565132858,
+      "grad_norm": 0.7686519026756287,
+      "learning_rate": 5.784081723554971e-07,
+      "loss": 0.0577,
+      "step": 14548
+    },
+    {
+      "epoch": 2.3572585871678546,
+      "grad_norm": 0.86738520860672,
+      "learning_rate": 5.781284646506008e-07,
+      "loss": 0.0601,
+      "step": 14549
+    },
+    {
+      "epoch": 2.3574206092028516,
+      "grad_norm": 0.8879827857017517,
+      "learning_rate": 5.778488157503223e-07,
+      "loss": 0.0595,
+      "step": 14550
+    },
+    {
+      "epoch": 2.3575826312378485,
+      "grad_norm": 0.8974021077156067,
+      "learning_rate": 5.775692256632187e-07,
+      "loss": 0.0587,
+      "step": 14551
+    },
+    {
+      "epoch": 2.357744653272845,
+      "grad_norm": 0.9620395302772522,
+      "learning_rate": 5.772896943978446e-07,
+      "loss": 0.0681,
+      "step": 14552
+    },
+    {
+      "epoch": 2.357906675307842,
+      "grad_norm": 0.8157269954681396,
+      "learning_rate": 5.770102219627526e-07,
+      "loss": 0.058,
+      "step": 14553
+    },
+    {
+      "epoch": 2.358068697342839,
+      "grad_norm": 0.857354998588562,
+      "learning_rate": 5.767308083664949e-07,
+      "loss": 0.0626,
+      "step": 14554
+    },
+    {
+      "epoch": 2.3582307193778353,
+      "grad_norm": 1.1273916959762573,
+      "learning_rate": 5.7645145361762e-07,
+      "loss": 0.0657,
+      "step": 14555
+    },
+    {
+      "epoch": 2.3583927414128323,
+      "grad_norm": 0.9196612238883972,
+      "learning_rate": 5.761721577246754e-07,
+      "loss": 0.0593,
+      "step": 14556
+    },
+    {
+      "epoch": 2.3585547634478288,
+      "grad_norm": 0.9983531832695007,
+      "learning_rate": 5.758929206962074e-07,
+      "loss": 0.0676,
+      "step": 14557
+    },
+    {
+      "epoch": 2.3587167854828257,
+      "grad_norm": 1.0570601224899292,
+      "learning_rate": 5.756137425407598e-07,
+      "loss": 0.0627,
+      "step": 14558
+    },
+    {
+      "epoch": 2.358878807517822,
+      "grad_norm": 1.1186915636062622,
+      "learning_rate": 5.753346232668758e-07,
+      "loss": 0.0683,
+      "step": 14559
+    },
+    {
+      "epoch": 2.359040829552819,
+      "grad_norm": 0.9919911623001099,
+      "learning_rate": 5.750555628830928e-07,
+      "loss": 0.0628,
+      "step": 14560
+    },
+    {
+      "epoch": 2.359202851587816,
+      "grad_norm": 0.9504989385604858,
+      "learning_rate": 5.747765613979523e-07,
+      "loss": 0.0648,
+      "step": 14561
+    },
+    {
+      "epoch": 2.3593648736228126,
+      "grad_norm": 0.9681611061096191,
+      "learning_rate": 5.744976188199905e-07,
+      "loss": 0.0653,
+      "step": 14562
+    },
+    {
+      "epoch": 2.3595268956578095,
+      "grad_norm": 0.9179617166519165,
+      "learning_rate": 5.742187351577416e-07,
+      "loss": 0.064,
+      "step": 14563
+    },
+    {
+      "epoch": 2.3596889176928064,
+      "grad_norm": 0.8961821794509888,
+      "learning_rate": 5.739399104197388e-07,
+      "loss": 0.0585,
+      "step": 14564
+    },
+    {
+      "epoch": 2.359850939727803,
+      "grad_norm": 0.9559543132781982,
+      "learning_rate": 5.73661144614514e-07,
+      "loss": 0.0606,
+      "step": 14565
+    },
+    {
+      "epoch": 2.3600129617628,
+      "grad_norm": 0.9991036653518677,
+      "learning_rate": 5.733824377505965e-07,
+      "loss": 0.0674,
+      "step": 14566
+    },
+    {
+      "epoch": 2.3601749837977963,
+      "grad_norm": 0.9018924832344055,
+      "learning_rate": 5.731037898365138e-07,
+      "loss": 0.0603,
+      "step": 14567
+    },
+    {
+      "epoch": 2.3603370058327933,
+      "grad_norm": 0.8355051279067993,
+      "learning_rate": 5.728252008807925e-07,
+      "loss": 0.0548,
+      "step": 14568
+    },
+    {
+      "epoch": 2.36049902786779,
+      "grad_norm": 0.8183480501174927,
+      "learning_rate": 5.725466708919561e-07,
+      "loss": 0.0614,
+      "step": 14569
+    },
+    {
+      "epoch": 2.3606610499027867,
+      "grad_norm": 0.9316161274909973,
+      "learning_rate": 5.722681998785273e-07,
+      "loss": 0.0618,
+      "step": 14570
+    },
+    {
+      "epoch": 2.3608230719377836,
+      "grad_norm": 0.8802493810653687,
+      "learning_rate": 5.719897878490265e-07,
+      "loss": 0.0595,
+      "step": 14571
+    },
+    {
+      "epoch": 2.36098509397278,
+      "grad_norm": 0.8617424368858337,
+      "learning_rate": 5.717114348119726e-07,
+      "loss": 0.0638,
+      "step": 14572
+    },
+    {
+      "epoch": 2.361147116007777,
+      "grad_norm": 0.9052625894546509,
+      "learning_rate": 5.71433140775883e-07,
+      "loss": 0.0619,
+      "step": 14573
+    },
+    {
+      "epoch": 2.361309138042774,
+      "grad_norm": 1.0340335369110107,
+      "learning_rate": 5.711549057492718e-07,
+      "loss": 0.0687,
+      "step": 14574
+    },
+    {
+      "epoch": 2.3614711600777705,
+      "grad_norm": 0.8974990844726562,
+      "learning_rate": 5.70876729740652e-07,
+      "loss": 0.056,
+      "step": 14575
+    },
+    {
+      "epoch": 2.3616331821127674,
+      "grad_norm": 0.8803661465644836,
+      "learning_rate": 5.705986127585364e-07,
+      "loss": 0.0533,
+      "step": 14576
+    },
+    {
+      "epoch": 2.3617952041477643,
+      "grad_norm": 0.917258620262146,
+      "learning_rate": 5.703205548114352e-07,
+      "loss": 0.0671,
+      "step": 14577
+    },
+    {
+      "epoch": 2.361957226182761,
+      "grad_norm": 0.8356657028198242,
+      "learning_rate": 5.700425559078543e-07,
+      "loss": 0.0538,
+      "step": 14578
+    },
+    {
+      "epoch": 2.3621192482177578,
+      "grad_norm": 0.7526644468307495,
+      "learning_rate": 5.697646160563001e-07,
+      "loss": 0.0544,
+      "step": 14579
+    },
+    {
+      "epoch": 2.3622812702527543,
+      "grad_norm": 0.8724152445793152,
+      "learning_rate": 5.694867352652791e-07,
+      "loss": 0.0571,
+      "step": 14580
+    },
+    {
+      "epoch": 2.362443292287751,
+      "grad_norm": 0.988396167755127,
+      "learning_rate": 5.692089135432913e-07,
+      "loss": 0.0603,
+      "step": 14581
+    },
+    {
+      "epoch": 2.3626053143227477,
+      "grad_norm": 0.8566252589225769,
+      "learning_rate": 5.689311508988385e-07,
+      "loss": 0.0614,
+      "step": 14582
+    },
+    {
+      "epoch": 2.3627673363577446,
+      "grad_norm": 0.9325825572013855,
+      "learning_rate": 5.686534473404195e-07,
+      "loss": 0.062,
+      "step": 14583
+    },
+    {
+      "epoch": 2.3629293583927415,
+      "grad_norm": 0.9369713664054871,
+      "learning_rate": 5.68375802876531e-07,
+      "loss": 0.0612,
+      "step": 14584
+    },
+    {
+      "epoch": 2.363091380427738,
+      "grad_norm": 0.9119452834129333,
+      "learning_rate": 5.680982175156688e-07,
+      "loss": 0.0552,
+      "step": 14585
+    },
+    {
+      "epoch": 2.363253402462735,
+      "grad_norm": 0.9514526128768921,
+      "learning_rate": 5.678206912663259e-07,
+      "loss": 0.0559,
+      "step": 14586
+    },
+    {
+      "epoch": 2.363415424497732,
+      "grad_norm": 0.9722010493278503,
+      "learning_rate": 5.675432241369938e-07,
+      "loss": 0.0602,
+      "step": 14587
+    },
+    {
+      "epoch": 2.3635774465327284,
+      "grad_norm": 0.9492889046669006,
+      "learning_rate": 5.672658161361636e-07,
+      "loss": 0.0601,
+      "step": 14588
+    },
+    {
+      "epoch": 2.3637394685677253,
+      "grad_norm": 0.9314302206039429,
+      "learning_rate": 5.669884672723208e-07,
+      "loss": 0.0622,
+      "step": 14589
+    },
+    {
+      "epoch": 2.363901490602722,
+      "grad_norm": 0.9510951638221741,
+      "learning_rate": 5.667111775539538e-07,
+      "loss": 0.0754,
+      "step": 14590
+    },
+    {
+      "epoch": 2.3640635126377187,
+      "grad_norm": 1.04404616355896,
+      "learning_rate": 5.664339469895472e-07,
+      "loss": 0.0693,
+      "step": 14591
+    },
+    {
+      "epoch": 2.3642255346727157,
+      "grad_norm": 0.877565860748291,
+      "learning_rate": 5.661567755875816e-07,
+      "loss": 0.061,
+      "step": 14592
+    },
+    {
+      "epoch": 2.364387556707712,
+      "grad_norm": 0.8133190274238586,
+      "learning_rate": 5.658796633565391e-07,
+      "loss": 0.0578,
+      "step": 14593
+    },
+    {
+      "epoch": 2.364549578742709,
+      "grad_norm": 0.9318547248840332,
+      "learning_rate": 5.656026103048975e-07,
+      "loss": 0.0612,
+      "step": 14594
+    },
+    {
+      "epoch": 2.3647116007777056,
+      "grad_norm": 1.0102546215057373,
+      "learning_rate": 5.653256164411366e-07,
+      "loss": 0.066,
+      "step": 14595
+    },
+    {
+      "epoch": 2.3648736228127025,
+      "grad_norm": 0.8324956297874451,
+      "learning_rate": 5.650486817737291e-07,
+      "loss": 0.0536,
+      "step": 14596
+    },
+    {
+      "epoch": 2.3650356448476995,
+      "grad_norm": 0.9316480755805969,
+      "learning_rate": 5.647718063111496e-07,
+      "loss": 0.0612,
+      "step": 14597
+    },
+    {
+      "epoch": 2.365197666882696,
+      "grad_norm": 1.2550431489944458,
+      "learning_rate": 5.644949900618696e-07,
+      "loss": 0.06,
+      "step": 14598
+    },
+    {
+      "epoch": 2.365359688917693,
+      "grad_norm": 0.8270633816719055,
+      "learning_rate": 5.64218233034359e-07,
+      "loss": 0.0493,
+      "step": 14599
+    },
+    {
+      "epoch": 2.3655217109526894,
+      "grad_norm": 0.9431858062744141,
+      "learning_rate": 5.639415352370858e-07,
+      "loss": 0.066,
+      "step": 14600
+    },
+    {
+      "epoch": 2.3656837329876863,
+      "grad_norm": 0.8048388361930847,
+      "learning_rate": 5.636648966785168e-07,
+      "loss": 0.0605,
+      "step": 14601
+    },
+    {
+      "epoch": 2.3658457550226832,
+      "grad_norm": 0.8069517016410828,
+      "learning_rate": 5.633883173671159e-07,
+      "loss": 0.0572,
+      "step": 14602
+    },
+    {
+      "epoch": 2.3660077770576797,
+      "grad_norm": 0.9145830273628235,
+      "learning_rate": 5.631117973113462e-07,
+      "loss": 0.0575,
+      "step": 14603
+    },
+    {
+      "epoch": 2.3661697990926767,
+      "grad_norm": 0.9402312636375427,
+      "learning_rate": 5.628353365196682e-07,
+      "loss": 0.064,
+      "step": 14604
+    },
+    {
+      "epoch": 2.366331821127673,
+      "grad_norm": 0.8871921896934509,
+      "learning_rate": 5.625589350005409e-07,
+      "loss": 0.0612,
+      "step": 14605
+    },
+    {
+      "epoch": 2.36649384316267,
+      "grad_norm": 0.9529529213905334,
+      "learning_rate": 5.622825927624226e-07,
+      "loss": 0.0685,
+      "step": 14606
+    },
+    {
+      "epoch": 2.366655865197667,
+      "grad_norm": 0.8908757567405701,
+      "learning_rate": 5.620063098137668e-07,
+      "loss": 0.0539,
+      "step": 14607
+    },
+    {
+      "epoch": 2.3668178872326635,
+      "grad_norm": 0.897894561290741,
+      "learning_rate": 5.617300861630276e-07,
+      "loss": 0.0623,
+      "step": 14608
+    },
+    {
+      "epoch": 2.3669799092676604,
+      "grad_norm": 1.1129932403564453,
+      "learning_rate": 5.61453921818658e-07,
+      "loss": 0.0598,
+      "step": 14609
+    },
+    {
+      "epoch": 2.3671419313026574,
+      "grad_norm": 0.9506595730781555,
+      "learning_rate": 5.611778167891077e-07,
+      "loss": 0.0658,
+      "step": 14610
+    },
+    {
+      "epoch": 2.367303953337654,
+      "grad_norm": 1.14960777759552,
+      "learning_rate": 5.609017710828238e-07,
+      "loss": 0.0709,
+      "step": 14611
+    },
+    {
+      "epoch": 2.367465975372651,
+      "grad_norm": 0.8241292834281921,
+      "learning_rate": 5.60625784708253e-07,
+      "loss": 0.0601,
+      "step": 14612
+    },
+    {
+      "epoch": 2.3676279974076473,
+      "grad_norm": 0.987915575504303,
+      "learning_rate": 5.6034985767384e-07,
+      "loss": 0.0651,
+      "step": 14613
+    },
+    {
+      "epoch": 2.3677900194426442,
+      "grad_norm": 0.8549913167953491,
+      "learning_rate": 5.600739899880275e-07,
+      "loss": 0.0588,
+      "step": 14614
+    },
+    {
+      "epoch": 2.3679520414776407,
+      "grad_norm": 0.9886636137962341,
+      "learning_rate": 5.597981816592565e-07,
+      "loss": 0.0615,
+      "step": 14615
+    },
+    {
+      "epoch": 2.3681140635126376,
+      "grad_norm": 0.8965216279029846,
+      "learning_rate": 5.595224326959662e-07,
+      "loss": 0.059,
+      "step": 14616
+    },
+    {
+      "epoch": 2.3682760855476346,
+      "grad_norm": 0.8499079346656799,
+      "learning_rate": 5.592467431065937e-07,
+      "loss": 0.0602,
+      "step": 14617
+    },
+    {
+      "epoch": 2.368438107582631,
+      "grad_norm": 0.9334196448326111,
+      "learning_rate": 5.589711128995734e-07,
+      "loss": 0.0587,
+      "step": 14618
+    },
+    {
+      "epoch": 2.368600129617628,
+      "grad_norm": 0.8634200692176819,
+      "learning_rate": 5.586955420833404e-07,
+      "loss": 0.0553,
+      "step": 14619
+    },
+    {
+      "epoch": 2.368762151652625,
+      "grad_norm": 0.9879969954490662,
+      "learning_rate": 5.584200306663259e-07,
+      "loss": 0.0591,
+      "step": 14620
+    },
+    {
+      "epoch": 2.3689241736876214,
+      "grad_norm": 0.8652029037475586,
+      "learning_rate": 5.581445786569606e-07,
+      "loss": 0.0591,
+      "step": 14621
+    },
+    {
+      "epoch": 2.3690861957226184,
+      "grad_norm": 1.0487457513809204,
+      "learning_rate": 5.578691860636706e-07,
+      "loss": 0.0652,
+      "step": 14622
+    },
+    {
+      "epoch": 2.369248217757615,
+      "grad_norm": 0.9185906648635864,
+      "learning_rate": 5.575938528948843e-07,
+      "loss": 0.0608,
+      "step": 14623
+    },
+    {
+      "epoch": 2.369410239792612,
+      "grad_norm": 0.8492238521575928,
+      "learning_rate": 5.573185791590266e-07,
+      "loss": 0.0612,
+      "step": 14624
+    },
+    {
+      "epoch": 2.3695722618276087,
+      "grad_norm": 0.8446729779243469,
+      "learning_rate": 5.570433648645182e-07,
+      "loss": 0.0578,
+      "step": 14625
+    },
+    {
+      "epoch": 2.369734283862605,
+      "grad_norm": 0.8767587542533875,
+      "learning_rate": 5.567682100197808e-07,
+      "loss": 0.0626,
+      "step": 14626
+    },
+    {
+      "epoch": 2.369896305897602,
+      "grad_norm": 0.8166986703872681,
+      "learning_rate": 5.564931146332334e-07,
+      "loss": 0.0594,
+      "step": 14627
+    },
+    {
+      "epoch": 2.3700583279325986,
+      "grad_norm": 0.84421306848526,
+      "learning_rate": 5.562180787132945e-07,
+      "loss": 0.06,
+      "step": 14628
+    },
+    {
+      "epoch": 2.3702203499675956,
+      "grad_norm": 0.8665793538093567,
+      "learning_rate": 5.559431022683779e-07,
+      "loss": 0.046,
+      "step": 14629
+    },
+    {
+      "epoch": 2.3703823720025925,
+      "grad_norm": 0.7502251267433167,
+      "learning_rate": 5.55668185306898e-07,
+      "loss": 0.0541,
+      "step": 14630
+    },
+    {
+      "epoch": 2.370544394037589,
+      "grad_norm": 0.9786513447761536,
+      "learning_rate": 5.553933278372664e-07,
+      "loss": 0.0676,
+      "step": 14631
+    },
+    {
+      "epoch": 2.370706416072586,
+      "grad_norm": 0.8539317846298218,
+      "learning_rate": 5.551185298678929e-07,
+      "loss": 0.059,
+      "step": 14632
+    },
+    {
+      "epoch": 2.370868438107583,
+      "grad_norm": 0.9358149766921997,
+      "learning_rate": 5.548437914071861e-07,
+      "loss": 0.0671,
+      "step": 14633
+    },
+    {
+      "epoch": 2.3710304601425793,
+      "grad_norm": 0.7582257986068726,
+      "learning_rate": 5.545691124635518e-07,
+      "loss": 0.0528,
+      "step": 14634
+    },
+    {
+      "epoch": 2.3711924821775763,
+      "grad_norm": 0.9727949500083923,
+      "learning_rate": 5.542944930453958e-07,
+      "loss": 0.0642,
+      "step": 14635
+    },
+    {
+      "epoch": 2.3713545042125728,
+      "grad_norm": 0.9191507697105408,
+      "learning_rate": 5.54019933161119e-07,
+      "loss": 0.0564,
+      "step": 14636
+    },
+    {
+      "epoch": 2.3715165262475697,
+      "grad_norm": 0.9941038489341736,
+      "learning_rate": 5.537454328191225e-07,
+      "loss": 0.0654,
+      "step": 14637
+    },
+    {
+      "epoch": 2.371678548282566,
+      "grad_norm": 1.0261110067367554,
+      "learning_rate": 5.534709920278064e-07,
+      "loss": 0.0663,
+      "step": 14638
+    },
+    {
+      "epoch": 2.371840570317563,
+      "grad_norm": 1.0194072723388672,
+      "learning_rate": 5.531966107955683e-07,
+      "loss": 0.058,
+      "step": 14639
+    },
+    {
+      "epoch": 2.37200259235256,
+      "grad_norm": 0.9304824471473694,
+      "learning_rate": 5.52922289130802e-07,
+      "loss": 0.0633,
+      "step": 14640
+    },
+    {
+      "epoch": 2.3721646143875565,
+      "grad_norm": 0.9960220456123352,
+      "learning_rate": 5.526480270419018e-07,
+      "loss": 0.0677,
+      "step": 14641
+    },
+    {
+      "epoch": 2.3723266364225535,
+      "grad_norm": 0.8424884080886841,
+      "learning_rate": 5.523738245372596e-07,
+      "loss": 0.06,
+      "step": 14642
+    },
+    {
+      "epoch": 2.3724886584575504,
+      "grad_norm": 0.851523220539093,
+      "learning_rate": 5.52099681625265e-07,
+      "loss": 0.0618,
+      "step": 14643
+    },
+    {
+      "epoch": 2.372650680492547,
+      "grad_norm": 0.8086889386177063,
+      "learning_rate": 5.518255983143061e-07,
+      "loss": 0.0611,
+      "step": 14644
+    },
+    {
+      "epoch": 2.372812702527544,
+      "grad_norm": 0.8051342368125916,
+      "learning_rate": 5.515515746127697e-07,
+      "loss": 0.0589,
+      "step": 14645
+    },
+    {
+      "epoch": 2.3729747245625403,
+      "grad_norm": 0.9982341527938843,
+      "learning_rate": 5.512776105290402e-07,
+      "loss": 0.0722,
+      "step": 14646
+    },
+    {
+      "epoch": 2.3731367465975373,
+      "grad_norm": 0.933042049407959,
+      "learning_rate": 5.510037060714995e-07,
+      "loss": 0.0621,
+      "step": 14647
+    },
+    {
+      "epoch": 2.373298768632534,
+      "grad_norm": 0.872920572757721,
+      "learning_rate": 5.507298612485293e-07,
+      "loss": 0.0582,
+      "step": 14648
+    },
+    {
+      "epoch": 2.3734607906675307,
+      "grad_norm": 0.8864262104034424,
+      "learning_rate": 5.504560760685079e-07,
+      "loss": 0.0587,
+      "step": 14649
+    },
+    {
+      "epoch": 2.3736228127025276,
+      "grad_norm": 0.8857452869415283,
+      "learning_rate": 5.501823505398137e-07,
+      "loss": 0.0643,
+      "step": 14650
+    },
+    {
+      "epoch": 2.373784834737524,
+      "grad_norm": 0.9040178060531616,
+      "learning_rate": 5.499086846708196e-07,
+      "loss": 0.06,
+      "step": 14651
+    },
+    {
+      "epoch": 2.373946856772521,
+      "grad_norm": 0.796747624874115,
+      "learning_rate": 5.496350784699015e-07,
+      "loss": 0.0553,
+      "step": 14652
+    },
+    {
+      "epoch": 2.374108878807518,
+      "grad_norm": 1.0335019826889038,
+      "learning_rate": 5.493615319454299e-07,
+      "loss": 0.0642,
+      "step": 14653
+    },
+    {
+      "epoch": 2.3742709008425145,
+      "grad_norm": 0.760349690914154,
+      "learning_rate": 5.490880451057759e-07,
+      "loss": 0.0557,
+      "step": 14654
+    },
+    {
+      "epoch": 2.3744329228775114,
+      "grad_norm": 0.7830966711044312,
+      "learning_rate": 5.488146179593057e-07,
+      "loss": 0.0559,
+      "step": 14655
+    },
+    {
+      "epoch": 2.3745949449125083,
+      "grad_norm": 1.081561803817749,
+      "learning_rate": 5.485412505143858e-07,
+      "loss": 0.0734,
+      "step": 14656
+    },
+    {
+      "epoch": 2.374756966947505,
+      "grad_norm": 0.9587316513061523,
+      "learning_rate": 5.482679427793827e-07,
+      "loss": 0.0659,
+      "step": 14657
+    },
+    {
+      "epoch": 2.3749189889825018,
+      "grad_norm": 0.8291387557983398,
+      "learning_rate": 5.479946947626566e-07,
+      "loss": 0.0584,
+      "step": 14658
+    },
+    {
+      "epoch": 2.3750810110174982,
+      "grad_norm": 0.8340387940406799,
+      "learning_rate": 5.477215064725692e-07,
+      "loss": 0.0572,
+      "step": 14659
+    },
+    {
+      "epoch": 2.375243033052495,
+      "grad_norm": 1.00160551071167,
+      "learning_rate": 5.474483779174791e-07,
+      "loss": 0.0646,
+      "step": 14660
+    },
+    {
+      "epoch": 2.3754050550874917,
+      "grad_norm": 0.9756659269332886,
+      "learning_rate": 5.471753091057438e-07,
+      "loss": 0.0626,
+      "step": 14661
+    },
+    {
+      "epoch": 2.3755670771224886,
+      "grad_norm": 0.8894532322883606,
+      "learning_rate": 5.469023000457183e-07,
+      "loss": 0.0557,
+      "step": 14662
+    },
+    {
+      "epoch": 2.3757290991574855,
+      "grad_norm": 1.0041909217834473,
+      "learning_rate": 5.466293507457557e-07,
+      "loss": 0.0645,
+      "step": 14663
+    },
+    {
+      "epoch": 2.375891121192482,
+      "grad_norm": 0.9752393364906311,
+      "learning_rate": 5.463564612142083e-07,
+      "loss": 0.0614,
+      "step": 14664
+    },
+    {
+      "epoch": 2.376053143227479,
+      "grad_norm": 0.8114173412322998,
+      "learning_rate": 5.460836314594259e-07,
+      "loss": 0.0578,
+      "step": 14665
+    },
+    {
+      "epoch": 2.376215165262476,
+      "grad_norm": 0.8527717590332031,
+      "learning_rate": 5.458108614897545e-07,
+      "loss": 0.0601,
+      "step": 14666
+    },
+    {
+      "epoch": 2.3763771872974724,
+      "grad_norm": 1.0959402322769165,
+      "learning_rate": 5.455381513135427e-07,
+      "loss": 0.0673,
+      "step": 14667
+    },
+    {
+      "epoch": 2.3765392093324693,
+      "grad_norm": 1.0152456760406494,
+      "learning_rate": 5.452655009391341e-07,
+      "loss": 0.0676,
+      "step": 14668
+    },
+    {
+      "epoch": 2.376701231367466,
+      "grad_norm": 0.9547176957130432,
+      "learning_rate": 5.449929103748705e-07,
+      "loss": 0.0598,
+      "step": 14669
+    },
+    {
+      "epoch": 2.3768632534024627,
+      "grad_norm": 0.8178269863128662,
+      "learning_rate": 5.447203796290918e-07,
+      "loss": 0.0586,
+      "step": 14670
+    },
+    {
+      "epoch": 2.3770252754374597,
+      "grad_norm": 1.0196658372879028,
+      "learning_rate": 5.444479087101387e-07,
+      "loss": 0.0674,
+      "step": 14671
+    },
+    {
+      "epoch": 2.377187297472456,
+      "grad_norm": 0.8403282761573792,
+      "learning_rate": 5.441754976263478e-07,
+      "loss": 0.0641,
+      "step": 14672
+    },
+    {
+      "epoch": 2.377349319507453,
+      "grad_norm": 1.0634124279022217,
+      "learning_rate": 5.43903146386053e-07,
+      "loss": 0.0671,
+      "step": 14673
+    },
+    {
+      "epoch": 2.3775113415424496,
+      "grad_norm": 0.9557974338531494,
+      "learning_rate": 5.436308549975883e-07,
+      "loss": 0.0666,
+      "step": 14674
+    },
+    {
+      "epoch": 2.3776733635774465,
+      "grad_norm": 0.9498515725135803,
+      "learning_rate": 5.433586234692853e-07,
+      "loss": 0.0594,
+      "step": 14675
+    },
+    {
+      "epoch": 2.3778353856124435,
+      "grad_norm": 0.8803273439407349,
+      "learning_rate": 5.430864518094731e-07,
+      "loss": 0.0602,
+      "step": 14676
+    },
+    {
+      "epoch": 2.37799740764744,
+      "grad_norm": 0.8781242966651917,
+      "learning_rate": 5.428143400264799e-07,
+      "loss": 0.0583,
+      "step": 14677
+    },
+    {
+      "epoch": 2.378159429682437,
+      "grad_norm": 0.9101912975311279,
+      "learning_rate": 5.425422881286319e-07,
+      "loss": 0.0566,
+      "step": 14678
+    },
+    {
+      "epoch": 2.378321451717434,
+      "grad_norm": 0.9364922046661377,
+      "learning_rate": 5.422702961242532e-07,
+      "loss": 0.0705,
+      "step": 14679
+    },
+    {
+      "epoch": 2.3784834737524303,
+      "grad_norm": 0.9377305507659912,
+      "learning_rate": 5.419983640216647e-07,
+      "loss": 0.0618,
+      "step": 14680
+    },
+    {
+      "epoch": 2.3786454957874272,
+      "grad_norm": 0.8949405550956726,
+      "learning_rate": 5.417264918291887e-07,
+      "loss": 0.059,
+      "step": 14681
+    },
+    {
+      "epoch": 2.3788075178224237,
+      "grad_norm": 1.1497116088867188,
+      "learning_rate": 5.414546795551429e-07,
+      "loss": 0.0691,
+      "step": 14682
+    },
+    {
+      "epoch": 2.3789695398574207,
+      "grad_norm": 0.9060092568397522,
+      "learning_rate": 5.411829272078453e-07,
+      "loss": 0.0627,
+      "step": 14683
+    },
+    {
+      "epoch": 2.379131561892417,
+      "grad_norm": 1.0954656600952148,
+      "learning_rate": 5.409112347956089e-07,
+      "loss": 0.0628,
+      "step": 14684
+    },
+    {
+      "epoch": 2.379293583927414,
+      "grad_norm": 0.9496614336967468,
+      "learning_rate": 5.406396023267473e-07,
+      "loss": 0.0644,
+      "step": 14685
+    },
+    {
+      "epoch": 2.379455605962411,
+      "grad_norm": 0.8147675395011902,
+      "learning_rate": 5.403680298095737e-07,
+      "loss": 0.0566,
+      "step": 14686
+    },
+    {
+      "epoch": 2.3796176279974075,
+      "grad_norm": 0.8436890244483948,
+      "learning_rate": 5.400965172523953e-07,
+      "loss": 0.0596,
+      "step": 14687
+    },
+    {
+      "epoch": 2.3797796500324044,
+      "grad_norm": 1.066830039024353,
+      "learning_rate": 5.398250646635209e-07,
+      "loss": 0.059,
+      "step": 14688
+    },
+    {
+      "epoch": 2.3799416720674014,
+      "grad_norm": 1.006394386291504,
+      "learning_rate": 5.395536720512551e-07,
+      "loss": 0.0612,
+      "step": 14689
+    },
+    {
+      "epoch": 2.380103694102398,
+      "grad_norm": 0.9849836230278015,
+      "learning_rate": 5.392823394239042e-07,
+      "loss": 0.0676,
+      "step": 14690
+    },
+    {
+      "epoch": 2.380265716137395,
+      "grad_norm": 1.1965252161026,
+      "learning_rate": 5.39011066789768e-07,
+      "loss": 0.0702,
+      "step": 14691
+    },
+    {
+      "epoch": 2.3804277381723913,
+      "grad_norm": 0.9115785360336304,
+      "learning_rate": 5.387398541571479e-07,
+      "loss": 0.0634,
+      "step": 14692
+    },
+    {
+      "epoch": 2.380589760207388,
+      "grad_norm": 1.0645346641540527,
+      "learning_rate": 5.384687015343418e-07,
+      "loss": 0.0629,
+      "step": 14693
+    },
+    {
+      "epoch": 2.380751782242385,
+      "grad_norm": 0.9696865677833557,
+      "learning_rate": 5.381976089296467e-07,
+      "loss": 0.0637,
+      "step": 14694
+    },
+    {
+      "epoch": 2.3809138042773816,
+      "grad_norm": 0.9118956327438354,
+      "learning_rate": 5.379265763513574e-07,
+      "loss": 0.0642,
+      "step": 14695
+    },
+    {
+      "epoch": 2.3810758263123786,
+      "grad_norm": 0.9417039155960083,
+      "learning_rate": 5.376556038077668e-07,
+      "loss": 0.0617,
+      "step": 14696
+    },
+    {
+      "epoch": 2.381237848347375,
+      "grad_norm": 0.9443559050559998,
+      "learning_rate": 5.373846913071659e-07,
+      "loss": 0.0653,
+      "step": 14697
+    },
+    {
+      "epoch": 2.381399870382372,
+      "grad_norm": 0.8240593671798706,
+      "learning_rate": 5.371138388578448e-07,
+      "loss": 0.0534,
+      "step": 14698
+    },
+    {
+      "epoch": 2.381561892417369,
+      "grad_norm": 1.074074625968933,
+      "learning_rate": 5.368430464680885e-07,
+      "loss": 0.0641,
+      "step": 14699
+    },
+    {
+      "epoch": 2.3817239144523654,
+      "grad_norm": 1.0060458183288574,
+      "learning_rate": 5.365723141461851e-07,
+      "loss": 0.0625,
+      "step": 14700
+    },
+    {
+      "epoch": 2.3818859364873624,
+      "grad_norm": 0.9213908314704895,
+      "learning_rate": 5.36301641900418e-07,
+      "loss": 0.0614,
+      "step": 14701
+    },
+    {
+      "epoch": 2.3820479585223593,
+      "grad_norm": 1.0687966346740723,
+      "learning_rate": 5.360310297390681e-07,
+      "loss": 0.0705,
+      "step": 14702
+    },
+    {
+      "epoch": 2.3822099805573558,
+      "grad_norm": 0.8176394104957581,
+      "learning_rate": 5.357604776704159e-07,
+      "loss": 0.0566,
+      "step": 14703
+    },
+    {
+      "epoch": 2.3823720025923527,
+      "grad_norm": 0.7344977259635925,
+      "learning_rate": 5.354899857027398e-07,
+      "loss": 0.052,
+      "step": 14704
+    },
+    {
+      "epoch": 2.382534024627349,
+      "grad_norm": 0.8186675310134888,
+      "learning_rate": 5.352195538443162e-07,
+      "loss": 0.0598,
+      "step": 14705
+    },
+    {
+      "epoch": 2.382696046662346,
+      "grad_norm": 0.8668035864830017,
+      "learning_rate": 5.349491821034192e-07,
+      "loss": 0.0565,
+      "step": 14706
+    },
+    {
+      "epoch": 2.3828580686973426,
+      "grad_norm": 0.8794363737106323,
+      "learning_rate": 5.346788704883222e-07,
+      "loss": 0.0653,
+      "step": 14707
+    },
+    {
+      "epoch": 2.3830200907323396,
+      "grad_norm": 0.8817334771156311,
+      "learning_rate": 5.344086190072955e-07,
+      "loss": 0.0602,
+      "step": 14708
+    },
+    {
+      "epoch": 2.3831821127673365,
+      "grad_norm": 1.023415446281433,
+      "learning_rate": 5.341384276686087e-07,
+      "loss": 0.0543,
+      "step": 14709
+    },
+    {
+      "epoch": 2.383344134802333,
+      "grad_norm": 0.9122980237007141,
+      "learning_rate": 5.338682964805286e-07,
+      "loss": 0.0592,
+      "step": 14710
+    },
+    {
+      "epoch": 2.38350615683733,
+      "grad_norm": 1.0508270263671875,
+      "learning_rate": 5.335982254513208e-07,
+      "loss": 0.0671,
+      "step": 14711
+    },
+    {
+      "epoch": 2.383668178872327,
+      "grad_norm": 0.8803588151931763,
+      "learning_rate": 5.333282145892493e-07,
+      "loss": 0.063,
+      "step": 14712
+    },
+    {
+      "epoch": 2.3838302009073233,
+      "grad_norm": 0.9654552340507507,
+      "learning_rate": 5.330582639025739e-07,
+      "loss": 0.0598,
+      "step": 14713
+    },
+    {
+      "epoch": 2.3839922229423203,
+      "grad_norm": 0.9189082384109497,
+      "learning_rate": 5.327883733995562e-07,
+      "loss": 0.0534,
+      "step": 14714
+    },
+    {
+      "epoch": 2.3841542449773168,
+      "grad_norm": 0.9922522306442261,
+      "learning_rate": 5.325185430884539e-07,
+      "loss": 0.0617,
+      "step": 14715
+    },
+    {
+      "epoch": 2.3843162670123137,
+      "grad_norm": 0.873923122882843,
+      "learning_rate": 5.322487729775233e-07,
+      "loss": 0.0668,
+      "step": 14716
+    },
+    {
+      "epoch": 2.3844782890473106,
+      "grad_norm": 1.0077275037765503,
+      "learning_rate": 5.319790630750182e-07,
+      "loss": 0.0597,
+      "step": 14717
+    },
+    {
+      "epoch": 2.384640311082307,
+      "grad_norm": 0.9447020888328552,
+      "learning_rate": 5.317094133891903e-07,
+      "loss": 0.0616,
+      "step": 14718
+    },
+    {
+      "epoch": 2.384802333117304,
+      "grad_norm": 1.0282572507858276,
+      "learning_rate": 5.314398239282926e-07,
+      "loss": 0.072,
+      "step": 14719
+    },
+    {
+      "epoch": 2.3849643551523005,
+      "grad_norm": 0.8626378178596497,
+      "learning_rate": 5.311702947005718e-07,
+      "loss": 0.0574,
+      "step": 14720
+    },
+    {
+      "epoch": 2.3851263771872975,
+      "grad_norm": 0.8774121403694153,
+      "learning_rate": 5.309008257142754e-07,
+      "loss": 0.0563,
+      "step": 14721
+    },
+    {
+      "epoch": 2.3852883992222944,
+      "grad_norm": 0.9489586353302002,
+      "learning_rate": 5.306314169776486e-07,
+      "loss": 0.0613,
+      "step": 14722
+    },
+    {
+      "epoch": 2.385450421257291,
+      "grad_norm": 1.0716079473495483,
+      "learning_rate": 5.303620684989347e-07,
+      "loss": 0.0602,
+      "step": 14723
+    },
+    {
+      "epoch": 2.385612443292288,
+      "grad_norm": 0.8634665012359619,
+      "learning_rate": 5.30092780286375e-07,
+      "loss": 0.0546,
+      "step": 14724
+    },
+    {
+      "epoch": 2.3857744653272843,
+      "grad_norm": 1.048065423965454,
+      "learning_rate": 5.298235523482093e-07,
+      "loss": 0.0685,
+      "step": 14725
+    },
+    {
+      "epoch": 2.3859364873622813,
+      "grad_norm": 0.8572198152542114,
+      "learning_rate": 5.295543846926752e-07,
+      "loss": 0.0559,
+      "step": 14726
+    },
+    {
+      "epoch": 2.386098509397278,
+      "grad_norm": 0.8316371440887451,
+      "learning_rate": 5.292852773280091e-07,
+      "loss": 0.0603,
+      "step": 14727
+    },
+    {
+      "epoch": 2.3862605314322747,
+      "grad_norm": 1.1684722900390625,
+      "learning_rate": 5.290162302624433e-07,
+      "loss": 0.0619,
+      "step": 14728
+    },
+    {
+      "epoch": 2.3864225534672716,
+      "grad_norm": 0.960731029510498,
+      "learning_rate": 5.287472435042116e-07,
+      "loss": 0.0612,
+      "step": 14729
+    },
+    {
+      "epoch": 2.386584575502268,
+      "grad_norm": 1.0198336839675903,
+      "learning_rate": 5.284783170615446e-07,
+      "loss": 0.0638,
+      "step": 14730
+    },
+    {
+      "epoch": 2.386746597537265,
+      "grad_norm": 0.9215663075447083,
+      "learning_rate": 5.282094509426694e-07,
+      "loss": 0.0627,
+      "step": 14731
+    },
+    {
+      "epoch": 2.386908619572262,
+      "grad_norm": 0.9803484678268433,
+      "learning_rate": 5.279406451558136e-07,
+      "loss": 0.0699,
+      "step": 14732
+    },
+    {
+      "epoch": 2.3870706416072585,
+      "grad_norm": 0.8979873061180115,
+      "learning_rate": 5.27671899709201e-07,
+      "loss": 0.053,
+      "step": 14733
+    },
+    {
+      "epoch": 2.3872326636422554,
+      "grad_norm": 1.023978352546692,
+      "learning_rate": 5.274032146110567e-07,
+      "loss": 0.0607,
+      "step": 14734
+    },
+    {
+      "epoch": 2.3873946856772523,
+      "grad_norm": 0.7679246664047241,
+      "learning_rate": 5.271345898695995e-07,
+      "loss": 0.053,
+      "step": 14735
+    },
+    {
+      "epoch": 2.387556707712249,
+      "grad_norm": 1.0310170650482178,
+      "learning_rate": 5.268660254930499e-07,
+      "loss": 0.062,
+      "step": 14736
+    },
+    {
+      "epoch": 2.3877187297472457,
+      "grad_norm": 0.8446293473243713,
+      "learning_rate": 5.265975214896249e-07,
+      "loss": 0.0596,
+      "step": 14737
+    },
+    {
+      "epoch": 2.3878807517822422,
+      "grad_norm": 1.072784662246704,
+      "learning_rate": 5.263290778675401e-07,
+      "loss": 0.0659,
+      "step": 14738
+    },
+    {
+      "epoch": 2.388042773817239,
+      "grad_norm": 0.93727707862854,
+      "learning_rate": 5.260606946350094e-07,
+      "loss": 0.0634,
+      "step": 14739
+    },
+    {
+      "epoch": 2.3882047958522357,
+      "grad_norm": 1.1457421779632568,
+      "learning_rate": 5.257923718002447e-07,
+      "loss": 0.0604,
+      "step": 14740
+    },
+    {
+      "epoch": 2.3883668178872326,
+      "grad_norm": 1.0058701038360596,
+      "learning_rate": 5.255241093714561e-07,
+      "loss": 0.0608,
+      "step": 14741
+    },
+    {
+      "epoch": 2.3885288399222295,
+      "grad_norm": 0.922507107257843,
+      "learning_rate": 5.252559073568514e-07,
+      "loss": 0.0557,
+      "step": 14742
+    },
+    {
+      "epoch": 2.388690861957226,
+      "grad_norm": 0.9533309936523438,
+      "learning_rate": 5.249877657646371e-07,
+      "loss": 0.0567,
+      "step": 14743
+    },
+    {
+      "epoch": 2.388852883992223,
+      "grad_norm": 0.8887631297111511,
+      "learning_rate": 5.247196846030178e-07,
+      "loss": 0.061,
+      "step": 14744
+    },
+    {
+      "epoch": 2.38901490602722,
+      "grad_norm": 1.0275040864944458,
+      "learning_rate": 5.244516638801966e-07,
+      "loss": 0.0604,
+      "step": 14745
+    },
+    {
+      "epoch": 2.3891769280622164,
+      "grad_norm": 0.7640145421028137,
+      "learning_rate": 5.241837036043731e-07,
+      "loss": 0.0586,
+      "step": 14746
+    },
+    {
+      "epoch": 2.3893389500972133,
+      "grad_norm": 1.0371057987213135,
+      "learning_rate": 5.239158037837464e-07,
+      "loss": 0.0581,
+      "step": 14747
+    },
+    {
+      "epoch": 2.38950097213221,
+      "grad_norm": 0.8980159163475037,
+      "learning_rate": 5.236479644265153e-07,
+      "loss": 0.0616,
+      "step": 14748
+    },
+    {
+      "epoch": 2.3896629941672067,
+      "grad_norm": 0.923527181148529,
+      "learning_rate": 5.233801855408733e-07,
+      "loss": 0.0555,
+      "step": 14749
+    },
+    {
+      "epoch": 2.3898250162022037,
+      "grad_norm": 0.9246334433555603,
+      "learning_rate": 5.231124671350141e-07,
+      "loss": 0.0641,
+      "step": 14750
+    },
+    {
+      "epoch": 2.3899870382372,
+      "grad_norm": 0.7615161538124084,
+      "learning_rate": 5.228448092171295e-07,
+      "loss": 0.0538,
+      "step": 14751
+    },
+    {
+      "epoch": 2.390149060272197,
+      "grad_norm": 0.9077004194259644,
+      "learning_rate": 5.225772117954089e-07,
+      "loss": 0.0647,
+      "step": 14752
+    },
+    {
+      "epoch": 2.3903110823071936,
+      "grad_norm": 0.9212455749511719,
+      "learning_rate": 5.223096748780407e-07,
+      "loss": 0.0559,
+      "step": 14753
+    },
+    {
+      "epoch": 2.3904731043421905,
+      "grad_norm": 0.8712791800498962,
+      "learning_rate": 5.220421984732104e-07,
+      "loss": 0.0621,
+      "step": 14754
+    },
+    {
+      "epoch": 2.3906351263771874,
+      "grad_norm": 0.9051830172538757,
+      "learning_rate": 5.217747825891023e-07,
+      "loss": 0.0596,
+      "step": 14755
+    },
+    {
+      "epoch": 2.390797148412184,
+      "grad_norm": 0.8532323241233826,
+      "learning_rate": 5.215074272338986e-07,
+      "loss": 0.0616,
+      "step": 14756
+    },
+    {
+      "epoch": 2.390959170447181,
+      "grad_norm": 0.9374991059303284,
+      "learning_rate": 5.212401324157795e-07,
+      "loss": 0.0571,
+      "step": 14757
+    },
+    {
+      "epoch": 2.391121192482178,
+      "grad_norm": 0.9487468600273132,
+      "learning_rate": 5.20972898142924e-07,
+      "loss": 0.0622,
+      "step": 14758
+    },
+    {
+      "epoch": 2.3912832145171743,
+      "grad_norm": 0.8730701208114624,
+      "learning_rate": 5.207057244235089e-07,
+      "loss": 0.0594,
+      "step": 14759
+    },
+    {
+      "epoch": 2.3914452365521712,
+      "grad_norm": 0.9014290571212769,
+      "learning_rate": 5.204386112657095e-07,
+      "loss": 0.0632,
+      "step": 14760
+    },
+    {
+      "epoch": 2.3916072585871677,
+      "grad_norm": 0.9186999797821045,
+      "learning_rate": 5.201715586776965e-07,
+      "loss": 0.0588,
+      "step": 14761
+    },
+    {
+      "epoch": 2.3917692806221647,
+      "grad_norm": 0.8001679182052612,
+      "learning_rate": 5.199045666676436e-07,
+      "loss": 0.0606,
+      "step": 14762
+    },
+    {
+      "epoch": 2.391931302657161,
+      "grad_norm": 0.8610774874687195,
+      "learning_rate": 5.196376352437199e-07,
+      "loss": 0.0522,
+      "step": 14763
+    },
+    {
+      "epoch": 2.392093324692158,
+      "grad_norm": 0.8187136650085449,
+      "learning_rate": 5.193707644140913e-07,
+      "loss": 0.0517,
+      "step": 14764
+    },
+    {
+      "epoch": 2.392255346727155,
+      "grad_norm": 1.040032148361206,
+      "learning_rate": 5.191039541869245e-07,
+      "loss": 0.0672,
+      "step": 14765
+    },
+    {
+      "epoch": 2.3924173687621515,
+      "grad_norm": 0.8961792588233948,
+      "learning_rate": 5.188372045703824e-07,
+      "loss": 0.0614,
+      "step": 14766
+    },
+    {
+      "epoch": 2.3925793907971484,
+      "grad_norm": 0.7686487436294556,
+      "learning_rate": 5.185705155726287e-07,
+      "loss": 0.0552,
+      "step": 14767
+    },
+    {
+      "epoch": 2.3927414128321454,
+      "grad_norm": 0.987948477268219,
+      "learning_rate": 5.183038872018215e-07,
+      "loss": 0.0584,
+      "step": 14768
+    },
+    {
+      "epoch": 2.392903434867142,
+      "grad_norm": 0.9744378328323364,
+      "learning_rate": 5.180373194661198e-07,
+      "loss": 0.0644,
+      "step": 14769
+    },
+    {
+      "epoch": 2.393065456902139,
+      "grad_norm": 0.9378809928894043,
+      "learning_rate": 5.1777081237368e-07,
+      "loss": 0.0642,
+      "step": 14770
+    },
+    {
+      "epoch": 2.3932274789371353,
+      "grad_norm": 1.1186909675598145,
+      "learning_rate": 5.175043659326564e-07,
+      "loss": 0.0673,
+      "step": 14771
+    },
+    {
+      "epoch": 2.393389500972132,
+      "grad_norm": 0.9698737263679504,
+      "learning_rate": 5.172379801512014e-07,
+      "loss": 0.0614,
+      "step": 14772
+    },
+    {
+      "epoch": 2.393551523007129,
+      "grad_norm": 0.985581636428833,
+      "learning_rate": 5.169716550374665e-07,
+      "loss": 0.0609,
+      "step": 14773
+    },
+    {
+      "epoch": 2.3937135450421256,
+      "grad_norm": 0.8977107405662537,
+      "learning_rate": 5.167053905996003e-07,
+      "loss": 0.0643,
+      "step": 14774
+    },
+    {
+      "epoch": 2.3938755670771226,
+      "grad_norm": 0.8801769018173218,
+      "learning_rate": 5.164391868457494e-07,
+      "loss": 0.0647,
+      "step": 14775
+    },
+    {
+      "epoch": 2.394037589112119,
+      "grad_norm": 1.0398766994476318,
+      "learning_rate": 5.161730437840585e-07,
+      "loss": 0.0658,
+      "step": 14776
+    },
+    {
+      "epoch": 2.394199611147116,
+      "grad_norm": 0.9638035297393799,
+      "learning_rate": 5.159069614226723e-07,
+      "loss": 0.0655,
+      "step": 14777
+    },
+    {
+      "epoch": 2.394361633182113,
+      "grad_norm": 1.0413271188735962,
+      "learning_rate": 5.15640939769732e-07,
+      "loss": 0.0587,
+      "step": 14778
+    },
+    {
+      "epoch": 2.3945236552171094,
+      "grad_norm": 0.9833051562309265,
+      "learning_rate": 5.153749788333767e-07,
+      "loss": 0.0633,
+      "step": 14779
+    },
+    {
+      "epoch": 2.3946856772521063,
+      "grad_norm": 0.9395384192466736,
+      "learning_rate": 5.151090786217433e-07,
+      "loss": 0.0647,
+      "step": 14780
+    },
+    {
+      "epoch": 2.3948476992871033,
+      "grad_norm": 0.9286221265792847,
+      "learning_rate": 5.148432391429703e-07,
+      "loss": 0.0617,
+      "step": 14781
+    },
+    {
+      "epoch": 2.3950097213220998,
+      "grad_norm": 0.8092193603515625,
+      "learning_rate": 5.145774604051895e-07,
+      "loss": 0.0587,
+      "step": 14782
+    },
+    {
+      "epoch": 2.3951717433570967,
+      "grad_norm": 0.9903756380081177,
+      "learning_rate": 5.143117424165339e-07,
+      "loss": 0.0611,
+      "step": 14783
+    },
+    {
+      "epoch": 2.395333765392093,
+      "grad_norm": 0.8078405261039734,
+      "learning_rate": 5.140460851851336e-07,
+      "loss": 0.0596,
+      "step": 14784
+    },
+    {
+      "epoch": 2.39549578742709,
+      "grad_norm": 1.076858639717102,
+      "learning_rate": 5.137804887191172e-07,
+      "loss": 0.0657,
+      "step": 14785
+    },
+    {
+      "epoch": 2.3956578094620866,
+      "grad_norm": 1.0425070524215698,
+      "learning_rate": 5.135149530266112e-07,
+      "loss": 0.0613,
+      "step": 14786
+    },
+    {
+      "epoch": 2.3958198314970836,
+      "grad_norm": 0.9560631513595581,
+      "learning_rate": 5.132494781157407e-07,
+      "loss": 0.064,
+      "step": 14787
+    },
+    {
+      "epoch": 2.3959818535320805,
+      "grad_norm": 0.765007495880127,
+      "learning_rate": 5.129840639946279e-07,
+      "loss": 0.0555,
+      "step": 14788
+    },
+    {
+      "epoch": 2.396143875567077,
+      "grad_norm": 0.9489015936851501,
+      "learning_rate": 5.127187106713951e-07,
+      "loss": 0.0601,
+      "step": 14789
+    },
+    {
+      "epoch": 2.396305897602074,
+      "grad_norm": 0.8428517580032349,
+      "learning_rate": 5.124534181541596e-07,
+      "loss": 0.0542,
+      "step": 14790
+    },
+    {
+      "epoch": 2.396467919637071,
+      "grad_norm": 0.8759004473686218,
+      "learning_rate": 5.1218818645104e-07,
+      "loss": 0.0604,
+      "step": 14791
+    },
+    {
+      "epoch": 2.3966299416720673,
+      "grad_norm": 0.9745606184005737,
+      "learning_rate": 5.119230155701515e-07,
+      "loss": 0.0633,
+      "step": 14792
+    },
+    {
+      "epoch": 2.3967919637070643,
+      "grad_norm": 0.8945474624633789,
+      "learning_rate": 5.116579055196085e-07,
+      "loss": 0.0614,
+      "step": 14793
+    },
+    {
+      "epoch": 2.3969539857420608,
+      "grad_norm": 0.9640982747077942,
+      "learning_rate": 5.113928563075213e-07,
+      "loss": 0.0664,
+      "step": 14794
+    },
+    {
+      "epoch": 2.3971160077770577,
+      "grad_norm": 0.8617674112319946,
+      "learning_rate": 5.111278679419996e-07,
+      "loss": 0.0607,
+      "step": 14795
+    },
+    {
+      "epoch": 2.3972780298120546,
+      "grad_norm": 1.115388035774231,
+      "learning_rate": 5.108629404311535e-07,
+      "loss": 0.0675,
+      "step": 14796
+    },
+    {
+      "epoch": 2.397440051847051,
+      "grad_norm": 0.9158341288566589,
+      "learning_rate": 5.105980737830871e-07,
+      "loss": 0.0631,
+      "step": 14797
+    },
+    {
+      "epoch": 2.397602073882048,
+      "grad_norm": 0.93878173828125,
+      "learning_rate": 5.103332680059053e-07,
+      "loss": 0.0581,
+      "step": 14798
+    },
+    {
+      "epoch": 2.3977640959170445,
+      "grad_norm": 0.8788214921951294,
+      "learning_rate": 5.100685231077107e-07,
+      "loss": 0.061,
+      "step": 14799
+    },
+    {
+      "epoch": 2.3979261179520415,
+      "grad_norm": 0.8120740056037903,
+      "learning_rate": 5.098038390966039e-07,
+      "loss": 0.0494,
+      "step": 14800
+    },
+    {
+      "epoch": 2.3980881399870384,
+      "grad_norm": 0.9346016645431519,
+      "learning_rate": 5.095392159806833e-07,
+      "loss": 0.0663,
+      "step": 14801
+    },
+    {
+      "epoch": 2.398250162022035,
+      "grad_norm": 1.0751339197158813,
+      "learning_rate": 5.09274653768046e-07,
+      "loss": 0.0646,
+      "step": 14802
+    },
+    {
+      "epoch": 2.398412184057032,
+      "grad_norm": 0.9000687003135681,
+      "learning_rate": 5.090101524667865e-07,
+      "loss": 0.0632,
+      "step": 14803
+    },
+    {
+      "epoch": 2.3985742060920288,
+      "grad_norm": 0.83973628282547,
+      "learning_rate": 5.087457120849984e-07,
+      "loss": 0.0593,
+      "step": 14804
+    },
+    {
+      "epoch": 2.3987362281270252,
+      "grad_norm": 0.8891993165016174,
+      "learning_rate": 5.084813326307728e-07,
+      "loss": 0.0607,
+      "step": 14805
+    },
+    {
+      "epoch": 2.398898250162022,
+      "grad_norm": 0.8002360463142395,
+      "learning_rate": 5.082170141121992e-07,
+      "loss": 0.0532,
+      "step": 14806
+    },
+    {
+      "epoch": 2.3990602721970187,
+      "grad_norm": 0.980747401714325,
+      "learning_rate": 5.079527565373654e-07,
+      "loss": 0.0631,
+      "step": 14807
+    },
+    {
+      "epoch": 2.3992222942320156,
+      "grad_norm": 0.9085080027580261,
+      "learning_rate": 5.076885599143558e-07,
+      "loss": 0.0628,
+      "step": 14808
+    },
+    {
+      "epoch": 2.399384316267012,
+      "grad_norm": 0.9033792614936829,
+      "learning_rate": 5.074244242512546e-07,
+      "loss": 0.0585,
+      "step": 14809
+    },
+    {
+      "epoch": 2.399546338302009,
+      "grad_norm": 0.9455671310424805,
+      "learning_rate": 5.071603495561444e-07,
+      "loss": 0.0657,
+      "step": 14810
+    },
+    {
+      "epoch": 2.399708360337006,
+      "grad_norm": 1.0358432531356812,
+      "learning_rate": 5.068963358371059e-07,
+      "loss": 0.064,
+      "step": 14811
+    },
+    {
+      "epoch": 2.3998703823720025,
+      "grad_norm": 0.968040943145752,
+      "learning_rate": 5.066323831022155e-07,
+      "loss": 0.0686,
+      "step": 14812
+    },
+    {
+      "epoch": 2.4000324044069994,
+      "grad_norm": 0.8224153518676758,
+      "learning_rate": 5.063684913595504e-07,
+      "loss": 0.057,
+      "step": 14813
+    },
+    {
+      "epoch": 2.4001944264419963,
+      "grad_norm": 0.9839720726013184,
+      "learning_rate": 5.061046606171849e-07,
+      "loss": 0.0606,
+      "step": 14814
+    },
+    {
+      "epoch": 2.400356448476993,
+      "grad_norm": 0.8918019533157349,
+      "learning_rate": 5.058408908831919e-07,
+      "loss": 0.0628,
+      "step": 14815
+    },
+    {
+      "epoch": 2.4005184705119897,
+      "grad_norm": 0.9705418348312378,
+      "learning_rate": 5.055771821656416e-07,
+      "loss": 0.0527,
+      "step": 14816
+    },
+    {
+      "epoch": 2.4006804925469862,
+      "grad_norm": 1.0187605619430542,
+      "learning_rate": 5.053135344726032e-07,
+      "loss": 0.0579,
+      "step": 14817
+    },
+    {
+      "epoch": 2.400842514581983,
+      "grad_norm": 0.8986850380897522,
+      "learning_rate": 5.05049947812144e-07,
+      "loss": 0.0617,
+      "step": 14818
+    },
+    {
+      "epoch": 2.40100453661698,
+      "grad_norm": 0.9489282965660095,
+      "learning_rate": 5.047864221923276e-07,
+      "loss": 0.0595,
+      "step": 14819
+    },
+    {
+      "epoch": 2.4011665586519766,
+      "grad_norm": 0.8312699794769287,
+      "learning_rate": 5.045229576212191e-07,
+      "loss": 0.0587,
+      "step": 14820
+    },
+    {
+      "epoch": 2.4013285806869735,
+      "grad_norm": 0.8826655149459839,
+      "learning_rate": 5.04259554106879e-07,
+      "loss": 0.062,
+      "step": 14821
+    },
+    {
+      "epoch": 2.40149060272197,
+      "grad_norm": 1.0575923919677734,
+      "learning_rate": 5.039962116573676e-07,
+      "loss": 0.0738,
+      "step": 14822
+    },
+    {
+      "epoch": 2.401652624756967,
+      "grad_norm": 0.8363692164421082,
+      "learning_rate": 5.037329302807409e-07,
+      "loss": 0.0583,
+      "step": 14823
+    },
+    {
+      "epoch": 2.401814646791964,
+      "grad_norm": 0.9132293462753296,
+      "learning_rate": 5.034697099850557e-07,
+      "loss": 0.066,
+      "step": 14824
+    },
+    {
+      "epoch": 2.4019766688269604,
+      "grad_norm": 0.8294099569320679,
+      "learning_rate": 5.032065507783671e-07,
+      "loss": 0.0549,
+      "step": 14825
+    },
+    {
+      "epoch": 2.4021386908619573,
+      "grad_norm": 0.8338202834129333,
+      "learning_rate": 5.029434526687249e-07,
+      "loss": 0.0594,
+      "step": 14826
+    },
+    {
+      "epoch": 2.402300712896954,
+      "grad_norm": 0.9756371378898621,
+      "learning_rate": 5.026804156641804e-07,
+      "loss": 0.0635,
+      "step": 14827
+    },
+    {
+      "epoch": 2.4024627349319507,
+      "grad_norm": 0.9361462593078613,
+      "learning_rate": 5.02417439772781e-07,
+      "loss": 0.0661,
+      "step": 14828
+    },
+    {
+      "epoch": 2.4026247569669477,
+      "grad_norm": 0.8787729144096375,
+      "learning_rate": 5.021545250025755e-07,
+      "loss": 0.0638,
+      "step": 14829
+    },
+    {
+      "epoch": 2.402786779001944,
+      "grad_norm": 0.8789749145507812,
+      "learning_rate": 5.01891671361606e-07,
+      "loss": 0.052,
+      "step": 14830
+    },
+    {
+      "epoch": 2.402948801036941,
+      "grad_norm": 0.8396589756011963,
+      "learning_rate": 5.01628878857916e-07,
+      "loss": 0.0594,
+      "step": 14831
+    },
+    {
+      "epoch": 2.4031108230719376,
+      "grad_norm": 0.9573028683662415,
+      "learning_rate": 5.013661474995463e-07,
+      "loss": 0.0675,
+      "step": 14832
+    },
+    {
+      "epoch": 2.4032728451069345,
+      "grad_norm": 0.9700629115104675,
+      "learning_rate": 5.011034772945359e-07,
+      "loss": 0.0628,
+      "step": 14833
+    },
+    {
+      "epoch": 2.4034348671419314,
+      "grad_norm": 1.1622406244277954,
+      "learning_rate": 5.008408682509219e-07,
+      "loss": 0.0643,
+      "step": 14834
+    },
+    {
+      "epoch": 2.403596889176928,
+      "grad_norm": 0.8823803067207336,
+      "learning_rate": 5.005783203767394e-07,
+      "loss": 0.0606,
+      "step": 14835
+    },
+    {
+      "epoch": 2.403758911211925,
+      "grad_norm": 0.8871647715568542,
+      "learning_rate": 5.003158336800218e-07,
+      "loss": 0.0567,
+      "step": 14836
+    },
+    {
+      "epoch": 2.403920933246922,
+      "grad_norm": 0.9583814740180969,
+      "learning_rate": 5.000534081688013e-07,
+      "loss": 0.0589,
+      "step": 14837
+    },
+    {
+      "epoch": 2.4040829552819183,
+      "grad_norm": 1.0450133085250854,
+      "learning_rate": 4.997910438511052e-07,
+      "loss": 0.0643,
+      "step": 14838
+    },
+    {
+      "epoch": 2.404244977316915,
+      "grad_norm": 1.106645107269287,
+      "learning_rate": 4.995287407349636e-07,
+      "loss": 0.069,
+      "step": 14839
+    },
+    {
+      "epoch": 2.4044069993519117,
+      "grad_norm": 0.9293127059936523,
+      "learning_rate": 4.992664988284021e-07,
+      "loss": 0.0619,
+      "step": 14840
+    },
+    {
+      "epoch": 2.4045690213869086,
+      "grad_norm": 0.8363195061683655,
+      "learning_rate": 4.990043181394433e-07,
+      "loss": 0.0611,
+      "step": 14841
+    },
+    {
+      "epoch": 2.404731043421905,
+      "grad_norm": 0.9512073993682861,
+      "learning_rate": 4.987421986761101e-07,
+      "loss": 0.0591,
+      "step": 14842
+    },
+    {
+      "epoch": 2.404893065456902,
+      "grad_norm": 0.8879616260528564,
+      "learning_rate": 4.984801404464229e-07,
+      "loss": 0.0593,
+      "step": 14843
+    },
+    {
+      "epoch": 2.405055087491899,
+      "grad_norm": 0.8494080901145935,
+      "learning_rate": 4.982181434583996e-07,
+      "loss": 0.0583,
+      "step": 14844
+    },
+    {
+      "epoch": 2.4052171095268955,
+      "grad_norm": 0.8918368816375732,
+      "learning_rate": 4.97956207720057e-07,
+      "loss": 0.0637,
+      "step": 14845
+    },
+    {
+      "epoch": 2.4053791315618924,
+      "grad_norm": 0.982917070388794,
+      "learning_rate": 4.976943332394093e-07,
+      "loss": 0.0662,
+      "step": 14846
+    },
+    {
+      "epoch": 2.4055411535968894,
+      "grad_norm": 0.9451669454574585,
+      "learning_rate": 4.974325200244698e-07,
+      "loss": 0.0609,
+      "step": 14847
+    },
+    {
+      "epoch": 2.405703175631886,
+      "grad_norm": 0.956188976764679,
+      "learning_rate": 4.971707680832491e-07,
+      "loss": 0.0558,
+      "step": 14848
+    },
+    {
+      "epoch": 2.405865197666883,
+      "grad_norm": 0.9561905860900879,
+      "learning_rate": 4.969090774237559e-07,
+      "loss": 0.0601,
+      "step": 14849
+    },
+    {
+      "epoch": 2.4060272197018793,
+      "grad_norm": 1.0489095449447632,
+      "learning_rate": 4.966474480539976e-07,
+      "loss": 0.064,
+      "step": 14850
+    },
+    {
+      "epoch": 2.406189241736876,
+      "grad_norm": 0.8939489126205444,
+      "learning_rate": 4.963858799819802e-07,
+      "loss": 0.0586,
+      "step": 14851
+    },
+    {
+      "epoch": 2.406351263771873,
+      "grad_norm": 0.8735487461090088,
+      "learning_rate": 4.961243732157048e-07,
+      "loss": 0.065,
+      "step": 14852
+    },
+    {
+      "epoch": 2.4065132858068696,
+      "grad_norm": 0.900790810585022,
+      "learning_rate": 4.958629277631749e-07,
+      "loss": 0.0603,
+      "step": 14853
+    },
+    {
+      "epoch": 2.4066753078418666,
+      "grad_norm": 0.991608738899231,
+      "learning_rate": 4.956015436323897e-07,
+      "loss": 0.0592,
+      "step": 14854
+    },
+    {
+      "epoch": 2.406837329876863,
+      "grad_norm": 0.8543813228607178,
+      "learning_rate": 4.953402208313471e-07,
+      "loss": 0.0651,
+      "step": 14855
+    },
+    {
+      "epoch": 2.40699935191186,
+      "grad_norm": 0.8871439695358276,
+      "learning_rate": 4.950789593680422e-07,
+      "loss": 0.0639,
+      "step": 14856
+    },
+    {
+      "epoch": 2.407161373946857,
+      "grad_norm": 0.9490445852279663,
+      "learning_rate": 4.948177592504682e-07,
+      "loss": 0.0669,
+      "step": 14857
+    },
+    {
+      "epoch": 2.4073233959818534,
+      "grad_norm": 0.8415353894233704,
+      "learning_rate": 4.945566204866201e-07,
+      "loss": 0.0585,
+      "step": 14858
+    },
+    {
+      "epoch": 2.4074854180168503,
+      "grad_norm": 1.0602666139602661,
+      "learning_rate": 4.942955430844856e-07,
+      "loss": 0.0623,
+      "step": 14859
+    },
+    {
+      "epoch": 2.4076474400518473,
+      "grad_norm": 0.8117191791534424,
+      "learning_rate": 4.940345270520536e-07,
+      "loss": 0.0591,
+      "step": 14860
+    },
+    {
+      "epoch": 2.4078094620868438,
+      "grad_norm": 0.9509507417678833,
+      "learning_rate": 4.937735723973109e-07,
+      "loss": 0.065,
+      "step": 14861
+    },
+    {
+      "epoch": 2.4079714841218407,
+      "grad_norm": 0.9605535268783569,
+      "learning_rate": 4.935126791282419e-07,
+      "loss": 0.0678,
+      "step": 14862
+    },
+    {
+      "epoch": 2.408133506156837,
+      "grad_norm": 0.9226017594337463,
+      "learning_rate": 4.932518472528292e-07,
+      "loss": 0.0626,
+      "step": 14863
+    },
+    {
+      "epoch": 2.408295528191834,
+      "grad_norm": 0.8513705730438232,
+      "learning_rate": 4.929910767790536e-07,
+      "loss": 0.0593,
+      "step": 14864
+    },
+    {
+      "epoch": 2.4084575502268306,
+      "grad_norm": 0.7933249473571777,
+      "learning_rate": 4.927303677148942e-07,
+      "loss": 0.054,
+      "step": 14865
+    },
+    {
+      "epoch": 2.4086195722618275,
+      "grad_norm": 0.8183956146240234,
+      "learning_rate": 4.92469720068329e-07,
+      "loss": 0.0534,
+      "step": 14866
+    },
+    {
+      "epoch": 2.4087815942968245,
+      "grad_norm": 1.0017709732055664,
+      "learning_rate": 4.922091338473309e-07,
+      "loss": 0.0691,
+      "step": 14867
+    },
+    {
+      "epoch": 2.408943616331821,
+      "grad_norm": 1.0862891674041748,
+      "learning_rate": 4.919486090598749e-07,
+      "loss": 0.0652,
+      "step": 14868
+    },
+    {
+      "epoch": 2.409105638366818,
+      "grad_norm": 0.8221504092216492,
+      "learning_rate": 4.91688145713933e-07,
+      "loss": 0.0558,
+      "step": 14869
+    },
+    {
+      "epoch": 2.409267660401815,
+      "grad_norm": 1.0607209205627441,
+      "learning_rate": 4.91427743817473e-07,
+      "loss": 0.0589,
+      "step": 14870
+    },
+    {
+      "epoch": 2.4094296824368113,
+      "grad_norm": 0.8373808264732361,
+      "learning_rate": 4.911674033784628e-07,
+      "loss": 0.0584,
+      "step": 14871
+    },
+    {
+      "epoch": 2.4095917044718083,
+      "grad_norm": 0.8616558909416199,
+      "learning_rate": 4.909071244048694e-07,
+      "loss": 0.0557,
+      "step": 14872
+    },
+    {
+      "epoch": 2.4097537265068047,
+      "grad_norm": 0.8790077567100525,
+      "learning_rate": 4.906469069046568e-07,
+      "loss": 0.0644,
+      "step": 14873
+    },
+    {
+      "epoch": 2.4099157485418017,
+      "grad_norm": 0.9295965433120728,
+      "learning_rate": 4.903867508857857e-07,
+      "loss": 0.0584,
+      "step": 14874
+    },
+    {
+      "epoch": 2.4100777705767986,
+      "grad_norm": 0.8102318048477173,
+      "learning_rate": 4.901266563562168e-07,
+      "loss": 0.0586,
+      "step": 14875
+    },
+    {
+      "epoch": 2.410239792611795,
+      "grad_norm": 0.8939180970191956,
+      "learning_rate": 4.898666233239083e-07,
+      "loss": 0.0602,
+      "step": 14876
+    },
+    {
+      "epoch": 2.410401814646792,
+      "grad_norm": 0.8693578839302063,
+      "learning_rate": 4.896066517968167e-07,
+      "loss": 0.0623,
+      "step": 14877
+    },
+    {
+      "epoch": 2.4105638366817885,
+      "grad_norm": 0.9918479323387146,
+      "learning_rate": 4.893467417828967e-07,
+      "loss": 0.0659,
+      "step": 14878
+    },
+    {
+      "epoch": 2.4107258587167855,
+      "grad_norm": 0.8719764351844788,
+      "learning_rate": 4.890868932901005e-07,
+      "loss": 0.0604,
+      "step": 14879
+    },
+    {
+      "epoch": 2.4108878807517824,
+      "grad_norm": 0.8317460417747498,
+      "learning_rate": 4.888271063263791e-07,
+      "loss": 0.0602,
+      "step": 14880
+    },
+    {
+      "epoch": 2.411049902786779,
+      "grad_norm": 1.0084984302520752,
+      "learning_rate": 4.885673808996816e-07,
+      "loss": 0.0651,
+      "step": 14881
+    },
+    {
+      "epoch": 2.411211924821776,
+      "grad_norm": 0.9814823269844055,
+      "learning_rate": 4.883077170179542e-07,
+      "loss": 0.0678,
+      "step": 14882
+    },
+    {
+      "epoch": 2.4113739468567728,
+      "grad_norm": 0.9129720330238342,
+      "learning_rate": 4.880481146891428e-07,
+      "loss": 0.0585,
+      "step": 14883
+    },
+    {
+      "epoch": 2.4115359688917692,
+      "grad_norm": 0.827803909778595,
+      "learning_rate": 4.877885739211907e-07,
+      "loss": 0.0594,
+      "step": 14884
+    },
+    {
+      "epoch": 2.411697990926766,
+      "grad_norm": 1.2343494892120361,
+      "learning_rate": 4.875290947220382e-07,
+      "loss": 0.0581,
+      "step": 14885
+    },
+    {
+      "epoch": 2.4118600129617627,
+      "grad_norm": 0.9479061961174011,
+      "learning_rate": 4.872696770996246e-07,
+      "loss": 0.0645,
+      "step": 14886
+    },
+    {
+      "epoch": 2.4120220349967596,
+      "grad_norm": 0.9281307458877563,
+      "learning_rate": 4.870103210618895e-07,
+      "loss": 0.0629,
+      "step": 14887
+    },
+    {
+      "epoch": 2.412184057031756,
+      "grad_norm": 0.9015679359436035,
+      "learning_rate": 4.867510266167669e-07,
+      "loss": 0.0598,
+      "step": 14888
+    },
+    {
+      "epoch": 2.412346079066753,
+      "grad_norm": 1.0281862020492554,
+      "learning_rate": 4.864917937721905e-07,
+      "loss": 0.0627,
+      "step": 14889
+    },
+    {
+      "epoch": 2.41250810110175,
+      "grad_norm": 0.7588191032409668,
+      "learning_rate": 4.862326225360927e-07,
+      "loss": 0.0536,
+      "step": 14890
+    },
+    {
+      "epoch": 2.4126701231367464,
+      "grad_norm": 0.9489138722419739,
+      "learning_rate": 4.859735129164036e-07,
+      "loss": 0.0613,
+      "step": 14891
+    },
+    {
+      "epoch": 2.4128321451717434,
+      "grad_norm": 0.9236519932746887,
+      "learning_rate": 4.85714464921051e-07,
+      "loss": 0.0605,
+      "step": 14892
+    },
+    {
+      "epoch": 2.4129941672067403,
+      "grad_norm": 0.9763721823692322,
+      "learning_rate": 4.854554785579613e-07,
+      "loss": 0.0568,
+      "step": 14893
+    },
+    {
+      "epoch": 2.413156189241737,
+      "grad_norm": 0.9076926112174988,
+      "learning_rate": 4.851965538350589e-07,
+      "loss": 0.0613,
+      "step": 14894
+    },
+    {
+      "epoch": 2.4133182112767337,
+      "grad_norm": 0.8788256645202637,
+      "learning_rate": 4.849376907602662e-07,
+      "loss": 0.0606,
+      "step": 14895
+    },
+    {
+      "epoch": 2.4134802333117302,
+      "grad_norm": 0.9264197945594788,
+      "learning_rate": 4.846788893415038e-07,
+      "loss": 0.0653,
+      "step": 14896
+    },
+    {
+      "epoch": 2.413642255346727,
+      "grad_norm": 0.8280066847801208,
+      "learning_rate": 4.844201495866904e-07,
+      "loss": 0.0616,
+      "step": 14897
+    },
+    {
+      "epoch": 2.413804277381724,
+      "grad_norm": 0.9356870055198669,
+      "learning_rate": 4.841614715037429e-07,
+      "loss": 0.0621,
+      "step": 14898
+    },
+    {
+      "epoch": 2.4139662994167206,
+      "grad_norm": 0.9848315715789795,
+      "learning_rate": 4.839028551005767e-07,
+      "loss": 0.0644,
+      "step": 14899
+    },
+    {
+      "epoch": 2.4141283214517175,
+      "grad_norm": 0.9987033605575562,
+      "learning_rate": 4.83644300385103e-07,
+      "loss": 0.0686,
+      "step": 14900
+    },
+    {
+      "epoch": 2.414290343486714,
+      "grad_norm": 0.8517066836357117,
+      "learning_rate": 4.833858073652351e-07,
+      "loss": 0.0565,
+      "step": 14901
+    },
+    {
+      "epoch": 2.414452365521711,
+      "grad_norm": 0.9564343094825745,
+      "learning_rate": 4.831273760488816e-07,
+      "loss": 0.0659,
+      "step": 14902
+    },
+    {
+      "epoch": 2.414614387556708,
+      "grad_norm": 0.9652625322341919,
+      "learning_rate": 4.828690064439492e-07,
+      "loss": 0.0632,
+      "step": 14903
+    },
+    {
+      "epoch": 2.4147764095917044,
+      "grad_norm": 0.9745001792907715,
+      "learning_rate": 4.82610698558344e-07,
+      "loss": 0.0628,
+      "step": 14904
+    },
+    {
+      "epoch": 2.4149384316267013,
+      "grad_norm": 0.9366600513458252,
+      "learning_rate": 4.823524523999685e-07,
+      "loss": 0.0614,
+      "step": 14905
+    },
+    {
+      "epoch": 2.4151004536616982,
+      "grad_norm": 0.8678472638130188,
+      "learning_rate": 4.820942679767268e-07,
+      "loss": 0.0553,
+      "step": 14906
+    },
+    {
+      "epoch": 2.4152624756966947,
+      "grad_norm": 1.0045222043991089,
+      "learning_rate": 4.818361452965165e-07,
+      "loss": 0.0626,
+      "step": 14907
+    },
+    {
+      "epoch": 2.4154244977316917,
+      "grad_norm": 0.8525367379188538,
+      "learning_rate": 4.815780843672366e-07,
+      "loss": 0.0577,
+      "step": 14908
+    },
+    {
+      "epoch": 2.415586519766688,
+      "grad_norm": 0.8319639563560486,
+      "learning_rate": 4.813200851967826e-07,
+      "loss": 0.052,
+      "step": 14909
+    },
+    {
+      "epoch": 2.415748541801685,
+      "grad_norm": 0.8100563287734985,
+      "learning_rate": 4.810621477930488e-07,
+      "loss": 0.0592,
+      "step": 14910
+    },
+    {
+      "epoch": 2.4159105638366816,
+      "grad_norm": 0.8080427050590515,
+      "learning_rate": 4.808042721639275e-07,
+      "loss": 0.0556,
+      "step": 14911
+    },
+    {
+      "epoch": 2.4160725858716785,
+      "grad_norm": 1.0545318126678467,
+      "learning_rate": 4.805464583173094e-07,
+      "loss": 0.0604,
+      "step": 14912
+    },
+    {
+      "epoch": 2.4162346079066754,
+      "grad_norm": 1.0038424730300903,
+      "learning_rate": 4.802887062610831e-07,
+      "loss": 0.064,
+      "step": 14913
+    },
+    {
+      "epoch": 2.416396629941672,
+      "grad_norm": 0.9150903224945068,
+      "learning_rate": 4.800310160031335e-07,
+      "loss": 0.0612,
+      "step": 14914
+    },
+    {
+      "epoch": 2.416558651976669,
+      "grad_norm": 1.0144057273864746,
+      "learning_rate": 4.797733875513475e-07,
+      "loss": 0.0604,
+      "step": 14915
+    },
+    {
+      "epoch": 2.416720674011666,
+      "grad_norm": 0.8084055781364441,
+      "learning_rate": 4.795158209136067e-07,
+      "loss": 0.0516,
+      "step": 14916
+    },
+    {
+      "epoch": 2.4168826960466623,
+      "grad_norm": 1.1354771852493286,
+      "learning_rate": 4.792583160977929e-07,
+      "loss": 0.0603,
+      "step": 14917
+    },
+    {
+      "epoch": 2.417044718081659,
+      "grad_norm": 0.9197603464126587,
+      "learning_rate": 4.79000873111784e-07,
+      "loss": 0.0604,
+      "step": 14918
+    },
+    {
+      "epoch": 2.4172067401166557,
+      "grad_norm": 0.8190104961395264,
+      "learning_rate": 4.787434919634571e-07,
+      "loss": 0.0541,
+      "step": 14919
+    },
+    {
+      "epoch": 2.4173687621516526,
+      "grad_norm": 0.9333441853523254,
+      "learning_rate": 4.784861726606893e-07,
+      "loss": 0.0566,
+      "step": 14920
+    },
+    {
+      "epoch": 2.4175307841866496,
+      "grad_norm": 0.7886890172958374,
+      "learning_rate": 4.782289152113518e-07,
+      "loss": 0.0567,
+      "step": 14921
+    },
+    {
+      "epoch": 2.417692806221646,
+      "grad_norm": 0.9538326859474182,
+      "learning_rate": 4.779717196233169e-07,
+      "loss": 0.0607,
+      "step": 14922
+    },
+    {
+      "epoch": 2.417854828256643,
+      "grad_norm": 0.8618488311767578,
+      "learning_rate": 4.777145859044543e-07,
+      "loss": 0.0589,
+      "step": 14923
+    },
+    {
+      "epoch": 2.4180168502916395,
+      "grad_norm": 0.9344967007637024,
+      "learning_rate": 4.774575140626317e-07,
+      "loss": 0.0666,
+      "step": 14924
+    },
+    {
+      "epoch": 2.4181788723266364,
+      "grad_norm": 1.032467007637024,
+      "learning_rate": 4.772005041057146e-07,
+      "loss": 0.0656,
+      "step": 14925
+    },
+    {
+      "epoch": 2.4183408943616334,
+      "grad_norm": 0.7819418907165527,
+      "learning_rate": 4.769435560415666e-07,
+      "loss": 0.0565,
+      "step": 14926
+    },
+    {
+      "epoch": 2.41850291639663,
+      "grad_norm": 0.8860273361206055,
+      "learning_rate": 4.766866698780506e-07,
+      "loss": 0.062,
+      "step": 14927
+    },
+    {
+      "epoch": 2.4186649384316268,
+      "grad_norm": 1.2574406862258911,
+      "learning_rate": 4.764298456230265e-07,
+      "loss": 0.059,
+      "step": 14928
+    },
+    {
+      "epoch": 2.4188269604666237,
+      "grad_norm": 0.9362139105796814,
+      "learning_rate": 4.7617308328435115e-07,
+      "loss": 0.0599,
+      "step": 14929
+    },
+    {
+      "epoch": 2.41898898250162,
+      "grad_norm": 1.0488052368164062,
+      "learning_rate": 4.7591638286988234e-07,
+      "loss": 0.0639,
+      "step": 14930
+    },
+    {
+      "epoch": 2.419151004536617,
+      "grad_norm": 0.9460102915763855,
+      "learning_rate": 4.756597443874747e-07,
+      "loss": 0.0642,
+      "step": 14931
+    },
+    {
+      "epoch": 2.4193130265716136,
+      "grad_norm": 0.952669084072113,
+      "learning_rate": 4.754031678449794e-07,
+      "loss": 0.0558,
+      "step": 14932
+    },
+    {
+      "epoch": 2.4194750486066106,
+      "grad_norm": 0.9806849360466003,
+      "learning_rate": 4.75146653250248e-07,
+      "loss": 0.066,
+      "step": 14933
+    },
+    {
+      "epoch": 2.419637070641607,
+      "grad_norm": 0.9641299247741699,
+      "learning_rate": 4.7489020061112805e-07,
+      "loss": 0.0566,
+      "step": 14934
+    },
+    {
+      "epoch": 2.419799092676604,
+      "grad_norm": 0.7902235984802246,
+      "learning_rate": 4.746338099354686e-07,
+      "loss": 0.0521,
+      "step": 14935
+    },
+    {
+      "epoch": 2.419961114711601,
+      "grad_norm": 0.8686903715133667,
+      "learning_rate": 4.743774812311125e-07,
+      "loss": 0.059,
+      "step": 14936
+    },
+    {
+      "epoch": 2.4201231367465974,
+      "grad_norm": 0.8258570432662964,
+      "learning_rate": 4.7412121450590374e-07,
+      "loss": 0.0544,
+      "step": 14937
+    },
+    {
+      "epoch": 2.4202851587815943,
+      "grad_norm": 0.9893523454666138,
+      "learning_rate": 4.7386500976768337e-07,
+      "loss": 0.0602,
+      "step": 14938
+    },
+    {
+      "epoch": 2.4204471808165913,
+      "grad_norm": 0.9651594758033752,
+      "learning_rate": 4.7360886702429056e-07,
+      "loss": 0.0649,
+      "step": 14939
+    },
+    {
+      "epoch": 2.4206092028515878,
+      "grad_norm": 0.8819870352745056,
+      "learning_rate": 4.733527862835624e-07,
+      "loss": 0.0593,
+      "step": 14940
+    },
+    {
+      "epoch": 2.4207712248865847,
+      "grad_norm": 0.9650877714157104,
+      "learning_rate": 4.730967675533346e-07,
+      "loss": 0.0641,
+      "step": 14941
+    },
+    {
+      "epoch": 2.420933246921581,
+      "grad_norm": 0.8652887344360352,
+      "learning_rate": 4.728408108414409e-07,
+      "loss": 0.0628,
+      "step": 14942
+    },
+    {
+      "epoch": 2.421095268956578,
+      "grad_norm": 1.117307186126709,
+      "learning_rate": 4.7258491615571277e-07,
+      "loss": 0.0695,
+      "step": 14943
+    },
+    {
+      "epoch": 2.4212572909915746,
+      "grad_norm": 0.9765766859054565,
+      "learning_rate": 4.7232908350397984e-07,
+      "loss": 0.0574,
+      "step": 14944
+    },
+    {
+      "epoch": 2.4214193130265715,
+      "grad_norm": 0.8039732575416565,
+      "learning_rate": 4.720733128940699e-07,
+      "loss": 0.0507,
+      "step": 14945
+    },
+    {
+      "epoch": 2.4215813350615685,
+      "grad_norm": 0.8668411374092102,
+      "learning_rate": 4.7181760433381017e-07,
+      "loss": 0.0637,
+      "step": 14946
+    },
+    {
+      "epoch": 2.421743357096565,
+      "grad_norm": 0.9708417654037476,
+      "learning_rate": 4.715619578310227e-07,
+      "loss": 0.0674,
+      "step": 14947
+    },
+    {
+      "epoch": 2.421905379131562,
+      "grad_norm": 0.8565574884414673,
+      "learning_rate": 4.7130637339352995e-07,
+      "loss": 0.0617,
+      "step": 14948
+    },
+    {
+      "epoch": 2.422067401166559,
+      "grad_norm": 0.8651789426803589,
+      "learning_rate": 4.7105085102915365e-07,
+      "loss": 0.0635,
+      "step": 14949
+    },
+    {
+      "epoch": 2.4222294232015553,
+      "grad_norm": 0.9995070695877075,
+      "learning_rate": 4.707953907457119e-07,
+      "loss": 0.0622,
+      "step": 14950
+    },
+    {
+      "epoch": 2.4223914452365523,
+      "grad_norm": 0.948974072933197,
+      "learning_rate": 4.7053999255101987e-07,
+      "loss": 0.0637,
+      "step": 14951
+    },
+    {
+      "epoch": 2.4225534672715487,
+      "grad_norm": 0.8724423050880432,
+      "learning_rate": 4.702846564528929e-07,
+      "loss": 0.0648,
+      "step": 14952
+    },
+    {
+      "epoch": 2.4227154893065457,
+      "grad_norm": 0.8870061039924622,
+      "learning_rate": 4.700293824591437e-07,
+      "loss": 0.0611,
+      "step": 14953
+    },
+    {
+      "epoch": 2.4228775113415426,
+      "grad_norm": 0.7316666841506958,
+      "learning_rate": 4.6977417057758297e-07,
+      "loss": 0.0513,
+      "step": 14954
+    },
+    {
+      "epoch": 2.423039533376539,
+      "grad_norm": 0.8627529740333557,
+      "learning_rate": 4.695190208160197e-07,
+      "loss": 0.0617,
+      "step": 14955
+    },
+    {
+      "epoch": 2.423201555411536,
+      "grad_norm": 0.8494905829429626,
+      "learning_rate": 4.6926393318226045e-07,
+      "loss": 0.057,
+      "step": 14956
+    },
+    {
+      "epoch": 2.4233635774465325,
+      "grad_norm": 1.1695611476898193,
+      "learning_rate": 4.6900890768411145e-07,
+      "loss": 0.0733,
+      "step": 14957
+    },
+    {
+      "epoch": 2.4235255994815295,
+      "grad_norm": 0.8258056640625,
+      "learning_rate": 4.6875394432937345e-07,
+      "loss": 0.0581,
+      "step": 14958
+    },
+    {
+      "epoch": 2.4236876215165264,
+      "grad_norm": 0.8477557897567749,
+      "learning_rate": 4.684990431258499e-07,
+      "loss": 0.0589,
+      "step": 14959
+    },
+    {
+      "epoch": 2.423849643551523,
+      "grad_norm": 0.9261993765830994,
+      "learning_rate": 4.6824420408133953e-07,
+      "loss": 0.0632,
+      "step": 14960
+    },
+    {
+      "epoch": 2.42401166558652,
+      "grad_norm": 1.014751672744751,
+      "learning_rate": 4.6798942720364063e-07,
+      "loss": 0.0599,
+      "step": 14961
+    },
+    {
+      "epoch": 2.4241736876215167,
+      "grad_norm": 1.0292463302612305,
+      "learning_rate": 4.677347125005463e-07,
+      "loss": 0.0717,
+      "step": 14962
+    },
+    {
+      "epoch": 2.4243357096565132,
+      "grad_norm": 0.988088846206665,
+      "learning_rate": 4.6748005997985264e-07,
+      "loss": 0.0617,
+      "step": 14963
+    },
+    {
+      "epoch": 2.42449773169151,
+      "grad_norm": 1.0865875482559204,
+      "learning_rate": 4.6722546964935114e-07,
+      "loss": 0.0694,
+      "step": 14964
+    },
+    {
+      "epoch": 2.4246597537265067,
+      "grad_norm": 0.8739269375801086,
+      "learning_rate": 4.6697094151683026e-07,
+      "loss": 0.0573,
+      "step": 14965
+    },
+    {
+      "epoch": 2.4248217757615036,
+      "grad_norm": 0.983854353427887,
+      "learning_rate": 4.6671647559007884e-07,
+      "loss": 0.0637,
+      "step": 14966
+    },
+    {
+      "epoch": 2.4249837977965,
+      "grad_norm": 0.8353926539421082,
+      "learning_rate": 4.6646207187688197e-07,
+      "loss": 0.0606,
+      "step": 14967
+    },
+    {
+      "epoch": 2.425145819831497,
+      "grad_norm": 0.9040305018424988,
+      "learning_rate": 4.6620773038502625e-07,
+      "loss": 0.0676,
+      "step": 14968
+    },
+    {
+      "epoch": 2.425307841866494,
+      "grad_norm": 0.89899080991745,
+      "learning_rate": 4.659534511222916e-07,
+      "loss": 0.0592,
+      "step": 14969
+    },
+    {
+      "epoch": 2.4254698639014904,
+      "grad_norm": 1.2018378973007202,
+      "learning_rate": 4.656992340964589e-07,
+      "loss": 0.0634,
+      "step": 14970
+    },
+    {
+      "epoch": 2.4256318859364874,
+      "grad_norm": 0.8837653398513794,
+      "learning_rate": 4.6544507931530676e-07,
+      "loss": 0.0666,
+      "step": 14971
+    },
+    {
+      "epoch": 2.4257939079714843,
+      "grad_norm": 0.8115970492362976,
+      "learning_rate": 4.651909867866117e-07,
+      "loss": 0.0603,
+      "step": 14972
+    },
+    {
+      "epoch": 2.425955930006481,
+      "grad_norm": 0.8570439219474792,
+      "learning_rate": 4.649369565181483e-07,
+      "loss": 0.0545,
+      "step": 14973
+    },
+    {
+      "epoch": 2.4261179520414777,
+      "grad_norm": 0.8621937036514282,
+      "learning_rate": 4.64682988517689e-07,
+      "loss": 0.0586,
+      "step": 14974
+    },
+    {
+      "epoch": 2.426279974076474,
+      "grad_norm": 0.8469997048377991,
+      "learning_rate": 4.6442908279300536e-07,
+      "loss": 0.0578,
+      "step": 14975
+    },
+    {
+      "epoch": 2.426441996111471,
+      "grad_norm": 0.898225724697113,
+      "learning_rate": 4.641752393518661e-07,
+      "loss": 0.0597,
+      "step": 14976
+    },
+    {
+      "epoch": 2.426604018146468,
+      "grad_norm": 0.8789613842964172,
+      "learning_rate": 4.639214582020368e-07,
+      "loss": 0.0572,
+      "step": 14977
+    },
+    {
+      "epoch": 2.4267660401814646,
+      "grad_norm": 0.9776742458343506,
+      "learning_rate": 4.6366773935128423e-07,
+      "loss": 0.062,
+      "step": 14978
+    },
+    {
+      "epoch": 2.4269280622164615,
+      "grad_norm": 0.9235649704933167,
+      "learning_rate": 4.634140828073716e-07,
+      "loss": 0.0606,
+      "step": 14979
+    },
+    {
+      "epoch": 2.427090084251458,
+      "grad_norm": 0.954810619354248,
+      "learning_rate": 4.631604885780591e-07,
+      "loss": 0.064,
+      "step": 14980
+    },
+    {
+      "epoch": 2.427252106286455,
+      "grad_norm": 0.9652833938598633,
+      "learning_rate": 4.629069566711059e-07,
+      "loss": 0.061,
+      "step": 14981
+    },
+    {
+      "epoch": 2.427414128321452,
+      "grad_norm": 0.9159372448921204,
+      "learning_rate": 4.6265348709427146e-07,
+      "loss": 0.0648,
+      "step": 14982
+    },
+    {
+      "epoch": 2.4275761503564484,
+      "grad_norm": 0.8039860725402832,
+      "learning_rate": 4.6240007985530913e-07,
+      "loss": 0.0529,
+      "step": 14983
+    },
+    {
+      "epoch": 2.4277381723914453,
+      "grad_norm": 0.8820045590400696,
+      "learning_rate": 4.621467349619738e-07,
+      "loss": 0.0562,
+      "step": 14984
+    },
+    {
+      "epoch": 2.4279001944264422,
+      "grad_norm": 0.8682768940925598,
+      "learning_rate": 4.6189345242201674e-07,
+      "loss": 0.0572,
+      "step": 14985
+    },
+    {
+      "epoch": 2.4280622164614387,
+      "grad_norm": 0.830511748790741,
+      "learning_rate": 4.6164023224318786e-07,
+      "loss": 0.0606,
+      "step": 14986
+    },
+    {
+      "epoch": 2.4282242384964356,
+      "grad_norm": 0.9582686424255371,
+      "learning_rate": 4.6138707443323523e-07,
+      "loss": 0.064,
+      "step": 14987
+    },
+    {
+      "epoch": 2.428386260531432,
+      "grad_norm": 0.7856261134147644,
+      "learning_rate": 4.6113397899990474e-07,
+      "loss": 0.0495,
+      "step": 14988
+    },
+    {
+      "epoch": 2.428548282566429,
+      "grad_norm": 1.0667169094085693,
+      "learning_rate": 4.6088094595094057e-07,
+      "loss": 0.0712,
+      "step": 14989
+    },
+    {
+      "epoch": 2.4287103046014256,
+      "grad_norm": 0.9162574410438538,
+      "learning_rate": 4.6062797529408537e-07,
+      "loss": 0.0666,
+      "step": 14990
+    },
+    {
+      "epoch": 2.4288723266364225,
+      "grad_norm": 0.8367456197738647,
+      "learning_rate": 4.603750670370777e-07,
+      "loss": 0.0615,
+      "step": 14991
+    },
+    {
+      "epoch": 2.4290343486714194,
+      "grad_norm": 1.0230121612548828,
+      "learning_rate": 4.6012222118765796e-07,
+      "loss": 0.063,
+      "step": 14992
+    },
+    {
+      "epoch": 2.429196370706416,
+      "grad_norm": 0.9114834666252136,
+      "learning_rate": 4.598694377535617e-07,
+      "loss": 0.0676,
+      "step": 14993
+    },
+    {
+      "epoch": 2.429358392741413,
+      "grad_norm": 1.0044584274291992,
+      "learning_rate": 4.5961671674252447e-07,
+      "loss": 0.0695,
+      "step": 14994
+    },
+    {
+      "epoch": 2.42952041477641,
+      "grad_norm": 1.0129789113998413,
+      "learning_rate": 4.593640581622771e-07,
+      "loss": 0.0668,
+      "step": 14995
+    },
+    {
+      "epoch": 2.4296824368114063,
+      "grad_norm": 1.0267726182937622,
+      "learning_rate": 4.5911146202055113e-07,
+      "loss": 0.0645,
+      "step": 14996
+    },
+    {
+      "epoch": 2.429844458846403,
+      "grad_norm": 0.8808178901672363,
+      "learning_rate": 4.588589283250763e-07,
+      "loss": 0.063,
+      "step": 14997
+    },
+    {
+      "epoch": 2.4300064808813997,
+      "grad_norm": 1.2315572500228882,
+      "learning_rate": 4.5860645708357855e-07,
+      "loss": 0.0587,
+      "step": 14998
+    },
+    {
+      "epoch": 2.4301685029163966,
+      "grad_norm": 0.8882023096084595,
+      "learning_rate": 4.5835404830378296e-07,
+      "loss": 0.0604,
+      "step": 14999
+    },
+    {
+      "epoch": 2.4303305249513936,
+      "grad_norm": 0.8703988790512085,
+      "learning_rate": 4.581017019934131e-07,
+      "loss": 0.0572,
+      "step": 15000
+    },
+    {
+      "epoch": 2.43049254698639,
+      "grad_norm": 0.9442451596260071,
+      "learning_rate": 4.578494181601895e-07,
+      "loss": 0.0641,
+      "step": 15001
+    },
+    {
+      "epoch": 2.430654569021387,
+      "grad_norm": 0.8746116757392883,
+      "learning_rate": 4.57597196811832e-07,
+      "loss": 0.0652,
+      "step": 15002
+    },
+    {
+      "epoch": 2.4308165910563835,
+      "grad_norm": 0.8572196364402771,
+      "learning_rate": 4.5734503795605763e-07,
+      "loss": 0.058,
+      "step": 15003
+    },
+    {
+      "epoch": 2.4309786130913804,
+      "grad_norm": 0.8255354166030884,
+      "learning_rate": 4.5709294160058204e-07,
+      "loss": 0.0576,
+      "step": 15004
+    },
+    {
+      "epoch": 2.4311406351263773,
+      "grad_norm": 0.918545126914978,
+      "learning_rate": 4.5684090775311855e-07,
+      "loss": 0.0653,
+      "step": 15005
+    },
+    {
+      "epoch": 2.431302657161374,
+      "grad_norm": 0.9600496292114258,
+      "learning_rate": 4.565889364213791e-07,
+      "loss": 0.0665,
+      "step": 15006
+    },
+    {
+      "epoch": 2.4314646791963708,
+      "grad_norm": 1.181902527809143,
+      "learning_rate": 4.5633702761307327e-07,
+      "loss": 0.0618,
+      "step": 15007
+    },
+    {
+      "epoch": 2.4316267012313677,
+      "grad_norm": 1.0709116458892822,
+      "learning_rate": 4.5608518133590933e-07,
+      "loss": 0.0652,
+      "step": 15008
+    },
+    {
+      "epoch": 2.431788723266364,
+      "grad_norm": 0.9790709018707275,
+      "learning_rate": 4.5583339759759203e-07,
+      "loss": 0.0615,
+      "step": 15009
+    },
+    {
+      "epoch": 2.431950745301361,
+      "grad_norm": 0.8114708662033081,
+      "learning_rate": 4.5558167640582545e-07,
+      "loss": 0.0577,
+      "step": 15010
+    },
+    {
+      "epoch": 2.4321127673363576,
+      "grad_norm": 0.8501421213150024,
+      "learning_rate": 4.553300177683129e-07,
+      "loss": 0.0592,
+      "step": 15011
+    },
+    {
+      "epoch": 2.4322747893713546,
+      "grad_norm": 0.9928768277168274,
+      "learning_rate": 4.550784216927542e-07,
+      "loss": 0.0652,
+      "step": 15012
+    },
+    {
+      "epoch": 2.432436811406351,
+      "grad_norm": 1.0321820974349976,
+      "learning_rate": 4.5482688818684646e-07,
+      "loss": 0.0617,
+      "step": 15013
+    },
+    {
+      "epoch": 2.432598833441348,
+      "grad_norm": 0.8856379985809326,
+      "learning_rate": 4.5457541725828696e-07,
+      "loss": 0.0677,
+      "step": 15014
+    },
+    {
+      "epoch": 2.432760855476345,
+      "grad_norm": 0.8795359134674072,
+      "learning_rate": 4.543240089147699e-07,
+      "loss": 0.0588,
+      "step": 15015
+    },
+    {
+      "epoch": 2.4329228775113414,
+      "grad_norm": 0.8717951774597168,
+      "learning_rate": 4.5407266316398745e-07,
+      "loss": 0.0664,
+      "step": 15016
+    },
+    {
+      "epoch": 2.4330848995463383,
+      "grad_norm": 0.9385889768600464,
+      "learning_rate": 4.5382138001363067e-07,
+      "loss": 0.0656,
+      "step": 15017
+    },
+    {
+      "epoch": 2.4332469215813353,
+      "grad_norm": 0.8398104906082153,
+      "learning_rate": 4.5357015947138786e-07,
+      "loss": 0.0579,
+      "step": 15018
+    },
+    {
+      "epoch": 2.4334089436163318,
+      "grad_norm": 0.9150863289833069,
+      "learning_rate": 4.5331900154494623e-07,
+      "loss": 0.0595,
+      "step": 15019
+    },
+    {
+      "epoch": 2.4335709656513287,
+      "grad_norm": 0.9102035164833069,
+      "learning_rate": 4.530679062419899e-07,
+      "loss": 0.059,
+      "step": 15020
+    },
+    {
+      "epoch": 2.433732987686325,
+      "grad_norm": 0.9387508034706116,
+      "learning_rate": 4.528168735702024e-07,
+      "loss": 0.0654,
+      "step": 15021
+    },
+    {
+      "epoch": 2.433895009721322,
+      "grad_norm": 0.9315344095230103,
+      "learning_rate": 4.5256590353726426e-07,
+      "loss": 0.0664,
+      "step": 15022
+    },
+    {
+      "epoch": 2.434057031756319,
+      "grad_norm": 1.0084915161132812,
+      "learning_rate": 4.523149961508558e-07,
+      "loss": 0.0587,
+      "step": 15023
+    },
+    {
+      "epoch": 2.4342190537913155,
+      "grad_norm": 0.8739719390869141,
+      "learning_rate": 4.520641514186522e-07,
+      "loss": 0.0623,
+      "step": 15024
+    },
+    {
+      "epoch": 2.4343810758263125,
+      "grad_norm": 1.0345163345336914,
+      "learning_rate": 4.5181336934832897e-07,
+      "loss": 0.0693,
+      "step": 15025
+    },
+    {
+      "epoch": 2.434543097861309,
+      "grad_norm": 0.8258783221244812,
+      "learning_rate": 4.5156264994756144e-07,
+      "loss": 0.065,
+      "step": 15026
+    },
+    {
+      "epoch": 2.434705119896306,
+      "grad_norm": 0.8560571074485779,
+      "learning_rate": 4.5131199322401926e-07,
+      "loss": 0.0637,
+      "step": 15027
+    },
+    {
+      "epoch": 2.434867141931303,
+      "grad_norm": 0.8309101462364197,
+      "learning_rate": 4.510613991853721e-07,
+      "loss": 0.0608,
+      "step": 15028
+    },
+    {
+      "epoch": 2.4350291639662993,
+      "grad_norm": 0.91250079870224,
+      "learning_rate": 4.5081086783928754e-07,
+      "loss": 0.0579,
+      "step": 15029
+    },
+    {
+      "epoch": 2.4351911860012962,
+      "grad_norm": 0.9563912749290466,
+      "learning_rate": 4.5056039919343236e-07,
+      "loss": 0.063,
+      "step": 15030
+    },
+    {
+      "epoch": 2.435353208036293,
+      "grad_norm": 1.026058554649353,
+      "learning_rate": 4.503099932554689e-07,
+      "loss": 0.0614,
+      "step": 15031
+    },
+    {
+      "epoch": 2.4355152300712897,
+      "grad_norm": 0.9105499386787415,
+      "learning_rate": 4.5005965003305953e-07,
+      "loss": 0.064,
+      "step": 15032
+    },
+    {
+      "epoch": 2.4356772521062866,
+      "grad_norm": 0.8013800978660583,
+      "learning_rate": 4.49809369533864e-07,
+      "loss": 0.0565,
+      "step": 15033
+    },
+    {
+      "epoch": 2.435839274141283,
+      "grad_norm": 1.1963187456130981,
+      "learning_rate": 4.4955915176554065e-07,
+      "loss": 0.0628,
+      "step": 15034
+    },
+    {
+      "epoch": 2.43600129617628,
+      "grad_norm": 0.8717741966247559,
+      "learning_rate": 4.49308996735745e-07,
+      "loss": 0.0616,
+      "step": 15035
+    },
+    {
+      "epoch": 2.4361633182112765,
+      "grad_norm": 0.8936408758163452,
+      "learning_rate": 4.490589044521315e-07,
+      "loss": 0.0577,
+      "step": 15036
+    },
+    {
+      "epoch": 2.4363253402462735,
+      "grad_norm": 0.8526644706726074,
+      "learning_rate": 4.4880887492235265e-07,
+      "loss": 0.0613,
+      "step": 15037
+    },
+    {
+      "epoch": 2.4364873622812704,
+      "grad_norm": 0.8698762059211731,
+      "learning_rate": 4.4855890815405867e-07,
+      "loss": 0.0637,
+      "step": 15038
+    },
+    {
+      "epoch": 2.436649384316267,
+      "grad_norm": 0.8664577007293701,
+      "learning_rate": 4.483090041548968e-07,
+      "loss": 0.0591,
+      "step": 15039
+    },
+    {
+      "epoch": 2.436811406351264,
+      "grad_norm": 0.8795555830001831,
+      "learning_rate": 4.4805916293251486e-07,
+      "loss": 0.0665,
+      "step": 15040
+    },
+    {
+      "epoch": 2.4369734283862607,
+      "grad_norm": 0.8652831315994263,
+      "learning_rate": 4.4780938449455747e-07,
+      "loss": 0.0586,
+      "step": 15041
+    },
+    {
+      "epoch": 2.4371354504212572,
+      "grad_norm": 0.8803925514221191,
+      "learning_rate": 4.4755966884866606e-07,
+      "loss": 0.0581,
+      "step": 15042
+    },
+    {
+      "epoch": 2.437297472456254,
+      "grad_norm": 0.9763633608818054,
+      "learning_rate": 4.4731001600248234e-07,
+      "loss": 0.0573,
+      "step": 15043
+    },
+    {
+      "epoch": 2.4374594944912507,
+      "grad_norm": 0.9532651305198669,
+      "learning_rate": 4.470604259636438e-07,
+      "loss": 0.0591,
+      "step": 15044
+    },
+    {
+      "epoch": 2.4376215165262476,
+      "grad_norm": 0.8046735525131226,
+      "learning_rate": 4.4681089873978957e-07,
+      "loss": 0.0601,
+      "step": 15045
+    },
+    {
+      "epoch": 2.4377835385612445,
+      "grad_norm": 0.9399983882904053,
+      "learning_rate": 4.465614343385524e-07,
+      "loss": 0.0584,
+      "step": 15046
+    },
+    {
+      "epoch": 2.437945560596241,
+      "grad_norm": 0.8739177584648132,
+      "learning_rate": 4.463120327675663e-07,
+      "loss": 0.0596,
+      "step": 15047
+    },
+    {
+      "epoch": 2.438107582631238,
+      "grad_norm": 0.8265982270240784,
+      "learning_rate": 4.46062694034462e-07,
+      "loss": 0.0565,
+      "step": 15048
+    },
+    {
+      "epoch": 2.4382696046662344,
+      "grad_norm": 0.8940845727920532,
+      "learning_rate": 4.45813418146869e-07,
+      "loss": 0.0627,
+      "step": 15049
+    },
+    {
+      "epoch": 2.4384316267012314,
+      "grad_norm": 0.8442987203598022,
+      "learning_rate": 4.455642051124143e-07,
+      "loss": 0.0598,
+      "step": 15050
+    },
+    {
+      "epoch": 2.4385936487362283,
+      "grad_norm": 1.1562384366989136,
+      "learning_rate": 4.4531505493872334e-07,
+      "loss": 0.0673,
+      "step": 15051
+    },
+    {
+      "epoch": 2.438755670771225,
+      "grad_norm": 0.9282644391059875,
+      "learning_rate": 4.4506596763341985e-07,
+      "loss": 0.059,
+      "step": 15052
+    },
+    {
+      "epoch": 2.4389176928062217,
+      "grad_norm": 0.8264767527580261,
+      "learning_rate": 4.4481694320412383e-07,
+      "loss": 0.0552,
+      "step": 15053
+    },
+    {
+      "epoch": 2.439079714841218,
+      "grad_norm": 0.8514079451560974,
+      "learning_rate": 4.445679816584567e-07,
+      "loss": 0.0586,
+      "step": 15054
+    },
+    {
+      "epoch": 2.439241736876215,
+      "grad_norm": 0.8398727178573608,
+      "learning_rate": 4.4431908300403506e-07,
+      "loss": 0.061,
+      "step": 15055
+    },
+    {
+      "epoch": 2.439403758911212,
+      "grad_norm": 0.8882549405097961,
+      "learning_rate": 4.4407024724847534e-07,
+      "loss": 0.0603,
+      "step": 15056
+    },
+    {
+      "epoch": 2.4395657809462086,
+      "grad_norm": 1.0399445295333862,
+      "learning_rate": 4.4382147439939045e-07,
+      "loss": 0.0656,
+      "step": 15057
+    },
+    {
+      "epoch": 2.4397278029812055,
+      "grad_norm": 0.9227023124694824,
+      "learning_rate": 4.4357276446439197e-07,
+      "loss": 0.064,
+      "step": 15058
+    },
+    {
+      "epoch": 2.439889825016202,
+      "grad_norm": 0.911526620388031,
+      "learning_rate": 4.4332411745109135e-07,
+      "loss": 0.0632,
+      "step": 15059
+    },
+    {
+      "epoch": 2.440051847051199,
+      "grad_norm": 0.8623772263526917,
+      "learning_rate": 4.4307553336709525e-07,
+      "loss": 0.0604,
+      "step": 15060
+    },
+    {
+      "epoch": 2.440213869086196,
+      "grad_norm": 0.9016939401626587,
+      "learning_rate": 4.428270122200104e-07,
+      "loss": 0.0576,
+      "step": 15061
+    },
+    {
+      "epoch": 2.4403758911211924,
+      "grad_norm": 0.9120055437088013,
+      "learning_rate": 4.4257855401744044e-07,
+      "loss": 0.0614,
+      "step": 15062
+    },
+    {
+      "epoch": 2.4405379131561893,
+      "grad_norm": 1.0015937089920044,
+      "learning_rate": 4.4233015876698787e-07,
+      "loss": 0.0607,
+      "step": 15063
+    },
+    {
+      "epoch": 2.440699935191186,
+      "grad_norm": 0.8989893794059753,
+      "learning_rate": 4.42081826476253e-07,
+      "loss": 0.0566,
+      "step": 15064
+    },
+    {
+      "epoch": 2.4408619572261827,
+      "grad_norm": 0.9549097418785095,
+      "learning_rate": 4.4183355715283425e-07,
+      "loss": 0.0642,
+      "step": 15065
+    },
+    {
+      "epoch": 2.4410239792611796,
+      "grad_norm": 0.9465216398239136,
+      "learning_rate": 4.4158535080432803e-07,
+      "loss": 0.0662,
+      "step": 15066
+    },
+    {
+      "epoch": 2.441186001296176,
+      "grad_norm": 0.9384377002716064,
+      "learning_rate": 4.413372074383293e-07,
+      "loss": 0.0565,
+      "step": 15067
+    },
+    {
+      "epoch": 2.441348023331173,
+      "grad_norm": 0.8981598019599915,
+      "learning_rate": 4.4108912706242876e-07,
+      "loss": 0.0571,
+      "step": 15068
+    },
+    {
+      "epoch": 2.4415100453661696,
+      "grad_norm": 0.9155302047729492,
+      "learning_rate": 4.408411096842194e-07,
+      "loss": 0.0575,
+      "step": 15069
+    },
+    {
+      "epoch": 2.4416720674011665,
+      "grad_norm": 0.8059252500534058,
+      "learning_rate": 4.405931553112894e-07,
+      "loss": 0.0568,
+      "step": 15070
+    },
+    {
+      "epoch": 2.4418340894361634,
+      "grad_norm": 0.778841495513916,
+      "learning_rate": 4.4034526395122485e-07,
+      "loss": 0.0521,
+      "step": 15071
+    },
+    {
+      "epoch": 2.44199611147116,
+      "grad_norm": 0.8791018724441528,
+      "learning_rate": 4.4009743561161e-07,
+      "loss": 0.0576,
+      "step": 15072
+    },
+    {
+      "epoch": 2.442158133506157,
+      "grad_norm": 1.0327953100204468,
+      "learning_rate": 4.3984967030002964e-07,
+      "loss": 0.0668,
+      "step": 15073
+    },
+    {
+      "epoch": 2.442320155541154,
+      "grad_norm": 0.899375319480896,
+      "learning_rate": 4.396019680240643e-07,
+      "loss": 0.0592,
+      "step": 15074
+    },
+    {
+      "epoch": 2.4424821775761503,
+      "grad_norm": 0.8751654028892517,
+      "learning_rate": 4.3935432879129215e-07,
+      "loss": 0.0572,
+      "step": 15075
+    },
+    {
+      "epoch": 2.442644199611147,
+      "grad_norm": 0.8992448449134827,
+      "learning_rate": 4.3910675260929096e-07,
+      "loss": 0.0603,
+      "step": 15076
+    },
+    {
+      "epoch": 2.4428062216461437,
+      "grad_norm": 0.8739854693412781,
+      "learning_rate": 4.3885923948563585e-07,
+      "loss": 0.0588,
+      "step": 15077
+    },
+    {
+      "epoch": 2.4429682436811406,
+      "grad_norm": 0.8920753598213196,
+      "learning_rate": 4.386117894278999e-07,
+      "loss": 0.0579,
+      "step": 15078
+    },
+    {
+      "epoch": 2.4431302657161376,
+      "grad_norm": 0.9651167988777161,
+      "learning_rate": 4.383644024436551e-07,
+      "loss": 0.0569,
+      "step": 15079
+    },
+    {
+      "epoch": 2.443292287751134,
+      "grad_norm": 0.9695246815681458,
+      "learning_rate": 4.381170785404704e-07,
+      "loss": 0.0606,
+      "step": 15080
+    },
+    {
+      "epoch": 2.443454309786131,
+      "grad_norm": 0.8437383770942688,
+      "learning_rate": 4.378698177259133e-07,
+      "loss": 0.0604,
+      "step": 15081
+    },
+    {
+      "epoch": 2.4436163318211275,
+      "grad_norm": 0.9582122564315796,
+      "learning_rate": 4.376226200075495e-07,
+      "loss": 0.0653,
+      "step": 15082
+    },
+    {
+      "epoch": 2.4437783538561244,
+      "grad_norm": 1.188623070716858,
+      "learning_rate": 4.3737548539294266e-07,
+      "loss": 0.0702,
+      "step": 15083
+    },
+    {
+      "epoch": 2.4439403758911213,
+      "grad_norm": 0.975605309009552,
+      "learning_rate": 4.3712841388965476e-07,
+      "loss": 0.0653,
+      "step": 15084
+    },
+    {
+      "epoch": 2.444102397926118,
+      "grad_norm": 0.9400991201400757,
+      "learning_rate": 4.368814055052459e-07,
+      "loss": 0.0586,
+      "step": 15085
+    },
+    {
+      "epoch": 2.4442644199611148,
+      "grad_norm": 0.957561194896698,
+      "learning_rate": 4.3663446024727247e-07,
+      "loss": 0.0627,
+      "step": 15086
+    },
+    {
+      "epoch": 2.4444264419961117,
+      "grad_norm": 0.8637754917144775,
+      "learning_rate": 4.3638757812329095e-07,
+      "loss": 0.0598,
+      "step": 15087
+    },
+    {
+      "epoch": 2.444588464031108,
+      "grad_norm": 0.7822928428649902,
+      "learning_rate": 4.3614075914085617e-07,
+      "loss": 0.0555,
+      "step": 15088
+    },
+    {
+      "epoch": 2.444750486066105,
+      "grad_norm": 0.9488714337348938,
+      "learning_rate": 4.358940033075207e-07,
+      "loss": 0.0554,
+      "step": 15089
+    },
+    {
+      "epoch": 2.4449125081011016,
+      "grad_norm": 1.0039563179016113,
+      "learning_rate": 4.356473106308326e-07,
+      "loss": 0.0636,
+      "step": 15090
+    },
+    {
+      "epoch": 2.4450745301360985,
+      "grad_norm": 1.1325535774230957,
+      "learning_rate": 4.3540068111834144e-07,
+      "loss": 0.0678,
+      "step": 15091
+    },
+    {
+      "epoch": 2.445236552171095,
+      "grad_norm": 0.962975263595581,
+      "learning_rate": 4.351541147775931e-07,
+      "loss": 0.0666,
+      "step": 15092
+    },
+    {
+      "epoch": 2.445398574206092,
+      "grad_norm": 1.0033221244812012,
+      "learning_rate": 4.3490761161613186e-07,
+      "loss": 0.0651,
+      "step": 15093
+    },
+    {
+      "epoch": 2.445560596241089,
+      "grad_norm": 1.0614312887191772,
+      "learning_rate": 4.346611716415006e-07,
+      "loss": 0.0623,
+      "step": 15094
+    },
+    {
+      "epoch": 2.4457226182760854,
+      "grad_norm": 0.8363829255104065,
+      "learning_rate": 4.344147948612393e-07,
+      "loss": 0.0593,
+      "step": 15095
+    },
+    {
+      "epoch": 2.4458846403110823,
+      "grad_norm": 0.7662982940673828,
+      "learning_rate": 4.341684812828867e-07,
+      "loss": 0.0506,
+      "step": 15096
+    },
+    {
+      "epoch": 2.4460466623460793,
+      "grad_norm": 0.8307511806488037,
+      "learning_rate": 4.3392223091397925e-07,
+      "loss": 0.059,
+      "step": 15097
+    },
+    {
+      "epoch": 2.4462086843810757,
+      "grad_norm": 0.9444789290428162,
+      "learning_rate": 4.336760437620519e-07,
+      "loss": 0.0681,
+      "step": 15098
+    },
+    {
+      "epoch": 2.4463707064160727,
+      "grad_norm": 0.9271620512008667,
+      "learning_rate": 4.33429919834637e-07,
+      "loss": 0.0625,
+      "step": 15099
+    },
+    {
+      "epoch": 2.446532728451069,
+      "grad_norm": 0.7332361936569214,
+      "learning_rate": 4.331838591392662e-07,
+      "loss": 0.0545,
+      "step": 15100
+    },
+    {
+      "epoch": 2.446694750486066,
+      "grad_norm": 0.8970035314559937,
+      "learning_rate": 4.3293786168346674e-07,
+      "loss": 0.0612,
+      "step": 15101
+    },
+    {
+      "epoch": 2.446856772521063,
+      "grad_norm": 0.9213421940803528,
+      "learning_rate": 4.326919274747668e-07,
+      "loss": 0.0575,
+      "step": 15102
+    },
+    {
+      "epoch": 2.4470187945560595,
+      "grad_norm": 0.8858630657196045,
+      "learning_rate": 4.32446056520692e-07,
+      "loss": 0.0578,
+      "step": 15103
+    },
+    {
+      "epoch": 2.4471808165910565,
+      "grad_norm": 1.067657709121704,
+      "learning_rate": 4.322002488287635e-07,
+      "loss": 0.0658,
+      "step": 15104
+    },
+    {
+      "epoch": 2.447342838626053,
+      "grad_norm": 0.9456945061683655,
+      "learning_rate": 4.319545044065038e-07,
+      "loss": 0.0584,
+      "step": 15105
+    },
+    {
+      "epoch": 2.44750486066105,
+      "grad_norm": 0.8441051244735718,
+      "learning_rate": 4.317088232614308e-07,
+      "loss": 0.0632,
+      "step": 15106
+    },
+    {
+      "epoch": 2.447666882696047,
+      "grad_norm": 0.9701780080795288,
+      "learning_rate": 4.3146320540106397e-07,
+      "loss": 0.0627,
+      "step": 15107
+    },
+    {
+      "epoch": 2.4478289047310433,
+      "grad_norm": 0.8207990527153015,
+      "learning_rate": 4.3121765083291663e-07,
+      "loss": 0.0608,
+      "step": 15108
+    },
+    {
+      "epoch": 2.4479909267660402,
+      "grad_norm": 0.8845500946044922,
+      "learning_rate": 4.3097215956450304e-07,
+      "loss": 0.0592,
+      "step": 15109
+    },
+    {
+      "epoch": 2.448152948801037,
+      "grad_norm": 0.8777700662612915,
+      "learning_rate": 4.307267316033342e-07,
+      "loss": 0.0663,
+      "step": 15110
+    },
+    {
+      "epoch": 2.4483149708360337,
+      "grad_norm": 0.9499847888946533,
+      "learning_rate": 4.3048136695691965e-07,
+      "loss": 0.0657,
+      "step": 15111
+    },
+    {
+      "epoch": 2.4484769928710306,
+      "grad_norm": 0.8840598464012146,
+      "learning_rate": 4.3023606563276753e-07,
+      "loss": 0.0586,
+      "step": 15112
+    },
+    {
+      "epoch": 2.448639014906027,
+      "grad_norm": 0.9703836441040039,
+      "learning_rate": 4.2999082763838293e-07,
+      "loss": 0.06,
+      "step": 15113
+    },
+    {
+      "epoch": 2.448801036941024,
+      "grad_norm": 1.0197018384933472,
+      "learning_rate": 4.297456529812702e-07,
+      "loss": 0.0734,
+      "step": 15114
+    },
+    {
+      "epoch": 2.4489630589760205,
+      "grad_norm": 0.9363842606544495,
+      "learning_rate": 4.2950054166892937e-07,
+      "loss": 0.0602,
+      "step": 15115
+    },
+    {
+      "epoch": 2.4491250810110174,
+      "grad_norm": 0.8683654069900513,
+      "learning_rate": 4.292554937088622e-07,
+      "loss": 0.0586,
+      "step": 15116
+    },
+    {
+      "epoch": 2.4492871030460144,
+      "grad_norm": 0.9922870397567749,
+      "learning_rate": 4.290105091085656e-07,
+      "loss": 0.0618,
+      "step": 15117
+    },
+    {
+      "epoch": 2.449449125081011,
+      "grad_norm": 0.9936143755912781,
+      "learning_rate": 4.287655878755365e-07,
+      "loss": 0.0596,
+      "step": 15118
+    },
+    {
+      "epoch": 2.449611147116008,
+      "grad_norm": 0.7621217370033264,
+      "learning_rate": 4.2852073001726754e-07,
+      "loss": 0.0545,
+      "step": 15119
+    },
+    {
+      "epoch": 2.4497731691510047,
+      "grad_norm": 1.023511290550232,
+      "learning_rate": 4.282759355412505e-07,
+      "loss": 0.0672,
+      "step": 15120
+    },
+    {
+      "epoch": 2.4499351911860012,
+      "grad_norm": 0.9315782189369202,
+      "learning_rate": 4.280312044549778e-07,
+      "loss": 0.0593,
+      "step": 15121
+    },
+    {
+      "epoch": 2.450097213220998,
+      "grad_norm": 0.8633824586868286,
+      "learning_rate": 4.2778653676593534e-07,
+      "loss": 0.0586,
+      "step": 15122
+    },
+    {
+      "epoch": 2.4502592352559946,
+      "grad_norm": 0.99221271276474,
+      "learning_rate": 4.275419324816105e-07,
+      "loss": 0.0601,
+      "step": 15123
+    },
+    {
+      "epoch": 2.4504212572909916,
+      "grad_norm": 0.7723276615142822,
+      "learning_rate": 4.272973916094872e-07,
+      "loss": 0.0519,
+      "step": 15124
+    },
+    {
+      "epoch": 2.4505832793259885,
+      "grad_norm": 0.8727956414222717,
+      "learning_rate": 4.2705291415704757e-07,
+      "loss": 0.0611,
+      "step": 15125
+    },
+    {
+      "epoch": 2.450745301360985,
+      "grad_norm": 1.0098901987075806,
+      "learning_rate": 4.268085001317726e-07,
+      "loss": 0.0693,
+      "step": 15126
+    },
+    {
+      "epoch": 2.450907323395982,
+      "grad_norm": 0.9148622751235962,
+      "learning_rate": 4.2656414954114044e-07,
+      "loss": 0.0559,
+      "step": 15127
+    },
+    {
+      "epoch": 2.4510693454309784,
+      "grad_norm": 1.0653668642044067,
+      "learning_rate": 4.263198623926279e-07,
+      "loss": 0.0658,
+      "step": 15128
+    },
+    {
+      "epoch": 2.4512313674659754,
+      "grad_norm": 1.062276005744934,
+      "learning_rate": 4.260756386937095e-07,
+      "loss": 0.0647,
+      "step": 15129
+    },
+    {
+      "epoch": 2.4513933895009723,
+      "grad_norm": 0.9811892509460449,
+      "learning_rate": 4.258314784518569e-07,
+      "loss": 0.0649,
+      "step": 15130
+    },
+    {
+      "epoch": 2.451555411535969,
+      "grad_norm": 0.8770061135292053,
+      "learning_rate": 4.2558738167454233e-07,
+      "loss": 0.0617,
+      "step": 15131
+    },
+    {
+      "epoch": 2.4517174335709657,
+      "grad_norm": 0.860925018787384,
+      "learning_rate": 4.253433483692337e-07,
+      "loss": 0.0617,
+      "step": 15132
+    },
+    {
+      "epoch": 2.4518794556059627,
+      "grad_norm": 0.9953809976577759,
+      "learning_rate": 4.250993785433988e-07,
+      "loss": 0.0604,
+      "step": 15133
+    },
+    {
+      "epoch": 2.452041477640959,
+      "grad_norm": 0.9298536777496338,
+      "learning_rate": 4.248554722045009e-07,
+      "loss": 0.0598,
+      "step": 15134
+    },
+    {
+      "epoch": 2.452203499675956,
+      "grad_norm": 0.8232978582382202,
+      "learning_rate": 4.246116293600033e-07,
+      "loss": 0.0578,
+      "step": 15135
+    },
+    {
+      "epoch": 2.4523655217109526,
+      "grad_norm": 0.9326602220535278,
+      "learning_rate": 4.2436785001736896e-07,
+      "loss": 0.0566,
+      "step": 15136
+    },
+    {
+      "epoch": 2.4525275437459495,
+      "grad_norm": 1.011151671409607,
+      "learning_rate": 4.241241341840546e-07,
+      "loss": 0.0606,
+      "step": 15137
+    },
+    {
+      "epoch": 2.452689565780946,
+      "grad_norm": 0.8477997183799744,
+      "learning_rate": 4.2388048186751823e-07,
+      "loss": 0.0556,
+      "step": 15138
+    },
+    {
+      "epoch": 2.452851587815943,
+      "grad_norm": 0.9269173741340637,
+      "learning_rate": 4.2363689307521494e-07,
+      "loss": 0.0573,
+      "step": 15139
+    },
+    {
+      "epoch": 2.45301360985094,
+      "grad_norm": 0.9142125844955444,
+      "learning_rate": 4.233933678145982e-07,
+      "loss": 0.0639,
+      "step": 15140
+    },
+    {
+      "epoch": 2.4531756318859363,
+      "grad_norm": 1.0160726308822632,
+      "learning_rate": 4.2314990609311905e-07,
+      "loss": 0.0579,
+      "step": 15141
+    },
+    {
+      "epoch": 2.4533376539209333,
+      "grad_norm": 0.950394332408905,
+      "learning_rate": 4.229065079182268e-07,
+      "loss": 0.0654,
+      "step": 15142
+    },
+    {
+      "epoch": 2.45349967595593,
+      "grad_norm": 0.9993029236793518,
+      "learning_rate": 4.2266317329736904e-07,
+      "loss": 0.0583,
+      "step": 15143
+    },
+    {
+      "epoch": 2.4536616979909267,
+      "grad_norm": 0.9236908555030823,
+      "learning_rate": 4.224199022379913e-07,
+      "loss": 0.0595,
+      "step": 15144
+    },
+    {
+      "epoch": 2.4538237200259236,
+      "grad_norm": 0.7894659042358398,
+      "learning_rate": 4.2217669474753644e-07,
+      "loss": 0.0562,
+      "step": 15145
+    },
+    {
+      "epoch": 2.45398574206092,
+      "grad_norm": 0.8922528028488159,
+      "learning_rate": 4.2193355083344684e-07,
+      "loss": 0.0649,
+      "step": 15146
+    },
+    {
+      "epoch": 2.454147764095917,
+      "grad_norm": 0.8606491088867188,
+      "learning_rate": 4.216904705031624e-07,
+      "loss": 0.0582,
+      "step": 15147
+    },
+    {
+      "epoch": 2.454309786130914,
+      "grad_norm": 0.8970129489898682,
+      "learning_rate": 4.2144745376411946e-07,
+      "loss": 0.0609,
+      "step": 15148
+    },
+    {
+      "epoch": 2.4544718081659105,
+      "grad_norm": 0.8637439012527466,
+      "learning_rate": 4.2120450062375364e-07,
+      "loss": 0.0556,
+      "step": 15149
+    },
+    {
+      "epoch": 2.4546338302009074,
+      "grad_norm": 0.9069430828094482,
+      "learning_rate": 4.2096161108950015e-07,
+      "loss": 0.0568,
+      "step": 15150
+    },
+    {
+      "epoch": 2.454795852235904,
+      "grad_norm": 1.070995807647705,
+      "learning_rate": 4.2071878516879107e-07,
+      "loss": 0.0597,
+      "step": 15151
+    },
+    {
+      "epoch": 2.454957874270901,
+      "grad_norm": 0.8751211762428284,
+      "learning_rate": 4.204760228690546e-07,
+      "loss": 0.0612,
+      "step": 15152
+    },
+    {
+      "epoch": 2.4551198963058978,
+      "grad_norm": 0.9568556547164917,
+      "learning_rate": 4.202333241977194e-07,
+      "loss": 0.0578,
+      "step": 15153
+    },
+    {
+      "epoch": 2.4552819183408943,
+      "grad_norm": 0.8110483288764954,
+      "learning_rate": 4.1999068916221184e-07,
+      "loss": 0.0536,
+      "step": 15154
+    },
+    {
+      "epoch": 2.455443940375891,
+      "grad_norm": 1.0597234964370728,
+      "learning_rate": 4.1974811776995526e-07,
+      "loss": 0.0536,
+      "step": 15155
+    },
+    {
+      "epoch": 2.4556059624108877,
+      "grad_norm": 0.8247944712638855,
+      "learning_rate": 4.1950561002837257e-07,
+      "loss": 0.0552,
+      "step": 15156
+    },
+    {
+      "epoch": 2.4557679844458846,
+      "grad_norm": 0.8759192228317261,
+      "learning_rate": 4.1926316594488315e-07,
+      "loss": 0.0639,
+      "step": 15157
+    },
+    {
+      "epoch": 2.4559300064808816,
+      "grad_norm": 0.8859750032424927,
+      "learning_rate": 4.1902078552690573e-07,
+      "loss": 0.055,
+      "step": 15158
+    },
+    {
+      "epoch": 2.456092028515878,
+      "grad_norm": 0.9089791774749756,
+      "learning_rate": 4.1877846878185635e-07,
+      "loss": 0.0567,
+      "step": 15159
+    },
+    {
+      "epoch": 2.456254050550875,
+      "grad_norm": 0.9586995244026184,
+      "learning_rate": 4.185362157171496e-07,
+      "loss": 0.0611,
+      "step": 15160
+    },
+    {
+      "epoch": 2.4564160725858715,
+      "grad_norm": 0.9053599238395691,
+      "learning_rate": 4.1829402634019746e-07,
+      "loss": 0.0628,
+      "step": 15161
+    },
+    {
+      "epoch": 2.4565780946208684,
+      "grad_norm": 0.8209922313690186,
+      "learning_rate": 4.1805190065841107e-07,
+      "loss": 0.0568,
+      "step": 15162
+    },
+    {
+      "epoch": 2.4567401166558653,
+      "grad_norm": 1.2450376749038696,
+      "learning_rate": 4.178098386791971e-07,
+      "loss": 0.0639,
+      "step": 15163
+    },
+    {
+      "epoch": 2.456902138690862,
+      "grad_norm": 0.9918776750564575,
+      "learning_rate": 4.175678404099637e-07,
+      "loss": 0.0621,
+      "step": 15164
+    },
+    {
+      "epoch": 2.4570641607258588,
+      "grad_norm": 0.916986346244812,
+      "learning_rate": 4.1732590585811586e-07,
+      "loss": 0.0672,
+      "step": 15165
+    },
+    {
+      "epoch": 2.4572261827608557,
+      "grad_norm": 1.3180897235870361,
+      "learning_rate": 4.1708403503105456e-07,
+      "loss": 0.0618,
+      "step": 15166
+    },
+    {
+      "epoch": 2.457388204795852,
+      "grad_norm": 0.946431577205658,
+      "learning_rate": 4.168422279361811e-07,
+      "loss": 0.059,
+      "step": 15167
+    },
+    {
+      "epoch": 2.457550226830849,
+      "grad_norm": 0.9107024669647217,
+      "learning_rate": 4.166004845808941e-07,
+      "loss": 0.0579,
+      "step": 15168
+    },
+    {
+      "epoch": 2.4577122488658456,
+      "grad_norm": 0.9876469969749451,
+      "learning_rate": 4.163588049725914e-07,
+      "loss": 0.0622,
+      "step": 15169
+    },
+    {
+      "epoch": 2.4578742709008425,
+      "grad_norm": 0.9634394645690918,
+      "learning_rate": 4.1611718911866663e-07,
+      "loss": 0.0572,
+      "step": 15170
+    },
+    {
+      "epoch": 2.458036292935839,
+      "grad_norm": 0.9489405155181885,
+      "learning_rate": 4.158756370265127e-07,
+      "loss": 0.0577,
+      "step": 15171
+    },
+    {
+      "epoch": 2.458198314970836,
+      "grad_norm": 0.8651463389396667,
+      "learning_rate": 4.1563414870352093e-07,
+      "loss": 0.0559,
+      "step": 15172
+    },
+    {
+      "epoch": 2.458360337005833,
+      "grad_norm": 0.8804042935371399,
+      "learning_rate": 4.1539272415708014e-07,
+      "loss": 0.0612,
+      "step": 15173
+    },
+    {
+      "epoch": 2.4585223590408294,
+      "grad_norm": 0.8089525103569031,
+      "learning_rate": 4.1515136339457725e-07,
+      "loss": 0.0591,
+      "step": 15174
+    },
+    {
+      "epoch": 2.4586843810758263,
+      "grad_norm": 0.8935924768447876,
+      "learning_rate": 4.1491006642339765e-07,
+      "loss": 0.0638,
+      "step": 15175
+    },
+    {
+      "epoch": 2.4588464031108233,
+      "grad_norm": 0.8347843289375305,
+      "learning_rate": 4.146688332509241e-07,
+      "loss": 0.0616,
+      "step": 15176
+    },
+    {
+      "epoch": 2.4590084251458197,
+      "grad_norm": 0.9167835712432861,
+      "learning_rate": 4.144276638845382e-07,
+      "loss": 0.0613,
+      "step": 15177
+    },
+    {
+      "epoch": 2.4591704471808167,
+      "grad_norm": 1.0505998134613037,
+      "learning_rate": 4.1418655833161794e-07,
+      "loss": 0.0596,
+      "step": 15178
+    },
+    {
+      "epoch": 2.459332469215813,
+      "grad_norm": 0.8533574938774109,
+      "learning_rate": 4.139455165995418e-07,
+      "loss": 0.0528,
+      "step": 15179
+    },
+    {
+      "epoch": 2.45949449125081,
+      "grad_norm": 0.8060251474380493,
+      "learning_rate": 4.137045386956853e-07,
+      "loss": 0.0542,
+      "step": 15180
+    },
+    {
+      "epoch": 2.459656513285807,
+      "grad_norm": 0.7060052752494812,
+      "learning_rate": 4.1346362462742067e-07,
+      "loss": 0.0521,
+      "step": 15181
+    },
+    {
+      "epoch": 2.4598185353208035,
+      "grad_norm": 1.0977946519851685,
+      "learning_rate": 4.1322277440211973e-07,
+      "loss": 0.062,
+      "step": 15182
+    },
+    {
+      "epoch": 2.4599805573558005,
+      "grad_norm": 0.7807490229606628,
+      "learning_rate": 4.129819880271516e-07,
+      "loss": 0.0566,
+      "step": 15183
+    },
+    {
+      "epoch": 2.460142579390797,
+      "grad_norm": 1.1565881967544556,
+      "learning_rate": 4.1274126550988505e-07,
+      "loss": 0.0633,
+      "step": 15184
+    },
+    {
+      "epoch": 2.460304601425794,
+      "grad_norm": 0.9105919003486633,
+      "learning_rate": 4.125006068576842e-07,
+      "loss": 0.0642,
+      "step": 15185
+    },
+    {
+      "epoch": 2.460466623460791,
+      "grad_norm": 0.8822482824325562,
+      "learning_rate": 4.1226001207791327e-07,
+      "loss": 0.0622,
+      "step": 15186
+    },
+    {
+      "epoch": 2.4606286454957873,
+      "grad_norm": 0.873638927936554,
+      "learning_rate": 4.120194811779335e-07,
+      "loss": 0.0659,
+      "step": 15187
+    },
+    {
+      "epoch": 2.4607906675307842,
+      "grad_norm": 0.9651090502738953,
+      "learning_rate": 4.1177901416510485e-07,
+      "loss": 0.0645,
+      "step": 15188
+    },
+    {
+      "epoch": 2.460952689565781,
+      "grad_norm": 0.8976024389266968,
+      "learning_rate": 4.1153861104678505e-07,
+      "loss": 0.0582,
+      "step": 15189
+    },
+    {
+      "epoch": 2.4611147116007777,
+      "grad_norm": 1.0942816734313965,
+      "learning_rate": 4.112982718303299e-07,
+      "loss": 0.0669,
+      "step": 15190
+    },
+    {
+      "epoch": 2.4612767336357746,
+      "grad_norm": 1.016480803489685,
+      "learning_rate": 4.1105799652309347e-07,
+      "loss": 0.0662,
+      "step": 15191
+    },
+    {
+      "epoch": 2.461438755670771,
+      "grad_norm": 0.9741160869598389,
+      "learning_rate": 4.1081778513242606e-07,
+      "loss": 0.0628,
+      "step": 15192
+    },
+    {
+      "epoch": 2.461600777705768,
+      "grad_norm": 0.8383246064186096,
+      "learning_rate": 4.105776376656795e-07,
+      "loss": 0.054,
+      "step": 15193
+    },
+    {
+      "epoch": 2.4617627997407645,
+      "grad_norm": 0.8395657539367676,
+      "learning_rate": 4.103375541302007e-07,
+      "loss": 0.0606,
+      "step": 15194
+    },
+    {
+      "epoch": 2.4619248217757614,
+      "grad_norm": 0.8531553745269775,
+      "learning_rate": 4.1009753453333636e-07,
+      "loss": 0.0621,
+      "step": 15195
+    },
+    {
+      "epoch": 2.4620868438107584,
+      "grad_norm": 0.9702677130699158,
+      "learning_rate": 4.0985757888242965e-07,
+      "loss": 0.0658,
+      "step": 15196
+    },
+    {
+      "epoch": 2.462248865845755,
+      "grad_norm": 0.9990732073783875,
+      "learning_rate": 4.09617687184822e-07,
+      "loss": 0.0656,
+      "step": 15197
+    },
+    {
+      "epoch": 2.462410887880752,
+      "grad_norm": 0.888516902923584,
+      "learning_rate": 4.0937785944785617e-07,
+      "loss": 0.0645,
+      "step": 15198
+    },
+    {
+      "epoch": 2.4625729099157487,
+      "grad_norm": 1.0981239080429077,
+      "learning_rate": 4.091380956788676e-07,
+      "loss": 0.0567,
+      "step": 15199
+    },
+    {
+      "epoch": 2.462734931950745,
+      "grad_norm": 0.9011585712432861,
+      "learning_rate": 4.0889839588519386e-07,
+      "loss": 0.0584,
+      "step": 15200
+    },
+    {
+      "epoch": 2.462896953985742,
+      "grad_norm": 0.8724988698959351,
+      "learning_rate": 4.086587600741687e-07,
+      "loss": 0.0592,
+      "step": 15201
+    },
+    {
+      "epoch": 2.4630589760207386,
+      "grad_norm": 1.0473746061325073,
+      "learning_rate": 4.0841918825312465e-07,
+      "loss": 0.0654,
+      "step": 15202
+    },
+    {
+      "epoch": 2.4632209980557356,
+      "grad_norm": 1.0286039113998413,
+      "learning_rate": 4.0817968042939165e-07,
+      "loss": 0.066,
+      "step": 15203
+    },
+    {
+      "epoch": 2.4633830200907325,
+      "grad_norm": 0.9677504301071167,
+      "learning_rate": 4.0794023661029856e-07,
+      "loss": 0.0646,
+      "step": 15204
+    },
+    {
+      "epoch": 2.463545042125729,
+      "grad_norm": 0.9991774559020996,
+      "learning_rate": 4.0770085680317153e-07,
+      "loss": 0.0683,
+      "step": 15205
+    },
+    {
+      "epoch": 2.463707064160726,
+      "grad_norm": 0.8651935458183289,
+      "learning_rate": 4.0746154101533485e-07,
+      "loss": 0.064,
+      "step": 15206
+    },
+    {
+      "epoch": 2.4638690861957224,
+      "grad_norm": 0.9634072184562683,
+      "learning_rate": 4.072222892541111e-07,
+      "loss": 0.068,
+      "step": 15207
+    },
+    {
+      "epoch": 2.4640311082307194,
+      "grad_norm": 0.7955970764160156,
+      "learning_rate": 4.0698310152682107e-07,
+      "loss": 0.0583,
+      "step": 15208
+    },
+    {
+      "epoch": 2.4641931302657163,
+      "grad_norm": 0.8827134966850281,
+      "learning_rate": 4.067439778407839e-07,
+      "loss": 0.058,
+      "step": 15209
+    },
+    {
+      "epoch": 2.464355152300713,
+      "grad_norm": 0.932346522808075,
+      "learning_rate": 4.065049182033146e-07,
+      "loss": 0.0603,
+      "step": 15210
+    },
+    {
+      "epoch": 2.4645171743357097,
+      "grad_norm": 0.8273137807846069,
+      "learning_rate": 4.0626592262172803e-07,
+      "loss": 0.0568,
+      "step": 15211
+    },
+    {
+      "epoch": 2.4646791963707066,
+      "grad_norm": 0.8896094560623169,
+      "learning_rate": 4.0602699110333795e-07,
+      "loss": 0.0552,
+      "step": 15212
+    },
+    {
+      "epoch": 2.464841218405703,
+      "grad_norm": 0.8498284220695496,
+      "learning_rate": 4.0578812365545533e-07,
+      "loss": 0.0577,
+      "step": 15213
+    },
+    {
+      "epoch": 2.4650032404407,
+      "grad_norm": 0.8612973690032959,
+      "learning_rate": 4.0554932028538774e-07,
+      "loss": 0.0596,
+      "step": 15214
+    },
+    {
+      "epoch": 2.4651652624756966,
+      "grad_norm": 0.9590808153152466,
+      "learning_rate": 4.0531058100044264e-07,
+      "loss": 0.0692,
+      "step": 15215
+    },
+    {
+      "epoch": 2.4653272845106935,
+      "grad_norm": 0.9362857341766357,
+      "learning_rate": 4.050719058079244e-07,
+      "loss": 0.0629,
+      "step": 15216
+    },
+    {
+      "epoch": 2.46548930654569,
+      "grad_norm": 1.062246322631836,
+      "learning_rate": 4.048332947151362e-07,
+      "loss": 0.0645,
+      "step": 15217
+    },
+    {
+      "epoch": 2.465651328580687,
+      "grad_norm": 0.8582957983016968,
+      "learning_rate": 4.045947477293791e-07,
+      "loss": 0.0562,
+      "step": 15218
+    },
+    {
+      "epoch": 2.465813350615684,
+      "grad_norm": 0.8503392934799194,
+      "learning_rate": 4.043562648579519e-07,
+      "loss": 0.0538,
+      "step": 15219
+    },
+    {
+      "epoch": 2.4659753726506803,
+      "grad_norm": 0.7822324633598328,
+      "learning_rate": 4.041178461081519e-07,
+      "loss": 0.0569,
+      "step": 15220
+    },
+    {
+      "epoch": 2.4661373946856773,
+      "grad_norm": 0.8472219705581665,
+      "learning_rate": 4.0387949148727343e-07,
+      "loss": 0.0584,
+      "step": 15221
+    },
+    {
+      "epoch": 2.466299416720674,
+      "grad_norm": 0.8778401613235474,
+      "learning_rate": 4.036412010026103e-07,
+      "loss": 0.0649,
+      "step": 15222
+    },
+    {
+      "epoch": 2.4664614387556707,
+      "grad_norm": 0.8573176860809326,
+      "learning_rate": 4.034029746614532e-07,
+      "loss": 0.0615,
+      "step": 15223
+    },
+    {
+      "epoch": 2.4666234607906676,
+      "grad_norm": 0.8038144707679749,
+      "learning_rate": 4.0316481247109215e-07,
+      "loss": 0.0562,
+      "step": 15224
+    },
+    {
+      "epoch": 2.466785482825664,
+      "grad_norm": 0.9153627753257751,
+      "learning_rate": 4.029267144388127e-07,
+      "loss": 0.0603,
+      "step": 15225
+    },
+    {
+      "epoch": 2.466947504860661,
+      "grad_norm": 0.8227776885032654,
+      "learning_rate": 4.0268868057190075e-07,
+      "loss": 0.0585,
+      "step": 15226
+    },
+    {
+      "epoch": 2.467109526895658,
+      "grad_norm": 0.880547285079956,
+      "learning_rate": 4.0245071087764015e-07,
+      "loss": 0.0633,
+      "step": 15227
+    },
+    {
+      "epoch": 2.4672715489306545,
+      "grad_norm": 0.9440403580665588,
+      "learning_rate": 4.022128053633123e-07,
+      "loss": 0.062,
+      "step": 15228
+    },
+    {
+      "epoch": 2.4674335709656514,
+      "grad_norm": 0.8492519855499268,
+      "learning_rate": 4.0197496403619557e-07,
+      "loss": 0.0534,
+      "step": 15229
+    },
+    {
+      "epoch": 2.467595593000648,
+      "grad_norm": 0.7454891800880432,
+      "learning_rate": 4.017371869035674e-07,
+      "loss": 0.051,
+      "step": 15230
+    },
+    {
+      "epoch": 2.467757615035645,
+      "grad_norm": 0.8574808239936829,
+      "learning_rate": 4.014994739727046e-07,
+      "loss": 0.0585,
+      "step": 15231
+    },
+    {
+      "epoch": 2.4679196370706418,
+      "grad_norm": 0.9869388937950134,
+      "learning_rate": 4.01261825250879e-07,
+      "loss": 0.063,
+      "step": 15232
+    },
+    {
+      "epoch": 2.4680816591056383,
+      "grad_norm": 0.9070796966552734,
+      "learning_rate": 4.0102424074536295e-07,
+      "loss": 0.0593,
+      "step": 15233
+    },
+    {
+      "epoch": 2.468243681140635,
+      "grad_norm": 1.0267338752746582,
+      "learning_rate": 4.0078672046342553e-07,
+      "loss": 0.0566,
+      "step": 15234
+    },
+    {
+      "epoch": 2.468405703175632,
+      "grad_norm": 1.009340763092041,
+      "learning_rate": 4.005492644123346e-07,
+      "loss": 0.0684,
+      "step": 15235
+    },
+    {
+      "epoch": 2.4685677252106286,
+      "grad_norm": 0.7549983263015747,
+      "learning_rate": 4.0031187259935546e-07,
+      "loss": 0.0585,
+      "step": 15236
+    },
+    {
+      "epoch": 2.4687297472456255,
+      "grad_norm": 1.0285526514053345,
+      "learning_rate": 4.0007454503175196e-07,
+      "loss": 0.0649,
+      "step": 15237
+    },
+    {
+      "epoch": 2.468891769280622,
+      "grad_norm": 0.9466545581817627,
+      "learning_rate": 3.998372817167856e-07,
+      "loss": 0.0668,
+      "step": 15238
+    },
+    {
+      "epoch": 2.469053791315619,
+      "grad_norm": 0.8367640376091003,
+      "learning_rate": 3.9960008266171663e-07,
+      "loss": 0.055,
+      "step": 15239
+    },
+    {
+      "epoch": 2.4692158133506155,
+      "grad_norm": 0.9086434841156006,
+      "learning_rate": 3.993629478738012e-07,
+      "loss": 0.0613,
+      "step": 15240
+    },
+    {
+      "epoch": 2.4693778353856124,
+      "grad_norm": 0.9851820468902588,
+      "learning_rate": 3.9912587736029656e-07,
+      "loss": 0.0644,
+      "step": 15241
+    },
+    {
+      "epoch": 2.4695398574206093,
+      "grad_norm": 0.8282864093780518,
+      "learning_rate": 3.988888711284569e-07,
+      "loss": 0.0498,
+      "step": 15242
+    },
+    {
+      "epoch": 2.469701879455606,
+      "grad_norm": 1.0456504821777344,
+      "learning_rate": 3.9865192918553256e-07,
+      "loss": 0.0621,
+      "step": 15243
+    },
+    {
+      "epoch": 2.4698639014906028,
+      "grad_norm": 1.0236272811889648,
+      "learning_rate": 3.9841505153877387e-07,
+      "loss": 0.0589,
+      "step": 15244
+    },
+    {
+      "epoch": 2.4700259235255997,
+      "grad_norm": 0.982083797454834,
+      "learning_rate": 3.981782381954283e-07,
+      "loss": 0.0691,
+      "step": 15245
+    },
+    {
+      "epoch": 2.470187945560596,
+      "grad_norm": 0.8207711577415466,
+      "learning_rate": 3.9794148916274365e-07,
+      "loss": 0.0525,
+      "step": 15246
+    },
+    {
+      "epoch": 2.470349967595593,
+      "grad_norm": 0.7450169324874878,
+      "learning_rate": 3.977048044479617e-07,
+      "loss": 0.0538,
+      "step": 15247
+    },
+    {
+      "epoch": 2.4705119896305896,
+      "grad_norm": 0.9385082125663757,
+      "learning_rate": 3.974681840583255e-07,
+      "loss": 0.0647,
+      "step": 15248
+    },
+    {
+      "epoch": 2.4706740116655865,
+      "grad_norm": 0.9451351165771484,
+      "learning_rate": 3.972316280010749e-07,
+      "loss": 0.0596,
+      "step": 15249
+    },
+    {
+      "epoch": 2.4708360337005835,
+      "grad_norm": 1.1026972532272339,
+      "learning_rate": 3.969951362834476e-07,
+      "loss": 0.0632,
+      "step": 15250
+    },
+    {
+      "epoch": 2.47099805573558,
+      "grad_norm": 0.82200026512146,
+      "learning_rate": 3.967587089126801e-07,
+      "loss": 0.0561,
+      "step": 15251
+    },
+    {
+      "epoch": 2.471160077770577,
+      "grad_norm": 0.9545662999153137,
+      "learning_rate": 3.965223458960063e-07,
+      "loss": 0.0698,
+      "step": 15252
+    },
+    {
+      "epoch": 2.4713220998055734,
+      "grad_norm": 1.0327435731887817,
+      "learning_rate": 3.9628604724065907e-07,
+      "loss": 0.0596,
+      "step": 15253
+    },
+    {
+      "epoch": 2.4714841218405703,
+      "grad_norm": 0.9295939207077026,
+      "learning_rate": 3.9604981295386673e-07,
+      "loss": 0.0572,
+      "step": 15254
+    },
+    {
+      "epoch": 2.4716461438755672,
+      "grad_norm": 0.9229394793510437,
+      "learning_rate": 3.958136430428594e-07,
+      "loss": 0.0621,
+      "step": 15255
+    },
+    {
+      "epoch": 2.4718081659105637,
+      "grad_norm": 0.9629893898963928,
+      "learning_rate": 3.9557753751486237e-07,
+      "loss": 0.0589,
+      "step": 15256
+    },
+    {
+      "epoch": 2.4719701879455607,
+      "grad_norm": 0.985286295413971,
+      "learning_rate": 3.9534149637710073e-07,
+      "loss": 0.0636,
+      "step": 15257
+    },
+    {
+      "epoch": 2.4721322099805576,
+      "grad_norm": 0.8284005522727966,
+      "learning_rate": 3.9510551963679534e-07,
+      "loss": 0.0588,
+      "step": 15258
+    },
+    {
+      "epoch": 2.472294232015554,
+      "grad_norm": 1.0502867698669434,
+      "learning_rate": 3.948696073011668e-07,
+      "loss": 0.0616,
+      "step": 15259
+    },
+    {
+      "epoch": 2.472456254050551,
+      "grad_norm": 0.9685097932815552,
+      "learning_rate": 3.9463375937743546e-07,
+      "loss": 0.0636,
+      "step": 15260
+    },
+    {
+      "epoch": 2.4726182760855475,
+      "grad_norm": 0.9071670174598694,
+      "learning_rate": 3.943979758728153e-07,
+      "loss": 0.0646,
+      "step": 15261
+    },
+    {
+      "epoch": 2.4727802981205445,
+      "grad_norm": 0.93815678358078,
+      "learning_rate": 3.941622567945216e-07,
+      "loss": 0.0576,
+      "step": 15262
+    },
+    {
+      "epoch": 2.472942320155541,
+      "grad_norm": 0.8597636222839355,
+      "learning_rate": 3.93926602149767e-07,
+      "loss": 0.0634,
+      "step": 15263
+    },
+    {
+      "epoch": 2.473104342190538,
+      "grad_norm": 0.9428747892379761,
+      "learning_rate": 3.9369101194576156e-07,
+      "loss": 0.058,
+      "step": 15264
+    },
+    {
+      "epoch": 2.473266364225535,
+      "grad_norm": 0.8766636252403259,
+      "learning_rate": 3.934554861897141e-07,
+      "loss": 0.0541,
+      "step": 15265
+    },
+    {
+      "epoch": 2.4734283862605313,
+      "grad_norm": 0.9514452219009399,
+      "learning_rate": 3.93220024888831e-07,
+      "loss": 0.0595,
+      "step": 15266
+    },
+    {
+      "epoch": 2.4735904082955282,
+      "grad_norm": 1.0236924886703491,
+      "learning_rate": 3.929846280503169e-07,
+      "loss": 0.0666,
+      "step": 15267
+    },
+    {
+      "epoch": 2.473752430330525,
+      "grad_norm": 0.8765333294868469,
+      "learning_rate": 3.927492956813747e-07,
+      "loss": 0.0601,
+      "step": 15268
+    },
+    {
+      "epoch": 2.4739144523655217,
+      "grad_norm": 0.871446967124939,
+      "learning_rate": 3.925140277892037e-07,
+      "loss": 0.0566,
+      "step": 15269
+    },
+    {
+      "epoch": 2.4740764744005186,
+      "grad_norm": 0.8625292181968689,
+      "learning_rate": 3.922788243810038e-07,
+      "loss": 0.0563,
+      "step": 15270
+    },
+    {
+      "epoch": 2.474238496435515,
+      "grad_norm": 0.8439027070999146,
+      "learning_rate": 3.9204368546397144e-07,
+      "loss": 0.0631,
+      "step": 15271
+    },
+    {
+      "epoch": 2.474400518470512,
+      "grad_norm": 0.9690973162651062,
+      "learning_rate": 3.918086110453015e-07,
+      "loss": 0.0647,
+      "step": 15272
+    },
+    {
+      "epoch": 2.4745625405055085,
+      "grad_norm": 0.9805171489715576,
+      "learning_rate": 3.915736011321855e-07,
+      "loss": 0.0608,
+      "step": 15273
+    },
+    {
+      "epoch": 2.4747245625405054,
+      "grad_norm": 0.8432701230049133,
+      "learning_rate": 3.9133865573181524e-07,
+      "loss": 0.0525,
+      "step": 15274
+    },
+    {
+      "epoch": 2.4748865845755024,
+      "grad_norm": 1.0578525066375732,
+      "learning_rate": 3.9110377485138017e-07,
+      "loss": 0.0603,
+      "step": 15275
+    },
+    {
+      "epoch": 2.475048606610499,
+      "grad_norm": 0.9432248473167419,
+      "learning_rate": 3.9086895849806547e-07,
+      "loss": 0.0623,
+      "step": 15276
+    },
+    {
+      "epoch": 2.475210628645496,
+      "grad_norm": 0.8474089503288269,
+      "learning_rate": 3.9063420667905637e-07,
+      "loss": 0.0598,
+      "step": 15277
+    },
+    {
+      "epoch": 2.4753726506804927,
+      "grad_norm": 0.940773069858551,
+      "learning_rate": 3.903995194015364e-07,
+      "loss": 0.053,
+      "step": 15278
+    },
+    {
+      "epoch": 2.475534672715489,
+      "grad_norm": 1.0299676656723022,
+      "learning_rate": 3.9016489667268563e-07,
+      "loss": 0.0628,
+      "step": 15279
+    },
+    {
+      "epoch": 2.475696694750486,
+      "grad_norm": 0.8737006187438965,
+      "learning_rate": 3.899303384996836e-07,
+      "loss": 0.0612,
+      "step": 15280
+    },
+    {
+      "epoch": 2.4758587167854826,
+      "grad_norm": 0.9336732029914856,
+      "learning_rate": 3.8969584488970675e-07,
+      "loss": 0.0607,
+      "step": 15281
+    },
+    {
+      "epoch": 2.4760207388204796,
+      "grad_norm": 0.8692206740379333,
+      "learning_rate": 3.894614158499302e-07,
+      "loss": 0.0597,
+      "step": 15282
+    },
+    {
+      "epoch": 2.4761827608554765,
+      "grad_norm": 1.1092135906219482,
+      "learning_rate": 3.892270513875271e-07,
+      "loss": 0.0599,
+      "step": 15283
+    },
+    {
+      "epoch": 2.476344782890473,
+      "grad_norm": 1.1713637113571167,
+      "learning_rate": 3.889927515096681e-07,
+      "loss": 0.0662,
+      "step": 15284
+    },
+    {
+      "epoch": 2.47650680492547,
+      "grad_norm": 1.050701379776001,
+      "learning_rate": 3.887585162235225e-07,
+      "loss": 0.0573,
+      "step": 15285
+    },
+    {
+      "epoch": 2.4766688269604664,
+      "grad_norm": 0.8351500630378723,
+      "learning_rate": 3.885243455362578e-07,
+      "loss": 0.0554,
+      "step": 15286
+    },
+    {
+      "epoch": 2.4768308489954634,
+      "grad_norm": 0.9191629886627197,
+      "learning_rate": 3.882902394550378e-07,
+      "loss": 0.0612,
+      "step": 15287
+    },
+    {
+      "epoch": 2.4769928710304603,
+      "grad_norm": 0.9809505939483643,
+      "learning_rate": 3.8805619798702565e-07,
+      "loss": 0.069,
+      "step": 15288
+    },
+    {
+      "epoch": 2.4771548930654568,
+      "grad_norm": 0.9045706391334534,
+      "learning_rate": 3.878222211393834e-07,
+      "loss": 0.0568,
+      "step": 15289
+    },
+    {
+      "epoch": 2.4773169151004537,
+      "grad_norm": 0.94025719165802,
+      "learning_rate": 3.8758830891927056e-07,
+      "loss": 0.0574,
+      "step": 15290
+    },
+    {
+      "epoch": 2.4774789371354506,
+      "grad_norm": 0.948002278804779,
+      "learning_rate": 3.8735446133384313e-07,
+      "loss": 0.0584,
+      "step": 15291
+    },
+    {
+      "epoch": 2.477640959170447,
+      "grad_norm": 1.0768437385559082,
+      "learning_rate": 3.8712067839025647e-07,
+      "loss": 0.0703,
+      "step": 15292
+    },
+    {
+      "epoch": 2.477802981205444,
+      "grad_norm": 1.0241899490356445,
+      "learning_rate": 3.8688696009566404e-07,
+      "loss": 0.07,
+      "step": 15293
+    },
+    {
+      "epoch": 2.4779650032404406,
+      "grad_norm": 0.8172342777252197,
+      "learning_rate": 3.86653306457217e-07,
+      "loss": 0.0562,
+      "step": 15294
+    },
+    {
+      "epoch": 2.4781270252754375,
+      "grad_norm": 0.9391677379608154,
+      "learning_rate": 3.864197174820647e-07,
+      "loss": 0.0619,
+      "step": 15295
+    },
+    {
+      "epoch": 2.478289047310434,
+      "grad_norm": 0.8953998684883118,
+      "learning_rate": 3.861861931773542e-07,
+      "loss": 0.0588,
+      "step": 15296
+    },
+    {
+      "epoch": 2.478451069345431,
+      "grad_norm": 0.8568360209465027,
+      "learning_rate": 3.8595273355023054e-07,
+      "loss": 0.0573,
+      "step": 15297
+    },
+    {
+      "epoch": 2.478613091380428,
+      "grad_norm": 1.2389211654663086,
+      "learning_rate": 3.8571933860783785e-07,
+      "loss": 0.0682,
+      "step": 15298
+    },
+    {
+      "epoch": 2.4787751134154243,
+      "grad_norm": 0.9053958058357239,
+      "learning_rate": 3.854860083573167e-07,
+      "loss": 0.0544,
+      "step": 15299
+    },
+    {
+      "epoch": 2.4789371354504213,
+      "grad_norm": 0.989845335483551,
+      "learning_rate": 3.8525274280580646e-07,
+      "loss": 0.0537,
+      "step": 15300
+    },
+    {
+      "epoch": 2.479099157485418,
+      "grad_norm": 0.8106503486633301,
+      "learning_rate": 3.850195419604455e-07,
+      "loss": 0.0611,
+      "step": 15301
+    },
+    {
+      "epoch": 2.4792611795204147,
+      "grad_norm": 0.9021776914596558,
+      "learning_rate": 3.8478640582836733e-07,
+      "loss": 0.0567,
+      "step": 15302
+    },
+    {
+      "epoch": 2.4794232015554116,
+      "grad_norm": 0.9458431601524353,
+      "learning_rate": 3.845533344167068e-07,
+      "loss": 0.0572,
+      "step": 15303
+    },
+    {
+      "epoch": 2.479585223590408,
+      "grad_norm": 1.0524413585662842,
+      "learning_rate": 3.8432032773259574e-07,
+      "loss": 0.0643,
+      "step": 15304
+    },
+    {
+      "epoch": 2.479747245625405,
+      "grad_norm": 0.8422672152519226,
+      "learning_rate": 3.84087385783162e-07,
+      "loss": 0.0622,
+      "step": 15305
+    },
+    {
+      "epoch": 2.479909267660402,
+      "grad_norm": 1.0135761499404907,
+      "learning_rate": 3.838545085755341e-07,
+      "loss": 0.0624,
+      "step": 15306
+    },
+    {
+      "epoch": 2.4800712896953985,
+      "grad_norm": 0.9610772728919983,
+      "learning_rate": 3.8362169611683655e-07,
+      "loss": 0.0637,
+      "step": 15307
+    },
+    {
+      "epoch": 2.4802333117303954,
+      "grad_norm": 0.903256356716156,
+      "learning_rate": 3.8338894841419476e-07,
+      "loss": 0.0601,
+      "step": 15308
+    },
+    {
+      "epoch": 2.480395333765392,
+      "grad_norm": 1.0476837158203125,
+      "learning_rate": 3.831562654747284e-07,
+      "loss": 0.0646,
+      "step": 15309
+    },
+    {
+      "epoch": 2.480557355800389,
+      "grad_norm": 0.9361180067062378,
+      "learning_rate": 3.8292364730555754e-07,
+      "loss": 0.057,
+      "step": 15310
+    },
+    {
+      "epoch": 2.4807193778353858,
+      "grad_norm": 0.9486135244369507,
+      "learning_rate": 3.826910939138001e-07,
+      "loss": 0.0601,
+      "step": 15311
+    },
+    {
+      "epoch": 2.4808813998703823,
+      "grad_norm": 0.8238722085952759,
+      "learning_rate": 3.8245860530657126e-07,
+      "loss": 0.0562,
+      "step": 15312
+    },
+    {
+      "epoch": 2.481043421905379,
+      "grad_norm": 0.9989494681358337,
+      "learning_rate": 3.8222618149098473e-07,
+      "loss": 0.0571,
+      "step": 15313
+    },
+    {
+      "epoch": 2.481205443940376,
+      "grad_norm": 0.9721057415008545,
+      "learning_rate": 3.8199382247415236e-07,
+      "loss": 0.0678,
+      "step": 15314
+    },
+    {
+      "epoch": 2.4813674659753726,
+      "grad_norm": 0.9013247489929199,
+      "learning_rate": 3.817615282631831e-07,
+      "loss": 0.0597,
+      "step": 15315
+    },
+    {
+      "epoch": 2.4815294880103695,
+      "grad_norm": 0.9429559707641602,
+      "learning_rate": 3.8152929886518587e-07,
+      "loss": 0.0692,
+      "step": 15316
+    },
+    {
+      "epoch": 2.481691510045366,
+      "grad_norm": 0.803132176399231,
+      "learning_rate": 3.8129713428726454e-07,
+      "loss": 0.0536,
+      "step": 15317
+    },
+    {
+      "epoch": 2.481853532080363,
+      "grad_norm": 0.7906658053398132,
+      "learning_rate": 3.810650345365241e-07,
+      "loss": 0.0508,
+      "step": 15318
+    },
+    {
+      "epoch": 2.4820155541153595,
+      "grad_norm": 0.9531766772270203,
+      "learning_rate": 3.808329996200663e-07,
+      "loss": 0.0648,
+      "step": 15319
+    },
+    {
+      "epoch": 2.4821775761503564,
+      "grad_norm": 0.8146340847015381,
+      "learning_rate": 3.8060102954499024e-07,
+      "loss": 0.0545,
+      "step": 15320
+    },
+    {
+      "epoch": 2.4823395981853533,
+      "grad_norm": 0.9485729932785034,
+      "learning_rate": 3.8036912431839297e-07,
+      "loss": 0.055,
+      "step": 15321
+    },
+    {
+      "epoch": 2.48250162022035,
+      "grad_norm": 0.8682623505592346,
+      "learning_rate": 3.8013728394737216e-07,
+      "loss": 0.0593,
+      "step": 15322
+    },
+    {
+      "epoch": 2.4826636422553467,
+      "grad_norm": 1.042887806892395,
+      "learning_rate": 3.7990550843902017e-07,
+      "loss": 0.0639,
+      "step": 15323
+    },
+    {
+      "epoch": 2.4828256642903437,
+      "grad_norm": 0.9474160075187683,
+      "learning_rate": 3.796737978004289e-07,
+      "loss": 0.0667,
+      "step": 15324
+    },
+    {
+      "epoch": 2.48298768632534,
+      "grad_norm": 0.9614787101745605,
+      "learning_rate": 3.7944215203868843e-07,
+      "loss": 0.0681,
+      "step": 15325
+    },
+    {
+      "epoch": 2.483149708360337,
+      "grad_norm": 0.7932602763175964,
+      "learning_rate": 3.792105711608865e-07,
+      "loss": 0.0587,
+      "step": 15326
+    },
+    {
+      "epoch": 2.4833117303953336,
+      "grad_norm": 0.8903249502182007,
+      "learning_rate": 3.7897905517410877e-07,
+      "loss": 0.0531,
+      "step": 15327
+    },
+    {
+      "epoch": 2.4834737524303305,
+      "grad_norm": 0.9789133667945862,
+      "learning_rate": 3.7874760408543933e-07,
+      "loss": 0.0701,
+      "step": 15328
+    },
+    {
+      "epoch": 2.4836357744653275,
+      "grad_norm": 0.8971397280693054,
+      "learning_rate": 3.785162179019597e-07,
+      "loss": 0.0692,
+      "step": 15329
+    },
+    {
+      "epoch": 2.483797796500324,
+      "grad_norm": 0.9187726974487305,
+      "learning_rate": 3.7828489663075065e-07,
+      "loss": 0.0591,
+      "step": 15330
+    },
+    {
+      "epoch": 2.483959818535321,
+      "grad_norm": 1.0862635374069214,
+      "learning_rate": 3.7805364027888787e-07,
+      "loss": 0.0704,
+      "step": 15331
+    },
+    {
+      "epoch": 2.4841218405703174,
+      "grad_norm": 0.8576655387878418,
+      "learning_rate": 3.778224488534496e-07,
+      "loss": 0.062,
+      "step": 15332
+    },
+    {
+      "epoch": 2.4842838626053143,
+      "grad_norm": 1.0012966394424438,
+      "learning_rate": 3.7759132236150854e-07,
+      "loss": 0.068,
+      "step": 15333
+    },
+    {
+      "epoch": 2.4844458846403112,
+      "grad_norm": 0.8457930088043213,
+      "learning_rate": 3.773602608101376e-07,
+      "loss": 0.0632,
+      "step": 15334
+    },
+    {
+      "epoch": 2.4846079066753077,
+      "grad_norm": 1.0266876220703125,
+      "learning_rate": 3.771292642064056e-07,
+      "loss": 0.0679,
+      "step": 15335
+    },
+    {
+      "epoch": 2.4847699287103047,
+      "grad_norm": 0.8954113721847534,
+      "learning_rate": 3.7689833255737995e-07,
+      "loss": 0.0574,
+      "step": 15336
+    },
+    {
+      "epoch": 2.4849319507453016,
+      "grad_norm": 0.9594757556915283,
+      "learning_rate": 3.7666746587012885e-07,
+      "loss": 0.0611,
+      "step": 15337
+    },
+    {
+      "epoch": 2.485093972780298,
+      "grad_norm": 0.8226574063301086,
+      "learning_rate": 3.764366641517145e-07,
+      "loss": 0.0606,
+      "step": 15338
+    },
+    {
+      "epoch": 2.485255994815295,
+      "grad_norm": 0.9588454365730286,
+      "learning_rate": 3.762059274091989e-07,
+      "loss": 0.0639,
+      "step": 15339
+    },
+    {
+      "epoch": 2.4854180168502915,
+      "grad_norm": 0.913917064666748,
+      "learning_rate": 3.759752556496421e-07,
+      "loss": 0.0661,
+      "step": 15340
+    },
+    {
+      "epoch": 2.4855800388852884,
+      "grad_norm": 0.7962075471878052,
+      "learning_rate": 3.7574464888010363e-07,
+      "loss": 0.0535,
+      "step": 15341
+    },
+    {
+      "epoch": 2.485742060920285,
+      "grad_norm": 0.8324793577194214,
+      "learning_rate": 3.7551410710763764e-07,
+      "loss": 0.0583,
+      "step": 15342
+    },
+    {
+      "epoch": 2.485904082955282,
+      "grad_norm": 1.2726354598999023,
+      "learning_rate": 3.75283630339299e-07,
+      "loss": 0.0629,
+      "step": 15343
+    },
+    {
+      "epoch": 2.486066104990279,
+      "grad_norm": 0.8221186995506287,
+      "learning_rate": 3.7505321858213926e-07,
+      "loss": 0.055,
+      "step": 15344
+    },
+    {
+      "epoch": 2.4862281270252753,
+      "grad_norm": 1.0081355571746826,
+      "learning_rate": 3.7482287184320897e-07,
+      "loss": 0.0632,
+      "step": 15345
+    },
+    {
+      "epoch": 2.4863901490602722,
+      "grad_norm": 0.920180082321167,
+      "learning_rate": 3.7459259012955606e-07,
+      "loss": 0.0622,
+      "step": 15346
+    },
+    {
+      "epoch": 2.486552171095269,
+      "grad_norm": 0.9343588948249817,
+      "learning_rate": 3.743623734482263e-07,
+      "loss": 0.0634,
+      "step": 15347
+    },
+    {
+      "epoch": 2.4867141931302656,
+      "grad_norm": 1.170159935951233,
+      "learning_rate": 3.7413222180626455e-07,
+      "loss": 0.0627,
+      "step": 15348
+    },
+    {
+      "epoch": 2.4868762151652626,
+      "grad_norm": 0.9837419390678406,
+      "learning_rate": 3.7390213521071193e-07,
+      "loss": 0.0612,
+      "step": 15349
+    },
+    {
+      "epoch": 2.487038237200259,
+      "grad_norm": 0.8149502873420715,
+      "learning_rate": 3.736721136686081e-07,
+      "loss": 0.0558,
+      "step": 15350
+    },
+    {
+      "epoch": 2.487200259235256,
+      "grad_norm": 0.8310353755950928,
+      "learning_rate": 3.7344215718699256e-07,
+      "loss": 0.0522,
+      "step": 15351
+    },
+    {
+      "epoch": 2.487362281270253,
+      "grad_norm": 0.8602123260498047,
+      "learning_rate": 3.7321226577290147e-07,
+      "loss": 0.0598,
+      "step": 15352
+    },
+    {
+      "epoch": 2.4875243033052494,
+      "grad_norm": 0.8709070682525635,
+      "learning_rate": 3.7298243943336784e-07,
+      "loss": 0.0583,
+      "step": 15353
+    },
+    {
+      "epoch": 2.4876863253402464,
+      "grad_norm": 0.9925591945648193,
+      "learning_rate": 3.7275267817542425e-07,
+      "loss": 0.0599,
+      "step": 15354
+    },
+    {
+      "epoch": 2.487848347375243,
+      "grad_norm": 0.8003130555152893,
+      "learning_rate": 3.725229820061008e-07,
+      "loss": 0.0569,
+      "step": 15355
+    },
+    {
+      "epoch": 2.48801036941024,
+      "grad_norm": 0.8514525890350342,
+      "learning_rate": 3.7229335093242587e-07,
+      "loss": 0.0528,
+      "step": 15356
+    },
+    {
+      "epoch": 2.4881723914452367,
+      "grad_norm": 0.8922498226165771,
+      "learning_rate": 3.720637849614253e-07,
+      "loss": 0.0614,
+      "step": 15357
+    },
+    {
+      "epoch": 2.488334413480233,
+      "grad_norm": 1.7786387205123901,
+      "learning_rate": 3.7183428410012326e-07,
+      "loss": 0.0647,
+      "step": 15358
+    },
+    {
+      "epoch": 2.48849643551523,
+      "grad_norm": 0.9049399495124817,
+      "learning_rate": 3.716048483555423e-07,
+      "loss": 0.063,
+      "step": 15359
+    },
+    {
+      "epoch": 2.488658457550227,
+      "grad_norm": 0.9355145692825317,
+      "learning_rate": 3.713754777347023e-07,
+      "loss": 0.0621,
+      "step": 15360
+    },
+    {
+      "epoch": 2.4888204795852236,
+      "grad_norm": 1.0823627710342407,
+      "learning_rate": 3.711461722446216e-07,
+      "loss": 0.0676,
+      "step": 15361
+    },
+    {
+      "epoch": 2.4889825016202205,
+      "grad_norm": 0.8411454558372498,
+      "learning_rate": 3.7091693189231615e-07,
+      "loss": 0.0586,
+      "step": 15362
+    },
+    {
+      "epoch": 2.489144523655217,
+      "grad_norm": 0.8815544843673706,
+      "learning_rate": 3.706877566848008e-07,
+      "loss": 0.0552,
+      "step": 15363
+    },
+    {
+      "epoch": 2.489306545690214,
+      "grad_norm": 0.96858811378479,
+      "learning_rate": 3.704586466290863e-07,
+      "loss": 0.0595,
+      "step": 15364
+    },
+    {
+      "epoch": 2.4894685677252104,
+      "grad_norm": 0.8106810450553894,
+      "learning_rate": 3.7022960173218437e-07,
+      "loss": 0.0564,
+      "step": 15365
+    },
+    {
+      "epoch": 2.4896305897602073,
+      "grad_norm": 0.8244809508323669,
+      "learning_rate": 3.7000062200110266e-07,
+      "loss": 0.0668,
+      "step": 15366
+    },
+    {
+      "epoch": 2.4897926117952043,
+      "grad_norm": 0.9649903178215027,
+      "learning_rate": 3.6977170744284805e-07,
+      "loss": 0.0636,
+      "step": 15367
+    },
+    {
+      "epoch": 2.4899546338302008,
+      "grad_norm": 1.087392807006836,
+      "learning_rate": 3.6954285806442337e-07,
+      "loss": 0.0681,
+      "step": 15368
+    },
+    {
+      "epoch": 2.4901166558651977,
+      "grad_norm": 1.129231333732605,
+      "learning_rate": 3.6931407387283126e-07,
+      "loss": 0.0688,
+      "step": 15369
+    },
+    {
+      "epoch": 2.4902786779001946,
+      "grad_norm": 0.8219298720359802,
+      "learning_rate": 3.6908535487507335e-07,
+      "loss": 0.0577,
+      "step": 15370
+    },
+    {
+      "epoch": 2.490440699935191,
+      "grad_norm": 0.8305365443229675,
+      "learning_rate": 3.688567010781463e-07,
+      "loss": 0.0571,
+      "step": 15371
+    },
+    {
+      "epoch": 2.490602721970188,
+      "grad_norm": 0.9187493324279785,
+      "learning_rate": 3.68628112489047e-07,
+      "loss": 0.0593,
+      "step": 15372
+    },
+    {
+      "epoch": 2.4907647440051845,
+      "grad_norm": 0.8878763914108276,
+      "learning_rate": 3.683995891147696e-07,
+      "loss": 0.0579,
+      "step": 15373
+    },
+    {
+      "epoch": 2.4909267660401815,
+      "grad_norm": 0.9653205275535583,
+      "learning_rate": 3.681711309623065e-07,
+      "loss": 0.0578,
+      "step": 15374
+    },
+    {
+      "epoch": 2.4910887880751784,
+      "grad_norm": 0.8415849208831787,
+      "learning_rate": 3.679427380386477e-07,
+      "loss": 0.0582,
+      "step": 15375
+    },
+    {
+      "epoch": 2.491250810110175,
+      "grad_norm": 0.9508287310600281,
+      "learning_rate": 3.677144103507818e-07,
+      "loss": 0.0608,
+      "step": 15376
+    },
+    {
+      "epoch": 2.491412832145172,
+      "grad_norm": 0.9753308892250061,
+      "learning_rate": 3.674861479056946e-07,
+      "loss": 0.0614,
+      "step": 15377
+    },
+    {
+      "epoch": 2.4915748541801683,
+      "grad_norm": 0.8694376349449158,
+      "learning_rate": 3.672579507103716e-07,
+      "loss": 0.0622,
+      "step": 15378
+    },
+    {
+      "epoch": 2.4917368762151653,
+      "grad_norm": 0.854770839214325,
+      "learning_rate": 3.67029818771793e-07,
+      "loss": 0.0606,
+      "step": 15379
+    },
+    {
+      "epoch": 2.491898898250162,
+      "grad_norm": 0.8874652981758118,
+      "learning_rate": 3.668017520969405e-07,
+      "loss": 0.0647,
+      "step": 15380
+    },
+    {
+      "epoch": 2.4920609202851587,
+      "grad_norm": 0.9091949462890625,
+      "learning_rate": 3.66573750692793e-07,
+      "loss": 0.0686,
+      "step": 15381
+    },
+    {
+      "epoch": 2.4922229423201556,
+      "grad_norm": 0.9150681495666504,
+      "learning_rate": 3.663458145663254e-07,
+      "loss": 0.0611,
+      "step": 15382
+    },
+    {
+      "epoch": 2.492384964355152,
+      "grad_norm": 1.12820565700531,
+      "learning_rate": 3.6611794372451244e-07,
+      "loss": 0.0716,
+      "step": 15383
+    },
+    {
+      "epoch": 2.492546986390149,
+      "grad_norm": 0.825421154499054,
+      "learning_rate": 3.65890138174326e-07,
+      "loss": 0.0642,
+      "step": 15384
+    },
+    {
+      "epoch": 2.492709008425146,
+      "grad_norm": 0.9262431263923645,
+      "learning_rate": 3.6566239792273775e-07,
+      "loss": 0.0597,
+      "step": 15385
+    },
+    {
+      "epoch": 2.4928710304601425,
+      "grad_norm": 0.9032964110374451,
+      "learning_rate": 3.6543472297671495e-07,
+      "loss": 0.0627,
+      "step": 15386
+    },
+    {
+      "epoch": 2.4930330524951394,
+      "grad_norm": 0.8417792320251465,
+      "learning_rate": 3.6520711334322387e-07,
+      "loss": 0.0637,
+      "step": 15387
+    },
+    {
+      "epoch": 2.493195074530136,
+      "grad_norm": 0.9112814664840698,
+      "learning_rate": 3.6497956902922904e-07,
+      "loss": 0.0601,
+      "step": 15388
+    },
+    {
+      "epoch": 2.493357096565133,
+      "grad_norm": 1.061842918395996,
+      "learning_rate": 3.6475209004169286e-07,
+      "loss": 0.0566,
+      "step": 15389
+    },
+    {
+      "epoch": 2.4935191186001298,
+      "grad_norm": 0.9109911918640137,
+      "learning_rate": 3.645246763875754e-07,
+      "loss": 0.0605,
+      "step": 15390
+    },
+    {
+      "epoch": 2.4936811406351262,
+      "grad_norm": 0.8097963929176331,
+      "learning_rate": 3.6429732807383517e-07,
+      "loss": 0.0591,
+      "step": 15391
+    },
+    {
+      "epoch": 2.493843162670123,
+      "grad_norm": 0.7953127026557922,
+      "learning_rate": 3.640700451074289e-07,
+      "loss": 0.0537,
+      "step": 15392
+    },
+    {
+      "epoch": 2.49400518470512,
+      "grad_norm": 1.0355387926101685,
+      "learning_rate": 3.63842827495309e-07,
+      "loss": 0.0673,
+      "step": 15393
+    },
+    {
+      "epoch": 2.4941672067401166,
+      "grad_norm": 0.8306580781936646,
+      "learning_rate": 3.636156752444303e-07,
+      "loss": 0.0578,
+      "step": 15394
+    },
+    {
+      "epoch": 2.4943292287751135,
+      "grad_norm": 1.0002905130386353,
+      "learning_rate": 3.633885883617416e-07,
+      "loss": 0.0626,
+      "step": 15395
+    },
+    {
+      "epoch": 2.49449125081011,
+      "grad_norm": 0.9213991165161133,
+      "learning_rate": 3.631615668541921e-07,
+      "loss": 0.0528,
+      "step": 15396
+    },
+    {
+      "epoch": 2.494653272845107,
+      "grad_norm": 1.1550705432891846,
+      "learning_rate": 3.6293461072872735e-07,
+      "loss": 0.066,
+      "step": 15397
+    },
+    {
+      "epoch": 2.4948152948801035,
+      "grad_norm": 0.9875110387802124,
+      "learning_rate": 3.6270771999229124e-07,
+      "loss": 0.0553,
+      "step": 15398
+    },
+    {
+      "epoch": 2.4949773169151004,
+      "grad_norm": 0.9158915281295776,
+      "learning_rate": 3.6248089465182797e-07,
+      "loss": 0.0621,
+      "step": 15399
+    },
+    {
+      "epoch": 2.4951393389500973,
+      "grad_norm": 0.9057533740997314,
+      "learning_rate": 3.622541347142758e-07,
+      "loss": 0.0537,
+      "step": 15400
+    },
+    {
+      "epoch": 2.495301360985094,
+      "grad_norm": 0.8839161396026611,
+      "learning_rate": 3.6202744018657393e-07,
+      "loss": 0.0623,
+      "step": 15401
+    },
+    {
+      "epoch": 2.4954633830200907,
+      "grad_norm": 0.9387931227684021,
+      "learning_rate": 3.618008110756588e-07,
+      "loss": 0.0662,
+      "step": 15402
+    },
+    {
+      "epoch": 2.4956254050550877,
+      "grad_norm": 1.0209976434707642,
+      "learning_rate": 3.6157424738846427e-07,
+      "loss": 0.0676,
+      "step": 15403
+    },
+    {
+      "epoch": 2.495787427090084,
+      "grad_norm": 0.9517818689346313,
+      "learning_rate": 3.6134774913192314e-07,
+      "loss": 0.0619,
+      "step": 15404
+    },
+    {
+      "epoch": 2.495949449125081,
+      "grad_norm": 0.9246603846549988,
+      "learning_rate": 3.6112131631296507e-07,
+      "loss": 0.0608,
+      "step": 15405
+    },
+    {
+      "epoch": 2.4961114711600776,
+      "grad_norm": 0.9202209711074829,
+      "learning_rate": 3.608949489385191e-07,
+      "loss": 0.0601,
+      "step": 15406
+    },
+    {
+      "epoch": 2.4962734931950745,
+      "grad_norm": 0.8076236248016357,
+      "learning_rate": 3.60668647015511e-07,
+      "loss": 0.0593,
+      "step": 15407
+    },
+    {
+      "epoch": 2.4964355152300715,
+      "grad_norm": 0.9734016060829163,
+      "learning_rate": 3.6044241055086525e-07,
+      "loss": 0.0606,
+      "step": 15408
+    },
+    {
+      "epoch": 2.496597537265068,
+      "grad_norm": 1.0602586269378662,
+      "learning_rate": 3.602162395515041e-07,
+      "loss": 0.066,
+      "step": 15409
+    },
+    {
+      "epoch": 2.496759559300065,
+      "grad_norm": 1.0129172801971436,
+      "learning_rate": 3.599901340243478e-07,
+      "loss": 0.0573,
+      "step": 15410
+    },
+    {
+      "epoch": 2.4969215813350614,
+      "grad_norm": 0.8970673084259033,
+      "learning_rate": 3.597640939763153e-07,
+      "loss": 0.0616,
+      "step": 15411
+    },
+    {
+      "epoch": 2.4970836033700583,
+      "grad_norm": 0.8650174140930176,
+      "learning_rate": 3.5953811941432104e-07,
+      "loss": 0.0544,
+      "step": 15412
+    },
+    {
+      "epoch": 2.4972456254050552,
+      "grad_norm": 0.920050323009491,
+      "learning_rate": 3.593122103452812e-07,
+      "loss": 0.0581,
+      "step": 15413
+    },
+    {
+      "epoch": 2.4974076474400517,
+      "grad_norm": 0.9447347521781921,
+      "learning_rate": 3.590863667761077e-07,
+      "loss": 0.0596,
+      "step": 15414
+    },
+    {
+      "epoch": 2.4975696694750487,
+      "grad_norm": 0.8913710713386536,
+      "learning_rate": 3.5886058871371005e-07,
+      "loss": 0.0622,
+      "step": 15415
+    },
+    {
+      "epoch": 2.4977316915100456,
+      "grad_norm": 1.1236344575881958,
+      "learning_rate": 3.5863487616499713e-07,
+      "loss": 0.064,
+      "step": 15416
+    },
+    {
+      "epoch": 2.497893713545042,
+      "grad_norm": 0.8553385138511658,
+      "learning_rate": 3.584092291368746e-07,
+      "loss": 0.0516,
+      "step": 15417
+    },
+    {
+      "epoch": 2.498055735580039,
+      "grad_norm": 0.8064867854118347,
+      "learning_rate": 3.581836476362474e-07,
+      "loss": 0.055,
+      "step": 15418
+    },
+    {
+      "epoch": 2.4982177576150355,
+      "grad_norm": 1.035109043121338,
+      "learning_rate": 3.579581316700173e-07,
+      "loss": 0.0661,
+      "step": 15419
+    },
+    {
+      "epoch": 2.4983797796500324,
+      "grad_norm": 0.9095247983932495,
+      "learning_rate": 3.5773268124508485e-07,
+      "loss": 0.0644,
+      "step": 15420
+    },
+    {
+      "epoch": 2.498541801685029,
+      "grad_norm": 0.7617493867874146,
+      "learning_rate": 3.575072963683482e-07,
+      "loss": 0.055,
+      "step": 15421
+    },
+    {
+      "epoch": 2.498703823720026,
+      "grad_norm": 0.863577663898468,
+      "learning_rate": 3.5728197704670344e-07,
+      "loss": 0.0569,
+      "step": 15422
+    },
+    {
+      "epoch": 2.498865845755023,
+      "grad_norm": 0.8674677610397339,
+      "learning_rate": 3.5705672328704503e-07,
+      "loss": 0.0629,
+      "step": 15423
+    },
+    {
+      "epoch": 2.4990278677900193,
+      "grad_norm": 0.8979498744010925,
+      "learning_rate": 3.5683153509626504e-07,
+      "loss": 0.0641,
+      "step": 15424
+    },
+    {
+      "epoch": 2.499189889825016,
+      "grad_norm": 1.075107216835022,
+      "learning_rate": 3.566064124812541e-07,
+      "loss": 0.0667,
+      "step": 15425
+    },
+    {
+      "epoch": 2.499351911860013,
+      "grad_norm": 0.913252055644989,
+      "learning_rate": 3.563813554488996e-07,
+      "loss": 0.0589,
+      "step": 15426
+    },
+    {
+      "epoch": 2.4995139338950096,
+      "grad_norm": 0.9128844141960144,
+      "learning_rate": 3.561563640060875e-07,
+      "loss": 0.0632,
+      "step": 15427
+    },
+    {
+      "epoch": 2.4996759559300066,
+      "grad_norm": 0.9388177394866943,
+      "learning_rate": 3.559314381597034e-07,
+      "loss": 0.0677,
+      "step": 15428
+    },
+    {
+      "epoch": 2.499837977965003,
+      "grad_norm": 0.8632997870445251,
+      "learning_rate": 3.557065779166291e-07,
+      "loss": 0.0533,
+      "step": 15429
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.9118877649307251,
+      "learning_rate": 3.55481783283744e-07,
+      "loss": 0.0619,
+      "step": 15430
+    },
+    {
+      "epoch": 2.5001620220349965,
+      "grad_norm": 1.0225169658660889,
+      "learning_rate": 3.5525705426792624e-07,
+      "loss": 0.0709,
+      "step": 15431
+    },
+    {
+      "epoch": 2.5003240440699934,
+      "grad_norm": 0.8662908673286438,
+      "learning_rate": 3.5503239087605337e-07,
+      "loss": 0.0596,
+      "step": 15432
+    },
+    {
+      "epoch": 2.5004860661049904,
+      "grad_norm": 0.7899366617202759,
+      "learning_rate": 3.548077931149982e-07,
+      "loss": 0.0581,
+      "step": 15433
+    },
+    {
+      "epoch": 2.500648088139987,
+      "grad_norm": 0.9061411023139954,
+      "learning_rate": 3.54583260991633e-07,
+      "loss": 0.0606,
+      "step": 15434
+    },
+    {
+      "epoch": 2.500810110174984,
+      "grad_norm": 0.8774332404136658,
+      "learning_rate": 3.543587945128285e-07,
+      "loss": 0.0537,
+      "step": 15435
+    },
+    {
+      "epoch": 2.5009721322099807,
+      "grad_norm": 1.0649468898773193,
+      "learning_rate": 3.541343936854524e-07,
+      "loss": 0.0656,
+      "step": 15436
+    },
+    {
+      "epoch": 2.501134154244977,
+      "grad_norm": 0.9664431810379028,
+      "learning_rate": 3.53910058516371e-07,
+      "loss": 0.0612,
+      "step": 15437
+    },
+    {
+      "epoch": 2.501296176279974,
+      "grad_norm": 1.0078061819076538,
+      "learning_rate": 3.5368578901244843e-07,
+      "loss": 0.0584,
+      "step": 15438
+    },
+    {
+      "epoch": 2.501458198314971,
+      "grad_norm": 1.0571532249450684,
+      "learning_rate": 3.5346158518054674e-07,
+      "loss": 0.0649,
+      "step": 15439
+    },
+    {
+      "epoch": 2.5016202203499676,
+      "grad_norm": 0.8381620049476624,
+      "learning_rate": 3.5323744702752657e-07,
+      "loss": 0.0511,
+      "step": 15440
+    },
+    {
+      "epoch": 2.5017822423849645,
+      "grad_norm": 0.9569916129112244,
+      "learning_rate": 3.5301337456024434e-07,
+      "loss": 0.0622,
+      "step": 15441
+    },
+    {
+      "epoch": 2.501944264419961,
+      "grad_norm": 0.8359852433204651,
+      "learning_rate": 3.5278936778555763e-07,
+      "loss": 0.0558,
+      "step": 15442
+    },
+    {
+      "epoch": 2.502106286454958,
+      "grad_norm": 0.9020277857780457,
+      "learning_rate": 3.525654267103207e-07,
+      "loss": 0.0602,
+      "step": 15443
+    },
+    {
+      "epoch": 2.5022683084899544,
+      "grad_norm": 0.7572386860847473,
+      "learning_rate": 3.523415513413847e-07,
+      "loss": 0.0566,
+      "step": 15444
+    },
+    {
+      "epoch": 2.5024303305249513,
+      "grad_norm": 0.9130625128746033,
+      "learning_rate": 3.5211774168559976e-07,
+      "loss": 0.0552,
+      "step": 15445
+    },
+    {
+      "epoch": 2.5025923525599483,
+      "grad_norm": 1.0692723989486694,
+      "learning_rate": 3.518939977498137e-07,
+      "loss": 0.0597,
+      "step": 15446
+    },
+    {
+      "epoch": 2.5027543745949448,
+      "grad_norm": 0.9212605357170105,
+      "learning_rate": 3.516703195408741e-07,
+      "loss": 0.0638,
+      "step": 15447
+    },
+    {
+      "epoch": 2.5029163966299417,
+      "grad_norm": 0.9287461042404175,
+      "learning_rate": 3.514467070656233e-07,
+      "loss": 0.0586,
+      "step": 15448
+    },
+    {
+      "epoch": 2.5030784186649386,
+      "grad_norm": 0.8843840956687927,
+      "learning_rate": 3.512231603309038e-07,
+      "loss": 0.063,
+      "step": 15449
+    },
+    {
+      "epoch": 2.503240440699935,
+      "grad_norm": 0.8561710119247437,
+      "learning_rate": 3.509996793435558e-07,
+      "loss": 0.0593,
+      "step": 15450
+    },
+    {
+      "epoch": 2.503402462734932,
+      "grad_norm": 1.0155240297317505,
+      "learning_rate": 3.5077626411041707e-07,
+      "loss": 0.0709,
+      "step": 15451
+    },
+    {
+      "epoch": 2.5035644847699285,
+      "grad_norm": 0.9709542393684387,
+      "learning_rate": 3.505529146383235e-07,
+      "loss": 0.0626,
+      "step": 15452
+    },
+    {
+      "epoch": 2.5037265068049255,
+      "grad_norm": 0.9691731929779053,
+      "learning_rate": 3.503296309341095e-07,
+      "loss": 0.0617,
+      "step": 15453
+    },
+    {
+      "epoch": 2.503888528839922,
+      "grad_norm": 0.9437772035598755,
+      "learning_rate": 3.501064130046064e-07,
+      "loss": 0.0593,
+      "step": 15454
+    },
+    {
+      "epoch": 2.504050550874919,
+      "grad_norm": 0.8508868217468262,
+      "learning_rate": 3.4988326085664463e-07,
+      "loss": 0.0586,
+      "step": 15455
+    },
+    {
+      "epoch": 2.504212572909916,
+      "grad_norm": 0.8837978839874268,
+      "learning_rate": 3.496601744970518e-07,
+      "loss": 0.0541,
+      "step": 15456
+    },
+    {
+      "epoch": 2.5043745949449123,
+      "grad_norm": 0.975632905960083,
+      "learning_rate": 3.494371539326538e-07,
+      "loss": 0.0556,
+      "step": 15457
+    },
+    {
+      "epoch": 2.5045366169799093,
+      "grad_norm": 1.0804563760757446,
+      "learning_rate": 3.492141991702752e-07,
+      "loss": 0.0585,
+      "step": 15458
+    },
+    {
+      "epoch": 2.504698639014906,
+      "grad_norm": 0.832747757434845,
+      "learning_rate": 3.4899131021673693e-07,
+      "loss": 0.0584,
+      "step": 15459
+    },
+    {
+      "epoch": 2.5048606610499027,
+      "grad_norm": 0.9667251110076904,
+      "learning_rate": 3.4876848707885854e-07,
+      "loss": 0.0595,
+      "step": 15460
+    },
+    {
+      "epoch": 2.5050226830848996,
+      "grad_norm": 0.9770776629447937,
+      "learning_rate": 3.4854572976345954e-07,
+      "loss": 0.0636,
+      "step": 15461
+    },
+    {
+      "epoch": 2.5051847051198965,
+      "grad_norm": 0.8421487212181091,
+      "learning_rate": 3.483230382773545e-07,
+      "loss": 0.0574,
+      "step": 15462
+    },
+    {
+      "epoch": 2.505346727154893,
+      "grad_norm": 1.0473026037216187,
+      "learning_rate": 3.481004126273574e-07,
+      "loss": 0.0623,
+      "step": 15463
+    },
+    {
+      "epoch": 2.50550874918989,
+      "grad_norm": 0.7978666424751282,
+      "learning_rate": 3.478778528202803e-07,
+      "loss": 0.0544,
+      "step": 15464
+    },
+    {
+      "epoch": 2.5056707712248865,
+      "grad_norm": 1.039564609527588,
+      "learning_rate": 3.476553588629328e-07,
+      "loss": 0.0655,
+      "step": 15465
+    },
+    {
+      "epoch": 2.5058327932598834,
+      "grad_norm": 0.7758963108062744,
+      "learning_rate": 3.474329307621227e-07,
+      "loss": 0.053,
+      "step": 15466
+    },
+    {
+      "epoch": 2.50599481529488,
+      "grad_norm": 0.9446272850036621,
+      "learning_rate": 3.4721056852465575e-07,
+      "loss": 0.0561,
+      "step": 15467
+    },
+    {
+      "epoch": 2.506156837329877,
+      "grad_norm": 0.9161103963851929,
+      "learning_rate": 3.469882721573356e-07,
+      "loss": 0.0607,
+      "step": 15468
+    },
+    {
+      "epoch": 2.5063188593648738,
+      "grad_norm": 0.8289987444877625,
+      "learning_rate": 3.467660416669649e-07,
+      "loss": 0.0606,
+      "step": 15469
+    },
+    {
+      "epoch": 2.5064808813998702,
+      "grad_norm": 0.985321581363678,
+      "learning_rate": 3.465438770603416e-07,
+      "loss": 0.0652,
+      "step": 15470
+    },
+    {
+      "epoch": 2.506642903434867,
+      "grad_norm": 0.8236386179924011,
+      "learning_rate": 3.463217783442649e-07,
+      "loss": 0.0561,
+      "step": 15471
+    },
+    {
+      "epoch": 2.506804925469864,
+      "grad_norm": 1.0456392765045166,
+      "learning_rate": 3.4609974552552993e-07,
+      "loss": 0.0686,
+      "step": 15472
+    },
+    {
+      "epoch": 2.5069669475048606,
+      "grad_norm": 1.1649935245513916,
+      "learning_rate": 3.4587777861093105e-07,
+      "loss": 0.0645,
+      "step": 15473
+    },
+    {
+      "epoch": 2.5071289695398575,
+      "grad_norm": 0.954638659954071,
+      "learning_rate": 3.456558776072585e-07,
+      "loss": 0.0642,
+      "step": 15474
+    },
+    {
+      "epoch": 2.507290991574854,
+      "grad_norm": 0.8497850298881531,
+      "learning_rate": 3.4543404252130234e-07,
+      "loss": 0.0587,
+      "step": 15475
+    },
+    {
+      "epoch": 2.507453013609851,
+      "grad_norm": 0.8479697108268738,
+      "learning_rate": 3.4521227335985146e-07,
+      "loss": 0.0582,
+      "step": 15476
+    },
+    {
+      "epoch": 2.5076150356448474,
+      "grad_norm": 0.8747241497039795,
+      "learning_rate": 3.449905701296902e-07,
+      "loss": 0.0555,
+      "step": 15477
+    },
+    {
+      "epoch": 2.5077770576798444,
+      "grad_norm": 0.9284306764602661,
+      "learning_rate": 3.447689328376022e-07,
+      "loss": 0.0651,
+      "step": 15478
+    },
+    {
+      "epoch": 2.5079390797148413,
+      "grad_norm": 0.9648377299308777,
+      "learning_rate": 3.445473614903688e-07,
+      "loss": 0.069,
+      "step": 15479
+    },
+    {
+      "epoch": 2.508101101749838,
+      "grad_norm": 0.9850466251373291,
+      "learning_rate": 3.4432585609477125e-07,
+      "loss": 0.0611,
+      "step": 15480
+    },
+    {
+      "epoch": 2.5082631237848347,
+      "grad_norm": 0.8666247725486755,
+      "learning_rate": 3.441044166575855e-07,
+      "loss": 0.0637,
+      "step": 15481
+    },
+    {
+      "epoch": 2.5084251458198317,
+      "grad_norm": 0.8115009069442749,
+      "learning_rate": 3.438830431855872e-07,
+      "loss": 0.052,
+      "step": 15482
+    },
+    {
+      "epoch": 2.508587167854828,
+      "grad_norm": 1.0275858640670776,
+      "learning_rate": 3.4366173568555013e-07,
+      "loss": 0.0657,
+      "step": 15483
+    },
+    {
+      "epoch": 2.508749189889825,
+      "grad_norm": 0.9150252342224121,
+      "learning_rate": 3.434404941642455e-07,
+      "loss": 0.0587,
+      "step": 15484
+    },
+    {
+      "epoch": 2.508911211924822,
+      "grad_norm": 1.8770604133605957,
+      "learning_rate": 3.4321931862844327e-07,
+      "loss": 0.0569,
+      "step": 15485
+    },
+    {
+      "epoch": 2.5090732339598185,
+      "grad_norm": 1.0579850673675537,
+      "learning_rate": 3.4299820908491045e-07,
+      "loss": 0.0671,
+      "step": 15486
+    },
+    {
+      "epoch": 2.5092352559948155,
+      "grad_norm": 0.8055444955825806,
+      "learning_rate": 3.427771655404133e-07,
+      "loss": 0.0595,
+      "step": 15487
+    },
+    {
+      "epoch": 2.509397278029812,
+      "grad_norm": 0.8405749797821045,
+      "learning_rate": 3.4255618800171366e-07,
+      "loss": 0.0574,
+      "step": 15488
+    },
+    {
+      "epoch": 2.509559300064809,
+      "grad_norm": 0.9373750686645508,
+      "learning_rate": 3.423352764755733e-07,
+      "loss": 0.0611,
+      "step": 15489
+    },
+    {
+      "epoch": 2.5097213220998054,
+      "grad_norm": 0.957573652267456,
+      "learning_rate": 3.421144309687527e-07,
+      "loss": 0.0701,
+      "step": 15490
+    },
+    {
+      "epoch": 2.5098833441348023,
+      "grad_norm": 0.9812716245651245,
+      "learning_rate": 3.418936514880092e-07,
+      "loss": 0.0618,
+      "step": 15491
+    },
+    {
+      "epoch": 2.5100453661697992,
+      "grad_norm": 0.9846394658088684,
+      "learning_rate": 3.4167293804009656e-07,
+      "loss": 0.0691,
+      "step": 15492
+    },
+    {
+      "epoch": 2.5102073882047957,
+      "grad_norm": 1.0045005083084106,
+      "learning_rate": 3.414522906317691e-07,
+      "loss": 0.0551,
+      "step": 15493
+    },
+    {
+      "epoch": 2.5103694102397927,
+      "grad_norm": 0.8848669528961182,
+      "learning_rate": 3.412317092697781e-07,
+      "loss": 0.0637,
+      "step": 15494
+    },
+    {
+      "epoch": 2.5105314322747896,
+      "grad_norm": 0.9748433232307434,
+      "learning_rate": 3.4101119396087237e-07,
+      "loss": 0.0644,
+      "step": 15495
+    },
+    {
+      "epoch": 2.510693454309786,
+      "grad_norm": 0.8080410957336426,
+      "learning_rate": 3.407907447117997e-07,
+      "loss": 0.0569,
+      "step": 15496
+    },
+    {
+      "epoch": 2.510855476344783,
+      "grad_norm": 0.9153392314910889,
+      "learning_rate": 3.405703615293052e-07,
+      "loss": 0.0655,
+      "step": 15497
+    },
+    {
+      "epoch": 2.5110174983797795,
+      "grad_norm": 0.8493557572364807,
+      "learning_rate": 3.4035004442013157e-07,
+      "loss": 0.0581,
+      "step": 15498
+    },
+    {
+      "epoch": 2.5111795204147764,
+      "grad_norm": 0.969572126865387,
+      "learning_rate": 3.4012979339102054e-07,
+      "loss": 0.0614,
+      "step": 15499
+    },
+    {
+      "epoch": 2.511341542449773,
+      "grad_norm": 1.0106775760650635,
+      "learning_rate": 3.399096084487108e-07,
+      "loss": 0.0603,
+      "step": 15500
+    },
+    {
+      "epoch": 2.51150356448477,
+      "grad_norm": 0.7856919765472412,
+      "learning_rate": 3.3968948959994004e-07,
+      "loss": 0.0533,
+      "step": 15501
+    },
+    {
+      "epoch": 2.511665586519767,
+      "grad_norm": 0.8515869975090027,
+      "learning_rate": 3.394694368514434e-07,
+      "loss": 0.0634,
+      "step": 15502
+    },
+    {
+      "epoch": 2.5118276085547633,
+      "grad_norm": 0.7295697927474976,
+      "learning_rate": 3.3924945020995277e-07,
+      "loss": 0.0557,
+      "step": 15503
+    },
+    {
+      "epoch": 2.51198963058976,
+      "grad_norm": 0.8382651805877686,
+      "learning_rate": 3.390295296822002e-07,
+      "loss": 0.055,
+      "step": 15504
+    },
+    {
+      "epoch": 2.512151652624757,
+      "grad_norm": 0.8761834502220154,
+      "learning_rate": 3.388096752749154e-07,
+      "loss": 0.0607,
+      "step": 15505
+    },
+    {
+      "epoch": 2.5123136746597536,
+      "grad_norm": 0.9966699481010437,
+      "learning_rate": 3.3858988699482397e-07,
+      "loss": 0.0619,
+      "step": 15506
+    },
+    {
+      "epoch": 2.5124756966947506,
+      "grad_norm": 0.8567604422569275,
+      "learning_rate": 3.3837016484865146e-07,
+      "loss": 0.0573,
+      "step": 15507
+    },
+    {
+      "epoch": 2.5126377187297475,
+      "grad_norm": 1.0470939874649048,
+      "learning_rate": 3.381505088431203e-07,
+      "loss": 0.0621,
+      "step": 15508
+    },
+    {
+      "epoch": 2.512799740764744,
+      "grad_norm": 0.8962857127189636,
+      "learning_rate": 3.3793091898495305e-07,
+      "loss": 0.0627,
+      "step": 15509
+    },
+    {
+      "epoch": 2.512961762799741,
+      "grad_norm": 1.0544072389602661,
+      "learning_rate": 3.377113952808669e-07,
+      "loss": 0.0638,
+      "step": 15510
+    },
+    {
+      "epoch": 2.5131237848347374,
+      "grad_norm": 0.7999473214149475,
+      "learning_rate": 3.374919377375796e-07,
+      "loss": 0.062,
+      "step": 15511
+    },
+    {
+      "epoch": 2.5132858068697344,
+      "grad_norm": 0.9653675556182861,
+      "learning_rate": 3.3727254636180597e-07,
+      "loss": 0.0689,
+      "step": 15512
+    },
+    {
+      "epoch": 2.513447828904731,
+      "grad_norm": 0.9286097288131714,
+      "learning_rate": 3.370532211602587e-07,
+      "loss": 0.0689,
+      "step": 15513
+    },
+    {
+      "epoch": 2.5136098509397278,
+      "grad_norm": 0.9723500609397888,
+      "learning_rate": 3.3683396213964826e-07,
+      "loss": 0.0592,
+      "step": 15514
+    },
+    {
+      "epoch": 2.5137718729747247,
+      "grad_norm": 0.8343885540962219,
+      "learning_rate": 3.3661476930668404e-07,
+      "loss": 0.0516,
+      "step": 15515
+    },
+    {
+      "epoch": 2.513933895009721,
+      "grad_norm": 0.8575674295425415,
+      "learning_rate": 3.363956426680728e-07,
+      "loss": 0.0581,
+      "step": 15516
+    },
+    {
+      "epoch": 2.514095917044718,
+      "grad_norm": 0.7917022109031677,
+      "learning_rate": 3.3617658223051935e-07,
+      "loss": 0.058,
+      "step": 15517
+    },
+    {
+      "epoch": 2.514257939079715,
+      "grad_norm": 0.8156291246414185,
+      "learning_rate": 3.3595758800072515e-07,
+      "loss": 0.0549,
+      "step": 15518
+    },
+    {
+      "epoch": 2.5144199611147116,
+      "grad_norm": 0.7907753586769104,
+      "learning_rate": 3.3573865998539236e-07,
+      "loss": 0.056,
+      "step": 15519
+    },
+    {
+      "epoch": 2.5145819831497085,
+      "grad_norm": 0.9458636045455933,
+      "learning_rate": 3.355197981912198e-07,
+      "loss": 0.0628,
+      "step": 15520
+    },
+    {
+      "epoch": 2.514744005184705,
+      "grad_norm": 0.9401422739028931,
+      "learning_rate": 3.3530100262490287e-07,
+      "loss": 0.0609,
+      "step": 15521
+    },
+    {
+      "epoch": 2.514906027219702,
+      "grad_norm": 0.8505311608314514,
+      "learning_rate": 3.350822732931361e-07,
+      "loss": 0.0601,
+      "step": 15522
+    },
+    {
+      "epoch": 2.5150680492546984,
+      "grad_norm": 0.9184473752975464,
+      "learning_rate": 3.3486361020261345e-07,
+      "loss": 0.0595,
+      "step": 15523
+    },
+    {
+      "epoch": 2.5152300712896953,
+      "grad_norm": 0.870324432849884,
+      "learning_rate": 3.3464501336002544e-07,
+      "loss": 0.0557,
+      "step": 15524
+    },
+    {
+      "epoch": 2.5153920933246923,
+      "grad_norm": 0.8141091465950012,
+      "learning_rate": 3.344264827720592e-07,
+      "loss": 0.0593,
+      "step": 15525
+    },
+    {
+      "epoch": 2.5155541153596888,
+      "grad_norm": 0.8419532775878906,
+      "learning_rate": 3.342080184454022e-07,
+      "loss": 0.0568,
+      "step": 15526
+    },
+    {
+      "epoch": 2.5157161373946857,
+      "grad_norm": 0.961286723613739,
+      "learning_rate": 3.339896203867385e-07,
+      "loss": 0.0676,
+      "step": 15527
+    },
+    {
+      "epoch": 2.5158781594296826,
+      "grad_norm": 0.8865793347358704,
+      "learning_rate": 3.337712886027511e-07,
+      "loss": 0.0601,
+      "step": 15528
+    },
+    {
+      "epoch": 2.516040181464679,
+      "grad_norm": 0.8175432682037354,
+      "learning_rate": 3.3355302310011996e-07,
+      "loss": 0.0587,
+      "step": 15529
+    },
+    {
+      "epoch": 2.516202203499676,
+      "grad_norm": 0.8112668991088867,
+      "learning_rate": 3.3333482388552356e-07,
+      "loss": 0.0549,
+      "step": 15530
+    },
+    {
+      "epoch": 2.516364225534673,
+      "grad_norm": 0.8561041951179504,
+      "learning_rate": 3.3311669096563886e-07,
+      "loss": 0.0598,
+      "step": 15531
+    },
+    {
+      "epoch": 2.5165262475696695,
+      "grad_norm": 0.9206536412239075,
+      "learning_rate": 3.3289862434713857e-07,
+      "loss": 0.0605,
+      "step": 15532
+    },
+    {
+      "epoch": 2.516688269604666,
+      "grad_norm": 1.009718418121338,
+      "learning_rate": 3.3268062403669646e-07,
+      "loss": 0.0722,
+      "step": 15533
+    },
+    {
+      "epoch": 2.516850291639663,
+      "grad_norm": 0.945552408695221,
+      "learning_rate": 3.3246269004098275e-07,
+      "loss": 0.0655,
+      "step": 15534
+    },
+    {
+      "epoch": 2.51701231367466,
+      "grad_norm": 0.9071747660636902,
+      "learning_rate": 3.322448223666658e-07,
+      "loss": 0.059,
+      "step": 15535
+    },
+    {
+      "epoch": 2.5171743357096563,
+      "grad_norm": 0.9418054223060608,
+      "learning_rate": 3.320270210204107e-07,
+      "loss": 0.0657,
+      "step": 15536
+    },
+    {
+      "epoch": 2.5173363577446533,
+      "grad_norm": 0.7941263914108276,
+      "learning_rate": 3.31809286008882e-07,
+      "loss": 0.0506,
+      "step": 15537
+    },
+    {
+      "epoch": 2.51749837977965,
+      "grad_norm": 0.8504292964935303,
+      "learning_rate": 3.3159161733874347e-07,
+      "loss": 0.0592,
+      "step": 15538
+    },
+    {
+      "epoch": 2.5176604018146467,
+      "grad_norm": 0.9112208485603333,
+      "learning_rate": 3.3137401501665334e-07,
+      "loss": 0.061,
+      "step": 15539
+    },
+    {
+      "epoch": 2.5178224238496436,
+      "grad_norm": 0.8395227789878845,
+      "learning_rate": 3.311564790492702e-07,
+      "loss": 0.0598,
+      "step": 15540
+    },
+    {
+      "epoch": 2.5179844458846405,
+      "grad_norm": 1.0928508043289185,
+      "learning_rate": 3.3093900944325046e-07,
+      "loss": 0.0652,
+      "step": 15541
+    },
+    {
+      "epoch": 2.518146467919637,
+      "grad_norm": 0.8456630110740662,
+      "learning_rate": 3.307216062052479e-07,
+      "loss": 0.0569,
+      "step": 15542
+    },
+    {
+      "epoch": 2.518308489954634,
+      "grad_norm": 0.9061066508293152,
+      "learning_rate": 3.305042693419147e-07,
+      "loss": 0.0579,
+      "step": 15543
+    },
+    {
+      "epoch": 2.5184705119896305,
+      "grad_norm": 0.9693790674209595,
+      "learning_rate": 3.3028699885990085e-07,
+      "loss": 0.0609,
+      "step": 15544
+    },
+    {
+      "epoch": 2.5186325340246274,
+      "grad_norm": 0.8443101644515991,
+      "learning_rate": 3.300697947658543e-07,
+      "loss": 0.0571,
+      "step": 15545
+    },
+    {
+      "epoch": 2.518794556059624,
+      "grad_norm": 0.828754723072052,
+      "learning_rate": 3.298526570664207e-07,
+      "loss": 0.0597,
+      "step": 15546
+    },
+    {
+      "epoch": 2.518956578094621,
+      "grad_norm": 0.961614727973938,
+      "learning_rate": 3.296355857682443e-07,
+      "loss": 0.0658,
+      "step": 15547
+    },
+    {
+      "epoch": 2.5191186001296177,
+      "grad_norm": 1.3140058517456055,
+      "learning_rate": 3.294185808779665e-07,
+      "loss": 0.0645,
+      "step": 15548
+    },
+    {
+      "epoch": 2.5192806221646142,
+      "grad_norm": 0.8643244504928589,
+      "learning_rate": 3.292016424022276e-07,
+      "loss": 0.0578,
+      "step": 15549
+    },
+    {
+      "epoch": 2.519442644199611,
+      "grad_norm": 0.9640627503395081,
+      "learning_rate": 3.289847703476659e-07,
+      "loss": 0.0605,
+      "step": 15550
+    },
+    {
+      "epoch": 2.519604666234608,
+      "grad_norm": 1.0605357885360718,
+      "learning_rate": 3.287679647209152e-07,
+      "loss": 0.0665,
+      "step": 15551
+    },
+    {
+      "epoch": 2.5197666882696046,
+      "grad_norm": 0.8427574038505554,
+      "learning_rate": 3.285512255286111e-07,
+      "loss": 0.0604,
+      "step": 15552
+    },
+    {
+      "epoch": 2.5199287103046015,
+      "grad_norm": 0.8638863563537598,
+      "learning_rate": 3.283345527773854e-07,
+      "loss": 0.0562,
+      "step": 15553
+    },
+    {
+      "epoch": 2.5200907323395985,
+      "grad_norm": 1.155843734741211,
+      "learning_rate": 3.2811794647386625e-07,
+      "loss": 0.0649,
+      "step": 15554
+    },
+    {
+      "epoch": 2.520252754374595,
+      "grad_norm": 0.9180019497871399,
+      "learning_rate": 3.279014066246822e-07,
+      "loss": 0.0602,
+      "step": 15555
+    },
+    {
+      "epoch": 2.5204147764095914,
+      "grad_norm": 0.9564993977546692,
+      "learning_rate": 3.276849332364587e-07,
+      "loss": 0.0619,
+      "step": 15556
+    },
+    {
+      "epoch": 2.5205767984445884,
+      "grad_norm": 0.8472095727920532,
+      "learning_rate": 3.2746852631581947e-07,
+      "loss": 0.0545,
+      "step": 15557
+    },
+    {
+      "epoch": 2.5207388204795853,
+      "grad_norm": 0.9872234463691711,
+      "learning_rate": 3.2725218586938584e-07,
+      "loss": 0.0583,
+      "step": 15558
+    },
+    {
+      "epoch": 2.520900842514582,
+      "grad_norm": 0.9279904961585999,
+      "learning_rate": 3.270359119037772e-07,
+      "loss": 0.0557,
+      "step": 15559
+    },
+    {
+      "epoch": 2.5210628645495787,
+      "grad_norm": 0.9269344806671143,
+      "learning_rate": 3.2681970442561134e-07,
+      "loss": 0.06,
+      "step": 15560
+    },
+    {
+      "epoch": 2.5212248865845757,
+      "grad_norm": 0.9857649803161621,
+      "learning_rate": 3.266035634415035e-07,
+      "loss": 0.0634,
+      "step": 15561
+    },
+    {
+      "epoch": 2.521386908619572,
+      "grad_norm": 0.7839047312736511,
+      "learning_rate": 3.2638748895806705e-07,
+      "loss": 0.0513,
+      "step": 15562
+    },
+    {
+      "epoch": 2.521548930654569,
+      "grad_norm": 0.9197927713394165,
+      "learning_rate": 3.261714809819136e-07,
+      "loss": 0.0632,
+      "step": 15563
+    },
+    {
+      "epoch": 2.521710952689566,
+      "grad_norm": 0.8434021472930908,
+      "learning_rate": 3.259555395196526e-07,
+      "loss": 0.0637,
+      "step": 15564
+    },
+    {
+      "epoch": 2.5218729747245625,
+      "grad_norm": 0.9010359048843384,
+      "learning_rate": 3.2573966457789014e-07,
+      "loss": 0.0669,
+      "step": 15565
+    },
+    {
+      "epoch": 2.5220349967595594,
+      "grad_norm": 1.0379987955093384,
+      "learning_rate": 3.255238561632326e-07,
+      "loss": 0.0633,
+      "step": 15566
+    },
+    {
+      "epoch": 2.522197018794556,
+      "grad_norm": 0.9836294651031494,
+      "learning_rate": 3.253081142822831e-07,
+      "loss": 0.0634,
+      "step": 15567
+    },
+    {
+      "epoch": 2.522359040829553,
+      "grad_norm": 1.0170096158981323,
+      "learning_rate": 3.250924389416432e-07,
+      "loss": 0.062,
+      "step": 15568
+    },
+    {
+      "epoch": 2.5225210628645494,
+      "grad_norm": 1.039481282234192,
+      "learning_rate": 3.2487683014791077e-07,
+      "loss": 0.0666,
+      "step": 15569
+    },
+    {
+      "epoch": 2.5226830848995463,
+      "grad_norm": 0.9465806484222412,
+      "learning_rate": 3.2466128790768327e-07,
+      "loss": 0.0606,
+      "step": 15570
+    },
+    {
+      "epoch": 2.5228451069345432,
+      "grad_norm": 1.00591242313385,
+      "learning_rate": 3.2444581222755733e-07,
+      "loss": 0.065,
+      "step": 15571
+    },
+    {
+      "epoch": 2.5230071289695397,
+      "grad_norm": 0.8654181957244873,
+      "learning_rate": 3.2423040311412384e-07,
+      "loss": 0.0602,
+      "step": 15572
+    },
+    {
+      "epoch": 2.5231691510045366,
+      "grad_norm": 0.86398845911026,
+      "learning_rate": 3.2401506057397503e-07,
+      "loss": 0.0563,
+      "step": 15573
+    },
+    {
+      "epoch": 2.5233311730395336,
+      "grad_norm": 0.8598949313163757,
+      "learning_rate": 3.2379978461369976e-07,
+      "loss": 0.0625,
+      "step": 15574
+    },
+    {
+      "epoch": 2.52349319507453,
+      "grad_norm": 0.9612644910812378,
+      "learning_rate": 3.2358457523988446e-07,
+      "loss": 0.0576,
+      "step": 15575
+    },
+    {
+      "epoch": 2.523655217109527,
+      "grad_norm": 0.9757301211357117,
+      "learning_rate": 3.233694324591144e-07,
+      "loss": 0.0639,
+      "step": 15576
+    },
+    {
+      "epoch": 2.5238172391445235,
+      "grad_norm": 1.0324627161026,
+      "learning_rate": 3.2315435627797266e-07,
+      "loss": 0.0676,
+      "step": 15577
+    },
+    {
+      "epoch": 2.5239792611795204,
+      "grad_norm": 0.9274498224258423,
+      "learning_rate": 3.229393467030395e-07,
+      "loss": 0.0603,
+      "step": 15578
+    },
+    {
+      "epoch": 2.524141283214517,
+      "grad_norm": 0.9407957196235657,
+      "learning_rate": 3.2272440374089443e-07,
+      "loss": 0.0647,
+      "step": 15579
+    },
+    {
+      "epoch": 2.524303305249514,
+      "grad_norm": 0.7908738255500793,
+      "learning_rate": 3.225095273981127e-07,
+      "loss": 0.0517,
+      "step": 15580
+    },
+    {
+      "epoch": 2.524465327284511,
+      "grad_norm": 0.8206102848052979,
+      "learning_rate": 3.2229471768127047e-07,
+      "loss": 0.0551,
+      "step": 15581
+    },
+    {
+      "epoch": 2.5246273493195073,
+      "grad_norm": 0.9348705410957336,
+      "learning_rate": 3.2207997459694053e-07,
+      "loss": 0.065,
+      "step": 15582
+    },
+    {
+      "epoch": 2.524789371354504,
+      "grad_norm": 0.9329501986503601,
+      "learning_rate": 3.218652981516923e-07,
+      "loss": 0.0654,
+      "step": 15583
+    },
+    {
+      "epoch": 2.524951393389501,
+      "grad_norm": 0.928276002407074,
+      "learning_rate": 3.2165068835209506e-07,
+      "loss": 0.0653,
+      "step": 15584
+    },
+    {
+      "epoch": 2.5251134154244976,
+      "grad_norm": 0.9858642220497131,
+      "learning_rate": 3.214361452047149e-07,
+      "loss": 0.0642,
+      "step": 15585
+    },
+    {
+      "epoch": 2.5252754374594946,
+      "grad_norm": 1.175571322441101,
+      "learning_rate": 3.2122166871611736e-07,
+      "loss": 0.0692,
+      "step": 15586
+    },
+    {
+      "epoch": 2.5254374594944915,
+      "grad_norm": 0.8295290470123291,
+      "learning_rate": 3.2100725889286393e-07,
+      "loss": 0.0581,
+      "step": 15587
+    },
+    {
+      "epoch": 2.525599481529488,
+      "grad_norm": 0.8364087343215942,
+      "learning_rate": 3.207929157415152e-07,
+      "loss": 0.0551,
+      "step": 15588
+    },
+    {
+      "epoch": 2.525761503564485,
+      "grad_norm": 0.9872838854789734,
+      "learning_rate": 3.205786392686297e-07,
+      "loss": 0.0602,
+      "step": 15589
+    },
+    {
+      "epoch": 2.5259235255994814,
+      "grad_norm": 0.8154653310775757,
+      "learning_rate": 3.2036442948076395e-07,
+      "loss": 0.0598,
+      "step": 15590
+    },
+    {
+      "epoch": 2.5260855476344783,
+      "grad_norm": 0.900958776473999,
+      "learning_rate": 3.201502863844716e-07,
+      "loss": 0.0608,
+      "step": 15591
+    },
+    {
+      "epoch": 2.526247569669475,
+      "grad_norm": 0.9498975276947021,
+      "learning_rate": 3.199362099863057e-07,
+      "loss": 0.0633,
+      "step": 15592
+    },
+    {
+      "epoch": 2.5264095917044718,
+      "grad_norm": 0.947784423828125,
+      "learning_rate": 3.1972220029281605e-07,
+      "loss": 0.0554,
+      "step": 15593
+    },
+    {
+      "epoch": 2.5265716137394687,
+      "grad_norm": 0.9299330115318298,
+      "learning_rate": 3.19508257310551e-07,
+      "loss": 0.0585,
+      "step": 15594
+    },
+    {
+      "epoch": 2.526733635774465,
+      "grad_norm": 0.9528511166572571,
+      "learning_rate": 3.1929438104605636e-07,
+      "loss": 0.06,
+      "step": 15595
+    },
+    {
+      "epoch": 2.526895657809462,
+      "grad_norm": 0.8668416142463684,
+      "learning_rate": 3.190805715058765e-07,
+      "loss": 0.0589,
+      "step": 15596
+    },
+    {
+      "epoch": 2.527057679844459,
+      "grad_norm": 0.8737120032310486,
+      "learning_rate": 3.1886682869655403e-07,
+      "loss": 0.0571,
+      "step": 15597
+    },
+    {
+      "epoch": 2.5272197018794555,
+      "grad_norm": 0.9954075813293457,
+      "learning_rate": 3.1865315262462783e-07,
+      "loss": 0.0658,
+      "step": 15598
+    },
+    {
+      "epoch": 2.5273817239144525,
+      "grad_norm": 0.8705222010612488,
+      "learning_rate": 3.1843954329663583e-07,
+      "loss": 0.0588,
+      "step": 15599
+    },
+    {
+      "epoch": 2.527543745949449,
+      "grad_norm": 0.8527966737747192,
+      "learning_rate": 3.182260007191157e-07,
+      "loss": 0.059,
+      "step": 15600
+    },
+    {
+      "epoch": 2.527705767984446,
+      "grad_norm": 1.4021953344345093,
+      "learning_rate": 3.1801252489859933e-07,
+      "loss": 0.0547,
+      "step": 15601
+    },
+    {
+      "epoch": 2.5278677900194424,
+      "grad_norm": 0.9866907596588135,
+      "learning_rate": 3.1779911584161963e-07,
+      "loss": 0.0631,
+      "step": 15602
+    },
+    {
+      "epoch": 2.5280298120544393,
+      "grad_norm": 0.8450327515602112,
+      "learning_rate": 3.175857735547061e-07,
+      "loss": 0.06,
+      "step": 15603
+    },
+    {
+      "epoch": 2.5281918340894363,
+      "grad_norm": 0.7806409001350403,
+      "learning_rate": 3.173724980443868e-07,
+      "loss": 0.0496,
+      "step": 15604
+    },
+    {
+      "epoch": 2.5283538561244328,
+      "grad_norm": 0.8109617233276367,
+      "learning_rate": 3.171592893171868e-07,
+      "loss": 0.0517,
+      "step": 15605
+    },
+    {
+      "epoch": 2.5285158781594297,
+      "grad_norm": 0.9940459132194519,
+      "learning_rate": 3.1694614737963036e-07,
+      "loss": 0.0662,
+      "step": 15606
+    },
+    {
+      "epoch": 2.5286779001944266,
+      "grad_norm": 0.9137442111968994,
+      "learning_rate": 3.167330722382389e-07,
+      "loss": 0.0588,
+      "step": 15607
+    },
+    {
+      "epoch": 2.528839922229423,
+      "grad_norm": 0.7500993013381958,
+      "learning_rate": 3.165200638995328e-07,
+      "loss": 0.0513,
+      "step": 15608
+    },
+    {
+      "epoch": 2.52900194426442,
+      "grad_norm": 0.8935854434967041,
+      "learning_rate": 3.163071223700273e-07,
+      "loss": 0.0569,
+      "step": 15609
+    },
+    {
+      "epoch": 2.529163966299417,
+      "grad_norm": 0.8556219935417175,
+      "learning_rate": 3.160942476562404e-07,
+      "loss": 0.0588,
+      "step": 15610
+    },
+    {
+      "epoch": 2.5293259883344135,
+      "grad_norm": 1.0256495475769043,
+      "learning_rate": 3.158814397646842e-07,
+      "loss": 0.0636,
+      "step": 15611
+    },
+    {
+      "epoch": 2.5294880103694104,
+      "grad_norm": 0.8018587827682495,
+      "learning_rate": 3.1566869870187115e-07,
+      "loss": 0.0522,
+      "step": 15612
+    },
+    {
+      "epoch": 2.529650032404407,
+      "grad_norm": 0.794672966003418,
+      "learning_rate": 3.1545602447430904e-07,
+      "loss": 0.0521,
+      "step": 15613
+    },
+    {
+      "epoch": 2.529812054439404,
+      "grad_norm": 0.8398905992507935,
+      "learning_rate": 3.1524341708850633e-07,
+      "loss": 0.0643,
+      "step": 15614
+    },
+    {
+      "epoch": 2.5299740764744003,
+      "grad_norm": 0.9044195413589478,
+      "learning_rate": 3.150308765509688e-07,
+      "loss": 0.0608,
+      "step": 15615
+    },
+    {
+      "epoch": 2.5301360985093972,
+      "grad_norm": 0.9135866761207581,
+      "learning_rate": 3.148184028681983e-07,
+      "loss": 0.0634,
+      "step": 15616
+    },
+    {
+      "epoch": 2.530298120544394,
+      "grad_norm": 0.8611244559288025,
+      "learning_rate": 3.1460599604669684e-07,
+      "loss": 0.0589,
+      "step": 15617
+    },
+    {
+      "epoch": 2.5304601425793907,
+      "grad_norm": 0.9294557571411133,
+      "learning_rate": 3.1439365609296253e-07,
+      "loss": 0.0573,
+      "step": 15618
+    },
+    {
+      "epoch": 2.5306221646143876,
+      "grad_norm": 0.9217930436134338,
+      "learning_rate": 3.141813830134943e-07,
+      "loss": 0.0631,
+      "step": 15619
+    },
+    {
+      "epoch": 2.5307841866493845,
+      "grad_norm": 0.8782767653465271,
+      "learning_rate": 3.1396917681478595e-07,
+      "loss": 0.0633,
+      "step": 15620
+    },
+    {
+      "epoch": 2.530946208684381,
+      "grad_norm": 0.9176865220069885,
+      "learning_rate": 3.137570375033305e-07,
+      "loss": 0.0532,
+      "step": 15621
+    },
+    {
+      "epoch": 2.531108230719378,
+      "grad_norm": 0.9333579540252686,
+      "learning_rate": 3.13544965085619e-07,
+      "loss": 0.0611,
+      "step": 15622
+    },
+    {
+      "epoch": 2.5312702527543745,
+      "grad_norm": 0.8954581618309021,
+      "learning_rate": 3.133329595681406e-07,
+      "loss": 0.0563,
+      "step": 15623
+    },
+    {
+      "epoch": 2.5314322747893714,
+      "grad_norm": 0.8141576051712036,
+      "learning_rate": 3.1312102095738205e-07,
+      "loss": 0.0513,
+      "step": 15624
+    },
+    {
+      "epoch": 2.531594296824368,
+      "grad_norm": 0.9502742886543274,
+      "learning_rate": 3.1290914925982794e-07,
+      "loss": 0.0651,
+      "step": 15625
+    },
+    {
+      "epoch": 2.531756318859365,
+      "grad_norm": 0.951263964176178,
+      "learning_rate": 3.12697344481962e-07,
+      "loss": 0.0565,
+      "step": 15626
+    },
+    {
+      "epoch": 2.5319183408943617,
+      "grad_norm": 0.9662578701972961,
+      "learning_rate": 3.124856066302634e-07,
+      "loss": 0.0623,
+      "step": 15627
+    },
+    {
+      "epoch": 2.5320803629293582,
+      "grad_norm": 1.0653865337371826,
+      "learning_rate": 3.1227393571121117e-07,
+      "loss": 0.0634,
+      "step": 15628
+    },
+    {
+      "epoch": 2.532242384964355,
+      "grad_norm": 0.9926478266716003,
+      "learning_rate": 3.1206233173128265e-07,
+      "loss": 0.0583,
+      "step": 15629
+    },
+    {
+      "epoch": 2.532404406999352,
+      "grad_norm": 0.9750377535820007,
+      "learning_rate": 3.1185079469695263e-07,
+      "loss": 0.0629,
+      "step": 15630
+    },
+    {
+      "epoch": 2.5325664290343486,
+      "grad_norm": 0.9622008800506592,
+      "learning_rate": 3.116393246146926e-07,
+      "loss": 0.0643,
+      "step": 15631
+    },
+    {
+      "epoch": 2.5327284510693455,
+      "grad_norm": 1.4896754026412964,
+      "learning_rate": 3.1142792149097297e-07,
+      "loss": 0.0577,
+      "step": 15632
+    },
+    {
+      "epoch": 2.5328904731043425,
+      "grad_norm": 1.024353265762329,
+      "learning_rate": 3.1121658533226376e-07,
+      "loss": 0.0706,
+      "step": 15633
+    },
+    {
+      "epoch": 2.533052495139339,
+      "grad_norm": 0.9080899953842163,
+      "learning_rate": 3.110053161450299e-07,
+      "loss": 0.0589,
+      "step": 15634
+    },
+    {
+      "epoch": 2.5332145171743354,
+      "grad_norm": 0.8674499988555908,
+      "learning_rate": 3.1079411393573597e-07,
+      "loss": 0.0575,
+      "step": 15635
+    },
+    {
+      "epoch": 2.5333765392093324,
+      "grad_norm": 0.9596990346908569,
+      "learning_rate": 3.105829787108444e-07,
+      "loss": 0.0649,
+      "step": 15636
+    },
+    {
+      "epoch": 2.5335385612443293,
+      "grad_norm": 0.9714479446411133,
+      "learning_rate": 3.103719104768155e-07,
+      "loss": 0.0638,
+      "step": 15637
+    },
+    {
+      "epoch": 2.533700583279326,
+      "grad_norm": 0.8182390332221985,
+      "learning_rate": 3.10160909240107e-07,
+      "loss": 0.0556,
+      "step": 15638
+    },
+    {
+      "epoch": 2.5338626053143227,
+      "grad_norm": 0.8322840332984924,
+      "learning_rate": 3.0994997500717575e-07,
+      "loss": 0.0517,
+      "step": 15639
+    },
+    {
+      "epoch": 2.5340246273493197,
+      "grad_norm": 0.9775702357292175,
+      "learning_rate": 3.0973910778447523e-07,
+      "loss": 0.0633,
+      "step": 15640
+    },
+    {
+      "epoch": 2.534186649384316,
+      "grad_norm": 0.8347146511077881,
+      "learning_rate": 3.0952830757845833e-07,
+      "loss": 0.0552,
+      "step": 15641
+    },
+    {
+      "epoch": 2.534348671419313,
+      "grad_norm": 0.7945565581321716,
+      "learning_rate": 3.0931757439557313e-07,
+      "loss": 0.0533,
+      "step": 15642
+    },
+    {
+      "epoch": 2.53451069345431,
+      "grad_norm": 0.970628023147583,
+      "learning_rate": 3.091069082422696e-07,
+      "loss": 0.0589,
+      "step": 15643
+    },
+    {
+      "epoch": 2.5346727154893065,
+      "grad_norm": 0.7898392677307129,
+      "learning_rate": 3.08896309124993e-07,
+      "loss": 0.0513,
+      "step": 15644
+    },
+    {
+      "epoch": 2.5348347375243034,
+      "grad_norm": 0.951751172542572,
+      "learning_rate": 3.086857770501867e-07,
+      "loss": 0.0617,
+      "step": 15645
+    },
+    {
+      "epoch": 2.5349967595593,
+      "grad_norm": 0.9525232911109924,
+      "learning_rate": 3.084753120242928e-07,
+      "loss": 0.0635,
+      "step": 15646
+    },
+    {
+      "epoch": 2.535158781594297,
+      "grad_norm": 0.8588886857032776,
+      "learning_rate": 3.0826491405375024e-07,
+      "loss": 0.0586,
+      "step": 15647
+    },
+    {
+      "epoch": 2.5353208036292934,
+      "grad_norm": 0.9159846901893616,
+      "learning_rate": 3.0805458314499855e-07,
+      "loss": 0.063,
+      "step": 15648
+    },
+    {
+      "epoch": 2.5354828256642903,
+      "grad_norm": 0.870168149471283,
+      "learning_rate": 3.078443193044717e-07,
+      "loss": 0.0574,
+      "step": 15649
+    },
+    {
+      "epoch": 2.535644847699287,
+      "grad_norm": 0.8437957763671875,
+      "learning_rate": 3.076341225386037e-07,
+      "loss": 0.0585,
+      "step": 15650
+    },
+    {
+      "epoch": 2.5358068697342837,
+      "grad_norm": 0.9445669054985046,
+      "learning_rate": 3.074239928538261e-07,
+      "loss": 0.0694,
+      "step": 15651
+    },
+    {
+      "epoch": 2.5359688917692806,
+      "grad_norm": 0.81999272108078,
+      "learning_rate": 3.0721393025656853e-07,
+      "loss": 0.0524,
+      "step": 15652
+    },
+    {
+      "epoch": 2.5361309138042776,
+      "grad_norm": 1.2289856672286987,
+      "learning_rate": 3.070039347532583e-07,
+      "loss": 0.0617,
+      "step": 15653
+    },
+    {
+      "epoch": 2.536292935839274,
+      "grad_norm": 1.0222786664962769,
+      "learning_rate": 3.0679400635032053e-07,
+      "loss": 0.0564,
+      "step": 15654
+    },
+    {
+      "epoch": 2.536454957874271,
+      "grad_norm": 0.8596547842025757,
+      "learning_rate": 3.065841450541787e-07,
+      "loss": 0.0636,
+      "step": 15655
+    },
+    {
+      "epoch": 2.536616979909268,
+      "grad_norm": 1.1208213567733765,
+      "learning_rate": 3.063743508712544e-07,
+      "loss": 0.0638,
+      "step": 15656
+    },
+    {
+      "epoch": 2.5367790019442644,
+      "grad_norm": 0.8171042799949646,
+      "learning_rate": 3.06164623807966e-07,
+      "loss": 0.058,
+      "step": 15657
+    },
+    {
+      "epoch": 2.536941023979261,
+      "grad_norm": 0.8748427033424377,
+      "learning_rate": 3.059549638707315e-07,
+      "loss": 0.0573,
+      "step": 15658
+    },
+    {
+      "epoch": 2.537103046014258,
+      "grad_norm": 1.00801420211792,
+      "learning_rate": 3.0574537106596606e-07,
+      "loss": 0.0647,
+      "step": 15659
+    },
+    {
+      "epoch": 2.537265068049255,
+      "grad_norm": 0.9508548974990845,
+      "learning_rate": 3.0553584540008176e-07,
+      "loss": 0.0621,
+      "step": 15660
+    },
+    {
+      "epoch": 2.5374270900842513,
+      "grad_norm": 0.862837553024292,
+      "learning_rate": 3.053263868794895e-07,
+      "loss": 0.0647,
+      "step": 15661
+    },
+    {
+      "epoch": 2.537589112119248,
+      "grad_norm": 0.9013253450393677,
+      "learning_rate": 3.0511699551059927e-07,
+      "loss": 0.0671,
+      "step": 15662
+    },
+    {
+      "epoch": 2.537751134154245,
+      "grad_norm": 1.0130845308303833,
+      "learning_rate": 3.049076712998181e-07,
+      "loss": 0.0585,
+      "step": 15663
+    },
+    {
+      "epoch": 2.5379131561892416,
+      "grad_norm": 1.0517815351486206,
+      "learning_rate": 3.0469841425354945e-07,
+      "loss": 0.0628,
+      "step": 15664
+    },
+    {
+      "epoch": 2.5380751782242386,
+      "grad_norm": 0.8955190777778625,
+      "learning_rate": 3.044892243781969e-07,
+      "loss": 0.0606,
+      "step": 15665
+    },
+    {
+      "epoch": 2.5382372002592355,
+      "grad_norm": 0.9051300287246704,
+      "learning_rate": 3.0428010168016107e-07,
+      "loss": 0.0605,
+      "step": 15666
+    },
+    {
+      "epoch": 2.538399222294232,
+      "grad_norm": 0.8366383910179138,
+      "learning_rate": 3.040710461658408e-07,
+      "loss": 0.0598,
+      "step": 15667
+    },
+    {
+      "epoch": 2.538561244329229,
+      "grad_norm": 0.8393315672874451,
+      "learning_rate": 3.0386205784163207e-07,
+      "loss": 0.0633,
+      "step": 15668
+    },
+    {
+      "epoch": 2.5387232663642254,
+      "grad_norm": 1.1416642665863037,
+      "learning_rate": 3.0365313671393e-07,
+      "loss": 0.0619,
+      "step": 15669
+    },
+    {
+      "epoch": 2.5388852883992223,
+      "grad_norm": 0.8364261984825134,
+      "learning_rate": 3.0344428278912765e-07,
+      "loss": 0.0606,
+      "step": 15670
+    },
+    {
+      "epoch": 2.539047310434219,
+      "grad_norm": 0.9973102807998657,
+      "learning_rate": 3.0323549607361333e-07,
+      "loss": 0.0586,
+      "step": 15671
+    },
+    {
+      "epoch": 2.5392093324692158,
+      "grad_norm": 0.8072327971458435,
+      "learning_rate": 3.030267765737774e-07,
+      "loss": 0.0571,
+      "step": 15672
+    },
+    {
+      "epoch": 2.5393713545042127,
+      "grad_norm": 1.0454763174057007,
+      "learning_rate": 3.0281812429600544e-07,
+      "loss": 0.0569,
+      "step": 15673
+    },
+    {
+      "epoch": 2.539533376539209,
+      "grad_norm": 1.0612105131149292,
+      "learning_rate": 3.026095392466824e-07,
+      "loss": 0.0655,
+      "step": 15674
+    },
+    {
+      "epoch": 2.539695398574206,
+      "grad_norm": 1.0224926471710205,
+      "learning_rate": 3.024010214321893e-07,
+      "loss": 0.0707,
+      "step": 15675
+    },
+    {
+      "epoch": 2.539857420609203,
+      "grad_norm": 0.8485469222068787,
+      "learning_rate": 3.021925708589066e-07,
+      "loss": 0.0567,
+      "step": 15676
+    },
+    {
+      "epoch": 2.5400194426441995,
+      "grad_norm": 1.0038886070251465,
+      "learning_rate": 3.019841875332133e-07,
+      "loss": 0.0614,
+      "step": 15677
+    },
+    {
+      "epoch": 2.5401814646791965,
+      "grad_norm": 0.8578863739967346,
+      "learning_rate": 3.0177587146148435e-07,
+      "loss": 0.061,
+      "step": 15678
+    },
+    {
+      "epoch": 2.540343486714193,
+      "grad_norm": 0.7758865356445312,
+      "learning_rate": 3.0156762265009437e-07,
+      "loss": 0.0567,
+      "step": 15679
+    },
+    {
+      "epoch": 2.54050550874919,
+      "grad_norm": 1.1475505828857422,
+      "learning_rate": 3.013594411054144e-07,
+      "loss": 0.0682,
+      "step": 15680
+    },
+    {
+      "epoch": 2.5406675307841864,
+      "grad_norm": 0.9241853952407837,
+      "learning_rate": 3.0115132683381565e-07,
+      "loss": 0.0679,
+      "step": 15681
+    },
+    {
+      "epoch": 2.5408295528191833,
+      "grad_norm": 0.9604292511940002,
+      "learning_rate": 3.0094327984166506e-07,
+      "loss": 0.0564,
+      "step": 15682
+    },
+    {
+      "epoch": 2.5409915748541803,
+      "grad_norm": 0.8330355882644653,
+      "learning_rate": 3.007353001353283e-07,
+      "loss": 0.0574,
+      "step": 15683
+    },
+    {
+      "epoch": 2.5411535968891767,
+      "grad_norm": 0.9024338126182556,
+      "learning_rate": 3.0052738772116925e-07,
+      "loss": 0.069,
+      "step": 15684
+    },
+    {
+      "epoch": 2.5413156189241737,
+      "grad_norm": 0.9222663640975952,
+      "learning_rate": 3.003195426055497e-07,
+      "loss": 0.0632,
+      "step": 15685
+    },
+    {
+      "epoch": 2.5414776409591706,
+      "grad_norm": 0.9716166257858276,
+      "learning_rate": 3.001117647948287e-07,
+      "loss": 0.0651,
+      "step": 15686
+    },
+    {
+      "epoch": 2.541639662994167,
+      "grad_norm": 0.904063880443573,
+      "learning_rate": 2.9990405429536433e-07,
+      "loss": 0.0624,
+      "step": 15687
+    },
+    {
+      "epoch": 2.541801685029164,
+      "grad_norm": 0.8916031122207642,
+      "learning_rate": 2.996964111135123e-07,
+      "loss": 0.0565,
+      "step": 15688
+    },
+    {
+      "epoch": 2.541963707064161,
+      "grad_norm": 0.8588840961456299,
+      "learning_rate": 2.9948883525562464e-07,
+      "loss": 0.0574,
+      "step": 15689
+    },
+    {
+      "epoch": 2.5421257290991575,
+      "grad_norm": 0.8580512404441833,
+      "learning_rate": 2.992813267280531e-07,
+      "loss": 0.0594,
+      "step": 15690
+    },
+    {
+      "epoch": 2.5422877511341544,
+      "grad_norm": 0.8236280083656311,
+      "learning_rate": 2.9907388553714806e-07,
+      "loss": 0.0582,
+      "step": 15691
+    },
+    {
+      "epoch": 2.542449773169151,
+      "grad_norm": 0.8864600658416748,
+      "learning_rate": 2.988665116892564e-07,
+      "loss": 0.0582,
+      "step": 15692
+    },
+    {
+      "epoch": 2.542611795204148,
+      "grad_norm": 0.934640109539032,
+      "learning_rate": 2.986592051907222e-07,
+      "loss": 0.066,
+      "step": 15693
+    },
+    {
+      "epoch": 2.5427738172391443,
+      "grad_norm": 0.8519915342330933,
+      "learning_rate": 2.9845196604788935e-07,
+      "loss": 0.0627,
+      "step": 15694
+    },
+    {
+      "epoch": 2.5429358392741412,
+      "grad_norm": 0.9214040637016296,
+      "learning_rate": 2.9824479426709853e-07,
+      "loss": 0.0572,
+      "step": 15695
+    },
+    {
+      "epoch": 2.543097861309138,
+      "grad_norm": 1.0581270456314087,
+      "learning_rate": 2.980376898546888e-07,
+      "loss": 0.0634,
+      "step": 15696
+    },
+    {
+      "epoch": 2.5432598833441347,
+      "grad_norm": 1.086905598640442,
+      "learning_rate": 2.9783065281699747e-07,
+      "loss": 0.0629,
+      "step": 15697
+    },
+    {
+      "epoch": 2.5434219053791316,
+      "grad_norm": 1.031447172164917,
+      "learning_rate": 2.976236831603588e-07,
+      "loss": 0.0629,
+      "step": 15698
+    },
+    {
+      "epoch": 2.5435839274141285,
+      "grad_norm": 0.9910357594490051,
+      "learning_rate": 2.9741678089110576e-07,
+      "loss": 0.0699,
+      "step": 15699
+    },
+    {
+      "epoch": 2.543745949449125,
+      "grad_norm": 0.838286817073822,
+      "learning_rate": 2.972099460155689e-07,
+      "loss": 0.0555,
+      "step": 15700
+    },
+    {
+      "epoch": 2.543907971484122,
+      "grad_norm": 0.939771294593811,
+      "learning_rate": 2.970031785400773e-07,
+      "loss": 0.0593,
+      "step": 15701
+    },
+    {
+      "epoch": 2.5440699935191184,
+      "grad_norm": 1.2176207304000854,
+      "learning_rate": 2.9679647847095735e-07,
+      "loss": 0.0637,
+      "step": 15702
+    },
+    {
+      "epoch": 2.5442320155541154,
+      "grad_norm": 0.8864500522613525,
+      "learning_rate": 2.965898458145339e-07,
+      "loss": 0.0614,
+      "step": 15703
+    },
+    {
+      "epoch": 2.544394037589112,
+      "grad_norm": 1.0630028247833252,
+      "learning_rate": 2.9638328057712775e-07,
+      "loss": 0.0688,
+      "step": 15704
+    },
+    {
+      "epoch": 2.544556059624109,
+      "grad_norm": 0.9076703786849976,
+      "learning_rate": 2.9617678276506136e-07,
+      "loss": 0.0578,
+      "step": 15705
+    },
+    {
+      "epoch": 2.5447180816591057,
+      "grad_norm": 0.9207355976104736,
+      "learning_rate": 2.9597035238465214e-07,
+      "loss": 0.0546,
+      "step": 15706
+    },
+    {
+      "epoch": 2.5448801036941022,
+      "grad_norm": 0.9597820043563843,
+      "learning_rate": 2.9576398944221707e-07,
+      "loss": 0.0627,
+      "step": 15707
+    },
+    {
+      "epoch": 2.545042125729099,
+      "grad_norm": 0.8808280229568481,
+      "learning_rate": 2.9555769394406934e-07,
+      "loss": 0.0615,
+      "step": 15708
+    },
+    {
+      "epoch": 2.545204147764096,
+      "grad_norm": 0.9642734527587891,
+      "learning_rate": 2.9535146589652093e-07,
+      "loss": 0.0668,
+      "step": 15709
+    },
+    {
+      "epoch": 2.5453661697990926,
+      "grad_norm": 0.897223711013794,
+      "learning_rate": 2.9514530530588367e-07,
+      "loss": 0.0609,
+      "step": 15710
+    },
+    {
+      "epoch": 2.5455281918340895,
+      "grad_norm": 0.9457272887229919,
+      "learning_rate": 2.949392121784636e-07,
+      "loss": 0.0596,
+      "step": 15711
+    },
+    {
+      "epoch": 2.5456902138690864,
+      "grad_norm": 0.9758214950561523,
+      "learning_rate": 2.947331865205677e-07,
+      "loss": 0.0564,
+      "step": 15712
+    },
+    {
+      "epoch": 2.545852235904083,
+      "grad_norm": 1.0195071697235107,
+      "learning_rate": 2.9452722833849976e-07,
+      "loss": 0.0663,
+      "step": 15713
+    },
+    {
+      "epoch": 2.54601425793908,
+      "grad_norm": 0.8964647054672241,
+      "learning_rate": 2.943213376385612e-07,
+      "loss": 0.0587,
+      "step": 15714
+    },
+    {
+      "epoch": 2.5461762799740764,
+      "grad_norm": 1.0028923749923706,
+      "learning_rate": 2.9411551442705243e-07,
+      "loss": 0.0637,
+      "step": 15715
+    },
+    {
+      "epoch": 2.5463383020090733,
+      "grad_norm": 0.8939936757087708,
+      "learning_rate": 2.9390975871027046e-07,
+      "loss": 0.0589,
+      "step": 15716
+    },
+    {
+      "epoch": 2.54650032404407,
+      "grad_norm": 0.9271273016929626,
+      "learning_rate": 2.9370407049451126e-07,
+      "loss": 0.0572,
+      "step": 15717
+    },
+    {
+      "epoch": 2.5466623460790667,
+      "grad_norm": 0.9877331256866455,
+      "learning_rate": 2.934984497860691e-07,
+      "loss": 0.0683,
+      "step": 15718
+    },
+    {
+      "epoch": 2.5468243681140637,
+      "grad_norm": 0.8610256910324097,
+      "learning_rate": 2.932928965912335e-07,
+      "loss": 0.0636,
+      "step": 15719
+    },
+    {
+      "epoch": 2.54698639014906,
+      "grad_norm": 0.9936156868934631,
+      "learning_rate": 2.9308741091629596e-07,
+      "loss": 0.059,
+      "step": 15720
+    },
+    {
+      "epoch": 2.547148412184057,
+      "grad_norm": 0.9047034382820129,
+      "learning_rate": 2.928819927675433e-07,
+      "loss": 0.064,
+      "step": 15721
+    },
+    {
+      "epoch": 2.547310434219054,
+      "grad_norm": 0.9128992557525635,
+      "learning_rate": 2.9267664215126e-07,
+      "loss": 0.0647,
+      "step": 15722
+    },
+    {
+      "epoch": 2.5474724562540505,
+      "grad_norm": 0.9131163358688354,
+      "learning_rate": 2.924713590737294e-07,
+      "loss": 0.0631,
+      "step": 15723
+    },
+    {
+      "epoch": 2.5476344782890474,
+      "grad_norm": 0.8168431520462036,
+      "learning_rate": 2.9226614354123356e-07,
+      "loss": 0.0582,
+      "step": 15724
+    },
+    {
+      "epoch": 2.547796500324044,
+      "grad_norm": 0.9571110606193542,
+      "learning_rate": 2.9206099556005145e-07,
+      "loss": 0.0636,
+      "step": 15725
+    },
+    {
+      "epoch": 2.547958522359041,
+      "grad_norm": 0.9737008213996887,
+      "learning_rate": 2.9185591513645947e-07,
+      "loss": 0.0674,
+      "step": 15726
+    },
+    {
+      "epoch": 2.5481205443940373,
+      "grad_norm": 0.8390480875968933,
+      "learning_rate": 2.9165090227673306e-07,
+      "loss": 0.0563,
+      "step": 15727
+    },
+    {
+      "epoch": 2.5482825664290343,
+      "grad_norm": 1.0809296369552612,
+      "learning_rate": 2.914459569871447e-07,
+      "loss": 0.0703,
+      "step": 15728
+    },
+    {
+      "epoch": 2.548444588464031,
+      "grad_norm": 0.7961978912353516,
+      "learning_rate": 2.9124107927396534e-07,
+      "loss": 0.0548,
+      "step": 15729
+    },
+    {
+      "epoch": 2.5486066104990277,
+      "grad_norm": 0.9523206949234009,
+      "learning_rate": 2.91036269143464e-07,
+      "loss": 0.0619,
+      "step": 15730
+    },
+    {
+      "epoch": 2.5487686325340246,
+      "grad_norm": 0.756062924861908,
+      "learning_rate": 2.908315266019074e-07,
+      "loss": 0.0545,
+      "step": 15731
+    },
+    {
+      "epoch": 2.5489306545690216,
+      "grad_norm": 0.991388201713562,
+      "learning_rate": 2.9062685165555963e-07,
+      "loss": 0.0585,
+      "step": 15732
+    },
+    {
+      "epoch": 2.549092676604018,
+      "grad_norm": 0.9171768426895142,
+      "learning_rate": 2.904222443106838e-07,
+      "loss": 0.0607,
+      "step": 15733
+    },
+    {
+      "epoch": 2.549254698639015,
+      "grad_norm": 1.0466676950454712,
+      "learning_rate": 2.9021770457354046e-07,
+      "loss": 0.0611,
+      "step": 15734
+    },
+    {
+      "epoch": 2.549416720674012,
+      "grad_norm": 1.0103275775909424,
+      "learning_rate": 2.9001323245038765e-07,
+      "loss": 0.0688,
+      "step": 15735
+    },
+    {
+      "epoch": 2.5495787427090084,
+      "grad_norm": 0.8275580406188965,
+      "learning_rate": 2.8980882794748227e-07,
+      "loss": 0.0535,
+      "step": 15736
+    },
+    {
+      "epoch": 2.5497407647440054,
+      "grad_norm": 0.885045051574707,
+      "learning_rate": 2.896044910710777e-07,
+      "loss": 0.0677,
+      "step": 15737
+    },
+    {
+      "epoch": 2.549902786779002,
+      "grad_norm": 0.9406101703643799,
+      "learning_rate": 2.894002218274261e-07,
+      "loss": 0.0581,
+      "step": 15738
+    },
+    {
+      "epoch": 2.5500648088139988,
+      "grad_norm": 0.9260234236717224,
+      "learning_rate": 2.8919602022277923e-07,
+      "loss": 0.0661,
+      "step": 15739
+    },
+    {
+      "epoch": 2.5502268308489953,
+      "grad_norm": 1.0107648372650146,
+      "learning_rate": 2.8899188626338363e-07,
+      "loss": 0.0613,
+      "step": 15740
+    },
+    {
+      "epoch": 2.550388852883992,
+      "grad_norm": 1.016796350479126,
+      "learning_rate": 2.8878781995548584e-07,
+      "loss": 0.0708,
+      "step": 15741
+    },
+    {
+      "epoch": 2.550550874918989,
+      "grad_norm": 0.8699153661727905,
+      "learning_rate": 2.8858382130532965e-07,
+      "loss": 0.0604,
+      "step": 15742
+    },
+    {
+      "epoch": 2.5507128969539856,
+      "grad_norm": 0.9409704804420471,
+      "learning_rate": 2.8837989031915674e-07,
+      "loss": 0.0604,
+      "step": 15743
+    },
+    {
+      "epoch": 2.5508749189889826,
+      "grad_norm": 0.884575605392456,
+      "learning_rate": 2.8817602700320747e-07,
+      "loss": 0.0657,
+      "step": 15744
+    },
+    {
+      "epoch": 2.5510369410239795,
+      "grad_norm": 0.9905060529708862,
+      "learning_rate": 2.879722313637193e-07,
+      "loss": 0.0693,
+      "step": 15745
+    },
+    {
+      "epoch": 2.551198963058976,
+      "grad_norm": 0.9149320125579834,
+      "learning_rate": 2.8776850340692777e-07,
+      "loss": 0.0633,
+      "step": 15746
+    },
+    {
+      "epoch": 2.551360985093973,
+      "grad_norm": 1.0935709476470947,
+      "learning_rate": 2.875648431390665e-07,
+      "loss": 0.0625,
+      "step": 15747
+    },
+    {
+      "epoch": 2.5515230071289694,
+      "grad_norm": 0.9420832991600037,
+      "learning_rate": 2.87361250566367e-07,
+      "loss": 0.0557,
+      "step": 15748
+    },
+    {
+      "epoch": 2.5516850291639663,
+      "grad_norm": 1.622456431388855,
+      "learning_rate": 2.871577256950589e-07,
+      "loss": 0.0592,
+      "step": 15749
+    },
+    {
+      "epoch": 2.551847051198963,
+      "grad_norm": 0.8848714232444763,
+      "learning_rate": 2.869542685313692e-07,
+      "loss": 0.0572,
+      "step": 15750
+    },
+    {
+      "epoch": 2.5520090732339598,
+      "grad_norm": 1.0072789192199707,
+      "learning_rate": 2.8675087908152407e-07,
+      "loss": 0.0661,
+      "step": 15751
+    },
+    {
+      "epoch": 2.5521710952689567,
+      "grad_norm": 0.8592365384101868,
+      "learning_rate": 2.865475573517451e-07,
+      "loss": 0.0597,
+      "step": 15752
+    },
+    {
+      "epoch": 2.552333117303953,
+      "grad_norm": 0.9718042612075806,
+      "learning_rate": 2.863443033482549e-07,
+      "loss": 0.0634,
+      "step": 15753
+    },
+    {
+      "epoch": 2.55249513933895,
+      "grad_norm": 1.0915923118591309,
+      "learning_rate": 2.8614111707727267e-07,
+      "loss": 0.0533,
+      "step": 15754
+    },
+    {
+      "epoch": 2.552657161373947,
+      "grad_norm": 0.9023035764694214,
+      "learning_rate": 2.859379985450142e-07,
+      "loss": 0.0615,
+      "step": 15755
+    },
+    {
+      "epoch": 2.5528191834089435,
+      "grad_norm": 0.9342001676559448,
+      "learning_rate": 2.8573494775769485e-07,
+      "loss": 0.0639,
+      "step": 15756
+    },
+    {
+      "epoch": 2.5529812054439405,
+      "grad_norm": 1.2013731002807617,
+      "learning_rate": 2.8553196472152794e-07,
+      "loss": 0.0641,
+      "step": 15757
+    },
+    {
+      "epoch": 2.5531432274789374,
+      "grad_norm": 0.8660479784011841,
+      "learning_rate": 2.853290494427238e-07,
+      "loss": 0.0601,
+      "step": 15758
+    },
+    {
+      "epoch": 2.553305249513934,
+      "grad_norm": 0.8665417432785034,
+      "learning_rate": 2.851262019274917e-07,
+      "loss": 0.066,
+      "step": 15759
+    },
+    {
+      "epoch": 2.5534672715489304,
+      "grad_norm": 0.9686926007270813,
+      "learning_rate": 2.8492342218203766e-07,
+      "loss": 0.061,
+      "step": 15760
+    },
+    {
+      "epoch": 2.5536292935839273,
+      "grad_norm": 0.931259274482727,
+      "learning_rate": 2.847207102125665e-07,
+      "loss": 0.0667,
+      "step": 15761
+    },
+    {
+      "epoch": 2.5537913156189243,
+      "grad_norm": 0.8071960210800171,
+      "learning_rate": 2.845180660252808e-07,
+      "loss": 0.0659,
+      "step": 15762
+    },
+    {
+      "epoch": 2.5539533376539207,
+      "grad_norm": 0.8331485986709595,
+      "learning_rate": 2.843154896263811e-07,
+      "loss": 0.0538,
+      "step": 15763
+    },
+    {
+      "epoch": 2.5541153596889177,
+      "grad_norm": 0.8786145448684692,
+      "learning_rate": 2.8411298102206524e-07,
+      "loss": 0.0623,
+      "step": 15764
+    },
+    {
+      "epoch": 2.5542773817239146,
+      "grad_norm": 0.9104515314102173,
+      "learning_rate": 2.839105402185305e-07,
+      "loss": 0.065,
+      "step": 15765
+    },
+    {
+      "epoch": 2.554439403758911,
+      "grad_norm": 0.935090959072113,
+      "learning_rate": 2.837081672219694e-07,
+      "loss": 0.0581,
+      "step": 15766
+    },
+    {
+      "epoch": 2.554601425793908,
+      "grad_norm": 0.8648780584335327,
+      "learning_rate": 2.835058620385756e-07,
+      "loss": 0.061,
+      "step": 15767
+    },
+    {
+      "epoch": 2.554763447828905,
+      "grad_norm": 0.8187403678894043,
+      "learning_rate": 2.833036246745385e-07,
+      "loss": 0.0566,
+      "step": 15768
+    },
+    {
+      "epoch": 2.5549254698639015,
+      "grad_norm": 0.8875253796577454,
+      "learning_rate": 2.831014551360467e-07,
+      "loss": 0.0608,
+      "step": 15769
+    },
+    {
+      "epoch": 2.5550874918988984,
+      "grad_norm": 0.8889281153678894,
+      "learning_rate": 2.828993534292851e-07,
+      "loss": 0.0608,
+      "step": 15770
+    },
+    {
+      "epoch": 2.555249513933895,
+      "grad_norm": 0.9261566400527954,
+      "learning_rate": 2.8269731956043736e-07,
+      "loss": 0.0626,
+      "step": 15771
+    },
+    {
+      "epoch": 2.555411535968892,
+      "grad_norm": 0.8873212337493896,
+      "learning_rate": 2.824953535356872e-07,
+      "loss": 0.0524,
+      "step": 15772
+    },
+    {
+      "epoch": 2.5555735580038883,
+      "grad_norm": 0.919251561164856,
+      "learning_rate": 2.8229345536121223e-07,
+      "loss": 0.0607,
+      "step": 15773
+    },
+    {
+      "epoch": 2.5557355800388852,
+      "grad_norm": 0.8338304162025452,
+      "learning_rate": 2.820916250431907e-07,
+      "loss": 0.0569,
+      "step": 15774
+    },
+    {
+      "epoch": 2.555897602073882,
+      "grad_norm": 0.7551842331886292,
+      "learning_rate": 2.818898625877983e-07,
+      "loss": 0.0512,
+      "step": 15775
+    },
+    {
+      "epoch": 2.5560596241088787,
+      "grad_norm": 0.929349422454834,
+      "learning_rate": 2.8168816800120845e-07,
+      "loss": 0.0586,
+      "step": 15776
+    },
+    {
+      "epoch": 2.5562216461438756,
+      "grad_norm": 0.8814112544059753,
+      "learning_rate": 2.814865412895926e-07,
+      "loss": 0.0554,
+      "step": 15777
+    },
+    {
+      "epoch": 2.5563836681788725,
+      "grad_norm": 0.9311420917510986,
+      "learning_rate": 2.812849824591196e-07,
+      "loss": 0.0615,
+      "step": 15778
+    },
+    {
+      "epoch": 2.556545690213869,
+      "grad_norm": 0.8256794810295105,
+      "learning_rate": 2.8108349151595713e-07,
+      "loss": 0.0577,
+      "step": 15779
+    },
+    {
+      "epoch": 2.556707712248866,
+      "grad_norm": 0.8068723082542419,
+      "learning_rate": 2.808820684662705e-07,
+      "loss": 0.0626,
+      "step": 15780
+    },
+    {
+      "epoch": 2.556869734283863,
+      "grad_norm": 0.75613933801651,
+      "learning_rate": 2.806807133162215e-07,
+      "loss": 0.0509,
+      "step": 15781
+    },
+    {
+      "epoch": 2.5570317563188594,
+      "grad_norm": 1.0102713108062744,
+      "learning_rate": 2.804794260719726e-07,
+      "loss": 0.0697,
+      "step": 15782
+    },
+    {
+      "epoch": 2.557193778353856,
+      "grad_norm": 0.8469998836517334,
+      "learning_rate": 2.8027820673968256e-07,
+      "loss": 0.0618,
+      "step": 15783
+    },
+    {
+      "epoch": 2.557355800388853,
+      "grad_norm": 1.0581094026565552,
+      "learning_rate": 2.800770553255072e-07,
+      "loss": 0.0644,
+      "step": 15784
+    },
+    {
+      "epoch": 2.5575178224238497,
+      "grad_norm": 0.8770332932472229,
+      "learning_rate": 2.7987597183560175e-07,
+      "loss": 0.0615,
+      "step": 15785
+    },
+    {
+      "epoch": 2.557679844458846,
+      "grad_norm": 0.9421771764755249,
+      "learning_rate": 2.796749562761186e-07,
+      "loss": 0.0575,
+      "step": 15786
+    },
+    {
+      "epoch": 2.557841866493843,
+      "grad_norm": 1.0569273233413696,
+      "learning_rate": 2.7947400865320966e-07,
+      "loss": 0.0666,
+      "step": 15787
+    },
+    {
+      "epoch": 2.55800388852884,
+      "grad_norm": 1.0118812322616577,
+      "learning_rate": 2.7927312897302217e-07,
+      "loss": 0.0615,
+      "step": 15788
+    },
+    {
+      "epoch": 2.5581659105638366,
+      "grad_norm": 1.168845772743225,
+      "learning_rate": 2.790723172417026e-07,
+      "loss": 0.065,
+      "step": 15789
+    },
+    {
+      "epoch": 2.5583279325988335,
+      "grad_norm": 0.9244213700294495,
+      "learning_rate": 2.7887157346539574e-07,
+      "loss": 0.0632,
+      "step": 15790
+    },
+    {
+      "epoch": 2.5584899546338304,
+      "grad_norm": 0.9718979001045227,
+      "learning_rate": 2.7867089765024365e-07,
+      "loss": 0.0678,
+      "step": 15791
+    },
+    {
+      "epoch": 2.558651976668827,
+      "grad_norm": 1.047033667564392,
+      "learning_rate": 2.7847028980238666e-07,
+      "loss": 0.0602,
+      "step": 15792
+    },
+    {
+      "epoch": 2.558813998703824,
+      "grad_norm": 1.0640023946762085,
+      "learning_rate": 2.7826974992796266e-07,
+      "loss": 0.0594,
+      "step": 15793
+    },
+    {
+      "epoch": 2.5589760207388204,
+      "grad_norm": 0.8724460005760193,
+      "learning_rate": 2.780692780331079e-07,
+      "loss": 0.059,
+      "step": 15794
+    },
+    {
+      "epoch": 2.5591380427738173,
+      "grad_norm": 0.7924855947494507,
+      "learning_rate": 2.778688741239563e-07,
+      "loss": 0.0529,
+      "step": 15795
+    },
+    {
+      "epoch": 2.559300064808814,
+      "grad_norm": 0.8275933861732483,
+      "learning_rate": 2.7766853820663963e-07,
+      "loss": 0.0577,
+      "step": 15796
+    },
+    {
+      "epoch": 2.5594620868438107,
+      "grad_norm": 1.010675072669983,
+      "learning_rate": 2.774682702872877e-07,
+      "loss": 0.0616,
+      "step": 15797
+    },
+    {
+      "epoch": 2.5596241088788076,
+      "grad_norm": 0.9357376098632812,
+      "learning_rate": 2.7726807037202903e-07,
+      "loss": 0.06,
+      "step": 15798
+    },
+    {
+      "epoch": 2.559786130913804,
+      "grad_norm": 0.8635546565055847,
+      "learning_rate": 2.7706793846698776e-07,
+      "loss": 0.0589,
+      "step": 15799
+    },
+    {
+      "epoch": 2.559948152948801,
+      "grad_norm": 0.8578700423240662,
+      "learning_rate": 2.7686787457828796e-07,
+      "loss": 0.0534,
+      "step": 15800
+    },
+    {
+      "epoch": 2.560110174983798,
+      "grad_norm": 1.073947787284851,
+      "learning_rate": 2.7666787871205135e-07,
+      "loss": 0.063,
+      "step": 15801
+    },
+    {
+      "epoch": 2.5602721970187945,
+      "grad_norm": 1.1108258962631226,
+      "learning_rate": 2.764679508743981e-07,
+      "loss": 0.0571,
+      "step": 15802
+    },
+    {
+      "epoch": 2.5604342190537914,
+      "grad_norm": 0.8796466588973999,
+      "learning_rate": 2.7626809107144435e-07,
+      "loss": 0.0547,
+      "step": 15803
+    },
+    {
+      "epoch": 2.560596241088788,
+      "grad_norm": 1.0041248798370361,
+      "learning_rate": 2.7606829930930555e-07,
+      "loss": 0.0646,
+      "step": 15804
+    },
+    {
+      "epoch": 2.560758263123785,
+      "grad_norm": 0.9209561944007874,
+      "learning_rate": 2.7586857559409484e-07,
+      "loss": 0.0549,
+      "step": 15805
+    },
+    {
+      "epoch": 2.5609202851587813,
+      "grad_norm": 0.8668680787086487,
+      "learning_rate": 2.7566891993192347e-07,
+      "loss": 0.0578,
+      "step": 15806
+    },
+    {
+      "epoch": 2.5610823071937783,
+      "grad_norm": 0.8909144401550293,
+      "learning_rate": 2.754693323289004e-07,
+      "loss": 0.0564,
+      "step": 15807
+    },
+    {
+      "epoch": 2.561244329228775,
+      "grad_norm": 0.8772615790367126,
+      "learning_rate": 2.752698127911327e-07,
+      "loss": 0.0613,
+      "step": 15808
+    },
+    {
+      "epoch": 2.5614063512637717,
+      "grad_norm": 0.8881198167800903,
+      "learning_rate": 2.750703613247252e-07,
+      "loss": 0.052,
+      "step": 15809
+    },
+    {
+      "epoch": 2.5615683732987686,
+      "grad_norm": 0.8577731847763062,
+      "learning_rate": 2.748709779357794e-07,
+      "loss": 0.0573,
+      "step": 15810
+    },
+    {
+      "epoch": 2.5617303953337656,
+      "grad_norm": 0.8215259313583374,
+      "learning_rate": 2.7467166263039776e-07,
+      "loss": 0.0586,
+      "step": 15811
+    },
+    {
+      "epoch": 2.561892417368762,
+      "grad_norm": 0.8955356478691101,
+      "learning_rate": 2.744724154146777e-07,
+      "loss": 0.0568,
+      "step": 15812
+    },
+    {
+      "epoch": 2.562054439403759,
+      "grad_norm": 0.9561681747436523,
+      "learning_rate": 2.7427323629471677e-07,
+      "loss": 0.0685,
+      "step": 15813
+    },
+    {
+      "epoch": 2.562216461438756,
+      "grad_norm": 0.9938089847564697,
+      "learning_rate": 2.740741252766077e-07,
+      "loss": 0.0621,
+      "step": 15814
+    },
+    {
+      "epoch": 2.5623784834737524,
+      "grad_norm": 0.9976528882980347,
+      "learning_rate": 2.7387508236644404e-07,
+      "loss": 0.0609,
+      "step": 15815
+    },
+    {
+      "epoch": 2.5625405055087493,
+      "grad_norm": 0.8575246334075928,
+      "learning_rate": 2.736761075703165e-07,
+      "loss": 0.0586,
+      "step": 15816
+    },
+    {
+      "epoch": 2.562702527543746,
+      "grad_norm": 0.9577282071113586,
+      "learning_rate": 2.734772008943118e-07,
+      "loss": 0.0626,
+      "step": 15817
+    },
+    {
+      "epoch": 2.5628645495787428,
+      "grad_norm": 0.8331069946289062,
+      "learning_rate": 2.732783623445168e-07,
+      "loss": 0.0563,
+      "step": 15818
+    },
+    {
+      "epoch": 2.5630265716137393,
+      "grad_norm": 0.8915289640426636,
+      "learning_rate": 2.730795919270149e-07,
+      "loss": 0.0608,
+      "step": 15819
+    },
+    {
+      "epoch": 2.563188593648736,
+      "grad_norm": 0.9112197160720825,
+      "learning_rate": 2.728808896478891e-07,
+      "loss": 0.0571,
+      "step": 15820
+    },
+    {
+      "epoch": 2.563350615683733,
+      "grad_norm": 1.0440930128097534,
+      "learning_rate": 2.726822555132183e-07,
+      "loss": 0.0612,
+      "step": 15821
+    },
+    {
+      "epoch": 2.5635126377187296,
+      "grad_norm": 1.0141581296920776,
+      "learning_rate": 2.7248368952908055e-07,
+      "loss": 0.0702,
+      "step": 15822
+    },
+    {
+      "epoch": 2.5636746597537265,
+      "grad_norm": 1.0160237550735474,
+      "learning_rate": 2.722851917015512e-07,
+      "loss": 0.0653,
+      "step": 15823
+    },
+    {
+      "epoch": 2.5638366817887235,
+      "grad_norm": 0.8820620775222778,
+      "learning_rate": 2.7208676203670406e-07,
+      "loss": 0.0551,
+      "step": 15824
+    },
+    {
+      "epoch": 2.56399870382372,
+      "grad_norm": 1.0316556692123413,
+      "learning_rate": 2.7188840054061084e-07,
+      "loss": 0.0625,
+      "step": 15825
+    },
+    {
+      "epoch": 2.564160725858717,
+      "grad_norm": 1.0389971733093262,
+      "learning_rate": 2.716901072193404e-07,
+      "loss": 0.0646,
+      "step": 15826
+    },
+    {
+      "epoch": 2.5643227478937134,
+      "grad_norm": 0.8673059940338135,
+      "learning_rate": 2.7149188207896084e-07,
+      "loss": 0.0631,
+      "step": 15827
+    },
+    {
+      "epoch": 2.5644847699287103,
+      "grad_norm": 0.8147962689399719,
+      "learning_rate": 2.71293725125536e-07,
+      "loss": 0.0547,
+      "step": 15828
+    },
+    {
+      "epoch": 2.564646791963707,
+      "grad_norm": 1.1430022716522217,
+      "learning_rate": 2.710956363651296e-07,
+      "loss": 0.0643,
+      "step": 15829
+    },
+    {
+      "epoch": 2.5648088139987038,
+      "grad_norm": 1.0345301628112793,
+      "learning_rate": 2.7089761580380346e-07,
+      "loss": 0.0694,
+      "step": 15830
+    },
+    {
+      "epoch": 2.5649708360337007,
+      "grad_norm": 1.059556245803833,
+      "learning_rate": 2.7069966344761636e-07,
+      "loss": 0.0555,
+      "step": 15831
+    },
+    {
+      "epoch": 2.565132858068697,
+      "grad_norm": 0.8952658176422119,
+      "learning_rate": 2.7050177930262406e-07,
+      "loss": 0.062,
+      "step": 15832
+    },
+    {
+      "epoch": 2.565294880103694,
+      "grad_norm": 0.9212254285812378,
+      "learning_rate": 2.703039633748822e-07,
+      "loss": 0.0525,
+      "step": 15833
+    },
+    {
+      "epoch": 2.565456902138691,
+      "grad_norm": 0.8648931980133057,
+      "learning_rate": 2.701062156704434e-07,
+      "loss": 0.0598,
+      "step": 15834
+    },
+    {
+      "epoch": 2.5656189241736875,
+      "grad_norm": 0.8723458647727966,
+      "learning_rate": 2.6990853619535793e-07,
+      "loss": 0.06,
+      "step": 15835
+    },
+    {
+      "epoch": 2.5657809462086845,
+      "grad_norm": 0.8057177662849426,
+      "learning_rate": 2.697109249556748e-07,
+      "loss": 0.0555,
+      "step": 15836
+    },
+    {
+      "epoch": 2.5659429682436814,
+      "grad_norm": 1.0397346019744873,
+      "learning_rate": 2.6951338195744e-07,
+      "loss": 0.0679,
+      "step": 15837
+    },
+    {
+      "epoch": 2.566104990278678,
+      "grad_norm": 0.824124276638031,
+      "learning_rate": 2.6931590720669807e-07,
+      "loss": 0.0582,
+      "step": 15838
+    },
+    {
+      "epoch": 2.566267012313675,
+      "grad_norm": 0.807733952999115,
+      "learning_rate": 2.6911850070949124e-07,
+      "loss": 0.0591,
+      "step": 15839
+    },
+    {
+      "epoch": 2.5664290343486713,
+      "grad_norm": 1.1735118627548218,
+      "learning_rate": 2.6892116247185964e-07,
+      "loss": 0.0717,
+      "step": 15840
+    },
+    {
+      "epoch": 2.5665910563836682,
+      "grad_norm": 0.878514289855957,
+      "learning_rate": 2.687238924998414e-07,
+      "loss": 0.0493,
+      "step": 15841
+    },
+    {
+      "epoch": 2.5667530784186647,
+      "grad_norm": 0.9837947487831116,
+      "learning_rate": 2.6852669079947294e-07,
+      "loss": 0.0616,
+      "step": 15842
+    },
+    {
+      "epoch": 2.5669151004536617,
+      "grad_norm": 0.9340665340423584,
+      "learning_rate": 2.683295573767866e-07,
+      "loss": 0.0597,
+      "step": 15843
+    },
+    {
+      "epoch": 2.5670771224886586,
+      "grad_norm": 0.9732357263565063,
+      "learning_rate": 2.681324922378159e-07,
+      "loss": 0.058,
+      "step": 15844
+    },
+    {
+      "epoch": 2.567239144523655,
+      "grad_norm": 0.8672887682914734,
+      "learning_rate": 2.679354953885899e-07,
+      "loss": 0.0621,
+      "step": 15845
+    },
+    {
+      "epoch": 2.567401166558652,
+      "grad_norm": 0.9062552452087402,
+      "learning_rate": 2.6773856683513677e-07,
+      "loss": 0.0574,
+      "step": 15846
+    },
+    {
+      "epoch": 2.567563188593649,
+      "grad_norm": 0.9597328305244446,
+      "learning_rate": 2.6754170658348094e-07,
+      "loss": 0.056,
+      "step": 15847
+    },
+    {
+      "epoch": 2.5677252106286454,
+      "grad_norm": 0.8402513265609741,
+      "learning_rate": 2.673449146396459e-07,
+      "loss": 0.0555,
+      "step": 15848
+    },
+    {
+      "epoch": 2.5678872326636424,
+      "grad_norm": 0.9047306776046753,
+      "learning_rate": 2.671481910096546e-07,
+      "loss": 0.0586,
+      "step": 15849
+    },
+    {
+      "epoch": 2.568049254698639,
+      "grad_norm": 0.9194369912147522,
+      "learning_rate": 2.6695153569952475e-07,
+      "loss": 0.0625,
+      "step": 15850
+    },
+    {
+      "epoch": 2.568211276733636,
+      "grad_norm": 0.9545932412147522,
+      "learning_rate": 2.6675494871527404e-07,
+      "loss": 0.0646,
+      "step": 15851
+    },
+    {
+      "epoch": 2.5683732987686323,
+      "grad_norm": 0.9350480437278748,
+      "learning_rate": 2.665584300629176e-07,
+      "loss": 0.0616,
+      "step": 15852
+    },
+    {
+      "epoch": 2.5685353208036292,
+      "grad_norm": 0.8764953017234802,
+      "learning_rate": 2.663619797484684e-07,
+      "loss": 0.0584,
+      "step": 15853
+    },
+    {
+      "epoch": 2.568697342838626,
+      "grad_norm": 0.904297947883606,
+      "learning_rate": 2.661655977779373e-07,
+      "loss": 0.0539,
+      "step": 15854
+    },
+    {
+      "epoch": 2.5688593648736227,
+      "grad_norm": 0.9557779431343079,
+      "learning_rate": 2.65969284157333e-07,
+      "loss": 0.0549,
+      "step": 15855
+    },
+    {
+      "epoch": 2.5690213869086196,
+      "grad_norm": 0.7922280430793762,
+      "learning_rate": 2.6577303889266244e-07,
+      "loss": 0.0576,
+      "step": 15856
+    },
+    {
+      "epoch": 2.5691834089436165,
+      "grad_norm": 1.1768953800201416,
+      "learning_rate": 2.655768619899302e-07,
+      "loss": 0.0574,
+      "step": 15857
+    },
+    {
+      "epoch": 2.569345430978613,
+      "grad_norm": 0.9072592854499817,
+      "learning_rate": 2.6538075345513864e-07,
+      "loss": 0.062,
+      "step": 15858
+    },
+    {
+      "epoch": 2.56950745301361,
+      "grad_norm": 0.9729146361351013,
+      "learning_rate": 2.651847132942886e-07,
+      "loss": 0.0647,
+      "step": 15859
+    },
+    {
+      "epoch": 2.569669475048607,
+      "grad_norm": 0.8979154825210571,
+      "learning_rate": 2.6498874151337865e-07,
+      "loss": 0.0649,
+      "step": 15860
+    },
+    {
+      "epoch": 2.5698314970836034,
+      "grad_norm": 1.0150434970855713,
+      "learning_rate": 2.6479283811840393e-07,
+      "loss": 0.0602,
+      "step": 15861
+    },
+    {
+      "epoch": 2.5699935191186,
+      "grad_norm": 0.9745667576789856,
+      "learning_rate": 2.6459700311535885e-07,
+      "loss": 0.0662,
+      "step": 15862
+    },
+    {
+      "epoch": 2.570155541153597,
+      "grad_norm": 0.8473979830741882,
+      "learning_rate": 2.6440123651023634e-07,
+      "loss": 0.0554,
+      "step": 15863
+    },
+    {
+      "epoch": 2.5703175631885937,
+      "grad_norm": 1.0695520639419556,
+      "learning_rate": 2.642055383090264e-07,
+      "loss": 0.0609,
+      "step": 15864
+    },
+    {
+      "epoch": 2.57047958522359,
+      "grad_norm": 0.8144644498825073,
+      "learning_rate": 2.6400990851771615e-07,
+      "loss": 0.0502,
+      "step": 15865
+    },
+    {
+      "epoch": 2.570641607258587,
+      "grad_norm": 0.8602494597434998,
+      "learning_rate": 2.638143471422916e-07,
+      "loss": 0.0585,
+      "step": 15866
+    },
+    {
+      "epoch": 2.570803629293584,
+      "grad_norm": 0.8381325006484985,
+      "learning_rate": 2.636188541887366e-07,
+      "loss": 0.0611,
+      "step": 15867
+    },
+    {
+      "epoch": 2.5709656513285806,
+      "grad_norm": 0.8610426783561707,
+      "learning_rate": 2.634234296630328e-07,
+      "loss": 0.0578,
+      "step": 15868
+    },
+    {
+      "epoch": 2.5711276733635775,
+      "grad_norm": 0.8060207366943359,
+      "learning_rate": 2.632280735711595e-07,
+      "loss": 0.0534,
+      "step": 15869
+    },
+    {
+      "epoch": 2.5712896953985744,
+      "grad_norm": 1.0370417833328247,
+      "learning_rate": 2.6303278591909426e-07,
+      "loss": 0.061,
+      "step": 15870
+    },
+    {
+      "epoch": 2.571451717433571,
+      "grad_norm": 0.9766287803649902,
+      "learning_rate": 2.62837566712813e-07,
+      "loss": 0.0637,
+      "step": 15871
+    },
+    {
+      "epoch": 2.571613739468568,
+      "grad_norm": 1.0253368616104126,
+      "learning_rate": 2.626424159582872e-07,
+      "loss": 0.0585,
+      "step": 15872
+    },
+    {
+      "epoch": 2.5717757615035644,
+      "grad_norm": 1.0771368741989136,
+      "learning_rate": 2.6244733366148994e-07,
+      "loss": 0.0607,
+      "step": 15873
+    },
+    {
+      "epoch": 2.5719377835385613,
+      "grad_norm": 0.9325494170188904,
+      "learning_rate": 2.622523198283894e-07,
+      "loss": 0.0618,
+      "step": 15874
+    },
+    {
+      "epoch": 2.5720998055735578,
+      "grad_norm": 0.887843906879425,
+      "learning_rate": 2.6205737446495296e-07,
+      "loss": 0.068,
+      "step": 15875
+    },
+    {
+      "epoch": 2.5722618276085547,
+      "grad_norm": 0.909279465675354,
+      "learning_rate": 2.6186249757714474e-07,
+      "loss": 0.0615,
+      "step": 15876
+    },
+    {
+      "epoch": 2.5724238496435516,
+      "grad_norm": 1.0336390733718872,
+      "learning_rate": 2.6166768917092746e-07,
+      "loss": 0.0653,
+      "step": 15877
+    },
+    {
+      "epoch": 2.572585871678548,
+      "grad_norm": 0.9292657375335693,
+      "learning_rate": 2.614729492522633e-07,
+      "loss": 0.0644,
+      "step": 15878
+    },
+    {
+      "epoch": 2.572747893713545,
+      "grad_norm": 0.995266854763031,
+      "learning_rate": 2.6127827782710916e-07,
+      "loss": 0.0614,
+      "step": 15879
+    },
+    {
+      "epoch": 2.572909915748542,
+      "grad_norm": 0.8683896660804749,
+      "learning_rate": 2.61083674901422e-07,
+      "loss": 0.0525,
+      "step": 15880
+    },
+    {
+      "epoch": 2.5730719377835385,
+      "grad_norm": 0.9227241277694702,
+      "learning_rate": 2.6088914048115585e-07,
+      "loss": 0.0609,
+      "step": 15881
+    },
+    {
+      "epoch": 2.5732339598185354,
+      "grad_norm": 0.9650192856788635,
+      "learning_rate": 2.6069467457226467e-07,
+      "loss": 0.0655,
+      "step": 15882
+    },
+    {
+      "epoch": 2.5733959818535324,
+      "grad_norm": 0.9700379371643066,
+      "learning_rate": 2.6050027718069694e-07,
+      "loss": 0.0588,
+      "step": 15883
+    },
+    {
+      "epoch": 2.573558003888529,
+      "grad_norm": 0.901326596736908,
+      "learning_rate": 2.6030594831240094e-07,
+      "loss": 0.064,
+      "step": 15884
+    },
+    {
+      "epoch": 2.5737200259235253,
+      "grad_norm": 0.947603166103363,
+      "learning_rate": 2.601116879733231e-07,
+      "loss": 0.0545,
+      "step": 15885
+    },
+    {
+      "epoch": 2.5738820479585223,
+      "grad_norm": 1.0234147310256958,
+      "learning_rate": 2.599174961694073e-07,
+      "loss": 0.0712,
+      "step": 15886
+    },
+    {
+      "epoch": 2.574044069993519,
+      "grad_norm": 0.9118129014968872,
+      "learning_rate": 2.597233729065951e-07,
+      "loss": 0.066,
+      "step": 15887
+    },
+    {
+      "epoch": 2.5742060920285157,
+      "grad_norm": 0.8596914410591125,
+      "learning_rate": 2.595293181908265e-07,
+      "loss": 0.0624,
+      "step": 15888
+    },
+    {
+      "epoch": 2.5743681140635126,
+      "grad_norm": 0.8012384176254272,
+      "learning_rate": 2.593353320280387e-07,
+      "loss": 0.0588,
+      "step": 15889
+    },
+    {
+      "epoch": 2.5745301360985096,
+      "grad_norm": 0.8562291860580444,
+      "learning_rate": 2.59141414424168e-07,
+      "loss": 0.0572,
+      "step": 15890
+    },
+    {
+      "epoch": 2.574692158133506,
+      "grad_norm": 1.0421948432922363,
+      "learning_rate": 2.5894756538514644e-07,
+      "loss": 0.0693,
+      "step": 15891
+    },
+    {
+      "epoch": 2.574854180168503,
+      "grad_norm": 0.880355715751648,
+      "learning_rate": 2.587537849169064e-07,
+      "loss": 0.0567,
+      "step": 15892
+    },
+    {
+      "epoch": 2.5750162022035,
+      "grad_norm": 0.8059505224227905,
+      "learning_rate": 2.585600730253773e-07,
+      "loss": 0.051,
+      "step": 15893
+    },
+    {
+      "epoch": 2.5751782242384964,
+      "grad_norm": 0.8571881651878357,
+      "learning_rate": 2.5836642971648534e-07,
+      "loss": 0.068,
+      "step": 15894
+    },
+    {
+      "epoch": 2.5753402462734933,
+      "grad_norm": 0.9805305600166321,
+      "learning_rate": 2.5817285499615624e-07,
+      "loss": 0.0608,
+      "step": 15895
+    },
+    {
+      "epoch": 2.57550226830849,
+      "grad_norm": 0.9158777594566345,
+      "learning_rate": 2.579793488703122e-07,
+      "loss": 0.0623,
+      "step": 15896
+    },
+    {
+      "epoch": 2.5756642903434868,
+      "grad_norm": 0.8620784878730774,
+      "learning_rate": 2.5778591134487494e-07,
+      "loss": 0.0629,
+      "step": 15897
+    },
+    {
+      "epoch": 2.5758263123784833,
+      "grad_norm": 0.8118161559104919,
+      "learning_rate": 2.5759254242576246e-07,
+      "loss": 0.0536,
+      "step": 15898
+    },
+    {
+      "epoch": 2.57598833441348,
+      "grad_norm": 0.9540441036224365,
+      "learning_rate": 2.5739924211889173e-07,
+      "loss": 0.0572,
+      "step": 15899
+    },
+    {
+      "epoch": 2.576150356448477,
+      "grad_norm": 0.7988532185554504,
+      "learning_rate": 2.572060104301771e-07,
+      "loss": 0.0562,
+      "step": 15900
+    },
+    {
+      "epoch": 2.5763123784834736,
+      "grad_norm": 0.9321579933166504,
+      "learning_rate": 2.5701284736553146e-07,
+      "loss": 0.0652,
+      "step": 15901
+    },
+    {
+      "epoch": 2.5764744005184705,
+      "grad_norm": 0.7826856970787048,
+      "learning_rate": 2.5681975293086443e-07,
+      "loss": 0.055,
+      "step": 15902
+    },
+    {
+      "epoch": 2.5766364225534675,
+      "grad_norm": 0.9522474408149719,
+      "learning_rate": 2.5662672713208465e-07,
+      "loss": 0.0601,
+      "step": 15903
+    },
+    {
+      "epoch": 2.576798444588464,
+      "grad_norm": 0.9039526581764221,
+      "learning_rate": 2.564337699750985e-07,
+      "loss": 0.0628,
+      "step": 15904
+    },
+    {
+      "epoch": 2.576960466623461,
+      "grad_norm": 0.9911723136901855,
+      "learning_rate": 2.5624088146580903e-07,
+      "loss": 0.07,
+      "step": 15905
+    },
+    {
+      "epoch": 2.5771224886584574,
+      "grad_norm": 0.9902855753898621,
+      "learning_rate": 2.560480616101191e-07,
+      "loss": 0.0602,
+      "step": 15906
+    },
+    {
+      "epoch": 2.5772845106934543,
+      "grad_norm": 1.157789707183838,
+      "learning_rate": 2.55855310413928e-07,
+      "loss": 0.0609,
+      "step": 15907
+    },
+    {
+      "epoch": 2.577446532728451,
+      "grad_norm": 0.9956571459770203,
+      "learning_rate": 2.556626278831345e-07,
+      "loss": 0.0597,
+      "step": 15908
+    },
+    {
+      "epoch": 2.5776085547634477,
+      "grad_norm": 0.8518067598342896,
+      "learning_rate": 2.554700140236327e-07,
+      "loss": 0.0552,
+      "step": 15909
+    },
+    {
+      "epoch": 2.5777705767984447,
+      "grad_norm": 0.8812946081161499,
+      "learning_rate": 2.552774688413165e-07,
+      "loss": 0.0525,
+      "step": 15910
+    },
+    {
+      "epoch": 2.577932598833441,
+      "grad_norm": 1.0784142017364502,
+      "learning_rate": 2.550849923420787e-07,
+      "loss": 0.0615,
+      "step": 15911
+    },
+    {
+      "epoch": 2.578094620868438,
+      "grad_norm": 1.0830011367797852,
+      "learning_rate": 2.5489258453180676e-07,
+      "loss": 0.0558,
+      "step": 15912
+    },
+    {
+      "epoch": 2.578256642903435,
+      "grad_norm": 0.8007397055625916,
+      "learning_rate": 2.547002454163888e-07,
+      "loss": 0.0524,
+      "step": 15913
+    },
+    {
+      "epoch": 2.5784186649384315,
+      "grad_norm": 0.9391987323760986,
+      "learning_rate": 2.545079750017099e-07,
+      "loss": 0.0546,
+      "step": 15914
+    },
+    {
+      "epoch": 2.5785806869734285,
+      "grad_norm": 0.9835785627365112,
+      "learning_rate": 2.543157732936527e-07,
+      "loss": 0.061,
+      "step": 15915
+    },
+    {
+      "epoch": 2.5787427090084254,
+      "grad_norm": 0.9381945729255676,
+      "learning_rate": 2.541236402980987e-07,
+      "loss": 0.0547,
+      "step": 15916
+    },
+    {
+      "epoch": 2.578904731043422,
+      "grad_norm": 1.088310956954956,
+      "learning_rate": 2.5393157602092626e-07,
+      "loss": 0.0656,
+      "step": 15917
+    },
+    {
+      "epoch": 2.579066753078419,
+      "grad_norm": 0.9402004480361938,
+      "learning_rate": 2.5373958046801207e-07,
+      "loss": 0.0596,
+      "step": 15918
+    },
+    {
+      "epoch": 2.5792287751134153,
+      "grad_norm": 0.8306174874305725,
+      "learning_rate": 2.5354765364523164e-07,
+      "loss": 0.0564,
+      "step": 15919
+    },
+    {
+      "epoch": 2.5793907971484122,
+      "grad_norm": 0.8160369396209717,
+      "learning_rate": 2.5335579555845563e-07,
+      "loss": 0.0584,
+      "step": 15920
+    },
+    {
+      "epoch": 2.5795528191834087,
+      "grad_norm": 1.0433648824691772,
+      "learning_rate": 2.531640062135557e-07,
+      "loss": 0.0592,
+      "step": 15921
+    },
+    {
+      "epoch": 2.5797148412184057,
+      "grad_norm": 1.0717437267303467,
+      "learning_rate": 2.5297228561640075e-07,
+      "loss": 0.0689,
+      "step": 15922
+    },
+    {
+      "epoch": 2.5798768632534026,
+      "grad_norm": 0.8935438394546509,
+      "learning_rate": 2.5278063377285556e-07,
+      "loss": 0.0655,
+      "step": 15923
+    },
+    {
+      "epoch": 2.580038885288399,
+      "grad_norm": 0.8454123735427856,
+      "learning_rate": 2.5258905068878433e-07,
+      "loss": 0.0582,
+      "step": 15924
+    },
+    {
+      "epoch": 2.580200907323396,
+      "grad_norm": 0.9273526072502136,
+      "learning_rate": 2.523975363700501e-07,
+      "loss": 0.0555,
+      "step": 15925
+    },
+    {
+      "epoch": 2.580362929358393,
+      "grad_norm": 1.308284878730774,
+      "learning_rate": 2.522060908225127e-07,
+      "loss": 0.0684,
+      "step": 15926
+    },
+    {
+      "epoch": 2.5805249513933894,
+      "grad_norm": 0.8910729885101318,
+      "learning_rate": 2.520147140520288e-07,
+      "loss": 0.0531,
+      "step": 15927
+    },
+    {
+      "epoch": 2.5806869734283864,
+      "grad_norm": 0.9772580862045288,
+      "learning_rate": 2.518234060644545e-07,
+      "loss": 0.0598,
+      "step": 15928
+    },
+    {
+      "epoch": 2.580848995463383,
+      "grad_norm": 0.845855712890625,
+      "learning_rate": 2.5163216686564354e-07,
+      "loss": 0.0567,
+      "step": 15929
+    },
+    {
+      "epoch": 2.58101101749838,
+      "grad_norm": 0.9278295040130615,
+      "learning_rate": 2.5144099646144724e-07,
+      "loss": 0.0579,
+      "step": 15930
+    },
+    {
+      "epoch": 2.5811730395333763,
+      "grad_norm": 1.0706762075424194,
+      "learning_rate": 2.512498948577152e-07,
+      "loss": 0.0644,
+      "step": 15931
+    },
+    {
+      "epoch": 2.5813350615683732,
+      "grad_norm": 0.9492532014846802,
+      "learning_rate": 2.510588620602947e-07,
+      "loss": 0.0567,
+      "step": 15932
+    },
+    {
+      "epoch": 2.58149708360337,
+      "grad_norm": 0.8904463648796082,
+      "learning_rate": 2.5086789807503036e-07,
+      "loss": 0.0576,
+      "step": 15933
+    },
+    {
+      "epoch": 2.5816591056383666,
+      "grad_norm": 0.8406658172607422,
+      "learning_rate": 2.506770029077657e-07,
+      "loss": 0.0558,
+      "step": 15934
+    },
+    {
+      "epoch": 2.5818211276733636,
+      "grad_norm": 0.919401228427887,
+      "learning_rate": 2.5048617656434127e-07,
+      "loss": 0.0607,
+      "step": 15935
+    },
+    {
+      "epoch": 2.5819831497083605,
+      "grad_norm": 0.9194125533103943,
+      "learning_rate": 2.502954190505963e-07,
+      "loss": 0.0582,
+      "step": 15936
+    },
+    {
+      "epoch": 2.582145171743357,
+      "grad_norm": 0.8466059565544128,
+      "learning_rate": 2.5010473037236776e-07,
+      "loss": 0.0589,
+      "step": 15937
+    },
+    {
+      "epoch": 2.582307193778354,
+      "grad_norm": 0.8999043107032776,
+      "learning_rate": 2.499141105354894e-07,
+      "loss": 0.0631,
+      "step": 15938
+    },
+    {
+      "epoch": 2.582469215813351,
+      "grad_norm": 0.8571875095367432,
+      "learning_rate": 2.4972355954579366e-07,
+      "loss": 0.0622,
+      "step": 15939
+    },
+    {
+      "epoch": 2.5826312378483474,
+      "grad_norm": 0.9047456383705139,
+      "learning_rate": 2.495330774091126e-07,
+      "loss": 0.0596,
+      "step": 15940
+    },
+    {
+      "epoch": 2.5827932598833443,
+      "grad_norm": 0.9654316902160645,
+      "learning_rate": 2.493426641312724e-07,
+      "loss": 0.0587,
+      "step": 15941
+    },
+    {
+      "epoch": 2.582955281918341,
+      "grad_norm": 0.8947680592536926,
+      "learning_rate": 2.4915231971810064e-07,
+      "loss": 0.0586,
+      "step": 15942
+    },
+    {
+      "epoch": 2.5831173039533377,
+      "grad_norm": 0.7643946409225464,
+      "learning_rate": 2.4896204417542066e-07,
+      "loss": 0.0537,
+      "step": 15943
+    },
+    {
+      "epoch": 2.583279325988334,
+      "grad_norm": 0.9256800413131714,
+      "learning_rate": 2.4877183750905475e-07,
+      "loss": 0.0596,
+      "step": 15944
+    },
+    {
+      "epoch": 2.583441348023331,
+      "grad_norm": 1.0002453327178955,
+      "learning_rate": 2.4858169972482276e-07,
+      "loss": 0.0602,
+      "step": 15945
+    },
+    {
+      "epoch": 2.583603370058328,
+      "grad_norm": 0.8987006545066833,
+      "learning_rate": 2.483916308285425e-07,
+      "loss": 0.0584,
+      "step": 15946
+    },
+    {
+      "epoch": 2.5837653920933246,
+      "grad_norm": 0.8947418332099915,
+      "learning_rate": 2.482016308260296e-07,
+      "loss": 0.0596,
+      "step": 15947
+    },
+    {
+      "epoch": 2.5839274141283215,
+      "grad_norm": 0.8880223631858826,
+      "learning_rate": 2.4801169972309745e-07,
+      "loss": 0.0575,
+      "step": 15948
+    },
+    {
+      "epoch": 2.5840894361633184,
+      "grad_norm": 0.8866935968399048,
+      "learning_rate": 2.4782183752555784e-07,
+      "loss": 0.0536,
+      "step": 15949
+    },
+    {
+      "epoch": 2.584251458198315,
+      "grad_norm": 0.8493706583976746,
+      "learning_rate": 2.4763204423921937e-07,
+      "loss": 0.0541,
+      "step": 15950
+    },
+    {
+      "epoch": 2.584413480233312,
+      "grad_norm": 0.8858264684677124,
+      "learning_rate": 2.4744231986988996e-07,
+      "loss": 0.0645,
+      "step": 15951
+    },
+    {
+      "epoch": 2.5845755022683083,
+      "grad_norm": 1.0235286951065063,
+      "learning_rate": 2.47252664423375e-07,
+      "loss": 0.0737,
+      "step": 15952
+    },
+    {
+      "epoch": 2.5847375243033053,
+      "grad_norm": 1.0129871368408203,
+      "learning_rate": 2.4706307790547614e-07,
+      "loss": 0.0563,
+      "step": 15953
+    },
+    {
+      "epoch": 2.5848995463383018,
+      "grad_norm": 0.9243896007537842,
+      "learning_rate": 2.4687356032199516e-07,
+      "loss": 0.0617,
+      "step": 15954
+    },
+    {
+      "epoch": 2.5850615683732987,
+      "grad_norm": 0.7274792790412903,
+      "learning_rate": 2.4668411167873165e-07,
+      "loss": 0.0506,
+      "step": 15955
+    },
+    {
+      "epoch": 2.5852235904082956,
+      "grad_norm": 0.7890907526016235,
+      "learning_rate": 2.464947319814806e-07,
+      "loss": 0.0544,
+      "step": 15956
+    },
+    {
+      "epoch": 2.585385612443292,
+      "grad_norm": 0.9411144852638245,
+      "learning_rate": 2.4630542123603775e-07,
+      "loss": 0.0586,
+      "step": 15957
+    },
+    {
+      "epoch": 2.585547634478289,
+      "grad_norm": 0.8673883080482483,
+      "learning_rate": 2.461161794481945e-07,
+      "loss": 0.0541,
+      "step": 15958
+    },
+    {
+      "epoch": 2.585709656513286,
+      "grad_norm": 1.1016530990600586,
+      "learning_rate": 2.4592700662374265e-07,
+      "loss": 0.0667,
+      "step": 15959
+    },
+    {
+      "epoch": 2.5858716785482825,
+      "grad_norm": 0.9959900975227356,
+      "learning_rate": 2.4573790276846947e-07,
+      "loss": 0.0651,
+      "step": 15960
+    },
+    {
+      "epoch": 2.5860337005832794,
+      "grad_norm": 0.895260751247406,
+      "learning_rate": 2.4554886788816094e-07,
+      "loss": 0.0672,
+      "step": 15961
+    },
+    {
+      "epoch": 2.5861957226182763,
+      "grad_norm": 1.0384845733642578,
+      "learning_rate": 2.453599019886016e-07,
+      "loss": 0.059,
+      "step": 15962
+    },
+    {
+      "epoch": 2.586357744653273,
+      "grad_norm": 0.8022738099098206,
+      "learning_rate": 2.451710050755732e-07,
+      "loss": 0.0506,
+      "step": 15963
+    },
+    {
+      "epoch": 2.5865197666882693,
+      "grad_norm": 0.8844985365867615,
+      "learning_rate": 2.449821771548552e-07,
+      "loss": 0.0591,
+      "step": 15964
+    },
+    {
+      "epoch": 2.5866817887232663,
+      "grad_norm": 0.8309991955757141,
+      "learning_rate": 2.4479341823222564e-07,
+      "loss": 0.0558,
+      "step": 15965
+    },
+    {
+      "epoch": 2.586843810758263,
+      "grad_norm": 0.9121332168579102,
+      "learning_rate": 2.446047283134606e-07,
+      "loss": 0.0673,
+      "step": 15966
+    },
+    {
+      "epoch": 2.5870058327932597,
+      "grad_norm": 0.8563621640205383,
+      "learning_rate": 2.444161074043325e-07,
+      "loss": 0.0605,
+      "step": 15967
+    },
+    {
+      "epoch": 2.5871678548282566,
+      "grad_norm": 0.834702730178833,
+      "learning_rate": 2.4422755551061246e-07,
+      "loss": 0.0561,
+      "step": 15968
+    },
+    {
+      "epoch": 2.5873298768632536,
+      "grad_norm": 0.9449539184570312,
+      "learning_rate": 2.4403907263807064e-07,
+      "loss": 0.0555,
+      "step": 15969
+    },
+    {
+      "epoch": 2.58749189889825,
+      "grad_norm": 1.0673609972000122,
+      "learning_rate": 2.4385065879247466e-07,
+      "loss": 0.0612,
+      "step": 15970
+    },
+    {
+      "epoch": 2.587653920933247,
+      "grad_norm": 0.8567180633544922,
+      "learning_rate": 2.4366231397958823e-07,
+      "loss": 0.0603,
+      "step": 15971
+    },
+    {
+      "epoch": 2.587815942968244,
+      "grad_norm": 0.8513876795768738,
+      "learning_rate": 2.4347403820517423e-07,
+      "loss": 0.058,
+      "step": 15972
+    },
+    {
+      "epoch": 2.5879779650032404,
+      "grad_norm": 0.8742343187332153,
+      "learning_rate": 2.4328583147499503e-07,
+      "loss": 0.0598,
+      "step": 15973
+    },
+    {
+      "epoch": 2.5881399870382373,
+      "grad_norm": 1.0191162824630737,
+      "learning_rate": 2.4309769379480764e-07,
+      "loss": 0.0576,
+      "step": 15974
+    },
+    {
+      "epoch": 2.588302009073234,
+      "grad_norm": 0.8935777544975281,
+      "learning_rate": 2.4290962517036915e-07,
+      "loss": 0.0568,
+      "step": 15975
+    },
+    {
+      "epoch": 2.5884640311082308,
+      "grad_norm": 0.8707433342933655,
+      "learning_rate": 2.427216256074341e-07,
+      "loss": 0.0537,
+      "step": 15976
+    },
+    {
+      "epoch": 2.5886260531432272,
+      "grad_norm": 1.0060011148452759,
+      "learning_rate": 2.425336951117549e-07,
+      "loss": 0.0622,
+      "step": 15977
+    },
+    {
+      "epoch": 2.588788075178224,
+      "grad_norm": 0.9074399471282959,
+      "learning_rate": 2.423458336890816e-07,
+      "loss": 0.0621,
+      "step": 15978
+    },
+    {
+      "epoch": 2.588950097213221,
+      "grad_norm": 0.931549072265625,
+      "learning_rate": 2.421580413451624e-07,
+      "loss": 0.0688,
+      "step": 15979
+    },
+    {
+      "epoch": 2.5891121192482176,
+      "grad_norm": 0.9901089072227478,
+      "learning_rate": 2.4197031808574327e-07,
+      "loss": 0.0644,
+      "step": 15980
+    },
+    {
+      "epoch": 2.5892741412832145,
+      "grad_norm": 0.8553241491317749,
+      "learning_rate": 2.417826639165688e-07,
+      "loss": 0.0572,
+      "step": 15981
+    },
+    {
+      "epoch": 2.5894361633182115,
+      "grad_norm": 0.7999852895736694,
+      "learning_rate": 2.4159507884337877e-07,
+      "loss": 0.0553,
+      "step": 15982
+    },
+    {
+      "epoch": 2.589598185353208,
+      "grad_norm": 0.7739843726158142,
+      "learning_rate": 2.414075628719145e-07,
+      "loss": 0.0565,
+      "step": 15983
+    },
+    {
+      "epoch": 2.589760207388205,
+      "grad_norm": 0.8175353407859802,
+      "learning_rate": 2.4122011600791334e-07,
+      "loss": 0.0576,
+      "step": 15984
+    },
+    {
+      "epoch": 2.589922229423202,
+      "grad_norm": 0.9324160218238831,
+      "learning_rate": 2.4103273825711094e-07,
+      "loss": 0.0601,
+      "step": 15985
+    },
+    {
+      "epoch": 2.5900842514581983,
+      "grad_norm": 1.0186549425125122,
+      "learning_rate": 2.408454296252397e-07,
+      "loss": 0.0605,
+      "step": 15986
+    },
+    {
+      "epoch": 2.590246273493195,
+      "grad_norm": 1.0774922370910645,
+      "learning_rate": 2.406581901180305e-07,
+      "loss": 0.0632,
+      "step": 15987
+    },
+    {
+      "epoch": 2.5904082955281917,
+      "grad_norm": 1.0663115978240967,
+      "learning_rate": 2.404710197412144e-07,
+      "loss": 0.0681,
+      "step": 15988
+    },
+    {
+      "epoch": 2.5905703175631887,
+      "grad_norm": 0.9276178479194641,
+      "learning_rate": 2.4028391850051654e-07,
+      "loss": 0.0622,
+      "step": 15989
+    },
+    {
+      "epoch": 2.590732339598185,
+      "grad_norm": 0.9000306725502014,
+      "learning_rate": 2.4009688640166257e-07,
+      "loss": 0.0572,
+      "step": 15990
+    },
+    {
+      "epoch": 2.590894361633182,
+      "grad_norm": 1.0842214822769165,
+      "learning_rate": 2.399099234503749e-07,
+      "loss": 0.0616,
+      "step": 15991
+    },
+    {
+      "epoch": 2.591056383668179,
+      "grad_norm": 0.9430128335952759,
+      "learning_rate": 2.397230296523742e-07,
+      "loss": 0.0669,
+      "step": 15992
+    },
+    {
+      "epoch": 2.5912184057031755,
+      "grad_norm": 1.052301049232483,
+      "learning_rate": 2.39536205013379e-07,
+      "loss": 0.07,
+      "step": 15993
+    },
+    {
+      "epoch": 2.5913804277381725,
+      "grad_norm": 0.9034858345985413,
+      "learning_rate": 2.3934944953910576e-07,
+      "loss": 0.0519,
+      "step": 15994
+    },
+    {
+      "epoch": 2.5915424497731694,
+      "grad_norm": 0.8782376050949097,
+      "learning_rate": 2.391627632352686e-07,
+      "loss": 0.0527,
+      "step": 15995
+    },
+    {
+      "epoch": 2.591704471808166,
+      "grad_norm": 0.8798543214797974,
+      "learning_rate": 2.3897614610757984e-07,
+      "loss": 0.0616,
+      "step": 15996
+    },
+    {
+      "epoch": 2.591866493843163,
+      "grad_norm": 0.9117831587791443,
+      "learning_rate": 2.387895981617497e-07,
+      "loss": 0.061,
+      "step": 15997
+    },
+    {
+      "epoch": 2.5920285158781593,
+      "grad_norm": 0.8840068578720093,
+      "learning_rate": 2.386031194034855e-07,
+      "loss": 0.0605,
+      "step": 15998
+    },
+    {
+      "epoch": 2.5921905379131562,
+      "grad_norm": 0.9171715378761292,
+      "learning_rate": 2.3841670983849402e-07,
+      "loss": 0.0562,
+      "step": 15999
+    },
+    {
+      "epoch": 2.5923525599481527,
+      "grad_norm": 0.9331315755844116,
+      "learning_rate": 2.3823036947247773e-07,
+      "loss": 0.0573,
+      "step": 16000
+    },
+    {
+      "epoch": 2.5925145819831497,
+      "grad_norm": 0.8709453344345093,
+      "learning_rate": 2.3804409831113817e-07,
+      "loss": 0.0554,
+      "step": 16001
+    },
+    {
+      "epoch": 2.5926766040181466,
+      "grad_norm": 0.8639891147613525,
+      "learning_rate": 2.3785789636017604e-07,
+      "loss": 0.0603,
+      "step": 16002
+    },
+    {
+      "epoch": 2.592838626053143,
+      "grad_norm": 0.9981393814086914,
+      "learning_rate": 2.3767176362528843e-07,
+      "loss": 0.0577,
+      "step": 16003
+    },
+    {
+      "epoch": 2.59300064808814,
+      "grad_norm": 0.8646852374076843,
+      "learning_rate": 2.374857001121697e-07,
+      "loss": 0.0618,
+      "step": 16004
+    },
+    {
+      "epoch": 2.593162670123137,
+      "grad_norm": 0.9664946794509888,
+      "learning_rate": 2.3729970582651307e-07,
+      "loss": 0.0676,
+      "step": 16005
+    },
+    {
+      "epoch": 2.5933246921581334,
+      "grad_norm": 0.8121821284294128,
+      "learning_rate": 2.371137807740101e-07,
+      "loss": 0.0553,
+      "step": 16006
+    },
+    {
+      "epoch": 2.5934867141931304,
+      "grad_norm": 1.0388915538787842,
+      "learning_rate": 2.3692792496034928e-07,
+      "loss": 0.0603,
+      "step": 16007
+    },
+    {
+      "epoch": 2.593648736228127,
+      "grad_norm": 0.9599438905715942,
+      "learning_rate": 2.3674213839121745e-07,
+      "loss": 0.0629,
+      "step": 16008
+    },
+    {
+      "epoch": 2.593810758263124,
+      "grad_norm": 0.8781747817993164,
+      "learning_rate": 2.3655642107229925e-07,
+      "loss": 0.0596,
+      "step": 16009
+    },
+    {
+      "epoch": 2.5939727802981203,
+      "grad_norm": 0.9688482284545898,
+      "learning_rate": 2.3637077300927762e-07,
+      "loss": 0.0646,
+      "step": 16010
+    },
+    {
+      "epoch": 2.594134802333117,
+      "grad_norm": 0.9513368606567383,
+      "learning_rate": 2.3618519420783137e-07,
+      "loss": 0.0647,
+      "step": 16011
+    },
+    {
+      "epoch": 2.594296824368114,
+      "grad_norm": 1.025209903717041,
+      "learning_rate": 2.3599968467364037e-07,
+      "loss": 0.0603,
+      "step": 16012
+    },
+    {
+      "epoch": 2.5944588464031106,
+      "grad_norm": 0.7293410897254944,
+      "learning_rate": 2.3581424441238038e-07,
+      "loss": 0.0476,
+      "step": 16013
+    },
+    {
+      "epoch": 2.5946208684381076,
+      "grad_norm": 0.8842470645904541,
+      "learning_rate": 2.3562887342972574e-07,
+      "loss": 0.0616,
+      "step": 16014
+    },
+    {
+      "epoch": 2.5947828904731045,
+      "grad_norm": 0.8669375777244568,
+      "learning_rate": 2.3544357173134691e-07,
+      "loss": 0.0557,
+      "step": 16015
+    },
+    {
+      "epoch": 2.594944912508101,
+      "grad_norm": 1.0717798471450806,
+      "learning_rate": 2.3525833932291491e-07,
+      "loss": 0.0648,
+      "step": 16016
+    },
+    {
+      "epoch": 2.595106934543098,
+      "grad_norm": 0.8745622038841248,
+      "learning_rate": 2.35073176210098e-07,
+      "loss": 0.063,
+      "step": 16017
+    },
+    {
+      "epoch": 2.595268956578095,
+      "grad_norm": 1.0365678071975708,
+      "learning_rate": 2.3488808239855998e-07,
+      "loss": 0.063,
+      "step": 16018
+    },
+    {
+      "epoch": 2.5954309786130914,
+      "grad_norm": 0.9520593285560608,
+      "learning_rate": 2.3470305789396546e-07,
+      "loss": 0.06,
+      "step": 16019
+    },
+    {
+      "epoch": 2.5955930006480883,
+      "grad_norm": 0.9252042174339294,
+      "learning_rate": 2.3451810270197494e-07,
+      "loss": 0.062,
+      "step": 16020
+    },
+    {
+      "epoch": 2.595755022683085,
+      "grad_norm": 0.9404069781303406,
+      "learning_rate": 2.3433321682824917e-07,
+      "loss": 0.0542,
+      "step": 16021
+    },
+    {
+      "epoch": 2.5959170447180817,
+      "grad_norm": 0.8857480883598328,
+      "learning_rate": 2.341484002784436e-07,
+      "loss": 0.0554,
+      "step": 16022
+    },
+    {
+      "epoch": 2.596079066753078,
+      "grad_norm": 0.9531388878822327,
+      "learning_rate": 2.3396365305821372e-07,
+      "loss": 0.0647,
+      "step": 16023
+    },
+    {
+      "epoch": 2.596241088788075,
+      "grad_norm": 0.8874648809432983,
+      "learning_rate": 2.3377897517321224e-07,
+      "loss": 0.0634,
+      "step": 16024
+    },
+    {
+      "epoch": 2.596403110823072,
+      "grad_norm": 0.8631939888000488,
+      "learning_rate": 2.3359436662909018e-07,
+      "loss": 0.0571,
+      "step": 16025
+    },
+    {
+      "epoch": 2.5965651328580686,
+      "grad_norm": 0.7769051194190979,
+      "learning_rate": 2.3340982743149582e-07,
+      "loss": 0.0585,
+      "step": 16026
+    },
+    {
+      "epoch": 2.5967271548930655,
+      "grad_norm": 0.8491709232330322,
+      "learning_rate": 2.3322535758607573e-07,
+      "loss": 0.0531,
+      "step": 16027
+    },
+    {
+      "epoch": 2.5968891769280624,
+      "grad_norm": 0.8236337304115295,
+      "learning_rate": 2.3304095709847402e-07,
+      "loss": 0.0529,
+      "step": 16028
+    },
+    {
+      "epoch": 2.597051198963059,
+      "grad_norm": 0.9366066455841064,
+      "learning_rate": 2.3285662597433368e-07,
+      "loss": 0.0658,
+      "step": 16029
+    },
+    {
+      "epoch": 2.597213220998056,
+      "grad_norm": 1.007311463356018,
+      "learning_rate": 2.3267236421929323e-07,
+      "loss": 0.0583,
+      "step": 16030
+    },
+    {
+      "epoch": 2.5973752430330523,
+      "grad_norm": 0.8656664490699768,
+      "learning_rate": 2.3248817183899209e-07,
+      "loss": 0.0572,
+      "step": 16031
+    },
+    {
+      "epoch": 2.5975372650680493,
+      "grad_norm": 0.909775972366333,
+      "learning_rate": 2.3230404883906626e-07,
+      "loss": 0.0573,
+      "step": 16032
+    },
+    {
+      "epoch": 2.5976992871030458,
+      "grad_norm": 0.9270328283309937,
+      "learning_rate": 2.321199952251482e-07,
+      "loss": 0.0578,
+      "step": 16033
+    },
+    {
+      "epoch": 2.5978613091380427,
+      "grad_norm": 1.056470274925232,
+      "learning_rate": 2.319360110028701e-07,
+      "loss": 0.0651,
+      "step": 16034
+    },
+    {
+      "epoch": 2.5980233311730396,
+      "grad_norm": 0.8853468894958496,
+      "learning_rate": 2.3175209617786133e-07,
+      "loss": 0.063,
+      "step": 16035
+    },
+    {
+      "epoch": 2.598185353208036,
+      "grad_norm": 0.9667817950248718,
+      "learning_rate": 2.3156825075574956e-07,
+      "loss": 0.06,
+      "step": 16036
+    },
+    {
+      "epoch": 2.598347375243033,
+      "grad_norm": 0.7612243890762329,
+      "learning_rate": 2.3138447474215981e-07,
+      "loss": 0.0556,
+      "step": 16037
+    },
+    {
+      "epoch": 2.59850939727803,
+      "grad_norm": 1.081939935684204,
+      "learning_rate": 2.312007681427153e-07,
+      "loss": 0.0573,
+      "step": 16038
+    },
+    {
+      "epoch": 2.5986714193130265,
+      "grad_norm": 0.9254032969474792,
+      "learning_rate": 2.3101713096303658e-07,
+      "loss": 0.0651,
+      "step": 16039
+    },
+    {
+      "epoch": 2.5988334413480234,
+      "grad_norm": 0.9094540476799011,
+      "learning_rate": 2.30833563208743e-07,
+      "loss": 0.0645,
+      "step": 16040
+    },
+    {
+      "epoch": 2.5989954633830203,
+      "grad_norm": 0.8876714706420898,
+      "learning_rate": 2.3065006488545122e-07,
+      "loss": 0.0662,
+      "step": 16041
+    },
+    {
+      "epoch": 2.599157485418017,
+      "grad_norm": 0.9631155729293823,
+      "learning_rate": 2.304666359987756e-07,
+      "loss": 0.0585,
+      "step": 16042
+    },
+    {
+      "epoch": 2.5993195074530138,
+      "grad_norm": 0.8992942571640015,
+      "learning_rate": 2.302832765543292e-07,
+      "loss": 0.0564,
+      "step": 16043
+    },
+    {
+      "epoch": 2.5994815294880103,
+      "grad_norm": 0.8657216429710388,
+      "learning_rate": 2.300999865577211e-07,
+      "loss": 0.0597,
+      "step": 16044
+    },
+    {
+      "epoch": 2.599643551523007,
+      "grad_norm": 0.8959026336669922,
+      "learning_rate": 2.2991676601456069e-07,
+      "loss": 0.0652,
+      "step": 16045
+    },
+    {
+      "epoch": 2.5998055735580037,
+      "grad_norm": 0.8713762760162354,
+      "learning_rate": 2.2973361493045382e-07,
+      "loss": 0.0621,
+      "step": 16046
+    },
+    {
+      "epoch": 2.5999675955930006,
+      "grad_norm": 0.9186781048774719,
+      "learning_rate": 2.2955053331100486e-07,
+      "loss": 0.0584,
+      "step": 16047
+    },
+    {
+      "epoch": 2.6001296176279975,
+      "grad_norm": 1.0051521062850952,
+      "learning_rate": 2.293675211618146e-07,
+      "loss": 0.065,
+      "step": 16048
+    },
+    {
+      "epoch": 2.600291639662994,
+      "grad_norm": 0.8411121964454651,
+      "learning_rate": 2.2918457848848303e-07,
+      "loss": 0.0585,
+      "step": 16049
+    },
+    {
+      "epoch": 2.600453661697991,
+      "grad_norm": 0.9174489974975586,
+      "learning_rate": 2.2900170529660898e-07,
+      "loss": 0.0634,
+      "step": 16050
+    },
+    {
+      "epoch": 2.600615683732988,
+      "grad_norm": 0.8582833409309387,
+      "learning_rate": 2.288189015917866e-07,
+      "loss": 0.0589,
+      "step": 16051
+    },
+    {
+      "epoch": 2.6007777057679844,
+      "grad_norm": 0.7952532768249512,
+      "learning_rate": 2.2863616737960976e-07,
+      "loss": 0.0533,
+      "step": 16052
+    },
+    {
+      "epoch": 2.6009397278029813,
+      "grad_norm": 0.9317299723625183,
+      "learning_rate": 2.2845350266566952e-07,
+      "loss": 0.0659,
+      "step": 16053
+    },
+    {
+      "epoch": 2.601101749837978,
+      "grad_norm": 0.8274461627006531,
+      "learning_rate": 2.2827090745555502e-07,
+      "loss": 0.0573,
+      "step": 16054
+    },
+    {
+      "epoch": 2.6012637718729748,
+      "grad_norm": 0.8277769088745117,
+      "learning_rate": 2.2808838175485321e-07,
+      "loss": 0.054,
+      "step": 16055
+    },
+    {
+      "epoch": 2.6014257939079712,
+      "grad_norm": 0.8775892853736877,
+      "learning_rate": 2.279059255691493e-07,
+      "loss": 0.061,
+      "step": 16056
+    },
+    {
+      "epoch": 2.601587815942968,
+      "grad_norm": 1.035484790802002,
+      "learning_rate": 2.2772353890402527e-07,
+      "loss": 0.0613,
+      "step": 16057
+    },
+    {
+      "epoch": 2.601749837977965,
+      "grad_norm": 0.9841363430023193,
+      "learning_rate": 2.2754122176506244e-07,
+      "loss": 0.0717,
+      "step": 16058
+    },
+    {
+      "epoch": 2.6019118600129616,
+      "grad_norm": 0.8622410297393799,
+      "learning_rate": 2.2735897415783888e-07,
+      "loss": 0.0613,
+      "step": 16059
+    },
+    {
+      "epoch": 2.6020738820479585,
+      "grad_norm": 0.7690012454986572,
+      "learning_rate": 2.271767960879312e-07,
+      "loss": 0.0528,
+      "step": 16060
+    },
+    {
+      "epoch": 2.6022359040829555,
+      "grad_norm": 0.8313643336296082,
+      "learning_rate": 2.2699468756091385e-07,
+      "loss": 0.0507,
+      "step": 16061
+    },
+    {
+      "epoch": 2.602397926117952,
+      "grad_norm": 0.8808419704437256,
+      "learning_rate": 2.2681264858235797e-07,
+      "loss": 0.0611,
+      "step": 16062
+    },
+    {
+      "epoch": 2.602559948152949,
+      "grad_norm": 0.8801603317260742,
+      "learning_rate": 2.2663067915783349e-07,
+      "loss": 0.0557,
+      "step": 16063
+    },
+    {
+      "epoch": 2.602721970187946,
+      "grad_norm": 0.8919458985328674,
+      "learning_rate": 2.2644877929290932e-07,
+      "loss": 0.0587,
+      "step": 16064
+    },
+    {
+      "epoch": 2.6028839922229423,
+      "grad_norm": 0.8611831665039062,
+      "learning_rate": 2.26266948993151e-07,
+      "loss": 0.0564,
+      "step": 16065
+    },
+    {
+      "epoch": 2.6030460142579392,
+      "grad_norm": 0.9475187659263611,
+      "learning_rate": 2.2608518826412128e-07,
+      "loss": 0.0631,
+      "step": 16066
+    },
+    {
+      "epoch": 2.6032080362929357,
+      "grad_norm": 0.8380742073059082,
+      "learning_rate": 2.2590349711138214e-07,
+      "loss": 0.0553,
+      "step": 16067
+    },
+    {
+      "epoch": 2.6033700583279327,
+      "grad_norm": 0.780228316783905,
+      "learning_rate": 2.2572187554049274e-07,
+      "loss": 0.0566,
+      "step": 16068
+    },
+    {
+      "epoch": 2.603532080362929,
+      "grad_norm": 0.9787698984146118,
+      "learning_rate": 2.2554032355701027e-07,
+      "loss": 0.0641,
+      "step": 16069
+    },
+    {
+      "epoch": 2.603694102397926,
+      "grad_norm": 0.9883422255516052,
+      "learning_rate": 2.2535884116648976e-07,
+      "loss": 0.0625,
+      "step": 16070
+    },
+    {
+      "epoch": 2.603856124432923,
+      "grad_norm": 1.032846212387085,
+      "learning_rate": 2.2517742837448425e-07,
+      "loss": 0.0589,
+      "step": 16071
+    },
+    {
+      "epoch": 2.6040181464679195,
+      "grad_norm": 0.8250254988670349,
+      "learning_rate": 2.2499608518654432e-07,
+      "loss": 0.0569,
+      "step": 16072
+    },
+    {
+      "epoch": 2.6041801685029164,
+      "grad_norm": 0.9005188345909119,
+      "learning_rate": 2.2481481160821883e-07,
+      "loss": 0.0606,
+      "step": 16073
+    },
+    {
+      "epoch": 2.6043421905379134,
+      "grad_norm": 0.9215274453163147,
+      "learning_rate": 2.2463360764505448e-07,
+      "loss": 0.0597,
+      "step": 16074
+    },
+    {
+      "epoch": 2.60450421257291,
+      "grad_norm": 0.7813541889190674,
+      "learning_rate": 2.244524733025952e-07,
+      "loss": 0.0521,
+      "step": 16075
+    },
+    {
+      "epoch": 2.604666234607907,
+      "grad_norm": 1.026735782623291,
+      "learning_rate": 2.2427140858638424e-07,
+      "loss": 0.0677,
+      "step": 16076
+    },
+    {
+      "epoch": 2.6048282566429033,
+      "grad_norm": 0.8096170425415039,
+      "learning_rate": 2.240904135019603e-07,
+      "loss": 0.0549,
+      "step": 16077
+    },
+    {
+      "epoch": 2.6049902786779002,
+      "grad_norm": 1.0487642288208008,
+      "learning_rate": 2.2390948805486174e-07,
+      "loss": 0.0671,
+      "step": 16078
+    },
+    {
+      "epoch": 2.6051523007128967,
+      "grad_norm": 0.9688891768455505,
+      "learning_rate": 2.2372863225062574e-07,
+      "loss": 0.058,
+      "step": 16079
+    },
+    {
+      "epoch": 2.6053143227478937,
+      "grad_norm": 0.9506645202636719,
+      "learning_rate": 2.2354784609478485e-07,
+      "loss": 0.0602,
+      "step": 16080
+    },
+    {
+      "epoch": 2.6054763447828906,
+      "grad_norm": 0.7836284041404724,
+      "learning_rate": 2.2336712959287077e-07,
+      "loss": 0.0532,
+      "step": 16081
+    },
+    {
+      "epoch": 2.605638366817887,
+      "grad_norm": 0.8392066359519958,
+      "learning_rate": 2.2318648275041267e-07,
+      "loss": 0.0583,
+      "step": 16082
+    },
+    {
+      "epoch": 2.605800388852884,
+      "grad_norm": 0.7897651195526123,
+      "learning_rate": 2.2300590557293944e-07,
+      "loss": 0.0497,
+      "step": 16083
+    },
+    {
+      "epoch": 2.605962410887881,
+      "grad_norm": 0.9836229085922241,
+      "learning_rate": 2.2282539806597476e-07,
+      "loss": 0.0629,
+      "step": 16084
+    },
+    {
+      "epoch": 2.6061244329228774,
+      "grad_norm": 1.0149391889572144,
+      "learning_rate": 2.2264496023504223e-07,
+      "loss": 0.0598,
+      "step": 16085
+    },
+    {
+      "epoch": 2.6062864549578744,
+      "grad_norm": 0.824090301990509,
+      "learning_rate": 2.22464592085663e-07,
+      "loss": 0.055,
+      "step": 16086
+    },
+    {
+      "epoch": 2.6064484769928713,
+      "grad_norm": 0.874244213104248,
+      "learning_rate": 2.2228429362335546e-07,
+      "loss": 0.0538,
+      "step": 16087
+    },
+    {
+      "epoch": 2.606610499027868,
+      "grad_norm": 0.8717989325523376,
+      "learning_rate": 2.2210406485363656e-07,
+      "loss": 0.063,
+      "step": 16088
+    },
+    {
+      "epoch": 2.6067725210628643,
+      "grad_norm": 1.0721566677093506,
+      "learning_rate": 2.2192390578202105e-07,
+      "loss": 0.0644,
+      "step": 16089
+    },
+    {
+      "epoch": 2.606934543097861,
+      "grad_norm": 0.8851465582847595,
+      "learning_rate": 2.217438164140212e-07,
+      "loss": 0.0536,
+      "step": 16090
+    },
+    {
+      "epoch": 2.607096565132858,
+      "grad_norm": 1.2500100135803223,
+      "learning_rate": 2.2156379675514762e-07,
+      "loss": 0.0662,
+      "step": 16091
+    },
+    {
+      "epoch": 2.6072585871678546,
+      "grad_norm": 0.930827260017395,
+      "learning_rate": 2.213838468109075e-07,
+      "loss": 0.0613,
+      "step": 16092
+    },
+    {
+      "epoch": 2.6074206092028516,
+      "grad_norm": 0.7674300074577332,
+      "learning_rate": 2.2120396658680765e-07,
+      "loss": 0.0532,
+      "step": 16093
+    },
+    {
+      "epoch": 2.6075826312378485,
+      "grad_norm": 0.9664437174797058,
+      "learning_rate": 2.210241560883525e-07,
+      "loss": 0.0676,
+      "step": 16094
+    },
+    {
+      "epoch": 2.607744653272845,
+      "grad_norm": 0.9895860552787781,
+      "learning_rate": 2.2084441532104262e-07,
+      "loss": 0.0644,
+      "step": 16095
+    },
+    {
+      "epoch": 2.607906675307842,
+      "grad_norm": 0.9111254811286926,
+      "learning_rate": 2.206647442903781e-07,
+      "loss": 0.0614,
+      "step": 16096
+    },
+    {
+      "epoch": 2.608068697342839,
+      "grad_norm": 1.104712724685669,
+      "learning_rate": 2.204851430018562e-07,
+      "loss": 0.069,
+      "step": 16097
+    },
+    {
+      "epoch": 2.6082307193778353,
+      "grad_norm": 1.0142390727996826,
+      "learning_rate": 2.2030561146097363e-07,
+      "loss": 0.0548,
+      "step": 16098
+    },
+    {
+      "epoch": 2.6083927414128323,
+      "grad_norm": 0.9152945876121521,
+      "learning_rate": 2.2012614967322182e-07,
+      "loss": 0.0608,
+      "step": 16099
+    },
+    {
+      "epoch": 2.6085547634478288,
+      "grad_norm": 0.9016642570495605,
+      "learning_rate": 2.199467576440928e-07,
+      "loss": 0.0629,
+      "step": 16100
+    },
+    {
+      "epoch": 2.6087167854828257,
+      "grad_norm": 0.7842551469802856,
+      "learning_rate": 2.1976743537907546e-07,
+      "loss": 0.055,
+      "step": 16101
+    },
+    {
+      "epoch": 2.608878807517822,
+      "grad_norm": 0.8693322539329529,
+      "learning_rate": 2.195881828836563e-07,
+      "loss": 0.0594,
+      "step": 16102
+    },
+    {
+      "epoch": 2.609040829552819,
+      "grad_norm": 1.128956913948059,
+      "learning_rate": 2.194090001633206e-07,
+      "loss": 0.0671,
+      "step": 16103
+    },
+    {
+      "epoch": 2.609202851587816,
+      "grad_norm": 0.8959473371505737,
+      "learning_rate": 2.1922988722355044e-07,
+      "loss": 0.0583,
+      "step": 16104
+    },
+    {
+      "epoch": 2.6093648736228126,
+      "grad_norm": 0.9709823131561279,
+      "learning_rate": 2.1905084406982663e-07,
+      "loss": 0.0631,
+      "step": 16105
+    },
+    {
+      "epoch": 2.6095268956578095,
+      "grad_norm": 1.0241434574127197,
+      "learning_rate": 2.188718707076265e-07,
+      "loss": 0.0605,
+      "step": 16106
+    },
+    {
+      "epoch": 2.6096889176928064,
+      "grad_norm": 1.0129493474960327,
+      "learning_rate": 2.1869296714242732e-07,
+      "loss": 0.0609,
+      "step": 16107
+    },
+    {
+      "epoch": 2.609850939727803,
+      "grad_norm": 0.9752632975578308,
+      "learning_rate": 2.185141333797025e-07,
+      "loss": 0.0576,
+      "step": 16108
+    },
+    {
+      "epoch": 2.6100129617628,
+      "grad_norm": 0.8120836019515991,
+      "learning_rate": 2.183353694249249e-07,
+      "loss": 0.048,
+      "step": 16109
+    },
+    {
+      "epoch": 2.6101749837977968,
+      "grad_norm": 0.946182370185852,
+      "learning_rate": 2.181566752835626e-07,
+      "loss": 0.0621,
+      "step": 16110
+    },
+    {
+      "epoch": 2.6103370058327933,
+      "grad_norm": 0.913953959941864,
+      "learning_rate": 2.1797805096108405e-07,
+      "loss": 0.0609,
+      "step": 16111
+    },
+    {
+      "epoch": 2.6104990278677898,
+      "grad_norm": 0.9735072255134583,
+      "learning_rate": 2.177994964629554e-07,
+      "loss": 0.0589,
+      "step": 16112
+    },
+    {
+      "epoch": 2.6106610499027867,
+      "grad_norm": 0.8608875274658203,
+      "learning_rate": 2.1762101179463896e-07,
+      "loss": 0.064,
+      "step": 16113
+    },
+    {
+      "epoch": 2.6108230719377836,
+      "grad_norm": 0.9533604979515076,
+      "learning_rate": 2.174425969615962e-07,
+      "loss": 0.0635,
+      "step": 16114
+    },
+    {
+      "epoch": 2.61098509397278,
+      "grad_norm": 0.88776695728302,
+      "learning_rate": 2.1726425196928663e-07,
+      "loss": 0.0543,
+      "step": 16115
+    },
+    {
+      "epoch": 2.611147116007777,
+      "grad_norm": 1.0538303852081299,
+      "learning_rate": 2.1708597682316645e-07,
+      "loss": 0.0626,
+      "step": 16116
+    },
+    {
+      "epoch": 2.611309138042774,
+      "grad_norm": 0.8502702713012695,
+      "learning_rate": 2.1690777152869103e-07,
+      "loss": 0.0554,
+      "step": 16117
+    },
+    {
+      "epoch": 2.6114711600777705,
+      "grad_norm": 0.9557256102561951,
+      "learning_rate": 2.1672963609131292e-07,
+      "loss": 0.0582,
+      "step": 16118
+    },
+    {
+      "epoch": 2.6116331821127674,
+      "grad_norm": 1.0968202352523804,
+      "learning_rate": 2.1655157051648223e-07,
+      "loss": 0.0612,
+      "step": 16119
+    },
+    {
+      "epoch": 2.6117952041477643,
+      "grad_norm": 0.9737009406089783,
+      "learning_rate": 2.1637357480964821e-07,
+      "loss": 0.0624,
+      "step": 16120
+    },
+    {
+      "epoch": 2.611957226182761,
+      "grad_norm": 0.9180959463119507,
+      "learning_rate": 2.1619564897625566e-07,
+      "loss": 0.0625,
+      "step": 16121
+    },
+    {
+      "epoch": 2.6121192482177578,
+      "grad_norm": 0.9389775395393372,
+      "learning_rate": 2.1601779302175026e-07,
+      "loss": 0.0632,
+      "step": 16122
+    },
+    {
+      "epoch": 2.6122812702527543,
+      "grad_norm": 0.9112357497215271,
+      "learning_rate": 2.158400069515734e-07,
+      "loss": 0.0631,
+      "step": 16123
+    },
+    {
+      "epoch": 2.612443292287751,
+      "grad_norm": 0.9230049848556519,
+      "learning_rate": 2.1566229077116445e-07,
+      "loss": 0.058,
+      "step": 16124
+    },
+    {
+      "epoch": 2.6126053143227477,
+      "grad_norm": 0.9439942240715027,
+      "learning_rate": 2.1548464448596123e-07,
+      "loss": 0.0643,
+      "step": 16125
+    },
+    {
+      "epoch": 2.6127673363577446,
+      "grad_norm": 0.8843783140182495,
+      "learning_rate": 2.1530706810139913e-07,
+      "loss": 0.064,
+      "step": 16126
+    },
+    {
+      "epoch": 2.6129293583927415,
+      "grad_norm": 0.980243980884552,
+      "learning_rate": 2.1512956162291294e-07,
+      "loss": 0.0633,
+      "step": 16127
+    },
+    {
+      "epoch": 2.613091380427738,
+      "grad_norm": 0.9535460472106934,
+      "learning_rate": 2.1495212505593221e-07,
+      "loss": 0.0688,
+      "step": 16128
+    },
+    {
+      "epoch": 2.613253402462735,
+      "grad_norm": 0.9649513959884644,
+      "learning_rate": 2.147747584058868e-07,
+      "loss": 0.0587,
+      "step": 16129
+    },
+    {
+      "epoch": 2.613415424497732,
+      "grad_norm": 0.9310510158538818,
+      "learning_rate": 2.1459746167820372e-07,
+      "loss": 0.0603,
+      "step": 16130
+    },
+    {
+      "epoch": 2.6135774465327284,
+      "grad_norm": 1.098154902458191,
+      "learning_rate": 2.1442023487830782e-07,
+      "loss": 0.0632,
+      "step": 16131
+    },
+    {
+      "epoch": 2.6137394685677253,
+      "grad_norm": 0.9631187915802002,
+      "learning_rate": 2.142430780116214e-07,
+      "loss": 0.0624,
+      "step": 16132
+    },
+    {
+      "epoch": 2.613901490602722,
+      "grad_norm": 1.261637568473816,
+      "learning_rate": 2.1406599108356573e-07,
+      "loss": 0.067,
+      "step": 16133
+    },
+    {
+      "epoch": 2.6140635126377187,
+      "grad_norm": 0.8249843120574951,
+      "learning_rate": 2.1388897409955867e-07,
+      "loss": 0.0528,
+      "step": 16134
+    },
+    {
+      "epoch": 2.6142255346727152,
+      "grad_norm": 0.8405267596244812,
+      "learning_rate": 2.1371202706501697e-07,
+      "loss": 0.0559,
+      "step": 16135
+    },
+    {
+      "epoch": 2.614387556707712,
+      "grad_norm": 0.8323154449462891,
+      "learning_rate": 2.1353514998535414e-07,
+      "loss": 0.0581,
+      "step": 16136
+    },
+    {
+      "epoch": 2.614549578742709,
+      "grad_norm": 0.9908446669578552,
+      "learning_rate": 2.13358342865983e-07,
+      "loss": 0.0634,
+      "step": 16137
+    },
+    {
+      "epoch": 2.6147116007777056,
+      "grad_norm": 0.9334249496459961,
+      "learning_rate": 2.1318160571231316e-07,
+      "loss": 0.0594,
+      "step": 16138
+    },
+    {
+      "epoch": 2.6148736228127025,
+      "grad_norm": 0.9564669728279114,
+      "learning_rate": 2.1300493852975167e-07,
+      "loss": 0.0613,
+      "step": 16139
+    },
+    {
+      "epoch": 2.6150356448476995,
+      "grad_norm": 1.0061742067337036,
+      "learning_rate": 2.128283413237045e-07,
+      "loss": 0.0643,
+      "step": 16140
+    },
+    {
+      "epoch": 2.615197666882696,
+      "grad_norm": 0.8926098942756653,
+      "learning_rate": 2.1265181409957537e-07,
+      "loss": 0.0605,
+      "step": 16141
+    },
+    {
+      "epoch": 2.615359688917693,
+      "grad_norm": 0.8778125047683716,
+      "learning_rate": 2.1247535686276632e-07,
+      "loss": 0.0568,
+      "step": 16142
+    },
+    {
+      "epoch": 2.61552171095269,
+      "grad_norm": 0.8341763615608215,
+      "learning_rate": 2.1229896961867475e-07,
+      "loss": 0.0596,
+      "step": 16143
+    },
+    {
+      "epoch": 2.6156837329876863,
+      "grad_norm": 1.0330524444580078,
+      "learning_rate": 2.121226523726988e-07,
+      "loss": 0.07,
+      "step": 16144
+    },
+    {
+      "epoch": 2.6158457550226832,
+      "grad_norm": 0.8376036286354065,
+      "learning_rate": 2.1194640513023306e-07,
+      "loss": 0.0545,
+      "step": 16145
+    },
+    {
+      "epoch": 2.6160077770576797,
+      "grad_norm": 0.8829439282417297,
+      "learning_rate": 2.1177022789667045e-07,
+      "loss": 0.059,
+      "step": 16146
+    },
+    {
+      "epoch": 2.6161697990926767,
+      "grad_norm": 0.9374154806137085,
+      "learning_rate": 2.1159412067740136e-07,
+      "loss": 0.0578,
+      "step": 16147
+    },
+    {
+      "epoch": 2.616331821127673,
+      "grad_norm": 0.9516973495483398,
+      "learning_rate": 2.1141808347781428e-07,
+      "loss": 0.0641,
+      "step": 16148
+    },
+    {
+      "epoch": 2.61649384316267,
+      "grad_norm": 0.885984480381012,
+      "learning_rate": 2.1124211630329571e-07,
+      "loss": 0.057,
+      "step": 16149
+    },
+    {
+      "epoch": 2.616655865197667,
+      "grad_norm": 0.9362042546272278,
+      "learning_rate": 2.110662191592297e-07,
+      "loss": 0.0587,
+      "step": 16150
+    },
+    {
+      "epoch": 2.6168178872326635,
+      "grad_norm": 1.1420210599899292,
+      "learning_rate": 2.1089039205099832e-07,
+      "loss": 0.0599,
+      "step": 16151
+    },
+    {
+      "epoch": 2.6169799092676604,
+      "grad_norm": 0.8978714346885681,
+      "learning_rate": 2.1071463498398114e-07,
+      "loss": 0.0628,
+      "step": 16152
+    },
+    {
+      "epoch": 2.6171419313026574,
+      "grad_norm": 0.8524075746536255,
+      "learning_rate": 2.1053894796355694e-07,
+      "loss": 0.0582,
+      "step": 16153
+    },
+    {
+      "epoch": 2.617303953337654,
+      "grad_norm": 0.8688123822212219,
+      "learning_rate": 2.103633309950995e-07,
+      "loss": 0.0616,
+      "step": 16154
+    },
+    {
+      "epoch": 2.617465975372651,
+      "grad_norm": 0.8266528844833374,
+      "learning_rate": 2.101877840839836e-07,
+      "loss": 0.0588,
+      "step": 16155
+    },
+    {
+      "epoch": 2.6176279974076473,
+      "grad_norm": 0.9983370900154114,
+      "learning_rate": 2.1001230723558087e-07,
+      "loss": 0.0625,
+      "step": 16156
+    },
+    {
+      "epoch": 2.6177900194426442,
+      "grad_norm": 0.835728645324707,
+      "learning_rate": 2.0983690045525944e-07,
+      "loss": 0.0593,
+      "step": 16157
+    },
+    {
+      "epoch": 2.6179520414776407,
+      "grad_norm": 0.8325420618057251,
+      "learning_rate": 2.0966156374838677e-07,
+      "loss": 0.0582,
+      "step": 16158
+    },
+    {
+      "epoch": 2.6181140635126376,
+      "grad_norm": 0.8607774376869202,
+      "learning_rate": 2.0948629712032738e-07,
+      "loss": 0.0552,
+      "step": 16159
+    },
+    {
+      "epoch": 2.6182760855476346,
+      "grad_norm": 1.0366528034210205,
+      "learning_rate": 2.0931110057644505e-07,
+      "loss": 0.0654,
+      "step": 16160
+    },
+    {
+      "epoch": 2.618438107582631,
+      "grad_norm": 1.09884512424469,
+      "learning_rate": 2.0913597412209941e-07,
+      "loss": 0.0653,
+      "step": 16161
+    },
+    {
+      "epoch": 2.618600129617628,
+      "grad_norm": 1.109409213066101,
+      "learning_rate": 2.089609177626492e-07,
+      "loss": 0.0675,
+      "step": 16162
+    },
+    {
+      "epoch": 2.618762151652625,
+      "grad_norm": 0.9305920004844666,
+      "learning_rate": 2.0878593150345043e-07,
+      "loss": 0.057,
+      "step": 16163
+    },
+    {
+      "epoch": 2.6189241736876214,
+      "grad_norm": 0.9153008460998535,
+      "learning_rate": 2.0861101534985774e-07,
+      "loss": 0.0565,
+      "step": 16164
+    },
+    {
+      "epoch": 2.6190861957226184,
+      "grad_norm": 0.9763879179954529,
+      "learning_rate": 2.0843616930722288e-07,
+      "loss": 0.0676,
+      "step": 16165
+    },
+    {
+      "epoch": 2.6192482177576153,
+      "grad_norm": 0.9485346078872681,
+      "learning_rate": 2.082613933808958e-07,
+      "loss": 0.0615,
+      "step": 16166
+    },
+    {
+      "epoch": 2.619410239792612,
+      "grad_norm": 0.7745640277862549,
+      "learning_rate": 2.0808668757622413e-07,
+      "loss": 0.0542,
+      "step": 16167
+    },
+    {
+      "epoch": 2.6195722618276087,
+      "grad_norm": 0.9337866902351379,
+      "learning_rate": 2.079120518985539e-07,
+      "loss": 0.0631,
+      "step": 16168
+    },
+    {
+      "epoch": 2.619734283862605,
+      "grad_norm": 0.7666480541229248,
+      "learning_rate": 2.077374863532275e-07,
+      "loss": 0.0533,
+      "step": 16169
+    },
+    {
+      "epoch": 2.619896305897602,
+      "grad_norm": 0.8683062791824341,
+      "learning_rate": 2.07562990945587e-07,
+      "loss": 0.0573,
+      "step": 16170
+    },
+    {
+      "epoch": 2.6200583279325986,
+      "grad_norm": 0.9523324370384216,
+      "learning_rate": 2.073885656809718e-07,
+      "loss": 0.061,
+      "step": 16171
+    },
+    {
+      "epoch": 2.6202203499675956,
+      "grad_norm": 0.855353832244873,
+      "learning_rate": 2.0721421056471818e-07,
+      "loss": 0.0556,
+      "step": 16172
+    },
+    {
+      "epoch": 2.6203823720025925,
+      "grad_norm": 0.924647331237793,
+      "learning_rate": 2.0703992560216075e-07,
+      "loss": 0.0608,
+      "step": 16173
+    },
+    {
+      "epoch": 2.620544394037589,
+      "grad_norm": 0.9063007831573486,
+      "learning_rate": 2.0686571079863383e-07,
+      "loss": 0.0636,
+      "step": 16174
+    },
+    {
+      "epoch": 2.620706416072586,
+      "grad_norm": 0.9852257370948792,
+      "learning_rate": 2.0669156615946623e-07,
+      "loss": 0.0637,
+      "step": 16175
+    },
+    {
+      "epoch": 2.620868438107583,
+      "grad_norm": 0.8728942275047302,
+      "learning_rate": 2.0651749168998703e-07,
+      "loss": 0.0603,
+      "step": 16176
+    },
+    {
+      "epoch": 2.6210304601425793,
+      "grad_norm": 0.9853529334068298,
+      "learning_rate": 2.0634348739552251e-07,
+      "loss": 0.065,
+      "step": 16177
+    },
+    {
+      "epoch": 2.6211924821775763,
+      "grad_norm": 0.9242932200431824,
+      "learning_rate": 2.0616955328139675e-07,
+      "loss": 0.0658,
+      "step": 16178
+    },
+    {
+      "epoch": 2.6213545042125728,
+      "grad_norm": 0.8449025750160217,
+      "learning_rate": 2.059956893529319e-07,
+      "loss": 0.0517,
+      "step": 16179
+    },
+    {
+      "epoch": 2.6215165262475697,
+      "grad_norm": 0.9864807724952698,
+      "learning_rate": 2.058218956154473e-07,
+      "loss": 0.063,
+      "step": 16180
+    },
+    {
+      "epoch": 2.621678548282566,
+      "grad_norm": 1.0352917909622192,
+      "learning_rate": 2.0564817207426092e-07,
+      "loss": 0.0568,
+      "step": 16181
+    },
+    {
+      "epoch": 2.621840570317563,
+      "grad_norm": 1.0253640413284302,
+      "learning_rate": 2.0547451873468877e-07,
+      "loss": 0.0694,
+      "step": 16182
+    },
+    {
+      "epoch": 2.62200259235256,
+      "grad_norm": 0.8204119205474854,
+      "learning_rate": 2.0530093560204272e-07,
+      "loss": 0.0532,
+      "step": 16183
+    },
+    {
+      "epoch": 2.6221646143875565,
+      "grad_norm": 0.9108116030693054,
+      "learning_rate": 2.051274226816355e-07,
+      "loss": 0.0631,
+      "step": 16184
+    },
+    {
+      "epoch": 2.6223266364225535,
+      "grad_norm": 0.8796073198318481,
+      "learning_rate": 2.0495397997877558e-07,
+      "loss": 0.0594,
+      "step": 16185
+    },
+    {
+      "epoch": 2.6224886584575504,
+      "grad_norm": 0.931476891040802,
+      "learning_rate": 2.0478060749877044e-07,
+      "loss": 0.0647,
+      "step": 16186
+    },
+    {
+      "epoch": 2.622650680492547,
+      "grad_norm": 0.9860643148422241,
+      "learning_rate": 2.0460730524692384e-07,
+      "loss": 0.061,
+      "step": 16187
+    },
+    {
+      "epoch": 2.622812702527544,
+      "grad_norm": 0.9249363541603088,
+      "learning_rate": 2.0443407322853882e-07,
+      "loss": 0.058,
+      "step": 16188
+    },
+    {
+      "epoch": 2.6229747245625408,
+      "grad_norm": 0.9313977956771851,
+      "learning_rate": 2.0426091144891664e-07,
+      "loss": 0.0585,
+      "step": 16189
+    },
+    {
+      "epoch": 2.6231367465975373,
+      "grad_norm": 0.8045008182525635,
+      "learning_rate": 2.0408781991335446e-07,
+      "loss": 0.0542,
+      "step": 16190
+    },
+    {
+      "epoch": 2.6232987686325338,
+      "grad_norm": 0.8684973120689392,
+      "learning_rate": 2.039147986271492e-07,
+      "loss": 0.0604,
+      "step": 16191
+    },
+    {
+      "epoch": 2.6234607906675307,
+      "grad_norm": 0.8771528005599976,
+      "learning_rate": 2.0374184759559463e-07,
+      "loss": 0.053,
+      "step": 16192
+    },
+    {
+      "epoch": 2.6236228127025276,
+      "grad_norm": 0.9225574731826782,
+      "learning_rate": 2.0356896682398264e-07,
+      "loss": 0.0605,
+      "step": 16193
+    },
+    {
+      "epoch": 2.623784834737524,
+      "grad_norm": 0.8193483948707581,
+      "learning_rate": 2.033961563176029e-07,
+      "loss": 0.0596,
+      "step": 16194
+    },
+    {
+      "epoch": 2.623946856772521,
+      "grad_norm": 1.0594878196716309,
+      "learning_rate": 2.0322341608174338e-07,
+      "loss": 0.0619,
+      "step": 16195
+    },
+    {
+      "epoch": 2.624108878807518,
+      "grad_norm": 0.7767667770385742,
+      "learning_rate": 2.0305074612168906e-07,
+      "loss": 0.0532,
+      "step": 16196
+    },
+    {
+      "epoch": 2.6242709008425145,
+      "grad_norm": 0.8942044377326965,
+      "learning_rate": 2.0287814644272347e-07,
+      "loss": 0.0574,
+      "step": 16197
+    },
+    {
+      "epoch": 2.6244329228775114,
+      "grad_norm": 0.8722692728042603,
+      "learning_rate": 2.0270561705012765e-07,
+      "loss": 0.0565,
+      "step": 16198
+    },
+    {
+      "epoch": 2.6245949449125083,
+      "grad_norm": 0.9796530604362488,
+      "learning_rate": 2.0253315794918043e-07,
+      "loss": 0.0698,
+      "step": 16199
+    },
+    {
+      "epoch": 2.624756966947505,
+      "grad_norm": 0.9763656854629517,
+      "learning_rate": 2.0236076914515956e-07,
+      "loss": 0.0721,
+      "step": 16200
+    },
+    {
+      "epoch": 2.6249189889825018,
+      "grad_norm": 1.099333643913269,
+      "learning_rate": 2.021884506433383e-07,
+      "loss": 0.0629,
+      "step": 16201
+    },
+    {
+      "epoch": 2.6250810110174982,
+      "grad_norm": 0.8187487721443176,
+      "learning_rate": 2.020162024489894e-07,
+      "loss": 0.059,
+      "step": 16202
+    },
+    {
+      "epoch": 2.625243033052495,
+      "grad_norm": 0.874085009098053,
+      "learning_rate": 2.0184402456738444e-07,
+      "loss": 0.0601,
+      "step": 16203
+    },
+    {
+      "epoch": 2.6254050550874917,
+      "grad_norm": 0.9488238096237183,
+      "learning_rate": 2.0167191700379092e-07,
+      "loss": 0.0593,
+      "step": 16204
+    },
+    {
+      "epoch": 2.6255670771224886,
+      "grad_norm": 1.0066373348236084,
+      "learning_rate": 2.0149987976347485e-07,
+      "loss": 0.0596,
+      "step": 16205
+    },
+    {
+      "epoch": 2.6257290991574855,
+      "grad_norm": 0.8716149926185608,
+      "learning_rate": 2.0132791285169985e-07,
+      "loss": 0.0615,
+      "step": 16206
+    },
+    {
+      "epoch": 2.625891121192482,
+      "grad_norm": 0.8857298493385315,
+      "learning_rate": 2.0115601627372832e-07,
+      "loss": 0.0551,
+      "step": 16207
+    },
+    {
+      "epoch": 2.626053143227479,
+      "grad_norm": 0.8469251394271851,
+      "learning_rate": 2.0098419003481946e-07,
+      "loss": 0.0573,
+      "step": 16208
+    },
+    {
+      "epoch": 2.626215165262476,
+      "grad_norm": 0.8539047837257385,
+      "learning_rate": 2.0081243414023067e-07,
+      "loss": 0.0561,
+      "step": 16209
+    },
+    {
+      "epoch": 2.6263771872974724,
+      "grad_norm": 0.9203428626060486,
+      "learning_rate": 2.0064074859521777e-07,
+      "loss": 0.0624,
+      "step": 16210
+    },
+    {
+      "epoch": 2.6265392093324693,
+      "grad_norm": 0.9050666093826294,
+      "learning_rate": 2.004691334050335e-07,
+      "loss": 0.0579,
+      "step": 16211
+    },
+    {
+      "epoch": 2.6267012313674662,
+      "grad_norm": 0.9167254567146301,
+      "learning_rate": 2.0029758857492893e-07,
+      "loss": 0.0663,
+      "step": 16212
+    },
+    {
+      "epoch": 2.6268632534024627,
+      "grad_norm": 0.9239572882652283,
+      "learning_rate": 2.001261141101532e-07,
+      "loss": 0.0646,
+      "step": 16213
+    },
+    {
+      "epoch": 2.6270252754374592,
+      "grad_norm": 0.7515432238578796,
+      "learning_rate": 1.9995471001595267e-07,
+      "loss": 0.0568,
+      "step": 16214
+    },
+    {
+      "epoch": 2.627187297472456,
+      "grad_norm": 0.834296703338623,
+      "learning_rate": 1.9978337629757233e-07,
+      "loss": 0.0564,
+      "step": 16215
+    },
+    {
+      "epoch": 2.627349319507453,
+      "grad_norm": 0.78092360496521,
+      "learning_rate": 1.9961211296025352e-07,
+      "loss": 0.0565,
+      "step": 16216
+    },
+    {
+      "epoch": 2.6275113415424496,
+      "grad_norm": 0.9915261268615723,
+      "learning_rate": 1.994409200092376e-07,
+      "loss": 0.0632,
+      "step": 16217
+    },
+    {
+      "epoch": 2.6276733635774465,
+      "grad_norm": 1.0370954275131226,
+      "learning_rate": 1.992697974497629e-07,
+      "loss": 0.066,
+      "step": 16218
+    },
+    {
+      "epoch": 2.6278353856124435,
+      "grad_norm": 0.905182957649231,
+      "learning_rate": 1.9909874528706407e-07,
+      "loss": 0.0663,
+      "step": 16219
+    },
+    {
+      "epoch": 2.62799740764744,
+      "grad_norm": 0.8637315034866333,
+      "learning_rate": 1.989277635263756e-07,
+      "loss": 0.0556,
+      "step": 16220
+    },
+    {
+      "epoch": 2.628159429682437,
+      "grad_norm": 0.8456125855445862,
+      "learning_rate": 1.9875685217292856e-07,
+      "loss": 0.0563,
+      "step": 16221
+    },
+    {
+      "epoch": 2.628321451717434,
+      "grad_norm": 0.8584468960762024,
+      "learning_rate": 1.9858601123195403e-07,
+      "loss": 0.0578,
+      "step": 16222
+    },
+    {
+      "epoch": 2.6284834737524303,
+      "grad_norm": 0.9558644890785217,
+      "learning_rate": 1.9841524070867784e-07,
+      "loss": 0.0656,
+      "step": 16223
+    },
+    {
+      "epoch": 2.6286454957874272,
+      "grad_norm": 1.1154943704605103,
+      "learning_rate": 1.9824454060832526e-07,
+      "loss": 0.0623,
+      "step": 16224
+    },
+    {
+      "epoch": 2.6288075178224237,
+      "grad_norm": 0.9311351776123047,
+      "learning_rate": 1.980739109361199e-07,
+      "loss": 0.0581,
+      "step": 16225
+    },
+    {
+      "epoch": 2.6289695398574207,
+      "grad_norm": 1.0172123908996582,
+      "learning_rate": 1.9790335169728197e-07,
+      "loss": 0.0644,
+      "step": 16226
+    },
+    {
+      "epoch": 2.629131561892417,
+      "grad_norm": 1.0837866067886353,
+      "learning_rate": 1.977328628970307e-07,
+      "loss": 0.0607,
+      "step": 16227
+    },
+    {
+      "epoch": 2.629293583927414,
+      "grad_norm": 0.8728600144386292,
+      "learning_rate": 1.9756244454058244e-07,
+      "loss": 0.06,
+      "step": 16228
+    },
+    {
+      "epoch": 2.629455605962411,
+      "grad_norm": 0.9363974332809448,
+      "learning_rate": 1.9739209663315162e-07,
+      "loss": 0.0597,
+      "step": 16229
+    },
+    {
+      "epoch": 2.6296176279974075,
+      "grad_norm": 0.9871296882629395,
+      "learning_rate": 1.9722181917995103e-07,
+      "loss": 0.0646,
+      "step": 16230
+    },
+    {
+      "epoch": 2.6297796500324044,
+      "grad_norm": 0.8740096092224121,
+      "learning_rate": 1.9705161218618902e-07,
+      "loss": 0.0625,
+      "step": 16231
+    },
+    {
+      "epoch": 2.6299416720674014,
+      "grad_norm": 1.0138083696365356,
+      "learning_rate": 1.9688147565707528e-07,
+      "loss": 0.0668,
+      "step": 16232
+    },
+    {
+      "epoch": 2.630103694102398,
+      "grad_norm": 0.7797446846961975,
+      "learning_rate": 1.967114095978151e-07,
+      "loss": 0.0515,
+      "step": 16233
+    },
+    {
+      "epoch": 2.630265716137395,
+      "grad_norm": 0.9009860754013062,
+      "learning_rate": 1.9654141401361183e-07,
+      "loss": 0.0583,
+      "step": 16234
+    },
+    {
+      "epoch": 2.6304277381723913,
+      "grad_norm": 0.9890360236167908,
+      "learning_rate": 1.9637148890966685e-07,
+      "loss": 0.0623,
+      "step": 16235
+    },
+    {
+      "epoch": 2.630589760207388,
+      "grad_norm": 0.7815765738487244,
+      "learning_rate": 1.9620163429117906e-07,
+      "loss": 0.0524,
+      "step": 16236
+    },
+    {
+      "epoch": 2.6307517822423847,
+      "grad_norm": 0.9130102396011353,
+      "learning_rate": 1.9603185016334737e-07,
+      "loss": 0.0572,
+      "step": 16237
+    },
+    {
+      "epoch": 2.6309138042773816,
+      "grad_norm": 0.9095372557640076,
+      "learning_rate": 1.958621365313648e-07,
+      "loss": 0.0609,
+      "step": 16238
+    },
+    {
+      "epoch": 2.6310758263123786,
+      "grad_norm": 0.7958774566650391,
+      "learning_rate": 1.9569249340042534e-07,
+      "loss": 0.0591,
+      "step": 16239
+    },
+    {
+      "epoch": 2.631237848347375,
+      "grad_norm": 0.8964948654174805,
+      "learning_rate": 1.9552292077571894e-07,
+      "loss": 0.0678,
+      "step": 16240
+    },
+    {
+      "epoch": 2.631399870382372,
+      "grad_norm": 0.9855856895446777,
+      "learning_rate": 1.953534186624345e-07,
+      "loss": 0.065,
+      "step": 16241
+    },
+    {
+      "epoch": 2.631561892417369,
+      "grad_norm": 1.1027717590332031,
+      "learning_rate": 1.9518398706575846e-07,
+      "loss": 0.064,
+      "step": 16242
+    },
+    {
+      "epoch": 2.6317239144523654,
+      "grad_norm": 0.8115701079368591,
+      "learning_rate": 1.9501462599087472e-07,
+      "loss": 0.0496,
+      "step": 16243
+    },
+    {
+      "epoch": 2.6318859364873624,
+      "grad_norm": 0.9368529915809631,
+      "learning_rate": 1.948453354429661e-07,
+      "loss": 0.0613,
+      "step": 16244
+    },
+    {
+      "epoch": 2.6320479585223593,
+      "grad_norm": 0.8521546721458435,
+      "learning_rate": 1.946761154272106e-07,
+      "loss": 0.059,
+      "step": 16245
+    },
+    {
+      "epoch": 2.6322099805573558,
+      "grad_norm": 0.9154667258262634,
+      "learning_rate": 1.9450696594878804e-07,
+      "loss": 0.0632,
+      "step": 16246
+    },
+    {
+      "epoch": 2.6323720025923527,
+      "grad_norm": 0.8996468782424927,
+      "learning_rate": 1.9433788701287288e-07,
+      "loss": 0.0639,
+      "step": 16247
+    },
+    {
+      "epoch": 2.632534024627349,
+      "grad_norm": 0.8377342820167542,
+      "learning_rate": 1.941688786246393e-07,
+      "loss": 0.0593,
+      "step": 16248
+    },
+    {
+      "epoch": 2.632696046662346,
+      "grad_norm": 0.7824660539627075,
+      "learning_rate": 1.939999407892576e-07,
+      "loss": 0.052,
+      "step": 16249
+    },
+    {
+      "epoch": 2.6328580686973426,
+      "grad_norm": 0.8129716515541077,
+      "learning_rate": 1.9383107351189672e-07,
+      "loss": 0.0569,
+      "step": 16250
+    },
+    {
+      "epoch": 2.6330200907323396,
+      "grad_norm": 0.8622326254844666,
+      "learning_rate": 1.936622767977253e-07,
+      "loss": 0.0588,
+      "step": 16251
+    },
+    {
+      "epoch": 2.6331821127673365,
+      "grad_norm": 0.9913507103919983,
+      "learning_rate": 1.9349355065190618e-07,
+      "loss": 0.0662,
+      "step": 16252
+    },
+    {
+      "epoch": 2.633344134802333,
+      "grad_norm": 0.8238366842269897,
+      "learning_rate": 1.9332489507960324e-07,
+      "loss": 0.0566,
+      "step": 16253
+    },
+    {
+      "epoch": 2.63350615683733,
+      "grad_norm": 0.9328113794326782,
+      "learning_rate": 1.9315631008597596e-07,
+      "loss": 0.0648,
+      "step": 16254
+    },
+    {
+      "epoch": 2.633668178872327,
+      "grad_norm": 0.9620225429534912,
+      "learning_rate": 1.9298779567618357e-07,
+      "loss": 0.0593,
+      "step": 16255
+    },
+    {
+      "epoch": 2.6338302009073233,
+      "grad_norm": 0.7656645774841309,
+      "learning_rate": 1.9281935185538141e-07,
+      "loss": 0.0511,
+      "step": 16256
+    },
+    {
+      "epoch": 2.6339922229423203,
+      "grad_norm": 0.8839242458343506,
+      "learning_rate": 1.9265097862872423e-07,
+      "loss": 0.0611,
+      "step": 16257
+    },
+    {
+      "epoch": 2.6341542449773168,
+      "grad_norm": 0.8976559638977051,
+      "learning_rate": 1.9248267600136317e-07,
+      "loss": 0.0574,
+      "step": 16258
+    },
+    {
+      "epoch": 2.6343162670123137,
+      "grad_norm": 1.1227093935012817,
+      "learning_rate": 1.9231444397844855e-07,
+      "loss": 0.0671,
+      "step": 16259
+    },
+    {
+      "epoch": 2.63447828904731,
+      "grad_norm": 0.8992955088615417,
+      "learning_rate": 1.9214628256512656e-07,
+      "loss": 0.0623,
+      "step": 16260
+    },
+    {
+      "epoch": 2.634640311082307,
+      "grad_norm": 1.0141760110855103,
+      "learning_rate": 1.919781917665439e-07,
+      "loss": 0.0653,
+      "step": 16261
+    },
+    {
+      "epoch": 2.634802333117304,
+      "grad_norm": 1.1703956127166748,
+      "learning_rate": 1.918101715878437e-07,
+      "loss": 0.0751,
+      "step": 16262
+    },
+    {
+      "epoch": 2.6349643551523005,
+      "grad_norm": 0.9227219820022583,
+      "learning_rate": 1.9164222203416627e-07,
+      "loss": 0.0644,
+      "step": 16263
+    },
+    {
+      "epoch": 2.6351263771872975,
+      "grad_norm": 0.804478645324707,
+      "learning_rate": 1.9147434311065028e-07,
+      "loss": 0.0484,
+      "step": 16264
+    },
+    {
+      "epoch": 2.6352883992222944,
+      "grad_norm": 0.8943800926208496,
+      "learning_rate": 1.91306534822433e-07,
+      "loss": 0.0589,
+      "step": 16265
+    },
+    {
+      "epoch": 2.635450421257291,
+      "grad_norm": 0.9125702977180481,
+      "learning_rate": 1.911387971746495e-07,
+      "loss": 0.0503,
+      "step": 16266
+    },
+    {
+      "epoch": 2.635612443292288,
+      "grad_norm": 0.9437099099159241,
+      "learning_rate": 1.9097113017243097e-07,
+      "loss": 0.0515,
+      "step": 16267
+    },
+    {
+      "epoch": 2.6357744653272848,
+      "grad_norm": 0.9175340533256531,
+      "learning_rate": 1.9080353382090798e-07,
+      "loss": 0.0571,
+      "step": 16268
+    },
+    {
+      "epoch": 2.6359364873622813,
+      "grad_norm": 1.21630859375,
+      "learning_rate": 1.9063600812520898e-07,
+      "loss": 0.0573,
+      "step": 16269
+    },
+    {
+      "epoch": 2.636098509397278,
+      "grad_norm": 1.0302997827529907,
+      "learning_rate": 1.9046855309045957e-07,
+      "loss": 0.0666,
+      "step": 16270
+    },
+    {
+      "epoch": 2.6362605314322747,
+      "grad_norm": 0.8015779256820679,
+      "learning_rate": 1.9030116872178317e-07,
+      "loss": 0.0546,
+      "step": 16271
+    },
+    {
+      "epoch": 2.6364225534672716,
+      "grad_norm": 1.0317645072937012,
+      "learning_rate": 1.9013385502430175e-07,
+      "loss": 0.0664,
+      "step": 16272
+    },
+    {
+      "epoch": 2.636584575502268,
+      "grad_norm": 0.9456698894500732,
+      "learning_rate": 1.899666120031349e-07,
+      "loss": 0.0589,
+      "step": 16273
+    },
+    {
+      "epoch": 2.636746597537265,
+      "grad_norm": 0.9725162386894226,
+      "learning_rate": 1.8979943966339924e-07,
+      "loss": 0.065,
+      "step": 16274
+    },
+    {
+      "epoch": 2.636908619572262,
+      "grad_norm": 0.8339110016822815,
+      "learning_rate": 1.8963233801021024e-07,
+      "loss": 0.0542,
+      "step": 16275
+    },
+    {
+      "epoch": 2.6370706416072585,
+      "grad_norm": 0.991547167301178,
+      "learning_rate": 1.8946530704868072e-07,
+      "loss": 0.0624,
+      "step": 16276
+    },
+    {
+      "epoch": 2.6372326636422554,
+      "grad_norm": 0.8456936478614807,
+      "learning_rate": 1.8929834678392184e-07,
+      "loss": 0.0566,
+      "step": 16277
+    },
+    {
+      "epoch": 2.6373946856772523,
+      "grad_norm": 0.9428996443748474,
+      "learning_rate": 1.891314572210412e-07,
+      "loss": 0.0575,
+      "step": 16278
+    },
+    {
+      "epoch": 2.637556707712249,
+      "grad_norm": 0.9978534579277039,
+      "learning_rate": 1.8896463836514556e-07,
+      "loss": 0.0676,
+      "step": 16279
+    },
+    {
+      "epoch": 2.6377187297472457,
+      "grad_norm": 0.837712824344635,
+      "learning_rate": 1.887978902213397e-07,
+      "loss": 0.0553,
+      "step": 16280
+    },
+    {
+      "epoch": 2.6378807517822422,
+      "grad_norm": 1.049025058746338,
+      "learning_rate": 1.8863121279472595e-07,
+      "loss": 0.0626,
+      "step": 16281
+    },
+    {
+      "epoch": 2.638042773817239,
+      "grad_norm": 1.0920323133468628,
+      "learning_rate": 1.8846460609040302e-07,
+      "loss": 0.0624,
+      "step": 16282
+    },
+    {
+      "epoch": 2.6382047958522357,
+      "grad_norm": 0.900133490562439,
+      "learning_rate": 1.88298070113469e-07,
+      "loss": 0.0582,
+      "step": 16283
+    },
+    {
+      "epoch": 2.6383668178872326,
+      "grad_norm": 0.9234185814857483,
+      "learning_rate": 1.881316048690207e-07,
+      "loss": 0.0614,
+      "step": 16284
+    },
+    {
+      "epoch": 2.6385288399222295,
+      "grad_norm": 0.9673925042152405,
+      "learning_rate": 1.8796521036215044e-07,
+      "loss": 0.0552,
+      "step": 16285
+    },
+    {
+      "epoch": 2.638690861957226,
+      "grad_norm": 0.9933484792709351,
+      "learning_rate": 1.8779888659794937e-07,
+      "loss": 0.0658,
+      "step": 16286
+    },
+    {
+      "epoch": 2.638852883992223,
+      "grad_norm": 0.8367605209350586,
+      "learning_rate": 1.8763263358150735e-07,
+      "loss": 0.0614,
+      "step": 16287
+    },
+    {
+      "epoch": 2.63901490602722,
+      "grad_norm": 0.9234703779220581,
+      "learning_rate": 1.874664513179106e-07,
+      "loss": 0.0558,
+      "step": 16288
+    },
+    {
+      "epoch": 2.6391769280622164,
+      "grad_norm": 1.0613017082214355,
+      "learning_rate": 1.8730033981224443e-07,
+      "loss": 0.0688,
+      "step": 16289
+    },
+    {
+      "epoch": 2.6393389500972133,
+      "grad_norm": 1.0138506889343262,
+      "learning_rate": 1.8713429906959097e-07,
+      "loss": 0.0574,
+      "step": 16290
+    },
+    {
+      "epoch": 2.6395009721322102,
+      "grad_norm": 0.9312925338745117,
+      "learning_rate": 1.869683290950311e-07,
+      "loss": 0.0671,
+      "step": 16291
+    },
+    {
+      "epoch": 2.6396629941672067,
+      "grad_norm": 0.90425705909729,
+      "learning_rate": 1.8680242989364327e-07,
+      "loss": 0.0607,
+      "step": 16292
+    },
+    {
+      "epoch": 2.6398250162022032,
+      "grad_norm": 1.0604413747787476,
+      "learning_rate": 1.8663660147050262e-07,
+      "loss": 0.0629,
+      "step": 16293
+    },
+    {
+      "epoch": 2.6399870382372,
+      "grad_norm": 0.7951275706291199,
+      "learning_rate": 1.8647084383068393e-07,
+      "loss": 0.051,
+      "step": 16294
+    },
+    {
+      "epoch": 2.640149060272197,
+      "grad_norm": 0.8711312413215637,
+      "learning_rate": 1.8630515697925927e-07,
+      "loss": 0.0563,
+      "step": 16295
+    },
+    {
+      "epoch": 2.6403110823071936,
+      "grad_norm": 1.0285757780075073,
+      "learning_rate": 1.8613954092129738e-07,
+      "loss": 0.0657,
+      "step": 16296
+    },
+    {
+      "epoch": 2.6404731043421905,
+      "grad_norm": 0.9499953985214233,
+      "learning_rate": 1.8597399566186615e-07,
+      "loss": 0.06,
+      "step": 16297
+    },
+    {
+      "epoch": 2.6406351263771874,
+      "grad_norm": 0.8965511918067932,
+      "learning_rate": 1.858085212060304e-07,
+      "loss": 0.0599,
+      "step": 16298
+    },
+    {
+      "epoch": 2.640797148412184,
+      "grad_norm": 1.0099302530288696,
+      "learning_rate": 1.8564311755885438e-07,
+      "loss": 0.0557,
+      "step": 16299
+    },
+    {
+      "epoch": 2.640959170447181,
+      "grad_norm": 0.9318955540657043,
+      "learning_rate": 1.85477784725398e-07,
+      "loss": 0.0622,
+      "step": 16300
+    },
+    {
+      "epoch": 2.641121192482178,
+      "grad_norm": 0.9583910703659058,
+      "learning_rate": 1.8531252271072025e-07,
+      "loss": 0.0625,
+      "step": 16301
+    },
+    {
+      "epoch": 2.6412832145171743,
+      "grad_norm": 0.8921169638633728,
+      "learning_rate": 1.851473315198782e-07,
+      "loss": 0.0591,
+      "step": 16302
+    },
+    {
+      "epoch": 2.6414452365521712,
+      "grad_norm": 0.9584546089172363,
+      "learning_rate": 1.8498221115792554e-07,
+      "loss": 0.0664,
+      "step": 16303
+    },
+    {
+      "epoch": 2.6416072585871677,
+      "grad_norm": 0.9304138422012329,
+      "learning_rate": 1.848171616299152e-07,
+      "loss": 0.06,
+      "step": 16304
+    },
+    {
+      "epoch": 2.6417692806221647,
+      "grad_norm": 0.9597314596176147,
+      "learning_rate": 1.8465218294089704e-07,
+      "loss": 0.0605,
+      "step": 16305
+    },
+    {
+      "epoch": 2.641931302657161,
+      "grad_norm": 0.9333338737487793,
+      "learning_rate": 1.8448727509591951e-07,
+      "loss": 0.0571,
+      "step": 16306
+    },
+    {
+      "epoch": 2.642093324692158,
+      "grad_norm": 0.9065783619880676,
+      "learning_rate": 1.843224381000272e-07,
+      "loss": 0.0621,
+      "step": 16307
+    },
+    {
+      "epoch": 2.642255346727155,
+      "grad_norm": 0.7040373086929321,
+      "learning_rate": 1.8415767195826468e-07,
+      "loss": 0.0506,
+      "step": 16308
+    },
+    {
+      "epoch": 2.6424173687621515,
+      "grad_norm": 0.825442373752594,
+      "learning_rate": 1.8399297667567317e-07,
+      "loss": 0.0603,
+      "step": 16309
+    },
+    {
+      "epoch": 2.6425793907971484,
+      "grad_norm": 0.9128686189651489,
+      "learning_rate": 1.8382835225729256e-07,
+      "loss": 0.0604,
+      "step": 16310
+    },
+    {
+      "epoch": 2.6427414128321454,
+      "grad_norm": 0.8846583962440491,
+      "learning_rate": 1.836637987081588e-07,
+      "loss": 0.06,
+      "step": 16311
+    },
+    {
+      "epoch": 2.642903434867142,
+      "grad_norm": 0.8874152302742004,
+      "learning_rate": 1.834993160333068e-07,
+      "loss": 0.0601,
+      "step": 16312
+    },
+    {
+      "epoch": 2.643065456902139,
+      "grad_norm": 1.2292134761810303,
+      "learning_rate": 1.8333490423777106e-07,
+      "loss": 0.0675,
+      "step": 16313
+    },
+    {
+      "epoch": 2.6432274789371357,
+      "grad_norm": 0.841103732585907,
+      "learning_rate": 1.831705633265804e-07,
+      "loss": 0.0571,
+      "step": 16314
+    },
+    {
+      "epoch": 2.643389500972132,
+      "grad_norm": 0.9429198503494263,
+      "learning_rate": 1.8300629330476383e-07,
+      "loss": 0.0633,
+      "step": 16315
+    },
+    {
+      "epoch": 2.6435515230071287,
+      "grad_norm": 1.0920491218566895,
+      "learning_rate": 1.8284209417734762e-07,
+      "loss": 0.0608,
+      "step": 16316
+    },
+    {
+      "epoch": 2.6437135450421256,
+      "grad_norm": 0.9739688634872437,
+      "learning_rate": 1.8267796594935606e-07,
+      "loss": 0.0637,
+      "step": 16317
+    },
+    {
+      "epoch": 2.6438755670771226,
+      "grad_norm": 1.1021599769592285,
+      "learning_rate": 1.8251390862581097e-07,
+      "loss": 0.0693,
+      "step": 16318
+    },
+    {
+      "epoch": 2.644037589112119,
+      "grad_norm": 0.9459995627403259,
+      "learning_rate": 1.8234992221173198e-07,
+      "loss": 0.0639,
+      "step": 16319
+    },
+    {
+      "epoch": 2.644199611147116,
+      "grad_norm": 0.9659364223480225,
+      "learning_rate": 1.8218600671213698e-07,
+      "loss": 0.06,
+      "step": 16320
+    },
+    {
+      "epoch": 2.644361633182113,
+      "grad_norm": 0.954434871673584,
+      "learning_rate": 1.8202216213204144e-07,
+      "loss": 0.0613,
+      "step": 16321
+    },
+    {
+      "epoch": 2.6445236552171094,
+      "grad_norm": 0.8224755525588989,
+      "learning_rate": 1.8185838847645743e-07,
+      "loss": 0.055,
+      "step": 16322
+    },
+    {
+      "epoch": 2.6446856772521063,
+      "grad_norm": 0.9007136821746826,
+      "learning_rate": 1.8169468575039735e-07,
+      "loss": 0.0645,
+      "step": 16323
+    },
+    {
+      "epoch": 2.6448476992871033,
+      "grad_norm": 1.1272404193878174,
+      "learning_rate": 1.8153105395886967e-07,
+      "loss": 0.0585,
+      "step": 16324
+    },
+    {
+      "epoch": 2.6450097213220998,
+      "grad_norm": 0.8485245108604431,
+      "learning_rate": 1.813674931068818e-07,
+      "loss": 0.0603,
+      "step": 16325
+    },
+    {
+      "epoch": 2.6451717433570967,
+      "grad_norm": 0.8735572099685669,
+      "learning_rate": 1.8120400319943692e-07,
+      "loss": 0.0557,
+      "step": 16326
+    },
+    {
+      "epoch": 2.645333765392093,
+      "grad_norm": 0.8615409731864929,
+      "learning_rate": 1.8104058424153802e-07,
+      "loss": 0.0594,
+      "step": 16327
+    },
+    {
+      "epoch": 2.64549578742709,
+      "grad_norm": 0.9410812258720398,
+      "learning_rate": 1.8087723623818608e-07,
+      "loss": 0.0596,
+      "step": 16328
+    },
+    {
+      "epoch": 2.6456578094620866,
+      "grad_norm": 0.7836385369300842,
+      "learning_rate": 1.8071395919437823e-07,
+      "loss": 0.0522,
+      "step": 16329
+    },
+    {
+      "epoch": 2.6458198314970836,
+      "grad_norm": 0.9764214158058167,
+      "learning_rate": 1.805507531151107e-07,
+      "loss": 0.067,
+      "step": 16330
+    },
+    {
+      "epoch": 2.6459818535320805,
+      "grad_norm": 0.8474244475364685,
+      "learning_rate": 1.8038761800537708e-07,
+      "loss": 0.0555,
+      "step": 16331
+    },
+    {
+      "epoch": 2.646143875567077,
+      "grad_norm": 0.8792239427566528,
+      "learning_rate": 1.8022455387016913e-07,
+      "loss": 0.0558,
+      "step": 16332
+    },
+    {
+      "epoch": 2.646305897602074,
+      "grad_norm": 0.8611176609992981,
+      "learning_rate": 1.8006156071447595e-07,
+      "loss": 0.0568,
+      "step": 16333
+    },
+    {
+      "epoch": 2.646467919637071,
+      "grad_norm": 0.9201486706733704,
+      "learning_rate": 1.7989863854328492e-07,
+      "loss": 0.0564,
+      "step": 16334
+    },
+    {
+      "epoch": 2.6466299416720673,
+      "grad_norm": 0.8337270021438599,
+      "learning_rate": 1.7973578736158098e-07,
+      "loss": 0.0522,
+      "step": 16335
+    },
+    {
+      "epoch": 2.6467919637070643,
+      "grad_norm": 0.9367399215698242,
+      "learning_rate": 1.7957300717434706e-07,
+      "loss": 0.0577,
+      "step": 16336
+    },
+    {
+      "epoch": 2.6469539857420608,
+      "grad_norm": 0.913486123085022,
+      "learning_rate": 1.794102979865639e-07,
+      "loss": 0.0639,
+      "step": 16337
+    },
+    {
+      "epoch": 2.6471160077770577,
+      "grad_norm": 0.9401398301124573,
+      "learning_rate": 1.7924765980320974e-07,
+      "loss": 0.0588,
+      "step": 16338
+    },
+    {
+      "epoch": 2.647278029812054,
+      "grad_norm": 0.8506883382797241,
+      "learning_rate": 1.790850926292617e-07,
+      "loss": 0.0583,
+      "step": 16339
+    },
+    {
+      "epoch": 2.647440051847051,
+      "grad_norm": 0.9261347055435181,
+      "learning_rate": 1.7892259646969278e-07,
+      "loss": 0.0591,
+      "step": 16340
+    },
+    {
+      "epoch": 2.647602073882048,
+      "grad_norm": 1.0030386447906494,
+      "learning_rate": 1.7876017132947483e-07,
+      "loss": 0.061,
+      "step": 16341
+    },
+    {
+      "epoch": 2.6477640959170445,
+      "grad_norm": 0.9110730290412903,
+      "learning_rate": 1.785978172135791e-07,
+      "loss": 0.0569,
+      "step": 16342
+    },
+    {
+      "epoch": 2.6479261179520415,
+      "grad_norm": 0.8250390887260437,
+      "learning_rate": 1.7843553412697278e-07,
+      "loss": 0.0486,
+      "step": 16343
+    },
+    {
+      "epoch": 2.6480881399870384,
+      "grad_norm": 1.0518966913223267,
+      "learning_rate": 1.782733220746205e-07,
+      "loss": 0.0639,
+      "step": 16344
+    },
+    {
+      "epoch": 2.648250162022035,
+      "grad_norm": 0.9728784561157227,
+      "learning_rate": 1.7811118106148633e-07,
+      "loss": 0.0679,
+      "step": 16345
+    },
+    {
+      "epoch": 2.648412184057032,
+      "grad_norm": 0.9852555990219116,
+      "learning_rate": 1.7794911109253105e-07,
+      "loss": 0.0634,
+      "step": 16346
+    },
+    {
+      "epoch": 2.6485742060920288,
+      "grad_norm": 0.9500349164009094,
+      "learning_rate": 1.777871121727137e-07,
+      "loss": 0.0619,
+      "step": 16347
+    },
+    {
+      "epoch": 2.6487362281270252,
+      "grad_norm": 0.9328777194023132,
+      "learning_rate": 1.7762518430699122e-07,
+      "loss": 0.0637,
+      "step": 16348
+    },
+    {
+      "epoch": 2.648898250162022,
+      "grad_norm": 1.0019419193267822,
+      "learning_rate": 1.774633275003179e-07,
+      "loss": 0.0678,
+      "step": 16349
+    },
+    {
+      "epoch": 2.6490602721970187,
+      "grad_norm": 1.070502758026123,
+      "learning_rate": 1.7730154175764623e-07,
+      "loss": 0.0578,
+      "step": 16350
+    },
+    {
+      "epoch": 2.6492222942320156,
+      "grad_norm": 0.82248455286026,
+      "learning_rate": 1.771398270839267e-07,
+      "loss": 0.054,
+      "step": 16351
+    },
+    {
+      "epoch": 2.649384316267012,
+      "grad_norm": 1.011540412902832,
+      "learning_rate": 1.7697818348410722e-07,
+      "loss": 0.066,
+      "step": 16352
+    },
+    {
+      "epoch": 2.649546338302009,
+      "grad_norm": 0.9603102207183838,
+      "learning_rate": 1.7681661096313364e-07,
+      "loss": 0.0643,
+      "step": 16353
+    },
+    {
+      "epoch": 2.649708360337006,
+      "grad_norm": 0.9704177379608154,
+      "learning_rate": 1.7665510952595027e-07,
+      "loss": 0.0703,
+      "step": 16354
+    },
+    {
+      "epoch": 2.6498703823720025,
+      "grad_norm": 0.9658411741256714,
+      "learning_rate": 1.764936791774974e-07,
+      "loss": 0.0621,
+      "step": 16355
+    },
+    {
+      "epoch": 2.6500324044069994,
+      "grad_norm": 0.9105367660522461,
+      "learning_rate": 1.7633231992271572e-07,
+      "loss": 0.0545,
+      "step": 16356
+    },
+    {
+      "epoch": 2.6501944264419963,
+      "grad_norm": 0.8852334022521973,
+      "learning_rate": 1.7617103176654187e-07,
+      "loss": 0.0669,
+      "step": 16357
+    },
+    {
+      "epoch": 2.650356448476993,
+      "grad_norm": 0.9062175154685974,
+      "learning_rate": 1.7600981471391083e-07,
+      "loss": 0.0594,
+      "step": 16358
+    },
+    {
+      "epoch": 2.6505184705119897,
+      "grad_norm": 0.8429781794548035,
+      "learning_rate": 1.7584866876975526e-07,
+      "loss": 0.0531,
+      "step": 16359
+    },
+    {
+      "epoch": 2.6506804925469862,
+      "grad_norm": 0.9941640496253967,
+      "learning_rate": 1.7568759393900597e-07,
+      "loss": 0.0622,
+      "step": 16360
+    },
+    {
+      "epoch": 2.650842514581983,
+      "grad_norm": 1.0209076404571533,
+      "learning_rate": 1.7552659022659206e-07,
+      "loss": 0.0636,
+      "step": 16361
+    },
+    {
+      "epoch": 2.6510045366169797,
+      "grad_norm": 1.1594321727752686,
+      "learning_rate": 1.7536565763743934e-07,
+      "loss": 0.0646,
+      "step": 16362
+    },
+    {
+      "epoch": 2.6511665586519766,
+      "grad_norm": 0.9234576225280762,
+      "learning_rate": 1.7520479617647163e-07,
+      "loss": 0.0573,
+      "step": 16363
+    },
+    {
+      "epoch": 2.6513285806869735,
+      "grad_norm": 0.9568017721176147,
+      "learning_rate": 1.7504400584861137e-07,
+      "loss": 0.065,
+      "step": 16364
+    },
+    {
+      "epoch": 2.65149060272197,
+      "grad_norm": 0.9261468052864075,
+      "learning_rate": 1.7488328665877823e-07,
+      "loss": 0.0632,
+      "step": 16365
+    },
+    {
+      "epoch": 2.651652624756967,
+      "grad_norm": 0.8637058734893799,
+      "learning_rate": 1.7472263861189e-07,
+      "loss": 0.0605,
+      "step": 16366
+    },
+    {
+      "epoch": 2.651814646791964,
+      "grad_norm": 0.9233385324478149,
+      "learning_rate": 1.7456206171286182e-07,
+      "loss": 0.061,
+      "step": 16367
+    },
+    {
+      "epoch": 2.6519766688269604,
+      "grad_norm": 0.8062421679496765,
+      "learning_rate": 1.7440155596660735e-07,
+      "loss": 0.0519,
+      "step": 16368
+    },
+    {
+      "epoch": 2.6521386908619573,
+      "grad_norm": 0.8380430340766907,
+      "learning_rate": 1.7424112137803763e-07,
+      "loss": 0.0575,
+      "step": 16369
+    },
+    {
+      "epoch": 2.6523007128969542,
+      "grad_norm": 0.9855707287788391,
+      "learning_rate": 1.7408075795206037e-07,
+      "loss": 0.0603,
+      "step": 16370
+    },
+    {
+      "epoch": 2.6524627349319507,
+      "grad_norm": 1.0595271587371826,
+      "learning_rate": 1.739204656935839e-07,
+      "loss": 0.0718,
+      "step": 16371
+    },
+    {
+      "epoch": 2.6526247569669477,
+      "grad_norm": 1.0735374689102173,
+      "learning_rate": 1.7376024460751262e-07,
+      "loss": 0.0748,
+      "step": 16372
+    },
+    {
+      "epoch": 2.652786779001944,
+      "grad_norm": 0.929942786693573,
+      "learning_rate": 1.7360009469874788e-07,
+      "loss": 0.0564,
+      "step": 16373
+    },
+    {
+      "epoch": 2.652948801036941,
+      "grad_norm": 0.8766565918922424,
+      "learning_rate": 1.7344001597219024e-07,
+      "loss": 0.0616,
+      "step": 16374
+    },
+    {
+      "epoch": 2.6531108230719376,
+      "grad_norm": 0.9398424029350281,
+      "learning_rate": 1.7328000843273879e-07,
+      "loss": 0.0622,
+      "step": 16375
+    },
+    {
+      "epoch": 2.6532728451069345,
+      "grad_norm": 0.9084738492965698,
+      "learning_rate": 1.7312007208528796e-07,
+      "loss": 0.0656,
+      "step": 16376
+    },
+    {
+      "epoch": 2.6534348671419314,
+      "grad_norm": 0.9480365514755249,
+      "learning_rate": 1.729602069347322e-07,
+      "loss": 0.061,
+      "step": 16377
+    },
+    {
+      "epoch": 2.653596889176928,
+      "grad_norm": 0.875299870967865,
+      "learning_rate": 1.7280041298596257e-07,
+      "loss": 0.0631,
+      "step": 16378
+    },
+    {
+      "epoch": 2.653758911211925,
+      "grad_norm": 0.9430073499679565,
+      "learning_rate": 1.7264069024386876e-07,
+      "loss": 0.066,
+      "step": 16379
+    },
+    {
+      "epoch": 2.653920933246922,
+      "grad_norm": 1.0011956691741943,
+      "learning_rate": 1.7248103871333743e-07,
+      "loss": 0.0557,
+      "step": 16380
+    },
+    {
+      "epoch": 2.6540829552819183,
+      "grad_norm": 0.783429741859436,
+      "learning_rate": 1.7232145839925413e-07,
+      "loss": 0.0582,
+      "step": 16381
+    },
+    {
+      "epoch": 2.654244977316915,
+      "grad_norm": 1.0384507179260254,
+      "learning_rate": 1.7216194930650105e-07,
+      "loss": 0.0629,
+      "step": 16382
+    },
+    {
+      "epoch": 2.6544069993519117,
+      "grad_norm": 0.9398157596588135,
+      "learning_rate": 1.7200251143995983e-07,
+      "loss": 0.0562,
+      "step": 16383
+    },
+    {
+      "epoch": 2.6545690213869086,
+      "grad_norm": 1.124776005744934,
+      "learning_rate": 1.7184314480450713e-07,
+      "loss": 0.0653,
+      "step": 16384
+    },
+    {
+      "epoch": 2.654731043421905,
+      "grad_norm": 0.9212331771850586,
+      "learning_rate": 1.7168384940502048e-07,
+      "loss": 0.0662,
+      "step": 16385
+    },
+    {
+      "epoch": 2.654893065456902,
+      "grad_norm": 1.0174168348312378,
+      "learning_rate": 1.715246252463737e-07,
+      "loss": 0.065,
+      "step": 16386
+    },
+    {
+      "epoch": 2.655055087491899,
+      "grad_norm": 0.8449579477310181,
+      "learning_rate": 1.7136547233343904e-07,
+      "loss": 0.0596,
+      "step": 16387
+    },
+    {
+      "epoch": 2.6552171095268955,
+      "grad_norm": 0.9600573182106018,
+      "learning_rate": 1.7120639067108508e-07,
+      "loss": 0.0591,
+      "step": 16388
+    },
+    {
+      "epoch": 2.6553791315618924,
+      "grad_norm": 0.9382137060165405,
+      "learning_rate": 1.7104738026417987e-07,
+      "loss": 0.0599,
+      "step": 16389
+    },
+    {
+      "epoch": 2.6555411535968894,
+      "grad_norm": 1.0673377513885498,
+      "learning_rate": 1.7088844111758956e-07,
+      "loss": 0.0638,
+      "step": 16390
+    },
+    {
+      "epoch": 2.655703175631886,
+      "grad_norm": 0.9802731871604919,
+      "learning_rate": 1.7072957323617635e-07,
+      "loss": 0.0656,
+      "step": 16391
+    },
+    {
+      "epoch": 2.655865197666883,
+      "grad_norm": 0.9394249320030212,
+      "learning_rate": 1.7057077662480131e-07,
+      "loss": 0.0533,
+      "step": 16392
+    },
+    {
+      "epoch": 2.6560272197018797,
+      "grad_norm": 0.9298037886619568,
+      "learning_rate": 1.7041205128832338e-07,
+      "loss": 0.059,
+      "step": 16393
+    },
+    {
+      "epoch": 2.656189241736876,
+      "grad_norm": 1.0378389358520508,
+      "learning_rate": 1.7025339723159924e-07,
+      "loss": 0.0637,
+      "step": 16394
+    },
+    {
+      "epoch": 2.656351263771873,
+      "grad_norm": 0.7979336380958557,
+      "learning_rate": 1.700948144594833e-07,
+      "loss": 0.059,
+      "step": 16395
+    },
+    {
+      "epoch": 2.6565132858068696,
+      "grad_norm": 0.8939142823219299,
+      "learning_rate": 1.6993630297682778e-07,
+      "loss": 0.0633,
+      "step": 16396
+    },
+    {
+      "epoch": 2.6566753078418666,
+      "grad_norm": 0.8715953230857849,
+      "learning_rate": 1.6977786278848275e-07,
+      "loss": 0.0577,
+      "step": 16397
+    },
+    {
+      "epoch": 2.656837329876863,
+      "grad_norm": 0.8175707459449768,
+      "learning_rate": 1.6961949389929593e-07,
+      "loss": 0.0571,
+      "step": 16398
+    },
+    {
+      "epoch": 2.65699935191186,
+      "grad_norm": 1.1204349994659424,
+      "learning_rate": 1.6946119631411352e-07,
+      "loss": 0.0584,
+      "step": 16399
+    },
+    {
+      "epoch": 2.657161373946857,
+      "grad_norm": 1.0144906044006348,
+      "learning_rate": 1.693029700377785e-07,
+      "loss": 0.0669,
+      "step": 16400
+    },
+    {
+      "epoch": 2.6573233959818534,
+      "grad_norm": 0.8621246218681335,
+      "learning_rate": 1.6914481507513263e-07,
+      "loss": 0.0629,
+      "step": 16401
+    },
+    {
+      "epoch": 2.6574854180168503,
+      "grad_norm": 0.8595054745674133,
+      "learning_rate": 1.6898673143101479e-07,
+      "loss": 0.0557,
+      "step": 16402
+    },
+    {
+      "epoch": 2.6576474400518473,
+      "grad_norm": 0.8768499493598938,
+      "learning_rate": 1.688287191102614e-07,
+      "loss": 0.0645,
+      "step": 16403
+    },
+    {
+      "epoch": 2.6578094620868438,
+      "grad_norm": 0.9614449143409729,
+      "learning_rate": 1.6867077811770826e-07,
+      "loss": 0.0583,
+      "step": 16404
+    },
+    {
+      "epoch": 2.6579714841218407,
+      "grad_norm": 0.9931366443634033,
+      "learning_rate": 1.6851290845818824e-07,
+      "loss": 0.0712,
+      "step": 16405
+    },
+    {
+      "epoch": 2.658133506156837,
+      "grad_norm": 0.8831222653388977,
+      "learning_rate": 1.683551101365305e-07,
+      "loss": 0.055,
+      "step": 16406
+    },
+    {
+      "epoch": 2.658295528191834,
+      "grad_norm": 0.9368479251861572,
+      "learning_rate": 1.6819738315756395e-07,
+      "loss": 0.0537,
+      "step": 16407
+    },
+    {
+      "epoch": 2.6584575502268306,
+      "grad_norm": 0.9596263766288757,
+      "learning_rate": 1.6803972752611475e-07,
+      "loss": 0.0548,
+      "step": 16408
+    },
+    {
+      "epoch": 2.6586195722618275,
+      "grad_norm": 0.9823015928268433,
+      "learning_rate": 1.6788214324700652e-07,
+      "loss": 0.0593,
+      "step": 16409
+    },
+    {
+      "epoch": 2.6587815942968245,
+      "grad_norm": 0.9856812357902527,
+      "learning_rate": 1.6772463032506126e-07,
+      "loss": 0.0581,
+      "step": 16410
+    },
+    {
+      "epoch": 2.658943616331821,
+      "grad_norm": 1.0176301002502441,
+      "learning_rate": 1.6756718876509815e-07,
+      "loss": 0.06,
+      "step": 16411
+    },
+    {
+      "epoch": 2.659105638366818,
+      "grad_norm": 0.8685017228126526,
+      "learning_rate": 1.6740981857193471e-07,
+      "loss": 0.0569,
+      "step": 16412
+    },
+    {
+      "epoch": 2.659267660401815,
+      "grad_norm": 0.8308966755867004,
+      "learning_rate": 1.672525197503863e-07,
+      "loss": 0.0605,
+      "step": 16413
+    },
+    {
+      "epoch": 2.6594296824368113,
+      "grad_norm": 0.7896578311920166,
+      "learning_rate": 1.6709529230526544e-07,
+      "loss": 0.0544,
+      "step": 16414
+    },
+    {
+      "epoch": 2.6595917044718083,
+      "grad_norm": 1.1066980361938477,
+      "learning_rate": 1.66938136241383e-07,
+      "loss": 0.0637,
+      "step": 16415
+    },
+    {
+      "epoch": 2.659753726506805,
+      "grad_norm": 1.2486014366149902,
+      "learning_rate": 1.667810515635482e-07,
+      "loss": 0.0629,
+      "step": 16416
+    },
+    {
+      "epoch": 2.6599157485418017,
+      "grad_norm": 0.92633056640625,
+      "learning_rate": 1.6662403827656603e-07,
+      "loss": 0.0627,
+      "step": 16417
+    },
+    {
+      "epoch": 2.660077770576798,
+      "grad_norm": 0.9085054397583008,
+      "learning_rate": 1.6646709638524216e-07,
+      "loss": 0.0594,
+      "step": 16418
+    },
+    {
+      "epoch": 2.660239792611795,
+      "grad_norm": 1.1926897764205933,
+      "learning_rate": 1.6631022589437828e-07,
+      "loss": 0.0662,
+      "step": 16419
+    },
+    {
+      "epoch": 2.660401814646792,
+      "grad_norm": 0.877842128276825,
+      "learning_rate": 1.6615342680877417e-07,
+      "loss": 0.0617,
+      "step": 16420
+    },
+    {
+      "epoch": 2.6605638366817885,
+      "grad_norm": 1.0418201684951782,
+      "learning_rate": 1.6599669913322708e-07,
+      "loss": 0.061,
+      "step": 16421
+    },
+    {
+      "epoch": 2.6607258587167855,
+      "grad_norm": 0.9494113326072693,
+      "learning_rate": 1.6584004287253235e-07,
+      "loss": 0.0586,
+      "step": 16422
+    },
+    {
+      "epoch": 2.6608878807517824,
+      "grad_norm": 1.1011654138565063,
+      "learning_rate": 1.6568345803148478e-07,
+      "loss": 0.0642,
+      "step": 16423
+    },
+    {
+      "epoch": 2.661049902786779,
+      "grad_norm": 1.0519500970840454,
+      "learning_rate": 1.6552694461487385e-07,
+      "loss": 0.0583,
+      "step": 16424
+    },
+    {
+      "epoch": 2.661211924821776,
+      "grad_norm": 0.9799832105636597,
+      "learning_rate": 1.6537050262748905e-07,
+      "loss": 0.0611,
+      "step": 16425
+    },
+    {
+      "epoch": 2.6613739468567728,
+      "grad_norm": 0.9446945190429688,
+      "learning_rate": 1.652141320741174e-07,
+      "loss": 0.0623,
+      "step": 16426
+    },
+    {
+      "epoch": 2.6615359688917692,
+      "grad_norm": 0.9712892770767212,
+      "learning_rate": 1.6505783295954314e-07,
+      "loss": 0.0589,
+      "step": 16427
+    },
+    {
+      "epoch": 2.661697990926766,
+      "grad_norm": 0.9126527905464172,
+      "learning_rate": 1.6490160528854855e-07,
+      "loss": 0.063,
+      "step": 16428
+    },
+    {
+      "epoch": 2.6618600129617627,
+      "grad_norm": 0.843527615070343,
+      "learning_rate": 1.6474544906591422e-07,
+      "loss": 0.0636,
+      "step": 16429
+    },
+    {
+      "epoch": 2.6620220349967596,
+      "grad_norm": 1.002021312713623,
+      "learning_rate": 1.6458936429641803e-07,
+      "loss": 0.0638,
+      "step": 16430
+    },
+    {
+      "epoch": 2.662184057031756,
+      "grad_norm": 0.8258562684059143,
+      "learning_rate": 1.6443335098483586e-07,
+      "loss": 0.0567,
+      "step": 16431
+    },
+    {
+      "epoch": 2.662346079066753,
+      "grad_norm": 0.987740159034729,
+      "learning_rate": 1.642774091359406e-07,
+      "loss": 0.0617,
+      "step": 16432
+    },
+    {
+      "epoch": 2.66250810110175,
+      "grad_norm": 1.08059823513031,
+      "learning_rate": 1.6412153875450448e-07,
+      "loss": 0.0681,
+      "step": 16433
+    },
+    {
+      "epoch": 2.6626701231367464,
+      "grad_norm": 0.9311217069625854,
+      "learning_rate": 1.6396573984529707e-07,
+      "loss": 0.063,
+      "step": 16434
+    },
+    {
+      "epoch": 2.6628321451717434,
+      "grad_norm": 0.8519605994224548,
+      "learning_rate": 1.6381001241308452e-07,
+      "loss": 0.0563,
+      "step": 16435
+    },
+    {
+      "epoch": 2.6629941672067403,
+      "grad_norm": 0.9323353171348572,
+      "learning_rate": 1.6365435646263223e-07,
+      "loss": 0.0625,
+      "step": 16436
+    },
+    {
+      "epoch": 2.663156189241737,
+      "grad_norm": 0.9580549597740173,
+      "learning_rate": 1.6349877199870218e-07,
+      "loss": 0.0612,
+      "step": 16437
+    },
+    {
+      "epoch": 2.6633182112767337,
+      "grad_norm": 0.9537264108657837,
+      "learning_rate": 1.6334325902605642e-07,
+      "loss": 0.0611,
+      "step": 16438
+    },
+    {
+      "epoch": 2.6634802333117307,
+      "grad_norm": 0.8179647922515869,
+      "learning_rate": 1.6318781754945168e-07,
+      "loss": 0.0574,
+      "step": 16439
+    },
+    {
+      "epoch": 2.663642255346727,
+      "grad_norm": 0.8439454436302185,
+      "learning_rate": 1.63032447573645e-07,
+      "loss": 0.0547,
+      "step": 16440
+    },
+    {
+      "epoch": 2.6638042773817237,
+      "grad_norm": 0.8650406002998352,
+      "learning_rate": 1.6287714910339008e-07,
+      "loss": 0.0558,
+      "step": 16441
+    },
+    {
+      "epoch": 2.6639662994167206,
+      "grad_norm": 1.0165598392486572,
+      "learning_rate": 1.6272192214343868e-07,
+      "loss": 0.069,
+      "step": 16442
+    },
+    {
+      "epoch": 2.6641283214517175,
+      "grad_norm": 0.9209080934524536,
+      "learning_rate": 1.6256676669854032e-07,
+      "loss": 0.0659,
+      "step": 16443
+    },
+    {
+      "epoch": 2.664290343486714,
+      "grad_norm": 1.0296342372894287,
+      "learning_rate": 1.6241168277344232e-07,
+      "loss": 0.0654,
+      "step": 16444
+    },
+    {
+      "epoch": 2.664452365521711,
+      "grad_norm": 0.9982236623764038,
+      "learning_rate": 1.6225667037289034e-07,
+      "loss": 0.0727,
+      "step": 16445
+    },
+    {
+      "epoch": 2.664614387556708,
+      "grad_norm": 0.8094567656517029,
+      "learning_rate": 1.6210172950162639e-07,
+      "loss": 0.0574,
+      "step": 16446
+    },
+    {
+      "epoch": 2.6647764095917044,
+      "grad_norm": 1.097362995147705,
+      "learning_rate": 1.6194686016439227e-07,
+      "loss": 0.0682,
+      "step": 16447
+    },
+    {
+      "epoch": 2.6649384316267013,
+      "grad_norm": 0.9655567407608032,
+      "learning_rate": 1.617920623659261e-07,
+      "loss": 0.0602,
+      "step": 16448
+    },
+    {
+      "epoch": 2.6651004536616982,
+      "grad_norm": 2.5268149375915527,
+      "learning_rate": 1.6163733611096495e-07,
+      "loss": 0.0604,
+      "step": 16449
+    },
+    {
+      "epoch": 2.6652624756966947,
+      "grad_norm": 0.8302810788154602,
+      "learning_rate": 1.6148268140424224e-07,
+      "loss": 0.0574,
+      "step": 16450
+    },
+    {
+      "epoch": 2.6654244977316917,
+      "grad_norm": 1.3241533041000366,
+      "learning_rate": 1.6132809825049e-07,
+      "loss": 0.0606,
+      "step": 16451
+    },
+    {
+      "epoch": 2.665586519766688,
+      "grad_norm": 0.991156280040741,
+      "learning_rate": 1.6117358665443922e-07,
+      "loss": 0.0633,
+      "step": 16452
+    },
+    {
+      "epoch": 2.665748541801685,
+      "grad_norm": 0.8456504344940186,
+      "learning_rate": 1.6101914662081665e-07,
+      "loss": 0.0609,
+      "step": 16453
+    },
+    {
+      "epoch": 2.6659105638366816,
+      "grad_norm": 1.1732730865478516,
+      "learning_rate": 1.6086477815434763e-07,
+      "loss": 0.0678,
+      "step": 16454
+    },
+    {
+      "epoch": 2.6660725858716785,
+      "grad_norm": 0.9648259878158569,
+      "learning_rate": 1.6071048125975598e-07,
+      "loss": 0.0613,
+      "step": 16455
+    },
+    {
+      "epoch": 2.6662346079066754,
+      "grad_norm": 1.1859503984451294,
+      "learning_rate": 1.6055625594176254e-07,
+      "loss": 0.0761,
+      "step": 16456
+    },
+    {
+      "epoch": 2.666396629941672,
+      "grad_norm": 0.9471797347068787,
+      "learning_rate": 1.604021022050864e-07,
+      "loss": 0.0643,
+      "step": 16457
+    },
+    {
+      "epoch": 2.666558651976669,
+      "grad_norm": 0.9277358651161194,
+      "learning_rate": 1.602480200544443e-07,
+      "loss": 0.0676,
+      "step": 16458
+    },
+    {
+      "epoch": 2.666720674011666,
+      "grad_norm": 0.9459307789802551,
+      "learning_rate": 1.600940094945505e-07,
+      "loss": 0.0658,
+      "step": 16459
+    },
+    {
+      "epoch": 2.6668826960466623,
+      "grad_norm": 0.8798080086708069,
+      "learning_rate": 1.5994007053011796e-07,
+      "loss": 0.0642,
+      "step": 16460
+    },
+    {
+      "epoch": 2.667044718081659,
+      "grad_norm": 0.9812191128730774,
+      "learning_rate": 1.5978620316585564e-07,
+      "loss": 0.0614,
+      "step": 16461
+    },
+    {
+      "epoch": 2.6672067401166557,
+      "grad_norm": 0.898577094078064,
+      "learning_rate": 1.5963240740647285e-07,
+      "loss": 0.0627,
+      "step": 16462
+    },
+    {
+      "epoch": 2.6673687621516526,
+      "grad_norm": 0.9194305539131165,
+      "learning_rate": 1.594786832566747e-07,
+      "loss": 0.0637,
+      "step": 16463
+    },
+    {
+      "epoch": 2.667530784186649,
+      "grad_norm": 0.895622730255127,
+      "learning_rate": 1.5932503072116524e-07,
+      "loss": 0.0625,
+      "step": 16464
+    },
+    {
+      "epoch": 2.667692806221646,
+      "grad_norm": 0.9284375309944153,
+      "learning_rate": 1.5917144980464483e-07,
+      "loss": 0.0645,
+      "step": 16465
+    },
+    {
+      "epoch": 2.667854828256643,
+      "grad_norm": 0.8722718358039856,
+      "learning_rate": 1.5901794051181362e-07,
+      "loss": 0.057,
+      "step": 16466
+    },
+    {
+      "epoch": 2.6680168502916395,
+      "grad_norm": 0.8681865930557251,
+      "learning_rate": 1.58864502847369e-07,
+      "loss": 0.059,
+      "step": 16467
+    },
+    {
+      "epoch": 2.6681788723266364,
+      "grad_norm": 0.9118895530700684,
+      "learning_rate": 1.5871113681600464e-07,
+      "loss": 0.0578,
+      "step": 16468
+    },
+    {
+      "epoch": 2.6683408943616334,
+      "grad_norm": 0.9743660092353821,
+      "learning_rate": 1.5855784242241352e-07,
+      "loss": 0.0638,
+      "step": 16469
+    },
+    {
+      "epoch": 2.66850291639663,
+      "grad_norm": 0.9498010277748108,
+      "learning_rate": 1.5840461967128628e-07,
+      "loss": 0.0605,
+      "step": 16470
+    },
+    {
+      "epoch": 2.6686649384316268,
+      "grad_norm": 0.9760184288024902,
+      "learning_rate": 1.5825146856731144e-07,
+      "loss": 0.0638,
+      "step": 16471
+    },
+    {
+      "epoch": 2.6688269604666237,
+      "grad_norm": 0.9990751147270203,
+      "learning_rate": 1.5809838911517438e-07,
+      "loss": 0.0612,
+      "step": 16472
+    },
+    {
+      "epoch": 2.66898898250162,
+      "grad_norm": 0.8862509727478027,
+      "learning_rate": 1.5794538131955944e-07,
+      "loss": 0.0579,
+      "step": 16473
+    },
+    {
+      "epoch": 2.669151004536617,
+      "grad_norm": 0.9482281804084778,
+      "learning_rate": 1.5779244518514813e-07,
+      "loss": 0.0669,
+      "step": 16474
+    },
+    {
+      "epoch": 2.6693130265716136,
+      "grad_norm": 0.9086149334907532,
+      "learning_rate": 1.5763958071662006e-07,
+      "loss": 0.059,
+      "step": 16475
+    },
+    {
+      "epoch": 2.6694750486066106,
+      "grad_norm": 0.8625389933586121,
+      "learning_rate": 1.574867879186523e-07,
+      "loss": 0.0598,
+      "step": 16476
+    },
+    {
+      "epoch": 2.669637070641607,
+      "grad_norm": 0.9149574041366577,
+      "learning_rate": 1.5733406679592028e-07,
+      "loss": 0.0646,
+      "step": 16477
+    },
+    {
+      "epoch": 2.669799092676604,
+      "grad_norm": 0.7897137999534607,
+      "learning_rate": 1.5718141735309695e-07,
+      "loss": 0.0554,
+      "step": 16478
+    },
+    {
+      "epoch": 2.669961114711601,
+      "grad_norm": 0.7885839939117432,
+      "learning_rate": 1.5702883959485215e-07,
+      "loss": 0.0547,
+      "step": 16479
+    },
+    {
+      "epoch": 2.6701231367465974,
+      "grad_norm": 0.9897783398628235,
+      "learning_rate": 1.5687633352585467e-07,
+      "loss": 0.0536,
+      "step": 16480
+    },
+    {
+      "epoch": 2.6702851587815943,
+      "grad_norm": 1.0975406169891357,
+      "learning_rate": 1.5672389915077162e-07,
+      "loss": 0.0666,
+      "step": 16481
+    },
+    {
+      "epoch": 2.6704471808165913,
+      "grad_norm": 0.8759128451347351,
+      "learning_rate": 1.5657153647426703e-07,
+      "loss": 0.0639,
+      "step": 16482
+    },
+    {
+      "epoch": 2.6706092028515878,
+      "grad_norm": 0.8712435364723206,
+      "learning_rate": 1.5641924550100218e-07,
+      "loss": 0.059,
+      "step": 16483
+    },
+    {
+      "epoch": 2.6707712248865847,
+      "grad_norm": 0.7633887529373169,
+      "learning_rate": 1.5626702623563694e-07,
+      "loss": 0.0531,
+      "step": 16484
+    },
+    {
+      "epoch": 2.670933246921581,
+      "grad_norm": 1.1077615022659302,
+      "learning_rate": 1.561148786828287e-07,
+      "loss": 0.0654,
+      "step": 16485
+    },
+    {
+      "epoch": 2.671095268956578,
+      "grad_norm": 0.95298832654953,
+      "learning_rate": 1.5596280284723348e-07,
+      "loss": 0.0593,
+      "step": 16486
+    },
+    {
+      "epoch": 2.6712572909915746,
+      "grad_norm": 1.1085457801818848,
+      "learning_rate": 1.558107987335039e-07,
+      "loss": 0.0647,
+      "step": 16487
+    },
+    {
+      "epoch": 2.6714193130265715,
+      "grad_norm": 0.8979312181472778,
+      "learning_rate": 1.5565886634629102e-07,
+      "loss": 0.06,
+      "step": 16488
+    },
+    {
+      "epoch": 2.6715813350615685,
+      "grad_norm": 0.8960966467857361,
+      "learning_rate": 1.5550700569024358e-07,
+      "loss": 0.0561,
+      "step": 16489
+    },
+    {
+      "epoch": 2.671743357096565,
+      "grad_norm": 0.8627614378929138,
+      "learning_rate": 1.5535521677000813e-07,
+      "loss": 0.0538,
+      "step": 16490
+    },
+    {
+      "epoch": 2.671905379131562,
+      "grad_norm": 0.9660282135009766,
+      "learning_rate": 1.5520349959022934e-07,
+      "loss": 0.0582,
+      "step": 16491
+    },
+    {
+      "epoch": 2.672067401166559,
+      "grad_norm": 0.9266077876091003,
+      "learning_rate": 1.5505185415554903e-07,
+      "loss": 0.063,
+      "step": 16492
+    },
+    {
+      "epoch": 2.6722294232015553,
+      "grad_norm": 0.9693909287452698,
+      "learning_rate": 1.5490028047060762e-07,
+      "loss": 0.0612,
+      "step": 16493
+    },
+    {
+      "epoch": 2.6723914452365523,
+      "grad_norm": 0.9419223666191101,
+      "learning_rate": 1.54748778540042e-07,
+      "loss": 0.0568,
+      "step": 16494
+    },
+    {
+      "epoch": 2.672553467271549,
+      "grad_norm": 0.9191377758979797,
+      "learning_rate": 1.5459734836848872e-07,
+      "loss": 0.0631,
+      "step": 16495
+    },
+    {
+      "epoch": 2.6727154893065457,
+      "grad_norm": 1.1649914979934692,
+      "learning_rate": 1.544459899605813e-07,
+      "loss": 0.063,
+      "step": 16496
+    },
+    {
+      "epoch": 2.6728775113415426,
+      "grad_norm": 0.7725280523300171,
+      "learning_rate": 1.5429470332094992e-07,
+      "loss": 0.0554,
+      "step": 16497
+    },
+    {
+      "epoch": 2.673039533376539,
+      "grad_norm": 0.9325549006462097,
+      "learning_rate": 1.5414348845422394e-07,
+      "loss": 0.0567,
+      "step": 16498
+    },
+    {
+      "epoch": 2.673201555411536,
+      "grad_norm": 1.1128214597702026,
+      "learning_rate": 1.5399234536503023e-07,
+      "loss": 0.0684,
+      "step": 16499
+    },
+    {
+      "epoch": 2.6733635774465325,
+      "grad_norm": 0.967528223991394,
+      "learning_rate": 1.538412740579942e-07,
+      "loss": 0.0644,
+      "step": 16500
+    },
+    {
+      "epoch": 2.6735255994815295,
+      "grad_norm": 0.893129825592041,
+      "learning_rate": 1.536902745377372e-07,
+      "loss": 0.0641,
+      "step": 16501
+    },
+    {
+      "epoch": 2.6736876215165264,
+      "grad_norm": 1.0466177463531494,
+      "learning_rate": 1.5353934680888e-07,
+      "loss": 0.0636,
+      "step": 16502
+    },
+    {
+      "epoch": 2.673849643551523,
+      "grad_norm": 1.234455943107605,
+      "learning_rate": 1.5338849087604025e-07,
+      "loss": 0.0587,
+      "step": 16503
+    },
+    {
+      "epoch": 2.67401166558652,
+      "grad_norm": 0.9519467353820801,
+      "learning_rate": 1.5323770674383398e-07,
+      "loss": 0.0616,
+      "step": 16504
+    },
+    {
+      "epoch": 2.6741736876215167,
+      "grad_norm": 0.8917698860168457,
+      "learning_rate": 1.5308699441687502e-07,
+      "loss": 0.0564,
+      "step": 16505
+    },
+    {
+      "epoch": 2.6743357096565132,
+      "grad_norm": 0.8450326919555664,
+      "learning_rate": 1.529363538997744e-07,
+      "loss": 0.0553,
+      "step": 16506
+    },
+    {
+      "epoch": 2.67449773169151,
+      "grad_norm": 0.9240355491638184,
+      "learning_rate": 1.5278578519714177e-07,
+      "loss": 0.063,
+      "step": 16507
+    },
+    {
+      "epoch": 2.6746597537265067,
+      "grad_norm": 0.9043802618980408,
+      "learning_rate": 1.526352883135837e-07,
+      "loss": 0.0639,
+      "step": 16508
+    },
+    {
+      "epoch": 2.6748217757615036,
+      "grad_norm": 0.8774120807647705,
+      "learning_rate": 1.5248486325370544e-07,
+      "loss": 0.0609,
+      "step": 16509
+    },
+    {
+      "epoch": 2.6749837977965,
+      "grad_norm": 0.9532600045204163,
+      "learning_rate": 1.5233451002210964e-07,
+      "loss": 0.0594,
+      "step": 16510
+    },
+    {
+      "epoch": 2.675145819831497,
+      "grad_norm": 0.8727222681045532,
+      "learning_rate": 1.5218422862339683e-07,
+      "loss": 0.0651,
+      "step": 16511
+    },
+    {
+      "epoch": 2.675307841866494,
+      "grad_norm": 0.8903719186782837,
+      "learning_rate": 1.520340190621647e-07,
+      "loss": 0.0586,
+      "step": 16512
+    },
+    {
+      "epoch": 2.6754698639014904,
+      "grad_norm": 1.0103119611740112,
+      "learning_rate": 1.5188388134300901e-07,
+      "loss": 0.0592,
+      "step": 16513
+    },
+    {
+      "epoch": 2.6756318859364874,
+      "grad_norm": 0.8567571043968201,
+      "learning_rate": 1.5173381547052528e-07,
+      "loss": 0.0577,
+      "step": 16514
+    },
+    {
+      "epoch": 2.6757939079714843,
+      "grad_norm": 0.874523401260376,
+      "learning_rate": 1.5158382144930344e-07,
+      "loss": 0.0564,
+      "step": 16515
+    },
+    {
+      "epoch": 2.675955930006481,
+      "grad_norm": 0.8860069513320923,
+      "learning_rate": 1.5143389928393398e-07,
+      "loss": 0.0573,
+      "step": 16516
+    },
+    {
+      "epoch": 2.6761179520414777,
+      "grad_norm": 0.8461055159568787,
+      "learning_rate": 1.512840489790035e-07,
+      "loss": 0.0578,
+      "step": 16517
+    },
+    {
+      "epoch": 2.6762799740764747,
+      "grad_norm": 0.9307635426521301,
+      "learning_rate": 1.5113427053909725e-07,
+      "loss": 0.0589,
+      "step": 16518
+    },
+    {
+      "epoch": 2.676441996111471,
+      "grad_norm": 0.9444718360900879,
+      "learning_rate": 1.5098456396879846e-07,
+      "loss": 0.0627,
+      "step": 16519
+    },
+    {
+      "epoch": 2.6766040181464676,
+      "grad_norm": 0.8695940375328064,
+      "learning_rate": 1.508349292726874e-07,
+      "loss": 0.0566,
+      "step": 16520
+    },
+    {
+      "epoch": 2.6767660401814646,
+      "grad_norm": 0.9241846203804016,
+      "learning_rate": 1.506853664553426e-07,
+      "loss": 0.0509,
+      "step": 16521
+    },
+    {
+      "epoch": 2.6769280622164615,
+      "grad_norm": 0.9394335150718689,
+      "learning_rate": 1.505358755213407e-07,
+      "loss": 0.0598,
+      "step": 16522
+    },
+    {
+      "epoch": 2.677090084251458,
+      "grad_norm": 0.865931510925293,
+      "learning_rate": 1.503864564752547e-07,
+      "loss": 0.0542,
+      "step": 16523
+    },
+    {
+      "epoch": 2.677252106286455,
+      "grad_norm": 0.9105206727981567,
+      "learning_rate": 1.5023710932165758e-07,
+      "loss": 0.051,
+      "step": 16524
+    },
+    {
+      "epoch": 2.677414128321452,
+      "grad_norm": 0.8169817328453064,
+      "learning_rate": 1.500878340651185e-07,
+      "loss": 0.0584,
+      "step": 16525
+    },
+    {
+      "epoch": 2.6775761503564484,
+      "grad_norm": 0.8245558142662048,
+      "learning_rate": 1.4993863071020548e-07,
+      "loss": 0.0581,
+      "step": 16526
+    },
+    {
+      "epoch": 2.6777381723914453,
+      "grad_norm": 0.9422661662101746,
+      "learning_rate": 1.4978949926148288e-07,
+      "loss": 0.0599,
+      "step": 16527
+    },
+    {
+      "epoch": 2.6779001944264422,
+      "grad_norm": 1.0110220909118652,
+      "learning_rate": 1.4964043972351377e-07,
+      "loss": 0.0624,
+      "step": 16528
+    },
+    {
+      "epoch": 2.6780622164614387,
+      "grad_norm": 1.0897401571273804,
+      "learning_rate": 1.494914521008603e-07,
+      "loss": 0.0539,
+      "step": 16529
+    },
+    {
+      "epoch": 2.6782242384964356,
+      "grad_norm": 0.8992266654968262,
+      "learning_rate": 1.4934253639807994e-07,
+      "loss": 0.0606,
+      "step": 16530
+    },
+    {
+      "epoch": 2.678386260531432,
+      "grad_norm": 0.9721251130104065,
+      "learning_rate": 1.4919369261972933e-07,
+      "loss": 0.0646,
+      "step": 16531
+    },
+    {
+      "epoch": 2.678548282566429,
+      "grad_norm": 0.8682243824005127,
+      "learning_rate": 1.4904492077036286e-07,
+      "loss": 0.0566,
+      "step": 16532
+    },
+    {
+      "epoch": 2.6787103046014256,
+      "grad_norm": 0.7980409860610962,
+      "learning_rate": 1.4889622085453304e-07,
+      "loss": 0.0563,
+      "step": 16533
+    },
+    {
+      "epoch": 2.6788723266364225,
+      "grad_norm": 0.8528122901916504,
+      "learning_rate": 1.4874759287678898e-07,
+      "loss": 0.0609,
+      "step": 16534
+    },
+    {
+      "epoch": 2.6790343486714194,
+      "grad_norm": 0.7880252599716187,
+      "learning_rate": 1.485990368416787e-07,
+      "loss": 0.0584,
+      "step": 16535
+    },
+    {
+      "epoch": 2.679196370706416,
+      "grad_norm": 1.048828363418579,
+      "learning_rate": 1.484505527537475e-07,
+      "loss": 0.0558,
+      "step": 16536
+    },
+    {
+      "epoch": 2.679358392741413,
+      "grad_norm": 0.9639270305633545,
+      "learning_rate": 1.483021406175389e-07,
+      "loss": 0.0636,
+      "step": 16537
+    },
+    {
+      "epoch": 2.67952041477641,
+      "grad_norm": 0.9279593825340271,
+      "learning_rate": 1.4815380043759374e-07,
+      "loss": 0.0578,
+      "step": 16538
+    },
+    {
+      "epoch": 2.6796824368114063,
+      "grad_norm": 0.8325791954994202,
+      "learning_rate": 1.4800553221845094e-07,
+      "loss": 0.0565,
+      "step": 16539
+    },
+    {
+      "epoch": 2.679844458846403,
+      "grad_norm": 0.8952010273933411,
+      "learning_rate": 1.4785733596464736e-07,
+      "loss": 0.0601,
+      "step": 16540
+    },
+    {
+      "epoch": 2.6800064808814,
+      "grad_norm": 0.8639797568321228,
+      "learning_rate": 1.4770921168071717e-07,
+      "loss": 0.0601,
+      "step": 16541
+    },
+    {
+      "epoch": 2.6801685029163966,
+      "grad_norm": 1.0438389778137207,
+      "learning_rate": 1.4756115937119202e-07,
+      "loss": 0.0681,
+      "step": 16542
+    },
+    {
+      "epoch": 2.680330524951393,
+      "grad_norm": 0.9621492028236389,
+      "learning_rate": 1.4741317904060304e-07,
+      "loss": 0.06,
+      "step": 16543
+    },
+    {
+      "epoch": 2.68049254698639,
+      "grad_norm": 0.814268946647644,
+      "learning_rate": 1.4726527069347796e-07,
+      "loss": 0.0555,
+      "step": 16544
+    },
+    {
+      "epoch": 2.680654569021387,
+      "grad_norm": 0.8960631489753723,
+      "learning_rate": 1.471174343343418e-07,
+      "loss": 0.0576,
+      "step": 16545
+    },
+    {
+      "epoch": 2.6808165910563835,
+      "grad_norm": 0.9540743827819824,
+      "learning_rate": 1.4696966996771838e-07,
+      "loss": 0.0691,
+      "step": 16546
+    },
+    {
+      "epoch": 2.6809786130913804,
+      "grad_norm": 1.125815987586975,
+      "learning_rate": 1.468219775981286e-07,
+      "loss": 0.0665,
+      "step": 16547
+    },
+    {
+      "epoch": 2.6811406351263773,
+      "grad_norm": 0.8713697195053101,
+      "learning_rate": 1.4667435723009187e-07,
+      "loss": 0.0554,
+      "step": 16548
+    },
+    {
+      "epoch": 2.681302657161374,
+      "grad_norm": 0.961420476436615,
+      "learning_rate": 1.4652680886812488e-07,
+      "loss": 0.0615,
+      "step": 16549
+    },
+    {
+      "epoch": 2.6814646791963708,
+      "grad_norm": 0.8258307576179504,
+      "learning_rate": 1.463793325167423e-07,
+      "loss": 0.0602,
+      "step": 16550
+    },
+    {
+      "epoch": 2.6816267012313677,
+      "grad_norm": 0.946225106716156,
+      "learning_rate": 1.4623192818045638e-07,
+      "loss": 0.0613,
+      "step": 16551
+    },
+    {
+      "epoch": 2.681788723266364,
+      "grad_norm": 1.008615255355835,
+      "learning_rate": 1.4608459586377743e-07,
+      "loss": 0.0708,
+      "step": 16552
+    },
+    {
+      "epoch": 2.681950745301361,
+      "grad_norm": 1.0377353429794312,
+      "learning_rate": 1.4593733557121347e-07,
+      "loss": 0.0649,
+      "step": 16553
+    },
+    {
+      "epoch": 2.6821127673363576,
+      "grad_norm": 0.9528126120567322,
+      "learning_rate": 1.4579014730727037e-07,
+      "loss": 0.0641,
+      "step": 16554
+    },
+    {
+      "epoch": 2.6822747893713546,
+      "grad_norm": 0.9190387725830078,
+      "learning_rate": 1.45643031076452e-07,
+      "loss": 0.0629,
+      "step": 16555
+    },
+    {
+      "epoch": 2.682436811406351,
+      "grad_norm": 0.855954647064209,
+      "learning_rate": 1.4549598688325896e-07,
+      "loss": 0.0566,
+      "step": 16556
+    },
+    {
+      "epoch": 2.682598833441348,
+      "grad_norm": 0.9234669208526611,
+      "learning_rate": 1.4534901473219093e-07,
+      "loss": 0.0671,
+      "step": 16557
+    },
+    {
+      "epoch": 2.682760855476345,
+      "grad_norm": 1.030083179473877,
+      "learning_rate": 1.4520211462774548e-07,
+      "loss": 0.0586,
+      "step": 16558
+    },
+    {
+      "epoch": 2.6829228775113414,
+      "grad_norm": 1.0558991432189941,
+      "learning_rate": 1.4505528657441648e-07,
+      "loss": 0.0627,
+      "step": 16559
+    },
+    {
+      "epoch": 2.6830848995463383,
+      "grad_norm": 1.0315167903900146,
+      "learning_rate": 1.4490853057669675e-07,
+      "loss": 0.061,
+      "step": 16560
+    },
+    {
+      "epoch": 2.6832469215813353,
+      "grad_norm": 0.8899813890457153,
+      "learning_rate": 1.4476184663907628e-07,
+      "loss": 0.0682,
+      "step": 16561
+    },
+    {
+      "epoch": 2.6834089436163318,
+      "grad_norm": 0.787449300289154,
+      "learning_rate": 1.4461523476604482e-07,
+      "loss": 0.0506,
+      "step": 16562
+    },
+    {
+      "epoch": 2.6835709656513287,
+      "grad_norm": 0.8441698551177979,
+      "learning_rate": 1.444686949620866e-07,
+      "loss": 0.0617,
+      "step": 16563
+    },
+    {
+      "epoch": 2.683732987686325,
+      "grad_norm": 0.8496869206428528,
+      "learning_rate": 1.4432222723168632e-07,
+      "loss": 0.0619,
+      "step": 16564
+    },
+    {
+      "epoch": 2.683895009721322,
+      "grad_norm": 1.025014042854309,
+      "learning_rate": 1.4417583157932485e-07,
+      "loss": 0.0568,
+      "step": 16565
+    },
+    {
+      "epoch": 2.6840570317563186,
+      "grad_norm": 0.9680132269859314,
+      "learning_rate": 1.4402950800948223e-07,
+      "loss": 0.0626,
+      "step": 16566
+    },
+    {
+      "epoch": 2.6842190537913155,
+      "grad_norm": 0.7993006110191345,
+      "learning_rate": 1.4388325652663542e-07,
+      "loss": 0.0592,
+      "step": 16567
+    },
+    {
+      "epoch": 2.6843810758263125,
+      "grad_norm": 0.8684203028678894,
+      "learning_rate": 1.437370771352589e-07,
+      "loss": 0.0611,
+      "step": 16568
+    },
+    {
+      "epoch": 2.684543097861309,
+      "grad_norm": 0.9955109357833862,
+      "learning_rate": 1.4359096983982607e-07,
+      "loss": 0.0586,
+      "step": 16569
+    },
+    {
+      "epoch": 2.684705119896306,
+      "grad_norm": 0.884691059589386,
+      "learning_rate": 1.4344493464480745e-07,
+      "loss": 0.0633,
+      "step": 16570
+    },
+    {
+      "epoch": 2.684867141931303,
+      "grad_norm": 0.8436540961265564,
+      "learning_rate": 1.4329897155467039e-07,
+      "loss": 0.0623,
+      "step": 16571
+    },
+    {
+      "epoch": 2.6850291639662993,
+      "grad_norm": 1.0994386672973633,
+      "learning_rate": 1.4315308057388206e-07,
+      "loss": 0.0638,
+      "step": 16572
+    },
+    {
+      "epoch": 2.6851911860012962,
+      "grad_norm": 0.9540835022926331,
+      "learning_rate": 1.4300726170690614e-07,
+      "loss": 0.056,
+      "step": 16573
+    },
+    {
+      "epoch": 2.685353208036293,
+      "grad_norm": 0.9536152482032776,
+      "learning_rate": 1.428615149582041e-07,
+      "loss": 0.0571,
+      "step": 16574
+    },
+    {
+      "epoch": 2.6855152300712897,
+      "grad_norm": 0.7673203349113464,
+      "learning_rate": 1.4271584033223512e-07,
+      "loss": 0.0578,
+      "step": 16575
+    },
+    {
+      "epoch": 2.6856772521062866,
+      "grad_norm": 0.899626612663269,
+      "learning_rate": 1.425702378334573e-07,
+      "loss": 0.0572,
+      "step": 16576
+    },
+    {
+      "epoch": 2.685839274141283,
+      "grad_norm": 0.9618022441864014,
+      "learning_rate": 1.4242470746632542e-07,
+      "loss": 0.0644,
+      "step": 16577
+    },
+    {
+      "epoch": 2.68600129617628,
+      "grad_norm": 0.9150751233100891,
+      "learning_rate": 1.4227924923529228e-07,
+      "loss": 0.0629,
+      "step": 16578
+    },
+    {
+      "epoch": 2.6861633182112765,
+      "grad_norm": 1.083203911781311,
+      "learning_rate": 1.4213386314480825e-07,
+      "loss": 0.0701,
+      "step": 16579
+    },
+    {
+      "epoch": 2.6863253402462735,
+      "grad_norm": 0.9395116567611694,
+      "learning_rate": 1.4198854919932225e-07,
+      "loss": 0.0624,
+      "step": 16580
+    },
+    {
+      "epoch": 2.6864873622812704,
+      "grad_norm": 0.8851693868637085,
+      "learning_rate": 1.4184330740328044e-07,
+      "loss": 0.0657,
+      "step": 16581
+    },
+    {
+      "epoch": 2.686649384316267,
+      "grad_norm": 0.8476728200912476,
+      "learning_rate": 1.4169813776112652e-07,
+      "loss": 0.0589,
+      "step": 16582
+    },
+    {
+      "epoch": 2.686811406351264,
+      "grad_norm": 1.0356652736663818,
+      "learning_rate": 1.4155304027730271e-07,
+      "loss": 0.0627,
+      "step": 16583
+    },
+    {
+      "epoch": 2.6869734283862607,
+      "grad_norm": 0.8928619623184204,
+      "learning_rate": 1.4140801495624913e-07,
+      "loss": 0.0626,
+      "step": 16584
+    },
+    {
+      "epoch": 2.6871354504212572,
+      "grad_norm": 1.1158393621444702,
+      "learning_rate": 1.412630618024016e-07,
+      "loss": 0.0641,
+      "step": 16585
+    },
+    {
+      "epoch": 2.687297472456254,
+      "grad_norm": 1.0287305116653442,
+      "learning_rate": 1.4111818082019696e-07,
+      "loss": 0.065,
+      "step": 16586
+    },
+    {
+      "epoch": 2.6874594944912507,
+      "grad_norm": 0.8559868931770325,
+      "learning_rate": 1.4097337201406742e-07,
+      "loss": 0.0576,
+      "step": 16587
+    },
+    {
+      "epoch": 2.6876215165262476,
+      "grad_norm": 0.9814032912254333,
+      "learning_rate": 1.4082863538844444e-07,
+      "loss": 0.0598,
+      "step": 16588
+    },
+    {
+      "epoch": 2.687783538561244,
+      "grad_norm": 1.0753200054168701,
+      "learning_rate": 1.406839709477556e-07,
+      "loss": 0.0605,
+      "step": 16589
+    },
+    {
+      "epoch": 2.687945560596241,
+      "grad_norm": 0.9408139586448669,
+      "learning_rate": 1.4053937869642737e-07,
+      "loss": 0.0624,
+      "step": 16590
+    },
+    {
+      "epoch": 2.688107582631238,
+      "grad_norm": 0.8320403695106506,
+      "learning_rate": 1.4039485863888537e-07,
+      "loss": 0.0557,
+      "step": 16591
+    },
+    {
+      "epoch": 2.6882696046662344,
+      "grad_norm": 0.9615926146507263,
+      "learning_rate": 1.402504107795502e-07,
+      "loss": 0.0627,
+      "step": 16592
+    },
+    {
+      "epoch": 2.6884316267012314,
+      "grad_norm": 0.8587374091148376,
+      "learning_rate": 1.401060351228417e-07,
+      "loss": 0.0631,
+      "step": 16593
+    },
+    {
+      "epoch": 2.6885936487362283,
+      "grad_norm": 0.8640631437301636,
+      "learning_rate": 1.39961731673178e-07,
+      "loss": 0.0605,
+      "step": 16594
+    },
+    {
+      "epoch": 2.688755670771225,
+      "grad_norm": 0.8059629797935486,
+      "learning_rate": 1.398175004349739e-07,
+      "loss": 0.0556,
+      "step": 16595
+    },
+    {
+      "epoch": 2.6889176928062217,
+      "grad_norm": 0.9349533319473267,
+      "learning_rate": 1.3967334141264277e-07,
+      "loss": 0.0646,
+      "step": 16596
+    },
+    {
+      "epoch": 2.6890797148412187,
+      "grad_norm": 0.9726789593696594,
+      "learning_rate": 1.3952925461059558e-07,
+      "loss": 0.0594,
+      "step": 16597
+    },
+    {
+      "epoch": 2.689241736876215,
+      "grad_norm": 0.9107815623283386,
+      "learning_rate": 1.39385240033241e-07,
+      "loss": 0.0614,
+      "step": 16598
+    },
+    {
+      "epoch": 2.689403758911212,
+      "grad_norm": 0.8884726762771606,
+      "learning_rate": 1.392412976849855e-07,
+      "loss": 0.0643,
+      "step": 16599
+    },
+    {
+      "epoch": 2.6895657809462086,
+      "grad_norm": 0.8729420900344849,
+      "learning_rate": 1.3909742757023336e-07,
+      "loss": 0.0585,
+      "step": 16600
+    },
+    {
+      "epoch": 2.6897278029812055,
+      "grad_norm": 0.8863007426261902,
+      "learning_rate": 1.3895362969338662e-07,
+      "loss": 0.0613,
+      "step": 16601
+    },
+    {
+      "epoch": 2.689889825016202,
+      "grad_norm": 0.9088746309280396,
+      "learning_rate": 1.3880990405884532e-07,
+      "loss": 0.056,
+      "step": 16602
+    },
+    {
+      "epoch": 2.690051847051199,
+      "grad_norm": 0.9081185460090637,
+      "learning_rate": 1.3866625067100707e-07,
+      "loss": 0.0599,
+      "step": 16603
+    },
+    {
+      "epoch": 2.690213869086196,
+      "grad_norm": 0.9388935565948486,
+      "learning_rate": 1.3852266953426674e-07,
+      "loss": 0.0642,
+      "step": 16604
+    },
+    {
+      "epoch": 2.6903758911211924,
+      "grad_norm": 0.7939647436141968,
+      "learning_rate": 1.3837916065301827e-07,
+      "loss": 0.0603,
+      "step": 16605
+    },
+    {
+      "epoch": 2.6905379131561893,
+      "grad_norm": 0.8875724077224731,
+      "learning_rate": 1.3823572403165285e-07,
+      "loss": 0.0599,
+      "step": 16606
+    },
+    {
+      "epoch": 2.690699935191186,
+      "grad_norm": 0.7508094906806946,
+      "learning_rate": 1.380923596745587e-07,
+      "loss": 0.0548,
+      "step": 16607
+    },
+    {
+      "epoch": 2.6908619572261827,
+      "grad_norm": 0.8791196942329407,
+      "learning_rate": 1.3794906758612252e-07,
+      "loss": 0.0589,
+      "step": 16608
+    },
+    {
+      "epoch": 2.6910239792611796,
+      "grad_norm": 0.9110643267631531,
+      "learning_rate": 1.3780584777072892e-07,
+      "loss": 0.061,
+      "step": 16609
+    },
+    {
+      "epoch": 2.691186001296176,
+      "grad_norm": 0.9413241744041443,
+      "learning_rate": 1.376627002327599e-07,
+      "loss": 0.0639,
+      "step": 16610
+    },
+    {
+      "epoch": 2.691348023331173,
+      "grad_norm": 0.9628585577011108,
+      "learning_rate": 1.375196249765956e-07,
+      "loss": 0.0696,
+      "step": 16611
+    },
+    {
+      "epoch": 2.6915100453661696,
+      "grad_norm": 0.8341220021247864,
+      "learning_rate": 1.373766220066136e-07,
+      "loss": 0.0596,
+      "step": 16612
+    },
+    {
+      "epoch": 2.6916720674011665,
+      "grad_norm": 0.8509212732315063,
+      "learning_rate": 1.372336913271896e-07,
+      "loss": 0.0554,
+      "step": 16613
+    },
+    {
+      "epoch": 2.6918340894361634,
+      "grad_norm": 0.8546708822250366,
+      "learning_rate": 1.3709083294269676e-07,
+      "loss": 0.0609,
+      "step": 16614
+    },
+    {
+      "epoch": 2.69199611147116,
+      "grad_norm": 0.9867240190505981,
+      "learning_rate": 1.369480468575063e-07,
+      "loss": 0.0511,
+      "step": 16615
+    },
+    {
+      "epoch": 2.692158133506157,
+      "grad_norm": 1.0250896215438843,
+      "learning_rate": 1.368053330759872e-07,
+      "loss": 0.0635,
+      "step": 16616
+    },
+    {
+      "epoch": 2.692320155541154,
+      "grad_norm": 1.0135303735733032,
+      "learning_rate": 1.366626916025063e-07,
+      "loss": 0.0621,
+      "step": 16617
+    },
+    {
+      "epoch": 2.6924821775761503,
+      "grad_norm": 0.852816104888916,
+      "learning_rate": 1.3652012244142754e-07,
+      "loss": 0.0576,
+      "step": 16618
+    },
+    {
+      "epoch": 2.692644199611147,
+      "grad_norm": 0.8429510593414307,
+      "learning_rate": 1.363776255971133e-07,
+      "loss": 0.0597,
+      "step": 16619
+    },
+    {
+      "epoch": 2.692806221646144,
+      "grad_norm": 0.8124983906745911,
+      "learning_rate": 1.362352010739243e-07,
+      "loss": 0.0522,
+      "step": 16620
+    },
+    {
+      "epoch": 2.6929682436811406,
+      "grad_norm": 0.836234450340271,
+      "learning_rate": 1.360928488762181e-07,
+      "loss": 0.0579,
+      "step": 16621
+    },
+    {
+      "epoch": 2.693130265716137,
+      "grad_norm": 0.9209657907485962,
+      "learning_rate": 1.3595056900834986e-07,
+      "loss": 0.0608,
+      "step": 16622
+    },
+    {
+      "epoch": 2.693292287751134,
+      "grad_norm": 0.9225515127182007,
+      "learning_rate": 1.3580836147467304e-07,
+      "loss": 0.0635,
+      "step": 16623
+    },
+    {
+      "epoch": 2.693454309786131,
+      "grad_norm": 0.9024338722229004,
+      "learning_rate": 1.3566622627953968e-07,
+      "loss": 0.0602,
+      "step": 16624
+    },
+    {
+      "epoch": 2.6936163318211275,
+      "grad_norm": 0.959028959274292,
+      "learning_rate": 1.3552416342729802e-07,
+      "loss": 0.0628,
+      "step": 16625
+    },
+    {
+      "epoch": 2.6937783538561244,
+      "grad_norm": 0.917083203792572,
+      "learning_rate": 1.3538217292229482e-07,
+      "loss": 0.0589,
+      "step": 16626
+    },
+    {
+      "epoch": 2.6939403758911213,
+      "grad_norm": 0.9200828671455383,
+      "learning_rate": 1.3524025476887527e-07,
+      "loss": 0.0577,
+      "step": 16627
+    },
+    {
+      "epoch": 2.694102397926118,
+      "grad_norm": 1.1384090185165405,
+      "learning_rate": 1.3509840897138083e-07,
+      "loss": 0.0641,
+      "step": 16628
+    },
+    {
+      "epoch": 2.6942644199611148,
+      "grad_norm": 0.9467897415161133,
+      "learning_rate": 1.349566355341525e-07,
+      "loss": 0.0635,
+      "step": 16629
+    },
+    {
+      "epoch": 2.6944264419961117,
+      "grad_norm": 0.8724448680877686,
+      "learning_rate": 1.3481493446152766e-07,
+      "loss": 0.0598,
+      "step": 16630
+    },
+    {
+      "epoch": 2.694588464031108,
+      "grad_norm": 0.9046519994735718,
+      "learning_rate": 1.3467330575784226e-07,
+      "loss": 0.0619,
+      "step": 16631
+    },
+    {
+      "epoch": 2.694750486066105,
+      "grad_norm": 0.7974745631217957,
+      "learning_rate": 1.3453174942743008e-07,
+      "loss": 0.0567,
+      "step": 16632
+    },
+    {
+      "epoch": 2.6949125081011016,
+      "grad_norm": 0.8849309682846069,
+      "learning_rate": 1.3439026547462126e-07,
+      "loss": 0.0636,
+      "step": 16633
+    },
+    {
+      "epoch": 2.6950745301360985,
+      "grad_norm": 1.4565906524658203,
+      "learning_rate": 1.3424885390374593e-07,
+      "loss": 0.0653,
+      "step": 16634
+    },
+    {
+      "epoch": 2.695236552171095,
+      "grad_norm": 0.9429144859313965,
+      "learning_rate": 1.341075147191312e-07,
+      "loss": 0.0632,
+      "step": 16635
+    },
+    {
+      "epoch": 2.695398574206092,
+      "grad_norm": 0.8843449354171753,
+      "learning_rate": 1.3396624792510082e-07,
+      "loss": 0.0587,
+      "step": 16636
+    },
+    {
+      "epoch": 2.695560596241089,
+      "grad_norm": 0.84376460313797,
+      "learning_rate": 1.3382505352597747e-07,
+      "loss": 0.0592,
+      "step": 16637
+    },
+    {
+      "epoch": 2.6957226182760854,
+      "grad_norm": 0.9161346554756165,
+      "learning_rate": 1.33683931526081e-07,
+      "loss": 0.0644,
+      "step": 16638
+    },
+    {
+      "epoch": 2.6958846403110823,
+      "grad_norm": 0.9942044019699097,
+      "learning_rate": 1.3354288192973074e-07,
+      "loss": 0.0626,
+      "step": 16639
+    },
+    {
+      "epoch": 2.6960466623460793,
+      "grad_norm": 0.9702102541923523,
+      "learning_rate": 1.3340190474124104e-07,
+      "loss": 0.0654,
+      "step": 16640
+    },
+    {
+      "epoch": 2.6962086843810757,
+      "grad_norm": 0.9594953060150146,
+      "learning_rate": 1.3326099996492618e-07,
+      "loss": 0.0608,
+      "step": 16641
+    },
+    {
+      "epoch": 2.6963707064160727,
+      "grad_norm": 0.9994754195213318,
+      "learning_rate": 1.3312016760509722e-07,
+      "loss": 0.0606,
+      "step": 16642
+    },
+    {
+      "epoch": 2.6965327284510696,
+      "grad_norm": 0.8460058569908142,
+      "learning_rate": 1.3297940766606344e-07,
+      "loss": 0.0541,
+      "step": 16643
+    },
+    {
+      "epoch": 2.696694750486066,
+      "grad_norm": 0.9044547080993652,
+      "learning_rate": 1.3283872015213168e-07,
+      "loss": 0.0607,
+      "step": 16644
+    },
+    {
+      "epoch": 2.6968567725210626,
+      "grad_norm": 0.807353675365448,
+      "learning_rate": 1.3269810506760683e-07,
+      "loss": 0.0583,
+      "step": 16645
+    },
+    {
+      "epoch": 2.6970187945560595,
+      "grad_norm": 1.019974708557129,
+      "learning_rate": 1.3255756241679102e-07,
+      "loss": 0.0644,
+      "step": 16646
+    },
+    {
+      "epoch": 2.6971808165910565,
+      "grad_norm": 0.818080723285675,
+      "learning_rate": 1.3241709220398467e-07,
+      "loss": 0.0579,
+      "step": 16647
+    },
+    {
+      "epoch": 2.697342838626053,
+      "grad_norm": 0.959563672542572,
+      "learning_rate": 1.3227669443348578e-07,
+      "loss": 0.0625,
+      "step": 16648
+    },
+    {
+      "epoch": 2.69750486066105,
+      "grad_norm": 0.9866206645965576,
+      "learning_rate": 1.321363691095906e-07,
+      "loss": 0.0584,
+      "step": 16649
+    },
+    {
+      "epoch": 2.697666882696047,
+      "grad_norm": 1.2002931833267212,
+      "learning_rate": 1.3199611623659235e-07,
+      "loss": 0.0728,
+      "step": 16650
+    },
+    {
+      "epoch": 2.6978289047310433,
+      "grad_norm": 0.8130374550819397,
+      "learning_rate": 1.3185593581878238e-07,
+      "loss": 0.0582,
+      "step": 16651
+    },
+    {
+      "epoch": 2.6979909267660402,
+      "grad_norm": 0.883966863155365,
+      "learning_rate": 1.3171582786044968e-07,
+      "loss": 0.0638,
+      "step": 16652
+    },
+    {
+      "epoch": 2.698152948801037,
+      "grad_norm": 1.0147669315338135,
+      "learning_rate": 1.3157579236588197e-07,
+      "loss": 0.0619,
+      "step": 16653
+    },
+    {
+      "epoch": 2.6983149708360337,
+      "grad_norm": 0.9203122854232788,
+      "learning_rate": 1.3143582933936333e-07,
+      "loss": 0.0581,
+      "step": 16654
+    },
+    {
+      "epoch": 2.6984769928710306,
+      "grad_norm": 0.9031851887702942,
+      "learning_rate": 1.3129593878517643e-07,
+      "loss": 0.0599,
+      "step": 16655
+    },
+    {
+      "epoch": 2.698639014906027,
+      "grad_norm": 0.8836839199066162,
+      "learning_rate": 1.3115612070760174e-07,
+      "loss": 0.0583,
+      "step": 16656
+    },
+    {
+      "epoch": 2.698801036941024,
+      "grad_norm": 0.9112008810043335,
+      "learning_rate": 1.3101637511091724e-07,
+      "loss": 0.0549,
+      "step": 16657
+    },
+    {
+      "epoch": 2.6989630589760205,
+      "grad_norm": 1.0826163291931152,
+      "learning_rate": 1.3087670199939894e-07,
+      "loss": 0.0597,
+      "step": 16658
+    },
+    {
+      "epoch": 2.6991250810110174,
+      "grad_norm": 0.8839635252952576,
+      "learning_rate": 1.3073710137732037e-07,
+      "loss": 0.0545,
+      "step": 16659
+    },
+    {
+      "epoch": 2.6992871030460144,
+      "grad_norm": 0.9589625000953674,
+      "learning_rate": 1.3059757324895283e-07,
+      "loss": 0.0638,
+      "step": 16660
+    },
+    {
+      "epoch": 2.699449125081011,
+      "grad_norm": 0.862845242023468,
+      "learning_rate": 1.3045811761856597e-07,
+      "loss": 0.0589,
+      "step": 16661
+    },
+    {
+      "epoch": 2.699611147116008,
+      "grad_norm": 0.8167102336883545,
+      "learning_rate": 1.303187344904261e-07,
+      "loss": 0.0543,
+      "step": 16662
+    },
+    {
+      "epoch": 2.6997731691510047,
+      "grad_norm": 1.040915608406067,
+      "learning_rate": 1.3017942386879867e-07,
+      "loss": 0.067,
+      "step": 16663
+    },
+    {
+      "epoch": 2.6999351911860012,
+      "grad_norm": 0.9312401413917542,
+      "learning_rate": 1.3004018575794586e-07,
+      "loss": 0.0596,
+      "step": 16664
+    },
+    {
+      "epoch": 2.700097213220998,
+      "grad_norm": 1.038403034210205,
+      "learning_rate": 1.2990102016212868e-07,
+      "loss": 0.0634,
+      "step": 16665
+    },
+    {
+      "epoch": 2.7002592352559946,
+      "grad_norm": 0.8530080914497375,
+      "learning_rate": 1.2976192708560432e-07,
+      "loss": 0.0642,
+      "step": 16666
+    },
+    {
+      "epoch": 2.7004212572909916,
+      "grad_norm": 1.1165688037872314,
+      "learning_rate": 1.2962290653262903e-07,
+      "loss": 0.0593,
+      "step": 16667
+    },
+    {
+      "epoch": 2.700583279325988,
+      "grad_norm": 0.9035430550575256,
+      "learning_rate": 1.2948395850745726e-07,
+      "loss": 0.0613,
+      "step": 16668
+    },
+    {
+      "epoch": 2.700745301360985,
+      "grad_norm": 0.8505802154541016,
+      "learning_rate": 1.293450830143392e-07,
+      "loss": 0.0609,
+      "step": 16669
+    },
+    {
+      "epoch": 2.700907323395982,
+      "grad_norm": 0.98024582862854,
+      "learning_rate": 1.29206280057525e-07,
+      "loss": 0.0601,
+      "step": 16670
+    },
+    {
+      "epoch": 2.7010693454309784,
+      "grad_norm": 0.8178975582122803,
+      "learning_rate": 1.2906754964126078e-07,
+      "loss": 0.0573,
+      "step": 16671
+    },
+    {
+      "epoch": 2.7012313674659754,
+      "grad_norm": 0.8594771027565002,
+      "learning_rate": 1.2892889176979284e-07,
+      "loss": 0.0607,
+      "step": 16672
+    },
+    {
+      "epoch": 2.7013933895009723,
+      "grad_norm": 0.9803131818771362,
+      "learning_rate": 1.2879030644736252e-07,
+      "loss": 0.0643,
+      "step": 16673
+    },
+    {
+      "epoch": 2.701555411535969,
+      "grad_norm": 0.8199018836021423,
+      "learning_rate": 1.2865179367821083e-07,
+      "loss": 0.0536,
+      "step": 16674
+    },
+    {
+      "epoch": 2.7017174335709657,
+      "grad_norm": 0.8338936567306519,
+      "learning_rate": 1.2851335346657557e-07,
+      "loss": 0.0597,
+      "step": 16675
+    },
+    {
+      "epoch": 2.7018794556059627,
+      "grad_norm": 0.974616527557373,
+      "learning_rate": 1.283749858166927e-07,
+      "loss": 0.0615,
+      "step": 16676
+    },
+    {
+      "epoch": 2.702041477640959,
+      "grad_norm": 0.9886922240257263,
+      "learning_rate": 1.2823669073279615e-07,
+      "loss": 0.0644,
+      "step": 16677
+    },
+    {
+      "epoch": 2.702203499675956,
+      "grad_norm": 0.7480154037475586,
+      "learning_rate": 1.280984682191172e-07,
+      "loss": 0.0548,
+      "step": 16678
+    },
+    {
+      "epoch": 2.7023655217109526,
+      "grad_norm": 0.8866371512413025,
+      "learning_rate": 1.2796031827988582e-07,
+      "loss": 0.0576,
+      "step": 16679
+    },
+    {
+      "epoch": 2.7025275437459495,
+      "grad_norm": 1.1099352836608887,
+      "learning_rate": 1.2782224091932775e-07,
+      "loss": 0.0609,
+      "step": 16680
+    },
+    {
+      "epoch": 2.702689565780946,
+      "grad_norm": 0.902736246585846,
+      "learning_rate": 1.276842361416686e-07,
+      "loss": 0.0601,
+      "step": 16681
+    },
+    {
+      "epoch": 2.702851587815943,
+      "grad_norm": 1.0859075784683228,
+      "learning_rate": 1.2754630395113098e-07,
+      "loss": 0.0672,
+      "step": 16682
+    },
+    {
+      "epoch": 2.70301360985094,
+      "grad_norm": 0.9454689025878906,
+      "learning_rate": 1.2740844435193578e-07,
+      "loss": 0.0651,
+      "step": 16683
+    },
+    {
+      "epoch": 2.7031756318859363,
+      "grad_norm": 0.8466532230377197,
+      "learning_rate": 1.2727065734830013e-07,
+      "loss": 0.0515,
+      "step": 16684
+    },
+    {
+      "epoch": 2.7033376539209333,
+      "grad_norm": 0.9573439359664917,
+      "learning_rate": 1.271329429444404e-07,
+      "loss": 0.0619,
+      "step": 16685
+    },
+    {
+      "epoch": 2.70349967595593,
+      "grad_norm": 0.9744999408721924,
+      "learning_rate": 1.269953011445707e-07,
+      "loss": 0.0584,
+      "step": 16686
+    },
+    {
+      "epoch": 2.7036616979909267,
+      "grad_norm": 0.9962242841720581,
+      "learning_rate": 1.2685773195290186e-07,
+      "loss": 0.0582,
+      "step": 16687
+    },
+    {
+      "epoch": 2.7038237200259236,
+      "grad_norm": 0.9875708818435669,
+      "learning_rate": 1.267202353736438e-07,
+      "loss": 0.0621,
+      "step": 16688
+    },
+    {
+      "epoch": 2.70398574206092,
+      "grad_norm": 0.8407350778579712,
+      "learning_rate": 1.265828114110032e-07,
+      "loss": 0.0556,
+      "step": 16689
+    },
+    {
+      "epoch": 2.704147764095917,
+      "grad_norm": 0.9471036791801453,
+      "learning_rate": 1.26445460069185e-07,
+      "loss": 0.0601,
+      "step": 16690
+    },
+    {
+      "epoch": 2.7043097861309136,
+      "grad_norm": 0.739356279373169,
+      "learning_rate": 1.2630818135239198e-07,
+      "loss": 0.0503,
+      "step": 16691
+    },
+    {
+      "epoch": 2.7044718081659105,
+      "grad_norm": 1.1428637504577637,
+      "learning_rate": 1.2617097526482407e-07,
+      "loss": 0.0649,
+      "step": 16692
+    },
+    {
+      "epoch": 2.7046338302009074,
+      "grad_norm": 0.9683777689933777,
+      "learning_rate": 1.2603384181068018e-07,
+      "loss": 0.0627,
+      "step": 16693
+    },
+    {
+      "epoch": 2.704795852235904,
+      "grad_norm": 1.0176392793655396,
+      "learning_rate": 1.2589678099415582e-07,
+      "loss": 0.0653,
+      "step": 16694
+    },
+    {
+      "epoch": 2.704957874270901,
+      "grad_norm": 0.9063898921012878,
+      "learning_rate": 1.2575979281944429e-07,
+      "loss": 0.0545,
+      "step": 16695
+    },
+    {
+      "epoch": 2.7051198963058978,
+      "grad_norm": 1.0391916036605835,
+      "learning_rate": 1.256228772907378e-07,
+      "loss": 0.0706,
+      "step": 16696
+    },
+    {
+      "epoch": 2.7052819183408943,
+      "grad_norm": 0.9358698129653931,
+      "learning_rate": 1.254860344122255e-07,
+      "loss": 0.0651,
+      "step": 16697
+    },
+    {
+      "epoch": 2.705443940375891,
+      "grad_norm": 1.0807019472122192,
+      "learning_rate": 1.2534926418809433e-07,
+      "loss": 0.0649,
+      "step": 16698
+    },
+    {
+      "epoch": 2.705605962410888,
+      "grad_norm": 0.9569528102874756,
+      "learning_rate": 1.2521256662252902e-07,
+      "loss": 0.0634,
+      "step": 16699
+    },
+    {
+      "epoch": 2.7057679844458846,
+      "grad_norm": 0.9983630180358887,
+      "learning_rate": 1.2507594171971198e-07,
+      "loss": 0.0618,
+      "step": 16700
+    },
+    {
+      "epoch": 2.7059300064808816,
+      "grad_norm": 1.1226799488067627,
+      "learning_rate": 1.2493938948382468e-07,
+      "loss": 0.0627,
+      "step": 16701
+    },
+    {
+      "epoch": 2.706092028515878,
+      "grad_norm": 0.9132358431816101,
+      "learning_rate": 1.2480290991904398e-07,
+      "loss": 0.064,
+      "step": 16702
+    },
+    {
+      "epoch": 2.706254050550875,
+      "grad_norm": 0.8795627951622009,
+      "learning_rate": 1.246665030295463e-07,
+      "loss": 0.0611,
+      "step": 16703
+    },
+    {
+      "epoch": 2.7064160725858715,
+      "grad_norm": 0.8634214997291565,
+      "learning_rate": 1.245301688195058e-07,
+      "loss": 0.0563,
+      "step": 16704
+    },
+    {
+      "epoch": 2.7065780946208684,
+      "grad_norm": 0.9130164384841919,
+      "learning_rate": 1.243939072930933e-07,
+      "loss": 0.057,
+      "step": 16705
+    },
+    {
+      "epoch": 2.7067401166558653,
+      "grad_norm": 1.1923558712005615,
+      "learning_rate": 1.2425771845447853e-07,
+      "loss": 0.0607,
+      "step": 16706
+    },
+    {
+      "epoch": 2.706902138690862,
+      "grad_norm": 1.020593523979187,
+      "learning_rate": 1.2412160230782844e-07,
+      "loss": 0.0604,
+      "step": 16707
+    },
+    {
+      "epoch": 2.7070641607258588,
+      "grad_norm": 0.8900664448738098,
+      "learning_rate": 1.2398555885730774e-07,
+      "loss": 0.0539,
+      "step": 16708
+    },
+    {
+      "epoch": 2.7072261827608557,
+      "grad_norm": 0.9280899167060852,
+      "learning_rate": 1.2384958810707892e-07,
+      "loss": 0.0645,
+      "step": 16709
+    },
+    {
+      "epoch": 2.707388204795852,
+      "grad_norm": 0.9784387350082397,
+      "learning_rate": 1.2371369006130256e-07,
+      "loss": 0.063,
+      "step": 16710
+    },
+    {
+      "epoch": 2.707550226830849,
+      "grad_norm": 0.920939564704895,
+      "learning_rate": 1.2357786472413702e-07,
+      "loss": 0.0656,
+      "step": 16711
+    },
+    {
+      "epoch": 2.7077122488658456,
+      "grad_norm": 0.858536958694458,
+      "learning_rate": 1.2344211209973811e-07,
+      "loss": 0.0582,
+      "step": 16712
+    },
+    {
+      "epoch": 2.7078742709008425,
+      "grad_norm": 0.9796631336212158,
+      "learning_rate": 1.2330643219225918e-07,
+      "loss": 0.0593,
+      "step": 16713
+    },
+    {
+      "epoch": 2.708036292935839,
+      "grad_norm": 0.9091916084289551,
+      "learning_rate": 1.2317082500585163e-07,
+      "loss": 0.06,
+      "step": 16714
+    },
+    {
+      "epoch": 2.708198314970836,
+      "grad_norm": 1.0876365900039673,
+      "learning_rate": 1.2303529054466522e-07,
+      "loss": 0.0648,
+      "step": 16715
+    },
+    {
+      "epoch": 2.708360337005833,
+      "grad_norm": 0.9896219968795776,
+      "learning_rate": 1.2289982881284718e-07,
+      "loss": 0.0657,
+      "step": 16716
+    },
+    {
+      "epoch": 2.7085223590408294,
+      "grad_norm": 0.9894326329231262,
+      "learning_rate": 1.2276443981454167e-07,
+      "loss": 0.0645,
+      "step": 16717
+    },
+    {
+      "epoch": 2.7086843810758263,
+      "grad_norm": 0.9388425946235657,
+      "learning_rate": 1.226291235538915e-07,
+      "loss": 0.0582,
+      "step": 16718
+    },
+    {
+      "epoch": 2.7088464031108233,
+      "grad_norm": 0.9230713844299316,
+      "learning_rate": 1.22493880035037e-07,
+      "loss": 0.0647,
+      "step": 16719
+    },
+    {
+      "epoch": 2.7090084251458197,
+      "grad_norm": 0.8833425045013428,
+      "learning_rate": 1.223587092621162e-07,
+      "loss": 0.0634,
+      "step": 16720
+    },
+    {
+      "epoch": 2.7091704471808167,
+      "grad_norm": 0.9370494484901428,
+      "learning_rate": 1.2222361123926525e-07,
+      "loss": 0.0603,
+      "step": 16721
+    },
+    {
+      "epoch": 2.7093324692158136,
+      "grad_norm": 0.8519914150238037,
+      "learning_rate": 1.2208858597061752e-07,
+      "loss": 0.0599,
+      "step": 16722
+    },
+    {
+      "epoch": 2.70949449125081,
+      "grad_norm": 0.9887134432792664,
+      "learning_rate": 1.2195363346030497e-07,
+      "loss": 0.0671,
+      "step": 16723
+    },
+    {
+      "epoch": 2.709656513285807,
+      "grad_norm": 0.8368325233459473,
+      "learning_rate": 1.21818753712456e-07,
+      "loss": 0.0527,
+      "step": 16724
+    },
+    {
+      "epoch": 2.7098185353208035,
+      "grad_norm": 0.8991503119468689,
+      "learning_rate": 1.2168394673119837e-07,
+      "loss": 0.0631,
+      "step": 16725
+    },
+    {
+      "epoch": 2.7099805573558005,
+      "grad_norm": 0.9804664850234985,
+      "learning_rate": 1.2154921252065633e-07,
+      "loss": 0.0622,
+      "step": 16726
+    },
+    {
+      "epoch": 2.710142579390797,
+      "grad_norm": 0.8143778443336487,
+      "learning_rate": 1.2141455108495321e-07,
+      "loss": 0.0536,
+      "step": 16727
+    },
+    {
+      "epoch": 2.710304601425794,
+      "grad_norm": 0.934648334980011,
+      "learning_rate": 1.2127996242820822e-07,
+      "loss": 0.0581,
+      "step": 16728
+    },
+    {
+      "epoch": 2.710466623460791,
+      "grad_norm": 1.0055309534072876,
+      "learning_rate": 1.2114544655454002e-07,
+      "loss": 0.0675,
+      "step": 16729
+    },
+    {
+      "epoch": 2.7106286454957873,
+      "grad_norm": 0.9444065690040588,
+      "learning_rate": 1.2101100346806478e-07,
+      "loss": 0.0661,
+      "step": 16730
+    },
+    {
+      "epoch": 2.7107906675307842,
+      "grad_norm": 0.7594142556190491,
+      "learning_rate": 1.2087663317289554e-07,
+      "loss": 0.0543,
+      "step": 16731
+    },
+    {
+      "epoch": 2.710952689565781,
+      "grad_norm": 0.9627096056938171,
+      "learning_rate": 1.2074233567314408e-07,
+      "loss": 0.0609,
+      "step": 16732
+    },
+    {
+      "epoch": 2.7111147116007777,
+      "grad_norm": 0.8099537491798401,
+      "learning_rate": 1.2060811097291874e-07,
+      "loss": 0.0559,
+      "step": 16733
+    },
+    {
+      "epoch": 2.7112767336357746,
+      "grad_norm": 1.0347872972488403,
+      "learning_rate": 1.2047395907632818e-07,
+      "loss": 0.06,
+      "step": 16734
+    },
+    {
+      "epoch": 2.711438755670771,
+      "grad_norm": 0.824565052986145,
+      "learning_rate": 1.2033987998747582e-07,
+      "loss": 0.062,
+      "step": 16735
+    },
+    {
+      "epoch": 2.711600777705768,
+      "grad_norm": 1.082157015800476,
+      "learning_rate": 1.2020587371046445e-07,
+      "loss": 0.0715,
+      "step": 16736
+    },
+    {
+      "epoch": 2.7117627997407645,
+      "grad_norm": 0.8374156355857849,
+      "learning_rate": 1.2007194024939412e-07,
+      "loss": 0.0589,
+      "step": 16737
+    },
+    {
+      "epoch": 2.7119248217757614,
+      "grad_norm": 1.058831810951233,
+      "learning_rate": 1.1993807960836322e-07,
+      "loss": 0.0694,
+      "step": 16738
+    },
+    {
+      "epoch": 2.7120868438107584,
+      "grad_norm": 1.0584897994995117,
+      "learning_rate": 1.198042917914677e-07,
+      "loss": 0.0624,
+      "step": 16739
+    },
+    {
+      "epoch": 2.712248865845755,
+      "grad_norm": 0.9175471067428589,
+      "learning_rate": 1.1967057680280058e-07,
+      "loss": 0.0571,
+      "step": 16740
+    },
+    {
+      "epoch": 2.712410887880752,
+      "grad_norm": 0.9468705058097839,
+      "learning_rate": 1.1953693464645395e-07,
+      "loss": 0.0601,
+      "step": 16741
+    },
+    {
+      "epoch": 2.7125729099157487,
+      "grad_norm": 0.8563498258590698,
+      "learning_rate": 1.1940336532651614e-07,
+      "loss": 0.0611,
+      "step": 16742
+    },
+    {
+      "epoch": 2.712734931950745,
+      "grad_norm": 1.0424282550811768,
+      "learning_rate": 1.192698688470742e-07,
+      "loss": 0.058,
+      "step": 16743
+    },
+    {
+      "epoch": 2.712896953985742,
+      "grad_norm": 0.9378633499145508,
+      "learning_rate": 1.1913644521221345e-07,
+      "loss": 0.0613,
+      "step": 16744
+    },
+    {
+      "epoch": 2.713058976020739,
+      "grad_norm": 0.874260425567627,
+      "learning_rate": 1.1900309442601593e-07,
+      "loss": 0.0591,
+      "step": 16745
+    },
+    {
+      "epoch": 2.7132209980557356,
+      "grad_norm": 1.0240492820739746,
+      "learning_rate": 1.1886981649256169e-07,
+      "loss": 0.0673,
+      "step": 16746
+    },
+    {
+      "epoch": 2.713383020090732,
+      "grad_norm": 0.8745399117469788,
+      "learning_rate": 1.1873661141592857e-07,
+      "loss": 0.0605,
+      "step": 16747
+    },
+    {
+      "epoch": 2.713545042125729,
+      "grad_norm": 0.8494243025779724,
+      "learning_rate": 1.1860347920019304e-07,
+      "loss": 0.0592,
+      "step": 16748
+    },
+    {
+      "epoch": 2.713707064160726,
+      "grad_norm": 0.9611318111419678,
+      "learning_rate": 1.184704198494277e-07,
+      "loss": 0.0642,
+      "step": 16749
+    },
+    {
+      "epoch": 2.7138690861957224,
+      "grad_norm": 0.9604257941246033,
+      "learning_rate": 1.1833743336770482e-07,
+      "loss": 0.0631,
+      "step": 16750
+    },
+    {
+      "epoch": 2.7140311082307194,
+      "grad_norm": 0.8594368696212769,
+      "learning_rate": 1.1820451975909253e-07,
+      "loss": 0.0571,
+      "step": 16751
+    },
+    {
+      "epoch": 2.7141931302657163,
+      "grad_norm": 0.8387914299964905,
+      "learning_rate": 1.1807167902765843e-07,
+      "loss": 0.0564,
+      "step": 16752
+    },
+    {
+      "epoch": 2.714355152300713,
+      "grad_norm": 1.1130517721176147,
+      "learning_rate": 1.1793891117746648e-07,
+      "loss": 0.0618,
+      "step": 16753
+    },
+    {
+      "epoch": 2.7145171743357097,
+      "grad_norm": 0.8822175860404968,
+      "learning_rate": 1.1780621621257953e-07,
+      "loss": 0.057,
+      "step": 16754
+    },
+    {
+      "epoch": 2.7146791963707066,
+      "grad_norm": 0.8818928599357605,
+      "learning_rate": 1.176735941370577e-07,
+      "loss": 0.0605,
+      "step": 16755
+    },
+    {
+      "epoch": 2.714841218405703,
+      "grad_norm": 0.932831883430481,
+      "learning_rate": 1.1754104495495882e-07,
+      "loss": 0.064,
+      "step": 16756
+    },
+    {
+      "epoch": 2.7150032404407,
+      "grad_norm": 0.8247097134590149,
+      "learning_rate": 1.1740856867033801e-07,
+      "loss": 0.0576,
+      "step": 16757
+    },
+    {
+      "epoch": 2.7151652624756966,
+      "grad_norm": 0.8611258864402771,
+      "learning_rate": 1.1727616528724949e-07,
+      "loss": 0.0581,
+      "step": 16758
+    },
+    {
+      "epoch": 2.7153272845106935,
+      "grad_norm": 0.8132169842720032,
+      "learning_rate": 1.171438348097445e-07,
+      "loss": 0.0533,
+      "step": 16759
+    },
+    {
+      "epoch": 2.71548930654569,
+      "grad_norm": 0.8210513591766357,
+      "learning_rate": 1.1701157724187173e-07,
+      "loss": 0.0576,
+      "step": 16760
+    },
+    {
+      "epoch": 2.715651328580687,
+      "grad_norm": 0.9274697303771973,
+      "learning_rate": 1.1687939258767795e-07,
+      "loss": 0.0642,
+      "step": 16761
+    },
+    {
+      "epoch": 2.715813350615684,
+      "grad_norm": 0.9596958160400391,
+      "learning_rate": 1.1674728085120713e-07,
+      "loss": 0.063,
+      "step": 16762
+    },
+    {
+      "epoch": 2.7159753726506803,
+      "grad_norm": 0.9137746691703796,
+      "learning_rate": 1.1661524203650298e-07,
+      "loss": 0.0604,
+      "step": 16763
+    },
+    {
+      "epoch": 2.7161373946856773,
+      "grad_norm": 0.9403409361839294,
+      "learning_rate": 1.1648327614760452e-07,
+      "loss": 0.0622,
+      "step": 16764
+    },
+    {
+      "epoch": 2.716299416720674,
+      "grad_norm": 0.9043160676956177,
+      "learning_rate": 1.1635138318854961e-07,
+      "loss": 0.0668,
+      "step": 16765
+    },
+    {
+      "epoch": 2.7164614387556707,
+      "grad_norm": 0.8880981802940369,
+      "learning_rate": 1.1621956316337391e-07,
+      "loss": 0.061,
+      "step": 16766
+    },
+    {
+      "epoch": 2.7166234607906676,
+      "grad_norm": 0.938401460647583,
+      "learning_rate": 1.1608781607611113e-07,
+      "loss": 0.0631,
+      "step": 16767
+    },
+    {
+      "epoch": 2.7167854828256646,
+      "grad_norm": 0.9028334021568298,
+      "learning_rate": 1.1595614193079224e-07,
+      "loss": 0.0553,
+      "step": 16768
+    },
+    {
+      "epoch": 2.716947504860661,
+      "grad_norm": 0.8787040710449219,
+      "learning_rate": 1.1582454073144623e-07,
+      "loss": 0.0611,
+      "step": 16769
+    },
+    {
+      "epoch": 2.7171095268956575,
+      "grad_norm": 0.8604116439819336,
+      "learning_rate": 1.1569301248209958e-07,
+      "loss": 0.0499,
+      "step": 16770
+    },
+    {
+      "epoch": 2.7172715489306545,
+      "grad_norm": 0.9343183636665344,
+      "learning_rate": 1.1556155718677714e-07,
+      "loss": 0.0587,
+      "step": 16771
+    },
+    {
+      "epoch": 2.7174335709656514,
+      "grad_norm": 0.7739042639732361,
+      "learning_rate": 1.1543017484950015e-07,
+      "loss": 0.0545,
+      "step": 16772
+    },
+    {
+      "epoch": 2.717595593000648,
+      "grad_norm": 1.046132206916809,
+      "learning_rate": 1.1529886547428954e-07,
+      "loss": 0.0585,
+      "step": 16773
+    },
+    {
+      "epoch": 2.717757615035645,
+      "grad_norm": 0.8855156302452087,
+      "learning_rate": 1.1516762906516322e-07,
+      "loss": 0.0581,
+      "step": 16774
+    },
+    {
+      "epoch": 2.7179196370706418,
+      "grad_norm": 0.8061611652374268,
+      "learning_rate": 1.1503646562613602e-07,
+      "loss": 0.0588,
+      "step": 16775
+    },
+    {
+      "epoch": 2.7180816591056383,
+      "grad_norm": 0.9897039532661438,
+      "learning_rate": 1.1490537516122141e-07,
+      "loss": 0.0622,
+      "step": 16776
+    },
+    {
+      "epoch": 2.718243681140635,
+      "grad_norm": 0.9542493224143982,
+      "learning_rate": 1.1477435767443007e-07,
+      "loss": 0.0619,
+      "step": 16777
+    },
+    {
+      "epoch": 2.718405703175632,
+      "grad_norm": 0.8548462986946106,
+      "learning_rate": 1.1464341316977184e-07,
+      "loss": 0.0585,
+      "step": 16778
+    },
+    {
+      "epoch": 2.7185677252106286,
+      "grad_norm": 0.810540497303009,
+      "learning_rate": 1.145125416512527e-07,
+      "loss": 0.053,
+      "step": 16779
+    },
+    {
+      "epoch": 2.7187297472456255,
+      "grad_norm": 0.8522582054138184,
+      "learning_rate": 1.1438174312287664e-07,
+      "loss": 0.0563,
+      "step": 16780
+    },
+    {
+      "epoch": 2.718891769280622,
+      "grad_norm": 0.9812653064727783,
+      "learning_rate": 1.1425101758864632e-07,
+      "loss": 0.0679,
+      "step": 16781
+    },
+    {
+      "epoch": 2.719053791315619,
+      "grad_norm": 1.0658005475997925,
+      "learning_rate": 1.1412036505256158e-07,
+      "loss": 0.057,
+      "step": 16782
+    },
+    {
+      "epoch": 2.7192158133506155,
+      "grad_norm": 0.930414080619812,
+      "learning_rate": 1.1398978551861978e-07,
+      "loss": 0.0626,
+      "step": 16783
+    },
+    {
+      "epoch": 2.7193778353856124,
+      "grad_norm": 0.9118586182594299,
+      "learning_rate": 1.1385927899081661e-07,
+      "loss": 0.0641,
+      "step": 16784
+    },
+    {
+      "epoch": 2.7195398574206093,
+      "grad_norm": 0.9204965829849243,
+      "learning_rate": 1.13728845473145e-07,
+      "loss": 0.0619,
+      "step": 16785
+    },
+    {
+      "epoch": 2.719701879455606,
+      "grad_norm": 0.9930737614631653,
+      "learning_rate": 1.1359848496959618e-07,
+      "loss": 0.0617,
+      "step": 16786
+    },
+    {
+      "epoch": 2.7198639014906028,
+      "grad_norm": 0.8213454484939575,
+      "learning_rate": 1.1346819748415893e-07,
+      "loss": 0.0568,
+      "step": 16787
+    },
+    {
+      "epoch": 2.7200259235255997,
+      "grad_norm": 0.8408246636390686,
+      "learning_rate": 1.1333798302081922e-07,
+      "loss": 0.0554,
+      "step": 16788
+    },
+    {
+      "epoch": 2.720187945560596,
+      "grad_norm": 0.8678555488586426,
+      "learning_rate": 1.1320784158356218e-07,
+      "loss": 0.0564,
+      "step": 16789
+    },
+    {
+      "epoch": 2.720349967595593,
+      "grad_norm": 0.8617439866065979,
+      "learning_rate": 1.1307777317636882e-07,
+      "loss": 0.0587,
+      "step": 16790
+    },
+    {
+      "epoch": 2.7205119896305896,
+      "grad_norm": 0.8767697811126709,
+      "learning_rate": 1.1294777780321898e-07,
+      "loss": 0.0613,
+      "step": 16791
+    },
+    {
+      "epoch": 2.7206740116655865,
+      "grad_norm": 1.116636872291565,
+      "learning_rate": 1.1281785546809115e-07,
+      "loss": 0.0709,
+      "step": 16792
+    },
+    {
+      "epoch": 2.720836033700583,
+      "grad_norm": 0.9495435357093811,
+      "learning_rate": 1.1268800617495995e-07,
+      "loss": 0.0596,
+      "step": 16793
+    },
+    {
+      "epoch": 2.72099805573558,
+      "grad_norm": 0.8045752644538879,
+      "learning_rate": 1.1255822992779858e-07,
+      "loss": 0.0519,
+      "step": 16794
+    },
+    {
+      "epoch": 2.721160077770577,
+      "grad_norm": 0.7897116541862488,
+      "learning_rate": 1.1242852673057774e-07,
+      "loss": 0.0559,
+      "step": 16795
+    },
+    {
+      "epoch": 2.7213220998055734,
+      "grad_norm": 0.9391621947288513,
+      "learning_rate": 1.1229889658726623e-07,
+      "loss": 0.0606,
+      "step": 16796
+    },
+    {
+      "epoch": 2.7214841218405703,
+      "grad_norm": 0.8378191590309143,
+      "learning_rate": 1.1216933950183028e-07,
+      "loss": 0.0553,
+      "step": 16797
+    },
+    {
+      "epoch": 2.7216461438755672,
+      "grad_norm": 0.8450056314468384,
+      "learning_rate": 1.1203985547823427e-07,
+      "loss": 0.0608,
+      "step": 16798
+    },
+    {
+      "epoch": 2.7218081659105637,
+      "grad_norm": 0.9059349894523621,
+      "learning_rate": 1.1191044452043998e-07,
+      "loss": 0.0581,
+      "step": 16799
+    },
+    {
+      "epoch": 2.7219701879455607,
+      "grad_norm": 0.9473249316215515,
+      "learning_rate": 1.1178110663240676e-07,
+      "loss": 0.0646,
+      "step": 16800
+    },
+    {
+      "epoch": 2.7221322099805576,
+      "grad_norm": 1.025634527206421,
+      "learning_rate": 1.1165184181809258e-07,
+      "loss": 0.0586,
+      "step": 16801
+    },
+    {
+      "epoch": 2.722294232015554,
+      "grad_norm": 0.9218292832374573,
+      "learning_rate": 1.1152265008145202e-07,
+      "loss": 0.0651,
+      "step": 16802
+    },
+    {
+      "epoch": 2.722456254050551,
+      "grad_norm": 0.8640052080154419,
+      "learning_rate": 1.1139353142643861e-07,
+      "loss": 0.0558,
+      "step": 16803
+    },
+    {
+      "epoch": 2.7226182760855475,
+      "grad_norm": 1.0886274576187134,
+      "learning_rate": 1.1126448585700306e-07,
+      "loss": 0.06,
+      "step": 16804
+    },
+    {
+      "epoch": 2.7227802981205445,
+      "grad_norm": 0.8292529582977295,
+      "learning_rate": 1.1113551337709305e-07,
+      "loss": 0.0591,
+      "step": 16805
+    },
+    {
+      "epoch": 2.722942320155541,
+      "grad_norm": 0.85284024477005,
+      "learning_rate": 1.110066139906557e-07,
+      "loss": 0.0607,
+      "step": 16806
+    },
+    {
+      "epoch": 2.723104342190538,
+      "grad_norm": 0.8384661674499512,
+      "learning_rate": 1.1087778770163482e-07,
+      "loss": 0.0543,
+      "step": 16807
+    },
+    {
+      "epoch": 2.723266364225535,
+      "grad_norm": 0.9519869089126587,
+      "learning_rate": 1.1074903451397195e-07,
+      "loss": 0.0633,
+      "step": 16808
+    },
+    {
+      "epoch": 2.7234283862605313,
+      "grad_norm": 0.8796502947807312,
+      "learning_rate": 1.1062035443160673e-07,
+      "loss": 0.0599,
+      "step": 16809
+    },
+    {
+      "epoch": 2.7235904082955282,
+      "grad_norm": 0.9636805057525635,
+      "learning_rate": 1.1049174745847657e-07,
+      "loss": 0.065,
+      "step": 16810
+    },
+    {
+      "epoch": 2.723752430330525,
+      "grad_norm": 0.8485105037689209,
+      "learning_rate": 1.1036321359851638e-07,
+      "loss": 0.0555,
+      "step": 16811
+    },
+    {
+      "epoch": 2.7239144523655217,
+      "grad_norm": 1.940087080001831,
+      "learning_rate": 1.1023475285565882e-07,
+      "loss": 0.0602,
+      "step": 16812
+    },
+    {
+      "epoch": 2.7240764744005186,
+      "grad_norm": 0.9121723175048828,
+      "learning_rate": 1.1010636523383494e-07,
+      "loss": 0.0591,
+      "step": 16813
+    },
+    {
+      "epoch": 2.724238496435515,
+      "grad_norm": 0.9082077145576477,
+      "learning_rate": 1.099780507369727e-07,
+      "loss": 0.062,
+      "step": 16814
+    },
+    {
+      "epoch": 2.724400518470512,
+      "grad_norm": 0.9480251669883728,
+      "learning_rate": 1.0984980936899842e-07,
+      "loss": 0.0608,
+      "step": 16815
+    },
+    {
+      "epoch": 2.7245625405055085,
+      "grad_norm": 1.0009106397628784,
+      "learning_rate": 1.0972164113383616e-07,
+      "loss": 0.0651,
+      "step": 16816
+    },
+    {
+      "epoch": 2.7247245625405054,
+      "grad_norm": 1.0686898231506348,
+      "learning_rate": 1.0959354603540695e-07,
+      "loss": 0.069,
+      "step": 16817
+    },
+    {
+      "epoch": 2.7248865845755024,
+      "grad_norm": 0.8673676252365112,
+      "learning_rate": 1.09465524077631e-07,
+      "loss": 0.0545,
+      "step": 16818
+    },
+    {
+      "epoch": 2.725048606610499,
+      "grad_norm": 0.9199877977371216,
+      "learning_rate": 1.093375752644249e-07,
+      "loss": 0.053,
+      "step": 16819
+    },
+    {
+      "epoch": 2.725210628645496,
+      "grad_norm": 0.8862485885620117,
+      "learning_rate": 1.0920969959970301e-07,
+      "loss": 0.0579,
+      "step": 16820
+    },
+    {
+      "epoch": 2.7253726506804927,
+      "grad_norm": 0.8255491256713867,
+      "learning_rate": 1.0908189708737942e-07,
+      "loss": 0.0614,
+      "step": 16821
+    },
+    {
+      "epoch": 2.725534672715489,
+      "grad_norm": 0.9784971475601196,
+      "learning_rate": 1.0895416773136408e-07,
+      "loss": 0.0637,
+      "step": 16822
+    },
+    {
+      "epoch": 2.725696694750486,
+      "grad_norm": 1.1451879739761353,
+      "learning_rate": 1.088265115355644e-07,
+      "loss": 0.0695,
+      "step": 16823
+    },
+    {
+      "epoch": 2.725858716785483,
+      "grad_norm": 0.827608585357666,
+      "learning_rate": 1.0869892850388697e-07,
+      "loss": 0.0603,
+      "step": 16824
+    },
+    {
+      "epoch": 2.7260207388204796,
+      "grad_norm": 0.9534388184547424,
+      "learning_rate": 1.0857141864023591e-07,
+      "loss": 0.0635,
+      "step": 16825
+    },
+    {
+      "epoch": 2.7261827608554765,
+      "grad_norm": 0.9384109973907471,
+      "learning_rate": 1.0844398194851197e-07,
+      "loss": 0.0566,
+      "step": 16826
+    },
+    {
+      "epoch": 2.726344782890473,
+      "grad_norm": 1.1413118839263916,
+      "learning_rate": 1.0831661843261482e-07,
+      "loss": 0.0673,
+      "step": 16827
+    },
+    {
+      "epoch": 2.72650680492547,
+      "grad_norm": 0.878052830696106,
+      "learning_rate": 1.0818932809644161e-07,
+      "loss": 0.0537,
+      "step": 16828
+    },
+    {
+      "epoch": 2.7266688269604664,
+      "grad_norm": 0.9519762992858887,
+      "learning_rate": 1.0806211094388647e-07,
+      "loss": 0.0624,
+      "step": 16829
+    },
+    {
+      "epoch": 2.7268308489954634,
+      "grad_norm": 0.9057498574256897,
+      "learning_rate": 1.0793496697884265e-07,
+      "loss": 0.0527,
+      "step": 16830
+    },
+    {
+      "epoch": 2.7269928710304603,
+      "grad_norm": 0.9844750761985779,
+      "learning_rate": 1.078078962052001e-07,
+      "loss": 0.0589,
+      "step": 16831
+    },
+    {
+      "epoch": 2.7271548930654568,
+      "grad_norm": 1.1606096029281616,
+      "learning_rate": 1.0768089862684684e-07,
+      "loss": 0.0655,
+      "step": 16832
+    },
+    {
+      "epoch": 2.7273169151004537,
+      "grad_norm": 0.8622628450393677,
+      "learning_rate": 1.0755397424766917e-07,
+      "loss": 0.0558,
+      "step": 16833
+    },
+    {
+      "epoch": 2.7274789371354506,
+      "grad_norm": 0.8796065449714661,
+      "learning_rate": 1.0742712307154957e-07,
+      "loss": 0.0685,
+      "step": 16834
+    },
+    {
+      "epoch": 2.727640959170447,
+      "grad_norm": 0.8955232501029968,
+      "learning_rate": 1.0730034510237048e-07,
+      "loss": 0.0654,
+      "step": 16835
+    },
+    {
+      "epoch": 2.727802981205444,
+      "grad_norm": 0.9430897235870361,
+      "learning_rate": 1.0717364034401073e-07,
+      "loss": 0.0662,
+      "step": 16836
+    },
+    {
+      "epoch": 2.7279650032404406,
+      "grad_norm": 0.849699079990387,
+      "learning_rate": 1.0704700880034696e-07,
+      "loss": 0.0577,
+      "step": 16837
+    },
+    {
+      "epoch": 2.7281270252754375,
+      "grad_norm": 0.9427576065063477,
+      "learning_rate": 1.0692045047525384e-07,
+      "loss": 0.0603,
+      "step": 16838
+    },
+    {
+      "epoch": 2.728289047310434,
+      "grad_norm": 0.8516416549682617,
+      "learning_rate": 1.0679396537260356e-07,
+      "loss": 0.0539,
+      "step": 16839
+    },
+    {
+      "epoch": 2.728451069345431,
+      "grad_norm": 0.9089876413345337,
+      "learning_rate": 1.066675534962669e-07,
+      "loss": 0.0621,
+      "step": 16840
+    },
+    {
+      "epoch": 2.728613091380428,
+      "grad_norm": 1.2640206813812256,
+      "learning_rate": 1.0654121485011131e-07,
+      "loss": 0.0632,
+      "step": 16841
+    },
+    {
+      "epoch": 2.7287751134154243,
+      "grad_norm": 0.8047393560409546,
+      "learning_rate": 1.0641494943800234e-07,
+      "loss": 0.0581,
+      "step": 16842
+    },
+    {
+      "epoch": 2.7289371354504213,
+      "grad_norm": 0.8546563982963562,
+      "learning_rate": 1.0628875726380355e-07,
+      "loss": 0.057,
+      "step": 16843
+    },
+    {
+      "epoch": 2.729099157485418,
+      "grad_norm": 1.0105202198028564,
+      "learning_rate": 1.0616263833137602e-07,
+      "loss": 0.0678,
+      "step": 16844
+    },
+    {
+      "epoch": 2.7292611795204147,
+      "grad_norm": 0.9958057999610901,
+      "learning_rate": 1.0603659264457888e-07,
+      "loss": 0.0581,
+      "step": 16845
+    },
+    {
+      "epoch": 2.7294232015554116,
+      "grad_norm": 0.9695627093315125,
+      "learning_rate": 1.0591062020726878e-07,
+      "loss": 0.059,
+      "step": 16846
+    },
+    {
+      "epoch": 2.7295852235904086,
+      "grad_norm": 0.8972888588905334,
+      "learning_rate": 1.0578472102330011e-07,
+      "loss": 0.0588,
+      "step": 16847
+    },
+    {
+      "epoch": 2.729747245625405,
+      "grad_norm": 1.0485844612121582,
+      "learning_rate": 1.0565889509652483e-07,
+      "loss": 0.0638,
+      "step": 16848
+    },
+    {
+      "epoch": 2.7299092676604015,
+      "grad_norm": 0.9550689458847046,
+      "learning_rate": 1.0553314243079343e-07,
+      "loss": 0.0621,
+      "step": 16849
+    },
+    {
+      "epoch": 2.7300712896953985,
+      "grad_norm": 0.8764699101448059,
+      "learning_rate": 1.0540746302995341e-07,
+      "loss": 0.0568,
+      "step": 16850
+    },
+    {
+      "epoch": 2.7302333117303954,
+      "grad_norm": 0.8220252394676208,
+      "learning_rate": 1.0528185689785031e-07,
+      "loss": 0.056,
+      "step": 16851
+    },
+    {
+      "epoch": 2.730395333765392,
+      "grad_norm": 1.0028624534606934,
+      "learning_rate": 1.0515632403832715e-07,
+      "loss": 0.0509,
+      "step": 16852
+    },
+    {
+      "epoch": 2.730557355800389,
+      "grad_norm": 0.84007328748703,
+      "learning_rate": 1.0503086445522476e-07,
+      "loss": 0.0505,
+      "step": 16853
+    },
+    {
+      "epoch": 2.7307193778353858,
+      "grad_norm": 0.9844392538070679,
+      "learning_rate": 1.0490547815238228e-07,
+      "loss": 0.0602,
+      "step": 16854
+    },
+    {
+      "epoch": 2.7308813998703823,
+      "grad_norm": 1.050789475440979,
+      "learning_rate": 1.0478016513363665e-07,
+      "loss": 0.0632,
+      "step": 16855
+    },
+    {
+      "epoch": 2.731043421905379,
+      "grad_norm": 1.0053508281707764,
+      "learning_rate": 1.0465492540282146e-07,
+      "loss": 0.0655,
+      "step": 16856
+    },
+    {
+      "epoch": 2.731205443940376,
+      "grad_norm": 0.9643328785896301,
+      "learning_rate": 1.0452975896376865e-07,
+      "loss": 0.0642,
+      "step": 16857
+    },
+    {
+      "epoch": 2.7313674659753726,
+      "grad_norm": 0.9741927981376648,
+      "learning_rate": 1.044046658203085e-07,
+      "loss": 0.0602,
+      "step": 16858
+    },
+    {
+      "epoch": 2.7315294880103695,
+      "grad_norm": 0.9322402477264404,
+      "learning_rate": 1.0427964597626822e-07,
+      "loss": 0.0567,
+      "step": 16859
+    },
+    {
+      "epoch": 2.731691510045366,
+      "grad_norm": 0.8425524234771729,
+      "learning_rate": 1.0415469943547335e-07,
+      "loss": 0.0608,
+      "step": 16860
+    },
+    {
+      "epoch": 2.731853532080363,
+      "grad_norm": 1.0030529499053955,
+      "learning_rate": 1.0402982620174696e-07,
+      "loss": 0.0638,
+      "step": 16861
+    },
+    {
+      "epoch": 2.7320155541153595,
+      "grad_norm": 0.8651495575904846,
+      "learning_rate": 1.0390502627890986e-07,
+      "loss": 0.0586,
+      "step": 16862
+    },
+    {
+      "epoch": 2.7321775761503564,
+      "grad_norm": 0.9395628571510315,
+      "learning_rate": 1.0378029967077985e-07,
+      "loss": 0.0558,
+      "step": 16863
+    },
+    {
+      "epoch": 2.7323395981853533,
+      "grad_norm": 0.8924857378005981,
+      "learning_rate": 1.0365564638117442e-07,
+      "loss": 0.0659,
+      "step": 16864
+    },
+    {
+      "epoch": 2.73250162022035,
+      "grad_norm": 0.860127329826355,
+      "learning_rate": 1.0353106641390693e-07,
+      "loss": 0.0651,
+      "step": 16865
+    },
+    {
+      "epoch": 2.7326636422553467,
+      "grad_norm": 0.846647322177887,
+      "learning_rate": 1.0340655977279012e-07,
+      "loss": 0.0592,
+      "step": 16866
+    },
+    {
+      "epoch": 2.7328256642903437,
+      "grad_norm": 0.871707022190094,
+      "learning_rate": 1.032821264616321e-07,
+      "loss": 0.0618,
+      "step": 16867
+    },
+    {
+      "epoch": 2.73298768632534,
+      "grad_norm": 0.89871746301651,
+      "learning_rate": 1.0315776648424119e-07,
+      "loss": 0.0655,
+      "step": 16868
+    },
+    {
+      "epoch": 2.733149708360337,
+      "grad_norm": 0.8250013589859009,
+      "learning_rate": 1.0303347984442297e-07,
+      "loss": 0.0585,
+      "step": 16869
+    },
+    {
+      "epoch": 2.733311730395334,
+      "grad_norm": 0.8901594281196594,
+      "learning_rate": 1.0290926654597938e-07,
+      "loss": 0.0568,
+      "step": 16870
+    },
+    {
+      "epoch": 2.7334737524303305,
+      "grad_norm": 0.9799373745918274,
+      "learning_rate": 1.0278512659271128e-07,
+      "loss": 0.0617,
+      "step": 16871
+    },
+    {
+      "epoch": 2.733635774465327,
+      "grad_norm": 0.9966387748718262,
+      "learning_rate": 1.0266105998841702e-07,
+      "loss": 0.058,
+      "step": 16872
+    },
+    {
+      "epoch": 2.733797796500324,
+      "grad_norm": 0.8289188742637634,
+      "learning_rate": 1.0253706673689328e-07,
+      "loss": 0.0542,
+      "step": 16873
+    },
+    {
+      "epoch": 2.733959818535321,
+      "grad_norm": 0.9346227645874023,
+      "learning_rate": 1.0241314684193343e-07,
+      "loss": 0.0611,
+      "step": 16874
+    },
+    {
+      "epoch": 2.7341218405703174,
+      "grad_norm": 0.9059098958969116,
+      "learning_rate": 1.0228930030732914e-07,
+      "loss": 0.0619,
+      "step": 16875
+    },
+    {
+      "epoch": 2.7342838626053143,
+      "grad_norm": 0.9077804684638977,
+      "learning_rate": 1.0216552713686989e-07,
+      "loss": 0.0595,
+      "step": 16876
+    },
+    {
+      "epoch": 2.7344458846403112,
+      "grad_norm": 0.9058060646057129,
+      "learning_rate": 1.0204182733434293e-07,
+      "loss": 0.0612,
+      "step": 16877
+    },
+    {
+      "epoch": 2.7346079066753077,
+      "grad_norm": 0.8321201205253601,
+      "learning_rate": 1.01918200903533e-07,
+      "loss": 0.0592,
+      "step": 16878
+    },
+    {
+      "epoch": 2.7347699287103047,
+      "grad_norm": 0.9400634765625,
+      "learning_rate": 1.017946478482229e-07,
+      "loss": 0.068,
+      "step": 16879
+    },
+    {
+      "epoch": 2.7349319507453016,
+      "grad_norm": 0.822931170463562,
+      "learning_rate": 1.0167116817219325e-07,
+      "loss": 0.0585,
+      "step": 16880
+    },
+    {
+      "epoch": 2.735093972780298,
+      "grad_norm": 1.0861873626708984,
+      "learning_rate": 1.0154776187922182e-07,
+      "loss": 0.068,
+      "step": 16881
+    },
+    {
+      "epoch": 2.735255994815295,
+      "grad_norm": 1.1060582399368286,
+      "learning_rate": 1.0142442897308453e-07,
+      "loss": 0.0665,
+      "step": 16882
+    },
+    {
+      "epoch": 2.7354180168502915,
+      "grad_norm": 0.9769384860992432,
+      "learning_rate": 1.0130116945755553e-07,
+      "loss": 0.0658,
+      "step": 16883
+    },
+    {
+      "epoch": 2.7355800388852884,
+      "grad_norm": 0.9157807230949402,
+      "learning_rate": 1.0117798333640627e-07,
+      "loss": 0.0583,
+      "step": 16884
+    },
+    {
+      "epoch": 2.735742060920285,
+      "grad_norm": 0.98633873462677,
+      "learning_rate": 1.0105487061340541e-07,
+      "loss": 0.0652,
+      "step": 16885
+    },
+    {
+      "epoch": 2.735904082955282,
+      "grad_norm": 0.9951062202453613,
+      "learning_rate": 1.0093183129231993e-07,
+      "loss": 0.0606,
+      "step": 16886
+    },
+    {
+      "epoch": 2.736066104990279,
+      "grad_norm": 0.8598393201828003,
+      "learning_rate": 1.0080886537691514e-07,
+      "loss": 0.0607,
+      "step": 16887
+    },
+    {
+      "epoch": 2.7362281270252753,
+      "grad_norm": 0.911261260509491,
+      "learning_rate": 1.0068597287095305e-07,
+      "loss": 0.0566,
+      "step": 16888
+    },
+    {
+      "epoch": 2.7363901490602722,
+      "grad_norm": 0.9654666185379028,
+      "learning_rate": 1.005631537781937e-07,
+      "loss": 0.0631,
+      "step": 16889
+    },
+    {
+      "epoch": 2.736552171095269,
+      "grad_norm": 1.0117939710617065,
+      "learning_rate": 1.0044040810239547e-07,
+      "loss": 0.063,
+      "step": 16890
+    },
+    {
+      "epoch": 2.7367141931302656,
+      "grad_norm": 0.8387407064437866,
+      "learning_rate": 1.0031773584731397e-07,
+      "loss": 0.0552,
+      "step": 16891
+    },
+    {
+      "epoch": 2.7368762151652626,
+      "grad_norm": 0.8349279165267944,
+      "learning_rate": 1.0019513701670285e-07,
+      "loss": 0.06,
+      "step": 16892
+    },
+    {
+      "epoch": 2.737038237200259,
+      "grad_norm": 0.9240767955780029,
+      "learning_rate": 1.0007261161431275e-07,
+      "loss": 0.0591,
+      "step": 16893
+    },
+    {
+      "epoch": 2.737200259235256,
+      "grad_norm": 0.9168483018875122,
+      "learning_rate": 9.995015964389315e-08,
+      "loss": 0.0643,
+      "step": 16894
+    },
+    {
+      "epoch": 2.7373622812702525,
+      "grad_norm": 1.0924078226089478,
+      "learning_rate": 9.982778110919106e-08,
+      "loss": 0.0587,
+      "step": 16895
+    },
+    {
+      "epoch": 2.7375243033052494,
+      "grad_norm": 0.9266462326049805,
+      "learning_rate": 9.970547601394986e-08,
+      "loss": 0.0645,
+      "step": 16896
+    },
+    {
+      "epoch": 2.7376863253402464,
+      "grad_norm": 1.2055165767669678,
+      "learning_rate": 9.958324436191297e-08,
+      "loss": 0.0617,
+      "step": 16897
+    },
+    {
+      "epoch": 2.737848347375243,
+      "grad_norm": 0.8799644708633423,
+      "learning_rate": 9.94610861568196e-08,
+      "loss": 0.0596,
+      "step": 16898
+    },
+    {
+      "epoch": 2.73801036941024,
+      "grad_norm": 0.9457053542137146,
+      "learning_rate": 9.933900140240843e-08,
+      "loss": 0.057,
+      "step": 16899
+    },
+    {
+      "epoch": 2.7381723914452367,
+      "grad_norm": 0.8883636593818665,
+      "learning_rate": 9.92169901024137e-08,
+      "loss": 0.0646,
+      "step": 16900
+    },
+    {
+      "epoch": 2.738334413480233,
+      "grad_norm": 0.9511093497276306,
+      "learning_rate": 9.90950522605691e-08,
+      "loss": 0.0616,
+      "step": 16901
+    },
+    {
+      "epoch": 2.73849643551523,
+      "grad_norm": 0.9124867916107178,
+      "learning_rate": 9.897318788060662e-08,
+      "loss": 0.0563,
+      "step": 16902
+    },
+    {
+      "epoch": 2.738658457550227,
+      "grad_norm": 1.006246566772461,
+      "learning_rate": 9.885139696625356e-08,
+      "loss": 0.0661,
+      "step": 16903
+    },
+    {
+      "epoch": 2.7388204795852236,
+      "grad_norm": 0.8624682426452637,
+      "learning_rate": 9.872967952123752e-08,
+      "loss": 0.0538,
+      "step": 16904
+    },
+    {
+      "epoch": 2.7389825016202205,
+      "grad_norm": 1.1840333938598633,
+      "learning_rate": 9.860803554928189e-08,
+      "loss": 0.0687,
+      "step": 16905
+    },
+    {
+      "epoch": 2.739144523655217,
+      "grad_norm": 0.8678603172302246,
+      "learning_rate": 9.848646505410953e-08,
+      "loss": 0.0524,
+      "step": 16906
+    },
+    {
+      "epoch": 2.739306545690214,
+      "grad_norm": 0.8984764218330383,
+      "learning_rate": 9.83649680394394e-08,
+      "loss": 0.0554,
+      "step": 16907
+    },
+    {
+      "epoch": 2.7394685677252104,
+      "grad_norm": 0.9465093016624451,
+      "learning_rate": 9.824354450898966e-08,
+      "loss": 0.0618,
+      "step": 16908
+    },
+    {
+      "epoch": 2.7396305897602073,
+      "grad_norm": 0.8658840656280518,
+      "learning_rate": 9.812219446647509e-08,
+      "loss": 0.0571,
+      "step": 16909
+    },
+    {
+      "epoch": 2.7397926117952043,
+      "grad_norm": 0.9028376340866089,
+      "learning_rate": 9.800091791560939e-08,
+      "loss": 0.0542,
+      "step": 16910
+    },
+    {
+      "epoch": 2.7399546338302008,
+      "grad_norm": 1.0068341493606567,
+      "learning_rate": 9.78797148601024e-08,
+      "loss": 0.0694,
+      "step": 16911
+    },
+    {
+      "epoch": 2.7401166558651977,
+      "grad_norm": 0.8635134696960449,
+      "learning_rate": 9.775858530366334e-08,
+      "loss": 0.06,
+      "step": 16912
+    },
+    {
+      "epoch": 2.7402786779001946,
+      "grad_norm": 0.8920632600784302,
+      "learning_rate": 9.763752924999842e-08,
+      "loss": 0.0559,
+      "step": 16913
+    },
+    {
+      "epoch": 2.740440699935191,
+      "grad_norm": 0.8138787746429443,
+      "learning_rate": 9.751654670281135e-08,
+      "loss": 0.0555,
+      "step": 16914
+    },
+    {
+      "epoch": 2.740602721970188,
+      "grad_norm": 0.9727160334587097,
+      "learning_rate": 9.739563766580362e-08,
+      "loss": 0.0632,
+      "step": 16915
+    },
+    {
+      "epoch": 2.7407647440051845,
+      "grad_norm": 0.830783486366272,
+      "learning_rate": 9.727480214267559e-08,
+      "loss": 0.0599,
+      "step": 16916
+    },
+    {
+      "epoch": 2.7409267660401815,
+      "grad_norm": 0.8595759272575378,
+      "learning_rate": 9.715404013712432e-08,
+      "loss": 0.0564,
+      "step": 16917
+    },
+    {
+      "epoch": 2.741088788075178,
+      "grad_norm": 1.019550085067749,
+      "learning_rate": 9.70333516528446e-08,
+      "loss": 0.0684,
+      "step": 16918
+    },
+    {
+      "epoch": 2.741250810110175,
+      "grad_norm": 1.0129154920578003,
+      "learning_rate": 9.691273669352908e-08,
+      "loss": 0.066,
+      "step": 16919
+    },
+    {
+      "epoch": 2.741412832145172,
+      "grad_norm": 1.0200157165527344,
+      "learning_rate": 9.679219526286837e-08,
+      "loss": 0.066,
+      "step": 16920
+    },
+    {
+      "epoch": 2.7415748541801683,
+      "grad_norm": 1.0694841146469116,
+      "learning_rate": 9.667172736455093e-08,
+      "loss": 0.0631,
+      "step": 16921
+    },
+    {
+      "epoch": 2.7417368762151653,
+      "grad_norm": 0.9911954402923584,
+      "learning_rate": 9.655133300226271e-08,
+      "loss": 0.0644,
+      "step": 16922
+    },
+    {
+      "epoch": 2.741898898250162,
+      "grad_norm": 1.1050474643707275,
+      "learning_rate": 9.643101217968743e-08,
+      "loss": 0.0572,
+      "step": 16923
+    },
+    {
+      "epoch": 2.7420609202851587,
+      "grad_norm": 0.8301832675933838,
+      "learning_rate": 9.631076490050684e-08,
+      "loss": 0.0577,
+      "step": 16924
+    },
+    {
+      "epoch": 2.7422229423201556,
+      "grad_norm": 1.2750616073608398,
+      "learning_rate": 9.619059116839968e-08,
+      "loss": 0.07,
+      "step": 16925
+    },
+    {
+      "epoch": 2.7423849643551526,
+      "grad_norm": 0.9567539691925049,
+      "learning_rate": 9.60704909870433e-08,
+      "loss": 0.0601,
+      "step": 16926
+    },
+    {
+      "epoch": 2.742546986390149,
+      "grad_norm": 1.1425048112869263,
+      "learning_rate": 9.59504643601128e-08,
+      "loss": 0.0629,
+      "step": 16927
+    },
+    {
+      "epoch": 2.742709008425146,
+      "grad_norm": 0.8750780820846558,
+      "learning_rate": 9.583051129128051e-08,
+      "loss": 0.0594,
+      "step": 16928
+    },
+    {
+      "epoch": 2.7428710304601425,
+      "grad_norm": 0.8770819306373596,
+      "learning_rate": 9.57106317842163e-08,
+      "loss": 0.0646,
+      "step": 16929
+    },
+    {
+      "epoch": 2.7430330524951394,
+      "grad_norm": 1.0766657590866089,
+      "learning_rate": 9.559082584258833e-08,
+      "loss": 0.0548,
+      "step": 16930
+    },
+    {
+      "epoch": 2.743195074530136,
+      "grad_norm": 0.9907754063606262,
+      "learning_rate": 9.547109347006312e-08,
+      "loss": 0.0658,
+      "step": 16931
+    },
+    {
+      "epoch": 2.743357096565133,
+      "grad_norm": 0.9115135669708252,
+      "learning_rate": 9.535143467030327e-08,
+      "loss": 0.0584,
+      "step": 16932
+    },
+    {
+      "epoch": 2.7435191186001298,
+      "grad_norm": 0.9572370052337646,
+      "learning_rate": 9.523184944697034e-08,
+      "loss": 0.0609,
+      "step": 16933
+    },
+    {
+      "epoch": 2.7436811406351262,
+      "grad_norm": 0.8873291015625,
+      "learning_rate": 9.511233780372303e-08,
+      "loss": 0.0609,
+      "step": 16934
+    },
+    {
+      "epoch": 2.743843162670123,
+      "grad_norm": 0.8288209438323975,
+      "learning_rate": 9.499289974421927e-08,
+      "loss": 0.0567,
+      "step": 16935
+    },
+    {
+      "epoch": 2.74400518470512,
+      "grad_norm": 0.8421540856361389,
+      "learning_rate": 9.487353527211223e-08,
+      "loss": 0.0597,
+      "step": 16936
+    },
+    {
+      "epoch": 2.7441672067401166,
+      "grad_norm": 0.884370744228363,
+      "learning_rate": 9.475424439105485e-08,
+      "loss": 0.0656,
+      "step": 16937
+    },
+    {
+      "epoch": 2.7443292287751135,
+      "grad_norm": 0.9282497763633728,
+      "learning_rate": 9.463502710469697e-08,
+      "loss": 0.0596,
+      "step": 16938
+    },
+    {
+      "epoch": 2.74449125081011,
+      "grad_norm": 0.9526280760765076,
+      "learning_rate": 9.45158834166865e-08,
+      "loss": 0.0618,
+      "step": 16939
+    },
+    {
+      "epoch": 2.744653272845107,
+      "grad_norm": 1.0486241579055786,
+      "learning_rate": 9.439681333066858e-08,
+      "loss": 0.0664,
+      "step": 16940
+    },
+    {
+      "epoch": 2.7448152948801035,
+      "grad_norm": 0.9274911284446716,
+      "learning_rate": 9.427781685028697e-08,
+      "loss": 0.0546,
+      "step": 16941
+    },
+    {
+      "epoch": 2.7449773169151004,
+      "grad_norm": 0.9596142768859863,
+      "learning_rate": 9.415889397918238e-08,
+      "loss": 0.0612,
+      "step": 16942
+    },
+    {
+      "epoch": 2.7451393389500973,
+      "grad_norm": 0.8194335699081421,
+      "learning_rate": 9.404004472099382e-08,
+      "loss": 0.0555,
+      "step": 16943
+    },
+    {
+      "epoch": 2.745301360985094,
+      "grad_norm": 0.9087086319923401,
+      "learning_rate": 9.3921269079357e-08,
+      "loss": 0.0567,
+      "step": 16944
+    },
+    {
+      "epoch": 2.7454633830200907,
+      "grad_norm": 0.8666279911994934,
+      "learning_rate": 9.380256705790708e-08,
+      "loss": 0.0553,
+      "step": 16945
+    },
+    {
+      "epoch": 2.7456254050550877,
+      "grad_norm": 0.9419772624969482,
+      "learning_rate": 9.368393866027614e-08,
+      "loss": 0.0666,
+      "step": 16946
+    },
+    {
+      "epoch": 2.745787427090084,
+      "grad_norm": 0.8581135869026184,
+      "learning_rate": 9.356538389009296e-08,
+      "loss": 0.0524,
+      "step": 16947
+    },
+    {
+      "epoch": 2.745949449125081,
+      "grad_norm": 0.846437394618988,
+      "learning_rate": 9.344690275098573e-08,
+      "loss": 0.0499,
+      "step": 16948
+    },
+    {
+      "epoch": 2.746111471160078,
+      "grad_norm": 0.8755781054496765,
+      "learning_rate": 9.332849524657961e-08,
+      "loss": 0.0586,
+      "step": 16949
+    },
+    {
+      "epoch": 2.7462734931950745,
+      "grad_norm": 0.9014527201652527,
+      "learning_rate": 9.321016138049727e-08,
+      "loss": 0.0562,
+      "step": 16950
+    },
+    {
+      "epoch": 2.7464355152300715,
+      "grad_norm": 0.9872837662696838,
+      "learning_rate": 9.309190115635996e-08,
+      "loss": 0.057,
+      "step": 16951
+    },
+    {
+      "epoch": 2.746597537265068,
+      "grad_norm": 0.9479248523712158,
+      "learning_rate": 9.297371457778565e-08,
+      "loss": 0.0581,
+      "step": 16952
+    },
+    {
+      "epoch": 2.746759559300065,
+      "grad_norm": 0.9341238737106323,
+      "learning_rate": 9.285560164839086e-08,
+      "loss": 0.0633,
+      "step": 16953
+    },
+    {
+      "epoch": 2.7469215813350614,
+      "grad_norm": 0.9351992011070251,
+      "learning_rate": 9.273756237178938e-08,
+      "loss": 0.0639,
+      "step": 16954
+    },
+    {
+      "epoch": 2.7470836033700583,
+      "grad_norm": 0.8418577313423157,
+      "learning_rate": 9.261959675159304e-08,
+      "loss": 0.0612,
+      "step": 16955
+    },
+    {
+      "epoch": 2.7472456254050552,
+      "grad_norm": 1.0418609380722046,
+      "learning_rate": 9.250170479141146e-08,
+      "loss": 0.0661,
+      "step": 16956
+    },
+    {
+      "epoch": 2.7474076474400517,
+      "grad_norm": 1.1158641576766968,
+      "learning_rate": 9.238388649485175e-08,
+      "loss": 0.065,
+      "step": 16957
+    },
+    {
+      "epoch": 2.7475696694750487,
+      "grad_norm": 0.9515754580497742,
+      "learning_rate": 9.226614186551852e-08,
+      "loss": 0.0578,
+      "step": 16958
+    },
+    {
+      "epoch": 2.7477316915100456,
+      "grad_norm": 0.8326787352561951,
+      "learning_rate": 9.214847090701474e-08,
+      "loss": 0.0537,
+      "step": 16959
+    },
+    {
+      "epoch": 2.747893713545042,
+      "grad_norm": 0.9199459552764893,
+      "learning_rate": 9.20308736229411e-08,
+      "loss": 0.0591,
+      "step": 16960
+    },
+    {
+      "epoch": 2.748055735580039,
+      "grad_norm": 0.9134218692779541,
+      "learning_rate": 9.19133500168956e-08,
+      "loss": 0.0604,
+      "step": 16961
+    },
+    {
+      "epoch": 2.7482177576150355,
+      "grad_norm": 0.8909247517585754,
+      "learning_rate": 9.179590009247397e-08,
+      "loss": 0.0587,
+      "step": 16962
+    },
+    {
+      "epoch": 2.7483797796500324,
+      "grad_norm": 0.8900055885314941,
+      "learning_rate": 9.167852385326969e-08,
+      "loss": 0.0579,
+      "step": 16963
+    },
+    {
+      "epoch": 2.748541801685029,
+      "grad_norm": 1.0126845836639404,
+      "learning_rate": 9.15612213028752e-08,
+      "loss": 0.0661,
+      "step": 16964
+    },
+    {
+      "epoch": 2.748703823720026,
+      "grad_norm": 0.9734671711921692,
+      "learning_rate": 9.144399244487873e-08,
+      "loss": 0.06,
+      "step": 16965
+    },
+    {
+      "epoch": 2.748865845755023,
+      "grad_norm": 0.8340158462524414,
+      "learning_rate": 9.132683728286767e-08,
+      "loss": 0.0552,
+      "step": 16966
+    },
+    {
+      "epoch": 2.7490278677900193,
+      "grad_norm": 0.81150883436203,
+      "learning_rate": 9.120975582042613e-08,
+      "loss": 0.057,
+      "step": 16967
+    },
+    {
+      "epoch": 2.749189889825016,
+      "grad_norm": 0.9296124577522278,
+      "learning_rate": 9.109274806113732e-08,
+      "loss": 0.0647,
+      "step": 16968
+    },
+    {
+      "epoch": 2.749351911860013,
+      "grad_norm": 0.8688836097717285,
+      "learning_rate": 9.097581400858064e-08,
+      "loss": 0.0588,
+      "step": 16969
+    },
+    {
+      "epoch": 2.7495139338950096,
+      "grad_norm": 1.00139319896698,
+      "learning_rate": 9.085895366633457e-08,
+      "loss": 0.0626,
+      "step": 16970
+    },
+    {
+      "epoch": 2.7496759559300066,
+      "grad_norm": 0.9233237504959106,
+      "learning_rate": 9.074216703797434e-08,
+      "loss": 0.0646,
+      "step": 16971
+    },
+    {
+      "epoch": 2.7498379779650035,
+      "grad_norm": 0.954022228717804,
+      "learning_rate": 9.062545412707375e-08,
+      "loss": 0.0637,
+      "step": 16972
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7831946611404419,
+      "learning_rate": 9.050881493720326e-08,
+      "loss": 0.0533,
+      "step": 16973
+    },
+    {
+      "epoch": 2.7501620220349965,
+      "grad_norm": 0.8610273599624634,
+      "learning_rate": 9.039224947193254e-08,
+      "loss": 0.0567,
+      "step": 16974
+    },
+    {
+      "epoch": 2.7503240440699934,
+      "grad_norm": 1.0053431987762451,
+      "learning_rate": 9.027575773482788e-08,
+      "loss": 0.063,
+      "step": 16975
+    },
+    {
+      "epoch": 2.7504860661049904,
+      "grad_norm": 1.0265311002731323,
+      "learning_rate": 9.01593397294534e-08,
+      "loss": 0.068,
+      "step": 16976
+    },
+    {
+      "epoch": 2.750648088139987,
+      "grad_norm": 1.0385494232177734,
+      "learning_rate": 9.004299545937151e-08,
+      "loss": 0.0615,
+      "step": 16977
+    },
+    {
+      "epoch": 2.750810110174984,
+      "grad_norm": 1.0085492134094238,
+      "learning_rate": 8.992672492814158e-08,
+      "loss": 0.0618,
+      "step": 16978
+    },
+    {
+      "epoch": 2.7509721322099807,
+      "grad_norm": 0.9866387248039246,
+      "learning_rate": 8.981052813932245e-08,
+      "loss": 0.0675,
+      "step": 16979
+    },
+    {
+      "epoch": 2.751134154244977,
+      "grad_norm": 1.0213267803192139,
+      "learning_rate": 8.969440509646821e-08,
+      "loss": 0.0586,
+      "step": 16980
+    },
+    {
+      "epoch": 2.751296176279974,
+      "grad_norm": 1.007440447807312,
+      "learning_rate": 8.957835580313212e-08,
+      "loss": 0.0656,
+      "step": 16981
+    },
+    {
+      "epoch": 2.751458198314971,
+      "grad_norm": 0.9071718454360962,
+      "learning_rate": 8.946238026286552e-08,
+      "loss": 0.0619,
+      "step": 16982
+    },
+    {
+      "epoch": 2.7516202203499676,
+      "grad_norm": 1.0188510417938232,
+      "learning_rate": 8.93464784792164e-08,
+      "loss": 0.0566,
+      "step": 16983
+    },
+    {
+      "epoch": 2.7517822423849645,
+      "grad_norm": 0.8976929187774658,
+      "learning_rate": 8.923065045573165e-08,
+      "loss": 0.0603,
+      "step": 16984
+    },
+    {
+      "epoch": 2.751944264419961,
+      "grad_norm": 0.7960435748100281,
+      "learning_rate": 8.911489619595482e-08,
+      "loss": 0.0537,
+      "step": 16985
+    },
+    {
+      "epoch": 2.752106286454958,
+      "grad_norm": 1.094245195388794,
+      "learning_rate": 8.899921570342807e-08,
+      "loss": 0.0682,
+      "step": 16986
+    },
+    {
+      "epoch": 2.7522683084899544,
+      "grad_norm": 0.8297873139381409,
+      "learning_rate": 8.888360898169079e-08,
+      "loss": 0.0515,
+      "step": 16987
+    },
+    {
+      "epoch": 2.7524303305249513,
+      "grad_norm": 0.8277057409286499,
+      "learning_rate": 8.876807603428017e-08,
+      "loss": 0.0569,
+      "step": 16988
+    },
+    {
+      "epoch": 2.7525923525599483,
+      "grad_norm": 0.8910994529724121,
+      "learning_rate": 8.865261686473143e-08,
+      "loss": 0.0529,
+      "step": 16989
+    },
+    {
+      "epoch": 2.7527543745949448,
+      "grad_norm": 0.798820436000824,
+      "learning_rate": 8.853723147657755e-08,
+      "loss": 0.0544,
+      "step": 16990
+    },
+    {
+      "epoch": 2.7529163966299417,
+      "grad_norm": 0.994156539440155,
+      "learning_rate": 8.842191987334853e-08,
+      "loss": 0.0661,
+      "step": 16991
+    },
+    {
+      "epoch": 2.7530784186649386,
+      "grad_norm": 0.8740126490592957,
+      "learning_rate": 8.830668205857263e-08,
+      "loss": 0.0599,
+      "step": 16992
+    },
+    {
+      "epoch": 2.753240440699935,
+      "grad_norm": 0.9140499830245972,
+      "learning_rate": 8.819151803577647e-08,
+      "loss": 0.0605,
+      "step": 16993
+    },
+    {
+      "epoch": 2.753402462734932,
+      "grad_norm": 0.9925417304039001,
+      "learning_rate": 8.807642780848335e-08,
+      "loss": 0.0623,
+      "step": 16994
+    },
+    {
+      "epoch": 2.7535644847699285,
+      "grad_norm": 0.8395823836326599,
+      "learning_rate": 8.796141138021464e-08,
+      "loss": 0.0575,
+      "step": 16995
+    },
+    {
+      "epoch": 2.7537265068049255,
+      "grad_norm": 1.012656331062317,
+      "learning_rate": 8.784646875448971e-08,
+      "loss": 0.0669,
+      "step": 16996
+    },
+    {
+      "epoch": 2.753888528839922,
+      "grad_norm": 0.8465946316719055,
+      "learning_rate": 8.77315999348255e-08,
+      "loss": 0.0572,
+      "step": 16997
+    },
+    {
+      "epoch": 2.754050550874919,
+      "grad_norm": 1.1134271621704102,
+      "learning_rate": 8.761680492473668e-08,
+      "loss": 0.058,
+      "step": 16998
+    },
+    {
+      "epoch": 2.754212572909916,
+      "grad_norm": 0.852190375328064,
+      "learning_rate": 8.7502083727736e-08,
+      "loss": 0.0576,
+      "step": 16999
+    },
+    {
+      "epoch": 2.7543745949449123,
+      "grad_norm": 0.9086182713508606,
+      "learning_rate": 8.738743634733316e-08,
+      "loss": 0.0631,
+      "step": 17000
+    },
+    {
+      "epoch": 2.7545366169799093,
+      "grad_norm": 0.9035629034042358,
+      "learning_rate": 8.727286278703672e-08,
+      "loss": 0.0589,
+      "step": 17001
+    },
+    {
+      "epoch": 2.754698639014906,
+      "grad_norm": 0.8812063932418823,
+      "learning_rate": 8.715836305035169e-08,
+      "loss": 0.0516,
+      "step": 17002
+    },
+    {
+      "epoch": 2.7548606610499027,
+      "grad_norm": 0.9071505069732666,
+      "learning_rate": 8.704393714078191e-08,
+      "loss": 0.0585,
+      "step": 17003
+    },
+    {
+      "epoch": 2.7550226830848996,
+      "grad_norm": 0.9533473253250122,
+      "learning_rate": 8.692958506182847e-08,
+      "loss": 0.0562,
+      "step": 17004
+    },
+    {
+      "epoch": 2.7551847051198965,
+      "grad_norm": 0.8007147312164307,
+      "learning_rate": 8.681530681699024e-08,
+      "loss": 0.0611,
+      "step": 17005
+    },
+    {
+      "epoch": 2.755346727154893,
+      "grad_norm": 0.9169260263442993,
+      "learning_rate": 8.67011024097636e-08,
+      "loss": 0.065,
+      "step": 17006
+    },
+    {
+      "epoch": 2.75550874918989,
+      "grad_norm": 0.8631471395492554,
+      "learning_rate": 8.658697184364323e-08,
+      "loss": 0.0504,
+      "step": 17007
+    },
+    {
+      "epoch": 2.7556707712248865,
+      "grad_norm": 0.9157608151435852,
+      "learning_rate": 8.647291512212136e-08,
+      "loss": 0.0572,
+      "step": 17008
+    },
+    {
+      "epoch": 2.7558327932598834,
+      "grad_norm": 0.9917179942131042,
+      "learning_rate": 8.635893224868769e-08,
+      "loss": 0.064,
+      "step": 17009
+    },
+    {
+      "epoch": 2.75599481529488,
+      "grad_norm": 0.9392203092575073,
+      "learning_rate": 8.624502322682942e-08,
+      "loss": 0.0655,
+      "step": 17010
+    },
+    {
+      "epoch": 2.756156837329877,
+      "grad_norm": 0.8594088554382324,
+      "learning_rate": 8.61311880600324e-08,
+      "loss": 0.0567,
+      "step": 17011
+    },
+    {
+      "epoch": 2.7563188593648738,
+      "grad_norm": 0.7704145312309265,
+      "learning_rate": 8.601742675177993e-08,
+      "loss": 0.0494,
+      "step": 17012
+    },
+    {
+      "epoch": 2.7564808813998702,
+      "grad_norm": 0.9152031540870667,
+      "learning_rate": 8.590373930555201e-08,
+      "loss": 0.0579,
+      "step": 17013
+    },
+    {
+      "epoch": 2.756642903434867,
+      "grad_norm": 0.8692240715026855,
+      "learning_rate": 8.57901257248278e-08,
+      "loss": 0.0578,
+      "step": 17014
+    },
+    {
+      "epoch": 2.756804925469864,
+      "grad_norm": 0.8714682459831238,
+      "learning_rate": 8.567658601308371e-08,
+      "loss": 0.0523,
+      "step": 17015
+    },
+    {
+      "epoch": 2.7569669475048606,
+      "grad_norm": 0.9357670545578003,
+      "learning_rate": 8.556312017379332e-08,
+      "loss": 0.0581,
+      "step": 17016
+    },
+    {
+      "epoch": 2.7571289695398575,
+      "grad_norm": 0.9461519718170166,
+      "learning_rate": 8.544972821042857e-08,
+      "loss": 0.0652,
+      "step": 17017
+    },
+    {
+      "epoch": 2.757290991574854,
+      "grad_norm": 0.7985438108444214,
+      "learning_rate": 8.533641012645921e-08,
+      "loss": 0.0567,
+      "step": 17018
+    },
+    {
+      "epoch": 2.757453013609851,
+      "grad_norm": 0.8582321405410767,
+      "learning_rate": 8.522316592535246e-08,
+      "loss": 0.0545,
+      "step": 17019
+    },
+    {
+      "epoch": 2.7576150356448474,
+      "grad_norm": 0.9717732071876526,
+      "learning_rate": 8.510999561057276e-08,
+      "loss": 0.0655,
+      "step": 17020
+    },
+    {
+      "epoch": 2.7577770576798444,
+      "grad_norm": 0.8554613590240479,
+      "learning_rate": 8.499689918558318e-08,
+      "loss": 0.0555,
+      "step": 17021
+    },
+    {
+      "epoch": 2.7579390797148413,
+      "grad_norm": 0.9778949618339539,
+      "learning_rate": 8.488387665384457e-08,
+      "loss": 0.0646,
+      "step": 17022
+    },
+    {
+      "epoch": 2.758101101749838,
+      "grad_norm": 1.0577203035354614,
+      "learning_rate": 8.477092801881525e-08,
+      "loss": 0.0594,
+      "step": 17023
+    },
+    {
+      "epoch": 2.7582631237848347,
+      "grad_norm": 0.9223877787590027,
+      "learning_rate": 8.465805328395055e-08,
+      "loss": 0.0574,
+      "step": 17024
+    },
+    {
+      "epoch": 2.7584251458198317,
+      "grad_norm": 0.7924177050590515,
+      "learning_rate": 8.454525245270378e-08,
+      "loss": 0.0518,
+      "step": 17025
+    },
+    {
+      "epoch": 2.758587167854828,
+      "grad_norm": 0.8869810700416565,
+      "learning_rate": 8.443252552852776e-08,
+      "loss": 0.0636,
+      "step": 17026
+    },
+    {
+      "epoch": 2.758749189889825,
+      "grad_norm": 0.927229106426239,
+      "learning_rate": 8.431987251487083e-08,
+      "loss": 0.0608,
+      "step": 17027
+    },
+    {
+      "epoch": 2.758911211924822,
+      "grad_norm": 1.047018051147461,
+      "learning_rate": 8.42072934151797e-08,
+      "loss": 0.0618,
+      "step": 17028
+    },
+    {
+      "epoch": 2.7590732339598185,
+      "grad_norm": 0.8683516979217529,
+      "learning_rate": 8.409478823289934e-08,
+      "loss": 0.0614,
+      "step": 17029
+    },
+    {
+      "epoch": 2.7592352559948155,
+      "grad_norm": 0.9671280384063721,
+      "learning_rate": 8.398235697147205e-08,
+      "loss": 0.0533,
+      "step": 17030
+    },
+    {
+      "epoch": 2.759397278029812,
+      "grad_norm": 0.8958953619003296,
+      "learning_rate": 8.386999963433812e-08,
+      "loss": 0.0573,
+      "step": 17031
+    },
+    {
+      "epoch": 2.759559300064809,
+      "grad_norm": 0.8084527850151062,
+      "learning_rate": 8.375771622493506e-08,
+      "loss": 0.0498,
+      "step": 17032
+    },
+    {
+      "epoch": 2.7597213220998054,
+      "grad_norm": 0.838610827922821,
+      "learning_rate": 8.364550674669875e-08,
+      "loss": 0.0635,
+      "step": 17033
+    },
+    {
+      "epoch": 2.7598833441348023,
+      "grad_norm": 0.9222990274429321,
+      "learning_rate": 8.353337120306282e-08,
+      "loss": 0.0647,
+      "step": 17034
+    },
+    {
+      "epoch": 2.7600453661697992,
+      "grad_norm": 1.0761842727661133,
+      "learning_rate": 8.342130959745731e-08,
+      "loss": 0.0665,
+      "step": 17035
+    },
+    {
+      "epoch": 2.7602073882047957,
+      "grad_norm": 0.8044500350952148,
+      "learning_rate": 8.330932193331226e-08,
+      "loss": 0.0568,
+      "step": 17036
+    },
+    {
+      "epoch": 2.7603694102397927,
+      "grad_norm": 0.8677234649658203,
+      "learning_rate": 8.319740821405354e-08,
+      "loss": 0.0588,
+      "step": 17037
+    },
+    {
+      "epoch": 2.7605314322747896,
+      "grad_norm": 1.054807424545288,
+      "learning_rate": 8.308556844310589e-08,
+      "loss": 0.0701,
+      "step": 17038
+    },
+    {
+      "epoch": 2.760693454309786,
+      "grad_norm": 1.1043615341186523,
+      "learning_rate": 8.297380262389077e-08,
+      "loss": 0.0624,
+      "step": 17039
+    },
+    {
+      "epoch": 2.760855476344783,
+      "grad_norm": 0.7604734301567078,
+      "learning_rate": 8.286211075982764e-08,
+      "loss": 0.0498,
+      "step": 17040
+    },
+    {
+      "epoch": 2.7610174983797795,
+      "grad_norm": 0.8906594514846802,
+      "learning_rate": 8.275049285433545e-08,
+      "loss": 0.0596,
+      "step": 17041
+    },
+    {
+      "epoch": 2.7611795204147764,
+      "grad_norm": 0.9391190409660339,
+      "learning_rate": 8.263894891082813e-08,
+      "loss": 0.0644,
+      "step": 17042
+    },
+    {
+      "epoch": 2.761341542449773,
+      "grad_norm": 0.9467473030090332,
+      "learning_rate": 8.252747893271906e-08,
+      "loss": 0.0614,
+      "step": 17043
+    },
+    {
+      "epoch": 2.76150356448477,
+      "grad_norm": 0.8915271162986755,
+      "learning_rate": 8.241608292341913e-08,
+      "loss": 0.0602,
+      "step": 17044
+    },
+    {
+      "epoch": 2.761665586519767,
+      "grad_norm": 0.9345802068710327,
+      "learning_rate": 8.230476088633644e-08,
+      "loss": 0.0603,
+      "step": 17045
+    },
+    {
+      "epoch": 2.7618276085547633,
+      "grad_norm": 0.8942722082138062,
+      "learning_rate": 8.219351282487742e-08,
+      "loss": 0.0602,
+      "step": 17046
+    },
+    {
+      "epoch": 2.76198963058976,
+      "grad_norm": 0.9697921276092529,
+      "learning_rate": 8.208233874244575e-08,
+      "loss": 0.0628,
+      "step": 17047
+    },
+    {
+      "epoch": 2.762151652624757,
+      "grad_norm": 0.877307116985321,
+      "learning_rate": 8.197123864244344e-08,
+      "loss": 0.0578,
+      "step": 17048
+    },
+    {
+      "epoch": 2.7623136746597536,
+      "grad_norm": 0.9268743395805359,
+      "learning_rate": 8.18602125282697e-08,
+      "loss": 0.0567,
+      "step": 17049
+    },
+    {
+      "epoch": 2.7624756966947506,
+      "grad_norm": 0.8506268262863159,
+      "learning_rate": 8.174926040332182e-08,
+      "loss": 0.0533,
+      "step": 17050
+    },
+    {
+      "epoch": 2.7626377187297475,
+      "grad_norm": 0.9653171896934509,
+      "learning_rate": 8.16383822709943e-08,
+      "loss": 0.0601,
+      "step": 17051
+    },
+    {
+      "epoch": 2.762799740764744,
+      "grad_norm": 0.9039610028266907,
+      "learning_rate": 8.152757813468027e-08,
+      "loss": 0.0585,
+      "step": 17052
+    },
+    {
+      "epoch": 2.762961762799741,
+      "grad_norm": 0.8565463423728943,
+      "learning_rate": 8.14168479977695e-08,
+      "loss": 0.0521,
+      "step": 17053
+    },
+    {
+      "epoch": 2.7631237848347374,
+      "grad_norm": 0.8485569357872009,
+      "learning_rate": 8.130619186365012e-08,
+      "loss": 0.0554,
+      "step": 17054
+    },
+    {
+      "epoch": 2.7632858068697344,
+      "grad_norm": 0.872444212436676,
+      "learning_rate": 8.119560973570834e-08,
+      "loss": 0.0548,
+      "step": 17055
+    },
+    {
+      "epoch": 2.763447828904731,
+      "grad_norm": 0.9443418979644775,
+      "learning_rate": 8.10851016173278e-08,
+      "loss": 0.0645,
+      "step": 17056
+    },
+    {
+      "epoch": 2.7636098509397278,
+      "grad_norm": 0.9589036703109741,
+      "learning_rate": 8.097466751188915e-08,
+      "loss": 0.064,
+      "step": 17057
+    },
+    {
+      "epoch": 2.7637718729747247,
+      "grad_norm": 0.9315710663795471,
+      "learning_rate": 8.086430742277191e-08,
+      "loss": 0.0634,
+      "step": 17058
+    },
+    {
+      "epoch": 2.763933895009721,
+      "grad_norm": 1.0759693384170532,
+      "learning_rate": 8.075402135335253e-08,
+      "loss": 0.0581,
+      "step": 17059
+    },
+    {
+      "epoch": 2.764095917044718,
+      "grad_norm": 0.9197863340377808,
+      "learning_rate": 8.064380930700556e-08,
+      "loss": 0.0607,
+      "step": 17060
+    },
+    {
+      "epoch": 2.764257939079715,
+      "grad_norm": 0.8800620436668396,
+      "learning_rate": 8.053367128710355e-08,
+      "loss": 0.0556,
+      "step": 17061
+    },
+    {
+      "epoch": 2.7644199611147116,
+      "grad_norm": 1.0651123523712158,
+      "learning_rate": 8.042360729701604e-08,
+      "loss": 0.0606,
+      "step": 17062
+    },
+    {
+      "epoch": 2.7645819831497085,
+      "grad_norm": 0.8733367323875427,
+      "learning_rate": 8.031361734011118e-08,
+      "loss": 0.0663,
+      "step": 17063
+    },
+    {
+      "epoch": 2.764744005184705,
+      "grad_norm": 1.1324882507324219,
+      "learning_rate": 8.020370141975347e-08,
+      "loss": 0.0627,
+      "step": 17064
+    },
+    {
+      "epoch": 2.764906027219702,
+      "grad_norm": 0.9091924428939819,
+      "learning_rate": 8.009385953930721e-08,
+      "loss": 0.0635,
+      "step": 17065
+    },
+    {
+      "epoch": 2.7650680492546984,
+      "grad_norm": 0.9367949962615967,
+      "learning_rate": 7.998409170213245e-08,
+      "loss": 0.0591,
+      "step": 17066
+    },
+    {
+      "epoch": 2.7652300712896953,
+      "grad_norm": 0.7516220211982727,
+      "learning_rate": 7.987439791158874e-08,
+      "loss": 0.0515,
+      "step": 17067
+    },
+    {
+      "epoch": 2.7653920933246923,
+      "grad_norm": 1.0401772260665894,
+      "learning_rate": 7.976477817103117e-08,
+      "loss": 0.0612,
+      "step": 17068
+    },
+    {
+      "epoch": 2.7655541153596888,
+      "grad_norm": 0.9251458644866943,
+      "learning_rate": 7.965523248381485e-08,
+      "loss": 0.0598,
+      "step": 17069
+    },
+    {
+      "epoch": 2.7657161373946857,
+      "grad_norm": 1.0257396697998047,
+      "learning_rate": 7.954576085329152e-08,
+      "loss": 0.0613,
+      "step": 17070
+    },
+    {
+      "epoch": 2.7658781594296826,
+      "grad_norm": 1.0854121446609497,
+      "learning_rate": 7.943636328281018e-08,
+      "loss": 0.0627,
+      "step": 17071
+    },
+    {
+      "epoch": 2.766040181464679,
+      "grad_norm": 0.9566676616668701,
+      "learning_rate": 7.93270397757187e-08,
+      "loss": 0.0564,
+      "step": 17072
+    },
+    {
+      "epoch": 2.766202203499676,
+      "grad_norm": 0.9054695963859558,
+      "learning_rate": 7.921779033536137e-08,
+      "loss": 0.0537,
+      "step": 17073
+    },
+    {
+      "epoch": 2.766364225534673,
+      "grad_norm": 1.10857093334198,
+      "learning_rate": 7.910861496508216e-08,
+      "loss": 0.0631,
+      "step": 17074
+    },
+    {
+      "epoch": 2.7665262475696695,
+      "grad_norm": 1.0174477100372314,
+      "learning_rate": 7.899951366822061e-08,
+      "loss": 0.0657,
+      "step": 17075
+    },
+    {
+      "epoch": 2.766688269604666,
+      "grad_norm": 0.9663045406341553,
+      "learning_rate": 7.88904864481152e-08,
+      "loss": 0.061,
+      "step": 17076
+    },
+    {
+      "epoch": 2.766850291639663,
+      "grad_norm": 0.9540015459060669,
+      "learning_rate": 7.878153330810184e-08,
+      "loss": 0.0563,
+      "step": 17077
+    },
+    {
+      "epoch": 2.76701231367466,
+      "grad_norm": 0.9780965447425842,
+      "learning_rate": 7.867265425151454e-08,
+      "loss": 0.0641,
+      "step": 17078
+    },
+    {
+      "epoch": 2.7671743357096563,
+      "grad_norm": 0.9400720596313477,
+      "learning_rate": 7.856384928168426e-08,
+      "loss": 0.0663,
+      "step": 17079
+    },
+    {
+      "epoch": 2.7673363577446533,
+      "grad_norm": 1.059554934501648,
+      "learning_rate": 7.845511840194081e-08,
+      "loss": 0.0611,
+      "step": 17080
+    },
+    {
+      "epoch": 2.76749837977965,
+      "grad_norm": 0.911750078201294,
+      "learning_rate": 7.834646161561044e-08,
+      "loss": 0.0582,
+      "step": 17081
+    },
+    {
+      "epoch": 2.7676604018146467,
+      "grad_norm": 1.0574133396148682,
+      "learning_rate": 7.823787892601825e-08,
+      "loss": 0.0648,
+      "step": 17082
+    },
+    {
+      "epoch": 2.7678224238496436,
+      "grad_norm": 0.8240780234336853,
+      "learning_rate": 7.812937033648604e-08,
+      "loss": 0.0583,
+      "step": 17083
+    },
+    {
+      "epoch": 2.7679844458846405,
+      "grad_norm": 1.0235298871994019,
+      "learning_rate": 7.802093585033449e-08,
+      "loss": 0.0694,
+      "step": 17084
+    },
+    {
+      "epoch": 2.768146467919637,
+      "grad_norm": 1.218329668045044,
+      "learning_rate": 7.79125754708815e-08,
+      "loss": 0.0731,
+      "step": 17085
+    },
+    {
+      "epoch": 2.768308489954634,
+      "grad_norm": 0.8903065919876099,
+      "learning_rate": 7.780428920144217e-08,
+      "loss": 0.0582,
+      "step": 17086
+    },
+    {
+      "epoch": 2.7684705119896305,
+      "grad_norm": 0.8183029294013977,
+      "learning_rate": 7.769607704532972e-08,
+      "loss": 0.0587,
+      "step": 17087
+    },
+    {
+      "epoch": 2.7686325340246274,
+      "grad_norm": 0.8426903486251831,
+      "learning_rate": 7.758793900585565e-08,
+      "loss": 0.0606,
+      "step": 17088
+    },
+    {
+      "epoch": 2.768794556059624,
+      "grad_norm": 1.1522109508514404,
+      "learning_rate": 7.747987508632871e-08,
+      "loss": 0.0576,
+      "step": 17089
+    },
+    {
+      "epoch": 2.768956578094621,
+      "grad_norm": 0.7906349897384644,
+      "learning_rate": 7.737188529005484e-08,
+      "loss": 0.0545,
+      "step": 17090
+    },
+    {
+      "epoch": 2.7691186001296177,
+      "grad_norm": 0.8786861896514893,
+      "learning_rate": 7.726396962033894e-08,
+      "loss": 0.0612,
+      "step": 17091
+    },
+    {
+      "epoch": 2.7692806221646142,
+      "grad_norm": 1.079187273979187,
+      "learning_rate": 7.715612808048251e-08,
+      "loss": 0.0612,
+      "step": 17092
+    },
+    {
+      "epoch": 2.769442644199611,
+      "grad_norm": 1.0963796377182007,
+      "learning_rate": 7.70483606737854e-08,
+      "loss": 0.0613,
+      "step": 17093
+    },
+    {
+      "epoch": 2.769604666234608,
+      "grad_norm": 0.9393141269683838,
+      "learning_rate": 7.6940667403545e-08,
+      "loss": 0.0641,
+      "step": 17094
+    },
+    {
+      "epoch": 2.7697666882696046,
+      "grad_norm": 0.9409939050674438,
+      "learning_rate": 7.683304827305644e-08,
+      "loss": 0.0603,
+      "step": 17095
+    },
+    {
+      "epoch": 2.7699287103046015,
+      "grad_norm": 0.8602232933044434,
+      "learning_rate": 7.672550328561318e-08,
+      "loss": 0.0574,
+      "step": 17096
+    },
+    {
+      "epoch": 2.7700907323395985,
+      "grad_norm": 1.0245064496994019,
+      "learning_rate": 7.661803244450455e-08,
+      "loss": 0.0628,
+      "step": 17097
+    },
+    {
+      "epoch": 2.770252754374595,
+      "grad_norm": 0.8060587048530579,
+      "learning_rate": 7.651063575301986e-08,
+      "loss": 0.0561,
+      "step": 17098
+    },
+    {
+      "epoch": 2.7704147764095914,
+      "grad_norm": 1.021560549736023,
+      "learning_rate": 7.64033132144451e-08,
+      "loss": 0.0612,
+      "step": 17099
+    },
+    {
+      "epoch": 2.7705767984445884,
+      "grad_norm": 1.0078121423721313,
+      "learning_rate": 7.62960648320643e-08,
+      "loss": 0.0581,
+      "step": 17100
+    },
+    {
+      "epoch": 2.7707388204795853,
+      "grad_norm": 0.9183746576309204,
+      "learning_rate": 7.618889060915819e-08,
+      "loss": 0.0544,
+      "step": 17101
+    },
+    {
+      "epoch": 2.770900842514582,
+      "grad_norm": 0.9616214632987976,
+      "learning_rate": 7.608179054900634e-08,
+      "loss": 0.0645,
+      "step": 17102
+    },
+    {
+      "epoch": 2.7710628645495787,
+      "grad_norm": 1.02703058719635,
+      "learning_rate": 7.597476465488668e-08,
+      "loss": 0.0657,
+      "step": 17103
+    },
+    {
+      "epoch": 2.7712248865845757,
+      "grad_norm": 0.9865126609802246,
+      "learning_rate": 7.586781293007273e-08,
+      "loss": 0.0653,
+      "step": 17104
+    },
+    {
+      "epoch": 2.771386908619572,
+      "grad_norm": 0.8494060039520264,
+      "learning_rate": 7.57609353778374e-08,
+      "loss": 0.0586,
+      "step": 17105
+    },
+    {
+      "epoch": 2.771548930654569,
+      "grad_norm": 0.9772338271141052,
+      "learning_rate": 7.565413200145089e-08,
+      "loss": 0.0611,
+      "step": 17106
+    },
+    {
+      "epoch": 2.771710952689566,
+      "grad_norm": 1.0353899002075195,
+      "learning_rate": 7.55474028041811e-08,
+      "loss": 0.0631,
+      "step": 17107
+    },
+    {
+      "epoch": 2.7718729747245625,
+      "grad_norm": 0.9081644415855408,
+      "learning_rate": 7.544074778929378e-08,
+      "loss": 0.0616,
+      "step": 17108
+    },
+    {
+      "epoch": 2.7720349967595594,
+      "grad_norm": 1.0636510848999023,
+      "learning_rate": 7.533416696005242e-08,
+      "loss": 0.0683,
+      "step": 17109
+    },
+    {
+      "epoch": 2.772197018794556,
+      "grad_norm": 0.981203556060791,
+      "learning_rate": 7.522766031971774e-08,
+      "loss": 0.0553,
+      "step": 17110
+    },
+    {
+      "epoch": 2.772359040829553,
+      "grad_norm": 0.9812530279159546,
+      "learning_rate": 7.512122787154908e-08,
+      "loss": 0.0583,
+      "step": 17111
+    },
+    {
+      "epoch": 2.7725210628645494,
+      "grad_norm": 0.9434771537780762,
+      "learning_rate": 7.501486961880245e-08,
+      "loss": 0.0635,
+      "step": 17112
+    },
+    {
+      "epoch": 2.7726830848995463,
+      "grad_norm": 0.9104921817779541,
+      "learning_rate": 7.490858556473246e-08,
+      "loss": 0.056,
+      "step": 17113
+    },
+    {
+      "epoch": 2.7728451069345432,
+      "grad_norm": 0.996190071105957,
+      "learning_rate": 7.480237571259153e-08,
+      "loss": 0.0633,
+      "step": 17114
+    },
+    {
+      "epoch": 2.7730071289695397,
+      "grad_norm": 0.8751932978630066,
+      "learning_rate": 7.469624006562898e-08,
+      "loss": 0.0574,
+      "step": 17115
+    },
+    {
+      "epoch": 2.7731691510045366,
+      "grad_norm": 0.8272261023521423,
+      "learning_rate": 7.459017862709194e-08,
+      "loss": 0.0515,
+      "step": 17116
+    },
+    {
+      "epoch": 2.7733311730395336,
+      "grad_norm": 0.916673481464386,
+      "learning_rate": 7.448419140022616e-08,
+      "loss": 0.0609,
+      "step": 17117
+    },
+    {
+      "epoch": 2.77349319507453,
+      "grad_norm": 1.0777217149734497,
+      "learning_rate": 7.437827838827488e-08,
+      "loss": 0.0615,
+      "step": 17118
+    },
+    {
+      "epoch": 2.773655217109527,
+      "grad_norm": 1.0199631452560425,
+      "learning_rate": 7.4272439594478e-08,
+      "loss": 0.0547,
+      "step": 17119
+    },
+    {
+      "epoch": 2.7738172391445235,
+      "grad_norm": 0.8210328817367554,
+      "learning_rate": 7.416667502207458e-08,
+      "loss": 0.0541,
+      "step": 17120
+    },
+    {
+      "epoch": 2.7739792611795204,
+      "grad_norm": 0.8746510148048401,
+      "learning_rate": 7.40609846743004e-08,
+      "loss": 0.0565,
+      "step": 17121
+    },
+    {
+      "epoch": 2.774141283214517,
+      "grad_norm": 0.8559312224388123,
+      "learning_rate": 7.395536855438923e-08,
+      "loss": 0.0582,
+      "step": 17122
+    },
+    {
+      "epoch": 2.774303305249514,
+      "grad_norm": 1.033912181854248,
+      "learning_rate": 7.384982666557322e-08,
+      "loss": 0.0567,
+      "step": 17123
+    },
+    {
+      "epoch": 2.774465327284511,
+      "grad_norm": 0.8947577476501465,
+      "learning_rate": 7.37443590110809e-08,
+      "loss": 0.0601,
+      "step": 17124
+    },
+    {
+      "epoch": 2.7746273493195073,
+      "grad_norm": 0.901314377784729,
+      "learning_rate": 7.363896559414024e-08,
+      "loss": 0.0593,
+      "step": 17125
+    },
+    {
+      "epoch": 2.774789371354504,
+      "grad_norm": 0.8654842972755432,
+      "learning_rate": 7.353364641797533e-08,
+      "loss": 0.0636,
+      "step": 17126
+    },
+    {
+      "epoch": 2.774951393389501,
+      "grad_norm": 0.9468207955360413,
+      "learning_rate": 7.342840148580888e-08,
+      "loss": 0.0626,
+      "step": 17127
+    },
+    {
+      "epoch": 2.7751134154244976,
+      "grad_norm": 0.8542966246604919,
+      "learning_rate": 7.332323080086106e-08,
+      "loss": 0.0542,
+      "step": 17128
+    },
+    {
+      "epoch": 2.7752754374594946,
+      "grad_norm": 0.9125382900238037,
+      "learning_rate": 7.321813436635044e-08,
+      "loss": 0.0621,
+      "step": 17129
+    },
+    {
+      "epoch": 2.7754374594944915,
+      "grad_norm": 0.9692714810371399,
+      "learning_rate": 7.311311218549166e-08,
+      "loss": 0.0604,
+      "step": 17130
+    },
+    {
+      "epoch": 2.775599481529488,
+      "grad_norm": 0.8632164597511292,
+      "learning_rate": 7.300816426149854e-08,
+      "loss": 0.0619,
+      "step": 17131
+    },
+    {
+      "epoch": 2.775761503564485,
+      "grad_norm": 0.9177026748657227,
+      "learning_rate": 7.290329059758294e-08,
+      "loss": 0.0648,
+      "step": 17132
+    },
+    {
+      "epoch": 2.7759235255994814,
+      "grad_norm": 0.9254752993583679,
+      "learning_rate": 7.279849119695314e-08,
+      "loss": 0.0649,
+      "step": 17133
+    },
+    {
+      "epoch": 2.7760855476344783,
+      "grad_norm": 1.0560814142227173,
+      "learning_rate": 7.269376606281547e-08,
+      "loss": 0.0711,
+      "step": 17134
+    },
+    {
+      "epoch": 2.776247569669475,
+      "grad_norm": 0.8399720788002014,
+      "learning_rate": 7.258911519837486e-08,
+      "loss": 0.0619,
+      "step": 17135
+    },
+    {
+      "epoch": 2.7764095917044718,
+      "grad_norm": 0.8254349827766418,
+      "learning_rate": 7.248453860683291e-08,
+      "loss": 0.0566,
+      "step": 17136
+    },
+    {
+      "epoch": 2.7765716137394687,
+      "grad_norm": 0.8527083992958069,
+      "learning_rate": 7.238003629138957e-08,
+      "loss": 0.0586,
+      "step": 17137
+    },
+    {
+      "epoch": 2.776733635774465,
+      "grad_norm": 1.0240825414657593,
+      "learning_rate": 7.227560825524255e-08,
+      "loss": 0.0664,
+      "step": 17138
+    },
+    {
+      "epoch": 2.776895657809462,
+      "grad_norm": 1.000952124595642,
+      "learning_rate": 7.21712545015868e-08,
+      "loss": 0.0619,
+      "step": 17139
+    },
+    {
+      "epoch": 2.777057679844459,
+      "grad_norm": 0.9190289378166199,
+      "learning_rate": 7.20669750336156e-08,
+      "loss": 0.064,
+      "step": 17140
+    },
+    {
+      "epoch": 2.7772197018794555,
+      "grad_norm": 0.967212975025177,
+      "learning_rate": 7.196276985451916e-08,
+      "loss": 0.0608,
+      "step": 17141
+    },
+    {
+      "epoch": 2.7773817239144525,
+      "grad_norm": 0.7784137725830078,
+      "learning_rate": 7.185863896748662e-08,
+      "loss": 0.0556,
+      "step": 17142
+    },
+    {
+      "epoch": 2.777543745949449,
+      "grad_norm": 0.9892643094062805,
+      "learning_rate": 7.175458237570349e-08,
+      "loss": 0.0673,
+      "step": 17143
+    },
+    {
+      "epoch": 2.777705767984446,
+      "grad_norm": 0.872578501701355,
+      "learning_rate": 7.165060008235414e-08,
+      "loss": 0.0567,
+      "step": 17144
+    },
+    {
+      "epoch": 2.7778677900194424,
+      "grad_norm": 0.8793370723724365,
+      "learning_rate": 7.154669209061965e-08,
+      "loss": 0.0538,
+      "step": 17145
+    },
+    {
+      "epoch": 2.7780298120544393,
+      "grad_norm": 1.0323148965835571,
+      "learning_rate": 7.14428584036797e-08,
+      "loss": 0.058,
+      "step": 17146
+    },
+    {
+      "epoch": 2.7781918340894363,
+      "grad_norm": 0.8175366520881653,
+      "learning_rate": 7.133909902471147e-08,
+      "loss": 0.0573,
+      "step": 17147
+    },
+    {
+      "epoch": 2.7783538561244328,
+      "grad_norm": 0.9729697108268738,
+      "learning_rate": 7.123541395688966e-08,
+      "loss": 0.0602,
+      "step": 17148
+    },
+    {
+      "epoch": 2.7785158781594297,
+      "grad_norm": 0.9307907223701477,
+      "learning_rate": 7.113180320338642e-08,
+      "loss": 0.0591,
+      "step": 17149
+    },
+    {
+      "epoch": 2.7786779001944266,
+      "grad_norm": 0.8893921375274658,
+      "learning_rate": 7.102826676737202e-08,
+      "loss": 0.0552,
+      "step": 17150
+    },
+    {
+      "epoch": 2.778839922229423,
+      "grad_norm": 0.8809131979942322,
+      "learning_rate": 7.0924804652015e-08,
+      "loss": 0.0638,
+      "step": 17151
+    },
+    {
+      "epoch": 2.77900194426442,
+      "grad_norm": 0.9402658939361572,
+      "learning_rate": 7.082141686048066e-08,
+      "loss": 0.0621,
+      "step": 17152
+    },
+    {
+      "epoch": 2.779163966299417,
+      "grad_norm": 1.108227014541626,
+      "learning_rate": 7.071810339593254e-08,
+      "loss": 0.0662,
+      "step": 17153
+    },
+    {
+      "epoch": 2.7793259883344135,
+      "grad_norm": 0.9297798871994019,
+      "learning_rate": 7.061486426153146e-08,
+      "loss": 0.0619,
+      "step": 17154
+    },
+    {
+      "epoch": 2.7794880103694104,
+      "grad_norm": 0.8598953485488892,
+      "learning_rate": 7.051169946043685e-08,
+      "loss": 0.065,
+      "step": 17155
+    },
+    {
+      "epoch": 2.779650032404407,
+      "grad_norm": 1.2274415493011475,
+      "learning_rate": 7.040860899580475e-08,
+      "loss": 0.0584,
+      "step": 17156
+    },
+    {
+      "epoch": 2.779812054439404,
+      "grad_norm": 0.8859155774116516,
+      "learning_rate": 7.030559287078992e-08,
+      "loss": 0.0601,
+      "step": 17157
+    },
+    {
+      "epoch": 2.7799740764744003,
+      "grad_norm": 0.8546262979507446,
+      "learning_rate": 7.020265108854423e-08,
+      "loss": 0.0584,
+      "step": 17158
+    },
+    {
+      "epoch": 2.7801360985093972,
+      "grad_norm": 0.8823912739753723,
+      "learning_rate": 7.009978365221687e-08,
+      "loss": 0.0592,
+      "step": 17159
+    },
+    {
+      "epoch": 2.780298120544394,
+      "grad_norm": 0.9547003507614136,
+      "learning_rate": 6.99969905649564e-08,
+      "loss": 0.0606,
+      "step": 17160
+    },
+    {
+      "epoch": 2.7804601425793907,
+      "grad_norm": 0.7352187633514404,
+      "learning_rate": 6.989427182990727e-08,
+      "loss": 0.0517,
+      "step": 17161
+    },
+    {
+      "epoch": 2.7806221646143876,
+      "grad_norm": 0.7668613791465759,
+      "learning_rate": 6.979162745021306e-08,
+      "loss": 0.0512,
+      "step": 17162
+    },
+    {
+      "epoch": 2.7807841866493845,
+      "grad_norm": 0.9652883410453796,
+      "learning_rate": 6.968905742901405e-08,
+      "loss": 0.0578,
+      "step": 17163
+    },
+    {
+      "epoch": 2.780946208684381,
+      "grad_norm": 0.8641253709793091,
+      "learning_rate": 6.958656176944801e-08,
+      "loss": 0.0599,
+      "step": 17164
+    },
+    {
+      "epoch": 2.781108230719378,
+      "grad_norm": 0.9378582835197449,
+      "learning_rate": 6.94841404746524e-08,
+      "loss": 0.0599,
+      "step": 17165
+    },
+    {
+      "epoch": 2.7812702527543745,
+      "grad_norm": 0.902350664138794,
+      "learning_rate": 6.938179354776003e-08,
+      "loss": 0.0588,
+      "step": 17166
+    },
+    {
+      "epoch": 2.7814322747893714,
+      "grad_norm": 0.9747065305709839,
+      "learning_rate": 6.927952099190282e-08,
+      "loss": 0.0606,
+      "step": 17167
+    },
+    {
+      "epoch": 2.781594296824368,
+      "grad_norm": 0.9471100568771362,
+      "learning_rate": 6.917732281020995e-08,
+      "loss": 0.0641,
+      "step": 17168
+    },
+    {
+      "epoch": 2.781756318859365,
+      "grad_norm": 0.9175786972045898,
+      "learning_rate": 6.907519900580862e-08,
+      "loss": 0.0576,
+      "step": 17169
+    },
+    {
+      "epoch": 2.7819183408943617,
+      "grad_norm": 0.8128597736358643,
+      "learning_rate": 6.897314958182327e-08,
+      "loss": 0.0569,
+      "step": 17170
+    },
+    {
+      "epoch": 2.7820803629293582,
+      "grad_norm": 1.0002559423446655,
+      "learning_rate": 6.887117454137698e-08,
+      "loss": 0.0681,
+      "step": 17171
+    },
+    {
+      "epoch": 2.782242384964355,
+      "grad_norm": 0.9145398736000061,
+      "learning_rate": 6.87692738875892e-08,
+      "loss": 0.0587,
+      "step": 17172
+    },
+    {
+      "epoch": 2.782404406999352,
+      "grad_norm": 0.9303719997406006,
+      "learning_rate": 6.866744762357852e-08,
+      "loss": 0.06,
+      "step": 17173
+    },
+    {
+      "epoch": 2.7825664290343486,
+      "grad_norm": 0.9759852886199951,
+      "learning_rate": 6.856569575245969e-08,
+      "loss": 0.062,
+      "step": 17174
+    },
+    {
+      "epoch": 2.7827284510693455,
+      "grad_norm": 0.9545977711677551,
+      "learning_rate": 6.846401827734689e-08,
+      "loss": 0.0615,
+      "step": 17175
+    },
+    {
+      "epoch": 2.7828904731043425,
+      "grad_norm": 0.8938077092170715,
+      "learning_rate": 6.836241520135123e-08,
+      "loss": 0.0616,
+      "step": 17176
+    },
+    {
+      "epoch": 2.783052495139339,
+      "grad_norm": 1.1049113273620605,
+      "learning_rate": 6.826088652758106e-08,
+      "loss": 0.0655,
+      "step": 17177
+    },
+    {
+      "epoch": 2.7832145171743354,
+      "grad_norm": 0.866930365562439,
+      "learning_rate": 6.815943225914278e-08,
+      "loss": 0.0569,
+      "step": 17178
+    },
+    {
+      "epoch": 2.7833765392093324,
+      "grad_norm": 0.9813682436943054,
+      "learning_rate": 6.805805239914087e-08,
+      "loss": 0.0615,
+      "step": 17179
+    },
+    {
+      "epoch": 2.7835385612443293,
+      "grad_norm": 1.0694559812545776,
+      "learning_rate": 6.795674695067783e-08,
+      "loss": 0.0656,
+      "step": 17180
+    },
+    {
+      "epoch": 2.783700583279326,
+      "grad_norm": 0.8712436556816101,
+      "learning_rate": 6.785551591685257e-08,
+      "loss": 0.0597,
+      "step": 17181
+    },
+    {
+      "epoch": 2.7838626053143227,
+      "grad_norm": 0.7749736905097961,
+      "learning_rate": 6.77543593007629e-08,
+      "loss": 0.0492,
+      "step": 17182
+    },
+    {
+      "epoch": 2.7840246273493197,
+      "grad_norm": 0.8556262850761414,
+      "learning_rate": 6.765327710550412e-08,
+      "loss": 0.062,
+      "step": 17183
+    },
+    {
+      "epoch": 2.784186649384316,
+      "grad_norm": 0.953547477722168,
+      "learning_rate": 6.755226933416876e-08,
+      "loss": 0.0589,
+      "step": 17184
+    },
+    {
+      "epoch": 2.784348671419313,
+      "grad_norm": 0.8861910104751587,
+      "learning_rate": 6.745133598984737e-08,
+      "loss": 0.0543,
+      "step": 17185
+    },
+    {
+      "epoch": 2.78451069345431,
+      "grad_norm": 1.0571638345718384,
+      "learning_rate": 6.735047707562863e-08,
+      "loss": 0.0595,
+      "step": 17186
+    },
+    {
+      "epoch": 2.7846727154893065,
+      "grad_norm": 1.2609553337097168,
+      "learning_rate": 6.72496925945984e-08,
+      "loss": 0.0701,
+      "step": 17187
+    },
+    {
+      "epoch": 2.7848347375243034,
+      "grad_norm": 0.9145808815956116,
+      "learning_rate": 6.714898254984031e-08,
+      "loss": 0.062,
+      "step": 17188
+    },
+    {
+      "epoch": 2.7849967595593,
+      "grad_norm": 0.9036248326301575,
+      "learning_rate": 6.704834694443608e-08,
+      "loss": 0.0625,
+      "step": 17189
+    },
+    {
+      "epoch": 2.785158781594297,
+      "grad_norm": 0.834865391254425,
+      "learning_rate": 6.69477857814646e-08,
+      "loss": 0.0523,
+      "step": 17190
+    },
+    {
+      "epoch": 2.7853208036292934,
+      "grad_norm": 0.823307991027832,
+      "learning_rate": 6.684729906400344e-08,
+      "loss": 0.0572,
+      "step": 17191
+    },
+    {
+      "epoch": 2.7854828256642903,
+      "grad_norm": 1.1076858043670654,
+      "learning_rate": 6.674688679512654e-08,
+      "loss": 0.0676,
+      "step": 17192
+    },
+    {
+      "epoch": 2.785644847699287,
+      "grad_norm": 0.8785462975502014,
+      "learning_rate": 6.66465489779064e-08,
+      "loss": 0.0596,
+      "step": 17193
+    },
+    {
+      "epoch": 2.7858068697342837,
+      "grad_norm": 0.8788356781005859,
+      "learning_rate": 6.654628561541337e-08,
+      "loss": 0.0625,
+      "step": 17194
+    },
+    {
+      "epoch": 2.7859688917692806,
+      "grad_norm": 0.8812581896781921,
+      "learning_rate": 6.644609671071556e-08,
+      "loss": 0.0609,
+      "step": 17195
+    },
+    {
+      "epoch": 2.7861309138042776,
+      "grad_norm": 0.9713631272315979,
+      "learning_rate": 6.634598226687772e-08,
+      "loss": 0.0644,
+      "step": 17196
+    },
+    {
+      "epoch": 2.786292935839274,
+      "grad_norm": 0.881287693977356,
+      "learning_rate": 6.624594228696323e-08,
+      "loss": 0.0618,
+      "step": 17197
+    },
+    {
+      "epoch": 2.786454957874271,
+      "grad_norm": 0.8775894641876221,
+      "learning_rate": 6.614597677403384e-08,
+      "loss": 0.0579,
+      "step": 17198
+    },
+    {
+      "epoch": 2.786616979909268,
+      "grad_norm": 0.877691388130188,
+      "learning_rate": 6.604608573114735e-08,
+      "loss": 0.0617,
+      "step": 17199
+    },
+    {
+      "epoch": 2.7867790019442644,
+      "grad_norm": 0.8632456064224243,
+      "learning_rate": 6.594626916136077e-08,
+      "loss": 0.0617,
+      "step": 17200
+    },
+    {
+      "epoch": 2.786941023979261,
+      "grad_norm": 1.1622703075408936,
+      "learning_rate": 6.584652706772804e-08,
+      "loss": 0.0661,
+      "step": 17201
+    },
+    {
+      "epoch": 2.787103046014258,
+      "grad_norm": 0.9076300263404846,
+      "learning_rate": 6.574685945330145e-08,
+      "loss": 0.0588,
+      "step": 17202
+    },
+    {
+      "epoch": 2.787265068049255,
+      "grad_norm": 0.931695282459259,
+      "learning_rate": 6.564726632112939e-08,
+      "loss": 0.0568,
+      "step": 17203
+    },
+    {
+      "epoch": 2.7874270900842513,
+      "grad_norm": 1.018306016921997,
+      "learning_rate": 6.554774767426026e-08,
+      "loss": 0.0644,
+      "step": 17204
+    },
+    {
+      "epoch": 2.787589112119248,
+      "grad_norm": 0.9979287981987,
+      "learning_rate": 6.544830351573883e-08,
+      "loss": 0.0625,
+      "step": 17205
+    },
+    {
+      "epoch": 2.787751134154245,
+      "grad_norm": 0.8496156334877014,
+      "learning_rate": 6.534893384860824e-08,
+      "loss": 0.0638,
+      "step": 17206
+    },
+    {
+      "epoch": 2.7879131561892416,
+      "grad_norm": 0.8916325569152832,
+      "learning_rate": 6.524963867590772e-08,
+      "loss": 0.0595,
+      "step": 17207
+    },
+    {
+      "epoch": 2.7880751782242386,
+      "grad_norm": 0.9312731027603149,
+      "learning_rate": 6.515041800067678e-08,
+      "loss": 0.0627,
+      "step": 17208
+    },
+    {
+      "epoch": 2.7882372002592355,
+      "grad_norm": 0.8041099309921265,
+      "learning_rate": 6.505127182595106e-08,
+      "loss": 0.0594,
+      "step": 17209
+    },
+    {
+      "epoch": 2.788399222294232,
+      "grad_norm": 0.9212023019790649,
+      "learning_rate": 6.495220015476366e-08,
+      "loss": 0.0539,
+      "step": 17210
+    },
+    {
+      "epoch": 2.788561244329229,
+      "grad_norm": 0.9335891008377075,
+      "learning_rate": 6.485320299014608e-08,
+      "loss": 0.0572,
+      "step": 17211
+    },
+    {
+      "epoch": 2.7887232663642254,
+      "grad_norm": 0.9345332384109497,
+      "learning_rate": 6.475428033512754e-08,
+      "loss": 0.0667,
+      "step": 17212
+    },
+    {
+      "epoch": 2.7888852883992223,
+      "grad_norm": 0.938679575920105,
+      "learning_rate": 6.465543219273507e-08,
+      "loss": 0.0602,
+      "step": 17213
+    },
+    {
+      "epoch": 2.789047310434219,
+      "grad_norm": 0.8519788980484009,
+      "learning_rate": 6.455665856599291e-08,
+      "loss": 0.05,
+      "step": 17214
+    },
+    {
+      "epoch": 2.7892093324692158,
+      "grad_norm": 1.0343631505966187,
+      "learning_rate": 6.445795945792338e-08,
+      "loss": 0.0649,
+      "step": 17215
+    },
+    {
+      "epoch": 2.7893713545042127,
+      "grad_norm": 0.935124933719635,
+      "learning_rate": 6.435933487154627e-08,
+      "loss": 0.0668,
+      "step": 17216
+    },
+    {
+      "epoch": 2.789533376539209,
+      "grad_norm": 0.9516078233718872,
+      "learning_rate": 6.426078480987947e-08,
+      "loss": 0.0628,
+      "step": 17217
+    },
+    {
+      "epoch": 2.789695398574206,
+      "grad_norm": 1.1208375692367554,
+      "learning_rate": 6.416230927593803e-08,
+      "loss": 0.0603,
+      "step": 17218
+    },
+    {
+      "epoch": 2.789857420609203,
+      "grad_norm": 0.8511807918548584,
+      "learning_rate": 6.406390827273567e-08,
+      "loss": 0.0597,
+      "step": 17219
+    },
+    {
+      "epoch": 2.7900194426441995,
+      "grad_norm": 0.8629888296127319,
+      "learning_rate": 6.39655818032825e-08,
+      "loss": 0.0586,
+      "step": 17220
+    },
+    {
+      "epoch": 2.7901814646791965,
+      "grad_norm": 1.064953088760376,
+      "learning_rate": 6.386732987058802e-08,
+      "loss": 0.0625,
+      "step": 17221
+    },
+    {
+      "epoch": 2.790343486714193,
+      "grad_norm": 0.9076513648033142,
+      "learning_rate": 6.376915247765735e-08,
+      "loss": 0.0551,
+      "step": 17222
+    },
+    {
+      "epoch": 2.79050550874919,
+      "grad_norm": 0.8662562966346741,
+      "learning_rate": 6.367104962749504e-08,
+      "loss": 0.0561,
+      "step": 17223
+    },
+    {
+      "epoch": 2.7906675307841864,
+      "grad_norm": 0.8259896636009216,
+      "learning_rate": 6.357302132310338e-08,
+      "loss": 0.054,
+      "step": 17224
+    },
+    {
+      "epoch": 2.7908295528191833,
+      "grad_norm": 0.8116698265075684,
+      "learning_rate": 6.347506756748084e-08,
+      "loss": 0.0535,
+      "step": 17225
+    },
+    {
+      "epoch": 2.7909915748541803,
+      "grad_norm": 0.8382045030593872,
+      "learning_rate": 6.337718836362473e-08,
+      "loss": 0.056,
+      "step": 17226
+    },
+    {
+      "epoch": 2.7911535968891767,
+      "grad_norm": 0.8277113437652588,
+      "learning_rate": 6.327938371453069e-08,
+      "loss": 0.0618,
+      "step": 17227
+    },
+    {
+      "epoch": 2.7913156189241737,
+      "grad_norm": 1.0887233018875122,
+      "learning_rate": 6.318165362319023e-08,
+      "loss": 0.0643,
+      "step": 17228
+    },
+    {
+      "epoch": 2.7914776409591706,
+      "grad_norm": 0.9287629723548889,
+      "learning_rate": 6.308399809259457e-08,
+      "loss": 0.0593,
+      "step": 17229
+    },
+    {
+      "epoch": 2.791639662994167,
+      "grad_norm": 0.8305715918540955,
+      "learning_rate": 6.298641712573105e-08,
+      "loss": 0.057,
+      "step": 17230
+    },
+    {
+      "epoch": 2.791801685029164,
+      "grad_norm": 0.8574482202529907,
+      "learning_rate": 6.28889107255859e-08,
+      "loss": 0.0548,
+      "step": 17231
+    },
+    {
+      "epoch": 2.791963707064161,
+      "grad_norm": 0.9703893661499023,
+      "learning_rate": 6.279147889514226e-08,
+      "loss": 0.0562,
+      "step": 17232
+    },
+    {
+      "epoch": 2.7921257290991575,
+      "grad_norm": 0.9155253767967224,
+      "learning_rate": 6.269412163738137e-08,
+      "loss": 0.0623,
+      "step": 17233
+    },
+    {
+      "epoch": 2.7922877511341544,
+      "grad_norm": 0.8451345562934875,
+      "learning_rate": 6.259683895528251e-08,
+      "loss": 0.0535,
+      "step": 17234
+    },
+    {
+      "epoch": 2.792449773169151,
+      "grad_norm": 0.997577965259552,
+      "learning_rate": 6.249963085182192e-08,
+      "loss": 0.0666,
+      "step": 17235
+    },
+    {
+      "epoch": 2.792611795204148,
+      "grad_norm": 1.17750883102417,
+      "learning_rate": 6.24024973299736e-08,
+      "loss": 0.0615,
+      "step": 17236
+    },
+    {
+      "epoch": 2.7927738172391443,
+      "grad_norm": 0.9555923938751221,
+      "learning_rate": 6.230543839271019e-08,
+      "loss": 0.0606,
+      "step": 17237
+    },
+    {
+      "epoch": 2.7929358392741412,
+      "grad_norm": 1.071610689163208,
+      "learning_rate": 6.220845404300124e-08,
+      "loss": 0.0687,
+      "step": 17238
+    },
+    {
+      "epoch": 2.793097861309138,
+      "grad_norm": 0.7882108092308044,
+      "learning_rate": 6.211154428381466e-08,
+      "loss": 0.0511,
+      "step": 17239
+    },
+    {
+      "epoch": 2.7932598833441347,
+      "grad_norm": 0.882218599319458,
+      "learning_rate": 6.201470911811474e-08,
+      "loss": 0.0581,
+      "step": 17240
+    },
+    {
+      "epoch": 2.7934219053791316,
+      "grad_norm": 0.8951191902160645,
+      "learning_rate": 6.191794854886496e-08,
+      "loss": 0.0643,
+      "step": 17241
+    },
+    {
+      "epoch": 2.7935839274141285,
+      "grad_norm": 0.8501166105270386,
+      "learning_rate": 6.182126257902626e-08,
+      "loss": 0.0549,
+      "step": 17242
+    },
+    {
+      "epoch": 2.793745949449125,
+      "grad_norm": 1.0171254873275757,
+      "learning_rate": 6.17246512115563e-08,
+      "loss": 0.0557,
+      "step": 17243
+    },
+    {
+      "epoch": 2.793907971484122,
+      "grad_norm": 0.9592586159706116,
+      "learning_rate": 6.162811444941159e-08,
+      "loss": 0.0647,
+      "step": 17244
+    },
+    {
+      "epoch": 2.7940699935191184,
+      "grad_norm": 0.7187346816062927,
+      "learning_rate": 6.153165229554587e-08,
+      "loss": 0.043,
+      "step": 17245
+    },
+    {
+      "epoch": 2.7942320155541154,
+      "grad_norm": 0.9293650984764099,
+      "learning_rate": 6.143526475291067e-08,
+      "loss": 0.0611,
+      "step": 17246
+    },
+    {
+      "epoch": 2.794394037589112,
+      "grad_norm": 0.919105052947998,
+      "learning_rate": 6.133895182445504e-08,
+      "loss": 0.0585,
+      "step": 17247
+    },
+    {
+      "epoch": 2.794556059624109,
+      "grad_norm": 0.9237651228904724,
+      "learning_rate": 6.124271351312605e-08,
+      "loss": 0.0615,
+      "step": 17248
+    },
+    {
+      "epoch": 2.7947180816591057,
+      "grad_norm": 0.9139578938484192,
+      "learning_rate": 6.114654982186829e-08,
+      "loss": 0.0635,
+      "step": 17249
+    },
+    {
+      "epoch": 2.7948801036941022,
+      "grad_norm": 1.0389397144317627,
+      "learning_rate": 6.105046075362441e-08,
+      "loss": 0.0565,
+      "step": 17250
+    },
+    {
+      "epoch": 2.795042125729099,
+      "grad_norm": 0.9643591046333313,
+      "learning_rate": 6.095444631133401e-08,
+      "loss": 0.0637,
+      "step": 17251
+    },
+    {
+      "epoch": 2.795204147764096,
+      "grad_norm": 0.9869605898857117,
+      "learning_rate": 6.085850649793529e-08,
+      "loss": 0.0612,
+      "step": 17252
+    },
+    {
+      "epoch": 2.7953661697990926,
+      "grad_norm": 1.068179965019226,
+      "learning_rate": 6.076264131636394e-08,
+      "loss": 0.0603,
+      "step": 17253
+    },
+    {
+      "epoch": 2.7955281918340895,
+      "grad_norm": 0.8888418674468994,
+      "learning_rate": 6.066685076955264e-08,
+      "loss": 0.0567,
+      "step": 17254
+    },
+    {
+      "epoch": 2.7956902138690864,
+      "grad_norm": 0.759746253490448,
+      "learning_rate": 6.057113486043236e-08,
+      "loss": 0.0525,
+      "step": 17255
+    },
+    {
+      "epoch": 2.795852235904083,
+      "grad_norm": 1.0633058547973633,
+      "learning_rate": 6.047549359193245e-08,
+      "loss": 0.0631,
+      "step": 17256
+    },
+    {
+      "epoch": 2.79601425793908,
+      "grad_norm": 0.906599223613739,
+      "learning_rate": 6.037992696697914e-08,
+      "loss": 0.0616,
+      "step": 17257
+    },
+    {
+      "epoch": 2.7961762799740764,
+      "grad_norm": 0.8136864900588989,
+      "learning_rate": 6.028443498849596e-08,
+      "loss": 0.0528,
+      "step": 17258
+    },
+    {
+      "epoch": 2.7963383020090733,
+      "grad_norm": 0.992445170879364,
+      "learning_rate": 6.018901765940499e-08,
+      "loss": 0.0618,
+      "step": 17259
+    },
+    {
+      "epoch": 2.79650032404407,
+      "grad_norm": 0.8325508236885071,
+      "learning_rate": 6.009367498262587e-08,
+      "loss": 0.05,
+      "step": 17260
+    },
+    {
+      "epoch": 2.7966623460790667,
+      "grad_norm": 0.7613760232925415,
+      "learning_rate": 5.999840696107595e-08,
+      "loss": 0.0488,
+      "step": 17261
+    },
+    {
+      "epoch": 2.7968243681140637,
+      "grad_norm": 1.0377253293991089,
+      "learning_rate": 5.990321359767015e-08,
+      "loss": 0.068,
+      "step": 17262
+    },
+    {
+      "epoch": 2.79698639014906,
+      "grad_norm": 1.0658155679702759,
+      "learning_rate": 5.980809489532085e-08,
+      "loss": 0.0616,
+      "step": 17263
+    },
+    {
+      "epoch": 2.797148412184057,
+      "grad_norm": 0.9770277142524719,
+      "learning_rate": 5.97130508569388e-08,
+      "loss": 0.0582,
+      "step": 17264
+    },
+    {
+      "epoch": 2.797310434219054,
+      "grad_norm": 0.9497054219245911,
+      "learning_rate": 5.961808148543219e-08,
+      "loss": 0.0561,
+      "step": 17265
+    },
+    {
+      "epoch": 2.7974724562540505,
+      "grad_norm": 0.9469033479690552,
+      "learning_rate": 5.95231867837065e-08,
+      "loss": 0.0573,
+      "step": 17266
+    },
+    {
+      "epoch": 2.7976344782890474,
+      "grad_norm": 0.9465834498405457,
+      "learning_rate": 5.9428366754665234e-08,
+      "loss": 0.061,
+      "step": 17267
+    },
+    {
+      "epoch": 2.797796500324044,
+      "grad_norm": 0.92009437084198,
+      "learning_rate": 5.933362140121052e-08,
+      "loss": 0.0604,
+      "step": 17268
+    },
+    {
+      "epoch": 2.797958522359041,
+      "grad_norm": 0.8519408702850342,
+      "learning_rate": 5.923895072624031e-08,
+      "loss": 0.0618,
+      "step": 17269
+    },
+    {
+      "epoch": 2.7981205443940373,
+      "grad_norm": 1.1452556848526,
+      "learning_rate": 5.9144354732651455e-08,
+      "loss": 0.0634,
+      "step": 17270
+    },
+    {
+      "epoch": 2.7982825664290343,
+      "grad_norm": 0.9965704679489136,
+      "learning_rate": 5.904983342333887e-08,
+      "loss": 0.0651,
+      "step": 17271
+    },
+    {
+      "epoch": 2.798444588464031,
+      "grad_norm": 0.9546822905540466,
+      "learning_rate": 5.8955386801194394e-08,
+      "loss": 0.0615,
+      "step": 17272
+    },
+    {
+      "epoch": 2.7986066104990277,
+      "grad_norm": 0.8720921277999878,
+      "learning_rate": 5.886101486910767e-08,
+      "loss": 0.058,
+      "step": 17273
+    },
+    {
+      "epoch": 2.7987686325340246,
+      "grad_norm": 0.9746670722961426,
+      "learning_rate": 5.8766717629966387e-08,
+      "loss": 0.0621,
+      "step": 17274
+    },
+    {
+      "epoch": 2.7989306545690216,
+      "grad_norm": 0.9597407579421997,
+      "learning_rate": 5.867249508665629e-08,
+      "loss": 0.062,
+      "step": 17275
+    },
+    {
+      "epoch": 2.799092676604018,
+      "grad_norm": 1.0109037160873413,
+      "learning_rate": 5.857834724205979e-08,
+      "loss": 0.0593,
+      "step": 17276
+    },
+    {
+      "epoch": 2.799254698639015,
+      "grad_norm": 0.8651391863822937,
+      "learning_rate": 5.848427409905766e-08,
+      "loss": 0.0586,
+      "step": 17277
+    },
+    {
+      "epoch": 2.799416720674012,
+      "grad_norm": 0.9493432641029358,
+      "learning_rate": 5.839027566052841e-08,
+      "loss": 0.0629,
+      "step": 17278
+    },
+    {
+      "epoch": 2.7995787427090084,
+      "grad_norm": 0.9215166568756104,
+      "learning_rate": 5.829635192934807e-08,
+      "loss": 0.0638,
+      "step": 17279
+    },
+    {
+      "epoch": 2.7997407647440054,
+      "grad_norm": 0.8210483193397522,
+      "learning_rate": 5.820250290839047e-08,
+      "loss": 0.0543,
+      "step": 17280
+    },
+    {
+      "epoch": 2.799902786779002,
+      "grad_norm": 0.8877785801887512,
+      "learning_rate": 5.810872860052747e-08,
+      "loss": 0.0562,
+      "step": 17281
+    },
+    {
+      "epoch": 2.8000648088139988,
+      "grad_norm": 0.9447657465934753,
+      "learning_rate": 5.801502900862788e-08,
+      "loss": 0.0624,
+      "step": 17282
+    },
+    {
+      "epoch": 2.8002268308489953,
+      "grad_norm": 1.0124446153640747,
+      "learning_rate": 5.7921404135559414e-08,
+      "loss": 0.0577,
+      "step": 17283
+    },
+    {
+      "epoch": 2.800388852883992,
+      "grad_norm": 1.1428158283233643,
+      "learning_rate": 5.782785398418561e-08,
+      "loss": 0.0618,
+      "step": 17284
+    },
+    {
+      "epoch": 2.800550874918989,
+      "grad_norm": 1.0584673881530762,
+      "learning_rate": 5.773437855736974e-08,
+      "loss": 0.0651,
+      "step": 17285
+    },
+    {
+      "epoch": 2.8007128969539856,
+      "grad_norm": 0.9388712048530579,
+      "learning_rate": 5.7640977857972016e-08,
+      "loss": 0.0651,
+      "step": 17286
+    },
+    {
+      "epoch": 2.8008749189889826,
+      "grad_norm": 0.9077460765838623,
+      "learning_rate": 5.7547651888849864e-08,
+      "loss": 0.0635,
+      "step": 17287
+    },
+    {
+      "epoch": 2.8010369410239795,
+      "grad_norm": 0.8643128275871277,
+      "learning_rate": 5.745440065285879e-08,
+      "loss": 0.0568,
+      "step": 17288
+    },
+    {
+      "epoch": 2.801198963058976,
+      "grad_norm": 0.9081558585166931,
+      "learning_rate": 5.736122415285206e-08,
+      "loss": 0.0654,
+      "step": 17289
+    },
+    {
+      "epoch": 2.801360985093973,
+      "grad_norm": 0.9258993864059448,
+      "learning_rate": 5.726812239168128e-08,
+      "loss": 0.0616,
+      "step": 17290
+    },
+    {
+      "epoch": 2.8015230071289694,
+      "grad_norm": 0.9604719281196594,
+      "learning_rate": 5.717509537219418e-08,
+      "loss": 0.0643,
+      "step": 17291
+    },
+    {
+      "epoch": 2.8016850291639663,
+      "grad_norm": 0.8643040060997009,
+      "learning_rate": 5.708214309723792e-08,
+      "loss": 0.0642,
+      "step": 17292
+    },
+    {
+      "epoch": 2.801847051198963,
+      "grad_norm": 0.8756396770477295,
+      "learning_rate": 5.6989265569656335e-08,
+      "loss": 0.0583,
+      "step": 17293
+    },
+    {
+      "epoch": 2.8020090732339598,
+      "grad_norm": 0.7664240598678589,
+      "learning_rate": 5.689646279229105e-08,
+      "loss": 0.0532,
+      "step": 17294
+    },
+    {
+      "epoch": 2.8021710952689567,
+      "grad_norm": 0.7147080898284912,
+      "learning_rate": 5.680373476798201e-08,
+      "loss": 0.0479,
+      "step": 17295
+    },
+    {
+      "epoch": 2.802333117303953,
+      "grad_norm": 0.7867782711982727,
+      "learning_rate": 5.671108149956611e-08,
+      "loss": 0.0516,
+      "step": 17296
+    },
+    {
+      "epoch": 2.80249513933895,
+      "grad_norm": 0.9071274399757385,
+      "learning_rate": 5.661850298987859e-08,
+      "loss": 0.0572,
+      "step": 17297
+    },
+    {
+      "epoch": 2.802657161373947,
+      "grad_norm": 0.8463209867477417,
+      "learning_rate": 5.6525999241751894e-08,
+      "loss": 0.056,
+      "step": 17298
+    },
+    {
+      "epoch": 2.8028191834089435,
+      "grad_norm": 1.1977580785751343,
+      "learning_rate": 5.643357025801655e-08,
+      "loss": 0.0617,
+      "step": 17299
+    },
+    {
+      "epoch": 2.8029812054439405,
+      "grad_norm": 0.9107047319412231,
+      "learning_rate": 5.6341216041500555e-08,
+      "loss": 0.061,
+      "step": 17300
+    },
+    {
+      "epoch": 2.8031432274789374,
+      "grad_norm": 0.907247006893158,
+      "learning_rate": 5.624893659503028e-08,
+      "loss": 0.0526,
+      "step": 17301
+    },
+    {
+      "epoch": 2.803305249513934,
+      "grad_norm": 0.8431805968284607,
+      "learning_rate": 5.6156731921428455e-08,
+      "loss": 0.0576,
+      "step": 17302
+    },
+    {
+      "epoch": 2.8034672715489304,
+      "grad_norm": 1.0197104215621948,
+      "learning_rate": 5.6064602023516154e-08,
+      "loss": 0.0609,
+      "step": 17303
+    },
+    {
+      "epoch": 2.8036292935839273,
+      "grad_norm": 0.908807098865509,
+      "learning_rate": 5.597254690411363e-08,
+      "loss": 0.061,
+      "step": 17304
+    },
+    {
+      "epoch": 2.8037913156189243,
+      "grad_norm": 0.8455592393875122,
+      "learning_rate": 5.588056656603641e-08,
+      "loss": 0.0588,
+      "step": 17305
+    },
+    {
+      "epoch": 2.8039533376539207,
+      "grad_norm": 0.9535689949989319,
+      "learning_rate": 5.5788661012099176e-08,
+      "loss": 0.0598,
+      "step": 17306
+    },
+    {
+      "epoch": 2.8041153596889177,
+      "grad_norm": 0.8808077573776245,
+      "learning_rate": 5.5696830245114134e-08,
+      "loss": 0.0658,
+      "step": 17307
+    },
+    {
+      "epoch": 2.8042773817239146,
+      "grad_norm": 0.8899673819541931,
+      "learning_rate": 5.560507426789069e-08,
+      "loss": 0.0557,
+      "step": 17308
+    },
+    {
+      "epoch": 2.804439403758911,
+      "grad_norm": 0.8185904622077942,
+      "learning_rate": 5.551339308323689e-08,
+      "loss": 0.0569,
+      "step": 17309
+    },
+    {
+      "epoch": 2.804601425793908,
+      "grad_norm": 0.8157827258110046,
+      "learning_rate": 5.5421786693957705e-08,
+      "loss": 0.0508,
+      "step": 17310
+    },
+    {
+      "epoch": 2.804763447828905,
+      "grad_norm": 0.9755039811134338,
+      "learning_rate": 5.533025510285617e-08,
+      "loss": 0.0592,
+      "step": 17311
+    },
+    {
+      "epoch": 2.8049254698639015,
+      "grad_norm": 0.9864017367362976,
+      "learning_rate": 5.523879831273282e-08,
+      "loss": 0.0625,
+      "step": 17312
+    },
+    {
+      "epoch": 2.8050874918988984,
+      "grad_norm": 0.8920227289199829,
+      "learning_rate": 5.514741632638571e-08,
+      "loss": 0.0634,
+      "step": 17313
+    },
+    {
+      "epoch": 2.805249513933895,
+      "grad_norm": 0.9825411438941956,
+      "learning_rate": 5.505610914661147e-08,
+      "loss": 0.0567,
+      "step": 17314
+    },
+    {
+      "epoch": 2.805411535968892,
+      "grad_norm": 1.1181998252868652,
+      "learning_rate": 5.496487677620399e-08,
+      "loss": 0.0629,
+      "step": 17315
+    },
+    {
+      "epoch": 2.8055735580038883,
+      "grad_norm": 1.0909833908081055,
+      "learning_rate": 5.487371921795381e-08,
+      "loss": 0.0687,
+      "step": 17316
+    },
+    {
+      "epoch": 2.8057355800388852,
+      "grad_norm": 0.9423796534538269,
+      "learning_rate": 5.478263647465093e-08,
+      "loss": 0.0582,
+      "step": 17317
+    },
+    {
+      "epoch": 2.805897602073882,
+      "grad_norm": 0.8542878031730652,
+      "learning_rate": 5.4691628549082e-08,
+      "loss": 0.0655,
+      "step": 17318
+    },
+    {
+      "epoch": 2.8060596241088787,
+      "grad_norm": 0.9813777208328247,
+      "learning_rate": 5.4600695444032014e-08,
+      "loss": 0.0615,
+      "step": 17319
+    },
+    {
+      "epoch": 2.8062216461438756,
+      "grad_norm": 0.9087584018707275,
+      "learning_rate": 5.450983716228292e-08,
+      "loss": 0.0576,
+      "step": 17320
+    },
+    {
+      "epoch": 2.8063836681788725,
+      "grad_norm": 0.9132459759712219,
+      "learning_rate": 5.441905370661471e-08,
+      "loss": 0.0615,
+      "step": 17321
+    },
+    {
+      "epoch": 2.806545690213869,
+      "grad_norm": 0.7650594115257263,
+      "learning_rate": 5.4328345079805164e-08,
+      "loss": 0.0522,
+      "step": 17322
+    },
+    {
+      "epoch": 2.806707712248866,
+      "grad_norm": 0.9708273410797119,
+      "learning_rate": 5.423771128462985e-08,
+      "loss": 0.0676,
+      "step": 17323
+    },
+    {
+      "epoch": 2.806869734283863,
+      "grad_norm": 0.949398934841156,
+      "learning_rate": 5.4147152323862085e-08,
+      "loss": 0.0591,
+      "step": 17324
+    },
+    {
+      "epoch": 2.8070317563188594,
+      "grad_norm": 0.8356457352638245,
+      "learning_rate": 5.405666820027272e-08,
+      "loss": 0.0577,
+      "step": 17325
+    },
+    {
+      "epoch": 2.807193778353856,
+      "grad_norm": 0.9790476560592651,
+      "learning_rate": 5.3966258916629824e-08,
+      "loss": 0.0585,
+      "step": 17326
+    },
+    {
+      "epoch": 2.807355800388853,
+      "grad_norm": 0.979407548904419,
+      "learning_rate": 5.387592447570061e-08,
+      "loss": 0.0618,
+      "step": 17327
+    },
+    {
+      "epoch": 2.8075178224238497,
+      "grad_norm": 0.8905614614486694,
+      "learning_rate": 5.378566488024817e-08,
+      "loss": 0.0569,
+      "step": 17328
+    },
+    {
+      "epoch": 2.807679844458846,
+      "grad_norm": 0.8756917715072632,
+      "learning_rate": 5.3695480133034994e-08,
+      "loss": 0.0593,
+      "step": 17329
+    },
+    {
+      "epoch": 2.807841866493843,
+      "grad_norm": 0.8476364016532898,
+      "learning_rate": 5.3605370236820276e-08,
+      "loss": 0.0571,
+      "step": 17330
+    },
+    {
+      "epoch": 2.80800388852884,
+      "grad_norm": 1.1375116109848022,
+      "learning_rate": 5.3515335194360694e-08,
+      "loss": 0.0673,
+      "step": 17331
+    },
+    {
+      "epoch": 2.8081659105638366,
+      "grad_norm": 0.9701889157295227,
+      "learning_rate": 5.3425375008411276e-08,
+      "loss": 0.0694,
+      "step": 17332
+    },
+    {
+      "epoch": 2.8083279325988335,
+      "grad_norm": 0.8260276913642883,
+      "learning_rate": 5.333548968172536e-08,
+      "loss": 0.0599,
+      "step": 17333
+    },
+    {
+      "epoch": 2.8084899546338304,
+      "grad_norm": 0.8484598994255066,
+      "learning_rate": 5.3245679217052424e-08,
+      "loss": 0.0548,
+      "step": 17334
+    },
+    {
+      "epoch": 2.808651976668827,
+      "grad_norm": 0.9745419025421143,
+      "learning_rate": 5.315594361714083e-08,
+      "loss": 0.062,
+      "step": 17335
+    },
+    {
+      "epoch": 2.808813998703824,
+      "grad_norm": 0.868465006351471,
+      "learning_rate": 5.3066282884735863e-08,
+      "loss": 0.0614,
+      "step": 17336
+    },
+    {
+      "epoch": 2.8089760207388204,
+      "grad_norm": 0.9748360514640808,
+      "learning_rate": 5.297669702258118e-08,
+      "loss": 0.0632,
+      "step": 17337
+    },
+    {
+      "epoch": 2.8091380427738173,
+      "grad_norm": 1.0200855731964111,
+      "learning_rate": 5.2887186033417914e-08,
+      "loss": 0.0625,
+      "step": 17338
+    },
+    {
+      "epoch": 2.809300064808814,
+      "grad_norm": 1.0314770936965942,
+      "learning_rate": 5.279774991998499e-08,
+      "loss": 0.0611,
+      "step": 17339
+    },
+    {
+      "epoch": 2.8094620868438107,
+      "grad_norm": 1.0516095161437988,
+      "learning_rate": 5.270838868501854e-08,
+      "loss": 0.0644,
+      "step": 17340
+    },
+    {
+      "epoch": 2.8096241088788076,
+      "grad_norm": 0.8740168809890747,
+      "learning_rate": 5.261910233125333e-08,
+      "loss": 0.0567,
+      "step": 17341
+    },
+    {
+      "epoch": 2.809786130913804,
+      "grad_norm": 0.9460188746452332,
+      "learning_rate": 5.252989086142107e-08,
+      "loss": 0.0538,
+      "step": 17342
+    },
+    {
+      "epoch": 2.809948152948801,
+      "grad_norm": 0.8696618676185608,
+      "learning_rate": 5.244075427825124e-08,
+      "loss": 0.0557,
+      "step": 17343
+    },
+    {
+      "epoch": 2.810110174983798,
+      "grad_norm": 1.0212994813919067,
+      "learning_rate": 5.235169258447137e-08,
+      "loss": 0.0699,
+      "step": 17344
+    },
+    {
+      "epoch": 2.8102721970187945,
+      "grad_norm": 1.036773443222046,
+      "learning_rate": 5.2262705782806513e-08,
+      "loss": 0.0664,
+      "step": 17345
+    },
+    {
+      "epoch": 2.8104342190537914,
+      "grad_norm": 0.875264585018158,
+      "learning_rate": 5.2173793875979204e-08,
+      "loss": 0.0647,
+      "step": 17346
+    },
+    {
+      "epoch": 2.810596241088788,
+      "grad_norm": 1.0210704803466797,
+      "learning_rate": 5.208495686671061e-08,
+      "loss": 0.0674,
+      "step": 17347
+    },
+    {
+      "epoch": 2.810758263123785,
+      "grad_norm": 1.0680155754089355,
+      "learning_rate": 5.199619475771856e-08,
+      "loss": 0.0707,
+      "step": 17348
+    },
+    {
+      "epoch": 2.8109202851587813,
+      "grad_norm": 0.8734906315803528,
+      "learning_rate": 5.190750755171864e-08,
+      "loss": 0.0533,
+      "step": 17349
+    },
+    {
+      "epoch": 2.8110823071937783,
+      "grad_norm": 0.8121834397315979,
+      "learning_rate": 5.181889525142453e-08,
+      "loss": 0.0516,
+      "step": 17350
+    },
+    {
+      "epoch": 2.811244329228775,
+      "grad_norm": 1.0527007579803467,
+      "learning_rate": 5.1730357859547666e-08,
+      "loss": 0.0645,
+      "step": 17351
+    },
+    {
+      "epoch": 2.8114063512637717,
+      "grad_norm": 0.9521415829658508,
+      "learning_rate": 5.164189537879782e-08,
+      "loss": 0.0601,
+      "step": 17352
+    },
+    {
+      "epoch": 2.8115683732987686,
+      "grad_norm": 0.9354329705238342,
+      "learning_rate": 5.155350781188062e-08,
+      "loss": 0.0609,
+      "step": 17353
+    },
+    {
+      "epoch": 2.8117303953337656,
+      "grad_norm": 0.8666455149650574,
+      "learning_rate": 5.146519516150084e-08,
+      "loss": 0.0575,
+      "step": 17354
+    },
+    {
+      "epoch": 2.811892417368762,
+      "grad_norm": 0.874804675579071,
+      "learning_rate": 5.137695743036103e-08,
+      "loss": 0.0606,
+      "step": 17355
+    },
+    {
+      "epoch": 2.812054439403759,
+      "grad_norm": 0.96415114402771,
+      "learning_rate": 5.128879462116071e-08,
+      "loss": 0.0546,
+      "step": 17356
+    },
+    {
+      "epoch": 2.812216461438756,
+      "grad_norm": 1.104056477546692,
+      "learning_rate": 5.1200706736597435e-08,
+      "loss": 0.0641,
+      "step": 17357
+    },
+    {
+      "epoch": 2.8123784834737524,
+      "grad_norm": 0.9852107763290405,
+      "learning_rate": 5.111269377936656e-08,
+      "loss": 0.0684,
+      "step": 17358
+    },
+    {
+      "epoch": 2.8125405055087493,
+      "grad_norm": 0.9444692730903625,
+      "learning_rate": 5.10247557521612e-08,
+      "loss": 0.0621,
+      "step": 17359
+    },
+    {
+      "epoch": 2.812702527543746,
+      "grad_norm": 0.8795128464698792,
+      "learning_rate": 5.093689265767143e-08,
+      "loss": 0.0592,
+      "step": 17360
+    },
+    {
+      "epoch": 2.8128645495787428,
+      "grad_norm": 0.9272993206977844,
+      "learning_rate": 5.084910449858649e-08,
+      "loss": 0.0582,
+      "step": 17361
+    },
+    {
+      "epoch": 2.8130265716137393,
+      "grad_norm": 1.0397619009017944,
+      "learning_rate": 5.0761391277591996e-08,
+      "loss": 0.0598,
+      "step": 17362
+    },
+    {
+      "epoch": 2.813188593648736,
+      "grad_norm": 0.9312973022460938,
+      "learning_rate": 5.0673752997372204e-08,
+      "loss": 0.0581,
+      "step": 17363
+    },
+    {
+      "epoch": 2.813350615683733,
+      "grad_norm": 0.9002386331558228,
+      "learning_rate": 5.05861896606083e-08,
+      "loss": 0.0614,
+      "step": 17364
+    },
+    {
+      "epoch": 2.8135126377187296,
+      "grad_norm": 0.9482325911521912,
+      "learning_rate": 5.049870126997897e-08,
+      "loss": 0.0642,
+      "step": 17365
+    },
+    {
+      "epoch": 2.8136746597537265,
+      "grad_norm": 0.8579296469688416,
+      "learning_rate": 5.0411287828162346e-08,
+      "loss": 0.0602,
+      "step": 17366
+    },
+    {
+      "epoch": 2.8138366817887235,
+      "grad_norm": 0.9267897605895996,
+      "learning_rate": 5.032394933783213e-08,
+      "loss": 0.0552,
+      "step": 17367
+    },
+    {
+      "epoch": 2.81399870382372,
+      "grad_norm": 0.8635824918746948,
+      "learning_rate": 5.023668580166091e-08,
+      "loss": 0.057,
+      "step": 17368
+    },
+    {
+      "epoch": 2.814160725858717,
+      "grad_norm": 1.0104823112487793,
+      "learning_rate": 5.014949722231876e-08,
+      "loss": 0.0652,
+      "step": 17369
+    },
+    {
+      "epoch": 2.8143227478937134,
+      "grad_norm": 0.9497948884963989,
+      "learning_rate": 5.0062383602473566e-08,
+      "loss": 0.0546,
+      "step": 17370
+    },
+    {
+      "epoch": 2.8144847699287103,
+      "grad_norm": 0.9637094140052795,
+      "learning_rate": 4.9975344944790674e-08,
+      "loss": 0.0576,
+      "step": 17371
+    },
+    {
+      "epoch": 2.814646791963707,
+      "grad_norm": 0.904549241065979,
+      "learning_rate": 4.9888381251933237e-08,
+      "loss": 0.0664,
+      "step": 17372
+    },
+    {
+      "epoch": 2.8148088139987038,
+      "grad_norm": 0.9141635894775391,
+      "learning_rate": 4.980149252656219e-08,
+      "loss": 0.0575,
+      "step": 17373
+    },
+    {
+      "epoch": 2.8149708360337007,
+      "grad_norm": 0.8708399534225464,
+      "learning_rate": 4.971467877133651e-08,
+      "loss": 0.0565,
+      "step": 17374
+    },
+    {
+      "epoch": 2.815132858068697,
+      "grad_norm": 0.9253936409950256,
+      "learning_rate": 4.962793998891158e-08,
+      "loss": 0.0568,
+      "step": 17375
+    },
+    {
+      "epoch": 2.815294880103694,
+      "grad_norm": 1.0665290355682373,
+      "learning_rate": 4.954127618194193e-08,
+      "loss": 0.0641,
+      "step": 17376
+    },
+    {
+      "epoch": 2.815456902138691,
+      "grad_norm": 0.9630182385444641,
+      "learning_rate": 4.945468735307934e-08,
+      "loss": 0.0655,
+      "step": 17377
+    },
+    {
+      "epoch": 2.8156189241736875,
+      "grad_norm": 0.8089599609375,
+      "learning_rate": 4.936817350497336e-08,
+      "loss": 0.057,
+      "step": 17378
+    },
+    {
+      "epoch": 2.8157809462086845,
+      "grad_norm": 0.85359787940979,
+      "learning_rate": 4.9281734640270476e-08,
+      "loss": 0.0535,
+      "step": 17379
+    },
+    {
+      "epoch": 2.8159429682436814,
+      "grad_norm": 0.9590152502059937,
+      "learning_rate": 4.919537076161579e-08,
+      "loss": 0.0589,
+      "step": 17380
+    },
+    {
+      "epoch": 2.816104990278678,
+      "grad_norm": 0.8613442182540894,
+      "learning_rate": 4.910908187165248e-08,
+      "loss": 0.0625,
+      "step": 17381
+    },
+    {
+      "epoch": 2.816267012313675,
+      "grad_norm": 0.9294003248214722,
+      "learning_rate": 4.90228679730198e-08,
+      "loss": 0.0623,
+      "step": 17382
+    },
+    {
+      "epoch": 2.8164290343486713,
+      "grad_norm": 0.906424343585968,
+      "learning_rate": 4.893672906835623e-08,
+      "loss": 0.058,
+      "step": 17383
+    },
+    {
+      "epoch": 2.8165910563836682,
+      "grad_norm": 0.8201852440834045,
+      "learning_rate": 4.8850665160297406e-08,
+      "loss": 0.0538,
+      "step": 17384
+    },
+    {
+      "epoch": 2.8167530784186647,
+      "grad_norm": 0.8930208086967468,
+      "learning_rate": 4.8764676251476237e-08,
+      "loss": 0.0556,
+      "step": 17385
+    },
+    {
+      "epoch": 2.8169151004536617,
+      "grad_norm": 0.8145630955696106,
+      "learning_rate": 4.867876234452423e-08,
+      "loss": 0.0543,
+      "step": 17386
+    },
+    {
+      "epoch": 2.8170771224886586,
+      "grad_norm": 1.011269211769104,
+      "learning_rate": 4.859292344207012e-08,
+      "loss": 0.0641,
+      "step": 17387
+    },
+    {
+      "epoch": 2.817239144523655,
+      "grad_norm": 0.9331755638122559,
+      "learning_rate": 4.85071595467404e-08,
+      "loss": 0.0623,
+      "step": 17388
+    },
+    {
+      "epoch": 2.817401166558652,
+      "grad_norm": 1.006993055343628,
+      "learning_rate": 4.842147066115882e-08,
+      "loss": 0.0664,
+      "step": 17389
+    },
+    {
+      "epoch": 2.817563188593649,
+      "grad_norm": 0.8741045594215393,
+      "learning_rate": 4.8335856787947447e-08,
+      "loss": 0.0569,
+      "step": 17390
+    },
+    {
+      "epoch": 2.8177252106286454,
+      "grad_norm": 0.9738301634788513,
+      "learning_rate": 4.825031792972612e-08,
+      "loss": 0.0626,
+      "step": 17391
+    },
+    {
+      "epoch": 2.8178872326636424,
+      "grad_norm": 0.9677324295043945,
+      "learning_rate": 4.81648540891122e-08,
+      "loss": 0.0639,
+      "step": 17392
+    },
+    {
+      "epoch": 2.818049254698639,
+      "grad_norm": 0.8875753283500671,
+      "learning_rate": 4.807946526872026e-08,
+      "loss": 0.0516,
+      "step": 17393
+    },
+    {
+      "epoch": 2.818211276733636,
+      "grad_norm": 0.8754310607910156,
+      "learning_rate": 4.799415147116265e-08,
+      "loss": 0.0591,
+      "step": 17394
+    },
+    {
+      "epoch": 2.8183732987686323,
+      "grad_norm": 0.9936445355415344,
+      "learning_rate": 4.7908912699050906e-08,
+      "loss": 0.0666,
+      "step": 17395
+    },
+    {
+      "epoch": 2.8185353208036292,
+      "grad_norm": 0.9375000596046448,
+      "learning_rate": 4.782374895499236e-08,
+      "loss": 0.0646,
+      "step": 17396
+    },
+    {
+      "epoch": 2.818697342838626,
+      "grad_norm": 0.9706060886383057,
+      "learning_rate": 4.773866024159274e-08,
+      "loss": 0.0558,
+      "step": 17397
+    },
+    {
+      "epoch": 2.8188593648736227,
+      "grad_norm": 0.8691913485527039,
+      "learning_rate": 4.7653646561455767e-08,
+      "loss": 0.063,
+      "step": 17398
+    },
+    {
+      "epoch": 2.8190213869086196,
+      "grad_norm": 1.1515320539474487,
+      "learning_rate": 4.756870791718271e-08,
+      "loss": 0.0665,
+      "step": 17399
+    },
+    {
+      "epoch": 2.8191834089436165,
+      "grad_norm": 1.0879017114639282,
+      "learning_rate": 4.7483844311372594e-08,
+      "loss": 0.0659,
+      "step": 17400
+    },
+    {
+      "epoch": 2.819345430978613,
+      "grad_norm": 0.8411441445350647,
+      "learning_rate": 4.739905574662168e-08,
+      "loss": 0.0525,
+      "step": 17401
+    },
+    {
+      "epoch": 2.81950745301361,
+      "grad_norm": 0.8254294395446777,
+      "learning_rate": 4.731434222552456e-08,
+      "loss": 0.0584,
+      "step": 17402
+    },
+    {
+      "epoch": 2.819669475048607,
+      "grad_norm": 0.8175220489501953,
+      "learning_rate": 4.722970375067304e-08,
+      "loss": 0.0581,
+      "step": 17403
+    },
+    {
+      "epoch": 2.8198314970836034,
+      "grad_norm": 0.9433904886245728,
+      "learning_rate": 4.7145140324657e-08,
+      "loss": 0.0595,
+      "step": 17404
+    },
+    {
+      "epoch": 2.8199935191186,
+      "grad_norm": 0.9785334467887878,
+      "learning_rate": 4.7060651950064094e-08,
+      "loss": 0.0633,
+      "step": 17405
+    },
+    {
+      "epoch": 2.820155541153597,
+      "grad_norm": 0.7298676371574402,
+      "learning_rate": 4.697623862947892e-08,
+      "loss": 0.0528,
+      "step": 17406
+    },
+    {
+      "epoch": 2.8203175631885937,
+      "grad_norm": 0.8374927639961243,
+      "learning_rate": 4.689190036548524e-08,
+      "loss": 0.0558,
+      "step": 17407
+    },
+    {
+      "epoch": 2.82047958522359,
+      "grad_norm": 0.8497073650360107,
+      "learning_rate": 4.680763716066239e-08,
+      "loss": 0.0592,
+      "step": 17408
+    },
+    {
+      "epoch": 2.820641607258587,
+      "grad_norm": 1.0084335803985596,
+      "learning_rate": 4.672344901758941e-08,
+      "loss": 0.0642,
+      "step": 17409
+    },
+    {
+      "epoch": 2.820803629293584,
+      "grad_norm": 0.8744716644287109,
+      "learning_rate": 4.663933593884229e-08,
+      "loss": 0.0627,
+      "step": 17410
+    },
+    {
+      "epoch": 2.8209656513285806,
+      "grad_norm": 0.8989715576171875,
+      "learning_rate": 4.655529792699426e-08,
+      "loss": 0.0603,
+      "step": 17411
+    },
+    {
+      "epoch": 2.8211276733635775,
+      "grad_norm": 1.0203579664230347,
+      "learning_rate": 4.6471334984616866e-08,
+      "loss": 0.0635,
+      "step": 17412
+    },
+    {
+      "epoch": 2.8212896953985744,
+      "grad_norm": 0.8238543272018433,
+      "learning_rate": 4.6387447114278897e-08,
+      "loss": 0.0526,
+      "step": 17413
+    },
+    {
+      "epoch": 2.821451717433571,
+      "grad_norm": 0.872983455657959,
+      "learning_rate": 4.6303634318548006e-08,
+      "loss": 0.0611,
+      "step": 17414
+    },
+    {
+      "epoch": 2.821613739468568,
+      "grad_norm": 0.9443795084953308,
+      "learning_rate": 4.6219896599987714e-08,
+      "loss": 0.0602,
+      "step": 17415
+    },
+    {
+      "epoch": 2.8217757615035644,
+      "grad_norm": 0.8985307812690735,
+      "learning_rate": 4.613623396116068e-08,
+      "loss": 0.0577,
+      "step": 17416
+    },
+    {
+      "epoch": 2.8219377835385613,
+      "grad_norm": 0.9005011916160583,
+      "learning_rate": 4.6052646404626814e-08,
+      "loss": 0.0627,
+      "step": 17417
+    },
+    {
+      "epoch": 2.8220998055735578,
+      "grad_norm": 1.0429044961929321,
+      "learning_rate": 4.596913393294322e-08,
+      "loss": 0.0734,
+      "step": 17418
+    },
+    {
+      "epoch": 2.8222618276085547,
+      "grad_norm": 0.9500260949134827,
+      "learning_rate": 4.5885696548665645e-08,
+      "loss": 0.0641,
+      "step": 17419
+    },
+    {
+      "epoch": 2.8224238496435516,
+      "grad_norm": 0.828524112701416,
+      "learning_rate": 4.580233425434677e-08,
+      "loss": 0.0574,
+      "step": 17420
+    },
+    {
+      "epoch": 2.822585871678548,
+      "grad_norm": 1.0398857593536377,
+      "learning_rate": 4.57190470525376e-08,
+      "loss": 0.0606,
+      "step": 17421
+    },
+    {
+      "epoch": 2.822747893713545,
+      "grad_norm": 0.9271135926246643,
+      "learning_rate": 4.563583494578638e-08,
+      "loss": 0.0619,
+      "step": 17422
+    },
+    {
+      "epoch": 2.822909915748542,
+      "grad_norm": 0.8910772800445557,
+      "learning_rate": 4.555269793663886e-08,
+      "loss": 0.0572,
+      "step": 17423
+    },
+    {
+      "epoch": 2.8230719377835385,
+      "grad_norm": 0.9423297047615051,
+      "learning_rate": 4.546963602763937e-08,
+      "loss": 0.0671,
+      "step": 17424
+    },
+    {
+      "epoch": 2.8232339598185354,
+      "grad_norm": 0.9918851256370544,
+      "learning_rate": 4.5386649221329516e-08,
+      "loss": 0.0641,
+      "step": 17425
+    },
+    {
+      "epoch": 2.8233959818535324,
+      "grad_norm": 0.9373489022254944,
+      "learning_rate": 4.530373752024753e-08,
+      "loss": 0.0565,
+      "step": 17426
+    },
+    {
+      "epoch": 2.823558003888529,
+      "grad_norm": 0.9793664216995239,
+      "learning_rate": 4.5220900926931374e-08,
+      "loss": 0.0552,
+      "step": 17427
+    },
+    {
+      "epoch": 2.8237200259235253,
+      "grad_norm": 0.9216907024383545,
+      "learning_rate": 4.51381394439146e-08,
+      "loss": 0.0623,
+      "step": 17428
+    },
+    {
+      "epoch": 2.8238820479585223,
+      "grad_norm": 0.9568885564804077,
+      "learning_rate": 4.5055453073730715e-08,
+      "loss": 0.0557,
+      "step": 17429
+    },
+    {
+      "epoch": 2.824044069993519,
+      "grad_norm": 0.8096397519111633,
+      "learning_rate": 4.497284181890882e-08,
+      "loss": 0.0505,
+      "step": 17430
+    },
+    {
+      "epoch": 2.8242060920285157,
+      "grad_norm": 0.8063415884971619,
+      "learning_rate": 4.4890305681977164e-08,
+      "loss": 0.0506,
+      "step": 17431
+    },
+    {
+      "epoch": 2.8243681140635126,
+      "grad_norm": 0.9902795553207397,
+      "learning_rate": 4.480784466546068e-08,
+      "loss": 0.0676,
+      "step": 17432
+    },
+    {
+      "epoch": 2.8245301360985096,
+      "grad_norm": 0.9459345936775208,
+      "learning_rate": 4.4725458771882615e-08,
+      "loss": 0.0634,
+      "step": 17433
+    },
+    {
+      "epoch": 2.824692158133506,
+      "grad_norm": 0.8988990783691406,
+      "learning_rate": 4.4643148003764015e-08,
+      "loss": 0.0587,
+      "step": 17434
+    },
+    {
+      "epoch": 2.824854180168503,
+      "grad_norm": 0.9229409694671631,
+      "learning_rate": 4.456091236362314e-08,
+      "loss": 0.0612,
+      "step": 17435
+    },
+    {
+      "epoch": 2.8250162022035,
+      "grad_norm": 1.0969082117080688,
+      "learning_rate": 4.44787518539766e-08,
+      "loss": 0.0609,
+      "step": 17436
+    },
+    {
+      "epoch": 2.8251782242384964,
+      "grad_norm": 0.863645076751709,
+      "learning_rate": 4.4396666477337645e-08,
+      "loss": 0.055,
+      "step": 17437
+    },
+    {
+      "epoch": 2.8253402462734933,
+      "grad_norm": 0.9219480156898499,
+      "learning_rate": 4.4314656236218444e-08,
+      "loss": 0.061,
+      "step": 17438
+    },
+    {
+      "epoch": 2.82550226830849,
+      "grad_norm": 1.2461076974868774,
+      "learning_rate": 4.423272113312782e-08,
+      "loss": 0.0572,
+      "step": 17439
+    },
+    {
+      "epoch": 2.8256642903434868,
+      "grad_norm": 0.7902315258979797,
+      "learning_rate": 4.415086117057377e-08,
+      "loss": 0.0539,
+      "step": 17440
+    },
+    {
+      "epoch": 2.8258263123784833,
+      "grad_norm": 0.904567301273346,
+      "learning_rate": 4.406907635105984e-08,
+      "loss": 0.0638,
+      "step": 17441
+    },
+    {
+      "epoch": 2.82598833441348,
+      "grad_norm": 0.8270232081413269,
+      "learning_rate": 4.398736667708875e-08,
+      "loss": 0.0571,
+      "step": 17442
+    },
+    {
+      "epoch": 2.826150356448477,
+      "grad_norm": 1.1331998109817505,
+      "learning_rate": 4.390573215116101e-08,
+      "loss": 0.0772,
+      "step": 17443
+    },
+    {
+      "epoch": 2.8263123784834736,
+      "grad_norm": 0.8062843680381775,
+      "learning_rate": 4.382417277577433e-08,
+      "loss": 0.0547,
+      "step": 17444
+    },
+    {
+      "epoch": 2.8264744005184705,
+      "grad_norm": 1.1029092073440552,
+      "learning_rate": 4.374268855342395e-08,
+      "loss": 0.065,
+      "step": 17445
+    },
+    {
+      "epoch": 2.8266364225534675,
+      "grad_norm": 0.9821694493293762,
+      "learning_rate": 4.3661279486603424e-08,
+      "loss": 0.0602,
+      "step": 17446
+    },
+    {
+      "epoch": 2.826798444588464,
+      "grad_norm": 0.8740617036819458,
+      "learning_rate": 4.357994557780354e-08,
+      "loss": 0.0604,
+      "step": 17447
+    },
+    {
+      "epoch": 2.826960466623461,
+      "grad_norm": 1.0357604026794434,
+      "learning_rate": 4.349868682951286e-08,
+      "loss": 0.0618,
+      "step": 17448
+    },
+    {
+      "epoch": 2.8271224886584574,
+      "grad_norm": 0.8919463753700256,
+      "learning_rate": 4.3417503244217726e-08,
+      "loss": 0.0608,
+      "step": 17449
+    },
+    {
+      "epoch": 2.8272845106934543,
+      "grad_norm": 0.9080489873886108,
+      "learning_rate": 4.333639482440199e-08,
+      "loss": 0.0629,
+      "step": 17450
+    },
+    {
+      "epoch": 2.827446532728451,
+      "grad_norm": 0.9267844557762146,
+      "learning_rate": 4.3255361572547836e-08,
+      "loss": 0.0615,
+      "step": 17451
+    },
+    {
+      "epoch": 2.8276085547634477,
+      "grad_norm": 0.9971922039985657,
+      "learning_rate": 4.3174403491134385e-08,
+      "loss": 0.0687,
+      "step": 17452
+    },
+    {
+      "epoch": 2.8277705767984447,
+      "grad_norm": 0.7694254517555237,
+      "learning_rate": 4.309352058263855e-08,
+      "loss": 0.0511,
+      "step": 17453
+    },
+    {
+      "epoch": 2.827932598833441,
+      "grad_norm": 0.826542854309082,
+      "learning_rate": 4.301271284953584e-08,
+      "loss": 0.0516,
+      "step": 17454
+    },
+    {
+      "epoch": 2.828094620868438,
+      "grad_norm": 0.7882339358329773,
+      "learning_rate": 4.29319802942979e-08,
+      "loss": 0.0527,
+      "step": 17455
+    },
+    {
+      "epoch": 2.828256642903435,
+      "grad_norm": 0.8655419945716858,
+      "learning_rate": 4.285132291939526e-08,
+      "loss": 0.059,
+      "step": 17456
+    },
+    {
+      "epoch": 2.8284186649384315,
+      "grad_norm": 0.8806518316268921,
+      "learning_rate": 4.27707407272962e-08,
+      "loss": 0.0605,
+      "step": 17457
+    },
+    {
+      "epoch": 2.8285806869734285,
+      "grad_norm": 0.9758058786392212,
+      "learning_rate": 4.2690233720466265e-08,
+      "loss": 0.0649,
+      "step": 17458
+    },
+    {
+      "epoch": 2.8287427090084254,
+      "grad_norm": 0.8468053340911865,
+      "learning_rate": 4.2609801901368485e-08,
+      "loss": 0.0556,
+      "step": 17459
+    },
+    {
+      "epoch": 2.828904731043422,
+      "grad_norm": 0.8715428709983826,
+      "learning_rate": 4.2529445272463946e-08,
+      "loss": 0.0595,
+      "step": 17460
+    },
+    {
+      "epoch": 2.829066753078419,
+      "grad_norm": 1.172911524772644,
+      "learning_rate": 4.2449163836211507e-08,
+      "loss": 0.0675,
+      "step": 17461
+    },
+    {
+      "epoch": 2.8292287751134153,
+      "grad_norm": 0.836900532245636,
+      "learning_rate": 4.2368957595067264e-08,
+      "loss": 0.0587,
+      "step": 17462
+    },
+    {
+      "epoch": 2.8293907971484122,
+      "grad_norm": 0.9022508263587952,
+      "learning_rate": 4.228882655148564e-08,
+      "loss": 0.0594,
+      "step": 17463
+    },
+    {
+      "epoch": 2.8295528191834087,
+      "grad_norm": 0.8521215915679932,
+      "learning_rate": 4.220877070791857e-08,
+      "loss": 0.0583,
+      "step": 17464
+    },
+    {
+      "epoch": 2.8297148412184057,
+      "grad_norm": 0.9187961220741272,
+      "learning_rate": 4.2128790066815195e-08,
+      "loss": 0.0621,
+      "step": 17465
+    },
+    {
+      "epoch": 2.8298768632534026,
+      "grad_norm": 0.8813655972480774,
+      "learning_rate": 4.204888463062273e-08,
+      "loss": 0.0555,
+      "step": 17466
+    },
+    {
+      "epoch": 2.830038885288399,
+      "grad_norm": 0.7997599840164185,
+      "learning_rate": 4.1969054401786724e-08,
+      "loss": 0.0553,
+      "step": 17467
+    },
+    {
+      "epoch": 2.830200907323396,
+      "grad_norm": 0.8015317916870117,
+      "learning_rate": 4.188929938274911e-08,
+      "loss": 0.0537,
+      "step": 17468
+    },
+    {
+      "epoch": 2.830362929358393,
+      "grad_norm": 0.8224532604217529,
+      "learning_rate": 4.1809619575950425e-08,
+      "loss": 0.0552,
+      "step": 17469
+    },
+    {
+      "epoch": 2.8305249513933894,
+      "grad_norm": 1.1566404104232788,
+      "learning_rate": 4.1730014983828724e-08,
+      "loss": 0.0624,
+      "step": 17470
+    },
+    {
+      "epoch": 2.8306869734283864,
+      "grad_norm": 0.8228906989097595,
+      "learning_rate": 4.165048560881929e-08,
+      "loss": 0.0598,
+      "step": 17471
+    },
+    {
+      "epoch": 2.830848995463383,
+      "grad_norm": 1.1967852115631104,
+      "learning_rate": 4.157103145335628e-08,
+      "loss": 0.0753,
+      "step": 17472
+    },
+    {
+      "epoch": 2.83101101749838,
+      "grad_norm": 0.9739968776702881,
+      "learning_rate": 4.149165251987053e-08,
+      "loss": 0.0575,
+      "step": 17473
+    },
+    {
+      "epoch": 2.8311730395333763,
+      "grad_norm": 0.9316312074661255,
+      "learning_rate": 4.141234881079065e-08,
+      "loss": 0.0653,
+      "step": 17474
+    },
+    {
+      "epoch": 2.8313350615683732,
+      "grad_norm": 1.3163434267044067,
+      "learning_rate": 4.1333120328542754e-08,
+      "loss": 0.0668,
+      "step": 17475
+    },
+    {
+      "epoch": 2.83149708360337,
+      "grad_norm": 0.9953809976577759,
+      "learning_rate": 4.125396707555213e-08,
+      "loss": 0.0615,
+      "step": 17476
+    },
+    {
+      "epoch": 2.8316591056383666,
+      "grad_norm": 1.2429362535476685,
+      "learning_rate": 4.1174889054239885e-08,
+      "loss": 0.0626,
+      "step": 17477
+    },
+    {
+      "epoch": 2.8318211276733636,
+      "grad_norm": 1.1035979986190796,
+      "learning_rate": 4.109588626702576e-08,
+      "loss": 0.0654,
+      "step": 17478
+    },
+    {
+      "epoch": 2.8319831497083605,
+      "grad_norm": 0.7999586462974548,
+      "learning_rate": 4.101695871632699e-08,
+      "loss": 0.0552,
+      "step": 17479
+    },
+    {
+      "epoch": 2.832145171743357,
+      "grad_norm": 1.007940649986267,
+      "learning_rate": 4.0938106404558864e-08,
+      "loss": 0.0631,
+      "step": 17480
+    },
+    {
+      "epoch": 2.832307193778354,
+      "grad_norm": 0.8191199898719788,
+      "learning_rate": 4.085932933413361e-08,
+      "loss": 0.0567,
+      "step": 17481
+    },
+    {
+      "epoch": 2.832469215813351,
+      "grad_norm": 1.0429675579071045,
+      "learning_rate": 4.078062750746209e-08,
+      "loss": 0.0633,
+      "step": 17482
+    },
+    {
+      "epoch": 2.8326312378483474,
+      "grad_norm": 0.9091582298278809,
+      "learning_rate": 4.070200092695209e-08,
+      "loss": 0.0591,
+      "step": 17483
+    },
+    {
+      "epoch": 2.8327932598833443,
+      "grad_norm": 0.9163491725921631,
+      "learning_rate": 4.062344959500947e-08,
+      "loss": 0.0672,
+      "step": 17484
+    },
+    {
+      "epoch": 2.832955281918341,
+      "grad_norm": 1.2758320569992065,
+      "learning_rate": 4.054497351403758e-08,
+      "loss": 0.0597,
+      "step": 17485
+    },
+    {
+      "epoch": 2.8331173039533377,
+      "grad_norm": 1.0452107191085815,
+      "learning_rate": 4.0466572686437833e-08,
+      "loss": 0.0654,
+      "step": 17486
+    },
+    {
+      "epoch": 2.833279325988334,
+      "grad_norm": 1.031578540802002,
+      "learning_rate": 4.038824711460943e-08,
+      "loss": 0.0656,
+      "step": 17487
+    },
+    {
+      "epoch": 2.833441348023331,
+      "grad_norm": 0.8796032071113586,
+      "learning_rate": 4.0309996800947936e-08,
+      "loss": 0.06,
+      "step": 17488
+    },
+    {
+      "epoch": 2.833603370058328,
+      "grad_norm": 0.7887482643127441,
+      "learning_rate": 4.02318217478484e-08,
+      "loss": 0.0564,
+      "step": 17489
+    },
+    {
+      "epoch": 2.8337653920933246,
+      "grad_norm": 0.9122143387794495,
+      "learning_rate": 4.0153721957702504e-08,
+      "loss": 0.0655,
+      "step": 17490
+    },
+    {
+      "epoch": 2.8339274141283215,
+      "grad_norm": 0.9661285877227783,
+      "learning_rate": 4.0075697432900295e-08,
+      "loss": 0.065,
+      "step": 17491
+    },
+    {
+      "epoch": 2.8340894361633184,
+      "grad_norm": 0.8843967318534851,
+      "learning_rate": 3.9997748175828467e-08,
+      "loss": 0.0567,
+      "step": 17492
+    },
+    {
+      "epoch": 2.834251458198315,
+      "grad_norm": 0.9385737180709839,
+      "learning_rate": 3.9919874188872607e-08,
+      "loss": 0.0606,
+      "step": 17493
+    },
+    {
+      "epoch": 2.834413480233312,
+      "grad_norm": 0.9699385166168213,
+      "learning_rate": 3.9842075474415545e-08,
+      "loss": 0.0575,
+      "step": 17494
+    },
+    {
+      "epoch": 2.8345755022683083,
+      "grad_norm": 0.8735253214836121,
+      "learning_rate": 3.976435203483703e-08,
+      "loss": 0.0592,
+      "step": 17495
+    },
+    {
+      "epoch": 2.8347375243033053,
+      "grad_norm": 0.9157827496528625,
+      "learning_rate": 3.9686703872516e-08,
+      "loss": 0.0653,
+      "step": 17496
+    },
+    {
+      "epoch": 2.8348995463383018,
+      "grad_norm": 0.9803964495658875,
+      "learning_rate": 3.960913098982805e-08,
+      "loss": 0.0605,
+      "step": 17497
+    },
+    {
+      "epoch": 2.8350615683732987,
+      "grad_norm": 0.8232876658439636,
+      "learning_rate": 3.953163338914656e-08,
+      "loss": 0.0562,
+      "step": 17498
+    },
+    {
+      "epoch": 2.8352235904082956,
+      "grad_norm": 0.7949116826057434,
+      "learning_rate": 3.94542110728427e-08,
+      "loss": 0.0577,
+      "step": 17499
+    },
+    {
+      "epoch": 2.835385612443292,
+      "grad_norm": 0.94376540184021,
+      "learning_rate": 3.9376864043285943e-08,
+      "loss": 0.0646,
+      "step": 17500
+    },
+    {
+      "epoch": 2.835547634478289,
+      "grad_norm": 1.0657342672348022,
+      "learning_rate": 3.9299592302842195e-08,
+      "loss": 0.061,
+      "step": 17501
+    },
+    {
+      "epoch": 2.835709656513286,
+      "grad_norm": 0.8984468579292297,
+      "learning_rate": 3.922239585387649e-08,
+      "loss": 0.0647,
+      "step": 17502
+    },
+    {
+      "epoch": 2.8358716785482825,
+      "grad_norm": 1.0442514419555664,
+      "learning_rate": 3.914527469875029e-08,
+      "loss": 0.06,
+      "step": 17503
+    },
+    {
+      "epoch": 2.8360337005832794,
+      "grad_norm": 0.9332334995269775,
+      "learning_rate": 3.906822883982336e-08,
+      "loss": 0.063,
+      "step": 17504
+    },
+    {
+      "epoch": 2.8361957226182763,
+      "grad_norm": 1.0261024236679077,
+      "learning_rate": 3.8991258279453544e-08,
+      "loss": 0.0672,
+      "step": 17505
+    },
+    {
+      "epoch": 2.836357744653273,
+      "grad_norm": 1.0123735666275024,
+      "learning_rate": 3.891436301999563e-08,
+      "loss": 0.0578,
+      "step": 17506
+    },
+    {
+      "epoch": 2.8365197666882693,
+      "grad_norm": 0.8809978365898132,
+      "learning_rate": 3.883754306380244e-08,
+      "loss": 0.0535,
+      "step": 17507
+    },
+    {
+      "epoch": 2.8366817887232663,
+      "grad_norm": 0.9029288291931152,
+      "learning_rate": 3.876079841322461e-08,
+      "loss": 0.0615,
+      "step": 17508
+    },
+    {
+      "epoch": 2.836843810758263,
+      "grad_norm": 0.9749417304992676,
+      "learning_rate": 3.868412907061026e-08,
+      "loss": 0.0564,
+      "step": 17509
+    },
+    {
+      "epoch": 2.8370058327932597,
+      "grad_norm": 0.9761818051338196,
+      "learning_rate": 3.8607535038305276e-08,
+      "loss": 0.0627,
+      "step": 17510
+    },
+    {
+      "epoch": 2.8371678548282566,
+      "grad_norm": 0.8667659759521484,
+      "learning_rate": 3.853101631865336e-08,
+      "loss": 0.0568,
+      "step": 17511
+    },
+    {
+      "epoch": 2.8373298768632536,
+      "grad_norm": 1.0688188076019287,
+      "learning_rate": 3.84545729139954e-08,
+      "loss": 0.0594,
+      "step": 17512
+    },
+    {
+      "epoch": 2.83749189889825,
+      "grad_norm": 0.9591261744499207,
+      "learning_rate": 3.837820482667121e-08,
+      "loss": 0.0625,
+      "step": 17513
+    },
+    {
+      "epoch": 2.837653920933247,
+      "grad_norm": 0.90581876039505,
+      "learning_rate": 3.83019120590164e-08,
+      "loss": 0.0537,
+      "step": 17514
+    },
+    {
+      "epoch": 2.837815942968244,
+      "grad_norm": 0.8603760004043579,
+      "learning_rate": 3.822569461336606e-08,
+      "loss": 0.0517,
+      "step": 17515
+    },
+    {
+      "epoch": 2.8379779650032404,
+      "grad_norm": 0.9439842104911804,
+      "learning_rate": 3.814955249205221e-08,
+      "loss": 0.0637,
+      "step": 17516
+    },
+    {
+      "epoch": 2.8381399870382373,
+      "grad_norm": 0.9294354915618896,
+      "learning_rate": 3.8073485697404655e-08,
+      "loss": 0.0601,
+      "step": 17517
+    },
+    {
+      "epoch": 2.838302009073234,
+      "grad_norm": 0.8488750457763672,
+      "learning_rate": 3.7997494231750145e-08,
+      "loss": 0.0617,
+      "step": 17518
+    },
+    {
+      "epoch": 2.8384640311082308,
+      "grad_norm": 0.9641651511192322,
+      "learning_rate": 3.792157809741459e-08,
+      "loss": 0.0648,
+      "step": 17519
+    },
+    {
+      "epoch": 2.8386260531432272,
+      "grad_norm": 1.0396709442138672,
+      "learning_rate": 3.784573729672086e-08,
+      "loss": 0.0551,
+      "step": 17520
+    },
+    {
+      "epoch": 2.838788075178224,
+      "grad_norm": 0.9486382603645325,
+      "learning_rate": 3.7769971831989325e-08,
+      "loss": 0.0659,
+      "step": 17521
+    },
+    {
+      "epoch": 2.838950097213221,
+      "grad_norm": 0.972588300704956,
+      "learning_rate": 3.769428170553785e-08,
+      "loss": 0.0649,
+      "step": 17522
+    },
+    {
+      "epoch": 2.8391121192482176,
+      "grad_norm": 1.0495613813400269,
+      "learning_rate": 3.761866691968291e-08,
+      "loss": 0.0679,
+      "step": 17523
+    },
+    {
+      "epoch": 2.8392741412832145,
+      "grad_norm": 0.8821991682052612,
+      "learning_rate": 3.754312747673766e-08,
+      "loss": 0.056,
+      "step": 17524
+    },
+    {
+      "epoch": 2.8394361633182115,
+      "grad_norm": 1.0493261814117432,
+      "learning_rate": 3.746766337901386e-08,
+      "loss": 0.0586,
+      "step": 17525
+    },
+    {
+      "epoch": 2.839598185353208,
+      "grad_norm": 0.923512876033783,
+      "learning_rate": 3.739227462882022e-08,
+      "loss": 0.0601,
+      "step": 17526
+    },
+    {
+      "epoch": 2.839760207388205,
+      "grad_norm": 0.8735133409500122,
+      "learning_rate": 3.731696122846379e-08,
+      "loss": 0.0589,
+      "step": 17527
+    },
+    {
+      "epoch": 2.839922229423202,
+      "grad_norm": 1.056152105331421,
+      "learning_rate": 3.724172318024854e-08,
+      "loss": 0.0684,
+      "step": 17528
+    },
+    {
+      "epoch": 2.8400842514581983,
+      "grad_norm": 0.8753458857536316,
+      "learning_rate": 3.71665604864771e-08,
+      "loss": 0.0581,
+      "step": 17529
+    },
+    {
+      "epoch": 2.840246273493195,
+      "grad_norm": 1.102301836013794,
+      "learning_rate": 3.709147314944872e-08,
+      "loss": 0.07,
+      "step": 17530
+    },
+    {
+      "epoch": 2.8404082955281917,
+      "grad_norm": 0.8725458383560181,
+      "learning_rate": 3.7016461171461296e-08,
+      "loss": 0.0613,
+      "step": 17531
+    },
+    {
+      "epoch": 2.8405703175631887,
+      "grad_norm": 0.938378095626831,
+      "learning_rate": 3.6941524554809924e-08,
+      "loss": 0.0622,
+      "step": 17532
+    },
+    {
+      "epoch": 2.840732339598185,
+      "grad_norm": 0.9458240270614624,
+      "learning_rate": 3.686666330178695e-08,
+      "loss": 0.0578,
+      "step": 17533
+    },
+    {
+      "epoch": 2.840894361633182,
+      "grad_norm": 0.9221711754798889,
+      "learning_rate": 3.6791877414683594e-08,
+      "loss": 0.0605,
+      "step": 17534
+    },
+    {
+      "epoch": 2.841056383668179,
+      "grad_norm": 0.9185110926628113,
+      "learning_rate": 3.6717166895788306e-08,
+      "loss": 0.0638,
+      "step": 17535
+    },
+    {
+      "epoch": 2.8412184057031755,
+      "grad_norm": 0.8876226544380188,
+      "learning_rate": 3.664253174738647e-08,
+      "loss": 0.0588,
+      "step": 17536
+    },
+    {
+      "epoch": 2.8413804277381725,
+      "grad_norm": 0.9121412634849548,
+      "learning_rate": 3.656797197176182e-08,
+      "loss": 0.0593,
+      "step": 17537
+    },
+    {
+      "epoch": 2.8415424497731694,
+      "grad_norm": 0.8094960451126099,
+      "learning_rate": 3.649348757119614e-08,
+      "loss": 0.0577,
+      "step": 17538
+    },
+    {
+      "epoch": 2.841704471808166,
+      "grad_norm": 0.8800841569900513,
+      "learning_rate": 3.641907854796789e-08,
+      "loss": 0.0586,
+      "step": 17539
+    },
+    {
+      "epoch": 2.841866493843163,
+      "grad_norm": 0.8749896883964539,
+      "learning_rate": 3.634474490435413e-08,
+      "loss": 0.0591,
+      "step": 17540
+    },
+    {
+      "epoch": 2.8420285158781593,
+      "grad_norm": 0.9184433221817017,
+      "learning_rate": 3.627048664262916e-08,
+      "loss": 0.0525,
+      "step": 17541
+    },
+    {
+      "epoch": 2.8421905379131562,
+      "grad_norm": 0.9051634669303894,
+      "learning_rate": 3.6196303765065333e-08,
+      "loss": 0.0676,
+      "step": 17542
+    },
+    {
+      "epoch": 2.8423525599481527,
+      "grad_norm": 0.8812767863273621,
+      "learning_rate": 3.612219627393221e-08,
+      "loss": 0.0601,
+      "step": 17543
+    },
+    {
+      "epoch": 2.8425145819831497,
+      "grad_norm": 0.9540053606033325,
+      "learning_rate": 3.60481641714977e-08,
+      "loss": 0.0591,
+      "step": 17544
+    },
+    {
+      "epoch": 2.8426766040181466,
+      "grad_norm": 1.0543007850646973,
+      "learning_rate": 3.597420746002639e-08,
+      "loss": 0.0655,
+      "step": 17545
+    },
+    {
+      "epoch": 2.842838626053143,
+      "grad_norm": 0.8972881436347961,
+      "learning_rate": 3.590032614178174e-08,
+      "loss": 0.06,
+      "step": 17546
+    },
+    {
+      "epoch": 2.84300064808814,
+      "grad_norm": 0.8661593794822693,
+      "learning_rate": 3.5826520219023887e-08,
+      "loss": 0.0607,
+      "step": 17547
+    },
+    {
+      "epoch": 2.843162670123137,
+      "grad_norm": 0.896032452583313,
+      "learning_rate": 3.57527896940113e-08,
+      "loss": 0.0592,
+      "step": 17548
+    },
+    {
+      "epoch": 2.8433246921581334,
+      "grad_norm": 0.822556734085083,
+      "learning_rate": 3.567913456900024e-08,
+      "loss": 0.0561,
+      "step": 17549
+    },
+    {
+      "epoch": 2.8434867141931304,
+      "grad_norm": 0.9163435101509094,
+      "learning_rate": 3.560555484624417e-08,
+      "loss": 0.0679,
+      "step": 17550
+    },
+    {
+      "epoch": 2.843648736228127,
+      "grad_norm": 0.8780238032341003,
+      "learning_rate": 3.5532050527994076e-08,
+      "loss": 0.0533,
+      "step": 17551
+    },
+    {
+      "epoch": 2.843810758263124,
+      "grad_norm": 0.9988797903060913,
+      "learning_rate": 3.545862161649927e-08,
+      "loss": 0.057,
+      "step": 17552
+    },
+    {
+      "epoch": 2.8439727802981203,
+      "grad_norm": 0.8687535524368286,
+      "learning_rate": 3.538526811400711e-08,
+      "loss": 0.0607,
+      "step": 17553
+    },
+    {
+      "epoch": 2.844134802333117,
+      "grad_norm": 1.180427074432373,
+      "learning_rate": 3.531199002276109e-08,
+      "loss": 0.0727,
+      "step": 17554
+    },
+    {
+      "epoch": 2.844296824368114,
+      "grad_norm": 0.8522838354110718,
+      "learning_rate": 3.5238787345003855e-08,
+      "loss": 0.0549,
+      "step": 17555
+    },
+    {
+      "epoch": 2.8444588464031106,
+      "grad_norm": 0.9009879231452942,
+      "learning_rate": 3.5165660082975006e-08,
+      "loss": 0.0628,
+      "step": 17556
+    },
+    {
+      "epoch": 2.8446208684381076,
+      "grad_norm": 0.8708153963088989,
+      "learning_rate": 3.50926082389122e-08,
+      "loss": 0.0574,
+      "step": 17557
+    },
+    {
+      "epoch": 2.8447828904731045,
+      "grad_norm": 0.8726041316986084,
+      "learning_rate": 3.501963181505058e-08,
+      "loss": 0.063,
+      "step": 17558
+    },
+    {
+      "epoch": 2.844944912508101,
+      "grad_norm": 0.9148244261741638,
+      "learning_rate": 3.4946730813623376e-08,
+      "loss": 0.0561,
+      "step": 17559
+    },
+    {
+      "epoch": 2.845106934543098,
+      "grad_norm": 0.9740058779716492,
+      "learning_rate": 3.487390523686074e-08,
+      "loss": 0.0645,
+      "step": 17560
+    },
+    {
+      "epoch": 2.845268956578095,
+      "grad_norm": 0.8795120716094971,
+      "learning_rate": 3.4801155086991165e-08,
+      "loss": 0.0604,
+      "step": 17561
+    },
+    {
+      "epoch": 2.8454309786130914,
+      "grad_norm": 0.8973474502563477,
+      "learning_rate": 3.472848036624038e-08,
+      "loss": 0.0601,
+      "step": 17562
+    },
+    {
+      "epoch": 2.8455930006480883,
+      "grad_norm": 0.9123368859291077,
+      "learning_rate": 3.4655881076832156e-08,
+      "loss": 0.0622,
+      "step": 17563
+    },
+    {
+      "epoch": 2.845755022683085,
+      "grad_norm": 0.8565255403518677,
+      "learning_rate": 3.4583357220988326e-08,
+      "loss": 0.0551,
+      "step": 17564
+    },
+    {
+      "epoch": 2.8459170447180817,
+      "grad_norm": 0.815396249294281,
+      "learning_rate": 3.451090880092739e-08,
+      "loss": 0.0529,
+      "step": 17565
+    },
+    {
+      "epoch": 2.846079066753078,
+      "grad_norm": 0.7915904521942139,
+      "learning_rate": 3.443853581886619e-08,
+      "loss": 0.0512,
+      "step": 17566
+    },
+    {
+      "epoch": 2.846241088788075,
+      "grad_norm": 1.000328779220581,
+      "learning_rate": 3.4366238277019625e-08,
+      "loss": 0.0591,
+      "step": 17567
+    },
+    {
+      "epoch": 2.846403110823072,
+      "grad_norm": 0.8379427194595337,
+      "learning_rate": 3.4294016177598974e-08,
+      "loss": 0.0601,
+      "step": 17568
+    },
+    {
+      "epoch": 2.8465651328580686,
+      "grad_norm": 0.8759309649467468,
+      "learning_rate": 3.42218695228147e-08,
+      "loss": 0.0594,
+      "step": 17569
+    },
+    {
+      "epoch": 2.8467271548930655,
+      "grad_norm": 0.9378208518028259,
+      "learning_rate": 3.4149798314874195e-08,
+      "loss": 0.0637,
+      "step": 17570
+    },
+    {
+      "epoch": 2.8468891769280624,
+      "grad_norm": 0.9910644888877869,
+      "learning_rate": 3.4077802555982645e-08,
+      "loss": 0.0633,
+      "step": 17571
+    },
+    {
+      "epoch": 2.847051198963059,
+      "grad_norm": 0.8594523668289185,
+      "learning_rate": 3.4005882248343e-08,
+      "loss": 0.057,
+      "step": 17572
+    },
+    {
+      "epoch": 2.847213220998056,
+      "grad_norm": 0.9198819994926453,
+      "learning_rate": 3.393403739415546e-08,
+      "loss": 0.0638,
+      "step": 17573
+    },
+    {
+      "epoch": 2.8473752430330523,
+      "grad_norm": 0.8322975039482117,
+      "learning_rate": 3.3862267995618817e-08,
+      "loss": 0.0572,
+      "step": 17574
+    },
+    {
+      "epoch": 2.8475372650680493,
+      "grad_norm": 0.9040446281433105,
+      "learning_rate": 3.379057405492908e-08,
+      "loss": 0.0641,
+      "step": 17575
+    },
+    {
+      "epoch": 2.8476992871030458,
+      "grad_norm": 0.8539929986000061,
+      "learning_rate": 3.3718955574279234e-08,
+      "loss": 0.0602,
+      "step": 17576
+    },
+    {
+      "epoch": 2.8478613091380427,
+      "grad_norm": 0.7834108471870422,
+      "learning_rate": 3.3647412555861126e-08,
+      "loss": 0.051,
+      "step": 17577
+    },
+    {
+      "epoch": 2.8480233311730396,
+      "grad_norm": 0.8313713073730469,
+      "learning_rate": 3.357594500186384e-08,
+      "loss": 0.0548,
+      "step": 17578
+    },
+    {
+      "epoch": 2.848185353208036,
+      "grad_norm": 0.8187208771705627,
+      "learning_rate": 3.3504552914474244e-08,
+      "loss": 0.0554,
+      "step": 17579
+    },
+    {
+      "epoch": 2.848347375243033,
+      "grad_norm": 0.8749431371688843,
+      "learning_rate": 3.3433236295876134e-08,
+      "loss": 0.054,
+      "step": 17580
+    },
+    {
+      "epoch": 2.84850939727803,
+      "grad_norm": 0.9242835640907288,
+      "learning_rate": 3.336199514825195e-08,
+      "loss": 0.0567,
+      "step": 17581
+    },
+    {
+      "epoch": 2.8486714193130265,
+      "grad_norm": 0.9290145635604858,
+      "learning_rate": 3.329082947378215e-08,
+      "loss": 0.0639,
+      "step": 17582
+    },
+    {
+      "epoch": 2.8488334413480234,
+      "grad_norm": 0.908678412437439,
+      "learning_rate": 3.3219739274643057e-08,
+      "loss": 0.0579,
+      "step": 17583
+    },
+    {
+      "epoch": 2.8489954633830203,
+      "grad_norm": 0.9778639078140259,
+      "learning_rate": 3.314872455301071e-08,
+      "loss": 0.0648,
+      "step": 17584
+    },
+    {
+      "epoch": 2.849157485418017,
+      "grad_norm": 0.6851035952568054,
+      "learning_rate": 3.3077785311057545e-08,
+      "loss": 0.0508,
+      "step": 17585
+    },
+    {
+      "epoch": 2.8493195074530138,
+      "grad_norm": 0.9486784338951111,
+      "learning_rate": 3.300692155095458e-08,
+      "loss": 0.0606,
+      "step": 17586
+    },
+    {
+      "epoch": 2.8494815294880103,
+      "grad_norm": 1.097536325454712,
+      "learning_rate": 3.293613327486983e-08,
+      "loss": 0.0656,
+      "step": 17587
+    },
+    {
+      "epoch": 2.849643551523007,
+      "grad_norm": 0.9391690492630005,
+      "learning_rate": 3.286542048496904e-08,
+      "loss": 0.0613,
+      "step": 17588
+    },
+    {
+      "epoch": 2.8498055735580037,
+      "grad_norm": 0.9068102836608887,
+      "learning_rate": 3.2794783183416055e-08,
+      "loss": 0.059,
+      "step": 17589
+    },
+    {
+      "epoch": 2.8499675955930006,
+      "grad_norm": 0.851341962814331,
+      "learning_rate": 3.272422137237219e-08,
+      "loss": 0.0532,
+      "step": 17590
+    },
+    {
+      "epoch": 2.8501296176279975,
+      "grad_norm": 0.9030383229255676,
+      "learning_rate": 3.265373505399627e-08,
+      "loss": 0.0536,
+      "step": 17591
+    },
+    {
+      "epoch": 2.850291639662994,
+      "grad_norm": 0.8289960026741028,
+      "learning_rate": 3.258332423044547e-08,
+      "loss": 0.0602,
+      "step": 17592
+    },
+    {
+      "epoch": 2.850453661697991,
+      "grad_norm": 0.8833999633789062,
+      "learning_rate": 3.2512988903873885e-08,
+      "loss": 0.0623,
+      "step": 17593
+    },
+    {
+      "epoch": 2.850615683732988,
+      "grad_norm": 1.016706943511963,
+      "learning_rate": 3.2442729076433697e-08,
+      "loss": 0.0629,
+      "step": 17594
+    },
+    {
+      "epoch": 2.8507777057679844,
+      "grad_norm": 0.903662383556366,
+      "learning_rate": 3.237254475027429e-08,
+      "loss": 0.0596,
+      "step": 17595
+    },
+    {
+      "epoch": 2.8509397278029813,
+      "grad_norm": 0.8987908959388733,
+      "learning_rate": 3.230243592754368e-08,
+      "loss": 0.0649,
+      "step": 17596
+    },
+    {
+      "epoch": 2.851101749837978,
+      "grad_norm": 0.9659387469291687,
+      "learning_rate": 3.223240261038707e-08,
+      "loss": 0.0661,
+      "step": 17597
+    },
+    {
+      "epoch": 2.8512637718729748,
+      "grad_norm": 0.9991119503974915,
+      "learning_rate": 3.2162444800946655e-08,
+      "loss": 0.0595,
+      "step": 17598
+    },
+    {
+      "epoch": 2.8514257939079712,
+      "grad_norm": 0.8747795224189758,
+      "learning_rate": 3.209256250136378e-08,
+      "loss": 0.0505,
+      "step": 17599
+    },
+    {
+      "epoch": 2.851587815942968,
+      "grad_norm": 0.9156426191329956,
+      "learning_rate": 3.202275571377589e-08,
+      "loss": 0.0553,
+      "step": 17600
+    },
+    {
+      "epoch": 2.851749837977965,
+      "grad_norm": 0.9192603826522827,
+      "learning_rate": 3.1953024440319334e-08,
+      "loss": 0.064,
+      "step": 17601
+    },
+    {
+      "epoch": 2.8519118600129616,
+      "grad_norm": 0.9150480031967163,
+      "learning_rate": 3.188336868312769e-08,
+      "loss": 0.0652,
+      "step": 17602
+    },
+    {
+      "epoch": 2.8520738820479585,
+      "grad_norm": 0.8279345035552979,
+      "learning_rate": 3.18137884443323e-08,
+      "loss": 0.0577,
+      "step": 17603
+    },
+    {
+      "epoch": 2.8522359040829555,
+      "grad_norm": 0.9174199104309082,
+      "learning_rate": 3.1744283726062306e-08,
+      "loss": 0.0614,
+      "step": 17604
+    },
+    {
+      "epoch": 2.852397926117952,
+      "grad_norm": 0.9263173937797546,
+      "learning_rate": 3.167485453044378e-08,
+      "loss": 0.0642,
+      "step": 17605
+    },
+    {
+      "epoch": 2.852559948152949,
+      "grad_norm": 0.8777797222137451,
+      "learning_rate": 3.160550085960168e-08,
+      "loss": 0.0556,
+      "step": 17606
+    },
+    {
+      "epoch": 2.852721970187946,
+      "grad_norm": 0.8794631958007812,
+      "learning_rate": 3.153622271565793e-08,
+      "loss": 0.0583,
+      "step": 17607
+    },
+    {
+      "epoch": 2.8528839922229423,
+      "grad_norm": 0.8909582495689392,
+      "learning_rate": 3.1467020100732215e-08,
+      "loss": 0.0647,
+      "step": 17608
+    },
+    {
+      "epoch": 2.8530460142579392,
+      "grad_norm": 0.9028645753860474,
+      "learning_rate": 3.139789301694146e-08,
+      "loss": 0.0583,
+      "step": 17609
+    },
+    {
+      "epoch": 2.8532080362929357,
+      "grad_norm": 0.9116567969322205,
+      "learning_rate": 3.1328841466401746e-08,
+      "loss": 0.0605,
+      "step": 17610
+    },
+    {
+      "epoch": 2.8533700583279327,
+      "grad_norm": 1.222704529762268,
+      "learning_rate": 3.125986545122528e-08,
+      "loss": 0.0621,
+      "step": 17611
+    },
+    {
+      "epoch": 2.853532080362929,
+      "grad_norm": 0.8335679769515991,
+      "learning_rate": 3.1190964973522865e-08,
+      "loss": 0.0562,
+      "step": 17612
+    },
+    {
+      "epoch": 2.853694102397926,
+      "grad_norm": 0.9687299728393555,
+      "learning_rate": 3.112214003540254e-08,
+      "loss": 0.057,
+      "step": 17613
+    },
+    {
+      "epoch": 2.853856124432923,
+      "grad_norm": 0.9810713529586792,
+      "learning_rate": 3.105339063896956e-08,
+      "loss": 0.07,
+      "step": 17614
+    },
+    {
+      "epoch": 2.8540181464679195,
+      "grad_norm": 1.0105944871902466,
+      "learning_rate": 3.098471678632892e-08,
+      "loss": 0.0691,
+      "step": 17615
+    },
+    {
+      "epoch": 2.8541801685029164,
+      "grad_norm": 0.9086531400680542,
+      "learning_rate": 3.0916118479580593e-08,
+      "loss": 0.0627,
+      "step": 17616
+    },
+    {
+      "epoch": 2.8543421905379134,
+      "grad_norm": 0.8142098188400269,
+      "learning_rate": 3.084759572082402e-08,
+      "loss": 0.0532,
+      "step": 17617
+    },
+    {
+      "epoch": 2.85450421257291,
+      "grad_norm": 0.9001931548118591,
+      "learning_rate": 3.077914851215585e-08,
+      "loss": 0.0685,
+      "step": 17618
+    },
+    {
+      "epoch": 2.854666234607907,
+      "grad_norm": 0.7494874000549316,
+      "learning_rate": 3.071077685567025e-08,
+      "loss": 0.0548,
+      "step": 17619
+    },
+    {
+      "epoch": 2.8548282566429033,
+      "grad_norm": 0.956156849861145,
+      "learning_rate": 3.064248075345916e-08,
+      "loss": 0.06,
+      "step": 17620
+    },
+    {
+      "epoch": 2.8549902786779002,
+      "grad_norm": 0.939167857170105,
+      "learning_rate": 3.057426020761256e-08,
+      "loss": 0.066,
+      "step": 17621
+    },
+    {
+      "epoch": 2.8551523007128967,
+      "grad_norm": 0.9294570088386536,
+      "learning_rate": 3.050611522021796e-08,
+      "loss": 0.0586,
+      "step": 17622
+    },
+    {
+      "epoch": 2.8553143227478937,
+      "grad_norm": 0.9607594609260559,
+      "learning_rate": 3.043804579336007e-08,
+      "loss": 0.0625,
+      "step": 17623
+    },
+    {
+      "epoch": 2.8554763447828906,
+      "grad_norm": 1.0955085754394531,
+      "learning_rate": 3.0370051929121405e-08,
+      "loss": 0.0652,
+      "step": 17624
+    },
+    {
+      "epoch": 2.855638366817887,
+      "grad_norm": 0.9362739324569702,
+      "learning_rate": 3.030213362958306e-08,
+      "loss": 0.0622,
+      "step": 17625
+    },
+    {
+      "epoch": 2.855800388852884,
+      "grad_norm": 0.8863419890403748,
+      "learning_rate": 3.023429089682284e-08,
+      "loss": 0.0599,
+      "step": 17626
+    },
+    {
+      "epoch": 2.855962410887881,
+      "grad_norm": 0.8718972206115723,
+      "learning_rate": 3.0166523732916564e-08,
+      "loss": 0.0588,
+      "step": 17627
+    },
+    {
+      "epoch": 2.8561244329228774,
+      "grad_norm": 1.004003643989563,
+      "learning_rate": 3.009883213993786e-08,
+      "loss": 0.0603,
+      "step": 17628
+    },
+    {
+      "epoch": 2.8562864549578744,
+      "grad_norm": 0.9552223682403564,
+      "learning_rate": 3.0031216119957576e-08,
+      "loss": 0.053,
+      "step": 17629
+    },
+    {
+      "epoch": 2.8564484769928713,
+      "grad_norm": 0.9536174535751343,
+      "learning_rate": 2.996367567504544e-08,
+      "loss": 0.0656,
+      "step": 17630
+    },
+    {
+      "epoch": 2.856610499027868,
+      "grad_norm": 1.1152795553207397,
+      "learning_rate": 2.989621080726701e-08,
+      "loss": 0.0635,
+      "step": 17631
+    },
+    {
+      "epoch": 2.8567725210628643,
+      "grad_norm": 0.9013298749923706,
+      "learning_rate": 2.9828821518687045e-08,
+      "loss": 0.0566,
+      "step": 17632
+    },
+    {
+      "epoch": 2.856934543097861,
+      "grad_norm": 0.9204021096229553,
+      "learning_rate": 2.9761507811367497e-08,
+      "loss": 0.0631,
+      "step": 17633
+    },
+    {
+      "epoch": 2.857096565132858,
+      "grad_norm": 0.9465929865837097,
+      "learning_rate": 2.9694269687367826e-08,
+      "loss": 0.0542,
+      "step": 17634
+    },
+    {
+      "epoch": 2.8572585871678546,
+      "grad_norm": 0.8179449439048767,
+      "learning_rate": 2.962710714874556e-08,
+      "loss": 0.0581,
+      "step": 17635
+    },
+    {
+      "epoch": 2.8574206092028516,
+      "grad_norm": 1.1688090562820435,
+      "learning_rate": 2.9560020197555716e-08,
+      "loss": 0.0635,
+      "step": 17636
+    },
+    {
+      "epoch": 2.8575826312378485,
+      "grad_norm": 1.0356324911117554,
+      "learning_rate": 2.9493008835850823e-08,
+      "loss": 0.0592,
+      "step": 17637
+    },
+    {
+      "epoch": 2.857744653272845,
+      "grad_norm": 0.9192071557044983,
+      "learning_rate": 2.9426073065681183e-08,
+      "loss": 0.0646,
+      "step": 17638
+    },
+    {
+      "epoch": 2.857906675307842,
+      "grad_norm": 0.8982980251312256,
+      "learning_rate": 2.9359212889095157e-08,
+      "loss": 0.0616,
+      "step": 17639
+    },
+    {
+      "epoch": 2.858068697342839,
+      "grad_norm": 0.8842819333076477,
+      "learning_rate": 2.929242830813861e-08,
+      "loss": 0.0607,
+      "step": 17640
+    },
+    {
+      "epoch": 2.8582307193778353,
+      "grad_norm": 0.8390337824821472,
+      "learning_rate": 2.9225719324854628e-08,
+      "loss": 0.0564,
+      "step": 17641
+    },
+    {
+      "epoch": 2.8583927414128323,
+      "grad_norm": 0.9694913029670715,
+      "learning_rate": 2.915908594128436e-08,
+      "loss": 0.0596,
+      "step": 17642
+    },
+    {
+      "epoch": 2.8585547634478288,
+      "grad_norm": 0.9143932461738586,
+      "learning_rate": 2.9092528159466727e-08,
+      "loss": 0.0585,
+      "step": 17643
+    },
+    {
+      "epoch": 2.8587167854828257,
+      "grad_norm": 0.9615548849105835,
+      "learning_rate": 2.9026045981438434e-08,
+      "loss": 0.0646,
+      "step": 17644
+    },
+    {
+      "epoch": 2.858878807517822,
+      "grad_norm": 0.8890393972396851,
+      "learning_rate": 2.895963940923341e-08,
+      "loss": 0.0605,
+      "step": 17645
+    },
+    {
+      "epoch": 2.859040829552819,
+      "grad_norm": 0.9828749299049377,
+      "learning_rate": 2.889330844488364e-08,
+      "loss": 0.0568,
+      "step": 17646
+    },
+    {
+      "epoch": 2.859202851587816,
+      "grad_norm": 0.8218222856521606,
+      "learning_rate": 2.882705309041861e-08,
+      "loss": 0.0532,
+      "step": 17647
+    },
+    {
+      "epoch": 2.8593648736228126,
+      "grad_norm": 1.0544509887695312,
+      "learning_rate": 2.8760873347865593e-08,
+      "loss": 0.0649,
+      "step": 17648
+    },
+    {
+      "epoch": 2.8595268956578095,
+      "grad_norm": 1.0283457040786743,
+      "learning_rate": 2.869476921924963e-08,
+      "loss": 0.0588,
+      "step": 17649
+    },
+    {
+      "epoch": 2.8596889176928064,
+      "grad_norm": 0.9277712106704712,
+      "learning_rate": 2.862874070659327e-08,
+      "loss": 0.0557,
+      "step": 17650
+    },
+    {
+      "epoch": 2.859850939727803,
+      "grad_norm": 0.8709115982055664,
+      "learning_rate": 2.8562787811916848e-08,
+      "loss": 0.0558,
+      "step": 17651
+    },
+    {
+      "epoch": 2.8600129617628,
+      "grad_norm": 0.9610093235969543,
+      "learning_rate": 2.8496910537238185e-08,
+      "loss": 0.0622,
+      "step": 17652
+    },
+    {
+      "epoch": 2.8601749837977968,
+      "grad_norm": 1.003127098083496,
+      "learning_rate": 2.8431108884573454e-08,
+      "loss": 0.0611,
+      "step": 17653
+    },
+    {
+      "epoch": 2.8603370058327933,
+      "grad_norm": 0.879693865776062,
+      "learning_rate": 2.8365382855935487e-08,
+      "loss": 0.0526,
+      "step": 17654
+    },
+    {
+      "epoch": 2.8604990278677898,
+      "grad_norm": 0.8258382678031921,
+      "learning_rate": 2.8299732453335725e-08,
+      "loss": 0.0576,
+      "step": 17655
+    },
+    {
+      "epoch": 2.8606610499027867,
+      "grad_norm": 0.9305282831192017,
+      "learning_rate": 2.8234157678782846e-08,
+      "loss": 0.0603,
+      "step": 17656
+    },
+    {
+      "epoch": 2.8608230719377836,
+      "grad_norm": 0.825541615486145,
+      "learning_rate": 2.8168658534282743e-08,
+      "loss": 0.0553,
+      "step": 17657
+    },
+    {
+      "epoch": 2.86098509397278,
+      "grad_norm": 0.8502270579338074,
+      "learning_rate": 2.8103235021840204e-08,
+      "loss": 0.0581,
+      "step": 17658
+    },
+    {
+      "epoch": 2.861147116007777,
+      "grad_norm": 0.9126089215278625,
+      "learning_rate": 2.8037887143456954e-08,
+      "loss": 0.0582,
+      "step": 17659
+    },
+    {
+      "epoch": 2.861309138042774,
+      "grad_norm": 0.8462595343589783,
+      "learning_rate": 2.7972614901132235e-08,
+      "loss": 0.0621,
+      "step": 17660
+    },
+    {
+      "epoch": 2.8614711600777705,
+      "grad_norm": 0.8996595144271851,
+      "learning_rate": 2.790741829686333e-08,
+      "loss": 0.0602,
+      "step": 17661
+    },
+    {
+      "epoch": 2.8616331821127674,
+      "grad_norm": 1.2317053079605103,
+      "learning_rate": 2.784229733264504e-08,
+      "loss": 0.0688,
+      "step": 17662
+    },
+    {
+      "epoch": 2.8617952041477643,
+      "grad_norm": 1.1419668197631836,
+      "learning_rate": 2.7777252010469657e-08,
+      "loss": 0.0643,
+      "step": 17663
+    },
+    {
+      "epoch": 2.861957226182761,
+      "grad_norm": 1.055248737335205,
+      "learning_rate": 2.771228233232809e-08,
+      "loss": 0.0607,
+      "step": 17664
+    },
+    {
+      "epoch": 2.8621192482177578,
+      "grad_norm": 0.8915876746177673,
+      "learning_rate": 2.7647388300207635e-08,
+      "loss": 0.0603,
+      "step": 17665
+    },
+    {
+      "epoch": 2.8622812702527543,
+      "grad_norm": 1.019177794456482,
+      "learning_rate": 2.7582569916094205e-08,
+      "loss": 0.062,
+      "step": 17666
+    },
+    {
+      "epoch": 2.862443292287751,
+      "grad_norm": 0.932235598564148,
+      "learning_rate": 2.7517827181970937e-08,
+      "loss": 0.0606,
+      "step": 17667
+    },
+    {
+      "epoch": 2.8626053143227477,
+      "grad_norm": 0.764743983745575,
+      "learning_rate": 2.745316009981902e-08,
+      "loss": 0.0513,
+      "step": 17668
+    },
+    {
+      "epoch": 2.8627673363577446,
+      "grad_norm": 0.9902213215827942,
+      "learning_rate": 2.7388568671616877e-08,
+      "loss": 0.0629,
+      "step": 17669
+    },
+    {
+      "epoch": 2.8629293583927415,
+      "grad_norm": 1.0022236108779907,
+      "learning_rate": 2.732405289934098e-08,
+      "loss": 0.0586,
+      "step": 17670
+    },
+    {
+      "epoch": 2.863091380427738,
+      "grad_norm": 1.2449359893798828,
+      "learning_rate": 2.7259612784965307e-08,
+      "loss": 0.0579,
+      "step": 17671
+    },
+    {
+      "epoch": 2.863253402462735,
+      "grad_norm": 0.9988813400268555,
+      "learning_rate": 2.719524833046133e-08,
+      "loss": 0.0626,
+      "step": 17672
+    },
+    {
+      "epoch": 2.863415424497732,
+      "grad_norm": 0.9619942307472229,
+      "learning_rate": 2.7130959537798874e-08,
+      "loss": 0.0637,
+      "step": 17673
+    },
+    {
+      "epoch": 2.8635774465327284,
+      "grad_norm": 0.9914158582687378,
+      "learning_rate": 2.7066746408944968e-08,
+      "loss": 0.0621,
+      "step": 17674
+    },
+    {
+      "epoch": 2.8637394685677253,
+      "grad_norm": 0.9497270584106445,
+      "learning_rate": 2.700260894586415e-08,
+      "loss": 0.0555,
+      "step": 17675
+    },
+    {
+      "epoch": 2.863901490602722,
+      "grad_norm": 0.8510124087333679,
+      "learning_rate": 2.6938547150518746e-08,
+      "loss": 0.0556,
+      "step": 17676
+    },
+    {
+      "epoch": 2.8640635126377187,
+      "grad_norm": 0.8168439269065857,
+      "learning_rate": 2.6874561024869407e-08,
+      "loss": 0.0514,
+      "step": 17677
+    },
+    {
+      "epoch": 2.8642255346727152,
+      "grad_norm": 0.8079927563667297,
+      "learning_rate": 2.6810650570873454e-08,
+      "loss": 0.0501,
+      "step": 17678
+    },
+    {
+      "epoch": 2.864387556707712,
+      "grad_norm": 0.9214690923690796,
+      "learning_rate": 2.6746815790486548e-08,
+      "loss": 0.061,
+      "step": 17679
+    },
+    {
+      "epoch": 2.864549578742709,
+      "grad_norm": 1.2173523902893066,
+      "learning_rate": 2.6683056685662122e-08,
+      "loss": 0.0676,
+      "step": 17680
+    },
+    {
+      "epoch": 2.8647116007777056,
+      "grad_norm": 0.9455307722091675,
+      "learning_rate": 2.6619373258350566e-08,
+      "loss": 0.0689,
+      "step": 17681
+    },
+    {
+      "epoch": 2.8648736228127025,
+      "grad_norm": 0.84743332862854,
+      "learning_rate": 2.6555765510500875e-08,
+      "loss": 0.0549,
+      "step": 17682
+    },
+    {
+      "epoch": 2.8650356448476995,
+      "grad_norm": 0.8518211245536804,
+      "learning_rate": 2.6492233444059267e-08,
+      "loss": 0.0595,
+      "step": 17683
+    },
+    {
+      "epoch": 2.865197666882696,
+      "grad_norm": 0.8848883509635925,
+      "learning_rate": 2.6428777060969468e-08,
+      "loss": 0.0615,
+      "step": 17684
+    },
+    {
+      "epoch": 2.865359688917693,
+      "grad_norm": 0.9524630308151245,
+      "learning_rate": 2.6365396363173256e-08,
+      "loss": 0.0634,
+      "step": 17685
+    },
+    {
+      "epoch": 2.86552171095269,
+      "grad_norm": 0.9205766916275024,
+      "learning_rate": 2.6302091352609637e-08,
+      "loss": 0.0672,
+      "step": 17686
+    },
+    {
+      "epoch": 2.8656837329876863,
+      "grad_norm": 0.821868360042572,
+      "learning_rate": 2.6238862031215672e-08,
+      "loss": 0.0562,
+      "step": 17687
+    },
+    {
+      "epoch": 2.8658457550226832,
+      "grad_norm": 0.9511677026748657,
+      "learning_rate": 2.617570840092648e-08,
+      "loss": 0.0646,
+      "step": 17688
+    },
+    {
+      "epoch": 2.8660077770576797,
+      "grad_norm": 0.9630586504936218,
+      "learning_rate": 2.611263046367385e-08,
+      "loss": 0.0589,
+      "step": 17689
+    },
+    {
+      "epoch": 2.8661697990926767,
+      "grad_norm": 0.8870193362236023,
+      "learning_rate": 2.604962822138818e-08,
+      "loss": 0.0544,
+      "step": 17690
+    },
+    {
+      "epoch": 2.866331821127673,
+      "grad_norm": 1.0679670572280884,
+      "learning_rate": 2.5986701675996816e-08,
+      "loss": 0.0634,
+      "step": 17691
+    },
+    {
+      "epoch": 2.86649384316267,
+      "grad_norm": 0.9083020091056824,
+      "learning_rate": 2.5923850829425723e-08,
+      "loss": 0.0639,
+      "step": 17692
+    },
+    {
+      "epoch": 2.866655865197667,
+      "grad_norm": 0.8310949802398682,
+      "learning_rate": 2.5861075683597526e-08,
+      "loss": 0.0558,
+      "step": 17693
+    },
+    {
+      "epoch": 2.8668178872326635,
+      "grad_norm": 0.956186830997467,
+      "learning_rate": 2.579837624043291e-08,
+      "loss": 0.0709,
+      "step": 17694
+    },
+    {
+      "epoch": 2.8669799092676604,
+      "grad_norm": 1.0341099500656128,
+      "learning_rate": 2.573575250185062e-08,
+      "loss": 0.0563,
+      "step": 17695
+    },
+    {
+      "epoch": 2.8671419313026574,
+      "grad_norm": 0.8617078065872192,
+      "learning_rate": 2.5673204469766898e-08,
+      "loss": 0.0621,
+      "step": 17696
+    },
+    {
+      "epoch": 2.867303953337654,
+      "grad_norm": 0.9676273465156555,
+      "learning_rate": 2.561073214609494e-08,
+      "loss": 0.0636,
+      "step": 17697
+    },
+    {
+      "epoch": 2.867465975372651,
+      "grad_norm": 0.9023331999778748,
+      "learning_rate": 2.5548335532747105e-08,
+      "loss": 0.0571,
+      "step": 17698
+    },
+    {
+      "epoch": 2.8676279974076473,
+      "grad_norm": 0.9330205321311951,
+      "learning_rate": 2.5486014631631862e-08,
+      "loss": 0.0592,
+      "step": 17699
+    },
+    {
+      "epoch": 2.8677900194426442,
+      "grad_norm": 1.1070448160171509,
+      "learning_rate": 2.5423769444656575e-08,
+      "loss": 0.0669,
+      "step": 17700
+    },
+    {
+      "epoch": 2.8679520414776407,
+      "grad_norm": 0.8804136514663696,
+      "learning_rate": 2.536159997372528e-08,
+      "loss": 0.0567,
+      "step": 17701
+    },
+    {
+      "epoch": 2.8681140635126376,
+      "grad_norm": 0.9040849208831787,
+      "learning_rate": 2.52995062207409e-08,
+      "loss": 0.0598,
+      "step": 17702
+    },
+    {
+      "epoch": 2.8682760855476346,
+      "grad_norm": 0.999211311340332,
+      "learning_rate": 2.5237488187602743e-08,
+      "loss": 0.0586,
+      "step": 17703
+    },
+    {
+      "epoch": 2.868438107582631,
+      "grad_norm": 0.9211670160293579,
+      "learning_rate": 2.517554587620874e-08,
+      "loss": 0.0648,
+      "step": 17704
+    },
+    {
+      "epoch": 2.868600129617628,
+      "grad_norm": 0.820000410079956,
+      "learning_rate": 2.511367928845404e-08,
+      "loss": 0.055,
+      "step": 17705
+    },
+    {
+      "epoch": 2.868762151652625,
+      "grad_norm": 0.8488945960998535,
+      "learning_rate": 2.5051888426231574e-08,
+      "loss": 0.0563,
+      "step": 17706
+    },
+    {
+      "epoch": 2.8689241736876214,
+      "grad_norm": 0.9143860936164856,
+      "learning_rate": 2.499017329143205e-08,
+      "loss": 0.0565,
+      "step": 17707
+    },
+    {
+      "epoch": 2.8690861957226184,
+      "grad_norm": 0.9244613647460938,
+      "learning_rate": 2.492853388594396e-08,
+      "loss": 0.0627,
+      "step": 17708
+    },
+    {
+      "epoch": 2.8692482177576153,
+      "grad_norm": 0.924808144569397,
+      "learning_rate": 2.486697021165302e-08,
+      "loss": 0.0617,
+      "step": 17709
+    },
+    {
+      "epoch": 2.869410239792612,
+      "grad_norm": 1.042020320892334,
+      "learning_rate": 2.480548227044327e-08,
+      "loss": 0.0593,
+      "step": 17710
+    },
+    {
+      "epoch": 2.8695722618276087,
+      "grad_norm": 1.0010639429092407,
+      "learning_rate": 2.4744070064195713e-08,
+      "loss": 0.0599,
+      "step": 17711
+    },
+    {
+      "epoch": 2.869734283862605,
+      "grad_norm": 1.0316102504730225,
+      "learning_rate": 2.4682733594789677e-08,
+      "loss": 0.0673,
+      "step": 17712
+    },
+    {
+      "epoch": 2.869896305897602,
+      "grad_norm": 1.0180681943893433,
+      "learning_rate": 2.4621472864101992e-08,
+      "loss": 0.0638,
+      "step": 17713
+    },
+    {
+      "epoch": 2.8700583279325986,
+      "grad_norm": 1.028444766998291,
+      "learning_rate": 2.4560287874006716e-08,
+      "loss": 0.0638,
+      "step": 17714
+    },
+    {
+      "epoch": 2.8702203499675956,
+      "grad_norm": 1.0102734565734863,
+      "learning_rate": 2.4499178626376243e-08,
+      "loss": 0.0612,
+      "step": 17715
+    },
+    {
+      "epoch": 2.8703823720025925,
+      "grad_norm": 0.8836684226989746,
+      "learning_rate": 2.443814512308018e-08,
+      "loss": 0.0563,
+      "step": 17716
+    },
+    {
+      "epoch": 2.870544394037589,
+      "grad_norm": 0.9856815934181213,
+      "learning_rate": 2.437718736598621e-08,
+      "loss": 0.0579,
+      "step": 17717
+    },
+    {
+      "epoch": 2.870706416072586,
+      "grad_norm": 0.9209667444229126,
+      "learning_rate": 2.431630535695978e-08,
+      "loss": 0.056,
+      "step": 17718
+    },
+    {
+      "epoch": 2.870868438107583,
+      "grad_norm": 0.9791855812072754,
+      "learning_rate": 2.4255499097863012e-08,
+      "loss": 0.0622,
+      "step": 17719
+    },
+    {
+      "epoch": 2.8710304601425793,
+      "grad_norm": 0.9314326047897339,
+      "learning_rate": 2.419476859055664e-08,
+      "loss": 0.0617,
+      "step": 17720
+    },
+    {
+      "epoch": 2.8711924821775763,
+      "grad_norm": 0.8641183376312256,
+      "learning_rate": 2.4134113836899455e-08,
+      "loss": 0.0557,
+      "step": 17721
+    },
+    {
+      "epoch": 2.8713545042125728,
+      "grad_norm": 0.8541434407234192,
+      "learning_rate": 2.4073534838746637e-08,
+      "loss": 0.0541,
+      "step": 17722
+    },
+    {
+      "epoch": 2.8715165262475697,
+      "grad_norm": 1.0766733884811401,
+      "learning_rate": 2.4013031597951985e-08,
+      "loss": 0.0597,
+      "step": 17723
+    },
+    {
+      "epoch": 2.871678548282566,
+      "grad_norm": 0.9844111204147339,
+      "learning_rate": 2.3952604116366795e-08,
+      "loss": 0.0637,
+      "step": 17724
+    },
+    {
+      "epoch": 2.871840570317563,
+      "grad_norm": 0.8357641696929932,
+      "learning_rate": 2.3892252395840143e-08,
+      "loss": 0.0598,
+      "step": 17725
+    },
+    {
+      "epoch": 2.87200259235256,
+      "grad_norm": 1.0578702688217163,
+      "learning_rate": 2.383197643821833e-08,
+      "loss": 0.0574,
+      "step": 17726
+    },
+    {
+      "epoch": 2.8721646143875565,
+      "grad_norm": 0.8286598920822144,
+      "learning_rate": 2.377177624534599e-08,
+      "loss": 0.0589,
+      "step": 17727
+    },
+    {
+      "epoch": 2.8723266364225535,
+      "grad_norm": 0.890817403793335,
+      "learning_rate": 2.3711651819064984e-08,
+      "loss": 0.0612,
+      "step": 17728
+    },
+    {
+      "epoch": 2.8724886584575504,
+      "grad_norm": 0.8068243265151978,
+      "learning_rate": 2.3651603161214677e-08,
+      "loss": 0.0547,
+      "step": 17729
+    },
+    {
+      "epoch": 2.872650680492547,
+      "grad_norm": 0.8931058049201965,
+      "learning_rate": 2.359163027363276e-08,
+      "loss": 0.0582,
+      "step": 17730
+    },
+    {
+      "epoch": 2.872812702527544,
+      "grad_norm": 0.9487268328666687,
+      "learning_rate": 2.3531733158154157e-08,
+      "loss": 0.0684,
+      "step": 17731
+    },
+    {
+      "epoch": 2.8729747245625408,
+      "grad_norm": 0.8844464421272278,
+      "learning_rate": 2.3471911816611846e-08,
+      "loss": 0.0576,
+      "step": 17732
+    },
+    {
+      "epoch": 2.8731367465975373,
+      "grad_norm": 0.7697759866714478,
+      "learning_rate": 2.3412166250835756e-08,
+      "loss": 0.0475,
+      "step": 17733
+    },
+    {
+      "epoch": 2.8732987686325338,
+      "grad_norm": 1.038557767868042,
+      "learning_rate": 2.335249646265414e-08,
+      "loss": 0.0675,
+      "step": 17734
+    },
+    {
+      "epoch": 2.8734607906675307,
+      "grad_norm": 0.8407655954360962,
+      "learning_rate": 2.3292902453892485e-08,
+      "loss": 0.0579,
+      "step": 17735
+    },
+    {
+      "epoch": 2.8736228127025276,
+      "grad_norm": 0.940645694732666,
+      "learning_rate": 2.3233384226375167e-08,
+      "loss": 0.0677,
+      "step": 17736
+    },
+    {
+      "epoch": 2.873784834737524,
+      "grad_norm": 0.9915009140968323,
+      "learning_rate": 2.3173941781922114e-08,
+      "loss": 0.0645,
+      "step": 17737
+    },
+    {
+      "epoch": 2.873946856772521,
+      "grad_norm": 1.1203358173370361,
+      "learning_rate": 2.311457512235271e-08,
+      "loss": 0.0607,
+      "step": 17738
+    },
+    {
+      "epoch": 2.874108878807518,
+      "grad_norm": 0.9407758116722107,
+      "learning_rate": 2.305528424948328e-08,
+      "loss": 0.0555,
+      "step": 17739
+    },
+    {
+      "epoch": 2.8742709008425145,
+      "grad_norm": 1.1099528074264526,
+      "learning_rate": 2.2996069165128198e-08,
+      "loss": 0.0638,
+      "step": 17740
+    },
+    {
+      "epoch": 2.8744329228775114,
+      "grad_norm": 0.8714383244514465,
+      "learning_rate": 2.2936929871099356e-08,
+      "loss": 0.0595,
+      "step": 17741
+    },
+    {
+      "epoch": 2.8745949449125083,
+      "grad_norm": 0.834579586982727,
+      "learning_rate": 2.2877866369205858e-08,
+      "loss": 0.0633,
+      "step": 17742
+    },
+    {
+      "epoch": 2.874756966947505,
+      "grad_norm": 0.9249007701873779,
+      "learning_rate": 2.281887866125515e-08,
+      "loss": 0.054,
+      "step": 17743
+    },
+    {
+      "epoch": 2.8749189889825018,
+      "grad_norm": 0.8720956444740295,
+      "learning_rate": 2.2759966749051897e-08,
+      "loss": 0.0561,
+      "step": 17744
+    },
+    {
+      "epoch": 2.8750810110174982,
+      "grad_norm": 1.0232727527618408,
+      "learning_rate": 2.2701130634399104e-08,
+      "loss": 0.063,
+      "step": 17745
+    },
+    {
+      "epoch": 2.875243033052495,
+      "grad_norm": 0.9553148746490479,
+      "learning_rate": 2.2642370319096718e-08,
+      "loss": 0.0691,
+      "step": 17746
+    },
+    {
+      "epoch": 2.8754050550874917,
+      "grad_norm": 0.820517361164093,
+      "learning_rate": 2.2583685804942746e-08,
+      "loss": 0.0578,
+      "step": 17747
+    },
+    {
+      "epoch": 2.8755670771224886,
+      "grad_norm": 0.9615774154663086,
+      "learning_rate": 2.2525077093732695e-08,
+      "loss": 0.0674,
+      "step": 17748
+    },
+    {
+      "epoch": 2.8757290991574855,
+      "grad_norm": 0.9988585114479065,
+      "learning_rate": 2.2466544187259852e-08,
+      "loss": 0.0686,
+      "step": 17749
+    },
+    {
+      "epoch": 2.875891121192482,
+      "grad_norm": 0.8221516013145447,
+      "learning_rate": 2.2408087087315567e-08,
+      "loss": 0.0566,
+      "step": 17750
+    },
+    {
+      "epoch": 2.876053143227479,
+      "grad_norm": 0.852265477180481,
+      "learning_rate": 2.234970579568785e-08,
+      "loss": 0.06,
+      "step": 17751
+    },
+    {
+      "epoch": 2.876215165262476,
+      "grad_norm": 0.9517872929573059,
+      "learning_rate": 2.2291400314163325e-08,
+      "loss": 0.0628,
+      "step": 17752
+    },
+    {
+      "epoch": 2.8763771872974724,
+      "grad_norm": 1.0439865589141846,
+      "learning_rate": 2.2233170644526126e-08,
+      "loss": 0.0619,
+      "step": 17753
+    },
+    {
+      "epoch": 2.8765392093324693,
+      "grad_norm": 0.8319445848464966,
+      "learning_rate": 2.217501678855788e-08,
+      "loss": 0.0548,
+      "step": 17754
+    },
+    {
+      "epoch": 2.8767012313674662,
+      "grad_norm": 0.9762924909591675,
+      "learning_rate": 2.211693874803772e-08,
+      "loss": 0.0599,
+      "step": 17755
+    },
+    {
+      "epoch": 2.8768632534024627,
+      "grad_norm": 0.831531822681427,
+      "learning_rate": 2.2058936524742835e-08,
+      "loss": 0.0537,
+      "step": 17756
+    },
+    {
+      "epoch": 2.8770252754374592,
+      "grad_norm": 0.8122191429138184,
+      "learning_rate": 2.2001010120448197e-08,
+      "loss": 0.0555,
+      "step": 17757
+    },
+    {
+      "epoch": 2.877187297472456,
+      "grad_norm": 1.2487119436264038,
+      "learning_rate": 2.1943159536925994e-08,
+      "loss": 0.0614,
+      "step": 17758
+    },
+    {
+      "epoch": 2.877349319507453,
+      "grad_norm": 1.0974977016448975,
+      "learning_rate": 2.1885384775946207e-08,
+      "loss": 0.0593,
+      "step": 17759
+    },
+    {
+      "epoch": 2.8775113415424496,
+      "grad_norm": 0.9182059168815613,
+      "learning_rate": 2.1827685839276856e-08,
+      "loss": 0.0623,
+      "step": 17760
+    },
+    {
+      "epoch": 2.8776733635774465,
+      "grad_norm": 1.0060678720474243,
+      "learning_rate": 2.177006272868293e-08,
+      "loss": 0.0623,
+      "step": 17761
+    },
+    {
+      "epoch": 2.8778353856124435,
+      "grad_norm": 0.8593238592147827,
+      "learning_rate": 2.1712515445928285e-08,
+      "loss": 0.0612,
+      "step": 17762
+    },
+    {
+      "epoch": 2.87799740764744,
+      "grad_norm": 0.8475682139396667,
+      "learning_rate": 2.1655043992773183e-08,
+      "loss": 0.0538,
+      "step": 17763
+    },
+    {
+      "epoch": 2.878159429682437,
+      "grad_norm": 1.3795852661132812,
+      "learning_rate": 2.159764837097622e-08,
+      "loss": 0.0681,
+      "step": 17764
+    },
+    {
+      "epoch": 2.878321451717434,
+      "grad_norm": 1.0932248830795288,
+      "learning_rate": 2.1540328582293767e-08,
+      "loss": 0.0699,
+      "step": 17765
+    },
+    {
+      "epoch": 2.8784834737524303,
+      "grad_norm": 0.8664324879646301,
+      "learning_rate": 2.1483084628479145e-08,
+      "loss": 0.0559,
+      "step": 17766
+    },
+    {
+      "epoch": 2.8786454957874272,
+      "grad_norm": 0.92646723985672,
+      "learning_rate": 2.142591651128456e-08,
+      "loss": 0.0559,
+      "step": 17767
+    },
+    {
+      "epoch": 2.8788075178224237,
+      "grad_norm": 0.7677503228187561,
+      "learning_rate": 2.1368824232458618e-08,
+      "loss": 0.0519,
+      "step": 17768
+    },
+    {
+      "epoch": 2.8789695398574207,
+      "grad_norm": 0.8314093351364136,
+      "learning_rate": 2.1311807793748805e-08,
+      "loss": 0.0546,
+      "step": 17769
+    },
+    {
+      "epoch": 2.879131561892417,
+      "grad_norm": 0.971030056476593,
+      "learning_rate": 2.125486719689929e-08,
+      "loss": 0.0632,
+      "step": 17770
+    },
+    {
+      "epoch": 2.879293583927414,
+      "grad_norm": 0.8812588453292847,
+      "learning_rate": 2.1198002443652276e-08,
+      "loss": 0.0607,
+      "step": 17771
+    },
+    {
+      "epoch": 2.879455605962411,
+      "grad_norm": 1.1112418174743652,
+      "learning_rate": 2.1141213535747772e-08,
+      "loss": 0.0659,
+      "step": 17772
+    },
+    {
+      "epoch": 2.8796176279974075,
+      "grad_norm": 0.976514995098114,
+      "learning_rate": 2.1084500474923554e-08,
+      "loss": 0.0672,
+      "step": 17773
+    },
+    {
+      "epoch": 2.8797796500324044,
+      "grad_norm": 0.9304304122924805,
+      "learning_rate": 2.1027863262914617e-08,
+      "loss": 0.0598,
+      "step": 17774
+    },
+    {
+      "epoch": 2.8799416720674014,
+      "grad_norm": 1.0370122194290161,
+      "learning_rate": 2.0971301901454023e-08,
+      "loss": 0.0622,
+      "step": 17775
+    },
+    {
+      "epoch": 2.880103694102398,
+      "grad_norm": 0.7911035418510437,
+      "learning_rate": 2.0914816392272608e-08,
+      "loss": 0.057,
+      "step": 17776
+    },
+    {
+      "epoch": 2.880265716137395,
+      "grad_norm": 0.8377315998077393,
+      "learning_rate": 2.0858406737098435e-08,
+      "loss": 0.0523,
+      "step": 17777
+    },
+    {
+      "epoch": 2.8804277381723913,
+      "grad_norm": 0.895128071308136,
+      "learning_rate": 2.0802072937657624e-08,
+      "loss": 0.0597,
+      "step": 17778
+    },
+    {
+      "epoch": 2.880589760207388,
+      "grad_norm": 0.9647876620292664,
+      "learning_rate": 2.0745814995673796e-08,
+      "loss": 0.0653,
+      "step": 17779
+    },
+    {
+      "epoch": 2.8807517822423847,
+      "grad_norm": 0.8965823650360107,
+      "learning_rate": 2.068963291286863e-08,
+      "loss": 0.0554,
+      "step": 17780
+    },
+    {
+      "epoch": 2.8809138042773816,
+      "grad_norm": 1.0152802467346191,
+      "learning_rate": 2.0633526690960747e-08,
+      "loss": 0.0599,
+      "step": 17781
+    },
+    {
+      "epoch": 2.8810758263123786,
+      "grad_norm": 0.886622428894043,
+      "learning_rate": 2.0577496331666837e-08,
+      "loss": 0.0631,
+      "step": 17782
+    },
+    {
+      "epoch": 2.881237848347375,
+      "grad_norm": 0.943684995174408,
+      "learning_rate": 2.052154183670163e-08,
+      "loss": 0.06,
+      "step": 17783
+    },
+    {
+      "epoch": 2.881399870382372,
+      "grad_norm": 0.9235215783119202,
+      "learning_rate": 2.046566320777682e-08,
+      "loss": 0.0594,
+      "step": 17784
+    },
+    {
+      "epoch": 2.881561892417369,
+      "grad_norm": 0.9773284196853638,
+      "learning_rate": 2.04098604466027e-08,
+      "loss": 0.0646,
+      "step": 17785
+    },
+    {
+      "epoch": 2.8817239144523654,
+      "grad_norm": 0.921341061592102,
+      "learning_rate": 2.0354133554885967e-08,
+      "loss": 0.059,
+      "step": 17786
+    },
+    {
+      "epoch": 2.8818859364873624,
+      "grad_norm": 0.9354360699653625,
+      "learning_rate": 2.0298482534332198e-08,
+      "loss": 0.0553,
+      "step": 17787
+    },
+    {
+      "epoch": 2.8820479585223593,
+      "grad_norm": 0.8430424332618713,
+      "learning_rate": 2.0242907386644195e-08,
+      "loss": 0.0548,
+      "step": 17788
+    },
+    {
+      "epoch": 2.8822099805573558,
+      "grad_norm": 0.8397956490516663,
+      "learning_rate": 2.0187408113522266e-08,
+      "loss": 0.054,
+      "step": 17789
+    },
+    {
+      "epoch": 2.8823720025923527,
+      "grad_norm": 0.967339813709259,
+      "learning_rate": 2.0131984716664776e-08,
+      "loss": 0.0617,
+      "step": 17790
+    },
+    {
+      "epoch": 2.882534024627349,
+      "grad_norm": 1.0861202478408813,
+      "learning_rate": 2.0076637197767312e-08,
+      "loss": 0.0653,
+      "step": 17791
+    },
+    {
+      "epoch": 2.882696046662346,
+      "grad_norm": 0.9953954219818115,
+      "learning_rate": 2.002136555852352e-08,
+      "loss": 0.0634,
+      "step": 17792
+    },
+    {
+      "epoch": 2.8828580686973426,
+      "grad_norm": 0.7926887273788452,
+      "learning_rate": 1.996616980062427e-08,
+      "loss": 0.0497,
+      "step": 17793
+    },
+    {
+      "epoch": 2.8830200907323396,
+      "grad_norm": 1.0522444248199463,
+      "learning_rate": 1.9911049925758765e-08,
+      "loss": 0.0647,
+      "step": 17794
+    },
+    {
+      "epoch": 2.8831821127673365,
+      "grad_norm": 0.9192836284637451,
+      "learning_rate": 1.9856005935613708e-08,
+      "loss": 0.0607,
+      "step": 17795
+    },
+    {
+      "epoch": 2.883344134802333,
+      "grad_norm": 0.9567740559577942,
+      "learning_rate": 1.9801037831872482e-08,
+      "loss": 0.0569,
+      "step": 17796
+    },
+    {
+      "epoch": 2.88350615683733,
+      "grad_norm": 0.8763161897659302,
+      "learning_rate": 1.9746145616217905e-08,
+      "loss": 0.0535,
+      "step": 17797
+    },
+    {
+      "epoch": 2.883668178872327,
+      "grad_norm": 0.9974742531776428,
+      "learning_rate": 1.9691329290329185e-08,
+      "loss": 0.0576,
+      "step": 17798
+    },
+    {
+      "epoch": 2.8838302009073233,
+      "grad_norm": 0.915104866027832,
+      "learning_rate": 1.9636588855883598e-08,
+      "loss": 0.062,
+      "step": 17799
+    },
+    {
+      "epoch": 2.8839922229423203,
+      "grad_norm": 0.8486490249633789,
+      "learning_rate": 1.958192431455591e-08,
+      "loss": 0.059,
+      "step": 17800
+    },
+    {
+      "epoch": 2.8841542449773168,
+      "grad_norm": 0.944681704044342,
+      "learning_rate": 1.9527335668018954e-08,
+      "loss": 0.0648,
+      "step": 17801
+    },
+    {
+      "epoch": 2.8843162670123137,
+      "grad_norm": 0.9253272414207458,
+      "learning_rate": 1.9472822917942778e-08,
+      "loss": 0.0609,
+      "step": 17802
+    },
+    {
+      "epoch": 2.88447828904731,
+      "grad_norm": 0.9688064455986023,
+      "learning_rate": 1.9418386065995222e-08,
+      "loss": 0.0623,
+      "step": 17803
+    },
+    {
+      "epoch": 2.884640311082307,
+      "grad_norm": 1.0113117694854736,
+      "learning_rate": 1.9364025113842444e-08,
+      "loss": 0.0597,
+      "step": 17804
+    },
+    {
+      "epoch": 2.884802333117304,
+      "grad_norm": 0.8749908804893494,
+      "learning_rate": 1.9309740063147566e-08,
+      "loss": 0.0587,
+      "step": 17805
+    },
+    {
+      "epoch": 2.8849643551523005,
+      "grad_norm": 1.0198384523391724,
+      "learning_rate": 1.9255530915571197e-08,
+      "loss": 0.0611,
+      "step": 17806
+    },
+    {
+      "epoch": 2.8851263771872975,
+      "grad_norm": 0.9575091600418091,
+      "learning_rate": 1.920139767277257e-08,
+      "loss": 0.061,
+      "step": 17807
+    },
+    {
+      "epoch": 2.8852883992222944,
+      "grad_norm": 0.9491891264915466,
+      "learning_rate": 1.9147340336407584e-08,
+      "loss": 0.0618,
+      "step": 17808
+    },
+    {
+      "epoch": 2.885450421257291,
+      "grad_norm": 0.9675207734107971,
+      "learning_rate": 1.9093358908130743e-08,
+      "loss": 0.0675,
+      "step": 17809
+    },
+    {
+      "epoch": 2.885612443292288,
+      "grad_norm": 0.8479520082473755,
+      "learning_rate": 1.9039453389592954e-08,
+      "loss": 0.0612,
+      "step": 17810
+    },
+    {
+      "epoch": 2.8857744653272848,
+      "grad_norm": 1.0039427280426025,
+      "learning_rate": 1.8985623782444284e-08,
+      "loss": 0.064,
+      "step": 17811
+    },
+    {
+      "epoch": 2.8859364873622813,
+      "grad_norm": 0.9611213207244873,
+      "learning_rate": 1.893187008833175e-08,
+      "loss": 0.0603,
+      "step": 17812
+    },
+    {
+      "epoch": 2.886098509397278,
+      "grad_norm": 0.8589776158332825,
+      "learning_rate": 1.8878192308899867e-08,
+      "loss": 0.0586,
+      "step": 17813
+    },
+    {
+      "epoch": 2.8862605314322747,
+      "grad_norm": 0.8623048067092896,
+      "learning_rate": 1.8824590445790935e-08,
+      "loss": 0.0626,
+      "step": 17814
+    },
+    {
+      "epoch": 2.8864225534672716,
+      "grad_norm": 0.9948340654373169,
+      "learning_rate": 1.877106450064531e-08,
+      "loss": 0.0649,
+      "step": 17815
+    },
+    {
+      "epoch": 2.886584575502268,
+      "grad_norm": 0.9719632863998413,
+      "learning_rate": 1.871761447510084e-08,
+      "loss": 0.0546,
+      "step": 17816
+    },
+    {
+      "epoch": 2.886746597537265,
+      "grad_norm": 0.8713729977607727,
+      "learning_rate": 1.866424037079234e-08,
+      "loss": 0.0602,
+      "step": 17817
+    },
+    {
+      "epoch": 2.886908619572262,
+      "grad_norm": 0.8655098676681519,
+      "learning_rate": 1.8610942189353777e-08,
+      "loss": 0.0556,
+      "step": 17818
+    },
+    {
+      "epoch": 2.8870706416072585,
+      "grad_norm": 0.830318808555603,
+      "learning_rate": 1.855771993241523e-08,
+      "loss": 0.0535,
+      "step": 17819
+    },
+    {
+      "epoch": 2.8872326636422554,
+      "grad_norm": 0.8426806926727295,
+      "learning_rate": 1.850457360160568e-08,
+      "loss": 0.0574,
+      "step": 17820
+    },
+    {
+      "epoch": 2.8873946856772523,
+      "grad_norm": 0.8310089111328125,
+      "learning_rate": 1.8451503198551047e-08,
+      "loss": 0.0546,
+      "step": 17821
+    },
+    {
+      "epoch": 2.887556707712249,
+      "grad_norm": 0.8731127381324768,
+      "learning_rate": 1.839850872487503e-08,
+      "loss": 0.0574,
+      "step": 17822
+    },
+    {
+      "epoch": 2.8877187297472457,
+      "grad_norm": 0.9887226819992065,
+      "learning_rate": 1.834559018219939e-08,
+      "loss": 0.0567,
+      "step": 17823
+    },
+    {
+      "epoch": 2.8878807517822422,
+      "grad_norm": 0.9463251829147339,
+      "learning_rate": 1.829274757214339e-08,
+      "loss": 0.0628,
+      "step": 17824
+    },
+    {
+      "epoch": 2.888042773817239,
+      "grad_norm": 0.8988006114959717,
+      "learning_rate": 1.8239980896323505e-08,
+      "loss": 0.0581,
+      "step": 17825
+    },
+    {
+      "epoch": 2.8882047958522357,
+      "grad_norm": 0.9556702375411987,
+      "learning_rate": 1.8187290156354565e-08,
+      "loss": 0.0587,
+      "step": 17826
+    },
+    {
+      "epoch": 2.8883668178872326,
+      "grad_norm": 0.8816114068031311,
+      "learning_rate": 1.8134675353848608e-08,
+      "loss": 0.0593,
+      "step": 17827
+    },
+    {
+      "epoch": 2.8885288399222295,
+      "grad_norm": 0.856730043888092,
+      "learning_rate": 1.808213649041546e-08,
+      "loss": 0.0554,
+      "step": 17828
+    },
+    {
+      "epoch": 2.888690861957226,
+      "grad_norm": 0.8481936454772949,
+      "learning_rate": 1.8029673567662997e-08,
+      "loss": 0.0543,
+      "step": 17829
+    },
+    {
+      "epoch": 2.888852883992223,
+      "grad_norm": 1.0299345254898071,
+      "learning_rate": 1.7977286587196053e-08,
+      "loss": 0.0651,
+      "step": 17830
+    },
+    {
+      "epoch": 2.88901490602722,
+      "grad_norm": 0.9097673296928406,
+      "learning_rate": 1.792497555061806e-08,
+      "loss": 0.0626,
+      "step": 17831
+    },
+    {
+      "epoch": 2.8891769280622164,
+      "grad_norm": 0.9130776524543762,
+      "learning_rate": 1.7872740459529135e-08,
+      "loss": 0.0627,
+      "step": 17832
+    },
+    {
+      "epoch": 2.8893389500972133,
+      "grad_norm": 0.9484906792640686,
+      "learning_rate": 1.7820581315527717e-08,
+      "loss": 0.0658,
+      "step": 17833
+    },
+    {
+      "epoch": 2.8895009721322102,
+      "grad_norm": 0.8442859649658203,
+      "learning_rate": 1.7768498120209755e-08,
+      "loss": 0.0559,
+      "step": 17834
+    },
+    {
+      "epoch": 2.8896629941672067,
+      "grad_norm": 0.9843007326126099,
+      "learning_rate": 1.771649087516897e-08,
+      "loss": 0.0648,
+      "step": 17835
+    },
+    {
+      "epoch": 2.8898250162022032,
+      "grad_norm": 0.8426430225372314,
+      "learning_rate": 1.766455958199631e-08,
+      "loss": 0.0564,
+      "step": 17836
+    },
+    {
+      "epoch": 2.8899870382372,
+      "grad_norm": 0.8461860418319702,
+      "learning_rate": 1.7612704242281342e-08,
+      "loss": 0.0614,
+      "step": 17837
+    },
+    {
+      "epoch": 2.890149060272197,
+      "grad_norm": 1.05631422996521,
+      "learning_rate": 1.7560924857610016e-08,
+      "loss": 0.064,
+      "step": 17838
+    },
+    {
+      "epoch": 2.8903110823071936,
+      "grad_norm": 0.8737570643424988,
+      "learning_rate": 1.750922142956718e-08,
+      "loss": 0.0607,
+      "step": 17839
+    },
+    {
+      "epoch": 2.8904731043421905,
+      "grad_norm": 0.8831677436828613,
+      "learning_rate": 1.745759395973462e-08,
+      "loss": 0.0525,
+      "step": 17840
+    },
+    {
+      "epoch": 2.8906351263771874,
+      "grad_norm": 0.9918739199638367,
+      "learning_rate": 1.7406042449691907e-08,
+      "loss": 0.0665,
+      "step": 17841
+    },
+    {
+      "epoch": 2.890797148412184,
+      "grad_norm": 0.8606711030006409,
+      "learning_rate": 1.7354566901016944e-08,
+      "loss": 0.0596,
+      "step": 17842
+    },
+    {
+      "epoch": 2.890959170447181,
+      "grad_norm": 0.8856733441352844,
+      "learning_rate": 1.730316731528403e-08,
+      "loss": 0.0629,
+      "step": 17843
+    },
+    {
+      "epoch": 2.891121192482178,
+      "grad_norm": 1.0403659343719482,
+      "learning_rate": 1.7251843694066074e-08,
+      "loss": 0.0601,
+      "step": 17844
+    },
+    {
+      "epoch": 2.8912832145171743,
+      "grad_norm": 0.8173930048942566,
+      "learning_rate": 1.7200596038934038e-08,
+      "loss": 0.0585,
+      "step": 17845
+    },
+    {
+      "epoch": 2.8914452365521712,
+      "grad_norm": 0.8533201813697815,
+      "learning_rate": 1.7149424351455003e-08,
+      "loss": 0.0593,
+      "step": 17846
+    },
+    {
+      "epoch": 2.8916072585871677,
+      "grad_norm": 0.941921591758728,
+      "learning_rate": 1.7098328633195493e-08,
+      "loss": 0.0658,
+      "step": 17847
+    },
+    {
+      "epoch": 2.8917692806221647,
+      "grad_norm": 1.0077810287475586,
+      "learning_rate": 1.7047308885718427e-08,
+      "loss": 0.0636,
+      "step": 17848
+    },
+    {
+      "epoch": 2.891931302657161,
+      "grad_norm": 1.1122187376022339,
+      "learning_rate": 1.6996365110585332e-08,
+      "loss": 0.0593,
+      "step": 17849
+    },
+    {
+      "epoch": 2.892093324692158,
+      "grad_norm": 1.093993902206421,
+      "learning_rate": 1.694549730935441e-08,
+      "loss": 0.073,
+      "step": 17850
+    },
+    {
+      "epoch": 2.892255346727155,
+      "grad_norm": 0.9843876957893372,
+      "learning_rate": 1.6894705483582464e-08,
+      "loss": 0.0657,
+      "step": 17851
+    },
+    {
+      "epoch": 2.8924173687621515,
+      "grad_norm": 0.9816562533378601,
+      "learning_rate": 1.684398963482381e-08,
+      "loss": 0.0609,
+      "step": 17852
+    },
+    {
+      "epoch": 2.8925793907971484,
+      "grad_norm": 1.04456627368927,
+      "learning_rate": 1.6793349764629707e-08,
+      "loss": 0.0619,
+      "step": 17853
+    },
+    {
+      "epoch": 2.8927414128321454,
+      "grad_norm": 0.8714144825935364,
+      "learning_rate": 1.674278587454975e-08,
+      "loss": 0.0546,
+      "step": 17854
+    },
+    {
+      "epoch": 2.892903434867142,
+      "grad_norm": 0.9265003204345703,
+      "learning_rate": 1.669229796613131e-08,
+      "loss": 0.0642,
+      "step": 17855
+    },
+    {
+      "epoch": 2.893065456902139,
+      "grad_norm": 0.916529655456543,
+      "learning_rate": 1.6641886040919263e-08,
+      "loss": 0.0613,
+      "step": 17856
+    },
+    {
+      "epoch": 2.8932274789371357,
+      "grad_norm": 0.8931515216827393,
+      "learning_rate": 1.659155010045571e-08,
+      "loss": 0.0584,
+      "step": 17857
+    },
+    {
+      "epoch": 2.893389500972132,
+      "grad_norm": 0.9384124279022217,
+      "learning_rate": 1.654129014628081e-08,
+      "loss": 0.0612,
+      "step": 17858
+    },
+    {
+      "epoch": 2.8935515230071287,
+      "grad_norm": 1.1038861274719238,
+      "learning_rate": 1.649110617993277e-08,
+      "loss": 0.0643,
+      "step": 17859
+    },
+    {
+      "epoch": 2.8937135450421256,
+      "grad_norm": 0.8668294548988342,
+      "learning_rate": 1.6440998202947034e-08,
+      "loss": 0.0565,
+      "step": 17860
+    },
+    {
+      "epoch": 2.8938755670771226,
+      "grad_norm": 0.8724477291107178,
+      "learning_rate": 1.639096621685654e-08,
+      "loss": 0.0577,
+      "step": 17861
+    },
+    {
+      "epoch": 2.894037589112119,
+      "grad_norm": 0.786655068397522,
+      "learning_rate": 1.634101022319229e-08,
+      "loss": 0.0581,
+      "step": 17862
+    },
+    {
+      "epoch": 2.894199611147116,
+      "grad_norm": 0.9810567498207092,
+      "learning_rate": 1.6291130223482498e-08,
+      "loss": 0.0551,
+      "step": 17863
+    },
+    {
+      "epoch": 2.894361633182113,
+      "grad_norm": 1.0155911445617676,
+      "learning_rate": 1.6241326219254006e-08,
+      "loss": 0.0584,
+      "step": 17864
+    },
+    {
+      "epoch": 2.8945236552171094,
+      "grad_norm": 1.0140700340270996,
+      "learning_rate": 1.6191598212030314e-08,
+      "loss": 0.0612,
+      "step": 17865
+    },
+    {
+      "epoch": 2.8946856772521063,
+      "grad_norm": 0.9874489903450012,
+      "learning_rate": 1.6141946203332703e-08,
+      "loss": 0.0627,
+      "step": 17866
+    },
+    {
+      "epoch": 2.8948476992871033,
+      "grad_norm": 0.938338041305542,
+      "learning_rate": 1.609237019468107e-08,
+      "loss": 0.0587,
+      "step": 17867
+    },
+    {
+      "epoch": 2.8950097213220998,
+      "grad_norm": 0.9551253318786621,
+      "learning_rate": 1.6042870187591985e-08,
+      "loss": 0.066,
+      "step": 17868
+    },
+    {
+      "epoch": 2.8951717433570967,
+      "grad_norm": 1.030510663986206,
+      "learning_rate": 1.5993446183579786e-08,
+      "loss": 0.0592,
+      "step": 17869
+    },
+    {
+      "epoch": 2.895333765392093,
+      "grad_norm": 0.9057535529136658,
+      "learning_rate": 1.5944098184156876e-08,
+      "loss": 0.0584,
+      "step": 17870
+    },
+    {
+      "epoch": 2.89549578742709,
+      "grad_norm": 0.8612263798713684,
+      "learning_rate": 1.5894826190833712e-08,
+      "loss": 0.0557,
+      "step": 17871
+    },
+    {
+      "epoch": 2.8956578094620866,
+      "grad_norm": 0.784995973110199,
+      "learning_rate": 1.5845630205117147e-08,
+      "loss": 0.0542,
+      "step": 17872
+    },
+    {
+      "epoch": 2.8958198314970836,
+      "grad_norm": 0.9127455353736877,
+      "learning_rate": 1.579651022851264e-08,
+      "loss": 0.0654,
+      "step": 17873
+    },
+    {
+      "epoch": 2.8959818535320805,
+      "grad_norm": 0.9822579026222229,
+      "learning_rate": 1.5747466262523438e-08,
+      "loss": 0.0597,
+      "step": 17874
+    },
+    {
+      "epoch": 2.896143875567077,
+      "grad_norm": 0.833354651927948,
+      "learning_rate": 1.569849830865e-08,
+      "loss": 0.0518,
+      "step": 17875
+    },
+    {
+      "epoch": 2.896305897602074,
+      "grad_norm": 0.9307559132575989,
+      "learning_rate": 1.5649606368390578e-08,
+      "loss": 0.0571,
+      "step": 17876
+    },
+    {
+      "epoch": 2.896467919637071,
+      "grad_norm": 0.8814070820808411,
+      "learning_rate": 1.56007904432412e-08,
+      "loss": 0.0577,
+      "step": 17877
+    },
+    {
+      "epoch": 2.8966299416720673,
+      "grad_norm": 0.9398168325424194,
+      "learning_rate": 1.5552050534695383e-08,
+      "loss": 0.0579,
+      "step": 17878
+    },
+    {
+      "epoch": 2.8967919637070643,
+      "grad_norm": 0.8523743152618408,
+      "learning_rate": 1.5503386644244724e-08,
+      "loss": 0.0567,
+      "step": 17879
+    },
+    {
+      "epoch": 2.8969539857420608,
+      "grad_norm": 0.9002468585968018,
+      "learning_rate": 1.5454798773378023e-08,
+      "loss": 0.0628,
+      "step": 17880
+    },
+    {
+      "epoch": 2.8971160077770577,
+      "grad_norm": 1.3708131313323975,
+      "learning_rate": 1.5406286923582148e-08,
+      "loss": 0.0594,
+      "step": 17881
+    },
+    {
+      "epoch": 2.897278029812054,
+      "grad_norm": 0.9118489027023315,
+      "learning_rate": 1.5357851096340915e-08,
+      "loss": 0.0566,
+      "step": 17882
+    },
+    {
+      "epoch": 2.897440051847051,
+      "grad_norm": 0.9221370816230774,
+      "learning_rate": 1.5309491293137026e-08,
+      "loss": 0.0616,
+      "step": 17883
+    },
+    {
+      "epoch": 2.897602073882048,
+      "grad_norm": 0.9659370183944702,
+      "learning_rate": 1.526120751544985e-08,
+      "loss": 0.0584,
+      "step": 17884
+    },
+    {
+      "epoch": 2.8977640959170445,
+      "grad_norm": 1.0934197902679443,
+      "learning_rate": 1.5212999764756543e-08,
+      "loss": 0.0669,
+      "step": 17885
+    },
+    {
+      "epoch": 2.8979261179520415,
+      "grad_norm": 0.854312539100647,
+      "learning_rate": 1.5164868042532864e-08,
+      "loss": 0.0521,
+      "step": 17886
+    },
+    {
+      "epoch": 2.8980881399870384,
+      "grad_norm": 1.0536237955093384,
+      "learning_rate": 1.5116812350250422e-08,
+      "loss": 0.0606,
+      "step": 17887
+    },
+    {
+      "epoch": 2.898250162022035,
+      "grad_norm": 0.9276726245880127,
+      "learning_rate": 1.506883268938053e-08,
+      "loss": 0.0588,
+      "step": 17888
+    },
+    {
+      "epoch": 2.898412184057032,
+      "grad_norm": 0.8733401894569397,
+      "learning_rate": 1.5020929061391188e-08,
+      "loss": 0.0603,
+      "step": 17889
+    },
+    {
+      "epoch": 2.8985742060920288,
+      "grad_norm": 0.8288245797157288,
+      "learning_rate": 1.4973101467747608e-08,
+      "loss": 0.0547,
+      "step": 17890
+    },
+    {
+      "epoch": 2.8987362281270252,
+      "grad_norm": 0.9966906905174255,
+      "learning_rate": 1.4925349909913346e-08,
+      "loss": 0.0615,
+      "step": 17891
+    },
+    {
+      "epoch": 2.898898250162022,
+      "grad_norm": 0.9769363403320312,
+      "learning_rate": 1.4877674389349728e-08,
+      "loss": 0.0607,
+      "step": 17892
+    },
+    {
+      "epoch": 2.8990602721970187,
+      "grad_norm": 0.8618332743644714,
+      "learning_rate": 1.4830074907515313e-08,
+      "loss": 0.0596,
+      "step": 17893
+    },
+    {
+      "epoch": 2.8992222942320156,
+      "grad_norm": 0.819103479385376,
+      "learning_rate": 1.4782551465866713e-08,
+      "loss": 0.0527,
+      "step": 17894
+    },
+    {
+      "epoch": 2.899384316267012,
+      "grad_norm": 0.8623654246330261,
+      "learning_rate": 1.4735104065858042e-08,
+      "loss": 0.0623,
+      "step": 17895
+    },
+    {
+      "epoch": 2.899546338302009,
+      "grad_norm": 0.8684125542640686,
+      "learning_rate": 1.4687732708940916e-08,
+      "loss": 0.0539,
+      "step": 17896
+    },
+    {
+      "epoch": 2.899708360337006,
+      "grad_norm": 0.8661196827888489,
+      "learning_rate": 1.4640437396564733e-08,
+      "loss": 0.056,
+      "step": 17897
+    },
+    {
+      "epoch": 2.8998703823720025,
+      "grad_norm": 0.9082359671592712,
+      "learning_rate": 1.4593218130176668e-08,
+      "loss": 0.0535,
+      "step": 17898
+    },
+    {
+      "epoch": 2.9000324044069994,
+      "grad_norm": 0.8829771876335144,
+      "learning_rate": 1.4546074911221675e-08,
+      "loss": 0.0582,
+      "step": 17899
+    },
+    {
+      "epoch": 2.9001944264419963,
+      "grad_norm": 1.0028576850891113,
+      "learning_rate": 1.4499007741141934e-08,
+      "loss": 0.063,
+      "step": 17900
+    },
+    {
+      "epoch": 2.900356448476993,
+      "grad_norm": 0.8748372793197632,
+      "learning_rate": 1.4452016621377961e-08,
+      "loss": 0.0625,
+      "step": 17901
+    },
+    {
+      "epoch": 2.9005184705119897,
+      "grad_norm": 0.986389696598053,
+      "learning_rate": 1.4405101553367218e-08,
+      "loss": 0.0616,
+      "step": 17902
+    },
+    {
+      "epoch": 2.9006804925469862,
+      "grad_norm": 1.0490912199020386,
+      "learning_rate": 1.4358262538545498e-08,
+      "loss": 0.058,
+      "step": 17903
+    },
+    {
+      "epoch": 2.900842514581983,
+      "grad_norm": 0.8131743669509888,
+      "learning_rate": 1.4311499578345821e-08,
+      "loss": 0.0559,
+      "step": 17904
+    },
+    {
+      "epoch": 2.9010045366169797,
+      "grad_norm": 0.8198305368423462,
+      "learning_rate": 1.4264812674198714e-08,
+      "loss": 0.0563,
+      "step": 17905
+    },
+    {
+      "epoch": 2.9011665586519766,
+      "grad_norm": 0.8010324239730835,
+      "learning_rate": 1.421820182753303e-08,
+      "loss": 0.0547,
+      "step": 17906
+    },
+    {
+      "epoch": 2.9013285806869735,
+      "grad_norm": 1.2426363229751587,
+      "learning_rate": 1.4171667039775128e-08,
+      "loss": 0.0607,
+      "step": 17907
+    },
+    {
+      "epoch": 2.90149060272197,
+      "grad_norm": 0.8623930811882019,
+      "learning_rate": 1.4125208312348593e-08,
+      "loss": 0.0633,
+      "step": 17908
+    },
+    {
+      "epoch": 2.901652624756967,
+      "grad_norm": 1.1092482805252075,
+      "learning_rate": 1.4078825646675065e-08,
+      "loss": 0.069,
+      "step": 17909
+    },
+    {
+      "epoch": 2.901814646791964,
+      "grad_norm": 0.8913895487785339,
+      "learning_rate": 1.403251904417341e-08,
+      "loss": 0.0588,
+      "step": 17910
+    },
+    {
+      "epoch": 2.9019766688269604,
+      "grad_norm": 1.0606086254119873,
+      "learning_rate": 1.3986288506260825e-08,
+      "loss": 0.0616,
+      "step": 17911
+    },
+    {
+      "epoch": 2.9021386908619573,
+      "grad_norm": 0.8608266115188599,
+      "learning_rate": 1.3940134034351738e-08,
+      "loss": 0.0547,
+      "step": 17912
+    },
+    {
+      "epoch": 2.9023007128969542,
+      "grad_norm": 1.2168540954589844,
+      "learning_rate": 1.3894055629858627e-08,
+      "loss": 0.072,
+      "step": 17913
+    },
+    {
+      "epoch": 2.9024627349319507,
+      "grad_norm": 0.862879753112793,
+      "learning_rate": 1.3848053294190922e-08,
+      "loss": 0.054,
+      "step": 17914
+    },
+    {
+      "epoch": 2.9026247569669477,
+      "grad_norm": 0.9422655701637268,
+      "learning_rate": 1.3802127028756662e-08,
+      "loss": 0.0596,
+      "step": 17915
+    },
+    {
+      "epoch": 2.902786779001944,
+      "grad_norm": 0.8769212961196899,
+      "learning_rate": 1.3756276834960558e-08,
+      "loss": 0.0569,
+      "step": 17916
+    },
+    {
+      "epoch": 2.902948801036941,
+      "grad_norm": 0.8366473913192749,
+      "learning_rate": 1.3710502714205654e-08,
+      "loss": 0.0578,
+      "step": 17917
+    },
+    {
+      "epoch": 2.9031108230719376,
+      "grad_norm": 0.9338627457618713,
+      "learning_rate": 1.366480466789305e-08,
+      "loss": 0.0608,
+      "step": 17918
+    },
+    {
+      "epoch": 2.9032728451069345,
+      "grad_norm": 0.9875946640968323,
+      "learning_rate": 1.3619182697420518e-08,
+      "loss": 0.0588,
+      "step": 17919
+    },
+    {
+      "epoch": 2.9034348671419314,
+      "grad_norm": 0.8782922029495239,
+      "learning_rate": 1.3573636804183887e-08,
+      "loss": 0.0565,
+      "step": 17920
+    },
+    {
+      "epoch": 2.903596889176928,
+      "grad_norm": 0.8396292924880981,
+      "learning_rate": 1.352816698957704e-08,
+      "loss": 0.0578,
+      "step": 17921
+    },
+    {
+      "epoch": 2.903758911211925,
+      "grad_norm": 1.0632270574569702,
+      "learning_rate": 1.3482773254991365e-08,
+      "loss": 0.0586,
+      "step": 17922
+    },
+    {
+      "epoch": 2.903920933246922,
+      "grad_norm": 0.9017321467399597,
+      "learning_rate": 1.3437455601815198e-08,
+      "loss": 0.0591,
+      "step": 17923
+    },
+    {
+      "epoch": 2.9040829552819183,
+      "grad_norm": 0.8652000427246094,
+      "learning_rate": 1.3392214031435757e-08,
+      "loss": 0.0583,
+      "step": 17924
+    },
+    {
+      "epoch": 2.904244977316915,
+      "grad_norm": 0.9219072461128235,
+      "learning_rate": 1.334704854523694e-08,
+      "loss": 0.0661,
+      "step": 17925
+    },
+    {
+      "epoch": 2.9044069993519117,
+      "grad_norm": 1.0006201267242432,
+      "learning_rate": 1.3301959144600974e-08,
+      "loss": 0.0635,
+      "step": 17926
+    },
+    {
+      "epoch": 2.9045690213869086,
+      "grad_norm": 0.8481730222702026,
+      "learning_rate": 1.325694583090731e-08,
+      "loss": 0.0576,
+      "step": 17927
+    },
+    {
+      "epoch": 2.904731043421905,
+      "grad_norm": 0.8712880611419678,
+      "learning_rate": 1.3212008605533177e-08,
+      "loss": 0.0556,
+      "step": 17928
+    },
+    {
+      "epoch": 2.904893065456902,
+      "grad_norm": 0.8364845514297485,
+      "learning_rate": 1.316714746985387e-08,
+      "loss": 0.058,
+      "step": 17929
+    },
+    {
+      "epoch": 2.905055087491899,
+      "grad_norm": 0.8850131630897522,
+      "learning_rate": 1.312236242524162e-08,
+      "loss": 0.0563,
+      "step": 17930
+    },
+    {
+      "epoch": 2.9052171095268955,
+      "grad_norm": 0.8984468579292297,
+      "learning_rate": 1.3077653473067276e-08,
+      "loss": 0.0587,
+      "step": 17931
+    },
+    {
+      "epoch": 2.9053791315618924,
+      "grad_norm": 0.9292992949485779,
+      "learning_rate": 1.3033020614698078e-08,
+      "loss": 0.0633,
+      "step": 17932
+    },
+    {
+      "epoch": 2.9055411535968894,
+      "grad_norm": 1.0178935527801514,
+      "learning_rate": 1.298846385150071e-08,
+      "loss": 0.0657,
+      "step": 17933
+    },
+    {
+      "epoch": 2.905703175631886,
+      "grad_norm": 0.9548495411872864,
+      "learning_rate": 1.2943983184837417e-08,
+      "loss": 0.0589,
+      "step": 17934
+    },
+    {
+      "epoch": 2.905865197666883,
+      "grad_norm": 0.9326539039611816,
+      "learning_rate": 1.2899578616069607e-08,
+      "loss": 0.0588,
+      "step": 17935
+    },
+    {
+      "epoch": 2.9060272197018797,
+      "grad_norm": 0.9059630036354065,
+      "learning_rate": 1.2855250146556197e-08,
+      "loss": 0.0605,
+      "step": 17936
+    },
+    {
+      "epoch": 2.906189241736876,
+      "grad_norm": 0.8523368835449219,
+      "learning_rate": 1.28109977776536e-08,
+      "loss": 0.051,
+      "step": 17937
+    },
+    {
+      "epoch": 2.906351263771873,
+      "grad_norm": 0.7979968786239624,
+      "learning_rate": 1.2766821510715177e-08,
+      "loss": 0.0501,
+      "step": 17938
+    },
+    {
+      "epoch": 2.9065132858068696,
+      "grad_norm": 0.8871493339538574,
+      "learning_rate": 1.272272134709318e-08,
+      "loss": 0.0613,
+      "step": 17939
+    },
+    {
+      "epoch": 2.9066753078418666,
+      "grad_norm": 0.8767561316490173,
+      "learning_rate": 1.2678697288136809e-08,
+      "loss": 0.0619,
+      "step": 17940
+    },
+    {
+      "epoch": 2.906837329876863,
+      "grad_norm": 0.8473535180091858,
+      "learning_rate": 1.2634749335193319e-08,
+      "loss": 0.0568,
+      "step": 17941
+    },
+    {
+      "epoch": 2.90699935191186,
+      "grad_norm": 0.9051182866096497,
+      "learning_rate": 1.2590877489606911e-08,
+      "loss": 0.0559,
+      "step": 17942
+    },
+    {
+      "epoch": 2.907161373946857,
+      "grad_norm": 0.928796112537384,
+      "learning_rate": 1.2547081752720402e-08,
+      "loss": 0.0609,
+      "step": 17943
+    },
+    {
+      "epoch": 2.9073233959818534,
+      "grad_norm": 0.9684448838233948,
+      "learning_rate": 1.2503362125873552e-08,
+      "loss": 0.0574,
+      "step": 17944
+    },
+    {
+      "epoch": 2.9074854180168503,
+      "grad_norm": 1.0342750549316406,
+      "learning_rate": 1.2459718610404182e-08,
+      "loss": 0.0629,
+      "step": 17945
+    },
+    {
+      "epoch": 2.9076474400518473,
+      "grad_norm": 0.8830204606056213,
+      "learning_rate": 1.241615120764761e-08,
+      "loss": 0.0574,
+      "step": 17946
+    },
+    {
+      "epoch": 2.9078094620868438,
+      "grad_norm": 0.8388495445251465,
+      "learning_rate": 1.2372659918937213e-08,
+      "loss": 0.0511,
+      "step": 17947
+    },
+    {
+      "epoch": 2.9079714841218407,
+      "grad_norm": 1.0486900806427002,
+      "learning_rate": 1.2329244745603596e-08,
+      "loss": 0.0685,
+      "step": 17948
+    },
+    {
+      "epoch": 2.908133506156837,
+      "grad_norm": 0.8908421993255615,
+      "learning_rate": 1.2285905688974587e-08,
+      "loss": 0.0535,
+      "step": 17949
+    },
+    {
+      "epoch": 2.908295528191834,
+      "grad_norm": 0.8851662278175354,
+      "learning_rate": 1.2242642750376899e-08,
+      "loss": 0.0572,
+      "step": 17950
+    },
+    {
+      "epoch": 2.9084575502268306,
+      "grad_norm": 0.9874460697174072,
+      "learning_rate": 1.2199455931134197e-08,
+      "loss": 0.0566,
+      "step": 17951
+    },
+    {
+      "epoch": 2.9086195722618275,
+      "grad_norm": 1.0689846277236938,
+      "learning_rate": 1.2156345232567923e-08,
+      "loss": 0.0702,
+      "step": 17952
+    },
+    {
+      "epoch": 2.9087815942968245,
+      "grad_norm": 1.057041049003601,
+      "learning_rate": 1.2113310655996746e-08,
+      "loss": 0.056,
+      "step": 17953
+    },
+    {
+      "epoch": 2.908943616331821,
+      "grad_norm": 0.8022787570953369,
+      "learning_rate": 1.2070352202737668e-08,
+      "loss": 0.0512,
+      "step": 17954
+    },
+    {
+      "epoch": 2.909105638366818,
+      "grad_norm": 0.9817745685577393,
+      "learning_rate": 1.202746987410519e-08,
+      "loss": 0.0612,
+      "step": 17955
+    },
+    {
+      "epoch": 2.909267660401815,
+      "grad_norm": 1.1243666410446167,
+      "learning_rate": 1.19846636714116e-08,
+      "loss": 0.0682,
+      "step": 17956
+    },
+    {
+      "epoch": 2.9094296824368113,
+      "grad_norm": 0.9882938861846924,
+      "learning_rate": 1.1941933595966127e-08,
+      "loss": 0.0599,
+      "step": 17957
+    },
+    {
+      "epoch": 2.9095917044718083,
+      "grad_norm": 0.9126664996147156,
+      "learning_rate": 1.1899279649076612e-08,
+      "loss": 0.0604,
+      "step": 17958
+    },
+    {
+      "epoch": 2.909753726506805,
+      "grad_norm": 0.8586945533752441,
+      "learning_rate": 1.1856701832047845e-08,
+      "loss": 0.0615,
+      "step": 17959
+    },
+    {
+      "epoch": 2.9099157485418017,
+      "grad_norm": 0.8310577273368835,
+      "learning_rate": 1.181420014618323e-08,
+      "loss": 0.0592,
+      "step": 17960
+    },
+    {
+      "epoch": 2.910077770576798,
+      "grad_norm": 0.9139492511749268,
+      "learning_rate": 1.1771774592782558e-08,
+      "loss": 0.0584,
+      "step": 17961
+    },
+    {
+      "epoch": 2.910239792611795,
+      "grad_norm": 1.0066767930984497,
+      "learning_rate": 1.172942517314396e-08,
+      "loss": 0.0588,
+      "step": 17962
+    },
+    {
+      "epoch": 2.910401814646792,
+      "grad_norm": 0.7591646909713745,
+      "learning_rate": 1.1687151888563897e-08,
+      "loss": 0.0507,
+      "step": 17963
+    },
+    {
+      "epoch": 2.9105638366817885,
+      "grad_norm": 0.8789999485015869,
+      "learning_rate": 1.1644954740334946e-08,
+      "loss": 0.0631,
+      "step": 17964
+    },
+    {
+      "epoch": 2.9107258587167855,
+      "grad_norm": 0.9494521617889404,
+      "learning_rate": 1.160283372974913e-08,
+      "loss": 0.0592,
+      "step": 17965
+    },
+    {
+      "epoch": 2.9108878807517824,
+      "grad_norm": 0.9379144310951233,
+      "learning_rate": 1.1560788858094584e-08,
+      "loss": 0.0607,
+      "step": 17966
+    },
+    {
+      "epoch": 2.911049902786779,
+      "grad_norm": 1.2142386436462402,
+      "learning_rate": 1.1518820126658058e-08,
+      "loss": 0.0623,
+      "step": 17967
+    },
+    {
+      "epoch": 2.911211924821776,
+      "grad_norm": 0.8130635619163513,
+      "learning_rate": 1.1476927536723248e-08,
+      "loss": 0.0544,
+      "step": 17968
+    },
+    {
+      "epoch": 2.9113739468567728,
+      "grad_norm": 0.9625205397605896,
+      "learning_rate": 1.1435111089572737e-08,
+      "loss": 0.0612,
+      "step": 17969
+    },
+    {
+      "epoch": 2.9115359688917692,
+      "grad_norm": 1.122702956199646,
+      "learning_rate": 1.1393370786485502e-08,
+      "loss": 0.0628,
+      "step": 17970
+    },
+    {
+      "epoch": 2.911697990926766,
+      "grad_norm": 0.9404149651527405,
+      "learning_rate": 1.1351706628738857e-08,
+      "loss": 0.0646,
+      "step": 17971
+    },
+    {
+      "epoch": 2.9118600129617627,
+      "grad_norm": 0.8580222725868225,
+      "learning_rate": 1.1310118617607613e-08,
+      "loss": 0.0536,
+      "step": 17972
+    },
+    {
+      "epoch": 2.9120220349967596,
+      "grad_norm": 0.8675228953361511,
+      "learning_rate": 1.1268606754364087e-08,
+      "loss": 0.056,
+      "step": 17973
+    },
+    {
+      "epoch": 2.912184057031756,
+      "grad_norm": 0.8910603523254395,
+      "learning_rate": 1.122717104027865e-08,
+      "loss": 0.0547,
+      "step": 17974
+    },
+    {
+      "epoch": 2.912346079066753,
+      "grad_norm": 0.8188005685806274,
+      "learning_rate": 1.1185811476619179e-08,
+      "loss": 0.0561,
+      "step": 17975
+    },
+    {
+      "epoch": 2.91250810110175,
+      "grad_norm": 0.8997900485992432,
+      "learning_rate": 1.1144528064650772e-08,
+      "loss": 0.0615,
+      "step": 17976
+    },
+    {
+      "epoch": 2.9126701231367464,
+      "grad_norm": 0.9417411684989929,
+      "learning_rate": 1.1103320805637141e-08,
+      "loss": 0.0616,
+      "step": 17977
+    },
+    {
+      "epoch": 2.9128321451717434,
+      "grad_norm": 0.7963204383850098,
+      "learning_rate": 1.1062189700838944e-08,
+      "loss": 0.0538,
+      "step": 17978
+    },
+    {
+      "epoch": 2.9129941672067403,
+      "grad_norm": 0.8257819414138794,
+      "learning_rate": 1.102113475151434e-08,
+      "loss": 0.0554,
+      "step": 17979
+    },
+    {
+      "epoch": 2.913156189241737,
+      "grad_norm": 1.1627333164215088,
+      "learning_rate": 1.0980155958920103e-08,
+      "loss": 0.0594,
+      "step": 17980
+    },
+    {
+      "epoch": 2.9133182112767337,
+      "grad_norm": 0.9248695373535156,
+      "learning_rate": 1.0939253324309673e-08,
+      "loss": 0.0618,
+      "step": 17981
+    },
+    {
+      "epoch": 2.9134802333117307,
+      "grad_norm": 0.9199352264404297,
+      "learning_rate": 1.089842684893455e-08,
+      "loss": 0.0585,
+      "step": 17982
+    },
+    {
+      "epoch": 2.913642255346727,
+      "grad_norm": 0.9159268140792847,
+      "learning_rate": 1.0857676534044014e-08,
+      "loss": 0.0591,
+      "step": 17983
+    },
+    {
+      "epoch": 2.9138042773817237,
+      "grad_norm": 1.0434057712554932,
+      "learning_rate": 1.0817002380885123e-08,
+      "loss": 0.0663,
+      "step": 17984
+    },
+    {
+      "epoch": 2.9139662994167206,
+      "grad_norm": 0.8765219449996948,
+      "learning_rate": 1.0776404390702434e-08,
+      "loss": 0.0646,
+      "step": 17985
+    },
+    {
+      "epoch": 2.9141283214517175,
+      "grad_norm": 0.9191032648086548,
+      "learning_rate": 1.0735882564737732e-08,
+      "loss": 0.0617,
+      "step": 17986
+    },
+    {
+      "epoch": 2.914290343486714,
+      "grad_norm": 1.1906532049179077,
+      "learning_rate": 1.0695436904231137e-08,
+      "loss": 0.0587,
+      "step": 17987
+    },
+    {
+      "epoch": 2.914452365521711,
+      "grad_norm": 0.9076985716819763,
+      "learning_rate": 1.0655067410419994e-08,
+      "loss": 0.0602,
+      "step": 17988
+    },
+    {
+      "epoch": 2.914614387556708,
+      "grad_norm": 0.9002047777175903,
+      "learning_rate": 1.061477408453998e-08,
+      "loss": 0.0568,
+      "step": 17989
+    },
+    {
+      "epoch": 2.9147764095917044,
+      "grad_norm": 0.9402999877929688,
+      "learning_rate": 1.057455692782372e-08,
+      "loss": 0.062,
+      "step": 17990
+    },
+    {
+      "epoch": 2.9149384316267013,
+      "grad_norm": 0.9988169074058533,
+      "learning_rate": 1.0534415941501341e-08,
+      "loss": 0.0629,
+      "step": 17991
+    },
+    {
+      "epoch": 2.9151004536616982,
+      "grad_norm": 0.8821046352386475,
+      "learning_rate": 1.049435112680186e-08,
+      "loss": 0.0601,
+      "step": 17992
+    },
+    {
+      "epoch": 2.9152624756966947,
+      "grad_norm": 0.8363478183746338,
+      "learning_rate": 1.0454362484950409e-08,
+      "loss": 0.0605,
+      "step": 17993
+    },
+    {
+      "epoch": 2.9154244977316917,
+      "grad_norm": 0.9162735939025879,
+      "learning_rate": 1.0414450017171007e-08,
+      "loss": 0.0612,
+      "step": 17994
+    },
+    {
+      "epoch": 2.915586519766688,
+      "grad_norm": 0.9638509750366211,
+      "learning_rate": 1.03746137246849e-08,
+      "loss": 0.0622,
+      "step": 17995
+    },
+    {
+      "epoch": 2.915748541801685,
+      "grad_norm": 0.8784467577934265,
+      "learning_rate": 1.0334853608710838e-08,
+      "loss": 0.0626,
+      "step": 17996
+    },
+    {
+      "epoch": 2.9159105638366816,
+      "grad_norm": 0.9395826458930969,
+      "learning_rate": 1.0295169670465066e-08,
+      "loss": 0.0605,
+      "step": 17997
+    },
+    {
+      "epoch": 2.9160725858716785,
+      "grad_norm": 1.00931978225708,
+      "learning_rate": 1.0255561911162449e-08,
+      "loss": 0.0597,
+      "step": 17998
+    },
+    {
+      "epoch": 2.9162346079066754,
+      "grad_norm": 0.8709773421287537,
+      "learning_rate": 1.0216030332014515e-08,
+      "loss": 0.0579,
+      "step": 17999
+    },
+    {
+      "epoch": 2.916396629941672,
+      "grad_norm": 0.8078761100769043,
+      "learning_rate": 1.0176574934230854e-08,
+      "loss": 0.0583,
+      "step": 18000
+    },
+    {
+      "epoch": 2.916558651976669,
+      "grad_norm": 0.8328400254249573,
+      "learning_rate": 1.0137195719018556e-08,
+      "loss": 0.0519,
+      "step": 18001
+    },
+    {
+      "epoch": 2.916720674011666,
+      "grad_norm": 0.9479650259017944,
+      "learning_rate": 1.0097892687583044e-08,
+      "loss": 0.0587,
+      "step": 18002
+    },
+    {
+      "epoch": 2.9168826960466623,
+      "grad_norm": 0.9351217150688171,
+      "learning_rate": 1.0058665841126414e-08,
+      "loss": 0.0608,
+      "step": 18003
+    },
+    {
+      "epoch": 2.917044718081659,
+      "grad_norm": 0.7649050354957581,
+      "learning_rate": 1.0019515180849094e-08,
+      "loss": 0.0538,
+      "step": 18004
+    },
+    {
+      "epoch": 2.9172067401166557,
+      "grad_norm": 0.9090438485145569,
+      "learning_rate": 9.980440707948735e-09,
+      "loss": 0.0559,
+      "step": 18005
+    },
+    {
+      "epoch": 2.9173687621516526,
+      "grad_norm": 0.8982040286064148,
+      "learning_rate": 9.941442423621606e-09,
+      "loss": 0.056,
+      "step": 18006
+    },
+    {
+      "epoch": 2.917530784186649,
+      "grad_norm": 0.9106990694999695,
+      "learning_rate": 9.902520329060083e-09,
+      "loss": 0.0535,
+      "step": 18007
+    },
+    {
+      "epoch": 2.917692806221646,
+      "grad_norm": 1.0003997087478638,
+      "learning_rate": 9.863674425455716e-09,
+      "loss": 0.0647,
+      "step": 18008
+    },
+    {
+      "epoch": 2.917854828256643,
+      "grad_norm": 1.027595043182373,
+      "learning_rate": 9.824904713996996e-09,
+      "loss": 0.0557,
+      "step": 18009
+    },
+    {
+      "epoch": 2.9180168502916395,
+      "grad_norm": 0.9183197021484375,
+      "learning_rate": 9.78621119586992e-09,
+      "loss": 0.0558,
+      "step": 18010
+    },
+    {
+      "epoch": 2.9181788723266364,
+      "grad_norm": 1.1512269973754883,
+      "learning_rate": 9.747593872258543e-09,
+      "loss": 0.0643,
+      "step": 18011
+    },
+    {
+      "epoch": 2.9183408943616334,
+      "grad_norm": 0.859645664691925,
+      "learning_rate": 9.709052744344694e-09,
+      "loss": 0.0649,
+      "step": 18012
+    },
+    {
+      "epoch": 2.91850291639663,
+      "grad_norm": 0.85866379737854,
+      "learning_rate": 9.670587813307153e-09,
+      "loss": 0.0537,
+      "step": 18013
+    },
+    {
+      "epoch": 2.9186649384316268,
+      "grad_norm": 0.8549041152000427,
+      "learning_rate": 9.63219908032359e-09,
+      "loss": 0.0596,
+      "step": 18014
+    },
+    {
+      "epoch": 2.9188269604666237,
+      "grad_norm": 0.8607486486434937,
+      "learning_rate": 9.593886546567787e-09,
+      "loss": 0.06,
+      "step": 18015
+    },
+    {
+      "epoch": 2.91898898250162,
+      "grad_norm": 0.9240609407424927,
+      "learning_rate": 9.55565021321242e-09,
+      "loss": 0.0589,
+      "step": 18016
+    },
+    {
+      "epoch": 2.919151004536617,
+      "grad_norm": 0.9318268895149231,
+      "learning_rate": 9.51749008142766e-09,
+      "loss": 0.0504,
+      "step": 18017
+    },
+    {
+      "epoch": 2.9193130265716136,
+      "grad_norm": 0.9120360016822815,
+      "learning_rate": 9.479406152380632e-09,
+      "loss": 0.0582,
+      "step": 18018
+    },
+    {
+      "epoch": 2.9194750486066106,
+      "grad_norm": 0.8662173748016357,
+      "learning_rate": 9.441398427236792e-09,
+      "loss": 0.0557,
+      "step": 18019
+    },
+    {
+      "epoch": 2.919637070641607,
+      "grad_norm": 0.9860486388206482,
+      "learning_rate": 9.403466907159375e-09,
+      "loss": 0.0609,
+      "step": 18020
+    },
+    {
+      "epoch": 2.919799092676604,
+      "grad_norm": 1.0033873319625854,
+      "learning_rate": 9.365611593308565e-09,
+      "loss": 0.0586,
+      "step": 18021
+    },
+    {
+      "epoch": 2.919961114711601,
+      "grad_norm": 0.8788847327232361,
+      "learning_rate": 9.327832486842603e-09,
+      "loss": 0.0595,
+      "step": 18022
+    },
+    {
+      "epoch": 2.9201231367465974,
+      "grad_norm": 1.0497198104858398,
+      "learning_rate": 9.290129588918062e-09,
+      "loss": 0.0669,
+      "step": 18023
+    },
+    {
+      "epoch": 2.9202851587815943,
+      "grad_norm": 0.9634641408920288,
+      "learning_rate": 9.25250290068791e-09,
+      "loss": 0.0572,
+      "step": 18024
+    },
+    {
+      "epoch": 2.9204471808165913,
+      "grad_norm": 0.8264623284339905,
+      "learning_rate": 9.214952423303724e-09,
+      "loss": 0.0575,
+      "step": 18025
+    },
+    {
+      "epoch": 2.9206092028515878,
+      "grad_norm": 0.8819963335990906,
+      "learning_rate": 9.17747815791431e-09,
+      "loss": 0.0554,
+      "step": 18026
+    },
+    {
+      "epoch": 2.9207712248865847,
+      "grad_norm": 0.8962211012840271,
+      "learning_rate": 9.140080105666527e-09,
+      "loss": 0.0608,
+      "step": 18027
+    },
+    {
+      "epoch": 2.920933246921581,
+      "grad_norm": 1.065934181213379,
+      "learning_rate": 9.102758267704736e-09,
+      "loss": 0.0623,
+      "step": 18028
+    },
+    {
+      "epoch": 2.921095268956578,
+      "grad_norm": 0.9114288687705994,
+      "learning_rate": 9.065512645170249e-09,
+      "loss": 0.0578,
+      "step": 18029
+    },
+    {
+      "epoch": 2.9212572909915746,
+      "grad_norm": 1.0408483743667603,
+      "learning_rate": 9.02834323920354e-09,
+      "loss": 0.0679,
+      "step": 18030
+    },
+    {
+      "epoch": 2.9214193130265715,
+      "grad_norm": 0.9269614815711975,
+      "learning_rate": 8.991250050941203e-09,
+      "loss": 0.0599,
+      "step": 18031
+    },
+    {
+      "epoch": 2.9215813350615685,
+      "grad_norm": 0.850156843662262,
+      "learning_rate": 8.954233081518438e-09,
+      "loss": 0.0532,
+      "step": 18032
+    },
+    {
+      "epoch": 2.921743357096565,
+      "grad_norm": 0.8286687135696411,
+      "learning_rate": 8.917292332068228e-09,
+      "loss": 0.0546,
+      "step": 18033
+    },
+    {
+      "epoch": 2.921905379131562,
+      "grad_norm": 0.8290266394615173,
+      "learning_rate": 8.880427803720226e-09,
+      "loss": 0.0591,
+      "step": 18034
+    },
+    {
+      "epoch": 2.922067401166559,
+      "grad_norm": 0.8320590853691101,
+      "learning_rate": 8.843639497602973e-09,
+      "loss": 0.0516,
+      "step": 18035
+    },
+    {
+      "epoch": 2.9222294232015553,
+      "grad_norm": 0.9579795002937317,
+      "learning_rate": 8.806927414841959e-09,
+      "loss": 0.0585,
+      "step": 18036
+    },
+    {
+      "epoch": 2.9223914452365523,
+      "grad_norm": 0.8604581356048584,
+      "learning_rate": 8.770291556560174e-09,
+      "loss": 0.0567,
+      "step": 18037
+    },
+    {
+      "epoch": 2.922553467271549,
+      "grad_norm": 0.8759823441505432,
+      "learning_rate": 8.73373192387894e-09,
+      "loss": 0.0628,
+      "step": 18038
+    },
+    {
+      "epoch": 2.9227154893065457,
+      "grad_norm": 0.9367085099220276,
+      "learning_rate": 8.697248517916535e-09,
+      "loss": 0.0609,
+      "step": 18039
+    },
+    {
+      "epoch": 2.9228775113415426,
+      "grad_norm": 0.9392814040184021,
+      "learning_rate": 8.660841339789561e-09,
+      "loss": 0.068,
+      "step": 18040
+    },
+    {
+      "epoch": 2.923039533376539,
+      "grad_norm": 0.8410159349441528,
+      "learning_rate": 8.62451039061213e-09,
+      "loss": 0.0539,
+      "step": 18041
+    },
+    {
+      "epoch": 2.923201555411536,
+      "grad_norm": 0.8481424450874329,
+      "learning_rate": 8.58825567149557e-09,
+      "loss": 0.0567,
+      "step": 18042
+    },
+    {
+      "epoch": 2.9233635774465325,
+      "grad_norm": 1.1205499172210693,
+      "learning_rate": 8.55207718354928e-09,
+      "loss": 0.0688,
+      "step": 18043
+    },
+    {
+      "epoch": 2.9235255994815295,
+      "grad_norm": 0.819627046585083,
+      "learning_rate": 8.515974927880144e-09,
+      "loss": 0.055,
+      "step": 18044
+    },
+    {
+      "epoch": 2.9236876215165264,
+      "grad_norm": 0.8424476981163025,
+      "learning_rate": 8.479948905592839e-09,
+      "loss": 0.0556,
+      "step": 18045
+    },
+    {
+      "epoch": 2.923849643551523,
+      "grad_norm": 0.9404410719871521,
+      "learning_rate": 8.443999117790091e-09,
+      "loss": 0.0572,
+      "step": 18046
+    },
+    {
+      "epoch": 2.92401166558652,
+      "grad_norm": 1.1736658811569214,
+      "learning_rate": 8.4081255655713e-09,
+      "loss": 0.0574,
+      "step": 18047
+    },
+    {
+      "epoch": 2.9241736876215167,
+      "grad_norm": 1.16831374168396,
+      "learning_rate": 8.3723282500342e-09,
+      "loss": 0.0615,
+      "step": 18048
+    },
+    {
+      "epoch": 2.9243357096565132,
+      "grad_norm": 0.8560138940811157,
+      "learning_rate": 8.33660717227458e-09,
+      "loss": 0.0581,
+      "step": 18049
+    },
+    {
+      "epoch": 2.92449773169151,
+      "grad_norm": 0.9814584255218506,
+      "learning_rate": 8.3009623333849e-09,
+      "loss": 0.0648,
+      "step": 18050
+    },
+    {
+      "epoch": 2.9246597537265067,
+      "grad_norm": 0.8727818131446838,
+      "learning_rate": 8.265393734455674e-09,
+      "loss": 0.0601,
+      "step": 18051
+    },
+    {
+      "epoch": 2.9248217757615036,
+      "grad_norm": 1.0411512851715088,
+      "learning_rate": 8.229901376575755e-09,
+      "loss": 0.0601,
+      "step": 18052
+    },
+    {
+      "epoch": 2.9249837977965,
+      "grad_norm": 1.0159389972686768,
+      "learning_rate": 8.194485260830943e-09,
+      "loss": 0.0582,
+      "step": 18053
+    },
+    {
+      "epoch": 2.925145819831497,
+      "grad_norm": 0.8708806037902832,
+      "learning_rate": 8.15914538830509e-09,
+      "loss": 0.0551,
+      "step": 18054
+    },
+    {
+      "epoch": 2.925307841866494,
+      "grad_norm": 0.8334715962409973,
+      "learning_rate": 8.123881760078723e-09,
+      "loss": 0.0575,
+      "step": 18055
+    },
+    {
+      "epoch": 2.9254698639014904,
+      "grad_norm": 0.9106010794639587,
+      "learning_rate": 8.088694377231532e-09,
+      "loss": 0.0626,
+      "step": 18056
+    },
+    {
+      "epoch": 2.9256318859364874,
+      "grad_norm": 0.9865289926528931,
+      "learning_rate": 8.053583240840157e-09,
+      "loss": 0.0614,
+      "step": 18057
+    },
+    {
+      "epoch": 2.9257939079714843,
+      "grad_norm": 0.9036328196525574,
+      "learning_rate": 8.018548351978738e-09,
+      "loss": 0.0612,
+      "step": 18058
+    },
+    {
+      "epoch": 2.925955930006481,
+      "grad_norm": 0.9329188466072083,
+      "learning_rate": 7.98358971171892e-09,
+      "loss": 0.0589,
+      "step": 18059
+    },
+    {
+      "epoch": 2.9261179520414777,
+      "grad_norm": 1.0107616186141968,
+      "learning_rate": 7.948707321130956e-09,
+      "loss": 0.0607,
+      "step": 18060
+    },
+    {
+      "epoch": 2.9262799740764747,
+      "grad_norm": 1.3307433128356934,
+      "learning_rate": 7.913901181281768e-09,
+      "loss": 0.0597,
+      "step": 18061
+    },
+    {
+      "epoch": 2.926441996111471,
+      "grad_norm": 0.8141980171203613,
+      "learning_rate": 7.879171293236621e-09,
+      "loss": 0.0541,
+      "step": 18062
+    },
+    {
+      "epoch": 2.9266040181464676,
+      "grad_norm": 0.894360363483429,
+      "learning_rate": 7.844517658057993e-09,
+      "loss": 0.0631,
+      "step": 18063
+    },
+    {
+      "epoch": 2.9267660401814646,
+      "grad_norm": 0.9647737145423889,
+      "learning_rate": 7.80994027680615e-09,
+      "loss": 0.0555,
+      "step": 18064
+    },
+    {
+      "epoch": 2.9269280622164615,
+      "grad_norm": 0.874281644821167,
+      "learning_rate": 7.775439150539132e-09,
+      "loss": 0.0584,
+      "step": 18065
+    },
+    {
+      "epoch": 2.927090084251458,
+      "grad_norm": 0.9083302617073059,
+      "learning_rate": 7.741014280312765e-09,
+      "loss": 0.0563,
+      "step": 18066
+    },
+    {
+      "epoch": 2.927252106286455,
+      "grad_norm": 0.9287460446357727,
+      "learning_rate": 7.70666566718009e-09,
+      "loss": 0.06,
+      "step": 18067
+    },
+    {
+      "epoch": 2.927414128321452,
+      "grad_norm": 0.9412106275558472,
+      "learning_rate": 7.672393312192218e-09,
+      "loss": 0.0596,
+      "step": 18068
+    },
+    {
+      "epoch": 2.9275761503564484,
+      "grad_norm": 0.7685844302177429,
+      "learning_rate": 7.638197216397748e-09,
+      "loss": 0.0508,
+      "step": 18069
+    },
+    {
+      "epoch": 2.9277381723914453,
+      "grad_norm": 0.9476762413978577,
+      "learning_rate": 7.604077380843067e-09,
+      "loss": 0.0544,
+      "step": 18070
+    },
+    {
+      "epoch": 2.9279001944264422,
+      "grad_norm": 0.9895046353340149,
+      "learning_rate": 7.57003380657234e-09,
+      "loss": 0.0626,
+      "step": 18071
+    },
+    {
+      "epoch": 2.9280622164614387,
+      "grad_norm": 0.825796365737915,
+      "learning_rate": 7.536066494626681e-09,
+      "loss": 0.0536,
+      "step": 18072
+    },
+    {
+      "epoch": 2.9282242384964356,
+      "grad_norm": 0.8940446972846985,
+      "learning_rate": 7.502175446046089e-09,
+      "loss": 0.0637,
+      "step": 18073
+    },
+    {
+      "epoch": 2.928386260531432,
+      "grad_norm": 1.003281831741333,
+      "learning_rate": 7.468360661866957e-09,
+      "loss": 0.0616,
+      "step": 18074
+    },
+    {
+      "epoch": 2.928548282566429,
+      "grad_norm": 0.901324987411499,
+      "learning_rate": 7.434622143124015e-09,
+      "loss": 0.0595,
+      "step": 18075
+    },
+    {
+      "epoch": 2.9287103046014256,
+      "grad_norm": 0.998007595539093,
+      "learning_rate": 7.400959890850046e-09,
+      "loss": 0.0623,
+      "step": 18076
+    },
+    {
+      "epoch": 2.9288723266364225,
+      "grad_norm": 0.9806587100028992,
+      "learning_rate": 7.367373906074782e-09,
+      "loss": 0.0575,
+      "step": 18077
+    },
+    {
+      "epoch": 2.9290343486714194,
+      "grad_norm": 0.963106095790863,
+      "learning_rate": 7.333864189825735e-09,
+      "loss": 0.0635,
+      "step": 18078
+    },
+    {
+      "epoch": 2.929196370706416,
+      "grad_norm": 0.9748607277870178,
+      "learning_rate": 7.3004307431281954e-09,
+      "loss": 0.0627,
+      "step": 18079
+    },
+    {
+      "epoch": 2.929358392741413,
+      "grad_norm": 0.8219755291938782,
+      "learning_rate": 7.267073567005234e-09,
+      "loss": 0.0534,
+      "step": 18080
+    },
+    {
+      "epoch": 2.92952041477641,
+      "grad_norm": 0.8908148407936096,
+      "learning_rate": 7.233792662477701e-09,
+      "loss": 0.0533,
+      "step": 18081
+    },
+    {
+      "epoch": 2.9296824368114063,
+      "grad_norm": 0.8585811853408813,
+      "learning_rate": 7.2005880305636714e-09,
+      "loss": 0.0563,
+      "step": 18082
+    },
+    {
+      "epoch": 2.929844458846403,
+      "grad_norm": 0.9972938895225525,
+      "learning_rate": 7.167459672278998e-09,
+      "loss": 0.0621,
+      "step": 18083
+    },
+    {
+      "epoch": 2.9300064808814,
+      "grad_norm": 0.8972851634025574,
+      "learning_rate": 7.134407588637871e-09,
+      "loss": 0.0567,
+      "step": 18084
+    },
+    {
+      "epoch": 2.9301685029163966,
+      "grad_norm": 0.903139591217041,
+      "learning_rate": 7.1014317806508696e-09,
+      "loss": 0.0603,
+      "step": 18085
+    },
+    {
+      "epoch": 2.930330524951393,
+      "grad_norm": 0.8544906377792358,
+      "learning_rate": 7.068532249327742e-09,
+      "loss": 0.06,
+      "step": 18086
+    },
+    {
+      "epoch": 2.93049254698639,
+      "grad_norm": 0.9023297429084778,
+      "learning_rate": 7.035708995674628e-09,
+      "loss": 0.0622,
+      "step": 18087
+    },
+    {
+      "epoch": 2.930654569021387,
+      "grad_norm": 0.9524878859519958,
+      "learning_rate": 7.002962020695725e-09,
+      "loss": 0.0581,
+      "step": 18088
+    },
+    {
+      "epoch": 2.9308165910563835,
+      "grad_norm": 1.0636718273162842,
+      "learning_rate": 6.970291325393286e-09,
+      "loss": 0.0624,
+      "step": 18089
+    },
+    {
+      "epoch": 2.9309786130913804,
+      "grad_norm": 0.9945453405380249,
+      "learning_rate": 6.937696910767067e-09,
+      "loss": 0.0608,
+      "step": 18090
+    },
+    {
+      "epoch": 2.9311406351263773,
+      "grad_norm": 0.8092207312583923,
+      "learning_rate": 6.905178777814326e-09,
+      "loss": 0.0559,
+      "step": 18091
+    },
+    {
+      "epoch": 2.931302657161374,
+      "grad_norm": 1.0313624143600464,
+      "learning_rate": 6.872736927529822e-09,
+      "loss": 0.0619,
+      "step": 18092
+    },
+    {
+      "epoch": 2.9314646791963708,
+      "grad_norm": 1.039193034172058,
+      "learning_rate": 6.840371360906095e-09,
+      "loss": 0.0611,
+      "step": 18093
+    },
+    {
+      "epoch": 2.9316267012313677,
+      "grad_norm": 0.9258902668952942,
+      "learning_rate": 6.8080820789340195e-09,
+      "loss": 0.0572,
+      "step": 18094
+    },
+    {
+      "epoch": 2.931788723266364,
+      "grad_norm": 0.9824669361114502,
+      "learning_rate": 6.775869082601139e-09,
+      "loss": 0.0573,
+      "step": 18095
+    },
+    {
+      "epoch": 2.931950745301361,
+      "grad_norm": 0.8684791922569275,
+      "learning_rate": 6.743732372893053e-09,
+      "loss": 0.0595,
+      "step": 18096
+    },
+    {
+      "epoch": 2.9321127673363576,
+      "grad_norm": 0.8853773474693298,
+      "learning_rate": 6.711671950793419e-09,
+      "loss": 0.0513,
+      "step": 18097
+    },
+    {
+      "epoch": 2.9322747893713546,
+      "grad_norm": 0.944572389125824,
+      "learning_rate": 6.679687817282843e-09,
+      "loss": 0.0517,
+      "step": 18098
+    },
+    {
+      "epoch": 2.932436811406351,
+      "grad_norm": 0.9606372117996216,
+      "learning_rate": 6.647779973339985e-09,
+      "loss": 0.0594,
+      "step": 18099
+    },
+    {
+      "epoch": 2.932598833441348,
+      "grad_norm": 0.8097171187400818,
+      "learning_rate": 6.615948419941565e-09,
+      "loss": 0.0531,
+      "step": 18100
+    },
+    {
+      "epoch": 2.932760855476345,
+      "grad_norm": 0.8059049844741821,
+      "learning_rate": 6.584193158060969e-09,
+      "loss": 0.0524,
+      "step": 18101
+    },
+    {
+      "epoch": 2.9329228775113414,
+      "grad_norm": 0.8145474791526794,
+      "learning_rate": 6.5525141886702005e-09,
+      "loss": 0.0572,
+      "step": 18102
+    },
+    {
+      "epoch": 2.9330848995463383,
+      "grad_norm": 1.0327907800674438,
+      "learning_rate": 6.520911512738481e-09,
+      "loss": 0.0654,
+      "step": 18103
+    },
+    {
+      "epoch": 2.9332469215813353,
+      "grad_norm": 0.9246883392333984,
+      "learning_rate": 6.489385131232817e-09,
+      "loss": 0.0663,
+      "step": 18104
+    },
+    {
+      "epoch": 2.9334089436163318,
+      "grad_norm": 0.9235041737556458,
+      "learning_rate": 6.4579350451177135e-09,
+      "loss": 0.0673,
+      "step": 18105
+    },
+    {
+      "epoch": 2.9335709656513287,
+      "grad_norm": 0.8101209998130798,
+      "learning_rate": 6.426561255355457e-09,
+      "loss": 0.0568,
+      "step": 18106
+    },
+    {
+      "epoch": 2.933732987686325,
+      "grad_norm": 0.8323474526405334,
+      "learning_rate": 6.395263762906112e-09,
+      "loss": 0.0514,
+      "step": 18107
+    },
+    {
+      "epoch": 2.933895009721322,
+      "grad_norm": 0.8345054984092712,
+      "learning_rate": 6.364042568727524e-09,
+      "loss": 0.05,
+      "step": 18108
+    },
+    {
+      "epoch": 2.9340570317563186,
+      "grad_norm": 0.903107225894928,
+      "learning_rate": 6.332897673774485e-09,
+      "loss": 0.0614,
+      "step": 18109
+    },
+    {
+      "epoch": 2.9342190537913155,
+      "grad_norm": 1.5786057710647583,
+      "learning_rate": 6.301829079000399e-09,
+      "loss": 0.0695,
+      "step": 18110
+    },
+    {
+      "epoch": 2.9343810758263125,
+      "grad_norm": 0.8032664060592651,
+      "learning_rate": 6.270836785355616e-09,
+      "loss": 0.05,
+      "step": 18111
+    },
+    {
+      "epoch": 2.934543097861309,
+      "grad_norm": 1.0438019037246704,
+      "learning_rate": 6.239920793788546e-09,
+      "loss": 0.0631,
+      "step": 18112
+    },
+    {
+      "epoch": 2.934705119896306,
+      "grad_norm": 0.9662144184112549,
+      "learning_rate": 6.209081105245096e-09,
+      "loss": 0.0647,
+      "step": 18113
+    },
+    {
+      "epoch": 2.934867141931303,
+      "grad_norm": 1.402650237083435,
+      "learning_rate": 6.178317720668958e-09,
+      "loss": 0.0615,
+      "step": 18114
+    },
+    {
+      "epoch": 2.9350291639662993,
+      "grad_norm": 0.9588105082511902,
+      "learning_rate": 6.147630641001323e-09,
+      "loss": 0.0585,
+      "step": 18115
+    },
+    {
+      "epoch": 2.9351911860012962,
+      "grad_norm": 0.9412093162536621,
+      "learning_rate": 6.117019867181162e-09,
+      "loss": 0.0613,
+      "step": 18116
+    },
+    {
+      "epoch": 2.935353208036293,
+      "grad_norm": 1.0334348678588867,
+      "learning_rate": 6.0864854001452255e-09,
+      "loss": 0.0649,
+      "step": 18117
+    },
+    {
+      "epoch": 2.9355152300712897,
+      "grad_norm": 0.9597827196121216,
+      "learning_rate": 6.056027240827489e-09,
+      "loss": 0.0642,
+      "step": 18118
+    },
+    {
+      "epoch": 2.9356772521062866,
+      "grad_norm": 0.927130401134491,
+      "learning_rate": 6.025645390160262e-09,
+      "loss": 0.0632,
+      "step": 18119
+    },
+    {
+      "epoch": 2.935839274141283,
+      "grad_norm": 0.8240948915481567,
+      "learning_rate": 5.995339849073079e-09,
+      "loss": 0.0607,
+      "step": 18120
+    },
+    {
+      "epoch": 2.93600129617628,
+      "grad_norm": 0.919998824596405,
+      "learning_rate": 5.965110618492698e-09,
+      "loss": 0.0637,
+      "step": 18121
+    },
+    {
+      "epoch": 2.9361633182112765,
+      "grad_norm": 0.9138144254684448,
+      "learning_rate": 5.9349576993447675e-09,
+      "loss": 0.0608,
+      "step": 18122
+    },
+    {
+      "epoch": 2.9363253402462735,
+      "grad_norm": 1.1568125486373901,
+      "learning_rate": 5.904881092551607e-09,
+      "loss": 0.0606,
+      "step": 18123
+    },
+    {
+      "epoch": 2.9364873622812704,
+      "grad_norm": 0.8754058480262756,
+      "learning_rate": 5.874880799033589e-09,
+      "loss": 0.0555,
+      "step": 18124
+    },
+    {
+      "epoch": 2.936649384316267,
+      "grad_norm": 0.8889382481575012,
+      "learning_rate": 5.844956819708314e-09,
+      "loss": 0.0585,
+      "step": 18125
+    },
+    {
+      "epoch": 2.936811406351264,
+      "grad_norm": 0.9085453748703003,
+      "learning_rate": 5.815109155491716e-09,
+      "loss": 0.0605,
+      "step": 18126
+    },
+    {
+      "epoch": 2.9369734283862607,
+      "grad_norm": 0.9168490767478943,
+      "learning_rate": 5.785337807297231e-09,
+      "loss": 0.0616,
+      "step": 18127
+    },
+    {
+      "epoch": 2.9371354504212572,
+      "grad_norm": 1.0325088500976562,
+      "learning_rate": 5.755642776035242e-09,
+      "loss": 0.0636,
+      "step": 18128
+    },
+    {
+      "epoch": 2.937297472456254,
+      "grad_norm": 0.9850202202796936,
+      "learning_rate": 5.726024062614466e-09,
+      "loss": 0.0607,
+      "step": 18129
+    },
+    {
+      "epoch": 2.9374594944912507,
+      "grad_norm": 1.0262207984924316,
+      "learning_rate": 5.696481667941678e-09,
+      "loss": 0.0581,
+      "step": 18130
+    },
+    {
+      "epoch": 2.9376215165262476,
+      "grad_norm": 0.9356762766838074,
+      "learning_rate": 5.667015592920322e-09,
+      "loss": 0.064,
+      "step": 18131
+    },
+    {
+      "epoch": 2.937783538561244,
+      "grad_norm": 0.9865776896476746,
+      "learning_rate": 5.637625838452176e-09,
+      "loss": 0.0695,
+      "step": 18132
+    },
+    {
+      "epoch": 2.937945560596241,
+      "grad_norm": 1.0751196146011353,
+      "learning_rate": 5.608312405436245e-09,
+      "loss": 0.0706,
+      "step": 18133
+    },
+    {
+      "epoch": 2.938107582631238,
+      "grad_norm": 0.8332173824310303,
+      "learning_rate": 5.579075294769864e-09,
+      "loss": 0.0523,
+      "step": 18134
+    },
+    {
+      "epoch": 2.9382696046662344,
+      "grad_norm": 0.9454176425933838,
+      "learning_rate": 5.54991450734732e-09,
+      "loss": 0.0566,
+      "step": 18135
+    },
+    {
+      "epoch": 2.9384316267012314,
+      "grad_norm": 0.8918482661247253,
+      "learning_rate": 5.520830044060677e-09,
+      "loss": 0.06,
+      "step": 18136
+    },
+    {
+      "epoch": 2.9385936487362283,
+      "grad_norm": 0.9533538818359375,
+      "learning_rate": 5.491821905800332e-09,
+      "loss": 0.0536,
+      "step": 18137
+    },
+    {
+      "epoch": 2.938755670771225,
+      "grad_norm": 0.8268304467201233,
+      "learning_rate": 5.46289009345391e-09,
+      "loss": 0.0584,
+      "step": 18138
+    },
+    {
+      "epoch": 2.9389176928062217,
+      "grad_norm": 0.7862445116043091,
+      "learning_rate": 5.43403460790598e-09,
+      "loss": 0.0561,
+      "step": 18139
+    },
+    {
+      "epoch": 2.9390797148412187,
+      "grad_norm": 1.113399624824524,
+      "learning_rate": 5.405255450040003e-09,
+      "loss": 0.0603,
+      "step": 18140
+    },
+    {
+      "epoch": 2.939241736876215,
+      "grad_norm": 1.0670251846313477,
+      "learning_rate": 5.376552620736664e-09,
+      "loss": 0.0631,
+      "step": 18141
+    },
+    {
+      "epoch": 2.939403758911212,
+      "grad_norm": 0.9453678727149963,
+      "learning_rate": 5.347926120873592e-09,
+      "loss": 0.0585,
+      "step": 18142
+    },
+    {
+      "epoch": 2.9395657809462086,
+      "grad_norm": 0.9201284050941467,
+      "learning_rate": 5.319375951327033e-09,
+      "loss": 0.0643,
+      "step": 18143
+    },
+    {
+      "epoch": 2.9397278029812055,
+      "grad_norm": 0.929400622844696,
+      "learning_rate": 5.290902112970731e-09,
+      "loss": 0.0587,
+      "step": 18144
+    },
+    {
+      "epoch": 2.939889825016202,
+      "grad_norm": 1.096695065498352,
+      "learning_rate": 5.262504606675656e-09,
+      "loss": 0.0624,
+      "step": 18145
+    },
+    {
+      "epoch": 2.940051847051199,
+      "grad_norm": 0.8255508542060852,
+      "learning_rate": 5.234183433310835e-09,
+      "loss": 0.0513,
+      "step": 18146
+    },
+    {
+      "epoch": 2.940213869086196,
+      "grad_norm": 1.058481216430664,
+      "learning_rate": 5.205938593742799e-09,
+      "loss": 0.0652,
+      "step": 18147
+    },
+    {
+      "epoch": 2.9403758911211924,
+      "grad_norm": 0.99068284034729,
+      "learning_rate": 5.177770088835854e-09,
+      "loss": 0.0587,
+      "step": 18148
+    },
+    {
+      "epoch": 2.9405379131561893,
+      "grad_norm": 0.9016160368919373,
+      "learning_rate": 5.149677919451535e-09,
+      "loss": 0.0575,
+      "step": 18149
+    },
+    {
+      "epoch": 2.940699935191186,
+      "grad_norm": 1.0625081062316895,
+      "learning_rate": 5.121662086449708e-09,
+      "loss": 0.071,
+      "step": 18150
+    },
+    {
+      "epoch": 2.9408619572261827,
+      "grad_norm": 1.1517467498779297,
+      "learning_rate": 5.093722590687744e-09,
+      "loss": 0.0692,
+      "step": 18151
+    },
+    {
+      "epoch": 2.9410239792611796,
+      "grad_norm": 0.8428022861480713,
+      "learning_rate": 5.065859433019959e-09,
+      "loss": 0.051,
+      "step": 18152
+    },
+    {
+      "epoch": 2.941186001296176,
+      "grad_norm": 0.9025364518165588,
+      "learning_rate": 5.038072614299561e-09,
+      "loss": 0.0583,
+      "step": 18153
+    },
+    {
+      "epoch": 2.941348023331173,
+      "grad_norm": 1.0714325904846191,
+      "learning_rate": 5.010362135376423e-09,
+      "loss": 0.0552,
+      "step": 18154
+    },
+    {
+      "epoch": 2.9415100453661696,
+      "grad_norm": 1.128408432006836,
+      "learning_rate": 4.9827279970982024e-09,
+      "loss": 0.0682,
+      "step": 18155
+    },
+    {
+      "epoch": 2.9416720674011665,
+      "grad_norm": 0.9572305083274841,
+      "learning_rate": 4.955170200310888e-09,
+      "loss": 0.0565,
+      "step": 18156
+    },
+    {
+      "epoch": 2.9418340894361634,
+      "grad_norm": 0.8969508409500122,
+      "learning_rate": 4.927688745857417e-09,
+      "loss": 0.0601,
+      "step": 18157
+    },
+    {
+      "epoch": 2.94199611147116,
+      "grad_norm": 1.0484877824783325,
+      "learning_rate": 4.9002836345787845e-09,
+      "loss": 0.0628,
+      "step": 18158
+    },
+    {
+      "epoch": 2.942158133506157,
+      "grad_norm": 0.9466218948364258,
+      "learning_rate": 4.872954867313484e-09,
+      "loss": 0.0631,
+      "step": 18159
+    },
+    {
+      "epoch": 2.942320155541154,
+      "grad_norm": 1.0803978443145752,
+      "learning_rate": 4.845702444897515e-09,
+      "loss": 0.0597,
+      "step": 18160
+    },
+    {
+      "epoch": 2.9424821775761503,
+      "grad_norm": 0.8907843232154846,
+      "learning_rate": 4.818526368164933e-09,
+      "loss": 0.0571,
+      "step": 18161
+    },
+    {
+      "epoch": 2.942644199611147,
+      "grad_norm": 1.0460045337677002,
+      "learning_rate": 4.791426637947294e-09,
+      "loss": 0.0617,
+      "step": 18162
+    },
+    {
+      "epoch": 2.942806221646144,
+      "grad_norm": 0.893657922744751,
+      "learning_rate": 4.764403255073657e-09,
+      "loss": 0.0601,
+      "step": 18163
+    },
+    {
+      "epoch": 2.9429682436811406,
+      "grad_norm": 1.0240740776062012,
+      "learning_rate": 4.7374562203708615e-09,
+      "loss": 0.0714,
+      "step": 18164
+    },
+    {
+      "epoch": 2.943130265716137,
+      "grad_norm": 0.8384248614311218,
+      "learning_rate": 4.710585534663526e-09,
+      "loss": 0.0593,
+      "step": 18165
+    },
+    {
+      "epoch": 2.943292287751134,
+      "grad_norm": 0.8110920786857605,
+      "learning_rate": 4.683791198773768e-09,
+      "loss": 0.0551,
+      "step": 18166
+    },
+    {
+      "epoch": 2.943454309786131,
+      "grad_norm": 0.8798488974571228,
+      "learning_rate": 4.657073213521768e-09,
+      "loss": 0.0547,
+      "step": 18167
+    },
+    {
+      "epoch": 2.9436163318211275,
+      "grad_norm": 0.8243097066879272,
+      "learning_rate": 4.630431579724371e-09,
+      "loss": 0.0606,
+      "step": 18168
+    },
+    {
+      "epoch": 2.9437783538561244,
+      "grad_norm": 1.1285996437072754,
+      "learning_rate": 4.603866298197035e-09,
+      "loss": 0.0602,
+      "step": 18169
+    },
+    {
+      "epoch": 2.9439403758911213,
+      "grad_norm": 0.9858548641204834,
+      "learning_rate": 4.577377369752722e-09,
+      "loss": 0.0655,
+      "step": 18170
+    },
+    {
+      "epoch": 2.944102397926118,
+      "grad_norm": 1.0158785581588745,
+      "learning_rate": 4.550964795202173e-09,
+      "loss": 0.0605,
+      "step": 18171
+    },
+    {
+      "epoch": 2.9442644199611148,
+      "grad_norm": 0.9173128604888916,
+      "learning_rate": 4.524628575352796e-09,
+      "loss": 0.057,
+      "step": 18172
+    },
+    {
+      "epoch": 2.9444264419961117,
+      "grad_norm": 1.0110288858413696,
+      "learning_rate": 4.4983687110111695e-09,
+      "loss": 0.0644,
+      "step": 18173
+    },
+    {
+      "epoch": 2.944588464031108,
+      "grad_norm": 0.8735485076904297,
+      "learning_rate": 4.472185202980261e-09,
+      "loss": 0.0587,
+      "step": 18174
+    },
+    {
+      "epoch": 2.944750486066105,
+      "grad_norm": 0.8773369193077087,
+      "learning_rate": 4.446078052061375e-09,
+      "loss": 0.0586,
+      "step": 18175
+    },
+    {
+      "epoch": 2.9449125081011016,
+      "grad_norm": 1.5746303796768188,
+      "learning_rate": 4.420047259053595e-09,
+      "loss": 0.0564,
+      "step": 18176
+    },
+    {
+      "epoch": 2.9450745301360985,
+      "grad_norm": 0.9835705161094666,
+      "learning_rate": 4.394092824752949e-09,
+      "loss": 0.0608,
+      "step": 18177
+    },
+    {
+      "epoch": 2.945236552171095,
+      "grad_norm": 0.9602859020233154,
+      "learning_rate": 4.36821474995408e-09,
+      "loss": 0.06,
+      "step": 18178
+    },
+    {
+      "epoch": 2.945398574206092,
+      "grad_norm": 0.9771770238876343,
+      "learning_rate": 4.342413035448301e-09,
+      "loss": 0.0584,
+      "step": 18179
+    },
+    {
+      "epoch": 2.945560596241089,
+      "grad_norm": 0.8505083322525024,
+      "learning_rate": 4.316687682025256e-09,
+      "loss": 0.0548,
+      "step": 18180
+    },
+    {
+      "epoch": 2.9457226182760854,
+      "grad_norm": 0.8177760243415833,
+      "learning_rate": 4.291038690472371e-09,
+      "loss": 0.0592,
+      "step": 18181
+    },
+    {
+      "epoch": 2.9458846403110823,
+      "grad_norm": 0.9233116507530212,
+      "learning_rate": 4.26546606157402e-09,
+      "loss": 0.0594,
+      "step": 18182
+    },
+    {
+      "epoch": 2.9460466623460793,
+      "grad_norm": 0.8462489247322083,
+      "learning_rate": 4.2399697961129084e-09,
+      "loss": 0.0502,
+      "step": 18183
+    },
+    {
+      "epoch": 2.9462086843810757,
+      "grad_norm": 0.8417542576789856,
+      "learning_rate": 4.2145498948692465e-09,
+      "loss": 0.0603,
+      "step": 18184
+    },
+    {
+      "epoch": 2.9463707064160727,
+      "grad_norm": 0.844294011592865,
+      "learning_rate": 4.189206358620468e-09,
+      "loss": 0.0534,
+      "step": 18185
+    },
+    {
+      "epoch": 2.9465327284510696,
+      "grad_norm": 0.9389594197273254,
+      "learning_rate": 4.163939188142341e-09,
+      "loss": 0.0582,
+      "step": 18186
+    },
+    {
+      "epoch": 2.946694750486066,
+      "grad_norm": 0.9046799540519714,
+      "learning_rate": 4.138748384207858e-09,
+      "loss": 0.0595,
+      "step": 18187
+    },
+    {
+      "epoch": 2.9468567725210626,
+      "grad_norm": 0.9170434474945068,
+      "learning_rate": 4.113633947587792e-09,
+      "loss": 0.059,
+      "step": 18188
+    },
+    {
+      "epoch": 2.9470187945560595,
+      "grad_norm": 0.7925693392753601,
+      "learning_rate": 4.0885958790504165e-09,
+      "loss": 0.0539,
+      "step": 18189
+    },
+    {
+      "epoch": 2.9471808165910565,
+      "grad_norm": 0.9091478586196899,
+      "learning_rate": 4.063634179362341e-09,
+      "loss": 0.0521,
+      "step": 18190
+    },
+    {
+      "epoch": 2.947342838626053,
+      "grad_norm": 0.9072052836418152,
+      "learning_rate": 4.0387488492868446e-09,
+      "loss": 0.0637,
+      "step": 18191
+    },
+    {
+      "epoch": 2.94750486066105,
+      "grad_norm": 1.005933165550232,
+      "learning_rate": 4.013939889585538e-09,
+      "loss": 0.0623,
+      "step": 18192
+    },
+    {
+      "epoch": 2.947666882696047,
+      "grad_norm": 0.9582807421684265,
+      "learning_rate": 3.989207301017262e-09,
+      "loss": 0.0613,
+      "step": 18193
+    },
+    {
+      "epoch": 2.9478289047310433,
+      "grad_norm": 1.0312175750732422,
+      "learning_rate": 3.964551084339463e-09,
+      "loss": 0.0714,
+      "step": 18194
+    },
+    {
+      "epoch": 2.9479909267660402,
+      "grad_norm": 0.9463022947311401,
+      "learning_rate": 3.939971240305707e-09,
+      "loss": 0.0649,
+      "step": 18195
+    },
+    {
+      "epoch": 2.948152948801037,
+      "grad_norm": 0.9024858474731445,
+      "learning_rate": 3.915467769668724e-09,
+      "loss": 0.0549,
+      "step": 18196
+    },
+    {
+      "epoch": 2.9483149708360337,
+      "grad_norm": 0.8559155464172363,
+      "learning_rate": 3.891040673177915e-09,
+      "loss": 0.0564,
+      "step": 18197
+    },
+    {
+      "epoch": 2.9484769928710306,
+      "grad_norm": 0.804895281791687,
+      "learning_rate": 3.866689951580738e-09,
+      "loss": 0.0563,
+      "step": 18198
+    },
+    {
+      "epoch": 2.948639014906027,
+      "grad_norm": 0.9484009742736816,
+      "learning_rate": 3.842415605622429e-09,
+      "loss": 0.056,
+      "step": 18199
+    },
+    {
+      "epoch": 2.948801036941024,
+      "grad_norm": 0.813424289226532,
+      "learning_rate": 3.818217636045729e-09,
+      "loss": 0.0606,
+      "step": 18200
+    },
+    {
+      "epoch": 2.9489630589760205,
+      "grad_norm": 0.9334542751312256,
+      "learning_rate": 3.794096043590878e-09,
+      "loss": 0.0611,
+      "step": 18201
+    },
+    {
+      "epoch": 2.9491250810110174,
+      "grad_norm": 0.9005117416381836,
+      "learning_rate": 3.770050828995897e-09,
+      "loss": 0.0569,
+      "step": 18202
+    },
+    {
+      "epoch": 2.9492871030460144,
+      "grad_norm": 0.8461850881576538,
+      "learning_rate": 3.746081992996587e-09,
+      "loss": 0.0572,
+      "step": 18203
+    },
+    {
+      "epoch": 2.949449125081011,
+      "grad_norm": 0.8409425616264343,
+      "learning_rate": 3.7221895363262485e-09,
+      "loss": 0.0623,
+      "step": 18204
+    },
+    {
+      "epoch": 2.949611147116008,
+      "grad_norm": 1.0652631521224976,
+      "learning_rate": 3.6983734597162423e-09,
+      "loss": 0.0653,
+      "step": 18205
+    },
+    {
+      "epoch": 2.9497731691510047,
+      "grad_norm": 0.9219419956207275,
+      "learning_rate": 3.674633763894875e-09,
+      "loss": 0.0573,
+      "step": 18206
+    },
+    {
+      "epoch": 2.9499351911860012,
+      "grad_norm": 1.0178706645965576,
+      "learning_rate": 3.6509704495887866e-09,
+      "loss": 0.0608,
+      "step": 18207
+    },
+    {
+      "epoch": 2.950097213220998,
+      "grad_norm": 0.853714644908905,
+      "learning_rate": 3.6273835175221204e-09,
+      "loss": 0.0556,
+      "step": 18208
+    },
+    {
+      "epoch": 2.9502592352559946,
+      "grad_norm": 0.9474278688430786,
+      "learning_rate": 3.6038729684162445e-09,
+      "loss": 0.0633,
+      "step": 18209
+    },
+    {
+      "epoch": 2.9504212572909916,
+      "grad_norm": 0.962617039680481,
+      "learning_rate": 3.58043880299086e-09,
+      "loss": 0.0638,
+      "step": 18210
+    },
+    {
+      "epoch": 2.950583279325988,
+      "grad_norm": 0.8814941048622131,
+      "learning_rate": 3.557081021962616e-09,
+      "loss": 0.0634,
+      "step": 18211
+    },
+    {
+      "epoch": 2.950745301360985,
+      "grad_norm": 0.8792970776557922,
+      "learning_rate": 3.533799626046497e-09,
+      "loss": 0.0634,
+      "step": 18212
+    },
+    {
+      "epoch": 2.950907323395982,
+      "grad_norm": 0.8841227889060974,
+      "learning_rate": 3.5105946159549874e-09,
+      "loss": 0.0564,
+      "step": 18213
+    },
+    {
+      "epoch": 2.9510693454309784,
+      "grad_norm": 1.0366162061691284,
+      "learning_rate": 3.487465992397521e-09,
+      "loss": 0.0656,
+      "step": 18214
+    },
+    {
+      "epoch": 2.9512313674659754,
+      "grad_norm": 0.873475968837738,
+      "learning_rate": 3.4644137560824187e-09,
+      "loss": 0.0582,
+      "step": 18215
+    },
+    {
+      "epoch": 2.9513933895009723,
+      "grad_norm": 0.8530957698822021,
+      "learning_rate": 3.4414379077146733e-09,
+      "loss": 0.0568,
+      "step": 18216
+    },
+    {
+      "epoch": 2.951555411535969,
+      "grad_norm": 0.9540935158729553,
+      "learning_rate": 3.418538447997333e-09,
+      "loss": 0.0599,
+      "step": 18217
+    },
+    {
+      "epoch": 2.9517174335709657,
+      "grad_norm": 0.8702645301818848,
+      "learning_rate": 3.3957153776312257e-09,
+      "loss": 0.0599,
+      "step": 18218
+    },
+    {
+      "epoch": 2.9518794556059627,
+      "grad_norm": 0.9750285744667053,
+      "learning_rate": 3.372968697314405e-09,
+      "loss": 0.0584,
+      "step": 18219
+    },
+    {
+      "epoch": 2.952041477640959,
+      "grad_norm": 1.2663254737854004,
+      "learning_rate": 3.3502984077429803e-09,
+      "loss": 0.0627,
+      "step": 18220
+    },
+    {
+      "epoch": 2.952203499675956,
+      "grad_norm": 0.9789478778839111,
+      "learning_rate": 3.3277045096108405e-09,
+      "loss": 0.0574,
+      "step": 18221
+    },
+    {
+      "epoch": 2.9523655217109526,
+      "grad_norm": 1.000509262084961,
+      "learning_rate": 3.3051870036091004e-09,
+      "loss": 0.0558,
+      "step": 18222
+    },
+    {
+      "epoch": 2.9525275437459495,
+      "grad_norm": 0.9567583799362183,
+      "learning_rate": 3.282745890426653e-09,
+      "loss": 0.0611,
+      "step": 18223
+    },
+    {
+      "epoch": 2.952689565780946,
+      "grad_norm": 0.9223018288612366,
+      "learning_rate": 3.260381170750171e-09,
+      "loss": 0.0581,
+      "step": 18224
+    },
+    {
+      "epoch": 2.952851587815943,
+      "grad_norm": 0.8357469439506531,
+      "learning_rate": 3.238092845264107e-09,
+      "loss": 0.0543,
+      "step": 18225
+    },
+    {
+      "epoch": 2.95301360985094,
+      "grad_norm": 0.7842499017715454,
+      "learning_rate": 3.215880914650693e-09,
+      "loss": 0.0563,
+      "step": 18226
+    },
+    {
+      "epoch": 2.9531756318859363,
+      "grad_norm": 1.091303825378418,
+      "learning_rate": 3.1937453795888306e-09,
+      "loss": 0.0625,
+      "step": 18227
+    },
+    {
+      "epoch": 2.9533376539209333,
+      "grad_norm": 0.9732781052589417,
+      "learning_rate": 3.171686240756033e-09,
+      "loss": 0.0619,
+      "step": 18228
+    },
+    {
+      "epoch": 2.95349967595593,
+      "grad_norm": 0.8895050883293152,
+      "learning_rate": 3.1497034988278717e-09,
+      "loss": 0.0539,
+      "step": 18229
+    },
+    {
+      "epoch": 2.9536616979909267,
+      "grad_norm": 0.9890386462211609,
+      "learning_rate": 3.1277971544763088e-09,
+      "loss": 0.0618,
+      "step": 18230
+    },
+    {
+      "epoch": 2.9538237200259236,
+      "grad_norm": 0.9173012375831604,
+      "learning_rate": 3.1059672083719183e-09,
+      "loss": 0.056,
+      "step": 18231
+    },
+    {
+      "epoch": 2.95398574206092,
+      "grad_norm": 0.9382359385490417,
+      "learning_rate": 3.0842136611825004e-09,
+      "loss": 0.0637,
+      "step": 18232
+    },
+    {
+      "epoch": 2.954147764095917,
+      "grad_norm": 0.8800296187400818,
+      "learning_rate": 3.062536513573633e-09,
+      "loss": 0.0596,
+      "step": 18233
+    },
+    {
+      "epoch": 2.9543097861309136,
+      "grad_norm": 1.0513750314712524,
+      "learning_rate": 3.0409357662086748e-09,
+      "loss": 0.0661,
+      "step": 18234
+    },
+    {
+      "epoch": 2.9544718081659105,
+      "grad_norm": 0.8666124939918518,
+      "learning_rate": 3.0194114197487635e-09,
+      "loss": 0.0535,
+      "step": 18235
+    },
+    {
+      "epoch": 2.9546338302009074,
+      "grad_norm": 0.9290871024131775,
+      "learning_rate": 2.997963474852261e-09,
+      "loss": 0.0626,
+      "step": 18236
+    },
+    {
+      "epoch": 2.954795852235904,
+      "grad_norm": 0.9381429553031921,
+      "learning_rate": 2.97659193217531e-09,
+      "loss": 0.0614,
+      "step": 18237
+    },
+    {
+      "epoch": 2.954957874270901,
+      "grad_norm": 0.8524567484855652,
+      "learning_rate": 2.9552967923721086e-09,
+      "loss": 0.0595,
+      "step": 18238
+    },
+    {
+      "epoch": 2.9551198963058978,
+      "grad_norm": 0.7893286943435669,
+      "learning_rate": 2.9340780560938032e-09,
+      "loss": 0.0515,
+      "step": 18239
+    },
+    {
+      "epoch": 2.9552819183408943,
+      "grad_norm": 0.8815685510635376,
+      "learning_rate": 2.9129357239901514e-09,
+      "loss": 0.0554,
+      "step": 18240
+    },
+    {
+      "epoch": 2.955443940375891,
+      "grad_norm": 0.7954118251800537,
+      "learning_rate": 2.8918697967078578e-09,
+      "loss": 0.0561,
+      "step": 18241
+    },
+    {
+      "epoch": 2.955605962410888,
+      "grad_norm": 0.8579531908035278,
+      "learning_rate": 2.8708802748914077e-09,
+      "loss": 0.0591,
+      "step": 18242
+    },
+    {
+      "epoch": 2.9557679844458846,
+      "grad_norm": 0.9822569489479065,
+      "learning_rate": 2.849967159183065e-09,
+      "loss": 0.0587,
+      "step": 18243
+    },
+    {
+      "epoch": 2.9559300064808816,
+      "grad_norm": 0.9963696002960205,
+      "learning_rate": 2.829130450222872e-09,
+      "loss": 0.0605,
+      "step": 18244
+    },
+    {
+      "epoch": 2.956092028515878,
+      "grad_norm": 1.0176767110824585,
+      "learning_rate": 2.8083701486480985e-09,
+      "loss": 0.0586,
+      "step": 18245
+    },
+    {
+      "epoch": 2.956254050550875,
+      "grad_norm": 0.9688636064529419,
+      "learning_rate": 2.7876862550940685e-09,
+      "loss": 0.0648,
+      "step": 18246
+    },
+    {
+      "epoch": 2.9564160725858715,
+      "grad_norm": 1.1327593326568604,
+      "learning_rate": 2.7670787701938874e-09,
+      "loss": 0.0649,
+      "step": 18247
+    },
+    {
+      "epoch": 2.9565780946208684,
+      "grad_norm": 0.8865358829498291,
+      "learning_rate": 2.7465476945778835e-09,
+      "loss": 0.0517,
+      "step": 18248
+    },
+    {
+      "epoch": 2.9567401166558653,
+      "grad_norm": 0.9270232319831848,
+      "learning_rate": 2.726093028874166e-09,
+      "loss": 0.063,
+      "step": 18249
+    },
+    {
+      "epoch": 2.956902138690862,
+      "grad_norm": 0.8303735852241516,
+      "learning_rate": 2.705714773708623e-09,
+      "loss": 0.0583,
+      "step": 18250
+    },
+    {
+      "epoch": 2.9570641607258588,
+      "grad_norm": 0.8890751004219055,
+      "learning_rate": 2.6854129297049225e-09,
+      "loss": 0.0557,
+      "step": 18251
+    },
+    {
+      "epoch": 2.9572261827608557,
+      "grad_norm": 0.9399266242980957,
+      "learning_rate": 2.6651874974845115e-09,
+      "loss": 0.0563,
+      "step": 18252
+    },
+    {
+      "epoch": 2.957388204795852,
+      "grad_norm": 0.7995545268058777,
+      "learning_rate": 2.645038477665507e-09,
+      "loss": 0.0526,
+      "step": 18253
+    },
+    {
+      "epoch": 2.957550226830849,
+      "grad_norm": 0.9148885011672974,
+      "learning_rate": 2.6249658708651928e-09,
+      "loss": 0.0664,
+      "step": 18254
+    },
+    {
+      "epoch": 2.9577122488658456,
+      "grad_norm": 0.847109317779541,
+      "learning_rate": 2.6049696776972443e-09,
+      "loss": 0.0579,
+      "step": 18255
+    },
+    {
+      "epoch": 2.9578742709008425,
+      "grad_norm": 0.858539342880249,
+      "learning_rate": 2.5850498987733952e-09,
+      "loss": 0.0644,
+      "step": 18256
+    },
+    {
+      "epoch": 2.958036292935839,
+      "grad_norm": 1.0084835290908813,
+      "learning_rate": 2.5652065347037126e-09,
+      "loss": 0.0673,
+      "step": 18257
+    },
+    {
+      "epoch": 2.958198314970836,
+      "grad_norm": 0.8510595560073853,
+      "learning_rate": 2.545439586094933e-09,
+      "loss": 0.0584,
+      "step": 18258
+    },
+    {
+      "epoch": 2.958360337005833,
+      "grad_norm": 0.9628735780715942,
+      "learning_rate": 2.525749053552129e-09,
+      "loss": 0.0601,
+      "step": 18259
+    },
+    {
+      "epoch": 2.9585223590408294,
+      "grad_norm": 0.8464257121086121,
+      "learning_rate": 2.506134937677318e-09,
+      "loss": 0.0567,
+      "step": 18260
+    },
+    {
+      "epoch": 2.9586843810758263,
+      "grad_norm": 0.838660478591919,
+      "learning_rate": 2.4865972390711314e-09,
+      "loss": 0.0583,
+      "step": 18261
+    },
+    {
+      "epoch": 2.9588464031108233,
+      "grad_norm": 0.9245243668556213,
+      "learning_rate": 2.4671359583314237e-09,
+      "loss": 0.0618,
+      "step": 18262
+    },
+    {
+      "epoch": 2.9590084251458197,
+      "grad_norm": 0.8043138980865479,
+      "learning_rate": 2.4477510960532747e-09,
+      "loss": 0.0579,
+      "step": 18263
+    },
+    {
+      "epoch": 2.9591704471808167,
+      "grad_norm": 0.8946799039840698,
+      "learning_rate": 2.4284426528298212e-09,
+      "loss": 0.0584,
+      "step": 18264
+    },
+    {
+      "epoch": 2.9593324692158136,
+      "grad_norm": 1.0274814367294312,
+      "learning_rate": 2.409210629251979e-09,
+      "loss": 0.063,
+      "step": 18265
+    },
+    {
+      "epoch": 2.95949449125081,
+      "grad_norm": 0.8745455145835876,
+      "learning_rate": 2.3900550259084445e-09,
+      "loss": 0.0599,
+      "step": 18266
+    },
+    {
+      "epoch": 2.959656513285807,
+      "grad_norm": 0.9556804895401001,
+      "learning_rate": 2.370975843385137e-09,
+      "loss": 0.0605,
+      "step": 18267
+    },
+    {
+      "epoch": 2.9598185353208035,
+      "grad_norm": 0.9654064774513245,
+      "learning_rate": 2.351973082265757e-09,
+      "loss": 0.0604,
+      "step": 18268
+    },
+    {
+      "epoch": 2.9599805573558005,
+      "grad_norm": 0.9703561067581177,
+      "learning_rate": 2.333046743131784e-09,
+      "loss": 0.0632,
+      "step": 18269
+    },
+    {
+      "epoch": 2.960142579390797,
+      "grad_norm": 1.048510193824768,
+      "learning_rate": 2.314196826562476e-09,
+      "loss": 0.0593,
+      "step": 18270
+    },
+    {
+      "epoch": 2.960304601425794,
+      "grad_norm": 0.8689060211181641,
+      "learning_rate": 2.295423333134317e-09,
+      "loss": 0.0659,
+      "step": 18271
+    },
+    {
+      "epoch": 2.960466623460791,
+      "grad_norm": 0.9351786971092224,
+      "learning_rate": 2.2767262634218466e-09,
+      "loss": 0.0653,
+      "step": 18272
+    },
+    {
+      "epoch": 2.9606286454957873,
+      "grad_norm": 0.7980687022209167,
+      "learning_rate": 2.2581056179971084e-09,
+      "loss": 0.0542,
+      "step": 18273
+    },
+    {
+      "epoch": 2.9607906675307842,
+      "grad_norm": 0.8763124346733093,
+      "learning_rate": 2.239561397430201e-09,
+      "loss": 0.0586,
+      "step": 18274
+    },
+    {
+      "epoch": 2.960952689565781,
+      "grad_norm": 0.9670480489730835,
+      "learning_rate": 2.221093602288171e-09,
+      "loss": 0.0565,
+      "step": 18275
+    },
+    {
+      "epoch": 2.9611147116007777,
+      "grad_norm": 0.8312298655509949,
+      "learning_rate": 2.2027022331361226e-09,
+      "loss": 0.0575,
+      "step": 18276
+    },
+    {
+      "epoch": 2.9612767336357746,
+      "grad_norm": 0.9964354038238525,
+      "learning_rate": 2.18438729053666e-09,
+      "loss": 0.06,
+      "step": 18277
+    },
+    {
+      "epoch": 2.961438755670771,
+      "grad_norm": 0.9789435863494873,
+      "learning_rate": 2.1661487750504473e-09,
+      "loss": 0.0577,
+      "step": 18278
+    },
+    {
+      "epoch": 2.961600777705768,
+      "grad_norm": 0.977944016456604,
+      "learning_rate": 2.147986687235648e-09,
+      "loss": 0.0589,
+      "step": 18279
+    },
+    {
+      "epoch": 2.9617627997407645,
+      "grad_norm": 0.9870699644088745,
+      "learning_rate": 2.129901027647652e-09,
+      "loss": 0.0603,
+      "step": 18280
+    },
+    {
+      "epoch": 2.9619248217757614,
+      "grad_norm": 0.8196133375167847,
+      "learning_rate": 2.1118917968399045e-09,
+      "loss": 0.0516,
+      "step": 18281
+    },
+    {
+      "epoch": 2.9620868438107584,
+      "grad_norm": 1.1633763313293457,
+      "learning_rate": 2.0939589953633542e-09,
+      "loss": 0.0574,
+      "step": 18282
+    },
+    {
+      "epoch": 2.962248865845755,
+      "grad_norm": 1.174547791481018,
+      "learning_rate": 2.076102623767007e-09,
+      "loss": 0.0649,
+      "step": 18283
+    },
+    {
+      "epoch": 2.962410887880752,
+      "grad_norm": 0.8780858516693115,
+      "learning_rate": 2.0583226825970915e-09,
+      "loss": 0.0636,
+      "step": 18284
+    },
+    {
+      "epoch": 2.9625729099157487,
+      "grad_norm": 0.8388274908065796,
+      "learning_rate": 2.0406191723976175e-09,
+      "loss": 0.0627,
+      "step": 18285
+    },
+    {
+      "epoch": 2.962734931950745,
+      "grad_norm": 0.9338237047195435,
+      "learning_rate": 2.022992093710097e-09,
+      "loss": 0.0605,
+      "step": 18286
+    },
+    {
+      "epoch": 2.962896953985742,
+      "grad_norm": 0.8622428774833679,
+      "learning_rate": 2.005441447074097e-09,
+      "loss": 0.0602,
+      "step": 18287
+    },
+    {
+      "epoch": 2.963058976020739,
+      "grad_norm": 0.9662334322929382,
+      "learning_rate": 1.9879672330266886e-09,
+      "loss": 0.0603,
+      "step": 18288
+    },
+    {
+      "epoch": 2.9632209980557356,
+      "grad_norm": 1.1020525693893433,
+      "learning_rate": 1.9705694521021666e-09,
+      "loss": 0.0663,
+      "step": 18289
+    },
+    {
+      "epoch": 2.963383020090732,
+      "grad_norm": 0.82396000623703,
+      "learning_rate": 1.9532481048334383e-09,
+      "loss": 0.0535,
+      "step": 18290
+    },
+    {
+      "epoch": 2.963545042125729,
+      "grad_norm": 0.8906350135803223,
+      "learning_rate": 1.9360031917498024e-09,
+      "loss": 0.06,
+      "step": 18291
+    },
+    {
+      "epoch": 2.963707064160726,
+      "grad_norm": 0.845971405506134,
+      "learning_rate": 1.918834713379447e-09,
+      "loss": 0.0633,
+      "step": 18292
+    },
+    {
+      "epoch": 2.9638690861957224,
+      "grad_norm": 0.9976989030838013,
+      "learning_rate": 1.9017426702475084e-09,
+      "loss": 0.0593,
+      "step": 18293
+    },
+    {
+      "epoch": 2.9640311082307194,
+      "grad_norm": 0.9086126685142517,
+      "learning_rate": 1.884727062876901e-09,
+      "loss": 0.0592,
+      "step": 18294
+    },
+    {
+      "epoch": 2.9641931302657163,
+      "grad_norm": 1.0157880783081055,
+      "learning_rate": 1.867787891788597e-09,
+      "loss": 0.0634,
+      "step": 18295
+    },
+    {
+      "epoch": 2.964355152300713,
+      "grad_norm": 1.0585784912109375,
+      "learning_rate": 1.8509251575002386e-09,
+      "loss": 0.0652,
+      "step": 18296
+    },
+    {
+      "epoch": 2.9645171743357097,
+      "grad_norm": 0.8272886276245117,
+      "learning_rate": 1.8341388605283562e-09,
+      "loss": 0.0594,
+      "step": 18297
+    },
+    {
+      "epoch": 2.9646791963707066,
+      "grad_norm": 0.8589320182800293,
+      "learning_rate": 1.8174290013864282e-09,
+      "loss": 0.0577,
+      "step": 18298
+    },
+    {
+      "epoch": 2.964841218405703,
+      "grad_norm": 0.8368940949440002,
+      "learning_rate": 1.800795580585435e-09,
+      "loss": 0.0585,
+      "step": 18299
+    },
+    {
+      "epoch": 2.9650032404407,
+      "grad_norm": 0.8517768383026123,
+      "learning_rate": 1.784238598634691e-09,
+      "loss": 0.0569,
+      "step": 18300
+    },
+    {
+      "epoch": 2.9651652624756966,
+      "grad_norm": 1.1536654233932495,
+      "learning_rate": 1.767758056040736e-09,
+      "loss": 0.0648,
+      "step": 18301
+    },
+    {
+      "epoch": 2.9653272845106935,
+      "grad_norm": 1.0494126081466675,
+      "learning_rate": 1.7513539533078882e-09,
+      "loss": 0.0679,
+      "step": 18302
+    },
+    {
+      "epoch": 2.96548930654569,
+      "grad_norm": 0.8454523682594299,
+      "learning_rate": 1.735026290937969e-09,
+      "loss": 0.0558,
+      "step": 18303
+    },
+    {
+      "epoch": 2.965651328580687,
+      "grad_norm": 0.9818223714828491,
+      "learning_rate": 1.7187750694303007e-09,
+      "loss": 0.0593,
+      "step": 18304
+    },
+    {
+      "epoch": 2.965813350615684,
+      "grad_norm": 0.9108536839485168,
+      "learning_rate": 1.7026002892825411e-09,
+      "loss": 0.0591,
+      "step": 18305
+    },
+    {
+      "epoch": 2.9659753726506803,
+      "grad_norm": 0.8628289699554443,
+      "learning_rate": 1.686501950989572e-09,
+      "loss": 0.0569,
+      "step": 18306
+    },
+    {
+      "epoch": 2.9661373946856773,
+      "grad_norm": 0.9580399990081787,
+      "learning_rate": 1.6704800550434997e-09,
+      "loss": 0.0551,
+      "step": 18307
+    },
+    {
+      "epoch": 2.966299416720674,
+      "grad_norm": 0.9089958667755127,
+      "learning_rate": 1.6545346019350427e-09,
+      "loss": 0.0599,
+      "step": 18308
+    },
+    {
+      "epoch": 2.9664614387556707,
+      "grad_norm": 0.875546932220459,
+      "learning_rate": 1.6386655921521443e-09,
+      "loss": 0.061,
+      "step": 18309
+    },
+    {
+      "epoch": 2.9666234607906676,
+      "grad_norm": 1.0294671058654785,
+      "learning_rate": 1.6228730261799718e-09,
+      "loss": 0.0663,
+      "step": 18310
+    },
+    {
+      "epoch": 2.9667854828256646,
+      "grad_norm": 0.8314631581306458,
+      "learning_rate": 1.6071569045020274e-09,
+      "loss": 0.0512,
+      "step": 18311
+    },
+    {
+      "epoch": 2.966947504860661,
+      "grad_norm": 1.001203179359436,
+      "learning_rate": 1.5915172275990375e-09,
+      "loss": 0.0555,
+      "step": 18312
+    },
+    {
+      "epoch": 2.9671095268956575,
+      "grad_norm": 0.8834933638572693,
+      "learning_rate": 1.5759539959495085e-09,
+      "loss": 0.0627,
+      "step": 18313
+    },
+    {
+      "epoch": 2.9672715489306545,
+      "grad_norm": 1.0192748308181763,
+      "learning_rate": 1.5604672100297258e-09,
+      "loss": 0.064,
+      "step": 18314
+    },
+    {
+      "epoch": 2.9674335709656514,
+      "grad_norm": 0.9845191240310669,
+      "learning_rate": 1.5450568703137547e-09,
+      "loss": 0.0635,
+      "step": 18315
+    },
+    {
+      "epoch": 2.967595593000648,
+      "grad_norm": 0.7708606123924255,
+      "learning_rate": 1.5297229772726075e-09,
+      "loss": 0.0567,
+      "step": 18316
+    },
+    {
+      "epoch": 2.967757615035645,
+      "grad_norm": 0.8803342580795288,
+      "learning_rate": 1.5144655313759082e-09,
+      "loss": 0.0596,
+      "step": 18317
+    },
+    {
+      "epoch": 2.9679196370706418,
+      "grad_norm": 0.9847882986068726,
+      "learning_rate": 1.499284533090506e-09,
+      "loss": 0.0609,
+      "step": 18318
+    },
+    {
+      "epoch": 2.9680816591056383,
+      "grad_norm": 0.9394886493682861,
+      "learning_rate": 1.4841799828804737e-09,
+      "loss": 0.0655,
+      "step": 18319
+    },
+    {
+      "epoch": 2.968243681140635,
+      "grad_norm": 0.8597686886787415,
+      "learning_rate": 1.469151881208497e-09,
+      "loss": 0.0533,
+      "step": 18320
+    },
+    {
+      "epoch": 2.968405703175632,
+      "grad_norm": 0.9775875806808472,
+      "learning_rate": 1.4542002285339307e-09,
+      "loss": 0.0652,
+      "step": 18321
+    },
+    {
+      "epoch": 2.9685677252106286,
+      "grad_norm": 0.9064924716949463,
+      "learning_rate": 1.4393250253144642e-09,
+      "loss": 0.0605,
+      "step": 18322
+    },
+    {
+      "epoch": 2.9687297472456255,
+      "grad_norm": 0.788995087146759,
+      "learning_rate": 1.4245262720052888e-09,
+      "loss": 0.055,
+      "step": 18323
+    },
+    {
+      "epoch": 2.968891769280622,
+      "grad_norm": 0.86928391456604,
+      "learning_rate": 1.4098039690593756e-09,
+      "loss": 0.0589,
+      "step": 18324
+    },
+    {
+      "epoch": 2.969053791315619,
+      "grad_norm": 0.894902765750885,
+      "learning_rate": 1.3951581169266426e-09,
+      "loss": 0.0581,
+      "step": 18325
+    },
+    {
+      "epoch": 2.9692158133506155,
+      "grad_norm": 0.9499050974845886,
+      "learning_rate": 1.3805887160558973e-09,
+      "loss": 0.0611,
+      "step": 18326
+    },
+    {
+      "epoch": 2.9693778353856124,
+      "grad_norm": 0.896775484085083,
+      "learning_rate": 1.3660957668923391e-09,
+      "loss": 0.0545,
+      "step": 18327
+    },
+    {
+      "epoch": 2.9695398574206093,
+      "grad_norm": 1.0583899021148682,
+      "learning_rate": 1.3516792698797797e-09,
+      "loss": 0.0639,
+      "step": 18328
+    },
+    {
+      "epoch": 2.969701879455606,
+      "grad_norm": 0.9759207963943481,
+      "learning_rate": 1.3373392254592554e-09,
+      "loss": 0.0665,
+      "step": 18329
+    },
+    {
+      "epoch": 2.9698639014906028,
+      "grad_norm": 0.8481022715568542,
+      "learning_rate": 1.323075634069304e-09,
+      "loss": 0.0598,
+      "step": 18330
+    },
+    {
+      "epoch": 2.9700259235255997,
+      "grad_norm": 1.0313864946365356,
+      "learning_rate": 1.3088884961467985e-09,
+      "loss": 0.0618,
+      "step": 18331
+    },
+    {
+      "epoch": 2.970187945560596,
+      "grad_norm": 0.9152479767799377,
+      "learning_rate": 1.2947778121255584e-09,
+      "loss": 0.0561,
+      "step": 18332
+    },
+    {
+      "epoch": 2.970349967595593,
+      "grad_norm": 0.9564821720123291,
+      "learning_rate": 1.2807435824371827e-09,
+      "loss": 0.0579,
+      "step": 18333
+    },
+    {
+      "epoch": 2.9705119896305896,
+      "grad_norm": 0.9317049384117126,
+      "learning_rate": 1.2667858075113281e-09,
+      "loss": 0.0563,
+      "step": 18334
+    },
+    {
+      "epoch": 2.9706740116655865,
+      "grad_norm": 1.058647632598877,
+      "learning_rate": 1.2529044877751528e-09,
+      "loss": 0.0618,
+      "step": 18335
+    },
+    {
+      "epoch": 2.970836033700583,
+      "grad_norm": 0.8454841375350952,
+      "learning_rate": 1.239099623653317e-09,
+      "loss": 0.0543,
+      "step": 18336
+    },
+    {
+      "epoch": 2.97099805573558,
+      "grad_norm": 1.0585004091262817,
+      "learning_rate": 1.2253712155679832e-09,
+      "loss": 0.0597,
+      "step": 18337
+    },
+    {
+      "epoch": 2.971160077770577,
+      "grad_norm": 0.9348391890525818,
+      "learning_rate": 1.2117192639393704e-09,
+      "loss": 0.0573,
+      "step": 18338
+    },
+    {
+      "epoch": 2.9713220998055734,
+      "grad_norm": 0.8744611740112305,
+      "learning_rate": 1.1981437691852004e-09,
+      "loss": 0.0567,
+      "step": 18339
+    },
+    {
+      "epoch": 2.9714841218405703,
+      "grad_norm": 0.9730286598205566,
+      "learning_rate": 1.1846447317206967e-09,
+      "loss": 0.0593,
+      "step": 18340
+    },
+    {
+      "epoch": 2.9716461438755672,
+      "grad_norm": 1.0506782531738281,
+      "learning_rate": 1.1712221519594168e-09,
+      "loss": 0.0634,
+      "step": 18341
+    },
+    {
+      "epoch": 2.9718081659105637,
+      "grad_norm": 0.9589892625808716,
+      "learning_rate": 1.1578760303113113e-09,
+      "loss": 0.0554,
+      "step": 18342
+    },
+    {
+      "epoch": 2.9719701879455607,
+      "grad_norm": 0.8297640085220337,
+      "learning_rate": 1.1446063671854969e-09,
+      "loss": 0.0528,
+      "step": 18343
+    },
+    {
+      "epoch": 2.9721322099805576,
+      "grad_norm": 0.8789884448051453,
+      "learning_rate": 1.131413162987205e-09,
+      "loss": 0.0607,
+      "step": 18344
+    },
+    {
+      "epoch": 2.972294232015554,
+      "grad_norm": 1.033374309539795,
+      "learning_rate": 1.1182964181208344e-09,
+      "loss": 0.0625,
+      "step": 18345
+    },
+    {
+      "epoch": 2.972456254050551,
+      "grad_norm": 0.8781412839889526,
+      "learning_rate": 1.1052561329871757e-09,
+      "loss": 0.0567,
+      "step": 18346
+    },
+    {
+      "epoch": 2.9726182760855475,
+      "grad_norm": 1.0838392972946167,
+      "learning_rate": 1.0922923079856319e-09,
+      "loss": 0.0611,
+      "step": 18347
+    },
+    {
+      "epoch": 2.9727802981205445,
+      "grad_norm": 0.891605019569397,
+      "learning_rate": 1.0794049435128296e-09,
+      "loss": 0.0582,
+      "step": 18348
+    },
+    {
+      "epoch": 2.972942320155541,
+      "grad_norm": 0.9720555543899536,
+      "learning_rate": 1.0665940399626207e-09,
+      "loss": 0.0635,
+      "step": 18349
+    },
+    {
+      "epoch": 2.973104342190538,
+      "grad_norm": 0.8500351309776306,
+      "learning_rate": 1.0538595977277466e-09,
+      "loss": 0.0587,
+      "step": 18350
+    },
+    {
+      "epoch": 2.973266364225535,
+      "grad_norm": 1.0722614526748657,
+      "learning_rate": 1.041201617197063e-09,
+      "loss": 0.0624,
+      "step": 18351
+    },
+    {
+      "epoch": 2.9734283862605313,
+      "grad_norm": 0.9611464738845825,
+      "learning_rate": 1.028620098758315e-09,
+      "loss": 0.0588,
+      "step": 18352
+    },
+    {
+      "epoch": 2.9735904082955282,
+      "grad_norm": 1.0209434032440186,
+      "learning_rate": 1.0161150427964727e-09,
+      "loss": 0.0619,
+      "step": 18353
+    },
+    {
+      "epoch": 2.973752430330525,
+      "grad_norm": 0.9883117079734802,
+      "learning_rate": 1.0036864496942856e-09,
+      "loss": 0.0645,
+      "step": 18354
+    },
+    {
+      "epoch": 2.9739144523655217,
+      "grad_norm": 0.890546977519989,
+      "learning_rate": 9.913343198314495e-10,
+      "loss": 0.0606,
+      "step": 18355
+    },
+    {
+      "epoch": 2.9740764744005186,
+      "grad_norm": 0.9300897121429443,
+      "learning_rate": 9.79058653586551e-10,
+      "loss": 0.0638,
+      "step": 18356
+    },
+    {
+      "epoch": 2.974238496435515,
+      "grad_norm": 0.7656763792037964,
+      "learning_rate": 9.66859451334845e-10,
+      "loss": 0.0535,
+      "step": 18357
+    },
+    {
+      "epoch": 2.974400518470512,
+      "grad_norm": 0.9918314814567566,
+      "learning_rate": 9.54736713449922e-10,
+      "loss": 0.0602,
+      "step": 18358
+    },
+    {
+      "epoch": 2.9745625405055085,
+      "grad_norm": 0.8427473306655884,
+      "learning_rate": 9.426904403023185e-10,
+      "loss": 0.061,
+      "step": 18359
+    },
+    {
+      "epoch": 2.9747245625405054,
+      "grad_norm": 0.7689868211746216,
+      "learning_rate": 9.307206322606288e-10,
+      "loss": 0.0537,
+      "step": 18360
+    },
+    {
+      "epoch": 2.9748865845755024,
+      "grad_norm": 0.8835668563842773,
+      "learning_rate": 9.188272896915041e-10,
+      "loss": 0.0591,
+      "step": 18361
+    },
+    {
+      "epoch": 2.975048606610499,
+      "grad_norm": 0.8549810647964478,
+      "learning_rate": 9.070104129582647e-10,
+      "loss": 0.0575,
+      "step": 18362
+    },
+    {
+      "epoch": 2.975210628645496,
+      "grad_norm": 0.9912297129631042,
+      "learning_rate": 8.952700024231209e-10,
+      "loss": 0.063,
+      "step": 18363
+    },
+    {
+      "epoch": 2.9753726506804927,
+      "grad_norm": 0.8764829635620117,
+      "learning_rate": 8.836060584449524e-10,
+      "loss": 0.0609,
+      "step": 18364
+    },
+    {
+      "epoch": 2.975534672715489,
+      "grad_norm": 0.7048518061637878,
+      "learning_rate": 8.720185813806958e-10,
+      "loss": 0.0481,
+      "step": 18365
+    },
+    {
+      "epoch": 2.975696694750486,
+      "grad_norm": 0.97881680727005,
+      "learning_rate": 8.6050757158479e-10,
+      "loss": 0.0604,
+      "step": 18366
+    },
+    {
+      "epoch": 2.975858716785483,
+      "grad_norm": 0.9841058850288391,
+      "learning_rate": 8.490730294097304e-10,
+      "loss": 0.0615,
+      "step": 18367
+    },
+    {
+      "epoch": 2.9760207388204796,
+      "grad_norm": 0.8247631788253784,
+      "learning_rate": 8.377149552049602e-10,
+      "loss": 0.0558,
+      "step": 18368
+    },
+    {
+      "epoch": 2.9761827608554765,
+      "grad_norm": 0.9710003137588501,
+      "learning_rate": 8.264333493185339e-10,
+      "loss": 0.0596,
+      "step": 18369
+    },
+    {
+      "epoch": 2.976344782890473,
+      "grad_norm": 1.0644299983978271,
+      "learning_rate": 8.15228212095176e-10,
+      "loss": 0.0599,
+      "step": 18370
+    },
+    {
+      "epoch": 2.97650680492547,
+      "grad_norm": 0.9960538148880005,
+      "learning_rate": 8.040995438779453e-10,
+      "loss": 0.07,
+      "step": 18371
+    },
+    {
+      "epoch": 2.9766688269604664,
+      "grad_norm": 0.9092425107955933,
+      "learning_rate": 7.930473450074028e-10,
+      "loss": 0.0665,
+      "step": 18372
+    },
+    {
+      "epoch": 2.9768308489954634,
+      "grad_norm": 0.7984481453895569,
+      "learning_rate": 7.820716158216113e-10,
+      "loss": 0.0556,
+      "step": 18373
+    },
+    {
+      "epoch": 2.9769928710304603,
+      "grad_norm": 0.8725264668464661,
+      "learning_rate": 7.711723566564133e-10,
+      "loss": 0.0587,
+      "step": 18374
+    },
+    {
+      "epoch": 2.9771548930654568,
+      "grad_norm": 0.873121976852417,
+      "learning_rate": 7.603495678451534e-10,
+      "loss": 0.0517,
+      "step": 18375
+    },
+    {
+      "epoch": 2.9773169151004537,
+      "grad_norm": 1.0072768926620483,
+      "learning_rate": 7.496032497195105e-10,
+      "loss": 0.0708,
+      "step": 18376
+    },
+    {
+      "epoch": 2.9774789371354506,
+      "grad_norm": 0.875882625579834,
+      "learning_rate": 7.389334026078332e-10,
+      "loss": 0.052,
+      "step": 18377
+    },
+    {
+      "epoch": 2.977640959170447,
+      "grad_norm": 0.9897551536560059,
+      "learning_rate": 7.283400268365271e-10,
+      "loss": 0.057,
+      "step": 18378
+    },
+    {
+      "epoch": 2.977802981205444,
+      "grad_norm": 0.9639883637428284,
+      "learning_rate": 7.178231227300548e-10,
+      "loss": 0.0682,
+      "step": 18379
+    },
+    {
+      "epoch": 2.9779650032404406,
+      "grad_norm": 0.8938714861869812,
+      "learning_rate": 7.073826906098258e-10,
+      "loss": 0.0534,
+      "step": 18380
+    },
+    {
+      "epoch": 2.9781270252754375,
+      "grad_norm": 0.8408438563346863,
+      "learning_rate": 6.97018730795862e-10,
+      "loss": 0.0569,
+      "step": 18381
+    },
+    {
+      "epoch": 2.978289047310434,
+      "grad_norm": 0.9069212079048157,
+      "learning_rate": 6.867312436045769e-10,
+      "loss": 0.0674,
+      "step": 18382
+    },
+    {
+      "epoch": 2.978451069345431,
+      "grad_norm": 0.9260680675506592,
+      "learning_rate": 6.76520229351274e-10,
+      "loss": 0.0562,
+      "step": 18383
+    },
+    {
+      "epoch": 2.978613091380428,
+      "grad_norm": 1.0838451385498047,
+      "learning_rate": 6.663856883482034e-10,
+      "loss": 0.0593,
+      "step": 18384
+    },
+    {
+      "epoch": 2.9787751134154243,
+      "grad_norm": 1.0117061138153076,
+      "learning_rate": 6.563276209053948e-10,
+      "loss": 0.0628,
+      "step": 18385
+    },
+    {
+      "epoch": 2.9789371354504213,
+      "grad_norm": 1.174195647239685,
+      "learning_rate": 6.463460273306577e-10,
+      "loss": 0.0687,
+      "step": 18386
+    },
+    {
+      "epoch": 2.979099157485418,
+      "grad_norm": 0.8610493540763855,
+      "learning_rate": 6.364409079295808e-10,
+      "loss": 0.0569,
+      "step": 18387
+    },
+    {
+      "epoch": 2.9792611795204147,
+      "grad_norm": 0.8971181511878967,
+      "learning_rate": 6.266122630049776e-10,
+      "loss": 0.061,
+      "step": 18388
+    },
+    {
+      "epoch": 2.9794232015554116,
+      "grad_norm": 1.0571962594985962,
+      "learning_rate": 6.168600928577184e-10,
+      "loss": 0.0602,
+      "step": 18389
+    },
+    {
+      "epoch": 2.9795852235904086,
+      "grad_norm": 0.8279181718826294,
+      "learning_rate": 6.071843977861758e-10,
+      "loss": 0.0559,
+      "step": 18390
+    },
+    {
+      "epoch": 2.979747245625405,
+      "grad_norm": 0.9802617430686951,
+      "learning_rate": 5.975851780862241e-10,
+      "loss": 0.0636,
+      "step": 18391
+    },
+    {
+      "epoch": 2.9799092676604015,
+      "grad_norm": 1.0149646997451782,
+      "learning_rate": 5.880624340517948e-10,
+      "loss": 0.0606,
+      "step": 18392
+    },
+    {
+      "epoch": 2.9800712896953985,
+      "grad_norm": 0.8134593367576599,
+      "learning_rate": 5.786161659740441e-10,
+      "loss": 0.0535,
+      "step": 18393
+    },
+    {
+      "epoch": 2.9802333117303954,
+      "grad_norm": 0.8985582590103149,
+      "learning_rate": 5.692463741424625e-10,
+      "loss": 0.0575,
+      "step": 18394
+    },
+    {
+      "epoch": 2.980395333765392,
+      "grad_norm": 1.0262095928192139,
+      "learning_rate": 5.5995305884321e-10,
+      "loss": 0.0618,
+      "step": 18395
+    },
+    {
+      "epoch": 2.980557355800389,
+      "grad_norm": 0.9016473293304443,
+      "learning_rate": 5.507362203607814e-10,
+      "loss": 0.0611,
+      "step": 18396
+    },
+    {
+      "epoch": 2.9807193778353858,
+      "grad_norm": 1.423275113105774,
+      "learning_rate": 5.415958589774506e-10,
+      "loss": 0.0656,
+      "step": 18397
+    },
+    {
+      "epoch": 2.9808813998703823,
+      "grad_norm": 1.001585602760315,
+      "learning_rate": 5.325319749727165e-10,
+      "loss": 0.059,
+      "step": 18398
+    },
+    {
+      "epoch": 2.981043421905379,
+      "grad_norm": 1.0050073862075806,
+      "learning_rate": 5.235445686238572e-10,
+      "loss": 0.0532,
+      "step": 18399
+    },
+    {
+      "epoch": 2.981205443940376,
+      "grad_norm": 1.142298698425293,
+      "learning_rate": 5.146336402059304e-10,
+      "loss": 0.0697,
+      "step": 18400
+    },
+    {
+      "epoch": 2.9813674659753726,
+      "grad_norm": 0.8341407775878906,
+      "learning_rate": 5.057991899917735e-10,
+      "loss": 0.0491,
+      "step": 18401
+    },
+    {
+      "epoch": 2.9815294880103695,
+      "grad_norm": 1.106696367263794,
+      "learning_rate": 4.970412182511708e-10,
+      "loss": 0.065,
+      "step": 18402
+    },
+    {
+      "epoch": 2.981691510045366,
+      "grad_norm": 0.9795176982879639,
+      "learning_rate": 4.883597252525185e-10,
+      "loss": 0.0658,
+      "step": 18403
+    },
+    {
+      "epoch": 2.981853532080363,
+      "grad_norm": 0.879523515701294,
+      "learning_rate": 4.797547112614376e-10,
+      "loss": 0.0624,
+      "step": 18404
+    },
+    {
+      "epoch": 2.9820155541153595,
+      "grad_norm": 0.9020334482192993,
+      "learning_rate": 4.712261765410509e-10,
+      "loss": 0.0638,
+      "step": 18405
+    },
+    {
+      "epoch": 2.9821775761503564,
+      "grad_norm": 0.8660544753074646,
+      "learning_rate": 4.627741213525383e-10,
+      "loss": 0.0582,
+      "step": 18406
+    },
+    {
+      "epoch": 2.9823395981853533,
+      "grad_norm": 0.871384859085083,
+      "learning_rate": 4.5439854595430435e-10,
+      "loss": 0.0613,
+      "step": 18407
+    },
+    {
+      "epoch": 2.98250162022035,
+      "grad_norm": 0.9486377835273743,
+      "learning_rate": 4.460994506028105e-10,
+      "loss": 0.0588,
+      "step": 18408
+    },
+    {
+      "epoch": 2.9826636422553467,
+      "grad_norm": 0.9920843839645386,
+      "learning_rate": 4.378768355514651e-10,
+      "loss": 0.0621,
+      "step": 18409
+    },
+    {
+      "epoch": 2.9828256642903437,
+      "grad_norm": 1.2648646831512451,
+      "learning_rate": 4.2973070105256643e-10,
+      "loss": 0.066,
+      "step": 18410
+    },
+    {
+      "epoch": 2.98298768632534,
+      "grad_norm": 0.9957455396652222,
+      "learning_rate": 4.2166104735508197e-10,
+      "loss": 0.0688,
+      "step": 18411
+    },
+    {
+      "epoch": 2.983149708360337,
+      "grad_norm": 0.8266428709030151,
+      "learning_rate": 4.136678747060363e-10,
+      "loss": 0.0555,
+      "step": 18412
+    },
+    {
+      "epoch": 2.983311730395334,
+      "grad_norm": 0.8543840050697327,
+      "learning_rate": 4.0575118334967854e-10,
+      "loss": 0.0543,
+      "step": 18413
+    },
+    {
+      "epoch": 2.9834737524303305,
+      "grad_norm": 0.8616589307785034,
+      "learning_rate": 3.9791097352831487e-10,
+      "loss": 0.0617,
+      "step": 18414
+    },
+    {
+      "epoch": 2.983635774465327,
+      "grad_norm": 1.0665658712387085,
+      "learning_rate": 3.90147245482031e-10,
+      "loss": 0.0605,
+      "step": 18415
+    },
+    {
+      "epoch": 2.983797796500324,
+      "grad_norm": 0.7606042623519897,
+      "learning_rate": 3.824599994484146e-10,
+      "loss": 0.0514,
+      "step": 18416
+    },
+    {
+      "epoch": 2.983959818535321,
+      "grad_norm": 1.0913400650024414,
+      "learning_rate": 3.748492356625555e-10,
+      "loss": 0.066,
+      "step": 18417
+    },
+    {
+      "epoch": 2.9841218405703174,
+      "grad_norm": 0.9275001883506775,
+      "learning_rate": 3.673149543573229e-10,
+      "loss": 0.0614,
+      "step": 18418
+    },
+    {
+      "epoch": 2.9842838626053143,
+      "grad_norm": 0.902944803237915,
+      "learning_rate": 3.598571557630881e-10,
+      "loss": 0.0639,
+      "step": 18419
+    },
+    {
+      "epoch": 2.9844458846403112,
+      "grad_norm": 0.9159849286079407,
+      "learning_rate": 3.5247584010827953e-10,
+      "loss": 0.0619,
+      "step": 18420
+    },
+    {
+      "epoch": 2.9846079066753077,
+      "grad_norm": 0.9164811372756958,
+      "learning_rate": 3.451710076188275e-10,
+      "loss": 0.059,
+      "step": 18421
+    },
+    {
+      "epoch": 2.9847699287103047,
+      "grad_norm": 0.8712190389633179,
+      "learning_rate": 3.3794265851816444e-10,
+      "loss": 0.0598,
+      "step": 18422
+    },
+    {
+      "epoch": 2.9849319507453016,
+      "grad_norm": 0.9287796020507812,
+      "learning_rate": 3.307907930272247e-10,
+      "loss": 0.0547,
+      "step": 18423
+    },
+    {
+      "epoch": 2.985093972780298,
+      "grad_norm": 1.1647449731826782,
+      "learning_rate": 3.237154113649998e-10,
+      "loss": 0.0669,
+      "step": 18424
+    },
+    {
+      "epoch": 2.985255994815295,
+      "grad_norm": 1.0105154514312744,
+      "learning_rate": 3.167165137479833e-10,
+      "loss": 0.055,
+      "step": 18425
+    },
+    {
+      "epoch": 2.9854180168502915,
+      "grad_norm": 0.9725024700164795,
+      "learning_rate": 3.0979410039017053e-10,
+      "loss": 0.0611,
+      "step": 18426
+    },
+    {
+      "epoch": 2.9855800388852884,
+      "grad_norm": 1.0524919033050537,
+      "learning_rate": 3.029481715038918e-10,
+      "loss": 0.0653,
+      "step": 18427
+    },
+    {
+      "epoch": 2.985742060920285,
+      "grad_norm": 1.088059425354004,
+      "learning_rate": 2.961787272978689e-10,
+      "loss": 0.0651,
+      "step": 18428
+    },
+    {
+      "epoch": 2.985904082955282,
+      "grad_norm": 0.8906368613243103,
+      "learning_rate": 2.8948576797971364e-10,
+      "loss": 0.0578,
+      "step": 18429
+    },
+    {
+      "epoch": 2.986066104990279,
+      "grad_norm": 0.9600349068641663,
+      "learning_rate": 2.828692937542621e-10,
+      "loss": 0.0644,
+      "step": 18430
+    },
+    {
+      "epoch": 2.9862281270252753,
+      "grad_norm": 0.9611825346946716,
+      "learning_rate": 2.7632930482385243e-10,
+      "loss": 0.0671,
+      "step": 18431
+    },
+    {
+      "epoch": 2.9863901490602722,
+      "grad_norm": 0.9350132346153259,
+      "learning_rate": 2.6986580138832487e-10,
+      "loss": 0.0597,
+      "step": 18432
+    },
+    {
+      "epoch": 2.986552171095269,
+      "grad_norm": 0.851214587688446,
+      "learning_rate": 2.634787836458541e-10,
+      "loss": 0.0572,
+      "step": 18433
+    },
+    {
+      "epoch": 2.9867141931302656,
+      "grad_norm": 1.1418606042861938,
+      "learning_rate": 2.571682517915619e-10,
+      "loss": 0.0617,
+      "step": 18434
+    },
+    {
+      "epoch": 2.9868762151652626,
+      "grad_norm": 0.9702183604240417,
+      "learning_rate": 2.5093420601862706e-10,
+      "loss": 0.0533,
+      "step": 18435
+    },
+    {
+      "epoch": 2.987038237200259,
+      "grad_norm": 0.8244792819023132,
+      "learning_rate": 2.447766465180079e-10,
+      "loss": 0.054,
+      "step": 18436
+    },
+    {
+      "epoch": 2.987200259235256,
+      "grad_norm": 1.0218993425369263,
+      "learning_rate": 2.386955734778873e-10,
+      "loss": 0.0629,
+      "step": 18437
+    },
+    {
+      "epoch": 2.9873622812702525,
+      "grad_norm": 0.794620156288147,
+      "learning_rate": 2.3269098708422754e-10,
+      "loss": 0.048,
+      "step": 18438
+    },
+    {
+      "epoch": 2.9875243033052494,
+      "grad_norm": 0.7882254719734192,
+      "learning_rate": 2.2676288752104814e-10,
+      "loss": 0.0506,
+      "step": 18439
+    },
+    {
+      "epoch": 2.9876863253402464,
+      "grad_norm": 1.0353506803512573,
+      "learning_rate": 2.2091127496959298e-10,
+      "loss": 0.0663,
+      "step": 18440
+    },
+    {
+      "epoch": 2.987848347375243,
+      "grad_norm": 0.8752200603485107,
+      "learning_rate": 2.1513614960888552e-10,
+      "loss": 0.0578,
+      "step": 18441
+    },
+    {
+      "epoch": 2.98801036941024,
+      "grad_norm": 0.8921242952346802,
+      "learning_rate": 2.0943751161545122e-10,
+      "loss": 0.0593,
+      "step": 18442
+    },
+    {
+      "epoch": 2.9881723914452367,
+      "grad_norm": 0.8822087049484253,
+      "learning_rate": 2.0381536116415025e-10,
+      "loss": 0.0559,
+      "step": 18443
+    },
+    {
+      "epoch": 2.988334413480233,
+      "grad_norm": 1.00393545627594,
+      "learning_rate": 1.98269698426512e-10,
+      "loss": 0.0679,
+      "step": 18444
+    },
+    {
+      "epoch": 2.98849643551523,
+      "grad_norm": 0.8842718005180359,
+      "learning_rate": 1.9280052357240065e-10,
+      "loss": 0.0635,
+      "step": 18445
+    },
+    {
+      "epoch": 2.988658457550227,
+      "grad_norm": 0.875261664390564,
+      "learning_rate": 1.8740783676945984e-10,
+      "loss": 0.0622,
+      "step": 18446
+    },
+    {
+      "epoch": 2.9888204795852236,
+      "grad_norm": 0.9737663865089417,
+      "learning_rate": 1.820916381820026e-10,
+      "loss": 0.0655,
+      "step": 18447
+    },
+    {
+      "epoch": 2.9889825016202205,
+      "grad_norm": 0.8469840884208679,
+      "learning_rate": 1.768519279732317e-10,
+      "loss": 0.059,
+      "step": 18448
+    },
+    {
+      "epoch": 2.989144523655217,
+      "grad_norm": 0.9026631116867065,
+      "learning_rate": 1.7168870630357437e-10,
+      "loss": 0.0613,
+      "step": 18449
+    },
+    {
+      "epoch": 2.989306545690214,
+      "grad_norm": 1.0350512266159058,
+      "learning_rate": 1.666019733306823e-10,
+      "loss": 0.0625,
+      "step": 18450
+    },
+    {
+      "epoch": 2.9894685677252104,
+      "grad_norm": 0.9230927228927612,
+      "learning_rate": 1.6159172920998667e-10,
+      "loss": 0.0618,
+      "step": 18451
+    },
+    {
+      "epoch": 2.9896305897602073,
+      "grad_norm": 0.933576226234436,
+      "learning_rate": 1.5665797409553097e-10,
+      "loss": 0.0561,
+      "step": 18452
+    },
+    {
+      "epoch": 2.9897926117952043,
+      "grad_norm": 0.8842147588729858,
+      "learning_rate": 1.5180070813747282e-10,
+      "loss": 0.0591,
+      "step": 18453
+    },
+    {
+      "epoch": 2.9899546338302008,
+      "grad_norm": 0.9423680305480957,
+      "learning_rate": 1.4701993148485972e-10,
+      "loss": 0.0578,
+      "step": 18454
+    },
+    {
+      "epoch": 2.9901166558651977,
+      "grad_norm": 0.9645872712135315,
+      "learning_rate": 1.4231564428424105e-10,
+      "loss": 0.0643,
+      "step": 18455
+    },
+    {
+      "epoch": 2.9902786779001946,
+      "grad_norm": 1.0968506336212158,
+      "learning_rate": 1.3768784667883562e-10,
+      "loss": 0.0557,
+      "step": 18456
+    },
+    {
+      "epoch": 2.990440699935191,
+      "grad_norm": 0.8200769424438477,
+      "learning_rate": 1.3313653881075195e-10,
+      "loss": 0.0579,
+      "step": 18457
+    },
+    {
+      "epoch": 2.990602721970188,
+      "grad_norm": 0.8951736688613892,
+      "learning_rate": 1.2866172081904548e-10,
+      "loss": 0.059,
+      "step": 18458
+    },
+    {
+      "epoch": 2.9907647440051845,
+      "grad_norm": 0.9866411685943604,
+      "learning_rate": 1.2426339284082877e-10,
+      "loss": 0.0585,
+      "step": 18459
+    },
+    {
+      "epoch": 2.9909267660401815,
+      "grad_norm": 0.8763075470924377,
+      "learning_rate": 1.1994155501071636e-10,
+      "loss": 0.0578,
+      "step": 18460
+    },
+    {
+      "epoch": 2.991088788075178,
+      "grad_norm": 0.9515385627746582,
+      "learning_rate": 1.1569620746054722e-10,
+      "loss": 0.0622,
+      "step": 18461
+    },
+    {
+      "epoch": 2.991250810110175,
+      "grad_norm": 0.9050562977790833,
+      "learning_rate": 1.1152735032077255e-10,
+      "loss": 0.0639,
+      "step": 18462
+    },
+    {
+      "epoch": 2.991412832145172,
+      "grad_norm": 0.997776985168457,
+      "learning_rate": 1.0743498371823535e-10,
+      "loss": 0.0671,
+      "step": 18463
+    },
+    {
+      "epoch": 2.9915748541801683,
+      "grad_norm": 0.8363643288612366,
+      "learning_rate": 1.0341910777894593e-10,
+      "loss": 0.0567,
+      "step": 18464
+    },
+    {
+      "epoch": 2.9917368762151653,
+      "grad_norm": 0.9793019890785217,
+      "learning_rate": 9.947972262502881e-11,
+      "loss": 0.0637,
+      "step": 18465
+    },
+    {
+      "epoch": 2.991898898250162,
+      "grad_norm": 0.83316969871521,
+      "learning_rate": 9.561682837777586e-11,
+      "loss": 0.0526,
+      "step": 18466
+    },
+    {
+      "epoch": 2.9920609202851587,
+      "grad_norm": 0.846834659576416,
+      "learning_rate": 9.183042515459317e-11,
+      "loss": 0.0572,
+      "step": 18467
+    },
+    {
+      "epoch": 2.9922229423201556,
+      "grad_norm": 0.9713281393051147,
+      "learning_rate": 8.812051307205416e-11,
+      "loss": 0.0651,
+      "step": 18468
+    },
+    {
+      "epoch": 2.9923849643551526,
+      "grad_norm": 0.8826894164085388,
+      "learning_rate": 8.448709224312402e-11,
+      "loss": 0.067,
+      "step": 18469
+    },
+    {
+      "epoch": 2.992546986390149,
+      "grad_norm": 1.1817312240600586,
+      "learning_rate": 8.093016277938015e-11,
+      "loss": 0.0597,
+      "step": 18470
+    },
+    {
+      "epoch": 2.992709008425146,
+      "grad_norm": 0.9918498992919922,
+      "learning_rate": 7.744972478962443e-11,
+      "loss": 0.0696,
+      "step": 18471
+    },
+    {
+      "epoch": 2.9928710304601425,
+      "grad_norm": 0.9315574765205383,
+      "learning_rate": 7.404577837988313e-11,
+      "loss": 0.0574,
+      "step": 18472
+    },
+    {
+      "epoch": 2.9930330524951394,
+      "grad_norm": 0.8410326838493347,
+      "learning_rate": 7.071832365479481e-11,
+      "loss": 0.0578,
+      "step": 18473
+    },
+    {
+      "epoch": 2.993195074530136,
+      "grad_norm": 1.0091389417648315,
+      "learning_rate": 6.746736071594484e-11,
+      "loss": 0.0696,
+      "step": 18474
+    },
+    {
+      "epoch": 2.993357096565133,
+      "grad_norm": 0.9494967460632324,
+      "learning_rate": 6.429288966297576e-11,
+      "loss": 0.0644,
+      "step": 18475
+    },
+    {
+      "epoch": 2.9935191186001298,
+      "grad_norm": 0.8485308885574341,
+      "learning_rate": 6.119491059303206e-11,
+      "loss": 0.0614,
+      "step": 18476
+    },
+    {
+      "epoch": 2.9936811406351262,
+      "grad_norm": 0.9347163438796997,
+      "learning_rate": 5.817342360048273e-11,
+      "loss": 0.0584,
+      "step": 18477
+    },
+    {
+      "epoch": 2.993843162670123,
+      "grad_norm": 0.8836044669151306,
+      "learning_rate": 5.522842877830892e-11,
+      "loss": 0.0605,
+      "step": 18478
+    },
+    {
+      "epoch": 2.99400518470512,
+      "grad_norm": 0.8401952981948853,
+      "learning_rate": 5.235992621616115e-11,
+      "loss": 0.0585,
+      "step": 18479
+    },
+    {
+      "epoch": 2.9941672067401166,
+      "grad_norm": 1.1539223194122314,
+      "learning_rate": 4.956791600230215e-11,
+      "loss": 0.0589,
+      "step": 18480
+    },
+    {
+      "epoch": 2.9943292287751135,
+      "grad_norm": 0.8261008262634277,
+      "learning_rate": 4.685239822166398e-11,
+      "loss": 0.0574,
+      "step": 18481
+    },
+    {
+      "epoch": 2.99449125081011,
+      "grad_norm": 0.8534218072891235,
+      "learning_rate": 4.4213372957790935e-11,
+      "loss": 0.0544,
+      "step": 18482
+    },
+    {
+      "epoch": 2.994653272845107,
+      "grad_norm": 0.8698385953903198,
+      "learning_rate": 4.165084029117417e-11,
+      "loss": 0.0586,
+      "step": 18483
+    },
+    {
+      "epoch": 2.9948152948801035,
+      "grad_norm": 1.0100245475769043,
+      "learning_rate": 3.9164800300084404e-11,
+      "loss": 0.0542,
+      "step": 18484
+    },
+    {
+      "epoch": 2.9949773169151004,
+      "grad_norm": 0.8979434967041016,
+      "learning_rate": 3.6755253060849484e-11,
+      "loss": 0.0589,
+      "step": 18485
+    },
+    {
+      "epoch": 2.9951393389500973,
+      "grad_norm": 0.9884781837463379,
+      "learning_rate": 3.442219864729923e-11,
+      "loss": 0.065,
+      "step": 18486
+    },
+    {
+      "epoch": 2.995301360985094,
+      "grad_norm": 1.027869701385498,
+      "learning_rate": 3.216563713048793e-11,
+      "loss": 0.0565,
+      "step": 18487
+    },
+    {
+      "epoch": 2.9954633830200907,
+      "grad_norm": 0.9127057790756226,
+      "learning_rate": 2.998556857952695e-11,
+      "loss": 0.06,
+      "step": 18488
+    },
+    {
+      "epoch": 2.9956254050550877,
+      "grad_norm": 0.9373353123664856,
+      "learning_rate": 2.7881993061307233e-11,
+      "loss": 0.0585,
+      "step": 18489
+    },
+    {
+      "epoch": 2.995787427090084,
+      "grad_norm": 1.021518588066101,
+      "learning_rate": 2.5854910639944165e-11,
+      "loss": 0.0629,
+      "step": 18490
+    },
+    {
+      "epoch": 2.995949449125081,
+      "grad_norm": 0.8520172238349915,
+      "learning_rate": 2.390432137761023e-11,
+      "loss": 0.0592,
+      "step": 18491
+    },
+    {
+      "epoch": 2.996111471160078,
+      "grad_norm": 0.7974404096603394,
+      "learning_rate": 2.203022533425747e-11,
+      "loss": 0.0554,
+      "step": 18492
+    },
+    {
+      "epoch": 2.9962734931950745,
+      "grad_norm": 0.8519773483276367,
+      "learning_rate": 2.023262256678482e-11,
+      "loss": 0.0582,
+      "step": 18493
+    },
+    {
+      "epoch": 2.9964355152300715,
+      "grad_norm": 0.904719352722168,
+      "learning_rate": 1.8511513130148317e-11,
+      "loss": 0.0618,
+      "step": 18494
+    },
+    {
+      "epoch": 2.996597537265068,
+      "grad_norm": 0.9865505695343018,
+      "learning_rate": 1.686689707736111e-11,
+      "loss": 0.0664,
+      "step": 18495
+    },
+    {
+      "epoch": 2.996759559300065,
+      "grad_norm": 0.965429425239563,
+      "learning_rate": 1.529877445866079e-11,
+      "loss": 0.0616,
+      "step": 18496
+    },
+    {
+      "epoch": 2.9969215813350614,
+      "grad_norm": 0.8916358351707458,
+      "learning_rate": 1.3807145322064508e-11,
+      "loss": 0.0618,
+      "step": 18497
+    },
+    {
+      "epoch": 2.9970836033700583,
+      "grad_norm": 0.8863734602928162,
+      "learning_rate": 1.2392009713091402e-11,
+      "loss": 0.0537,
+      "step": 18498
+    },
+    {
+      "epoch": 2.9972456254050552,
+      "grad_norm": 0.9996553659439087,
+      "learning_rate": 1.1053367674762617e-11,
+      "loss": 0.0576,
+      "step": 18499
+    },
+    {
+      "epoch": 2.9974076474400517,
+      "grad_norm": 0.8960807919502258,
+      "learning_rate": 9.791219248711515e-12,
+      "loss": 0.0567,
+      "step": 18500
+    },
+    {
+      "epoch": 2.9975696694750487,
+      "grad_norm": 0.9851888418197632,
+      "learning_rate": 8.605564472963235e-12,
+      "loss": 0.0613,
+      "step": 18501
+    },
+    {
+      "epoch": 2.9977316915100456,
+      "grad_norm": 1.055681824684143,
+      "learning_rate": 7.496403384155137e-12,
+      "loss": 0.0598,
+      "step": 18502
+    },
+    {
+      "epoch": 2.997893713545042,
+      "grad_norm": 0.9974924921989441,
+      "learning_rate": 6.463736015871469e-12,
+      "loss": 0.0644,
+      "step": 18503
+    },
+    {
+      "epoch": 2.998055735580039,
+      "grad_norm": 0.9039761424064636,
+      "learning_rate": 5.507562400308697e-12,
+      "loss": 0.0685,
+      "step": 18504
+    },
+    {
+      "epoch": 2.9982177576150355,
+      "grad_norm": 0.9217861890792847,
+      "learning_rate": 4.6278825660550645e-12,
+      "loss": 0.0597,
+      "step": 18505
+    },
+    {
+      "epoch": 2.9983797796500324,
+      "grad_norm": 0.8809846043586731,
+      "learning_rate": 3.8246965403110344e-12,
+      "loss": 0.0548,
+      "step": 18506
+    },
+    {
+      "epoch": 2.998541801685029,
+      "grad_norm": 0.8097039461135864,
+      "learning_rate": 3.098004347779071e-12,
+      "loss": 0.0532,
+      "step": 18507
+    },
+    {
+      "epoch": 2.998703823720026,
+      "grad_norm": 0.847151517868042,
+      "learning_rate": 2.4478060103860777e-12,
+      "loss": 0.0547,
+      "step": 18508
+    },
+    {
+      "epoch": 2.998865845755023,
+      "grad_norm": 0.8286563754081726,
+      "learning_rate": 1.8741015483936253e-12,
+      "loss": 0.0566,
+      "step": 18509
+    },
+    {
+      "epoch": 2.9990278677900193,
+      "grad_norm": 0.9556106925010681,
+      "learning_rate": 1.376890979287726e-12,
+      "loss": 0.0555,
+      "step": 18510
+    },
+    {
+      "epoch": 2.999189889825016,
+      "grad_norm": 0.7847691774368286,
+      "learning_rate": 9.56174318056391e-13,
+      "loss": 0.0503,
+      "step": 18511
+    },
+    {
+      "epoch": 2.999351911860013,
+      "grad_norm": 1.074299931526184,
+      "learning_rate": 6.119515774671847e-13,
+      "loss": 0.0612,
+      "step": 18512
+    },
+    {
+      "epoch": 2.9995139338950096,
+      "grad_norm": 0.9344704747200012,
+      "learning_rate": 3.442227686223376e-13,
+      "loss": 0.0638,
+      "step": 18513
+    },
+    {
+      "epoch": 2.9996759559300066,
+      "grad_norm": 1.0170155763626099,
+      "learning_rate": 1.529878990158551e-13,
+      "loss": 0.0624,
+      "step": 18514
+    },
+    {
+      "epoch": 2.9998379779650035,
+      "grad_norm": 0.8656219840049744,
+      "learning_rate": 3.824697503151953e-14,
+      "loss": 0.0574,
+      "step": 18515
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.9637875556945801,
+      "learning_rate": 0.0,
+      "loss": 0.0613,
+      "step": 18516
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 18516,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.8941809923882746e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}