Nipa / trainer_state.json
Minchael's picture
Upload folder using huggingface_hub
653dff1 verified
Raw
History Blame Contribute Delete
186 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1072,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0009328358208955224,
"grad_norm": 4.623331097115583,
"learning_rate": 3.0303030303030305e-07,
"loss": 1.4919,
"step": 1
},
{
"epoch": 0.0018656716417910447,
"grad_norm": 4.982360852693379,
"learning_rate": 6.060606060606061e-07,
"loss": 1.4712,
"step": 2
},
{
"epoch": 0.002798507462686567,
"grad_norm": 4.571334201521145,
"learning_rate": 9.090909090909091e-07,
"loss": 1.5044,
"step": 3
},
{
"epoch": 0.0037313432835820895,
"grad_norm": 4.6058164590866895,
"learning_rate": 1.2121212121212122e-06,
"loss": 1.6696,
"step": 4
},
{
"epoch": 0.0046641791044776115,
"grad_norm": 4.682187183049512,
"learning_rate": 1.5151515151515152e-06,
"loss": 1.6968,
"step": 5
},
{
"epoch": 0.005597014925373134,
"grad_norm": 4.4099526844325645,
"learning_rate": 1.8181818181818183e-06,
"loss": 1.5847,
"step": 6
},
{
"epoch": 0.0065298507462686565,
"grad_norm": 4.651975444128333,
"learning_rate": 2.1212121212121216e-06,
"loss": 1.4428,
"step": 7
},
{
"epoch": 0.007462686567164179,
"grad_norm": 4.475877377017541,
"learning_rate": 2.4242424242424244e-06,
"loss": 1.604,
"step": 8
},
{
"epoch": 0.008395522388059701,
"grad_norm": 3.531627244041,
"learning_rate": 2.7272727272727272e-06,
"loss": 1.5281,
"step": 9
},
{
"epoch": 0.009328358208955223,
"grad_norm": 3.026154013401852,
"learning_rate": 3.0303030303030305e-06,
"loss": 1.6254,
"step": 10
},
{
"epoch": 0.010261194029850746,
"grad_norm": 2.5306366546848325,
"learning_rate": 3.3333333333333333e-06,
"loss": 1.543,
"step": 11
},
{
"epoch": 0.011194029850746268,
"grad_norm": 0.7830830708177523,
"learning_rate": 3.6363636363636366e-06,
"loss": 1.5326,
"step": 12
},
{
"epoch": 0.012126865671641791,
"grad_norm": 0.9126609732248719,
"learning_rate": 3.93939393939394e-06,
"loss": 1.6445,
"step": 13
},
{
"epoch": 0.013059701492537313,
"grad_norm": 0.7489268830197376,
"learning_rate": 4.242424242424243e-06,
"loss": 1.6683,
"step": 14
},
{
"epoch": 0.013992537313432836,
"grad_norm": 0.8481617349590951,
"learning_rate": 4.5454545454545455e-06,
"loss": 1.6811,
"step": 15
},
{
"epoch": 0.014925373134328358,
"grad_norm": 0.8612646174536928,
"learning_rate": 4.848484848484849e-06,
"loss": 1.6458,
"step": 16
},
{
"epoch": 0.01585820895522388,
"grad_norm": 0.9939599683563156,
"learning_rate": 5.151515151515152e-06,
"loss": 1.4456,
"step": 17
},
{
"epoch": 0.016791044776119403,
"grad_norm": 0.9223232693840682,
"learning_rate": 5.4545454545454545e-06,
"loss": 1.6541,
"step": 18
},
{
"epoch": 0.017723880597014924,
"grad_norm": 1.2351119204349634,
"learning_rate": 5.7575757575757586e-06,
"loss": 1.5795,
"step": 19
},
{
"epoch": 0.018656716417910446,
"grad_norm": 1.1426043898847984,
"learning_rate": 6.060606060606061e-06,
"loss": 1.5607,
"step": 20
},
{
"epoch": 0.01958955223880597,
"grad_norm": 1.3138707224029016,
"learning_rate": 6.363636363636364e-06,
"loss": 1.7212,
"step": 21
},
{
"epoch": 0.020522388059701493,
"grad_norm": 1.3476881569285653,
"learning_rate": 6.666666666666667e-06,
"loss": 1.5182,
"step": 22
},
{
"epoch": 0.021455223880597014,
"grad_norm": 1.1747712135103985,
"learning_rate": 6.969696969696971e-06,
"loss": 1.6133,
"step": 23
},
{
"epoch": 0.022388059701492536,
"grad_norm": 1.1026687894833662,
"learning_rate": 7.272727272727273e-06,
"loss": 1.5381,
"step": 24
},
{
"epoch": 0.02332089552238806,
"grad_norm": 1.0632773382340268,
"learning_rate": 7.5757575757575764e-06,
"loss": 1.578,
"step": 25
},
{
"epoch": 0.024253731343283583,
"grad_norm": 1.1101437273963808,
"learning_rate": 7.87878787878788e-06,
"loss": 1.4186,
"step": 26
},
{
"epoch": 0.025186567164179104,
"grad_norm": 1.098308717381125,
"learning_rate": 8.181818181818183e-06,
"loss": 1.4988,
"step": 27
},
{
"epoch": 0.026119402985074626,
"grad_norm": 1.031737418357339,
"learning_rate": 8.484848484848486e-06,
"loss": 1.5504,
"step": 28
},
{
"epoch": 0.027052238805970148,
"grad_norm": 1.0396912057014163,
"learning_rate": 8.787878787878788e-06,
"loss": 1.3785,
"step": 29
},
{
"epoch": 0.027985074626865673,
"grad_norm": 1.035021027504079,
"learning_rate": 9.090909090909091e-06,
"loss": 1.2845,
"step": 30
},
{
"epoch": 0.028917910447761194,
"grad_norm": 1.2405272996923293,
"learning_rate": 9.393939393939396e-06,
"loss": 1.1829,
"step": 31
},
{
"epoch": 0.029850746268656716,
"grad_norm": 1.3402273059559975,
"learning_rate": 9.696969696969698e-06,
"loss": 1.5617,
"step": 32
},
{
"epoch": 0.030783582089552237,
"grad_norm": 1.4809015830648808,
"learning_rate": 1e-05,
"loss": 1.3494,
"step": 33
},
{
"epoch": 0.03171641791044776,
"grad_norm": 1.6430652134394867,
"learning_rate": 9.999977143573675e-06,
"loss": 1.2355,
"step": 34
},
{
"epoch": 0.03264925373134328,
"grad_norm": 1.658273939478843,
"learning_rate": 9.999908574503663e-06,
"loss": 1.3639,
"step": 35
},
{
"epoch": 0.033582089552238806,
"grad_norm": 1.475689573597717,
"learning_rate": 9.999794293416862e-06,
"loss": 1.3803,
"step": 36
},
{
"epoch": 0.03451492537313433,
"grad_norm": 1.3948448920407295,
"learning_rate": 9.999634301358096e-06,
"loss": 1.3542,
"step": 37
},
{
"epoch": 0.03544776119402985,
"grad_norm": 1.1055350589637938,
"learning_rate": 9.999428599790104e-06,
"loss": 1.21,
"step": 38
},
{
"epoch": 0.036380597014925374,
"grad_norm": 1.0964623155536093,
"learning_rate": 9.999177190593526e-06,
"loss": 1.3331,
"step": 39
},
{
"epoch": 0.03731343283582089,
"grad_norm": 1.138702649832535,
"learning_rate": 9.998880076066888e-06,
"loss": 1.2336,
"step": 40
},
{
"epoch": 0.03824626865671642,
"grad_norm": 1.1602536807007948,
"learning_rate": 9.998537258926582e-06,
"loss": 1.3064,
"step": 41
},
{
"epoch": 0.03917910447761194,
"grad_norm": 1.1606903920376512,
"learning_rate": 9.998148742306837e-06,
"loss": 1.2213,
"step": 42
},
{
"epoch": 0.04011194029850746,
"grad_norm": 0.9554450926381938,
"learning_rate": 9.997714529759695e-06,
"loss": 1.1901,
"step": 43
},
{
"epoch": 0.041044776119402986,
"grad_norm": 1.0980822624201496,
"learning_rate": 9.99723462525497e-06,
"loss": 1.2791,
"step": 44
},
{
"epoch": 0.04197761194029851,
"grad_norm": 1.1336205018297356,
"learning_rate": 9.99670903318023e-06,
"loss": 1.315,
"step": 45
},
{
"epoch": 0.04291044776119403,
"grad_norm": 1.2438571622825052,
"learning_rate": 9.99613775834073e-06,
"loss": 1.2003,
"step": 46
},
{
"epoch": 0.043843283582089554,
"grad_norm": 1.5570563526741161,
"learning_rate": 9.995520805959397e-06,
"loss": 1.3063,
"step": 47
},
{
"epoch": 0.04477611940298507,
"grad_norm": 1.040804654408347,
"learning_rate": 9.994858181676761e-06,
"loss": 1.0705,
"step": 48
},
{
"epoch": 0.0457089552238806,
"grad_norm": 1.1446266838469341,
"learning_rate": 9.994149891550906e-06,
"loss": 1.0339,
"step": 49
},
{
"epoch": 0.04664179104477612,
"grad_norm": 1.4756627769149095,
"learning_rate": 9.99339594205743e-06,
"loss": 1.2113,
"step": 50
},
{
"epoch": 0.04757462686567164,
"grad_norm": 1.259849985029313,
"learning_rate": 9.992596340089365e-06,
"loss": 1.2258,
"step": 51
},
{
"epoch": 0.048507462686567165,
"grad_norm": 1.2956614391809322,
"learning_rate": 9.991751092957132e-06,
"loss": 1.0924,
"step": 52
},
{
"epoch": 0.049440298507462684,
"grad_norm": 1.3420708281759948,
"learning_rate": 9.99086020838846e-06,
"loss": 1.2384,
"step": 53
},
{
"epoch": 0.05037313432835821,
"grad_norm": 1.2783069282531003,
"learning_rate": 9.989923694528328e-06,
"loss": 1.237,
"step": 54
},
{
"epoch": 0.051305970149253734,
"grad_norm": 1.1964785607053976,
"learning_rate": 9.988941559938874e-06,
"loss": 1.234,
"step": 55
},
{
"epoch": 0.05223880597014925,
"grad_norm": 1.5076011818777009,
"learning_rate": 9.987913813599337e-06,
"loss": 1.1272,
"step": 56
},
{
"epoch": 0.05317164179104478,
"grad_norm": 1.4257157033309842,
"learning_rate": 9.986840464905958e-06,
"loss": 1.1892,
"step": 57
},
{
"epoch": 0.054104477611940295,
"grad_norm": 1.2314445136173628,
"learning_rate": 9.985721523671905e-06,
"loss": 1.1463,
"step": 58
},
{
"epoch": 0.05503731343283582,
"grad_norm": 1.2664691798672598,
"learning_rate": 9.984557000127177e-06,
"loss": 1.0108,
"step": 59
},
{
"epoch": 0.055970149253731345,
"grad_norm": 1.5682727332612316,
"learning_rate": 9.983346904918514e-06,
"loss": 1.1416,
"step": 60
},
{
"epoch": 0.05690298507462686,
"grad_norm": 1.2584808693683118,
"learning_rate": 9.982091249109292e-06,
"loss": 1.1738,
"step": 61
},
{
"epoch": 0.05783582089552239,
"grad_norm": 1.9941784528084425,
"learning_rate": 9.980790044179438e-06,
"loss": 1.1491,
"step": 62
},
{
"epoch": 0.058768656716417914,
"grad_norm": 1.7426892569648058,
"learning_rate": 9.979443302025305e-06,
"loss": 1.1863,
"step": 63
},
{
"epoch": 0.05970149253731343,
"grad_norm": 2.3866935427080316,
"learning_rate": 9.978051034959584e-06,
"loss": 1.0804,
"step": 64
},
{
"epoch": 0.06063432835820896,
"grad_norm": 1.4814993552160667,
"learning_rate": 9.976613255711169e-06,
"loss": 1.199,
"step": 65
},
{
"epoch": 0.061567164179104475,
"grad_norm": 2.1987363361630354,
"learning_rate": 9.975129977425062e-06,
"loss": 1.1756,
"step": 66
},
{
"epoch": 0.0625,
"grad_norm": 1.7300443575307491,
"learning_rate": 9.973601213662238e-06,
"loss": 1.3492,
"step": 67
},
{
"epoch": 0.06343283582089553,
"grad_norm": 1.9374465570694885,
"learning_rate": 9.972026978399527e-06,
"loss": 1.2828,
"step": 68
},
{
"epoch": 0.06436567164179105,
"grad_norm": 1.5316063167848697,
"learning_rate": 9.970407286029488e-06,
"loss": 1.0642,
"step": 69
},
{
"epoch": 0.06529850746268656,
"grad_norm": 1.9704284794142983,
"learning_rate": 9.96874215136027e-06,
"loss": 1.1965,
"step": 70
},
{
"epoch": 0.06623134328358209,
"grad_norm": 1.7100091683197662,
"learning_rate": 9.967031589615485e-06,
"loss": 1.0393,
"step": 71
},
{
"epoch": 0.06716417910447761,
"grad_norm": 1.646993364396328,
"learning_rate": 9.965275616434068e-06,
"loss": 1.0583,
"step": 72
},
{
"epoch": 0.06809701492537314,
"grad_norm": 1.7098440572928602,
"learning_rate": 9.963474247870122e-06,
"loss": 0.9654,
"step": 73
},
{
"epoch": 0.06902985074626866,
"grad_norm": 3.526199431595036,
"learning_rate": 9.961627500392788e-06,
"loss": 1.0218,
"step": 74
},
{
"epoch": 0.06996268656716417,
"grad_norm": 2.3375191039323076,
"learning_rate": 9.959735390886087e-06,
"loss": 1.2331,
"step": 75
},
{
"epoch": 0.0708955223880597,
"grad_norm": 1.7320808995339354,
"learning_rate": 9.957797936648762e-06,
"loss": 1.1667,
"step": 76
},
{
"epoch": 0.07182835820895522,
"grad_norm": 2.20959336972768,
"learning_rate": 9.955815155394127e-06,
"loss": 1.068,
"step": 77
},
{
"epoch": 0.07276119402985075,
"grad_norm": 3.655938124780615,
"learning_rate": 9.953787065249896e-06,
"loss": 0.9298,
"step": 78
},
{
"epoch": 0.07369402985074627,
"grad_norm": 1.752176732796501,
"learning_rate": 9.951713684758027e-06,
"loss": 1.1321,
"step": 79
},
{
"epoch": 0.07462686567164178,
"grad_norm": 1.8900374372083335,
"learning_rate": 9.94959503287455e-06,
"loss": 1.2267,
"step": 80
},
{
"epoch": 0.07555970149253731,
"grad_norm": 1.40826386921375,
"learning_rate": 9.947431128969388e-06,
"loss": 1.1507,
"step": 81
},
{
"epoch": 0.07649253731343283,
"grad_norm": 2.0612058870077106,
"learning_rate": 9.945221992826184e-06,
"loss": 1.13,
"step": 82
},
{
"epoch": 0.07742537313432836,
"grad_norm": 1.6728432435742655,
"learning_rate": 9.942967644642122e-06,
"loss": 1.0518,
"step": 83
},
{
"epoch": 0.07835820895522388,
"grad_norm": 2.04585984401706,
"learning_rate": 9.940668105027739e-06,
"loss": 1.0699,
"step": 84
},
{
"epoch": 0.07929104477611941,
"grad_norm": 2.2613028108311672,
"learning_rate": 9.938323395006738e-06,
"loss": 1.2273,
"step": 85
},
{
"epoch": 0.08022388059701492,
"grad_norm": 1.886970172402171,
"learning_rate": 9.935933536015796e-06,
"loss": 1.0182,
"step": 86
},
{
"epoch": 0.08115671641791045,
"grad_norm": 1.7210327196268325,
"learning_rate": 9.933498549904367e-06,
"loss": 1.1456,
"step": 87
},
{
"epoch": 0.08208955223880597,
"grad_norm": 1.6019281223765343,
"learning_rate": 9.931018458934483e-06,
"loss": 1.0957,
"step": 88
},
{
"epoch": 0.0830223880597015,
"grad_norm": 2.0598869196648026,
"learning_rate": 9.928493285780554e-06,
"loss": 1.1574,
"step": 89
},
{
"epoch": 0.08395522388059702,
"grad_norm": 1.829587044623219,
"learning_rate": 9.925923053529147e-06,
"loss": 1.0266,
"step": 90
},
{
"epoch": 0.08488805970149253,
"grad_norm": 1.9026921384653142,
"learning_rate": 9.923307785678798e-06,
"loss": 1.0742,
"step": 91
},
{
"epoch": 0.08582089552238806,
"grad_norm": 3.308998014443621,
"learning_rate": 9.920647506139774e-06,
"loss": 1.0584,
"step": 92
},
{
"epoch": 0.08675373134328358,
"grad_norm": 3.181939637762634,
"learning_rate": 9.917942239233872e-06,
"loss": 1.0856,
"step": 93
},
{
"epoch": 0.08768656716417911,
"grad_norm": 1.6857641583070804,
"learning_rate": 9.91519200969418e-06,
"loss": 0.9718,
"step": 94
},
{
"epoch": 0.08861940298507463,
"grad_norm": 1.6927218531466517,
"learning_rate": 9.912396842664869e-06,
"loss": 0.9848,
"step": 95
},
{
"epoch": 0.08955223880597014,
"grad_norm": 2.4760532702119473,
"learning_rate": 9.909556763700952e-06,
"loss": 0.9073,
"step": 96
},
{
"epoch": 0.09048507462686567,
"grad_norm": 2.76097279385013,
"learning_rate": 9.906671798768048e-06,
"loss": 1.1054,
"step": 97
},
{
"epoch": 0.0914179104477612,
"grad_norm": 2.4800181227532194,
"learning_rate": 9.903741974242153e-06,
"loss": 1.0943,
"step": 98
},
{
"epoch": 0.09235074626865672,
"grad_norm": 2.8791817641876567,
"learning_rate": 9.900767316909396e-06,
"loss": 1.2024,
"step": 99
},
{
"epoch": 0.09328358208955224,
"grad_norm": 2.3806272090834866,
"learning_rate": 9.897747853965792e-06,
"loss": 1.1172,
"step": 100
},
{
"epoch": 0.09421641791044776,
"grad_norm": 2.0401751602162057,
"learning_rate": 9.894683613016992e-06,
"loss": 1.0038,
"step": 101
},
{
"epoch": 0.09514925373134328,
"grad_norm": 1.9939968473502965,
"learning_rate": 9.891574622078035e-06,
"loss": 1.0242,
"step": 102
},
{
"epoch": 0.0960820895522388,
"grad_norm": 1.5906419548399748,
"learning_rate": 9.888420909573093e-06,
"loss": 1.093,
"step": 103
},
{
"epoch": 0.09701492537313433,
"grad_norm": 2.328618711595177,
"learning_rate": 9.885222504335199e-06,
"loss": 1.1711,
"step": 104
},
{
"epoch": 0.09794776119402986,
"grad_norm": 2.543789241213799,
"learning_rate": 9.881979435606004e-06,
"loss": 1.0522,
"step": 105
},
{
"epoch": 0.09888059701492537,
"grad_norm": 2.3077555647920263,
"learning_rate": 9.87869173303549e-06,
"loss": 1.0827,
"step": 106
},
{
"epoch": 0.09981343283582089,
"grad_norm": 2.8085629524802145,
"learning_rate": 9.875359426681712e-06,
"loss": 1.2257,
"step": 107
},
{
"epoch": 0.10074626865671642,
"grad_norm": 4.080074811896897,
"learning_rate": 9.871982547010512e-06,
"loss": 1.1325,
"step": 108
},
{
"epoch": 0.10167910447761194,
"grad_norm": 3.525137852609373,
"learning_rate": 9.868561124895254e-06,
"loss": 0.9509,
"step": 109
},
{
"epoch": 0.10261194029850747,
"grad_norm": 2.060191839459078,
"learning_rate": 9.86509519161653e-06,
"loss": 1.0463,
"step": 110
},
{
"epoch": 0.10354477611940298,
"grad_norm": 2.370226146823881,
"learning_rate": 9.861584778861879e-06,
"loss": 1.0441,
"step": 111
},
{
"epoch": 0.1044776119402985,
"grad_norm": 2.0558026864855483,
"learning_rate": 9.858029918725496e-06,
"loss": 0.8712,
"step": 112
},
{
"epoch": 0.10541044776119403,
"grad_norm": 2.3546315759778094,
"learning_rate": 9.854430643707944e-06,
"loss": 0.9987,
"step": 113
},
{
"epoch": 0.10634328358208955,
"grad_norm": 1.6009020203762028,
"learning_rate": 9.850786986715845e-06,
"loss": 1.0343,
"step": 114
},
{
"epoch": 0.10727611940298508,
"grad_norm": 2.1073706794362206,
"learning_rate": 9.847098981061593e-06,
"loss": 1.0723,
"step": 115
},
{
"epoch": 0.10820895522388059,
"grad_norm": 2.099824528934914,
"learning_rate": 9.84336666046304e-06,
"loss": 1.0451,
"step": 116
},
{
"epoch": 0.10914179104477612,
"grad_norm": 2.0856461411542564,
"learning_rate": 9.839590059043185e-06,
"loss": 1.0091,
"step": 117
},
{
"epoch": 0.11007462686567164,
"grad_norm": 2.0779177939133566,
"learning_rate": 9.835769211329877e-06,
"loss": 1.1118,
"step": 118
},
{
"epoch": 0.11100746268656717,
"grad_norm": 2.4144651841775446,
"learning_rate": 9.831904152255488e-06,
"loss": 1.193,
"step": 119
},
{
"epoch": 0.11194029850746269,
"grad_norm": 1.7710703476726444,
"learning_rate": 9.827994917156587e-06,
"loss": 1.0564,
"step": 120
},
{
"epoch": 0.11287313432835822,
"grad_norm": 1.9169143595161509,
"learning_rate": 9.824041541773639e-06,
"loss": 1.0704,
"step": 121
},
{
"epoch": 0.11380597014925373,
"grad_norm": 2.123955424059729,
"learning_rate": 9.820044062250651e-06,
"loss": 1.0216,
"step": 122
},
{
"epoch": 0.11473880597014925,
"grad_norm": 2.3381060064187653,
"learning_rate": 9.816002515134865e-06,
"loss": 0.9453,
"step": 123
},
{
"epoch": 0.11567164179104478,
"grad_norm": 1.604489297951601,
"learning_rate": 9.81191693737641e-06,
"loss": 0.9268,
"step": 124
},
{
"epoch": 0.1166044776119403,
"grad_norm": 1.810562916363095,
"learning_rate": 9.807787366327966e-06,
"loss": 1.0254,
"step": 125
},
{
"epoch": 0.11753731343283583,
"grad_norm": 1.6019925401429236,
"learning_rate": 9.80361383974443e-06,
"loss": 1.043,
"step": 126
},
{
"epoch": 0.11847014925373134,
"grad_norm": 1.8301994695576136,
"learning_rate": 9.799396395782566e-06,
"loss": 0.9354,
"step": 127
},
{
"epoch": 0.11940298507462686,
"grad_norm": 2.7542713561480263,
"learning_rate": 9.795135073000648e-06,
"loss": 1.167,
"step": 128
},
{
"epoch": 0.12033582089552239,
"grad_norm": 2.231557818577589,
"learning_rate": 9.790829910358122e-06,
"loss": 1.0162,
"step": 129
},
{
"epoch": 0.12126865671641791,
"grad_norm": 1.6961865784763224,
"learning_rate": 9.78648094721524e-06,
"loss": 0.9084,
"step": 130
},
{
"epoch": 0.12220149253731344,
"grad_norm": 2.0310453271301565,
"learning_rate": 9.782088223332706e-06,
"loss": 1.0897,
"step": 131
},
{
"epoch": 0.12313432835820895,
"grad_norm": 1.708672585928847,
"learning_rate": 9.777651778871309e-06,
"loss": 0.8224,
"step": 132
},
{
"epoch": 0.12406716417910447,
"grad_norm": 2.2023115732815337,
"learning_rate": 9.773171654391552e-06,
"loss": 1.0109,
"step": 133
},
{
"epoch": 0.125,
"grad_norm": 2.4511631720574347,
"learning_rate": 9.76864789085329e-06,
"loss": 1.1535,
"step": 134
},
{
"epoch": 0.1259328358208955,
"grad_norm": 1.8203861496583766,
"learning_rate": 9.764080529615352e-06,
"loss": 1.1892,
"step": 135
},
{
"epoch": 0.12686567164179105,
"grad_norm": 1.990633930022591,
"learning_rate": 9.759469612435159e-06,
"loss": 0.9757,
"step": 136
},
{
"epoch": 0.12779850746268656,
"grad_norm": 1.7436861573162867,
"learning_rate": 9.754815181468347e-06,
"loss": 0.9317,
"step": 137
},
{
"epoch": 0.1287313432835821,
"grad_norm": 1.993761658704838,
"learning_rate": 9.750117279268379e-06,
"loss": 0.999,
"step": 138
},
{
"epoch": 0.1296641791044776,
"grad_norm": 2.609235796598445,
"learning_rate": 9.745375948786158e-06,
"loss": 1.0582,
"step": 139
},
{
"epoch": 0.13059701492537312,
"grad_norm": 2.189630740891414,
"learning_rate": 9.74059123336963e-06,
"loss": 1.0529,
"step": 140
},
{
"epoch": 0.13152985074626866,
"grad_norm": 3.0572322880709493,
"learning_rate": 9.735763176763395e-06,
"loss": 0.9776,
"step": 141
},
{
"epoch": 0.13246268656716417,
"grad_norm": 4.4579121734576415,
"learning_rate": 9.730891823108302e-06,
"loss": 0.8224,
"step": 142
},
{
"epoch": 0.1333955223880597,
"grad_norm": 2.3219146900818384,
"learning_rate": 9.725977216941044e-06,
"loss": 1.107,
"step": 143
},
{
"epoch": 0.13432835820895522,
"grad_norm": 1.5476063052400744,
"learning_rate": 9.721019403193754e-06,
"loss": 0.9352,
"step": 144
},
{
"epoch": 0.13526119402985073,
"grad_norm": 1.967189037662775,
"learning_rate": 9.716018427193593e-06,
"loss": 1.0125,
"step": 145
},
{
"epoch": 0.13619402985074627,
"grad_norm": 3.5351798602227573,
"learning_rate": 9.710974334662343e-06,
"loss": 1.2184,
"step": 146
},
{
"epoch": 0.13712686567164178,
"grad_norm": 1.6646686771782597,
"learning_rate": 9.705887171715966e-06,
"loss": 0.9975,
"step": 147
},
{
"epoch": 0.13805970149253732,
"grad_norm": 2.35983108448416,
"learning_rate": 9.700756984864217e-06,
"loss": 0.9784,
"step": 148
},
{
"epoch": 0.13899253731343283,
"grad_norm": 1.7934212397910025,
"learning_rate": 9.695583821010185e-06,
"loss": 1.0577,
"step": 149
},
{
"epoch": 0.13992537313432835,
"grad_norm": 2.065375772525035,
"learning_rate": 9.690367727449889e-06,
"loss": 1.0138,
"step": 150
},
{
"epoch": 0.14085820895522388,
"grad_norm": 1.9010891532908825,
"learning_rate": 9.68510875187183e-06,
"loss": 0.9123,
"step": 151
},
{
"epoch": 0.1417910447761194,
"grad_norm": 2.7834751029164977,
"learning_rate": 9.679806942356564e-06,
"loss": 0.9653,
"step": 152
},
{
"epoch": 0.14272388059701493,
"grad_norm": 1.9663317960471303,
"learning_rate": 9.674462347376258e-06,
"loss": 1.0148,
"step": 153
},
{
"epoch": 0.14365671641791045,
"grad_norm": 2.095644677239545,
"learning_rate": 9.669075015794251e-06,
"loss": 1.0235,
"step": 154
},
{
"epoch": 0.14458955223880596,
"grad_norm": 2.0969936719001163,
"learning_rate": 9.663644996864599e-06,
"loss": 0.9598,
"step": 155
},
{
"epoch": 0.1455223880597015,
"grad_norm": 3.0427928790811265,
"learning_rate": 9.658172340231636e-06,
"loss": 0.8697,
"step": 156
},
{
"epoch": 0.146455223880597,
"grad_norm": 2.167164339282768,
"learning_rate": 9.652657095929508e-06,
"loss": 0.986,
"step": 157
},
{
"epoch": 0.14738805970149255,
"grad_norm": 2.896141578178114,
"learning_rate": 9.647099314381725e-06,
"loss": 1.0231,
"step": 158
},
{
"epoch": 0.14832089552238806,
"grad_norm": 2.8127936561708053,
"learning_rate": 9.6414990464007e-06,
"loss": 0.9793,
"step": 159
},
{
"epoch": 0.14925373134328357,
"grad_norm": 3.1199046389759304,
"learning_rate": 9.635856343187274e-06,
"loss": 0.9846,
"step": 160
},
{
"epoch": 0.1501865671641791,
"grad_norm": 2.322147358995242,
"learning_rate": 9.630171256330263e-06,
"loss": 0.9202,
"step": 161
},
{
"epoch": 0.15111940298507462,
"grad_norm": 3.926995677306718,
"learning_rate": 9.624443837805972e-06,
"loss": 0.856,
"step": 162
},
{
"epoch": 0.15205223880597016,
"grad_norm": 2.00446814758781,
"learning_rate": 9.618674139977731e-06,
"loss": 0.9153,
"step": 163
},
{
"epoch": 0.15298507462686567,
"grad_norm": 1.8724604300412608,
"learning_rate": 9.612862215595406e-06,
"loss": 1.0332,
"step": 164
},
{
"epoch": 0.15391791044776118,
"grad_norm": 2.009928500014447,
"learning_rate": 9.607008117794929e-06,
"loss": 1.0563,
"step": 165
},
{
"epoch": 0.15485074626865672,
"grad_norm": 2.2046221834819453,
"learning_rate": 9.6011119000978e-06,
"loss": 1.1383,
"step": 166
},
{
"epoch": 0.15578358208955223,
"grad_norm": 1.9399189538924912,
"learning_rate": 9.595173616410606e-06,
"loss": 0.9033,
"step": 167
},
{
"epoch": 0.15671641791044777,
"grad_norm": 2.3326968099593617,
"learning_rate": 9.589193321024525e-06,
"loss": 0.9787,
"step": 168
},
{
"epoch": 0.15764925373134328,
"grad_norm": 1.3559013021975819,
"learning_rate": 9.583171068614828e-06,
"loss": 1.0368,
"step": 169
},
{
"epoch": 0.15858208955223882,
"grad_norm": 2.349477279622384,
"learning_rate": 9.577106914240383e-06,
"loss": 1.0096,
"step": 170
},
{
"epoch": 0.15951492537313433,
"grad_norm": 1.7642782782278263,
"learning_rate": 9.571000913343149e-06,
"loss": 1.0168,
"step": 171
},
{
"epoch": 0.16044776119402984,
"grad_norm": 2.322236297791432,
"learning_rate": 9.56485312174767e-06,
"loss": 0.923,
"step": 172
},
{
"epoch": 0.16138059701492538,
"grad_norm": 1.7162327809199747,
"learning_rate": 9.558663595660563e-06,
"loss": 0.9361,
"step": 173
},
{
"epoch": 0.1623134328358209,
"grad_norm": 1.8143119672908885,
"learning_rate": 9.55243239167001e-06,
"loss": 0.9505,
"step": 174
},
{
"epoch": 0.16324626865671643,
"grad_norm": 1.5769678120648798,
"learning_rate": 9.54615956674523e-06,
"loss": 1.0512,
"step": 175
},
{
"epoch": 0.16417910447761194,
"grad_norm": 2.478618905991099,
"learning_rate": 9.539845178235967e-06,
"loss": 0.9235,
"step": 176
},
{
"epoch": 0.16511194029850745,
"grad_norm": 2.1039850929847805,
"learning_rate": 9.533489283871967e-06,
"loss": 0.9107,
"step": 177
},
{
"epoch": 0.166044776119403,
"grad_norm": 2.523459615927365,
"learning_rate": 9.527091941762438e-06,
"loss": 0.9992,
"step": 178
},
{
"epoch": 0.1669776119402985,
"grad_norm": 2.307174159686461,
"learning_rate": 9.520653210395534e-06,
"loss": 0.9855,
"step": 179
},
{
"epoch": 0.16791044776119404,
"grad_norm": 1.8315119143513632,
"learning_rate": 9.514173148637811e-06,
"loss": 0.995,
"step": 180
},
{
"epoch": 0.16884328358208955,
"grad_norm": 1.7885561978708464,
"learning_rate": 9.507651815733689e-06,
"loss": 0.9773,
"step": 181
},
{
"epoch": 0.16977611940298507,
"grad_norm": 2.0020183664498497,
"learning_rate": 9.501089271304917e-06,
"loss": 0.9709,
"step": 182
},
{
"epoch": 0.1707089552238806,
"grad_norm": 2.064309516233632,
"learning_rate": 9.494485575350015e-06,
"loss": 1.0244,
"step": 183
},
{
"epoch": 0.17164179104477612,
"grad_norm": 1.3914227726419255,
"learning_rate": 9.487840788243743e-06,
"loss": 0.9135,
"step": 184
},
{
"epoch": 0.17257462686567165,
"grad_norm": 1.6590157091887638,
"learning_rate": 9.481154970736537e-06,
"loss": 0.9874,
"step": 185
},
{
"epoch": 0.17350746268656717,
"grad_norm": 3.551271956693789,
"learning_rate": 9.474428183953952e-06,
"loss": 1.0356,
"step": 186
},
{
"epoch": 0.17444029850746268,
"grad_norm": 2.4118930971041164,
"learning_rate": 9.467660489396111e-06,
"loss": 1.1508,
"step": 187
},
{
"epoch": 0.17537313432835822,
"grad_norm": 2.0254037236254887,
"learning_rate": 9.46085194893714e-06,
"loss": 0.8872,
"step": 188
},
{
"epoch": 0.17630597014925373,
"grad_norm": 2.140405189372085,
"learning_rate": 9.454002624824599e-06,
"loss": 1.068,
"step": 189
},
{
"epoch": 0.17723880597014927,
"grad_norm": 2.3045997285979944,
"learning_rate": 9.44711257967892e-06,
"loss": 0.9524,
"step": 190
},
{
"epoch": 0.17817164179104478,
"grad_norm": 1.865403905813375,
"learning_rate": 9.44018187649282e-06,
"loss": 0.9426,
"step": 191
},
{
"epoch": 0.1791044776119403,
"grad_norm": 2.5947298919369133,
"learning_rate": 9.43321057863075e-06,
"loss": 1.1064,
"step": 192
},
{
"epoch": 0.18003731343283583,
"grad_norm": 2.3492341421351135,
"learning_rate": 9.426198749828287e-06,
"loss": 0.9569,
"step": 193
},
{
"epoch": 0.18097014925373134,
"grad_norm": 1.8925588653745464,
"learning_rate": 9.419146454191572e-06,
"loss": 0.9658,
"step": 194
},
{
"epoch": 0.18190298507462688,
"grad_norm": 2.7043466672136027,
"learning_rate": 9.412053756196715e-06,
"loss": 0.9402,
"step": 195
},
{
"epoch": 0.1828358208955224,
"grad_norm": 2.0138566825233637,
"learning_rate": 9.404920720689208e-06,
"loss": 0.8864,
"step": 196
},
{
"epoch": 0.1837686567164179,
"grad_norm": 2.091495247463038,
"learning_rate": 9.397747412883332e-06,
"loss": 0.8885,
"step": 197
},
{
"epoch": 0.18470149253731344,
"grad_norm": 1.9285112451282858,
"learning_rate": 9.390533898361556e-06,
"loss": 0.9167,
"step": 198
},
{
"epoch": 0.18563432835820895,
"grad_norm": 2.040083797186546,
"learning_rate": 9.383280243073949e-06,
"loss": 0.9345,
"step": 199
},
{
"epoch": 0.1865671641791045,
"grad_norm": 1.723952314577276,
"learning_rate": 9.375986513337565e-06,
"loss": 0.9979,
"step": 200
},
{
"epoch": 0.1875,
"grad_norm": 1.8095760255511828,
"learning_rate": 9.368652775835843e-06,
"loss": 1.0768,
"step": 201
},
{
"epoch": 0.1884328358208955,
"grad_norm": 1.4462126418115793,
"learning_rate": 9.361279097617993e-06,
"loss": 0.9146,
"step": 202
},
{
"epoch": 0.18936567164179105,
"grad_norm": 1.8368261621996567,
"learning_rate": 9.353865546098392e-06,
"loss": 0.9299,
"step": 203
},
{
"epoch": 0.19029850746268656,
"grad_norm": 1.9314517270088316,
"learning_rate": 9.346412189055955e-06,
"loss": 0.9867,
"step": 204
},
{
"epoch": 0.1912313432835821,
"grad_norm": 2.1324814081206833,
"learning_rate": 9.338919094633524e-06,
"loss": 0.9836,
"step": 205
},
{
"epoch": 0.1921641791044776,
"grad_norm": 1.8184695050995958,
"learning_rate": 9.331386331337246e-06,
"loss": 0.9877,
"step": 206
},
{
"epoch": 0.19309701492537312,
"grad_norm": 2.933429678892775,
"learning_rate": 9.323813968035938e-06,
"loss": 0.8239,
"step": 207
},
{
"epoch": 0.19402985074626866,
"grad_norm": 2.245164271135303,
"learning_rate": 9.316202073960467e-06,
"loss": 1.0364,
"step": 208
},
{
"epoch": 0.19496268656716417,
"grad_norm": 2.3318898657858456,
"learning_rate": 9.308550718703111e-06,
"loss": 1.0335,
"step": 209
},
{
"epoch": 0.1958955223880597,
"grad_norm": 2.245392642278823,
"learning_rate": 9.300859972216925e-06,
"loss": 0.9421,
"step": 210
},
{
"epoch": 0.19682835820895522,
"grad_norm": 1.5548786916387438,
"learning_rate": 9.2931299048151e-06,
"loss": 0.9184,
"step": 211
},
{
"epoch": 0.19776119402985073,
"grad_norm": 1.9528951378361923,
"learning_rate": 9.285360587170325e-06,
"loss": 1.0494,
"step": 212
},
{
"epoch": 0.19869402985074627,
"grad_norm": 2.4189426136463164,
"learning_rate": 9.277552090314135e-06,
"loss": 1.1506,
"step": 213
},
{
"epoch": 0.19962686567164178,
"grad_norm": 1.5616943661544846,
"learning_rate": 9.26970448563626e-06,
"loss": 0.9589,
"step": 214
},
{
"epoch": 0.20055970149253732,
"grad_norm": 3.177665424953949,
"learning_rate": 9.26181784488398e-06,
"loss": 0.955,
"step": 215
},
{
"epoch": 0.20149253731343283,
"grad_norm": 2.468126988505505,
"learning_rate": 9.253892240161467e-06,
"loss": 1.0144,
"step": 216
},
{
"epoch": 0.20242537313432835,
"grad_norm": 1.7973297516841404,
"learning_rate": 9.245927743929117e-06,
"loss": 0.967,
"step": 217
},
{
"epoch": 0.20335820895522388,
"grad_norm": 2.03124911054653,
"learning_rate": 9.237924429002902e-06,
"loss": 1.0787,
"step": 218
},
{
"epoch": 0.2042910447761194,
"grad_norm": 1.829770314666033,
"learning_rate": 9.229882368553692e-06,
"loss": 1.0683,
"step": 219
},
{
"epoch": 0.20522388059701493,
"grad_norm": 2.2091528675897276,
"learning_rate": 9.221801636106591e-06,
"loss": 0.8892,
"step": 220
},
{
"epoch": 0.20615671641791045,
"grad_norm": 1.6090699382243439,
"learning_rate": 9.213682305540268e-06,
"loss": 0.8335,
"step": 221
},
{
"epoch": 0.20708955223880596,
"grad_norm": 2.058858293918553,
"learning_rate": 9.205524451086274e-06,
"loss": 1.0616,
"step": 222
},
{
"epoch": 0.2080223880597015,
"grad_norm": 2.2150003152441142,
"learning_rate": 9.197328147328366e-06,
"loss": 0.9852,
"step": 223
},
{
"epoch": 0.208955223880597,
"grad_norm": 2.763235888971465,
"learning_rate": 9.189093469201833e-06,
"loss": 1.0235,
"step": 224
},
{
"epoch": 0.20988805970149255,
"grad_norm": 2.269320138276552,
"learning_rate": 9.180820491992799e-06,
"loss": 1.023,
"step": 225
},
{
"epoch": 0.21082089552238806,
"grad_norm": 1.9534213070497515,
"learning_rate": 9.172509291337542e-06,
"loss": 0.9222,
"step": 226
},
{
"epoch": 0.21175373134328357,
"grad_norm": 2.1612665075029756,
"learning_rate": 9.1641599432218e-06,
"loss": 0.9607,
"step": 227
},
{
"epoch": 0.2126865671641791,
"grad_norm": 2.0805954285258683,
"learning_rate": 9.155772523980075e-06,
"loss": 0.9319,
"step": 228
},
{
"epoch": 0.21361940298507462,
"grad_norm": 1.9877131069568363,
"learning_rate": 9.147347110294942e-06,
"loss": 0.9474,
"step": 229
},
{
"epoch": 0.21455223880597016,
"grad_norm": 2.0919535577926354,
"learning_rate": 9.138883779196338e-06,
"loss": 1.0033,
"step": 230
},
{
"epoch": 0.21548507462686567,
"grad_norm": 1.8767396854481073,
"learning_rate": 9.130382608060869e-06,
"loss": 0.886,
"step": 231
},
{
"epoch": 0.21641791044776118,
"grad_norm": 1.6699251558936072,
"learning_rate": 9.121843674611084e-06,
"loss": 0.9349,
"step": 232
},
{
"epoch": 0.21735074626865672,
"grad_norm": 2.2207505642954404,
"learning_rate": 9.11326705691479e-06,
"loss": 0.859,
"step": 233
},
{
"epoch": 0.21828358208955223,
"grad_norm": 2.3306641619719852,
"learning_rate": 9.104652833384317e-06,
"loss": 0.8588,
"step": 234
},
{
"epoch": 0.21921641791044777,
"grad_norm": 2.891674900731344,
"learning_rate": 9.096001082775812e-06,
"loss": 1.0341,
"step": 235
},
{
"epoch": 0.22014925373134328,
"grad_norm": 2.2160521300050102,
"learning_rate": 9.087311884188514e-06,
"loss": 0.9502,
"step": 236
},
{
"epoch": 0.22108208955223882,
"grad_norm": 2.16372146616088,
"learning_rate": 9.078585317064036e-06,
"loss": 0.8136,
"step": 237
},
{
"epoch": 0.22201492537313433,
"grad_norm": 1.8857188847190862,
"learning_rate": 9.069821461185631e-06,
"loss": 0.8954,
"step": 238
},
{
"epoch": 0.22294776119402984,
"grad_norm": 2.090168020722269,
"learning_rate": 9.06102039667747e-06,
"loss": 1.0462,
"step": 239
},
{
"epoch": 0.22388059701492538,
"grad_norm": 1.8536192892022236,
"learning_rate": 9.05218220400391e-06,
"loss": 1.0036,
"step": 240
},
{
"epoch": 0.2248134328358209,
"grad_norm": 2.0125369564595434,
"learning_rate": 9.043306963968746e-06,
"loss": 0.8521,
"step": 241
},
{
"epoch": 0.22574626865671643,
"grad_norm": 1.9102023168120987,
"learning_rate": 9.034394757714486e-06,
"loss": 0.9069,
"step": 242
},
{
"epoch": 0.22667910447761194,
"grad_norm": 2.3022257071988546,
"learning_rate": 9.025445666721608e-06,
"loss": 0.9178,
"step": 243
},
{
"epoch": 0.22761194029850745,
"grad_norm": 2.7687688603397347,
"learning_rate": 9.016459772807804e-06,
"loss": 0.9129,
"step": 244
},
{
"epoch": 0.228544776119403,
"grad_norm": 2.605616028099984,
"learning_rate": 9.007437158127245e-06,
"loss": 0.9363,
"step": 245
},
{
"epoch": 0.2294776119402985,
"grad_norm": 2.0962501496749133,
"learning_rate": 8.998377905169822e-06,
"loss": 1.0257,
"step": 246
},
{
"epoch": 0.23041044776119404,
"grad_norm": 2.599232721625264,
"learning_rate": 8.989282096760392e-06,
"loss": 0.8775,
"step": 247
},
{
"epoch": 0.23134328358208955,
"grad_norm": 3.486722452035239,
"learning_rate": 8.980149816058028e-06,
"loss": 0.9283,
"step": 248
},
{
"epoch": 0.23227611940298507,
"grad_norm": 2.4728817550913966,
"learning_rate": 8.970981146555247e-06,
"loss": 0.9831,
"step": 249
},
{
"epoch": 0.2332089552238806,
"grad_norm": 1.9030163618632787,
"learning_rate": 8.96177617207726e-06,
"loss": 0.912,
"step": 250
},
{
"epoch": 0.23414179104477612,
"grad_norm": 1.7460671180608425,
"learning_rate": 8.952534976781192e-06,
"loss": 0.7962,
"step": 251
},
{
"epoch": 0.23507462686567165,
"grad_norm": 1.8677246630068516,
"learning_rate": 8.943257645155328e-06,
"loss": 1.122,
"step": 252
},
{
"epoch": 0.23600746268656717,
"grad_norm": 3.1493127049065612,
"learning_rate": 8.93394426201832e-06,
"loss": 0.9272,
"step": 253
},
{
"epoch": 0.23694029850746268,
"grad_norm": 2.481845465700941,
"learning_rate": 8.924594912518434e-06,
"loss": 0.883,
"step": 254
},
{
"epoch": 0.23787313432835822,
"grad_norm": 1.9320126110426468,
"learning_rate": 8.91520968213276e-06,
"loss": 1.0268,
"step": 255
},
{
"epoch": 0.23880597014925373,
"grad_norm": 1.9208139211156101,
"learning_rate": 8.905788656666422e-06,
"loss": 0.9004,
"step": 256
},
{
"epoch": 0.23973880597014927,
"grad_norm": 1.9988736528574007,
"learning_rate": 8.896331922251814e-06,
"loss": 1.067,
"step": 257
},
{
"epoch": 0.24067164179104478,
"grad_norm": 1.591776467196882,
"learning_rate": 8.886839565347798e-06,
"loss": 0.8825,
"step": 258
},
{
"epoch": 0.2416044776119403,
"grad_norm": 2.0127327114460196,
"learning_rate": 8.877311672738912e-06,
"loss": 0.9677,
"step": 259
},
{
"epoch": 0.24253731343283583,
"grad_norm": 1.9950286523997451,
"learning_rate": 8.867748331534594e-06,
"loss": 1.0592,
"step": 260
},
{
"epoch": 0.24347014925373134,
"grad_norm": 1.645962767369582,
"learning_rate": 8.858149629168358e-06,
"loss": 0.8434,
"step": 261
},
{
"epoch": 0.24440298507462688,
"grad_norm": 1.6880970673344027,
"learning_rate": 8.84851565339702e-06,
"loss": 0.8338,
"step": 262
},
{
"epoch": 0.2453358208955224,
"grad_norm": 2.044547855939332,
"learning_rate": 8.838846492299884e-06,
"loss": 0.9467,
"step": 263
},
{
"epoch": 0.2462686567164179,
"grad_norm": 1.4790794616226546,
"learning_rate": 8.829142234277937e-06,
"loss": 0.782,
"step": 264
},
{
"epoch": 0.24720149253731344,
"grad_norm": 1.6229748443500185,
"learning_rate": 8.819402968053042e-06,
"loss": 0.9459,
"step": 265
},
{
"epoch": 0.24813432835820895,
"grad_norm": 2.5879188323457716,
"learning_rate": 8.809628782667126e-06,
"loss": 0.9055,
"step": 266
},
{
"epoch": 0.2490671641791045,
"grad_norm": 2.504414171815065,
"learning_rate": 8.79981976748137e-06,
"loss": 0.9977,
"step": 267
},
{
"epoch": 0.25,
"grad_norm": 1.9334666176986381,
"learning_rate": 8.789976012175387e-06,
"loss": 1.0395,
"step": 268
},
{
"epoch": 0.25093283582089554,
"grad_norm": 2.512156270984628,
"learning_rate": 8.780097606746405e-06,
"loss": 1.0137,
"step": 269
},
{
"epoch": 0.251865671641791,
"grad_norm": 3.5975180765676837,
"learning_rate": 8.770184641508438e-06,
"loss": 0.9284,
"step": 270
},
{
"epoch": 0.25279850746268656,
"grad_norm": 2.1814492163835215,
"learning_rate": 8.760237207091476e-06,
"loss": 1.0393,
"step": 271
},
{
"epoch": 0.2537313432835821,
"grad_norm": 2.064833117172027,
"learning_rate": 8.750255394440636e-06,
"loss": 0.9769,
"step": 272
},
{
"epoch": 0.25466417910447764,
"grad_norm": 2.250919299708563,
"learning_rate": 8.740239294815347e-06,
"loss": 0.9556,
"step": 273
},
{
"epoch": 0.2555970149253731,
"grad_norm": 1.5724008209128304,
"learning_rate": 8.7301889997885e-06,
"loss": 0.8722,
"step": 274
},
{
"epoch": 0.25652985074626866,
"grad_norm": 2.2121608038310563,
"learning_rate": 8.720104601245633e-06,
"loss": 0.9967,
"step": 275
},
{
"epoch": 0.2574626865671642,
"grad_norm": 2.1030845827991653,
"learning_rate": 8.70998619138407e-06,
"loss": 0.9816,
"step": 276
},
{
"epoch": 0.2583955223880597,
"grad_norm": 2.2568912793949423,
"learning_rate": 8.699833862712082e-06,
"loss": 0.98,
"step": 277
},
{
"epoch": 0.2593283582089552,
"grad_norm": 1.7084191830548623,
"learning_rate": 8.689647708048055e-06,
"loss": 0.8993,
"step": 278
},
{
"epoch": 0.26026119402985076,
"grad_norm": 1.8719899221024068,
"learning_rate": 8.679427820519624e-06,
"loss": 0.8837,
"step": 279
},
{
"epoch": 0.26119402985074625,
"grad_norm": 1.7765508873071,
"learning_rate": 8.669174293562833e-06,
"loss": 0.8079,
"step": 280
},
{
"epoch": 0.2621268656716418,
"grad_norm": 2.437870484337082,
"learning_rate": 8.658887220921274e-06,
"loss": 0.9467,
"step": 281
},
{
"epoch": 0.2630597014925373,
"grad_norm": 2.0167732153244042,
"learning_rate": 8.648566696645235e-06,
"loss": 0.9259,
"step": 282
},
{
"epoch": 0.26399253731343286,
"grad_norm": 1.8642239840548487,
"learning_rate": 8.638212815090836e-06,
"loss": 1.1184,
"step": 283
},
{
"epoch": 0.26492537313432835,
"grad_norm": 1.4079237547497494,
"learning_rate": 8.62782567091917e-06,
"loss": 0.8599,
"step": 284
},
{
"epoch": 0.2658582089552239,
"grad_norm": 1.7958771300433856,
"learning_rate": 8.617405359095437e-06,
"loss": 1.0577,
"step": 285
},
{
"epoch": 0.2667910447761194,
"grad_norm": 2.25228508654143,
"learning_rate": 8.60695197488807e-06,
"loss": 1.0558,
"step": 286
},
{
"epoch": 0.2677238805970149,
"grad_norm": 1.772555185517165,
"learning_rate": 8.596465613867873e-06,
"loss": 0.9224,
"step": 287
},
{
"epoch": 0.26865671641791045,
"grad_norm": 1.6508144490006655,
"learning_rate": 8.585946371907138e-06,
"loss": 0.9541,
"step": 288
},
{
"epoch": 0.269589552238806,
"grad_norm": 1.887287406309347,
"learning_rate": 8.575394345178782e-06,
"loss": 1.072,
"step": 289
},
{
"epoch": 0.27052238805970147,
"grad_norm": 1.429310047148272,
"learning_rate": 8.564809630155451e-06,
"loss": 0.948,
"step": 290
},
{
"epoch": 0.271455223880597,
"grad_norm": 1.6884617477334563,
"learning_rate": 8.55419232360865e-06,
"loss": 1.0353,
"step": 291
},
{
"epoch": 0.27238805970149255,
"grad_norm": 1.6255061845136924,
"learning_rate": 8.54354252260785e-06,
"loss": 0.938,
"step": 292
},
{
"epoch": 0.2733208955223881,
"grad_norm": 1.745741718742768,
"learning_rate": 8.532860324519608e-06,
"loss": 0.9495,
"step": 293
},
{
"epoch": 0.27425373134328357,
"grad_norm": 1.6147722089793846,
"learning_rate": 8.522145827006677e-06,
"loss": 0.9453,
"step": 294
},
{
"epoch": 0.2751865671641791,
"grad_norm": 1.906507804355843,
"learning_rate": 8.511399128027102e-06,
"loss": 0.9012,
"step": 295
},
{
"epoch": 0.27611940298507465,
"grad_norm": 2.431164575884372,
"learning_rate": 8.50062032583334e-06,
"loss": 0.9444,
"step": 296
},
{
"epoch": 0.27705223880597013,
"grad_norm": 1.7562757659193227,
"learning_rate": 8.489809518971348e-06,
"loss": 0.8541,
"step": 297
},
{
"epoch": 0.27798507462686567,
"grad_norm": 1.733001460510656,
"learning_rate": 8.478966806279692e-06,
"loss": 0.9295,
"step": 298
},
{
"epoch": 0.2789179104477612,
"grad_norm": 1.8970129456513103,
"learning_rate": 8.468092286888635e-06,
"loss": 0.8855,
"step": 299
},
{
"epoch": 0.2798507462686567,
"grad_norm": 1.6189355764214455,
"learning_rate": 8.45718606021924e-06,
"loss": 0.7952,
"step": 300
},
{
"epoch": 0.28078358208955223,
"grad_norm": 1.8782089217783915,
"learning_rate": 8.446248225982451e-06,
"loss": 1.0526,
"step": 301
},
{
"epoch": 0.28171641791044777,
"grad_norm": 1.85669864853271,
"learning_rate": 8.43527888417819e-06,
"loss": 0.8406,
"step": 302
},
{
"epoch": 0.2826492537313433,
"grad_norm": 2.0233731293922856,
"learning_rate": 8.424278135094442e-06,
"loss": 0.9898,
"step": 303
},
{
"epoch": 0.2835820895522388,
"grad_norm": 1.7506877023354923,
"learning_rate": 8.413246079306325e-06,
"loss": 0.9166,
"step": 304
},
{
"epoch": 0.28451492537313433,
"grad_norm": 2.6164694885781365,
"learning_rate": 8.40218281767519e-06,
"loss": 0.8588,
"step": 305
},
{
"epoch": 0.28544776119402987,
"grad_norm": 2.2081786711173534,
"learning_rate": 8.391088451347688e-06,
"loss": 0.9596,
"step": 306
},
{
"epoch": 0.28638059701492535,
"grad_norm": 2.2686483504606896,
"learning_rate": 8.379963081754846e-06,
"loss": 1.011,
"step": 307
},
{
"epoch": 0.2873134328358209,
"grad_norm": 1.7619205073101016,
"learning_rate": 8.368806810611137e-06,
"loss": 1.115,
"step": 308
},
{
"epoch": 0.28824626865671643,
"grad_norm": 1.855062981871054,
"learning_rate": 8.357619739913557e-06,
"loss": 0.7489,
"step": 309
},
{
"epoch": 0.2891791044776119,
"grad_norm": 2.4423739150764514,
"learning_rate": 8.346401971940693e-06,
"loss": 1.0022,
"step": 310
},
{
"epoch": 0.29011194029850745,
"grad_norm": 1.9388105492754617,
"learning_rate": 8.335153609251775e-06,
"loss": 0.9711,
"step": 311
},
{
"epoch": 0.291044776119403,
"grad_norm": 1.6908869608474741,
"learning_rate": 8.323874754685755e-06,
"loss": 0.8164,
"step": 312
},
{
"epoch": 0.29197761194029853,
"grad_norm": 1.698334579707066,
"learning_rate": 8.312565511360358e-06,
"loss": 0.8647,
"step": 313
},
{
"epoch": 0.292910447761194,
"grad_norm": 1.8576380540310609,
"learning_rate": 8.301225982671135e-06,
"loss": 0.8849,
"step": 314
},
{
"epoch": 0.29384328358208955,
"grad_norm": 2.411407068513981,
"learning_rate": 8.289856272290526e-06,
"loss": 0.9628,
"step": 315
},
{
"epoch": 0.2947761194029851,
"grad_norm": 2.14902824556323,
"learning_rate": 8.278456484166916e-06,
"loss": 0.8121,
"step": 316
},
{
"epoch": 0.2957089552238806,
"grad_norm": 1.6830177215781728,
"learning_rate": 8.267026722523668e-06,
"loss": 0.9127,
"step": 317
},
{
"epoch": 0.2966417910447761,
"grad_norm": 2.2335773892409083,
"learning_rate": 8.255567091858183e-06,
"loss": 0.8889,
"step": 318
},
{
"epoch": 0.29757462686567165,
"grad_norm": 1.917043431825934,
"learning_rate": 8.244077696940943e-06,
"loss": 0.9322,
"step": 319
},
{
"epoch": 0.29850746268656714,
"grad_norm": 2.8083628235911333,
"learning_rate": 8.232558642814555e-06,
"loss": 0.8618,
"step": 320
},
{
"epoch": 0.2994402985074627,
"grad_norm": 2.2830062180381683,
"learning_rate": 8.22101003479278e-06,
"loss": 0.9826,
"step": 321
},
{
"epoch": 0.3003731343283582,
"grad_norm": 1.863385974933436,
"learning_rate": 8.209431978459583e-06,
"loss": 0.7484,
"step": 322
},
{
"epoch": 0.30130597014925375,
"grad_norm": 1.9037251570357387,
"learning_rate": 8.197824579668162e-06,
"loss": 0.8393,
"step": 323
},
{
"epoch": 0.30223880597014924,
"grad_norm": 1.970350941287213,
"learning_rate": 8.186187944539974e-06,
"loss": 0.9493,
"step": 324
},
{
"epoch": 0.3031716417910448,
"grad_norm": 1.7207755907664881,
"learning_rate": 8.174522179463779e-06,
"loss": 0.8589,
"step": 325
},
{
"epoch": 0.3041044776119403,
"grad_norm": 1.913770719899056,
"learning_rate": 8.162827391094661e-06,
"loss": 0.9051,
"step": 326
},
{
"epoch": 0.3050373134328358,
"grad_norm": 2.128879522890074,
"learning_rate": 8.151103686353042e-06,
"loss": 0.9286,
"step": 327
},
{
"epoch": 0.30597014925373134,
"grad_norm": 1.6528929953416232,
"learning_rate": 8.139351172423724e-06,
"loss": 0.9118,
"step": 328
},
{
"epoch": 0.3069029850746269,
"grad_norm": 1.705144002071713,
"learning_rate": 8.12756995675489e-06,
"loss": 0.8845,
"step": 329
},
{
"epoch": 0.30783582089552236,
"grad_norm": 1.7976517389999764,
"learning_rate": 8.115760147057139e-06,
"loss": 0.849,
"step": 330
},
{
"epoch": 0.3087686567164179,
"grad_norm": 1.715727842212141,
"learning_rate": 8.103921851302486e-06,
"loss": 0.9495,
"step": 331
},
{
"epoch": 0.30970149253731344,
"grad_norm": 2.3645538013845697,
"learning_rate": 8.092055177723388e-06,
"loss": 1.1365,
"step": 332
},
{
"epoch": 0.310634328358209,
"grad_norm": 1.7530071119951403,
"learning_rate": 8.080160234811743e-06,
"loss": 0.9161,
"step": 333
},
{
"epoch": 0.31156716417910446,
"grad_norm": 2.5293950983882585,
"learning_rate": 8.068237131317904e-06,
"loss": 0.9196,
"step": 334
},
{
"epoch": 0.3125,
"grad_norm": 1.8880808404888578,
"learning_rate": 8.056285976249689e-06,
"loss": 0.8595,
"step": 335
},
{
"epoch": 0.31343283582089554,
"grad_norm": 1.7156563395679425,
"learning_rate": 8.044306878871376e-06,
"loss": 0.9858,
"step": 336
},
{
"epoch": 0.314365671641791,
"grad_norm": 2.038408228577418,
"learning_rate": 8.032299948702707e-06,
"loss": 0.9764,
"step": 337
},
{
"epoch": 0.31529850746268656,
"grad_norm": 1.8920869091847787,
"learning_rate": 8.020265295517885e-06,
"loss": 0.9897,
"step": 338
},
{
"epoch": 0.3162313432835821,
"grad_norm": 2.1084547627584405,
"learning_rate": 8.00820302934458e-06,
"loss": 0.9011,
"step": 339
},
{
"epoch": 0.31716417910447764,
"grad_norm": 1.6546851803017,
"learning_rate": 7.996113260462908e-06,
"loss": 0.8704,
"step": 340
},
{
"epoch": 0.3180970149253731,
"grad_norm": 1.8506283384473063,
"learning_rate": 7.983996099404435e-06,
"loss": 0.9303,
"step": 341
},
{
"epoch": 0.31902985074626866,
"grad_norm": 1.5917776557212344,
"learning_rate": 7.971851656951161e-06,
"loss": 0.9069,
"step": 342
},
{
"epoch": 0.3199626865671642,
"grad_norm": 1.7646698034565214,
"learning_rate": 7.959680044134507e-06,
"loss": 0.9495,
"step": 343
},
{
"epoch": 0.3208955223880597,
"grad_norm": 1.7362424976562112,
"learning_rate": 7.947481372234303e-06,
"loss": 0.9294,
"step": 344
},
{
"epoch": 0.3218283582089552,
"grad_norm": 1.686682523846604,
"learning_rate": 7.935255752777766e-06,
"loss": 0.8355,
"step": 345
},
{
"epoch": 0.32276119402985076,
"grad_norm": 1.7532452698781866,
"learning_rate": 7.923003297538483e-06,
"loss": 0.8755,
"step": 346
},
{
"epoch": 0.32369402985074625,
"grad_norm": 1.9972043755079667,
"learning_rate": 7.91072411853539e-06,
"loss": 0.8211,
"step": 347
},
{
"epoch": 0.3246268656716418,
"grad_norm": 2.1501352087687837,
"learning_rate": 7.898418328031752e-06,
"loss": 0.9662,
"step": 348
},
{
"epoch": 0.3255597014925373,
"grad_norm": 2.1265999284907458,
"learning_rate": 7.886086038534121e-06,
"loss": 1.0094,
"step": 349
},
{
"epoch": 0.32649253731343286,
"grad_norm": 1.8055986277182625,
"learning_rate": 7.873727362791327e-06,
"loss": 0.7899,
"step": 350
},
{
"epoch": 0.32742537313432835,
"grad_norm": 2.2023331183096593,
"learning_rate": 7.861342413793434e-06,
"loss": 1.0772,
"step": 351
},
{
"epoch": 0.3283582089552239,
"grad_norm": 2.2202439506484954,
"learning_rate": 7.848931304770708e-06,
"loss": 0.8794,
"step": 352
},
{
"epoch": 0.3292910447761194,
"grad_norm": 2.1649079820557056,
"learning_rate": 7.836494149192594e-06,
"loss": 0.9592,
"step": 353
},
{
"epoch": 0.3302238805970149,
"grad_norm": 1.9065020496052099,
"learning_rate": 7.824031060766662e-06,
"loss": 0.9679,
"step": 354
},
{
"epoch": 0.33115671641791045,
"grad_norm": 1.940252041161397,
"learning_rate": 7.811542153437575e-06,
"loss": 0.9213,
"step": 355
},
{
"epoch": 0.332089552238806,
"grad_norm": 2.5445959881108933,
"learning_rate": 7.799027541386051e-06,
"loss": 0.9476,
"step": 356
},
{
"epoch": 0.33302238805970147,
"grad_norm": 2.51964982347319,
"learning_rate": 7.786487339027816e-06,
"loss": 0.8827,
"step": 357
},
{
"epoch": 0.333955223880597,
"grad_norm": 2.202259661366704,
"learning_rate": 7.773921661012548e-06,
"loss": 1.0652,
"step": 358
},
{
"epoch": 0.33488805970149255,
"grad_norm": 1.9551688723481973,
"learning_rate": 7.76133062222285e-06,
"loss": 0.9138,
"step": 359
},
{
"epoch": 0.3358208955223881,
"grad_norm": 1.877406637888787,
"learning_rate": 7.748714337773179e-06,
"loss": 0.8569,
"step": 360
},
{
"epoch": 0.33675373134328357,
"grad_norm": 2.101276666768221,
"learning_rate": 7.736072923008808e-06,
"loss": 0.9315,
"step": 361
},
{
"epoch": 0.3376865671641791,
"grad_norm": 2.303078664855918,
"learning_rate": 7.72340649350476e-06,
"loss": 0.984,
"step": 362
},
{
"epoch": 0.33861940298507465,
"grad_norm": 1.7278437564029068,
"learning_rate": 7.710715165064766e-06,
"loss": 0.8641,
"step": 363
},
{
"epoch": 0.33955223880597013,
"grad_norm": 2.0453846803766993,
"learning_rate": 7.697999053720185e-06,
"loss": 0.8223,
"step": 364
},
{
"epoch": 0.34048507462686567,
"grad_norm": 2.068984375550709,
"learning_rate": 7.685258275728964e-06,
"loss": 0.9798,
"step": 365
},
{
"epoch": 0.3414179104477612,
"grad_norm": 1.6643321382011356,
"learning_rate": 7.672492947574565e-06,
"loss": 0.8928,
"step": 366
},
{
"epoch": 0.3423507462686567,
"grad_norm": 2.017626586317116,
"learning_rate": 7.659703185964902e-06,
"loss": 0.9495,
"step": 367
},
{
"epoch": 0.34328358208955223,
"grad_norm": 1.7031289587415466,
"learning_rate": 7.64688910783127e-06,
"loss": 0.9091,
"step": 368
},
{
"epoch": 0.34421641791044777,
"grad_norm": 2.176542644587759,
"learning_rate": 7.634050830327283e-06,
"loss": 1.0041,
"step": 369
},
{
"epoch": 0.3451492537313433,
"grad_norm": 2.0210306937739966,
"learning_rate": 7.621188470827798e-06,
"loss": 0.9553,
"step": 370
},
{
"epoch": 0.3460820895522388,
"grad_norm": 2.229747936074299,
"learning_rate": 7.608302146927846e-06,
"loss": 0.9718,
"step": 371
},
{
"epoch": 0.34701492537313433,
"grad_norm": 1.930909614739677,
"learning_rate": 7.5953919764415505e-06,
"loss": 0.8035,
"step": 372
},
{
"epoch": 0.34794776119402987,
"grad_norm": 1.9440423782079594,
"learning_rate": 7.582458077401056e-06,
"loss": 0.9574,
"step": 373
},
{
"epoch": 0.34888059701492535,
"grad_norm": 1.7615577130413262,
"learning_rate": 7.569500568055448e-06,
"loss": 0.9762,
"step": 374
},
{
"epoch": 0.3498134328358209,
"grad_norm": 2.085756679036281,
"learning_rate": 7.556519566869667e-06,
"loss": 0.9487,
"step": 375
},
{
"epoch": 0.35074626865671643,
"grad_norm": 1.637952494853108,
"learning_rate": 7.543515192523435e-06,
"loss": 0.9455,
"step": 376
},
{
"epoch": 0.3516791044776119,
"grad_norm": 1.827371896754954,
"learning_rate": 7.530487563910158e-06,
"loss": 0.9388,
"step": 377
},
{
"epoch": 0.35261194029850745,
"grad_norm": 2.1536349961344867,
"learning_rate": 7.517436800135853e-06,
"loss": 0.9127,
"step": 378
},
{
"epoch": 0.353544776119403,
"grad_norm": 2.1827533787644753,
"learning_rate": 7.504363020518046e-06,
"loss": 0.9547,
"step": 379
},
{
"epoch": 0.35447761194029853,
"grad_norm": 1.9541453185705413,
"learning_rate": 7.491266344584691e-06,
"loss": 0.933,
"step": 380
},
{
"epoch": 0.355410447761194,
"grad_norm": 2.0067062007726117,
"learning_rate": 7.47814689207307e-06,
"loss": 0.8734,
"step": 381
},
{
"epoch": 0.35634328358208955,
"grad_norm": 1.877793451278765,
"learning_rate": 7.465004782928704e-06,
"loss": 0.9441,
"step": 382
},
{
"epoch": 0.3572761194029851,
"grad_norm": 1.755541114266372,
"learning_rate": 7.4518401373042515e-06,
"loss": 1.017,
"step": 383
},
{
"epoch": 0.3582089552238806,
"grad_norm": 1.6400394142968535,
"learning_rate": 7.438653075558413e-06,
"loss": 0.7383,
"step": 384
},
{
"epoch": 0.3591417910447761,
"grad_norm": 1.6382115462830131,
"learning_rate": 7.425443718254835e-06,
"loss": 0.816,
"step": 385
},
{
"epoch": 0.36007462686567165,
"grad_norm": 1.8854717526347602,
"learning_rate": 7.412212186160992e-06,
"loss": 0.8473,
"step": 386
},
{
"epoch": 0.36100746268656714,
"grad_norm": 2.134552360205562,
"learning_rate": 7.398958600247104e-06,
"loss": 0.8186,
"step": 387
},
{
"epoch": 0.3619402985074627,
"grad_norm": 2.1734907749465933,
"learning_rate": 7.3856830816850135e-06,
"loss": 0.8275,
"step": 388
},
{
"epoch": 0.3628731343283582,
"grad_norm": 2.038476866044781,
"learning_rate": 7.372385751847084e-06,
"loss": 0.9774,
"step": 389
},
{
"epoch": 0.36380597014925375,
"grad_norm": 2.0706583942991954,
"learning_rate": 7.359066732305095e-06,
"loss": 0.9216,
"step": 390
},
{
"epoch": 0.36473880597014924,
"grad_norm": 1.598860724693399,
"learning_rate": 7.345726144829117e-06,
"loss": 0.9447,
"step": 391
},
{
"epoch": 0.3656716417910448,
"grad_norm": 1.947712796975702,
"learning_rate": 7.3323641113864175e-06,
"loss": 1.0293,
"step": 392
},
{
"epoch": 0.3666044776119403,
"grad_norm": 2.097366389001302,
"learning_rate": 7.318980754140325e-06,
"loss": 0.9885,
"step": 393
},
{
"epoch": 0.3675373134328358,
"grad_norm": 1.9130039062016222,
"learning_rate": 7.305576195449131e-06,
"loss": 0.9028,
"step": 394
},
{
"epoch": 0.36847014925373134,
"grad_norm": 2.088828050146677,
"learning_rate": 7.292150557864956e-06,
"loss": 0.8572,
"step": 395
},
{
"epoch": 0.3694029850746269,
"grad_norm": 2.6202856117447757,
"learning_rate": 7.2787039641326384e-06,
"loss": 0.9327,
"step": 396
},
{
"epoch": 0.37033582089552236,
"grad_norm": 1.9100495303587661,
"learning_rate": 7.2652365371886116e-06,
"loss": 0.9424,
"step": 397
},
{
"epoch": 0.3712686567164179,
"grad_norm": 2.031809261849809,
"learning_rate": 7.251748400159774e-06,
"loss": 0.9691,
"step": 398
},
{
"epoch": 0.37220149253731344,
"grad_norm": 1.637054065959286,
"learning_rate": 7.238239676362372e-06,
"loss": 0.8291,
"step": 399
},
{
"epoch": 0.373134328358209,
"grad_norm": 1.9391335125756002,
"learning_rate": 7.224710489300866e-06,
"loss": 0.9655,
"step": 400
},
{
"epoch": 0.37406716417910446,
"grad_norm": 2.2429364771281213,
"learning_rate": 7.211160962666799e-06,
"loss": 0.9495,
"step": 401
},
{
"epoch": 0.375,
"grad_norm": 1.7739588268699233,
"learning_rate": 7.197591220337678e-06,
"loss": 0.8931,
"step": 402
},
{
"epoch": 0.37593283582089554,
"grad_norm": 1.7021658836128999,
"learning_rate": 7.1840013863758276e-06,
"loss": 0.8805,
"step": 403
},
{
"epoch": 0.376865671641791,
"grad_norm": 2.110147754671862,
"learning_rate": 7.170391585027263e-06,
"loss": 0.9405,
"step": 404
},
{
"epoch": 0.37779850746268656,
"grad_norm": 2.145511313961903,
"learning_rate": 7.156761940720556e-06,
"loss": 0.9645,
"step": 405
},
{
"epoch": 0.3787313432835821,
"grad_norm": 3.116448577793034,
"learning_rate": 7.1431125780656865e-06,
"loss": 0.9367,
"step": 406
},
{
"epoch": 0.37966417910447764,
"grad_norm": 1.795206170170663,
"learning_rate": 7.129443621852919e-06,
"loss": 0.9071,
"step": 407
},
{
"epoch": 0.3805970149253731,
"grad_norm": 2.207838189209194,
"learning_rate": 7.115755197051645e-06,
"loss": 1.0321,
"step": 408
},
{
"epoch": 0.38152985074626866,
"grad_norm": 2.332739139249594,
"learning_rate": 7.102047428809259e-06,
"loss": 0.9345,
"step": 409
},
{
"epoch": 0.3824626865671642,
"grad_norm": 2.3419234739820776,
"learning_rate": 7.088320442449996e-06,
"loss": 0.8461,
"step": 410
},
{
"epoch": 0.3833955223880597,
"grad_norm": 1.6907774521730226,
"learning_rate": 7.074574363473798e-06,
"loss": 0.8996,
"step": 411
},
{
"epoch": 0.3843283582089552,
"grad_norm": 1.6768703250954178,
"learning_rate": 7.0608093175551615e-06,
"loss": 0.919,
"step": 412
},
{
"epoch": 0.38526119402985076,
"grad_norm": 2.028286986690474,
"learning_rate": 7.047025430541989e-06,
"loss": 0.8693,
"step": 413
},
{
"epoch": 0.38619402985074625,
"grad_norm": 1.7740338735678833,
"learning_rate": 7.033222828454441e-06,
"loss": 0.9147,
"step": 414
},
{
"epoch": 0.3871268656716418,
"grad_norm": 1.901074030626858,
"learning_rate": 7.019401637483778e-06,
"loss": 0.9429,
"step": 415
},
{
"epoch": 0.3880597014925373,
"grad_norm": 2.158647660074332,
"learning_rate": 7.005561983991217e-06,
"loss": 0.9571,
"step": 416
},
{
"epoch": 0.38899253731343286,
"grad_norm": 1.7345485133289813,
"learning_rate": 6.991703994506762e-06,
"loss": 1.0339,
"step": 417
},
{
"epoch": 0.38992537313432835,
"grad_norm": 1.869146874378101,
"learning_rate": 6.9778277957280626e-06,
"loss": 0.8652,
"step": 418
},
{
"epoch": 0.3908582089552239,
"grad_norm": 1.8548963353684862,
"learning_rate": 6.963933514519244e-06,
"loss": 0.8897,
"step": 419
},
{
"epoch": 0.3917910447761194,
"grad_norm": 1.4325589273690118,
"learning_rate": 6.95002127790975e-06,
"loss": 0.8719,
"step": 420
},
{
"epoch": 0.3927238805970149,
"grad_norm": 2.6787920317470193,
"learning_rate": 6.936091213093188e-06,
"loss": 0.973,
"step": 421
},
{
"epoch": 0.39365671641791045,
"grad_norm": 2.03046527097446,
"learning_rate": 6.9221434474261574e-06,
"loss": 0.9624,
"step": 422
},
{
"epoch": 0.394589552238806,
"grad_norm": 2.19017984975356,
"learning_rate": 6.908178108427088e-06,
"loss": 0.934,
"step": 423
},
{
"epoch": 0.39552238805970147,
"grad_norm": 1.918678289470153,
"learning_rate": 6.894195323775078e-06,
"loss": 1.0283,
"step": 424
},
{
"epoch": 0.396455223880597,
"grad_norm": 1.6376477303991932,
"learning_rate": 6.880195221308723e-06,
"loss": 0.9558,
"step": 425
},
{
"epoch": 0.39738805970149255,
"grad_norm": 1.962830865725424,
"learning_rate": 6.8661779290249454e-06,
"loss": 0.8597,
"step": 426
},
{
"epoch": 0.3983208955223881,
"grad_norm": 1.870598796582494,
"learning_rate": 6.852143575077827e-06,
"loss": 0.9093,
"step": 427
},
{
"epoch": 0.39925373134328357,
"grad_norm": 2.166256633653315,
"learning_rate": 6.838092287777444e-06,
"loss": 0.9471,
"step": 428
},
{
"epoch": 0.4001865671641791,
"grad_norm": 1.9841536048083146,
"learning_rate": 6.824024195588678e-06,
"loss": 0.9602,
"step": 429
},
{
"epoch": 0.40111940298507465,
"grad_norm": 1.8945610124449401,
"learning_rate": 6.809939427130053e-06,
"loss": 0.8068,
"step": 430
},
{
"epoch": 0.40205223880597013,
"grad_norm": 1.9990134301621933,
"learning_rate": 6.795838111172561e-06,
"loss": 1.0632,
"step": 431
},
{
"epoch": 0.40298507462686567,
"grad_norm": 2.078783114186299,
"learning_rate": 6.7817203766384765e-06,
"loss": 0.9388,
"step": 432
},
{
"epoch": 0.4039179104477612,
"grad_norm": 1.5922860503007858,
"learning_rate": 6.767586352600183e-06,
"loss": 0.9102,
"step": 433
},
{
"epoch": 0.4048507462686567,
"grad_norm": 1.9536235648016507,
"learning_rate": 6.753436168278991e-06,
"loss": 0.8566,
"step": 434
},
{
"epoch": 0.40578358208955223,
"grad_norm": 1.6762350926648455,
"learning_rate": 6.739269953043959e-06,
"loss": 1.0009,
"step": 435
},
{
"epoch": 0.40671641791044777,
"grad_norm": 1.6516383301978195,
"learning_rate": 6.725087836410711e-06,
"loss": 0.8613,
"step": 436
},
{
"epoch": 0.4076492537313433,
"grad_norm": 1.929547975575555,
"learning_rate": 6.7108899480402476e-06,
"loss": 0.8907,
"step": 437
},
{
"epoch": 0.4085820895522388,
"grad_norm": 1.6194826649406242,
"learning_rate": 6.696676417737764e-06,
"loss": 0.8092,
"step": 438
},
{
"epoch": 0.40951492537313433,
"grad_norm": 1.875563730133676,
"learning_rate": 6.682447375451464e-06,
"loss": 0.878,
"step": 439
},
{
"epoch": 0.41044776119402987,
"grad_norm": 2.00019586112551,
"learning_rate": 6.668202951271371e-06,
"loss": 0.8648,
"step": 440
},
{
"epoch": 0.41138059701492535,
"grad_norm": 1.7527076035517881,
"learning_rate": 6.653943275428135e-06,
"loss": 0.9589,
"step": 441
},
{
"epoch": 0.4123134328358209,
"grad_norm": 1.9819356775801926,
"learning_rate": 6.6396684782918515e-06,
"loss": 0.9469,
"step": 442
},
{
"epoch": 0.41324626865671643,
"grad_norm": 1.9987608985386922,
"learning_rate": 6.625378690370859e-06,
"loss": 1.0344,
"step": 443
},
{
"epoch": 0.4141791044776119,
"grad_norm": 1.7249083150854618,
"learning_rate": 6.61107404231055e-06,
"loss": 0.8517,
"step": 444
},
{
"epoch": 0.41511194029850745,
"grad_norm": 2.1994675670117045,
"learning_rate": 6.596754664892179e-06,
"loss": 0.9265,
"step": 445
},
{
"epoch": 0.416044776119403,
"grad_norm": 1.7618989234329876,
"learning_rate": 6.582420689031665e-06,
"loss": 0.7673,
"step": 446
},
{
"epoch": 0.41697761194029853,
"grad_norm": 1.7214244183240965,
"learning_rate": 6.568072245778394e-06,
"loss": 1.0299,
"step": 447
},
{
"epoch": 0.417910447761194,
"grad_norm": 1.529996651023868,
"learning_rate": 6.553709466314017e-06,
"loss": 0.874,
"step": 448
},
{
"epoch": 0.41884328358208955,
"grad_norm": 1.9043551968707333,
"learning_rate": 6.539332481951262e-06,
"loss": 0.9497,
"step": 449
},
{
"epoch": 0.4197761194029851,
"grad_norm": 1.746984541380808,
"learning_rate": 6.524941424132719e-06,
"loss": 0.84,
"step": 450
},
{
"epoch": 0.4207089552238806,
"grad_norm": 1.8091037295843981,
"learning_rate": 6.510536424429652e-06,
"loss": 0.9888,
"step": 451
},
{
"epoch": 0.4216417910447761,
"grad_norm": 2.051403638209665,
"learning_rate": 6.496117614540786e-06,
"loss": 0.9529,
"step": 452
},
{
"epoch": 0.42257462686567165,
"grad_norm": 1.8969156719791262,
"learning_rate": 6.481685126291106e-06,
"loss": 1.0051,
"step": 453
},
{
"epoch": 0.42350746268656714,
"grad_norm": 1.947141623395184,
"learning_rate": 6.467239091630657e-06,
"loss": 0.8591,
"step": 454
},
{
"epoch": 0.4244402985074627,
"grad_norm": 1.8695643936299324,
"learning_rate": 6.4527796426333265e-06,
"loss": 0.9397,
"step": 455
},
{
"epoch": 0.4253731343283582,
"grad_norm": 1.8070122940649653,
"learning_rate": 6.438306911495649e-06,
"loss": 0.9402,
"step": 456
},
{
"epoch": 0.42630597014925375,
"grad_norm": 2.0476848029098518,
"learning_rate": 6.423821030535586e-06,
"loss": 0.9463,
"step": 457
},
{
"epoch": 0.42723880597014924,
"grad_norm": 2.2095466436651887,
"learning_rate": 6.409322132191331e-06,
"loss": 0.9039,
"step": 458
},
{
"epoch": 0.4281716417910448,
"grad_norm": 1.6595177564958277,
"learning_rate": 6.394810349020083e-06,
"loss": 1.0574,
"step": 459
},
{
"epoch": 0.4291044776119403,
"grad_norm": 1.5855661781953958,
"learning_rate": 6.38028581369684e-06,
"loss": 0.9724,
"step": 460
},
{
"epoch": 0.4300373134328358,
"grad_norm": 1.8107744585533756,
"learning_rate": 6.365748659013195e-06,
"loss": 0.8836,
"step": 461
},
{
"epoch": 0.43097014925373134,
"grad_norm": 2.0865039314010816,
"learning_rate": 6.351199017876106e-06,
"loss": 0.8753,
"step": 462
},
{
"epoch": 0.4319029850746269,
"grad_norm": 2.0288572419261297,
"learning_rate": 6.336637023306698e-06,
"loss": 0.9211,
"step": 463
},
{
"epoch": 0.43283582089552236,
"grad_norm": 1.7840599388258436,
"learning_rate": 6.322062808439029e-06,
"loss": 0.882,
"step": 464
},
{
"epoch": 0.4337686567164179,
"grad_norm": 1.5914382245964365,
"learning_rate": 6.307476506518889e-06,
"loss": 0.8462,
"step": 465
},
{
"epoch": 0.43470149253731344,
"grad_norm": 1.9250852636412363,
"learning_rate": 6.2928782509025695e-06,
"loss": 0.8577,
"step": 466
},
{
"epoch": 0.435634328358209,
"grad_norm": 2.199601564923834,
"learning_rate": 6.278268175055655e-06,
"loss": 0.9589,
"step": 467
},
{
"epoch": 0.43656716417910446,
"grad_norm": 1.9771386419581694,
"learning_rate": 6.2636464125517935e-06,
"loss": 0.8283,
"step": 468
},
{
"epoch": 0.4375,
"grad_norm": 1.6833942327425135,
"learning_rate": 6.2490130970714805e-06,
"loss": 0.8326,
"step": 469
},
{
"epoch": 0.43843283582089554,
"grad_norm": 2.2022249256914637,
"learning_rate": 6.234368362400833e-06,
"loss": 0.9041,
"step": 470
},
{
"epoch": 0.439365671641791,
"grad_norm": 1.6838869253771795,
"learning_rate": 6.2197123424303715e-06,
"loss": 0.8966,
"step": 471
},
{
"epoch": 0.44029850746268656,
"grad_norm": 1.6450845469957913,
"learning_rate": 6.205045171153793e-06,
"loss": 0.8953,
"step": 472
},
{
"epoch": 0.4412313432835821,
"grad_norm": 1.6754452637768948,
"learning_rate": 6.1903669826667445e-06,
"loss": 0.7866,
"step": 473
},
{
"epoch": 0.44216417910447764,
"grad_norm": 1.8277651927125815,
"learning_rate": 6.175677911165599e-06,
"loss": 0.9096,
"step": 474
},
{
"epoch": 0.4430970149253731,
"grad_norm": 1.8127434443359205,
"learning_rate": 6.160978090946231e-06,
"loss": 0.8031,
"step": 475
},
{
"epoch": 0.44402985074626866,
"grad_norm": 1.7524364483607422,
"learning_rate": 6.1462676564027814e-06,
"loss": 0.9784,
"step": 476
},
{
"epoch": 0.4449626865671642,
"grad_norm": 2.4328429693976243,
"learning_rate": 6.131546742026438e-06,
"loss": 0.898,
"step": 477
},
{
"epoch": 0.4458955223880597,
"grad_norm": 2.0086755076593463,
"learning_rate": 6.116815482404194e-06,
"loss": 0.8733,
"step": 478
},
{
"epoch": 0.4468283582089552,
"grad_norm": 2.1290678275748967,
"learning_rate": 6.102074012217635e-06,
"loss": 0.961,
"step": 479
},
{
"epoch": 0.44776119402985076,
"grad_norm": 1.8662798874171085,
"learning_rate": 6.08732246624169e-06,
"loss": 0.7988,
"step": 480
},
{
"epoch": 0.44869402985074625,
"grad_norm": 1.7484457627769991,
"learning_rate": 6.072560979343405e-06,
"loss": 0.9724,
"step": 481
},
{
"epoch": 0.4496268656716418,
"grad_norm": 2.6645334910759413,
"learning_rate": 6.05778968648072e-06,
"loss": 0.9064,
"step": 482
},
{
"epoch": 0.4505597014925373,
"grad_norm": 2.3025731221178845,
"learning_rate": 6.04300872270122e-06,
"loss": 1.0169,
"step": 483
},
{
"epoch": 0.45149253731343286,
"grad_norm": 2.0724485364513034,
"learning_rate": 6.028218223140908e-06,
"loss": 0.9275,
"step": 484
},
{
"epoch": 0.45242537313432835,
"grad_norm": 2.0066349926421654,
"learning_rate": 6.013418323022971e-06,
"loss": 0.922,
"step": 485
},
{
"epoch": 0.4533582089552239,
"grad_norm": 2.3304480567335855,
"learning_rate": 5.998609157656538e-06,
"loss": 0.944,
"step": 486
},
{
"epoch": 0.4542910447761194,
"grad_norm": 2.0750642576704044,
"learning_rate": 5.98379086243545e-06,
"loss": 0.9545,
"step": 487
},
{
"epoch": 0.4552238805970149,
"grad_norm": 2.0016185879287693,
"learning_rate": 5.968963572837015e-06,
"loss": 0.9974,
"step": 488
},
{
"epoch": 0.45615671641791045,
"grad_norm": 2.037344252419743,
"learning_rate": 5.954127424420774e-06,
"loss": 0.8487,
"step": 489
},
{
"epoch": 0.457089552238806,
"grad_norm": 1.9100538020516062,
"learning_rate": 5.9392825528272575e-06,
"loss": 0.8407,
"step": 490
},
{
"epoch": 0.45802238805970147,
"grad_norm": 2.4683934454665266,
"learning_rate": 5.924429093776757e-06,
"loss": 1.0248,
"step": 491
},
{
"epoch": 0.458955223880597,
"grad_norm": 1.8517438035145544,
"learning_rate": 5.909567183068065e-06,
"loss": 0.8783,
"step": 492
},
{
"epoch": 0.45988805970149255,
"grad_norm": 1.62584427498455,
"learning_rate": 5.89469695657725e-06,
"loss": 0.8687,
"step": 493
},
{
"epoch": 0.4608208955223881,
"grad_norm": 1.7425326435090562,
"learning_rate": 5.879818550256405e-06,
"loss": 0.8203,
"step": 494
},
{
"epoch": 0.46175373134328357,
"grad_norm": 1.6533647771980664,
"learning_rate": 5.864932100132411e-06,
"loss": 0.8041,
"step": 495
},
{
"epoch": 0.4626865671641791,
"grad_norm": 2.234820603917402,
"learning_rate": 5.850037742305687e-06,
"loss": 0.9257,
"step": 496
},
{
"epoch": 0.46361940298507465,
"grad_norm": 2.608509246144018,
"learning_rate": 5.835135612948951e-06,
"loss": 0.9392,
"step": 497
},
{
"epoch": 0.46455223880597013,
"grad_norm": 3.3076446105800485,
"learning_rate": 5.820225848305971e-06,
"loss": 1.0212,
"step": 498
},
{
"epoch": 0.46548507462686567,
"grad_norm": 1.7124578407016553,
"learning_rate": 5.805308584690322e-06,
"loss": 0.9321,
"step": 499
},
{
"epoch": 0.4664179104477612,
"grad_norm": 1.706562510198135,
"learning_rate": 5.790383958484138e-06,
"loss": 0.8978,
"step": 500
},
{
"epoch": 0.4673507462686567,
"grad_norm": 1.6510791484595355,
"learning_rate": 5.775452106136868e-06,
"loss": 0.9983,
"step": 501
},
{
"epoch": 0.46828358208955223,
"grad_norm": 2.090478419375231,
"learning_rate": 5.760513164164025e-06,
"loss": 1.0134,
"step": 502
},
{
"epoch": 0.46921641791044777,
"grad_norm": 2.5040099516652385,
"learning_rate": 5.745567269145939e-06,
"loss": 0.8524,
"step": 503
},
{
"epoch": 0.4701492537313433,
"grad_norm": 2.086614459436081,
"learning_rate": 5.7306145577265095e-06,
"loss": 1.0267,
"step": 504
},
{
"epoch": 0.4710820895522388,
"grad_norm": 1.84128903342119,
"learning_rate": 5.7156551666119554e-06,
"loss": 0.9144,
"step": 505
},
{
"epoch": 0.47201492537313433,
"grad_norm": 1.727295453523586,
"learning_rate": 5.700689232569565e-06,
"loss": 0.9162,
"step": 506
},
{
"epoch": 0.47294776119402987,
"grad_norm": 1.8526717448902457,
"learning_rate": 5.685716892426445e-06,
"loss": 0.9511,
"step": 507
},
{
"epoch": 0.47388059701492535,
"grad_norm": 1.944797120944845,
"learning_rate": 5.670738283068273e-06,
"loss": 0.8841,
"step": 508
},
{
"epoch": 0.4748134328358209,
"grad_norm": 1.6851933993114967,
"learning_rate": 5.655753541438041e-06,
"loss": 0.854,
"step": 509
},
{
"epoch": 0.47574626865671643,
"grad_norm": 1.7907677673401778,
"learning_rate": 5.640762804534806e-06,
"loss": 0.888,
"step": 510
},
{
"epoch": 0.4766791044776119,
"grad_norm": 2.030804382568102,
"learning_rate": 5.625766209412436e-06,
"loss": 0.8986,
"step": 511
},
{
"epoch": 0.47761194029850745,
"grad_norm": 2.529821890626977,
"learning_rate": 5.610763893178361e-06,
"loss": 0.9144,
"step": 512
},
{
"epoch": 0.478544776119403,
"grad_norm": 1.8699455112109173,
"learning_rate": 5.595755992992316e-06,
"loss": 0.8751,
"step": 513
},
{
"epoch": 0.47947761194029853,
"grad_norm": 1.9628191367766852,
"learning_rate": 5.580742646065085e-06,
"loss": 1.0019,
"step": 514
},
{
"epoch": 0.480410447761194,
"grad_norm": 1.634184474501889,
"learning_rate": 5.565723989657253e-06,
"loss": 0.86,
"step": 515
},
{
"epoch": 0.48134328358208955,
"grad_norm": 1.975013170460317,
"learning_rate": 5.550700161077946e-06,
"loss": 0.9182,
"step": 516
},
{
"epoch": 0.4822761194029851,
"grad_norm": 2.490918313313446,
"learning_rate": 5.535671297683574e-06,
"loss": 0.9191,
"step": 517
},
{
"epoch": 0.4832089552238806,
"grad_norm": 2.0588421516614415,
"learning_rate": 5.520637536876582e-06,
"loss": 0.9655,
"step": 518
},
{
"epoch": 0.4841417910447761,
"grad_norm": 1.9623593516769542,
"learning_rate": 5.505599016104187e-06,
"loss": 0.8234,
"step": 519
},
{
"epoch": 0.48507462686567165,
"grad_norm": 1.521511329909949,
"learning_rate": 5.490555872857128e-06,
"loss": 0.9535,
"step": 520
},
{
"epoch": 0.48600746268656714,
"grad_norm": 1.7528138342813324,
"learning_rate": 5.475508244668402e-06,
"loss": 0.882,
"step": 521
},
{
"epoch": 0.4869402985074627,
"grad_norm": 1.5878229265875845,
"learning_rate": 5.4604562691120135e-06,
"loss": 0.816,
"step": 522
},
{
"epoch": 0.4878731343283582,
"grad_norm": 1.9107605806765566,
"learning_rate": 5.445400083801705e-06,
"loss": 0.8465,
"step": 523
},
{
"epoch": 0.48880597014925375,
"grad_norm": 1.9978785486674788,
"learning_rate": 5.430339826389719e-06,
"loss": 0.8569,
"step": 524
},
{
"epoch": 0.48973880597014924,
"grad_norm": 2.450877546959106,
"learning_rate": 5.415275634565518e-06,
"loss": 0.9272,
"step": 525
},
{
"epoch": 0.4906716417910448,
"grad_norm": 1.4965007733169913,
"learning_rate": 5.400207646054536e-06,
"loss": 0.8477,
"step": 526
},
{
"epoch": 0.4916044776119403,
"grad_norm": 1.9872847925668968,
"learning_rate": 5.385135998616925e-06,
"loss": 0.9806,
"step": 527
},
{
"epoch": 0.4925373134328358,
"grad_norm": 1.7164040867277939,
"learning_rate": 5.370060830046282e-06,
"loss": 0.8864,
"step": 528
},
{
"epoch": 0.49347014925373134,
"grad_norm": 1.7892820474543287,
"learning_rate": 5.354982278168399e-06,
"loss": 0.9467,
"step": 529
},
{
"epoch": 0.4944029850746269,
"grad_norm": 1.9702553885624257,
"learning_rate": 5.339900480840002e-06,
"loss": 0.8643,
"step": 530
},
{
"epoch": 0.49533582089552236,
"grad_norm": 1.5725480446784155,
"learning_rate": 5.324815575947485e-06,
"loss": 0.8288,
"step": 531
},
{
"epoch": 0.4962686567164179,
"grad_norm": 1.9514380999571526,
"learning_rate": 5.309727701405657e-06,
"loss": 0.7732,
"step": 532
},
{
"epoch": 0.49720149253731344,
"grad_norm": 2.2694608310303654,
"learning_rate": 5.294636995156472e-06,
"loss": 0.9683,
"step": 533
},
{
"epoch": 0.498134328358209,
"grad_norm": 1.985052396559636,
"learning_rate": 5.2795435951677785e-06,
"loss": 0.8325,
"step": 534
},
{
"epoch": 0.49906716417910446,
"grad_norm": 2.727373176363925,
"learning_rate": 5.26444763943205e-06,
"loss": 1.0093,
"step": 535
},
{
"epoch": 0.5,
"grad_norm": 1.7026723585379113,
"learning_rate": 5.2493492659651265e-06,
"loss": 0.8063,
"step": 536
},
{
"epoch": 0.5009328358208955,
"grad_norm": 1.750858377275619,
"learning_rate": 5.234248612804952e-06,
"loss": 0.9045,
"step": 537
},
{
"epoch": 0.5018656716417911,
"grad_norm": 1.791800803946349,
"learning_rate": 5.219145818010312e-06,
"loss": 0.9286,
"step": 538
},
{
"epoch": 0.5027985074626866,
"grad_norm": 2.003785764098383,
"learning_rate": 5.204041019659577e-06,
"loss": 0.9712,
"step": 539
},
{
"epoch": 0.503731343283582,
"grad_norm": 1.7383132635839407,
"learning_rate": 5.1889343558494264e-06,
"loss": 0.8558,
"step": 540
},
{
"epoch": 0.5046641791044776,
"grad_norm": 2.0718901945578776,
"learning_rate": 5.173825964693603e-06,
"loss": 0.7397,
"step": 541
},
{
"epoch": 0.5055970149253731,
"grad_norm": 2.067452750056043,
"learning_rate": 5.158715984321636e-06,
"loss": 0.9138,
"step": 542
},
{
"epoch": 0.5065298507462687,
"grad_norm": 1.9055606022139955,
"learning_rate": 5.14360455287759e-06,
"loss": 0.8938,
"step": 543
},
{
"epoch": 0.5074626865671642,
"grad_norm": 1.8047059130283511,
"learning_rate": 5.128491808518791e-06,
"loss": 0.8714,
"step": 544
},
{
"epoch": 0.5083955223880597,
"grad_norm": 1.7159779139487976,
"learning_rate": 5.113377889414566e-06,
"loss": 0.8728,
"step": 545
},
{
"epoch": 0.5093283582089553,
"grad_norm": 1.660721220265496,
"learning_rate": 5.098262933744994e-06,
"loss": 0.799,
"step": 546
},
{
"epoch": 0.5102611940298507,
"grad_norm": 1.8430669159038027,
"learning_rate": 5.083147079699619e-06,
"loss": 0.935,
"step": 547
},
{
"epoch": 0.5111940298507462,
"grad_norm": 1.6196526751363667,
"learning_rate": 5.068030465476204e-06,
"loss": 0.8224,
"step": 548
},
{
"epoch": 0.5121268656716418,
"grad_norm": 2.110923069753341,
"learning_rate": 5.052913229279459e-06,
"loss": 0.8359,
"step": 549
},
{
"epoch": 0.5130597014925373,
"grad_norm": 2.520721003177226,
"learning_rate": 5.037795509319784e-06,
"loss": 0.9608,
"step": 550
},
{
"epoch": 0.5139925373134329,
"grad_norm": 1.8235087233455562,
"learning_rate": 5.0226774438119995e-06,
"loss": 0.8895,
"step": 551
},
{
"epoch": 0.5149253731343284,
"grad_norm": 2.1491344144113786,
"learning_rate": 5.0075591709740845e-06,
"loss": 0.8201,
"step": 552
},
{
"epoch": 0.5158582089552238,
"grad_norm": 1.7119474042771239,
"learning_rate": 4.9924408290259154e-06,
"loss": 0.8886,
"step": 553
},
{
"epoch": 0.5167910447761194,
"grad_norm": 2.2271591254890297,
"learning_rate": 4.977322556188001e-06,
"loss": 0.9185,
"step": 554
},
{
"epoch": 0.5177238805970149,
"grad_norm": 1.7581937074489065,
"learning_rate": 4.962204490680216e-06,
"loss": 0.8442,
"step": 555
},
{
"epoch": 0.5186567164179104,
"grad_norm": 1.866949874199332,
"learning_rate": 4.947086770720542e-06,
"loss": 0.9444,
"step": 556
},
{
"epoch": 0.519589552238806,
"grad_norm": 1.6783538225763326,
"learning_rate": 4.931969534523797e-06,
"loss": 0.869,
"step": 557
},
{
"epoch": 0.5205223880597015,
"grad_norm": 1.7813678602051846,
"learning_rate": 4.916852920300381e-06,
"loss": 0.9838,
"step": 558
},
{
"epoch": 0.5214552238805971,
"grad_norm": 2.8002408761767246,
"learning_rate": 4.9017370662550065e-06,
"loss": 0.9121,
"step": 559
},
{
"epoch": 0.5223880597014925,
"grad_norm": 1.9805275928128727,
"learning_rate": 4.8866221105854335e-06,
"loss": 0.9314,
"step": 560
},
{
"epoch": 0.523320895522388,
"grad_norm": 1.5865517436736651,
"learning_rate": 4.871508191481211e-06,
"loss": 0.9136,
"step": 561
},
{
"epoch": 0.5242537313432836,
"grad_norm": 2.0523766290038608,
"learning_rate": 4.856395447122411e-06,
"loss": 0.8393,
"step": 562
},
{
"epoch": 0.5251865671641791,
"grad_norm": 1.7594125814878478,
"learning_rate": 4.841284015678366e-06,
"loss": 0.8743,
"step": 563
},
{
"epoch": 0.5261194029850746,
"grad_norm": 2.1083143750056434,
"learning_rate": 4.8261740353063985e-06,
"loss": 0.8261,
"step": 564
},
{
"epoch": 0.5270522388059702,
"grad_norm": 1.9843753067575551,
"learning_rate": 4.811065644150575e-06,
"loss": 0.8294,
"step": 565
},
{
"epoch": 0.5279850746268657,
"grad_norm": 1.7437925408023613,
"learning_rate": 4.795958980340425e-06,
"loss": 0.8749,
"step": 566
},
{
"epoch": 0.5289179104477612,
"grad_norm": 1.8166643625951588,
"learning_rate": 4.7808541819896885e-06,
"loss": 0.8535,
"step": 567
},
{
"epoch": 0.5298507462686567,
"grad_norm": 1.9221152453064072,
"learning_rate": 4.76575138719505e-06,
"loss": 1.0323,
"step": 568
},
{
"epoch": 0.5307835820895522,
"grad_norm": 1.8983867340838674,
"learning_rate": 4.750650734034875e-06,
"loss": 0.8647,
"step": 569
},
{
"epoch": 0.5317164179104478,
"grad_norm": 1.9109383994089197,
"learning_rate": 4.735552360567952e-06,
"loss": 0.8301,
"step": 570
},
{
"epoch": 0.5326492537313433,
"grad_norm": 1.986965496840367,
"learning_rate": 4.720456404832222e-06,
"loss": 0.8854,
"step": 571
},
{
"epoch": 0.5335820895522388,
"grad_norm": 1.989224053860767,
"learning_rate": 4.7053630048435295e-06,
"loss": 0.9118,
"step": 572
},
{
"epoch": 0.5345149253731343,
"grad_norm": 1.8885324464279352,
"learning_rate": 4.690272298594344e-06,
"loss": 0.9404,
"step": 573
},
{
"epoch": 0.5354477611940298,
"grad_norm": 1.6606744504568167,
"learning_rate": 4.675184424052516e-06,
"loss": 0.8939,
"step": 574
},
{
"epoch": 0.5363805970149254,
"grad_norm": 1.7212809784010423,
"learning_rate": 4.660099519159999e-06,
"loss": 0.769,
"step": 575
},
{
"epoch": 0.5373134328358209,
"grad_norm": 1.9180279899058936,
"learning_rate": 4.645017721831602e-06,
"loss": 0.8694,
"step": 576
},
{
"epoch": 0.5382462686567164,
"grad_norm": 1.7598266570482806,
"learning_rate": 4.629939169953721e-06,
"loss": 0.9149,
"step": 577
},
{
"epoch": 0.539179104477612,
"grad_norm": 1.9100499143852558,
"learning_rate": 4.614864001383077e-06,
"loss": 0.8531,
"step": 578
},
{
"epoch": 0.5401119402985075,
"grad_norm": 2.140002365235594,
"learning_rate": 4.599792353945467e-06,
"loss": 0.9375,
"step": 579
},
{
"epoch": 0.5410447761194029,
"grad_norm": 1.6600056631310014,
"learning_rate": 4.584724365434485e-06,
"loss": 0.9454,
"step": 580
},
{
"epoch": 0.5419776119402985,
"grad_norm": 1.860009284660804,
"learning_rate": 4.569660173610284e-06,
"loss": 0.9537,
"step": 581
},
{
"epoch": 0.542910447761194,
"grad_norm": 1.758252494437625,
"learning_rate": 4.554599916198296e-06,
"loss": 0.9364,
"step": 582
},
{
"epoch": 0.5438432835820896,
"grad_norm": 1.8993222008926254,
"learning_rate": 4.53954373088799e-06,
"loss": 0.9033,
"step": 583
},
{
"epoch": 0.5447761194029851,
"grad_norm": 1.8415365373484056,
"learning_rate": 4.524491755331598e-06,
"loss": 0.9143,
"step": 584
},
{
"epoch": 0.5457089552238806,
"grad_norm": 2.179281682198713,
"learning_rate": 4.509444127142872e-06,
"loss": 0.9253,
"step": 585
},
{
"epoch": 0.5466417910447762,
"grad_norm": 1.7842191263951328,
"learning_rate": 4.494400983895813e-06,
"loss": 0.951,
"step": 586
},
{
"epoch": 0.5475746268656716,
"grad_norm": 1.6569592925105656,
"learning_rate": 4.47936246312342e-06,
"loss": 0.8698,
"step": 587
},
{
"epoch": 0.5485074626865671,
"grad_norm": 2.134800358190759,
"learning_rate": 4.464328702316427e-06,
"loss": 0.8816,
"step": 588
},
{
"epoch": 0.5494402985074627,
"grad_norm": 2.0678977367273768,
"learning_rate": 4.449299838922055e-06,
"loss": 0.9757,
"step": 589
},
{
"epoch": 0.5503731343283582,
"grad_norm": 2.1042271025651136,
"learning_rate": 4.434276010342747e-06,
"loss": 0.8478,
"step": 590
},
{
"epoch": 0.5513059701492538,
"grad_norm": 1.8471715160967321,
"learning_rate": 4.419257353934915e-06,
"loss": 1.0154,
"step": 591
},
{
"epoch": 0.5522388059701493,
"grad_norm": 1.9110264372269008,
"learning_rate": 4.4042440070076845e-06,
"loss": 0.9741,
"step": 592
},
{
"epoch": 0.5531716417910447,
"grad_norm": 1.7710140302314843,
"learning_rate": 4.3892361068216395e-06,
"loss": 0.98,
"step": 593
},
{
"epoch": 0.5541044776119403,
"grad_norm": 2.1598975465316697,
"learning_rate": 4.374233790587565e-06,
"loss": 0.9513,
"step": 594
},
{
"epoch": 0.5550373134328358,
"grad_norm": 2.1161425003393295,
"learning_rate": 4.359237195465196e-06,
"loss": 0.9691,
"step": 595
},
{
"epoch": 0.5559701492537313,
"grad_norm": 1.6123122364284703,
"learning_rate": 4.34424645856196e-06,
"loss": 0.7345,
"step": 596
},
{
"epoch": 0.5569029850746269,
"grad_norm": 1.6823152785721938,
"learning_rate": 4.329261716931728e-06,
"loss": 0.9259,
"step": 597
},
{
"epoch": 0.5578358208955224,
"grad_norm": 1.9787408040491625,
"learning_rate": 4.314283107573556e-06,
"loss": 1.0071,
"step": 598
},
{
"epoch": 0.558768656716418,
"grad_norm": 1.9732873087126463,
"learning_rate": 4.299310767430436e-06,
"loss": 0.8261,
"step": 599
},
{
"epoch": 0.5597014925373134,
"grad_norm": 1.7365689803587974,
"learning_rate": 4.284344833388046e-06,
"loss": 0.8984,
"step": 600
},
{
"epoch": 0.5606343283582089,
"grad_norm": 1.9781114524095245,
"learning_rate": 4.269385442273492e-06,
"loss": 0.9354,
"step": 601
},
{
"epoch": 0.5615671641791045,
"grad_norm": 2.0353017348197966,
"learning_rate": 4.254432730854063e-06,
"loss": 1.0476,
"step": 602
},
{
"epoch": 0.5625,
"grad_norm": 1.949312569327919,
"learning_rate": 4.239486835835978e-06,
"loss": 0.8222,
"step": 603
},
{
"epoch": 0.5634328358208955,
"grad_norm": 1.70023789657344,
"learning_rate": 4.224547893863133e-06,
"loss": 0.8493,
"step": 604
},
{
"epoch": 0.5643656716417911,
"grad_norm": 2.509857860628613,
"learning_rate": 4.209616041515864e-06,
"loss": 0.8796,
"step": 605
},
{
"epoch": 0.5652985074626866,
"grad_norm": 2.0749765119260632,
"learning_rate": 4.194691415309679e-06,
"loss": 0.8386,
"step": 606
},
{
"epoch": 0.566231343283582,
"grad_norm": 2.2947419302081435,
"learning_rate": 4.179774151694031e-06,
"loss": 0.9921,
"step": 607
},
{
"epoch": 0.5671641791044776,
"grad_norm": 2.118776309950458,
"learning_rate": 4.16486438705105e-06,
"loss": 0.954,
"step": 608
},
{
"epoch": 0.5680970149253731,
"grad_norm": 1.8396643565667856,
"learning_rate": 4.149962257694315e-06,
"loss": 1.0045,
"step": 609
},
{
"epoch": 0.5690298507462687,
"grad_norm": 1.9743349682846876,
"learning_rate": 4.135067899867592e-06,
"loss": 0.8452,
"step": 610
},
{
"epoch": 0.5699626865671642,
"grad_norm": 1.9500134401495213,
"learning_rate": 4.120181449743597e-06,
"loss": 0.9711,
"step": 611
},
{
"epoch": 0.5708955223880597,
"grad_norm": 1.882253020134391,
"learning_rate": 4.105303043422753e-06,
"loss": 0.931,
"step": 612
},
{
"epoch": 0.5718283582089553,
"grad_norm": 1.94819411872849,
"learning_rate": 4.090432816931937e-06,
"loss": 0.9984,
"step": 613
},
{
"epoch": 0.5727611940298507,
"grad_norm": 1.734833979859039,
"learning_rate": 4.075570906223245e-06,
"loss": 0.9455,
"step": 614
},
{
"epoch": 0.5736940298507462,
"grad_norm": 1.7826473345948533,
"learning_rate": 4.060717447172743e-06,
"loss": 0.8289,
"step": 615
},
{
"epoch": 0.5746268656716418,
"grad_norm": 1.687816762746417,
"learning_rate": 4.045872575579227e-06,
"loss": 0.7877,
"step": 616
},
{
"epoch": 0.5755597014925373,
"grad_norm": 1.8263410352435114,
"learning_rate": 4.031036427162986e-06,
"loss": 0.8942,
"step": 617
},
{
"epoch": 0.5764925373134329,
"grad_norm": 1.6511118051483709,
"learning_rate": 4.016209137564549e-06,
"loss": 0.921,
"step": 618
},
{
"epoch": 0.5774253731343284,
"grad_norm": 1.692660185784582,
"learning_rate": 4.0013908423434625e-06,
"loss": 0.921,
"step": 619
},
{
"epoch": 0.5783582089552238,
"grad_norm": 1.8491031015078039,
"learning_rate": 3.986581676977031e-06,
"loss": 0.8987,
"step": 620
},
{
"epoch": 0.5792910447761194,
"grad_norm": 1.6933128812399956,
"learning_rate": 3.971781776859093e-06,
"loss": 0.9109,
"step": 621
},
{
"epoch": 0.5802238805970149,
"grad_norm": 1.9111461743617122,
"learning_rate": 3.9569912772987815e-06,
"loss": 0.8082,
"step": 622
},
{
"epoch": 0.5811567164179104,
"grad_norm": 2.196290943574567,
"learning_rate": 3.942210313519281e-06,
"loss": 0.9847,
"step": 623
},
{
"epoch": 0.582089552238806,
"grad_norm": 1.8563549161648525,
"learning_rate": 3.927439020656596e-06,
"loss": 0.9361,
"step": 624
},
{
"epoch": 0.5830223880597015,
"grad_norm": 1.7524961576981606,
"learning_rate": 3.9126775337583126e-06,
"loss": 0.9692,
"step": 625
},
{
"epoch": 0.5839552238805971,
"grad_norm": 1.7885927847420795,
"learning_rate": 3.897925987782367e-06,
"loss": 0.9647,
"step": 626
},
{
"epoch": 0.5848880597014925,
"grad_norm": 1.8858514349433209,
"learning_rate": 3.883184517595807e-06,
"loss": 1.0914,
"step": 627
},
{
"epoch": 0.585820895522388,
"grad_norm": 2.106670018742413,
"learning_rate": 3.868453257973565e-06,
"loss": 0.8689,
"step": 628
},
{
"epoch": 0.5867537313432836,
"grad_norm": 1.8329999058514763,
"learning_rate": 3.85373234359722e-06,
"loss": 0.8808,
"step": 629
},
{
"epoch": 0.5876865671641791,
"grad_norm": 2.2313483321243495,
"learning_rate": 3.83902190905377e-06,
"loss": 0.9455,
"step": 630
},
{
"epoch": 0.5886194029850746,
"grad_norm": 1.9001706240267824,
"learning_rate": 3.824322088834402e-06,
"loss": 0.917,
"step": 631
},
{
"epoch": 0.5895522388059702,
"grad_norm": 1.990755570644533,
"learning_rate": 3.8096330173332568e-06,
"loss": 0.9183,
"step": 632
},
{
"epoch": 0.5904850746268657,
"grad_norm": 2.0492424458349987,
"learning_rate": 3.7949548288462085e-06,
"loss": 1.0354,
"step": 633
},
{
"epoch": 0.5914179104477612,
"grad_norm": 1.7144413032639272,
"learning_rate": 3.780287657569629e-06,
"loss": 0.8389,
"step": 634
},
{
"epoch": 0.5923507462686567,
"grad_norm": 1.8158938328142435,
"learning_rate": 3.7656316375991685e-06,
"loss": 0.8235,
"step": 635
},
{
"epoch": 0.5932835820895522,
"grad_norm": 2.773221116138112,
"learning_rate": 3.7509869029285216e-06,
"loss": 0.9057,
"step": 636
},
{
"epoch": 0.5942164179104478,
"grad_norm": 1.997002414798992,
"learning_rate": 3.7363535874482073e-06,
"loss": 0.9415,
"step": 637
},
{
"epoch": 0.5951492537313433,
"grad_norm": 1.650133033698427,
"learning_rate": 3.7217318249443457e-06,
"loss": 0.9279,
"step": 638
},
{
"epoch": 0.5960820895522388,
"grad_norm": 2.224138336233344,
"learning_rate": 3.707121749097431e-06,
"loss": 0.9785,
"step": 639
},
{
"epoch": 0.5970149253731343,
"grad_norm": 1.9910431855006252,
"learning_rate": 3.6925234934811132e-06,
"loss": 0.898,
"step": 640
},
{
"epoch": 0.5979477611940298,
"grad_norm": 1.9984941352254728,
"learning_rate": 3.6779371915609724e-06,
"loss": 0.9936,
"step": 641
},
{
"epoch": 0.5988805970149254,
"grad_norm": 1.7970006218494434,
"learning_rate": 3.6633629766933044e-06,
"loss": 0.909,
"step": 642
},
{
"epoch": 0.5998134328358209,
"grad_norm": 2.2684804618941787,
"learning_rate": 3.6488009821238956e-06,
"loss": 0.8899,
"step": 643
},
{
"epoch": 0.6007462686567164,
"grad_norm": 2.4892389354691553,
"learning_rate": 3.6342513409868075e-06,
"loss": 0.9505,
"step": 644
},
{
"epoch": 0.601679104477612,
"grad_norm": 2.3507788988464546,
"learning_rate": 3.6197141863031622e-06,
"loss": 0.9347,
"step": 645
},
{
"epoch": 0.6026119402985075,
"grad_norm": 2.200207420306531,
"learning_rate": 3.60518965097992e-06,
"loss": 0.9796,
"step": 646
},
{
"epoch": 0.6035447761194029,
"grad_norm": 1.8300040302653535,
"learning_rate": 3.590677867808669e-06,
"loss": 0.9412,
"step": 647
},
{
"epoch": 0.6044776119402985,
"grad_norm": 2.0896943094080154,
"learning_rate": 3.5761789694644138e-06,
"loss": 0.839,
"step": 648
},
{
"epoch": 0.605410447761194,
"grad_norm": 2.0012576222410394,
"learning_rate": 3.5616930885043523e-06,
"loss": 0.9367,
"step": 649
},
{
"epoch": 0.6063432835820896,
"grad_norm": 2.0135621275696454,
"learning_rate": 3.547220357366674e-06,
"loss": 0.9503,
"step": 650
},
{
"epoch": 0.6072761194029851,
"grad_norm": 1.8913561413759046,
"learning_rate": 3.5327609083693436e-06,
"loss": 0.8302,
"step": 651
},
{
"epoch": 0.6082089552238806,
"grad_norm": 2.078704359837769,
"learning_rate": 3.518314873708894e-06,
"loss": 1.0425,
"step": 652
},
{
"epoch": 0.6091417910447762,
"grad_norm": 2.1879480781342315,
"learning_rate": 3.503882385459215e-06,
"loss": 0.9555,
"step": 653
},
{
"epoch": 0.6100746268656716,
"grad_norm": 1.719682709307889,
"learning_rate": 3.4894635755703487e-06,
"loss": 0.817,
"step": 654
},
{
"epoch": 0.6110074626865671,
"grad_norm": 1.7000348890091614,
"learning_rate": 3.4750585758672827e-06,
"loss": 0.9009,
"step": 655
},
{
"epoch": 0.6119402985074627,
"grad_norm": 1.831130134138469,
"learning_rate": 3.46066751804874e-06,
"loss": 0.7928,
"step": 656
},
{
"epoch": 0.6128731343283582,
"grad_norm": 2.5624007184614475,
"learning_rate": 3.4462905336859842e-06,
"loss": 0.89,
"step": 657
},
{
"epoch": 0.6138059701492538,
"grad_norm": 2.242301460753576,
"learning_rate": 3.4319277542216078e-06,
"loss": 0.9334,
"step": 658
},
{
"epoch": 0.6147388059701493,
"grad_norm": 2.2386890730209488,
"learning_rate": 3.417579310968336e-06,
"loss": 0.8543,
"step": 659
},
{
"epoch": 0.6156716417910447,
"grad_norm": 2.1309891941524643,
"learning_rate": 3.403245335107822e-06,
"loss": 0.9932,
"step": 660
},
{
"epoch": 0.6166044776119403,
"grad_norm": 1.4937457768694915,
"learning_rate": 3.3889259576894517e-06,
"loss": 0.733,
"step": 661
},
{
"epoch": 0.6175373134328358,
"grad_norm": 2.1708679167599536,
"learning_rate": 3.3746213096291434e-06,
"loss": 0.8271,
"step": 662
},
{
"epoch": 0.6184701492537313,
"grad_norm": 1.7806973110074789,
"learning_rate": 3.3603315217081493e-06,
"loss": 0.9178,
"step": 663
},
{
"epoch": 0.6194029850746269,
"grad_norm": 2.3184863168303607,
"learning_rate": 3.3460567245718656e-06,
"loss": 0.8413,
"step": 664
},
{
"epoch": 0.6203358208955224,
"grad_norm": 2.044569134657629,
"learning_rate": 3.3317970487286303e-06,
"loss": 0.8839,
"step": 665
},
{
"epoch": 0.621268656716418,
"grad_norm": 2.0481419569683412,
"learning_rate": 3.3175526245485367e-06,
"loss": 0.8879,
"step": 666
},
{
"epoch": 0.6222014925373134,
"grad_norm": 1.8235909602014189,
"learning_rate": 3.3033235822622367e-06,
"loss": 0.9026,
"step": 667
},
{
"epoch": 0.6231343283582089,
"grad_norm": 1.9348802394539901,
"learning_rate": 3.289110051959754e-06,
"loss": 0.9344,
"step": 668
},
{
"epoch": 0.6240671641791045,
"grad_norm": 2.037046038835384,
"learning_rate": 3.274912163589291e-06,
"loss": 0.8829,
"step": 669
},
{
"epoch": 0.625,
"grad_norm": 2.0382109243893973,
"learning_rate": 3.260730046956042e-06,
"loss": 0.9389,
"step": 670
},
{
"epoch": 0.6259328358208955,
"grad_norm": 1.9183269539571592,
"learning_rate": 3.2465638317210124e-06,
"loss": 0.819,
"step": 671
},
{
"epoch": 0.6268656716417911,
"grad_norm": 2.02595437823034,
"learning_rate": 3.23241364739982e-06,
"loss": 0.8691,
"step": 672
},
{
"epoch": 0.6277985074626866,
"grad_norm": 2.064970384460898,
"learning_rate": 3.2182796233615264e-06,
"loss": 0.8363,
"step": 673
},
{
"epoch": 0.628731343283582,
"grad_norm": 1.9144993307498372,
"learning_rate": 3.2041618888274406e-06,
"loss": 0.9091,
"step": 674
},
{
"epoch": 0.6296641791044776,
"grad_norm": 2.307325028011396,
"learning_rate": 3.1900605728699484e-06,
"loss": 0.9417,
"step": 675
},
{
"epoch": 0.6305970149253731,
"grad_norm": 2.064546015202059,
"learning_rate": 3.175975804411325e-06,
"loss": 0.8797,
"step": 676
},
{
"epoch": 0.6315298507462687,
"grad_norm": 1.7772998710245742,
"learning_rate": 3.1619077122225585e-06,
"loss": 0.907,
"step": 677
},
{
"epoch": 0.6324626865671642,
"grad_norm": 1.465528428312256,
"learning_rate": 3.1478564249221744e-06,
"loss": 0.8058,
"step": 678
},
{
"epoch": 0.6333955223880597,
"grad_norm": 1.6922668947538908,
"learning_rate": 3.1338220709750567e-06,
"loss": 0.8482,
"step": 679
},
{
"epoch": 0.6343283582089553,
"grad_norm": 2.127665551858506,
"learning_rate": 3.119804778691278e-06,
"loss": 0.9115,
"step": 680
},
{
"epoch": 0.6352611940298507,
"grad_norm": 2.0133025467166537,
"learning_rate": 3.105804676224923e-06,
"loss": 1.029,
"step": 681
},
{
"epoch": 0.6361940298507462,
"grad_norm": 1.8967378743972476,
"learning_rate": 3.091821891572912e-06,
"loss": 0.7481,
"step": 682
},
{
"epoch": 0.6371268656716418,
"grad_norm": 1.6475953286654188,
"learning_rate": 3.0778565525738434e-06,
"loss": 0.8159,
"step": 683
},
{
"epoch": 0.6380597014925373,
"grad_norm": 1.548248188771747,
"learning_rate": 3.0639087869068118e-06,
"loss": 0.882,
"step": 684
},
{
"epoch": 0.6389925373134329,
"grad_norm": 2.0656940896231433,
"learning_rate": 3.0499787220902506e-06,
"loss": 0.8187,
"step": 685
},
{
"epoch": 0.6399253731343284,
"grad_norm": 1.9933996512476804,
"learning_rate": 3.036066485480758e-06,
"loss": 0.9361,
"step": 686
},
{
"epoch": 0.6408582089552238,
"grad_norm": 1.8495442843536554,
"learning_rate": 3.0221722042719383e-06,
"loss": 0.8517,
"step": 687
},
{
"epoch": 0.6417910447761194,
"grad_norm": 2.249895473561155,
"learning_rate": 3.008296005493239e-06,
"loss": 0.9217,
"step": 688
},
{
"epoch": 0.6427238805970149,
"grad_norm": 2.7473706077420568,
"learning_rate": 2.994438016008785e-06,
"loss": 0.9299,
"step": 689
},
{
"epoch": 0.6436567164179104,
"grad_norm": 1.8025374398755805,
"learning_rate": 2.9805983625162226e-06,
"loss": 0.9771,
"step": 690
},
{
"epoch": 0.644589552238806,
"grad_norm": 1.7725202098877957,
"learning_rate": 2.966777171545561e-06,
"loss": 0.9531,
"step": 691
},
{
"epoch": 0.6455223880597015,
"grad_norm": 2.2296703942457334,
"learning_rate": 2.9529745694580126e-06,
"loss": 0.8184,
"step": 692
},
{
"epoch": 0.6464552238805971,
"grad_norm": 2.1973659498108233,
"learning_rate": 2.93919068244484e-06,
"loss": 0.8653,
"step": 693
},
{
"epoch": 0.6473880597014925,
"grad_norm": 1.9445919744644677,
"learning_rate": 2.925425636526203e-06,
"loss": 0.8848,
"step": 694
},
{
"epoch": 0.648320895522388,
"grad_norm": 2.239647524059142,
"learning_rate": 2.9116795575500055e-06,
"loss": 0.9447,
"step": 695
},
{
"epoch": 0.6492537313432836,
"grad_norm": 1.89769964435654,
"learning_rate": 2.897952571190743e-06,
"loss": 0.8467,
"step": 696
},
{
"epoch": 0.6501865671641791,
"grad_norm": 1.9042482614938654,
"learning_rate": 2.884244802948355e-06,
"loss": 0.9096,
"step": 697
},
{
"epoch": 0.6511194029850746,
"grad_norm": 2.134927616953055,
"learning_rate": 2.8705563781470835e-06,
"loss": 0.8369,
"step": 698
},
{
"epoch": 0.6520522388059702,
"grad_norm": 2.009303968056513,
"learning_rate": 2.8568874219343156e-06,
"loss": 0.7767,
"step": 699
},
{
"epoch": 0.6529850746268657,
"grad_norm": 1.8830074044058998,
"learning_rate": 2.8432380592794466e-06,
"loss": 0.8732,
"step": 700
},
{
"epoch": 0.6539179104477612,
"grad_norm": 1.9593930211561184,
"learning_rate": 2.8296084149727388e-06,
"loss": 0.905,
"step": 701
},
{
"epoch": 0.6548507462686567,
"grad_norm": 1.7941773935555672,
"learning_rate": 2.8159986136241737e-06,
"loss": 1.0238,
"step": 702
},
{
"epoch": 0.6557835820895522,
"grad_norm": 1.820750536323812,
"learning_rate": 2.802408779662324e-06,
"loss": 0.9645,
"step": 703
},
{
"epoch": 0.6567164179104478,
"grad_norm": 1.744260946462034,
"learning_rate": 2.788839037333203e-06,
"loss": 0.899,
"step": 704
},
{
"epoch": 0.6576492537313433,
"grad_norm": 2.0267374963485145,
"learning_rate": 2.7752895106991382e-06,
"loss": 0.9394,
"step": 705
},
{
"epoch": 0.6585820895522388,
"grad_norm": 2.124106923485193,
"learning_rate": 2.7617603236376288e-06,
"loss": 0.8849,
"step": 706
},
{
"epoch": 0.6595149253731343,
"grad_norm": 1.8312346662306787,
"learning_rate": 2.7482515998402278e-06,
"loss": 0.8738,
"step": 707
},
{
"epoch": 0.6604477611940298,
"grad_norm": 1.8949545236475038,
"learning_rate": 2.7347634628113918e-06,
"loss": 0.7779,
"step": 708
},
{
"epoch": 0.6613805970149254,
"grad_norm": 2.605517253000102,
"learning_rate": 2.7212960358673645e-06,
"loss": 0.9746,
"step": 709
},
{
"epoch": 0.6623134328358209,
"grad_norm": 2.0859322722269265,
"learning_rate": 2.7078494421350456e-06,
"loss": 0.9587,
"step": 710
},
{
"epoch": 0.6632462686567164,
"grad_norm": 2.024279071168403,
"learning_rate": 2.6944238045508708e-06,
"loss": 0.8974,
"step": 711
},
{
"epoch": 0.664179104477612,
"grad_norm": 2.0499514215618393,
"learning_rate": 2.681019245859674e-06,
"loss": 0.8346,
"step": 712
},
{
"epoch": 0.6651119402985075,
"grad_norm": 1.9047652858706219,
"learning_rate": 2.6676358886135833e-06,
"loss": 0.9204,
"step": 713
},
{
"epoch": 0.6660447761194029,
"grad_norm": 2.241906329351152,
"learning_rate": 2.654273855170883e-06,
"loss": 0.8705,
"step": 714
},
{
"epoch": 0.6669776119402985,
"grad_norm": 1.7778456432011511,
"learning_rate": 2.640933267694907e-06,
"loss": 0.9813,
"step": 715
},
{
"epoch": 0.667910447761194,
"grad_norm": 1.918650078467954,
"learning_rate": 2.6276142481529153e-06,
"loss": 0.9596,
"step": 716
},
{
"epoch": 0.6688432835820896,
"grad_norm": 1.9106725825328377,
"learning_rate": 2.6143169183149873e-06,
"loss": 0.9596,
"step": 717
},
{
"epoch": 0.6697761194029851,
"grad_norm": 1.956195352996724,
"learning_rate": 2.601041399752897e-06,
"loss": 0.8664,
"step": 718
},
{
"epoch": 0.6707089552238806,
"grad_norm": 2.2120635522149446,
"learning_rate": 2.5877878138390096e-06,
"loss": 0.9621,
"step": 719
},
{
"epoch": 0.6716417910447762,
"grad_norm": 2.0670028578244493,
"learning_rate": 2.5745562817451688e-06,
"loss": 0.8436,
"step": 720
},
{
"epoch": 0.6725746268656716,
"grad_norm": 1.7236220934363757,
"learning_rate": 2.5613469244415866e-06,
"loss": 0.8234,
"step": 721
},
{
"epoch": 0.6735074626865671,
"grad_norm": 1.7642538158974197,
"learning_rate": 2.54815986269575e-06,
"loss": 0.9517,
"step": 722
},
{
"epoch": 0.6744402985074627,
"grad_norm": 1.921538429637906,
"learning_rate": 2.5349952170712975e-06,
"loss": 0.8499,
"step": 723
},
{
"epoch": 0.6753731343283582,
"grad_norm": 2.212881483906911,
"learning_rate": 2.5218531079269316e-06,
"loss": 1.1493,
"step": 724
},
{
"epoch": 0.6763059701492538,
"grad_norm": 1.9156345337974212,
"learning_rate": 2.5087336554153087e-06,
"loss": 0.7918,
"step": 725
},
{
"epoch": 0.6772388059701493,
"grad_norm": 1.7732161254922916,
"learning_rate": 2.4956369794819536e-06,
"loss": 0.8194,
"step": 726
},
{
"epoch": 0.6781716417910447,
"grad_norm": 2.1529400844771778,
"learning_rate": 2.4825631998641474e-06,
"loss": 0.6585,
"step": 727
},
{
"epoch": 0.6791044776119403,
"grad_norm": 2.0842707596753676,
"learning_rate": 2.4695124360898427e-06,
"loss": 0.9324,
"step": 728
},
{
"epoch": 0.6800373134328358,
"grad_norm": 1.7298207955887015,
"learning_rate": 2.456484807476567e-06,
"loss": 1.0246,
"step": 729
},
{
"epoch": 0.6809701492537313,
"grad_norm": 2.144816245773138,
"learning_rate": 2.443480433130333e-06,
"loss": 0.9002,
"step": 730
},
{
"epoch": 0.6819029850746269,
"grad_norm": 2.103899226982945,
"learning_rate": 2.4304994319445536e-06,
"loss": 0.8248,
"step": 731
},
{
"epoch": 0.6828358208955224,
"grad_norm": 2.5743688892037926,
"learning_rate": 2.417541922598945e-06,
"loss": 0.8529,
"step": 732
},
{
"epoch": 0.683768656716418,
"grad_norm": 2.1404771304749777,
"learning_rate": 2.404608023558451e-06,
"loss": 1.0364,
"step": 733
},
{
"epoch": 0.6847014925373134,
"grad_norm": 2.1524378679136,
"learning_rate": 2.3916978530721567e-06,
"loss": 0.9255,
"step": 734
},
{
"epoch": 0.6856343283582089,
"grad_norm": 1.9370575876285199,
"learning_rate": 2.378811529172203e-06,
"loss": 0.88,
"step": 735
},
{
"epoch": 0.6865671641791045,
"grad_norm": 1.7592262086367514,
"learning_rate": 2.3659491696727195e-06,
"loss": 0.8402,
"step": 736
},
{
"epoch": 0.6875,
"grad_norm": 2.124655171135172,
"learning_rate": 2.353110892168733e-06,
"loss": 0.8638,
"step": 737
},
{
"epoch": 0.6884328358208955,
"grad_norm": 2.0169184318530244,
"learning_rate": 2.3402968140351014e-06,
"loss": 0.9468,
"step": 738
},
{
"epoch": 0.6893656716417911,
"grad_norm": 1.9933778143802932,
"learning_rate": 2.3275070524254358e-06,
"loss": 0.8651,
"step": 739
},
{
"epoch": 0.6902985074626866,
"grad_norm": 2.1418428498152733,
"learning_rate": 2.3147417242710368e-06,
"loss": 0.9288,
"step": 740
},
{
"epoch": 0.691231343283582,
"grad_norm": 2.7442678181643694,
"learning_rate": 2.3020009462798165e-06,
"loss": 0.8109,
"step": 741
},
{
"epoch": 0.6921641791044776,
"grad_norm": 2.023692061029926,
"learning_rate": 2.289284834935234e-06,
"loss": 0.8164,
"step": 742
},
{
"epoch": 0.6930970149253731,
"grad_norm": 2.5044659967795866,
"learning_rate": 2.276593506495239e-06,
"loss": 0.8497,
"step": 743
},
{
"epoch": 0.6940298507462687,
"grad_norm": 2.222654221009305,
"learning_rate": 2.263927076991193e-06,
"loss": 0.8217,
"step": 744
},
{
"epoch": 0.6949626865671642,
"grad_norm": 2.315623339821215,
"learning_rate": 2.2512856622268204e-06,
"loss": 0.8996,
"step": 745
},
{
"epoch": 0.6958955223880597,
"grad_norm": 2.3351357248755558,
"learning_rate": 2.238669377777151e-06,
"loss": 0.8406,
"step": 746
},
{
"epoch": 0.6968283582089553,
"grad_norm": 1.9504829743925767,
"learning_rate": 2.226078338987453e-06,
"loss": 0.9406,
"step": 747
},
{
"epoch": 0.6977611940298507,
"grad_norm": 2.205381071299871,
"learning_rate": 2.213512660972187e-06,
"loss": 0.9402,
"step": 748
},
{
"epoch": 0.6986940298507462,
"grad_norm": 1.9589970297609565,
"learning_rate": 2.2009724586139486e-06,
"loss": 0.8048,
"step": 749
},
{
"epoch": 0.6996268656716418,
"grad_norm": 2.058178874128284,
"learning_rate": 2.1884578465624257e-06,
"loss": 0.892,
"step": 750
},
{
"epoch": 0.7005597014925373,
"grad_norm": 2.526913972445055,
"learning_rate": 2.17596893923334e-06,
"loss": 0.8973,
"step": 751
},
{
"epoch": 0.7014925373134329,
"grad_norm": 2.1252483981220087,
"learning_rate": 2.1635058508074074e-06,
"loss": 0.8724,
"step": 752
},
{
"epoch": 0.7024253731343284,
"grad_norm": 2.0077730378750243,
"learning_rate": 2.1510686952292936e-06,
"loss": 0.8304,
"step": 753
},
{
"epoch": 0.7033582089552238,
"grad_norm": 1.9613384497067794,
"learning_rate": 2.1386575862065677e-06,
"loss": 0.8092,
"step": 754
},
{
"epoch": 0.7042910447761194,
"grad_norm": 1.634409115799417,
"learning_rate": 2.126272637208674e-06,
"loss": 0.7804,
"step": 755
},
{
"epoch": 0.7052238805970149,
"grad_norm": 2.13346609729719,
"learning_rate": 2.11391396146588e-06,
"loss": 0.9695,
"step": 756
},
{
"epoch": 0.7061567164179104,
"grad_norm": 2.547238461994279,
"learning_rate": 2.1015816719682497e-06,
"loss": 0.9961,
"step": 757
},
{
"epoch": 0.707089552238806,
"grad_norm": 1.5781702063469576,
"learning_rate": 2.089275881464609e-06,
"loss": 0.7731,
"step": 758
},
{
"epoch": 0.7080223880597015,
"grad_norm": 2.173966506499643,
"learning_rate": 2.0769967024615185e-06,
"loss": 0.9578,
"step": 759
},
{
"epoch": 0.7089552238805971,
"grad_norm": 1.843650078543119,
"learning_rate": 2.064744247222236e-06,
"loss": 0.8981,
"step": 760
},
{
"epoch": 0.7098880597014925,
"grad_norm": 1.8787256412153666,
"learning_rate": 2.052518627765699e-06,
"loss": 0.8481,
"step": 761
},
{
"epoch": 0.710820895522388,
"grad_norm": 2.395612321162723,
"learning_rate": 2.0403199558654947e-06,
"loss": 0.8929,
"step": 762
},
{
"epoch": 0.7117537313432836,
"grad_norm": 2.0192512292025087,
"learning_rate": 2.0281483430488393e-06,
"loss": 0.8338,
"step": 763
},
{
"epoch": 0.7126865671641791,
"grad_norm": 1.6241502382373327,
"learning_rate": 2.016003900595566e-06,
"loss": 0.8173,
"step": 764
},
{
"epoch": 0.7136194029850746,
"grad_norm": 2.2650305185997284,
"learning_rate": 2.0038867395370937e-06,
"loss": 0.8381,
"step": 765
},
{
"epoch": 0.7145522388059702,
"grad_norm": 1.8330361802805586,
"learning_rate": 1.9917969706554226e-06,
"loss": 0.8554,
"step": 766
},
{
"epoch": 0.7154850746268657,
"grad_norm": 2.20919453483712,
"learning_rate": 1.9797347044821176e-06,
"loss": 0.8639,
"step": 767
},
{
"epoch": 0.7164179104477612,
"grad_norm": 3.2646044051136953,
"learning_rate": 1.967700051297295e-06,
"loss": 0.8328,
"step": 768
},
{
"epoch": 0.7173507462686567,
"grad_norm": 2.0800183552431086,
"learning_rate": 1.9556931211286255e-06,
"loss": 0.908,
"step": 769
},
{
"epoch": 0.7182835820895522,
"grad_norm": 1.9883048826975065,
"learning_rate": 1.9437140237503122e-06,
"loss": 0.8798,
"step": 770
},
{
"epoch": 0.7192164179104478,
"grad_norm": 1.8968524059275165,
"learning_rate": 1.931762868682098e-06,
"loss": 0.7726,
"step": 771
},
{
"epoch": 0.7201492537313433,
"grad_norm": 2.161378053347994,
"learning_rate": 1.9198397651882594e-06,
"loss": 1.0271,
"step": 772
},
{
"epoch": 0.7210820895522388,
"grad_norm": 2.208792592850636,
"learning_rate": 1.9079448222766127e-06,
"loss": 0.8829,
"step": 773
},
{
"epoch": 0.7220149253731343,
"grad_norm": 2.2266876324454756,
"learning_rate": 1.8960781486975143e-06,
"loss": 0.8873,
"step": 774
},
{
"epoch": 0.7229477611940298,
"grad_norm": 1.9678799515621732,
"learning_rate": 1.8842398529428625e-06,
"loss": 0.8655,
"step": 775
},
{
"epoch": 0.7238805970149254,
"grad_norm": 1.846054114752651,
"learning_rate": 1.8724300432451109e-06,
"loss": 0.8888,
"step": 776
},
{
"epoch": 0.7248134328358209,
"grad_norm": 1.8259322923247736,
"learning_rate": 1.860648827576278e-06,
"loss": 0.8726,
"step": 777
},
{
"epoch": 0.7257462686567164,
"grad_norm": 2.09958949295602,
"learning_rate": 1.8488963136469573e-06,
"loss": 0.9246,
"step": 778
},
{
"epoch": 0.726679104477612,
"grad_norm": 1.821216380202577,
"learning_rate": 1.8371726089053399e-06,
"loss": 0.8441,
"step": 779
},
{
"epoch": 0.7276119402985075,
"grad_norm": 1.874742026539673,
"learning_rate": 1.8254778205362205e-06,
"loss": 0.7805,
"step": 780
},
{
"epoch": 0.7285447761194029,
"grad_norm": 2.080728361890937,
"learning_rate": 1.813812055460028e-06,
"loss": 0.8345,
"step": 781
},
{
"epoch": 0.7294776119402985,
"grad_norm": 1.8098226808576026,
"learning_rate": 1.8021754203318403e-06,
"loss": 0.7921,
"step": 782
},
{
"epoch": 0.730410447761194,
"grad_norm": 1.9048931836633622,
"learning_rate": 1.7905680215404176e-06,
"loss": 0.9203,
"step": 783
},
{
"epoch": 0.7313432835820896,
"grad_norm": 2.0163104331428623,
"learning_rate": 1.7789899652072206e-06,
"loss": 0.9927,
"step": 784
},
{
"epoch": 0.7322761194029851,
"grad_norm": 2.0865194974641024,
"learning_rate": 1.767441357185446e-06,
"loss": 0.9128,
"step": 785
},
{
"epoch": 0.7332089552238806,
"grad_norm": 2.378284170292539,
"learning_rate": 1.7559223030590578e-06,
"loss": 0.972,
"step": 786
},
{
"epoch": 0.7341417910447762,
"grad_norm": 2.290689066691905,
"learning_rate": 1.7444329081418176e-06,
"loss": 1.006,
"step": 787
},
{
"epoch": 0.7350746268656716,
"grad_norm": 1.984870845374516,
"learning_rate": 1.7329732774763342e-06,
"loss": 0.8681,
"step": 788
},
{
"epoch": 0.7360074626865671,
"grad_norm": 1.8442748485520641,
"learning_rate": 1.7215435158330856e-06,
"loss": 0.9045,
"step": 789
},
{
"epoch": 0.7369402985074627,
"grad_norm": 1.8317682619238482,
"learning_rate": 1.7101437277094746e-06,
"loss": 0.9174,
"step": 790
},
{
"epoch": 0.7378731343283582,
"grad_norm": 2.0828052658229654,
"learning_rate": 1.698774017328867e-06,
"loss": 0.7999,
"step": 791
},
{
"epoch": 0.7388059701492538,
"grad_norm": 2.0509264510010996,
"learning_rate": 1.6874344886396438e-06,
"loss": 0.8701,
"step": 792
},
{
"epoch": 0.7397388059701493,
"grad_norm": 2.038563928777798,
"learning_rate": 1.6761252453142458e-06,
"loss": 0.9522,
"step": 793
},
{
"epoch": 0.7406716417910447,
"grad_norm": 1.7710083345510803,
"learning_rate": 1.6648463907482264e-06,
"loss": 0.9183,
"step": 794
},
{
"epoch": 0.7416044776119403,
"grad_norm": 2.0634078284870574,
"learning_rate": 1.65359802805931e-06,
"loss": 0.8284,
"step": 795
},
{
"epoch": 0.7425373134328358,
"grad_norm": 2.048676474874421,
"learning_rate": 1.6423802600864436e-06,
"loss": 0.8972,
"step": 796
},
{
"epoch": 0.7434701492537313,
"grad_norm": 1.8524635718033073,
"learning_rate": 1.631193189388865e-06,
"loss": 0.8437,
"step": 797
},
{
"epoch": 0.7444029850746269,
"grad_norm": 1.8299093925447154,
"learning_rate": 1.6200369182451564e-06,
"loss": 0.8899,
"step": 798
},
{
"epoch": 0.7453358208955224,
"grad_norm": 2.4249559792426543,
"learning_rate": 1.6089115486523133e-06,
"loss": 0.9083,
"step": 799
},
{
"epoch": 0.746268656716418,
"grad_norm": 2.0521527734845124,
"learning_rate": 1.5978171823248111e-06,
"loss": 0.8584,
"step": 800
},
{
"epoch": 0.7472014925373134,
"grad_norm": 1.7922796138710044,
"learning_rate": 1.586753920693676e-06,
"loss": 0.9391,
"step": 801
},
{
"epoch": 0.7481343283582089,
"grad_norm": 2.090798511928835,
"learning_rate": 1.5757218649055605e-06,
"loss": 0.8557,
"step": 802
},
{
"epoch": 0.7490671641791045,
"grad_norm": 2.3675386182155367,
"learning_rate": 1.5647211158218107e-06,
"loss": 0.8192,
"step": 803
},
{
"epoch": 0.75,
"grad_norm": 2.4067523019278565,
"learning_rate": 1.5537517740175512e-06,
"loss": 0.9228,
"step": 804
},
{
"epoch": 0.7509328358208955,
"grad_norm": 2.136851935872237,
"learning_rate": 1.5428139397807619e-06,
"loss": 1.0393,
"step": 805
},
{
"epoch": 0.7518656716417911,
"grad_norm": 2.2343128147507416,
"learning_rate": 1.5319077131113652e-06,
"loss": 0.944,
"step": 806
},
{
"epoch": 0.7527985074626866,
"grad_norm": 1.7797343977412017,
"learning_rate": 1.5210331937203087e-06,
"loss": 0.8214,
"step": 807
},
{
"epoch": 0.753731343283582,
"grad_norm": 1.9314774771052028,
"learning_rate": 1.5101904810286522e-06,
"loss": 0.8783,
"step": 808
},
{
"epoch": 0.7546641791044776,
"grad_norm": 1.878598232035822,
"learning_rate": 1.499379674166661e-06,
"loss": 0.8567,
"step": 809
},
{
"epoch": 0.7555970149253731,
"grad_norm": 1.71338723783142,
"learning_rate": 1.488600871972899e-06,
"loss": 0.8019,
"step": 810
},
{
"epoch": 0.7565298507462687,
"grad_norm": 1.7185851197007331,
"learning_rate": 1.4778541729933239e-06,
"loss": 0.8188,
"step": 811
},
{
"epoch": 0.7574626865671642,
"grad_norm": 1.9120353501989646,
"learning_rate": 1.467139675480392e-06,
"loss": 0.8561,
"step": 812
},
{
"epoch": 0.7583955223880597,
"grad_norm": 1.9076171880604034,
"learning_rate": 1.4564574773921514e-06,
"loss": 0.9199,
"step": 813
},
{
"epoch": 0.7593283582089553,
"grad_norm": 1.8168563080400306,
"learning_rate": 1.445807676391352e-06,
"loss": 0.8822,
"step": 814
},
{
"epoch": 0.7602611940298507,
"grad_norm": 1.970361344874575,
"learning_rate": 1.435190369844548e-06,
"loss": 0.8485,
"step": 815
},
{
"epoch": 0.7611940298507462,
"grad_norm": 1.8000972392963275,
"learning_rate": 1.4246056548212173e-06,
"loss": 0.8063,
"step": 816
},
{
"epoch": 0.7621268656716418,
"grad_norm": 1.8236958009027526,
"learning_rate": 1.4140536280928614e-06,
"loss": 0.7569,
"step": 817
},
{
"epoch": 0.7630597014925373,
"grad_norm": 1.6239179928588776,
"learning_rate": 1.4035343861321293e-06,
"loss": 0.8573,
"step": 818
},
{
"epoch": 0.7639925373134329,
"grad_norm": 1.8407412539189518,
"learning_rate": 1.393048025111932e-06,
"loss": 0.916,
"step": 819
},
{
"epoch": 0.7649253731343284,
"grad_norm": 1.9539831506616605,
"learning_rate": 1.3825946409045642e-06,
"loss": 0.8699,
"step": 820
},
{
"epoch": 0.7658582089552238,
"grad_norm": 2.1746988422173033,
"learning_rate": 1.3721743290808304e-06,
"loss": 0.8868,
"step": 821
},
{
"epoch": 0.7667910447761194,
"grad_norm": 1.9681731149696198,
"learning_rate": 1.3617871849091657e-06,
"loss": 0.9456,
"step": 822
},
{
"epoch": 0.7677238805970149,
"grad_norm": 2.123765369759133,
"learning_rate": 1.3514333033547671e-06,
"loss": 0.8701,
"step": 823
},
{
"epoch": 0.7686567164179104,
"grad_norm": 1.8826900475134194,
"learning_rate": 1.3411127790787265e-06,
"loss": 0.7898,
"step": 824
},
{
"epoch": 0.769589552238806,
"grad_norm": 1.651498724244396,
"learning_rate": 1.3308257064371678e-06,
"loss": 0.8729,
"step": 825
},
{
"epoch": 0.7705223880597015,
"grad_norm": 1.7478518320861869,
"learning_rate": 1.3205721794803766e-06,
"loss": 0.8918,
"step": 826
},
{
"epoch": 0.7714552238805971,
"grad_norm": 2.0213033182709164,
"learning_rate": 1.3103522919519463e-06,
"loss": 0.9061,
"step": 827
},
{
"epoch": 0.7723880597014925,
"grad_norm": 1.9093938260582595,
"learning_rate": 1.3001661372879194e-06,
"loss": 0.9188,
"step": 828
},
{
"epoch": 0.773320895522388,
"grad_norm": 1.7216945296247417,
"learning_rate": 1.2900138086159315e-06,
"loss": 0.8822,
"step": 829
},
{
"epoch": 0.7742537313432836,
"grad_norm": 1.9179031896968435,
"learning_rate": 1.2798953987543672e-06,
"loss": 0.8487,
"step": 830
},
{
"epoch": 0.7751865671641791,
"grad_norm": 1.7145754323703066,
"learning_rate": 1.2698110002115005e-06,
"loss": 0.8311,
"step": 831
},
{
"epoch": 0.7761194029850746,
"grad_norm": 1.8161538237089532,
"learning_rate": 1.259760705184656e-06,
"loss": 0.8545,
"step": 832
},
{
"epoch": 0.7770522388059702,
"grad_norm": 1.7974548107543482,
"learning_rate": 1.249744605559366e-06,
"loss": 0.819,
"step": 833
},
{
"epoch": 0.7779850746268657,
"grad_norm": 1.943050686225747,
"learning_rate": 1.2397627929085249e-06,
"loss": 0.8519,
"step": 834
},
{
"epoch": 0.7789179104477612,
"grad_norm": 2.393825224277449,
"learning_rate": 1.2298153584915623e-06,
"loss": 0.9091,
"step": 835
},
{
"epoch": 0.7798507462686567,
"grad_norm": 2.6021337150994124,
"learning_rate": 1.2199023932535964e-06,
"loss": 0.9185,
"step": 836
},
{
"epoch": 0.7807835820895522,
"grad_norm": 2.0227686089340895,
"learning_rate": 1.2100239878246134e-06,
"loss": 0.9101,
"step": 837
},
{
"epoch": 0.7817164179104478,
"grad_norm": 1.8921631606844307,
"learning_rate": 1.2001802325186307e-06,
"loss": 0.8704,
"step": 838
},
{
"epoch": 0.7826492537313433,
"grad_norm": 2.000647559245294,
"learning_rate": 1.1903712173328742e-06,
"loss": 0.8882,
"step": 839
},
{
"epoch": 0.7835820895522388,
"grad_norm": 1.8498233704812672,
"learning_rate": 1.180597031946959e-06,
"loss": 0.8857,
"step": 840
},
{
"epoch": 0.7845149253731343,
"grad_norm": 1.8298531421677997,
"learning_rate": 1.1708577657220637e-06,
"loss": 0.883,
"step": 841
},
{
"epoch": 0.7854477611940298,
"grad_norm": 1.94239728000116,
"learning_rate": 1.1611535077001168e-06,
"loss": 0.8234,
"step": 842
},
{
"epoch": 0.7863805970149254,
"grad_norm": 1.999625903422434,
"learning_rate": 1.1514843466029806e-06,
"loss": 0.8962,
"step": 843
},
{
"epoch": 0.7873134328358209,
"grad_norm": 2.394867814962891,
"learning_rate": 1.1418503708316426e-06,
"loss": 0.984,
"step": 844
},
{
"epoch": 0.7882462686567164,
"grad_norm": 1.6483774472377923,
"learning_rate": 1.1322516684654073e-06,
"loss": 0.7192,
"step": 845
},
{
"epoch": 0.789179104477612,
"grad_norm": 1.7917597473668263,
"learning_rate": 1.1226883272610877e-06,
"loss": 0.9523,
"step": 846
},
{
"epoch": 0.7901119402985075,
"grad_norm": 2.0785274480833533,
"learning_rate": 1.113160434652204e-06,
"loss": 1.0035,
"step": 847
},
{
"epoch": 0.7910447761194029,
"grad_norm": 1.863827394028924,
"learning_rate": 1.1036680777481862e-06,
"loss": 0.8988,
"step": 848
},
{
"epoch": 0.7919776119402985,
"grad_norm": 1.7898497868674286,
"learning_rate": 1.0942113433335787e-06,
"loss": 0.81,
"step": 849
},
{
"epoch": 0.792910447761194,
"grad_norm": 2.238815981519986,
"learning_rate": 1.0847903178672426e-06,
"loss": 0.8414,
"step": 850
},
{
"epoch": 0.7938432835820896,
"grad_norm": 2.220247721243224,
"learning_rate": 1.0754050874815664e-06,
"loss": 0.9011,
"step": 851
},
{
"epoch": 0.7947761194029851,
"grad_norm": 2.2267779206060068,
"learning_rate": 1.0660557379816816e-06,
"loss": 0.9774,
"step": 852
},
{
"epoch": 0.7957089552238806,
"grad_norm": 1.9281376811985778,
"learning_rate": 1.0567423548446736e-06,
"loss": 0.8592,
"step": 853
},
{
"epoch": 0.7966417910447762,
"grad_norm": 2.229505139699696,
"learning_rate": 1.0474650232188083e-06,
"loss": 0.9678,
"step": 854
},
{
"epoch": 0.7975746268656716,
"grad_norm": 2.099331064470327,
"learning_rate": 1.0382238279227418e-06,
"loss": 0.9765,
"step": 855
},
{
"epoch": 0.7985074626865671,
"grad_norm": 1.9908860160518558,
"learning_rate": 1.0290188534447542e-06,
"loss": 0.9471,
"step": 856
},
{
"epoch": 0.7994402985074627,
"grad_norm": 2.028268236909121,
"learning_rate": 1.0198501839419733e-06,
"loss": 0.9985,
"step": 857
},
{
"epoch": 0.8003731343283582,
"grad_norm": 2.170737985360099,
"learning_rate": 1.0107179032396086e-06,
"loss": 0.8475,
"step": 858
},
{
"epoch": 0.8013059701492538,
"grad_norm": 2.006177667446587,
"learning_rate": 1.0016220948301797e-06,
"loss": 0.8487,
"step": 859
},
{
"epoch": 0.8022388059701493,
"grad_norm": 1.6272093583111642,
"learning_rate": 9.925628418727563e-07,
"loss": 1.0067,
"step": 860
},
{
"epoch": 0.8031716417910447,
"grad_norm": 1.6832575070536506,
"learning_rate": 9.835402271921974e-07,
"loss": 0.8219,
"step": 861
},
{
"epoch": 0.8041044776119403,
"grad_norm": 2.3428893240431985,
"learning_rate": 9.74554333278393e-07,
"loss": 0.8559,
"step": 862
},
{
"epoch": 0.8050373134328358,
"grad_norm": 2.0034097300461506,
"learning_rate": 9.656052422855143e-07,
"loss": 0.7971,
"step": 863
},
{
"epoch": 0.8059701492537313,
"grad_norm": 1.9590721352474731,
"learning_rate": 9.56693036031256e-07,
"loss": 0.8244,
"step": 864
},
{
"epoch": 0.8069029850746269,
"grad_norm": 1.9266670860267583,
"learning_rate": 9.478177959960916e-07,
"loss": 0.8773,
"step": 865
},
{
"epoch": 0.8078358208955224,
"grad_norm": 2.298474766736844,
"learning_rate": 9.389796033225301e-07,
"loss": 0.9344,
"step": 866
},
{
"epoch": 0.808768656716418,
"grad_norm": 1.8377844202977682,
"learning_rate": 9.301785388143697e-07,
"loss": 0.7904,
"step": 867
},
{
"epoch": 0.8097014925373134,
"grad_norm": 1.8130734819322119,
"learning_rate": 9.214146829359644e-07,
"loss": 0.8956,
"step": 868
},
{
"epoch": 0.8106343283582089,
"grad_norm": 2.1231702459173127,
"learning_rate": 9.126881158114859e-07,
"loss": 0.8578,
"step": 869
},
{
"epoch": 0.8115671641791045,
"grad_norm": 1.781928253428776,
"learning_rate": 9.039989172241887e-07,
"loss": 0.9027,
"step": 870
},
{
"epoch": 0.8125,
"grad_norm": 1.96092850928182,
"learning_rate": 8.953471666156843e-07,
"loss": 0.8284,
"step": 871
},
{
"epoch": 0.8134328358208955,
"grad_norm": 2.1300468443484686,
"learning_rate": 8.867329430852101e-07,
"loss": 0.9251,
"step": 872
},
{
"epoch": 0.8143656716417911,
"grad_norm": 1.7910403904713712,
"learning_rate": 8.781563253889164e-07,
"loss": 0.7739,
"step": 873
},
{
"epoch": 0.8152985074626866,
"grad_norm": 2.315797508659485,
"learning_rate": 8.696173919391326e-07,
"loss": 0.8794,
"step": 874
},
{
"epoch": 0.816231343283582,
"grad_norm": 1.8802418177229716,
"learning_rate": 8.611162208036617e-07,
"loss": 0.9443,
"step": 875
},
{
"epoch": 0.8171641791044776,
"grad_norm": 2.3203711918379177,
"learning_rate": 8.52652889705059e-07,
"loss": 0.7846,
"step": 876
},
{
"epoch": 0.8180970149253731,
"grad_norm": 1.7143710254250017,
"learning_rate": 8.442274760199254e-07,
"loss": 0.9062,
"step": 877
},
{
"epoch": 0.8190298507462687,
"grad_norm": 2.174355310633949,
"learning_rate": 8.358400567782021e-07,
"loss": 0.8813,
"step": 878
},
{
"epoch": 0.8199626865671642,
"grad_norm": 2.0882641649773017,
"learning_rate": 8.274907086624595e-07,
"loss": 0.8697,
"step": 879
},
{
"epoch": 0.8208955223880597,
"grad_norm": 1.8963249388701662,
"learning_rate": 8.191795080072023e-07,
"loss": 0.8245,
"step": 880
},
{
"epoch": 0.8218283582089553,
"grad_norm": 2.1540141669507467,
"learning_rate": 8.109065307981673e-07,
"loss": 0.9356,
"step": 881
},
{
"epoch": 0.8227611940298507,
"grad_norm": 2.001302098385953,
"learning_rate": 8.026718526716343e-07,
"loss": 0.7907,
"step": 882
},
{
"epoch": 0.8236940298507462,
"grad_norm": 1.876240915642379,
"learning_rate": 7.944755489137279e-07,
"loss": 0.8852,
"step": 883
},
{
"epoch": 0.8246268656716418,
"grad_norm": 1.7837734368702272,
"learning_rate": 7.863176944597329e-07,
"loss": 0.8433,
"step": 884
},
{
"epoch": 0.8255597014925373,
"grad_norm": 2.0434835671727556,
"learning_rate": 7.781983638934093e-07,
"loss": 0.8648,
"step": 885
},
{
"epoch": 0.8264925373134329,
"grad_norm": 2.1341691595298555,
"learning_rate": 7.701176314463088e-07,
"loss": 0.8456,
"step": 886
},
{
"epoch": 0.8274253731343284,
"grad_norm": 1.9640000882309503,
"learning_rate": 7.62075570997099e-07,
"loss": 0.9038,
"step": 887
},
{
"epoch": 0.8283582089552238,
"grad_norm": 2.4597138975960777,
"learning_rate": 7.540722560708841e-07,
"loss": 0.9993,
"step": 888
},
{
"epoch": 0.8292910447761194,
"grad_norm": 2.464339245702074,
"learning_rate": 7.461077598385357e-07,
"loss": 0.9215,
"step": 889
},
{
"epoch": 0.8302238805970149,
"grad_norm": 1.7717405215501345,
"learning_rate": 7.381821551160207e-07,
"loss": 0.9129,
"step": 890
},
{
"epoch": 0.8311567164179104,
"grad_norm": 2.6578705461487373,
"learning_rate": 7.302955143637419e-07,
"loss": 0.8705,
"step": 891
},
{
"epoch": 0.832089552238806,
"grad_norm": 2.052066691153167,
"learning_rate": 7.224479096858672e-07,
"loss": 0.8443,
"step": 892
},
{
"epoch": 0.8330223880597015,
"grad_norm": 1.8659593232579255,
"learning_rate": 7.146394128296758e-07,
"loss": 0.878,
"step": 893
},
{
"epoch": 0.8339552238805971,
"grad_norm": 1.7932296021511334,
"learning_rate": 7.068700951849012e-07,
"loss": 0.9038,
"step": 894
},
{
"epoch": 0.8348880597014925,
"grad_norm": 1.7153801494008183,
"learning_rate": 6.991400277830768e-07,
"loss": 0.9808,
"step": 895
},
{
"epoch": 0.835820895522388,
"grad_norm": 2.066402980737668,
"learning_rate": 6.91449281296891e-07,
"loss": 0.8694,
"step": 896
},
{
"epoch": 0.8367537313432836,
"grad_norm": 1.8596656505868,
"learning_rate": 6.83797926039535e-07,
"loss": 0.7881,
"step": 897
},
{
"epoch": 0.8376865671641791,
"grad_norm": 1.9262125814612836,
"learning_rate": 6.761860319640639e-07,
"loss": 0.8184,
"step": 898
},
{
"epoch": 0.8386194029850746,
"grad_norm": 1.7725965184719716,
"learning_rate": 6.686136686627554e-07,
"loss": 0.8323,
"step": 899
},
{
"epoch": 0.8395522388059702,
"grad_norm": 2.197637474079969,
"learning_rate": 6.610809053664768e-07,
"loss": 0.9281,
"step": 900
},
{
"epoch": 0.8404850746268657,
"grad_norm": 1.4250596331019252,
"learning_rate": 6.535878109440458e-07,
"loss": 0.8846,
"step": 901
},
{
"epoch": 0.8414179104477612,
"grad_norm": 2.065810300580414,
"learning_rate": 6.46134453901609e-07,
"loss": 0.9134,
"step": 902
},
{
"epoch": 0.8423507462686567,
"grad_norm": 2.4999189638656962,
"learning_rate": 6.387209023820073e-07,
"loss": 0.863,
"step": 903
},
{
"epoch": 0.8432835820895522,
"grad_norm": 1.8767963799434837,
"learning_rate": 6.313472241641588e-07,
"loss": 0.7337,
"step": 904
},
{
"epoch": 0.8442164179104478,
"grad_norm": 1.7266702096213713,
"learning_rate": 6.240134866624353e-07,
"loss": 0.8397,
"step": 905
},
{
"epoch": 0.8451492537313433,
"grad_norm": 1.9806154533087825,
"learning_rate": 6.16719756926052e-07,
"loss": 0.8235,
"step": 906
},
{
"epoch": 0.8460820895522388,
"grad_norm": 1.9383073484543576,
"learning_rate": 6.094661016384451e-07,
"loss": 0.857,
"step": 907
},
{
"epoch": 0.8470149253731343,
"grad_norm": 1.9344031420617773,
"learning_rate": 6.022525871166706e-07,
"loss": 0.8795,
"step": 908
},
{
"epoch": 0.8479477611940298,
"grad_norm": 1.8048413622452402,
"learning_rate": 5.950792793107935e-07,
"loss": 0.7982,
"step": 909
},
{
"epoch": 0.8488805970149254,
"grad_norm": 1.898080825549606,
"learning_rate": 5.879462438032857e-07,
"loss": 0.8603,
"step": 910
},
{
"epoch": 0.8498134328358209,
"grad_norm": 2.417606287406364,
"learning_rate": 5.808535458084292e-07,
"loss": 0.9173,
"step": 911
},
{
"epoch": 0.8507462686567164,
"grad_norm": 1.8118897515244685,
"learning_rate": 5.738012501717144e-07,
"loss": 0.8185,
"step": 912
},
{
"epoch": 0.851679104477612,
"grad_norm": 1.8894736947825495,
"learning_rate": 5.667894213692515e-07,
"loss": 0.7958,
"step": 913
},
{
"epoch": 0.8526119402985075,
"grad_norm": 1.4917980399619162,
"learning_rate": 5.598181235071798e-07,
"loss": 0.7658,
"step": 914
},
{
"epoch": 0.8535447761194029,
"grad_norm": 2.08842168384019,
"learning_rate": 5.528874203210827e-07,
"loss": 0.8593,
"step": 915
},
{
"epoch": 0.8544776119402985,
"grad_norm": 2.212260840870776,
"learning_rate": 5.459973751754022e-07,
"loss": 0.8659,
"step": 916
},
{
"epoch": 0.855410447761194,
"grad_norm": 1.8547395390203762,
"learning_rate": 5.391480510628622e-07,
"loss": 0.879,
"step": 917
},
{
"epoch": 0.8563432835820896,
"grad_norm": 2.40394669113281,
"learning_rate": 5.323395106038909e-07,
"loss": 0.7668,
"step": 918
},
{
"epoch": 0.8572761194029851,
"grad_norm": 2.2337334020187236,
"learning_rate": 5.255718160460493e-07,
"loss": 0.8322,
"step": 919
},
{
"epoch": 0.8582089552238806,
"grad_norm": 2.1222049086348442,
"learning_rate": 5.188450292634645e-07,
"loss": 0.7797,
"step": 920
},
{
"epoch": 0.8591417910447762,
"grad_norm": 2.2363739742720896,
"learning_rate": 5.121592117562574e-07,
"loss": 0.868,
"step": 921
},
{
"epoch": 0.8600746268656716,
"grad_norm": 1.7846637435651487,
"learning_rate": 5.05514424649986e-07,
"loss": 0.8542,
"step": 922
},
{
"epoch": 0.8610074626865671,
"grad_norm": 2.0060433449251045,
"learning_rate": 4.989107286950845e-07,
"loss": 0.8534,
"step": 923
},
{
"epoch": 0.8619402985074627,
"grad_norm": 2.1793841674247236,
"learning_rate": 4.923481842663114e-07,
"loss": 0.8672,
"step": 924
},
{
"epoch": 0.8628731343283582,
"grad_norm": 1.6275952597263428,
"learning_rate": 4.8582685136219e-07,
"loss": 0.9215,
"step": 925
},
{
"epoch": 0.8638059701492538,
"grad_norm": 2.4348476023632823,
"learning_rate": 4.793467896044668e-07,
"loss": 0.8519,
"step": 926
},
{
"epoch": 0.8647388059701493,
"grad_norm": 1.7103808468811068,
"learning_rate": 4.729080582375634e-07,
"loss": 0.7784,
"step": 927
},
{
"epoch": 0.8656716417910447,
"grad_norm": 1.9075245448017746,
"learning_rate": 4.6651071612803466e-07,
"loss": 1.0887,
"step": 928
},
{
"epoch": 0.8666044776119403,
"grad_norm": 1.9943448724873647,
"learning_rate": 4.6015482176403337e-07,
"loss": 0.8878,
"step": 929
},
{
"epoch": 0.8675373134328358,
"grad_norm": 2.174968494708173,
"learning_rate": 4.5384043325477187e-07,
"loss": 0.8469,
"step": 930
},
{
"epoch": 0.8684701492537313,
"grad_norm": 2.1933052981119747,
"learning_rate": 4.4756760832999126e-07,
"loss": 0.8703,
"step": 931
},
{
"epoch": 0.8694029850746269,
"grad_norm": 1.689522579329061,
"learning_rate": 4.413364043394375e-07,
"loss": 0.9106,
"step": 932
},
{
"epoch": 0.8703358208955224,
"grad_norm": 2.699741149016076,
"learning_rate": 4.3514687825233157e-07,
"loss": 0.9033,
"step": 933
},
{
"epoch": 0.871268656716418,
"grad_norm": 2.0316558756604706,
"learning_rate": 4.289990866568516e-07,
"loss": 0.8633,
"step": 934
},
{
"epoch": 0.8722014925373134,
"grad_norm": 1.9980641212316246,
"learning_rate": 4.2289308575961805e-07,
"loss": 0.7647,
"step": 935
},
{
"epoch": 0.8731343283582089,
"grad_norm": 2.2705293720345665,
"learning_rate": 4.168289313851731e-07,
"loss": 0.88,
"step": 936
},
{
"epoch": 0.8740671641791045,
"grad_norm": 2.3106889755154967,
"learning_rate": 4.108066789754761e-07,
"loss": 0.9409,
"step": 937
},
{
"epoch": 0.875,
"grad_norm": 2.308052910572831,
"learning_rate": 4.0482638358939397e-07,
"loss": 0.9241,
"step": 938
},
{
"epoch": 0.8759328358208955,
"grad_norm": 2.1082950679412686,
"learning_rate": 3.988880999021999e-07,
"loss": 0.9785,
"step": 939
},
{
"epoch": 0.8768656716417911,
"grad_norm": 2.1756272495242674,
"learning_rate": 3.9299188220507157e-07,
"loss": 0.9698,
"step": 940
},
{
"epoch": 0.8777985074626866,
"grad_norm": 1.750550617288851,
"learning_rate": 3.8713778440459484e-07,
"loss": 0.9104,
"step": 941
},
{
"epoch": 0.878731343283582,
"grad_norm": 1.9014273448809333,
"learning_rate": 3.813258600222708e-07,
"loss": 0.9114,
"step": 942
},
{
"epoch": 0.8796641791044776,
"grad_norm": 2.0014605395097864,
"learning_rate": 3.7555616219402847e-07,
"loss": 0.9117,
"step": 943
},
{
"epoch": 0.8805970149253731,
"grad_norm": 1.689037556973581,
"learning_rate": 3.6982874366973785e-07,
"loss": 0.8836,
"step": 944
},
{
"epoch": 0.8815298507462687,
"grad_norm": 2.041328908856196,
"learning_rate": 3.641436568127271e-07,
"loss": 0.7733,
"step": 945
},
{
"epoch": 0.8824626865671642,
"grad_norm": 1.9218992196508813,
"learning_rate": 3.585009535993023e-07,
"loss": 0.9352,
"step": 946
},
{
"epoch": 0.8833955223880597,
"grad_norm": 2.0440098461795846,
"learning_rate": 3.5290068561827596e-07,
"loss": 0.834,
"step": 947
},
{
"epoch": 0.8843283582089553,
"grad_norm": 1.9228061615751502,
"learning_rate": 3.47342904070494e-07,
"loss": 0.9896,
"step": 948
},
{
"epoch": 0.8852611940298507,
"grad_norm": 1.9774017102870258,
"learning_rate": 3.418276597683662e-07,
"loss": 0.7365,
"step": 949
},
{
"epoch": 0.8861940298507462,
"grad_norm": 1.741318099023886,
"learning_rate": 3.3635500313540124e-07,
"loss": 0.8504,
"step": 950
},
{
"epoch": 0.8871268656716418,
"grad_norm": 2.169478451517118,
"learning_rate": 3.3092498420574994e-07,
"loss": 0.9195,
"step": 951
},
{
"epoch": 0.8880597014925373,
"grad_norm": 2.0346999191062185,
"learning_rate": 3.255376526237414e-07,
"loss": 0.8532,
"step": 952
},
{
"epoch": 0.8889925373134329,
"grad_norm": 1.793056843111332,
"learning_rate": 3.2019305764343644e-07,
"loss": 0.887,
"step": 953
},
{
"epoch": 0.8899253731343284,
"grad_norm": 2.101549538125784,
"learning_rate": 3.148912481281713e-07,
"loss": 0.8808,
"step": 954
},
{
"epoch": 0.8908582089552238,
"grad_norm": 2.13132339760495,
"learning_rate": 3.0963227255011273e-07,
"loss": 0.8865,
"step": 955
},
{
"epoch": 0.8917910447761194,
"grad_norm": 1.8995553856143481,
"learning_rate": 3.044161789898159e-07,
"loss": 0.8628,
"step": 956
},
{
"epoch": 0.8927238805970149,
"grad_norm": 1.6984488229016401,
"learning_rate": 2.99243015135785e-07,
"loss": 0.8563,
"step": 957
},
{
"epoch": 0.8936567164179104,
"grad_norm": 1.8077447985448474,
"learning_rate": 2.941128282840344e-07,
"loss": 0.807,
"step": 958
},
{
"epoch": 0.894589552238806,
"grad_norm": 1.980086980444361,
"learning_rate": 2.8902566533766e-07,
"loss": 0.9991,
"step": 959
},
{
"epoch": 0.8955223880597015,
"grad_norm": 1.912346123465066,
"learning_rate": 2.839815728064077e-07,
"loss": 0.8962,
"step": 960
},
{
"epoch": 0.8964552238805971,
"grad_norm": 2.1676291843097917,
"learning_rate": 2.7898059680624754e-07,
"loss": 0.8638,
"step": 961
},
{
"epoch": 0.8973880597014925,
"grad_norm": 1.772732380775385,
"learning_rate": 2.7402278305895757e-07,
"loss": 0.8279,
"step": 962
},
{
"epoch": 0.898320895522388,
"grad_norm": 1.8917275549784411,
"learning_rate": 2.691081768916992e-07,
"loss": 0.8788,
"step": 963
},
{
"epoch": 0.8992537313432836,
"grad_norm": 2.2247358406402173,
"learning_rate": 2.642368232366055e-07,
"loss": 1.0051,
"step": 964
},
{
"epoch": 0.9001865671641791,
"grad_norm": 2.093537102689636,
"learning_rate": 2.594087666303713e-07,
"loss": 1.0053,
"step": 965
},
{
"epoch": 0.9011194029850746,
"grad_norm": 1.9519723376240363,
"learning_rate": 2.546240512138443e-07,
"loss": 0.9138,
"step": 966
},
{
"epoch": 0.9020522388059702,
"grad_norm": 1.933758795475823,
"learning_rate": 2.498827207316218e-07,
"loss": 0.8041,
"step": 967
},
{
"epoch": 0.9029850746268657,
"grad_norm": 1.9099066129540319,
"learning_rate": 2.4518481853165355e-07,
"loss": 0.836,
"step": 968
},
{
"epoch": 0.9039179104477612,
"grad_norm": 2.092341393617281,
"learning_rate": 2.4053038756484183e-07,
"loss": 0.9311,
"step": 969
},
{
"epoch": 0.9048507462686567,
"grad_norm": 1.9685703294789092,
"learning_rate": 2.35919470384649e-07,
"loss": 0.8652,
"step": 970
},
{
"epoch": 0.9057835820895522,
"grad_norm": 1.9532043078444,
"learning_rate": 2.3135210914671025e-07,
"loss": 0.7688,
"step": 971
},
{
"epoch": 0.9067164179104478,
"grad_norm": 2.087686981807759,
"learning_rate": 2.2682834560844913e-07,
"loss": 0.8688,
"step": 972
},
{
"epoch": 0.9076492537313433,
"grad_norm": 2.0014552742842078,
"learning_rate": 2.2234822112869258e-07,
"loss": 1.0069,
"step": 973
},
{
"epoch": 0.9085820895522388,
"grad_norm": 1.9627894834696793,
"learning_rate": 2.1791177666729447e-07,
"loss": 0.873,
"step": 974
},
{
"epoch": 0.9095149253731343,
"grad_norm": 1.8883434502300744,
"learning_rate": 2.135190527847608e-07,
"loss": 0.8597,
"step": 975
},
{
"epoch": 0.9104477611940298,
"grad_norm": 2.0723568449568566,
"learning_rate": 2.0917008964187968e-07,
"loss": 0.9055,
"step": 976
},
{
"epoch": 0.9113805970149254,
"grad_norm": 1.8625715167467092,
"learning_rate": 2.0486492699935356e-07,
"loss": 0.8991,
"step": 977
},
{
"epoch": 0.9123134328358209,
"grad_norm": 1.8760337114692769,
"learning_rate": 2.0060360421743585e-07,
"loss": 1.0026,
"step": 978
},
{
"epoch": 0.9132462686567164,
"grad_norm": 2.2028741299249814,
"learning_rate": 1.9638616025557e-07,
"loss": 0.9046,
"step": 979
},
{
"epoch": 0.914179104477612,
"grad_norm": 1.8669104047533285,
"learning_rate": 1.922126336720348e-07,
"loss": 0.9072,
"step": 980
},
{
"epoch": 0.9151119402985075,
"grad_norm": 2.3404781141447377,
"learning_rate": 1.8808306262359245e-07,
"loss": 0.8037,
"step": 981
},
{
"epoch": 0.9160447761194029,
"grad_norm": 1.9108882089574675,
"learning_rate": 1.8399748486513657e-07,
"loss": 0.8884,
"step": 982
},
{
"epoch": 0.9169776119402985,
"grad_norm": 2.2460321766985327,
"learning_rate": 1.799559377493504e-07,
"loss": 0.8783,
"step": 983
},
{
"epoch": 0.917910447761194,
"grad_norm": 1.7418493175968153,
"learning_rate": 1.7595845822636292e-07,
"loss": 0.8471,
"step": 984
},
{
"epoch": 0.9188432835820896,
"grad_norm": 1.811710094313894,
"learning_rate": 1.7200508284341277e-07,
"loss": 0.9025,
"step": 985
},
{
"epoch": 0.9197761194029851,
"grad_norm": 2.081295731898405,
"learning_rate": 1.6809584774451382e-07,
"loss": 0.8154,
"step": 986
},
{
"epoch": 0.9207089552238806,
"grad_norm": 1.8779412512299882,
"learning_rate": 1.6423078867012388e-07,
"loss": 0.8678,
"step": 987
},
{
"epoch": 0.9216417910447762,
"grad_norm": 2.1230406942943962,
"learning_rate": 1.6040994095681716e-07,
"loss": 0.8317,
"step": 988
},
{
"epoch": 0.9225746268656716,
"grad_norm": 2.222130965564891,
"learning_rate": 1.5663333953696291e-07,
"loss": 0.8583,
"step": 989
},
{
"epoch": 0.9235074626865671,
"grad_norm": 2.1998660901992375,
"learning_rate": 1.529010189384078e-07,
"loss": 0.869,
"step": 990
},
{
"epoch": 0.9244402985074627,
"grad_norm": 1.908067530032784,
"learning_rate": 1.4921301328415573e-07,
"loss": 0.9142,
"step": 991
},
{
"epoch": 0.9253731343283582,
"grad_norm": 2.180856825863901,
"learning_rate": 1.4556935629205738e-07,
"loss": 0.9341,
"step": 992
},
{
"epoch": 0.9263059701492538,
"grad_norm": 1.8022278093380901,
"learning_rate": 1.419700812745045e-07,
"loss": 0.89,
"step": 993
},
{
"epoch": 0.9272388059701493,
"grad_norm": 2.1781591687837527,
"learning_rate": 1.3841522113812235e-07,
"loss": 0.9304,
"step": 994
},
{
"epoch": 0.9281716417910447,
"grad_norm": 1.9585332024755246,
"learning_rate": 1.349048083834703e-07,
"loss": 0.8885,
"step": 995
},
{
"epoch": 0.9291044776119403,
"grad_norm": 2.0590825929325978,
"learning_rate": 1.3143887510474628e-07,
"loss": 0.8007,
"step": 996
},
{
"epoch": 0.9300373134328358,
"grad_norm": 1.7807811290722644,
"learning_rate": 1.2801745298948787e-07,
"loss": 0.8718,
"step": 997
},
{
"epoch": 0.9309701492537313,
"grad_norm": 2.2114270797857776,
"learning_rate": 1.2464057331828928e-07,
"loss": 0.9755,
"step": 998
},
{
"epoch": 0.9319029850746269,
"grad_norm": 1.9875850788448337,
"learning_rate": 1.2130826696450993e-07,
"loss": 0.9572,
"step": 999
},
{
"epoch": 0.9328358208955224,
"grad_norm": 2.1663068221160198,
"learning_rate": 1.1802056439399634e-07,
"loss": 0.9524,
"step": 1000
},
{
"epoch": 0.933768656716418,
"grad_norm": 2.023421006307843,
"learning_rate": 1.1477749566480123e-07,
"loss": 0.9418,
"step": 1001
},
{
"epoch": 0.9347014925373134,
"grad_norm": 1.9263847019446345,
"learning_rate": 1.1157909042690928e-07,
"loss": 0.9746,
"step": 1002
},
{
"epoch": 0.9356343283582089,
"grad_norm": 1.9462492504234383,
"learning_rate": 1.0842537792196517e-07,
"loss": 0.8791,
"step": 1003
},
{
"epoch": 0.9365671641791045,
"grad_norm": 1.8712941198092845,
"learning_rate": 1.0531638698300872e-07,
"loss": 0.8482,
"step": 1004
},
{
"epoch": 0.9375,
"grad_norm": 2.0856680857593504,
"learning_rate": 1.0225214603420852e-07,
"loss": 0.8921,
"step": 1005
},
{
"epoch": 0.9384328358208955,
"grad_norm": 2.798704963400358,
"learning_rate": 9.923268309060374e-08,
"loss": 0.8363,
"step": 1006
},
{
"epoch": 0.9393656716417911,
"grad_norm": 1.9664179110532811,
"learning_rate": 9.625802575784771e-08,
"loss": 0.8266,
"step": 1007
},
{
"epoch": 0.9402985074626866,
"grad_norm": 2.020167079286138,
"learning_rate": 9.332820123195419e-08,
"loss": 1.055,
"step": 1008
},
{
"epoch": 0.941231343283582,
"grad_norm": 1.7637095033779218,
"learning_rate": 9.044323629904983e-08,
"loss": 0.8537,
"step": 1009
},
{
"epoch": 0.9421641791044776,
"grad_norm": 2.1059395595551007,
"learning_rate": 8.760315733513214e-08,
"loss": 0.9053,
"step": 1010
},
{
"epoch": 0.9430970149253731,
"grad_norm": 1.9942677985623445,
"learning_rate": 8.480799030582187e-08,
"loss": 0.9253,
"step": 1011
},
{
"epoch": 0.9440298507462687,
"grad_norm": 1.8568937850669813,
"learning_rate": 8.205776076613048e-08,
"loss": 0.8701,
"step": 1012
},
{
"epoch": 0.9449626865671642,
"grad_norm": 2.634956637886057,
"learning_rate": 7.935249386022637e-08,
"loss": 0.9487,
"step": 1013
},
{
"epoch": 0.9458955223880597,
"grad_norm": 1.5449737246600166,
"learning_rate": 7.669221432120288e-08,
"loss": 0.7607,
"step": 1014
},
{
"epoch": 0.9468283582089553,
"grad_norm": 1.9094520890156184,
"learning_rate": 7.40769464708535e-08,
"loss": 0.8649,
"step": 1015
},
{
"epoch": 0.9477611940298507,
"grad_norm": 1.9023157399520554,
"learning_rate": 7.15067142194481e-08,
"loss": 0.8369,
"step": 1016
},
{
"epoch": 0.9486940298507462,
"grad_norm": 1.9028901934022668,
"learning_rate": 6.898154106551702e-08,
"loss": 0.9476,
"step": 1017
},
{
"epoch": 0.9496268656716418,
"grad_norm": 1.8701554024689881,
"learning_rate": 6.650145009563347e-08,
"loss": 0.8783,
"step": 1018
},
{
"epoch": 0.9505597014925373,
"grad_norm": 1.7767602763747259,
"learning_rate": 6.40664639842048e-08,
"loss": 0.9441,
"step": 1019
},
{
"epoch": 0.9514925373134329,
"grad_norm": 1.8838693803099766,
"learning_rate": 6.167660499326322e-08,
"loss": 0.937,
"step": 1020
},
{
"epoch": 0.9524253731343284,
"grad_norm": 2.692563901612505,
"learning_rate": 5.933189497226211e-08,
"loss": 0.9834,
"step": 1021
},
{
"epoch": 0.9533582089552238,
"grad_norm": 1.7114062950851714,
"learning_rate": 5.703235535787888e-08,
"loss": 0.9326,
"step": 1022
},
{
"epoch": 0.9542910447761194,
"grad_norm": 1.9452192334747598,
"learning_rate": 5.477800717381687e-08,
"loss": 0.8999,
"step": 1023
},
{
"epoch": 0.9552238805970149,
"grad_norm": 2.1234599148224893,
"learning_rate": 5.2568871030613254e-08,
"loss": 0.9231,
"step": 1024
},
{
"epoch": 0.9561567164179104,
"grad_norm": 1.9592999631827852,
"learning_rate": 5.040496712545084e-08,
"loss": 0.8179,
"step": 1025
},
{
"epoch": 0.957089552238806,
"grad_norm": 1.8289245300871522,
"learning_rate": 4.828631524197325e-08,
"loss": 0.7899,
"step": 1026
},
{
"epoch": 0.9580223880597015,
"grad_norm": 1.9613592138918752,
"learning_rate": 4.621293475010613e-08,
"loss": 0.892,
"step": 1027
},
{
"epoch": 0.9589552238805971,
"grad_norm": 2.604672889185253,
"learning_rate": 4.4184844605875135e-08,
"loss": 0.8499,
"step": 1028
},
{
"epoch": 0.9598880597014925,
"grad_norm": 2.454723162477931,
"learning_rate": 4.2202063351238775e-08,
"loss": 0.8605,
"step": 1029
},
{
"epoch": 0.960820895522388,
"grad_norm": 2.1695154260419485,
"learning_rate": 4.026460911391361e-08,
"loss": 0.7738,
"step": 1030
},
{
"epoch": 0.9617537313432836,
"grad_norm": 1.964964950666423,
"learning_rate": 3.837249960721268e-08,
"loss": 0.8122,
"step": 1031
},
{
"epoch": 0.9626865671641791,
"grad_norm": 1.9820734458032918,
"learning_rate": 3.652575212987952e-08,
"loss": 0.9472,
"step": 1032
},
{
"epoch": 0.9636194029850746,
"grad_norm": 2.219528339090599,
"learning_rate": 3.472438356593388e-08,
"loss": 0.9492,
"step": 1033
},
{
"epoch": 0.9645522388059702,
"grad_norm": 1.7771138657874757,
"learning_rate": 3.2968410384514596e-08,
"loss": 0.7797,
"step": 1034
},
{
"epoch": 0.9654850746268657,
"grad_norm": 2.084512264922557,
"learning_rate": 3.1257848639730804e-08,
"loss": 0.9858,
"step": 1035
},
{
"epoch": 0.9664179104477612,
"grad_norm": 2.4967335118149228,
"learning_rate": 2.9592713970513222e-08,
"loss": 0.9236,
"step": 1036
},
{
"epoch": 0.9673507462686567,
"grad_norm": 1.8854307922159739,
"learning_rate": 2.7973021600473106e-08,
"loss": 0.8051,
"step": 1037
},
{
"epoch": 0.9682835820895522,
"grad_norm": 2.0016257711944516,
"learning_rate": 2.639878633776294e-08,
"loss": 0.8581,
"step": 1038
},
{
"epoch": 0.9692164179104478,
"grad_norm": 1.875451506704151,
"learning_rate": 2.4870022574938758e-08,
"loss": 0.878,
"step": 1039
},
{
"epoch": 0.9701492537313433,
"grad_norm": 1.999066584814321,
"learning_rate": 2.338674428883192e-08,
"loss": 0.8706,
"step": 1040
},
{
"epoch": 0.9710820895522388,
"grad_norm": 1.9542132100319682,
"learning_rate": 2.1948965040417546e-08,
"loss": 0.7331,
"step": 1041
},
{
"epoch": 0.9720149253731343,
"grad_norm": 1.8855472930999928,
"learning_rate": 2.0556697974695172e-08,
"loss": 0.8969,
"step": 1042
},
{
"epoch": 0.9729477611940298,
"grad_norm": 2.508547890467829,
"learning_rate": 1.9209955820563287e-08,
"loss": 0.8011,
"step": 1043
},
{
"epoch": 0.9738805970149254,
"grad_norm": 2.216111571172185,
"learning_rate": 1.790875089070887e-08,
"loss": 0.9244,
"step": 1044
},
{
"epoch": 0.9748134328358209,
"grad_norm": 2.065192032762422,
"learning_rate": 1.6653095081488048e-08,
"loss": 0.8858,
"step": 1045
},
{
"epoch": 0.9757462686567164,
"grad_norm": 1.836026549730265,
"learning_rate": 1.544299987282283e-08,
"loss": 0.8686,
"step": 1046
},
{
"epoch": 0.976679104477612,
"grad_norm": 1.9705872664319461,
"learning_rate": 1.4278476328095092e-08,
"loss": 0.8346,
"step": 1047
},
{
"epoch": 0.9776119402985075,
"grad_norm": 2.0804065505365985,
"learning_rate": 1.3159535094042775e-08,
"loss": 0.9699,
"step": 1048
},
{
"epoch": 0.9785447761194029,
"grad_norm": 1.8974138173958837,
"learning_rate": 1.208618640066439e-08,
"loss": 0.7586,
"step": 1049
},
{
"epoch": 0.9794776119402985,
"grad_norm": 1.7704013826483223,
"learning_rate": 1.105844006112744e-08,
"loss": 0.8322,
"step": 1050
},
{
"epoch": 0.980410447761194,
"grad_norm": 1.9546666204025989,
"learning_rate": 1.0076305471674042e-08,
"loss": 0.8277,
"step": 1051
},
{
"epoch": 0.9813432835820896,
"grad_norm": 2.17984099967704,
"learning_rate": 9.139791611540438e-09,
"loss": 0.8333,
"step": 1052
},
{
"epoch": 0.9822761194029851,
"grad_norm": 2.1232784074059565,
"learning_rate": 8.248907042868736e-09,
"loss": 0.9736,
"step": 1053
},
{
"epoch": 0.9832089552238806,
"grad_norm": 2.221022045548849,
"learning_rate": 7.403659910635852e-09,
"loss": 0.8901,
"step": 1054
},
{
"epoch": 0.9841417910447762,
"grad_norm": 2.011276585552664,
"learning_rate": 6.604057942571907e-09,
"loss": 0.9003,
"step": 1055
},
{
"epoch": 0.9850746268656716,
"grad_norm": 2.140492194504944,
"learning_rate": 5.850108449094727e-09,
"loss": 1.0668,
"step": 1056
},
{
"epoch": 0.9860074626865671,
"grad_norm": 2.0292742913803368,
"learning_rate": 5.141818323240455e-09,
"loss": 0.925,
"step": 1057
},
{
"epoch": 0.9869402985074627,
"grad_norm": 2.081858842600173,
"learning_rate": 4.479194040602486e-09,
"loss": 0.8006,
"step": 1058
},
{
"epoch": 0.9878731343283582,
"grad_norm": 1.7511409044327173,
"learning_rate": 3.862241659269294e-09,
"loss": 0.9453,
"step": 1059
},
{
"epoch": 0.9888059701492538,
"grad_norm": 2.120753447932695,
"learning_rate": 3.2909668197717016e-09,
"loss": 0.9983,
"step": 1060
},
{
"epoch": 0.9897388059701493,
"grad_norm": 1.773956270685399,
"learning_rate": 2.765374745029581e-09,
"loss": 0.9507,
"step": 1061
},
{
"epoch": 0.9906716417910447,
"grad_norm": 1.7760344822774348,
"learning_rate": 2.2854702403068975e-09,
"loss": 0.7667,
"step": 1062
},
{
"epoch": 0.9916044776119403,
"grad_norm": 1.6223977390069488,
"learning_rate": 1.8512576931634107e-09,
"loss": 0.804,
"step": 1063
},
{
"epoch": 0.9925373134328358,
"grad_norm": 1.9104883497047482,
"learning_rate": 1.4627410734185942e-09,
"loss": 0.7733,
"step": 1064
},
{
"epoch": 0.9934701492537313,
"grad_norm": 2.0159659350161108,
"learning_rate": 1.1199239331122213e-09,
"loss": 0.8762,
"step": 1065
},
{
"epoch": 0.9944029850746269,
"grad_norm": 2.172254823139684,
"learning_rate": 8.228094064749448e-10,
"loss": 0.8189,
"step": 1066
},
{
"epoch": 0.9953358208955224,
"grad_norm": 1.7815552559433994,
"learning_rate": 5.714002098966554e-10,
"loss": 0.8087,
"step": 1067
},
{
"epoch": 0.996268656716418,
"grad_norm": 2.285101015754093,
"learning_rate": 3.656986419042774e-10,
"loss": 0.9498,
"step": 1068
},
{
"epoch": 0.9972014925373134,
"grad_norm": 2.055347146771299,
"learning_rate": 2.0570658313845414e-10,
"loss": 0.9073,
"step": 1069
},
{
"epoch": 0.9981343283582089,
"grad_norm": 2.126909984025199,
"learning_rate": 9.142549633800457e-11,
"loss": 0.9089,
"step": 1070
},
{
"epoch": 0.9990671641791045,
"grad_norm": 1.9854452602787231,
"learning_rate": 2.2856426326045566e-11,
"loss": 1.0285,
"step": 1071
},
{
"epoch": 1.0,
"grad_norm": 2.0313067725482985,
"learning_rate": 0.0,
"loss": 0.8507,
"step": 1072
},
{
"epoch": 1.0,
"step": 1072,
"total_flos": 1130020696555520.0,
"train_loss": 0.947343766911706,
"train_runtime": 38010.843,
"train_samples_per_second": 0.903,
"train_steps_per_second": 0.028
}
],
"logging_steps": 1.0,
"max_steps": 1072,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2000,
"total_flos": 1130020696555520.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}