Download trainer_state.json from dejanai/QDF-large: direct link, hf CLI and curl.
- Browser
- Download file 301 kB
-
https://huggingface.co/dejanai/QDF-large/resolve/main/trainer_state.json
- Command line
-
hf download hf://dejanai/QDF-large/trainer_state.json
-
curl -L -o trainer_state.json https://huggingface.co/dejanai/QDF-large/resolve/main/trainer_state.json
301 kB
{
"best_global_step": 334,
"best_metric": 0.11946738511323929,
"best_model_checkpoint": "./albert-large-v2-checkpoints\\checkpoint-334",
"epoch": 4.985767790262172,
"eval_steps": 500,
"global_step": 1665,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00299625468164794,
"grad_norm": 14.905317306518555,
"learning_rate": 9.993993993993994e-06,
"loss": 0.7434,
"step": 1
},
{
"epoch": 0.00599250936329588,
"grad_norm": 14.136712074279785,
"learning_rate": 9.987987987987989e-06,
"loss": 0.6318,
"step": 2
},
{
"epoch": 0.008988764044943821,
"grad_norm": 14.850204467773438,
"learning_rate": 9.981981981981982e-06,
"loss": 0.6954,
"step": 3
},
{
"epoch": 0.01198501872659176,
"grad_norm": 14.774368286132812,
"learning_rate": 9.975975975975977e-06,
"loss": 0.6352,
"step": 4
},
{
"epoch": 0.0149812734082397,
"grad_norm": 14.21596908569336,
"learning_rate": 9.96996996996997e-06,
"loss": 0.5841,
"step": 5
},
{
"epoch": 0.017977528089887642,
"grad_norm": 11.977082252502441,
"learning_rate": 9.963963963963965e-06,
"loss": 0.5649,
"step": 6
},
{
"epoch": 0.020973782771535582,
"grad_norm": 15.396467208862305,
"learning_rate": 9.957957957957959e-06,
"loss": 0.6076,
"step": 7
},
{
"epoch": 0.02397003745318352,
"grad_norm": 14.53940486907959,
"learning_rate": 9.951951951951952e-06,
"loss": 0.5034,
"step": 8
},
{
"epoch": 0.02696629213483146,
"grad_norm": 11.599512100219727,
"learning_rate": 9.945945945945947e-06,
"loss": 0.4279,
"step": 9
},
{
"epoch": 0.0299625468164794,
"grad_norm": 32.86847686767578,
"learning_rate": 9.93993993993994e-06,
"loss": 0.467,
"step": 10
},
{
"epoch": 0.03295880149812734,
"grad_norm": 28.4532527923584,
"learning_rate": 9.933933933933935e-06,
"loss": 0.524,
"step": 11
},
{
"epoch": 0.035955056179775284,
"grad_norm": 44.46889877319336,
"learning_rate": 9.927927927927928e-06,
"loss": 0.5606,
"step": 12
},
{
"epoch": 0.03895131086142322,
"grad_norm": 21.121667861938477,
"learning_rate": 9.921921921921923e-06,
"loss": 0.5904,
"step": 13
},
{
"epoch": 0.041947565543071164,
"grad_norm": 58.1203727722168,
"learning_rate": 9.915915915915917e-06,
"loss": 0.4462,
"step": 14
},
{
"epoch": 0.0449438202247191,
"grad_norm": 196.31521606445312,
"learning_rate": 9.90990990990991e-06,
"loss": 0.5616,
"step": 15
},
{
"epoch": 0.04794007490636704,
"grad_norm": 71.38382720947266,
"learning_rate": 9.903903903903905e-06,
"loss": 0.744,
"step": 16
},
{
"epoch": 0.05093632958801498,
"grad_norm": 43.24066925048828,
"learning_rate": 9.8978978978979e-06,
"loss": 0.7099,
"step": 17
},
{
"epoch": 0.05393258426966292,
"grad_norm": 51.185211181640625,
"learning_rate": 9.891891891891893e-06,
"loss": 0.6348,
"step": 18
},
{
"epoch": 0.05692883895131086,
"grad_norm": 25.794620513916016,
"learning_rate": 9.885885885885886e-06,
"loss": 0.6368,
"step": 19
},
{
"epoch": 0.0599250936329588,
"grad_norm": 25.217397689819336,
"learning_rate": 9.879879879879881e-06,
"loss": 0.7707,
"step": 20
},
{
"epoch": 0.06292134831460675,
"grad_norm": 34.317901611328125,
"learning_rate": 9.873873873873875e-06,
"loss": 0.403,
"step": 21
},
{
"epoch": 0.06591760299625468,
"grad_norm": 33.57453918457031,
"learning_rate": 9.867867867867868e-06,
"loss": 0.5658,
"step": 22
},
{
"epoch": 0.06891385767790262,
"grad_norm": 14.95702838897705,
"learning_rate": 9.861861861861863e-06,
"loss": 0.5581,
"step": 23
},
{
"epoch": 0.07191011235955057,
"grad_norm": 15.038543701171875,
"learning_rate": 9.855855855855858e-06,
"loss": 0.459,
"step": 24
},
{
"epoch": 0.0749063670411985,
"grad_norm": 12.593106269836426,
"learning_rate": 9.849849849849851e-06,
"loss": 0.4268,
"step": 25
},
{
"epoch": 0.07790262172284644,
"grad_norm": 13.807997703552246,
"learning_rate": 9.843843843843844e-06,
"loss": 0.4778,
"step": 26
},
{
"epoch": 0.08089887640449438,
"grad_norm": 11.103631973266602,
"learning_rate": 9.83783783783784e-06,
"loss": 0.3946,
"step": 27
},
{
"epoch": 0.08389513108614233,
"grad_norm": 23.803539276123047,
"learning_rate": 9.831831831831833e-06,
"loss": 0.5929,
"step": 28
},
{
"epoch": 0.08689138576779026,
"grad_norm": 23.249616622924805,
"learning_rate": 9.825825825825826e-06,
"loss": 0.4351,
"step": 29
},
{
"epoch": 0.0898876404494382,
"grad_norm": 33.72541427612305,
"learning_rate": 9.81981981981982e-06,
"loss": 0.4335,
"step": 30
},
{
"epoch": 0.09288389513108614,
"grad_norm": 13.924614906311035,
"learning_rate": 9.813813813813816e-06,
"loss": 0.3411,
"step": 31
},
{
"epoch": 0.09588014981273409,
"grad_norm": 14.99662971496582,
"learning_rate": 9.807807807807809e-06,
"loss": 0.3439,
"step": 32
},
{
"epoch": 0.09887640449438202,
"grad_norm": 27.40151023864746,
"learning_rate": 9.801801801801802e-06,
"loss": 0.4916,
"step": 33
},
{
"epoch": 0.10187265917602996,
"grad_norm": 18.182722091674805,
"learning_rate": 9.795795795795795e-06,
"loss": 0.4784,
"step": 34
},
{
"epoch": 0.10486891385767791,
"grad_norm": 12.109911918640137,
"learning_rate": 9.78978978978979e-06,
"loss": 0.3813,
"step": 35
},
{
"epoch": 0.10786516853932585,
"grad_norm": 27.76490592956543,
"learning_rate": 9.783783783783785e-06,
"loss": 0.4398,
"step": 36
},
{
"epoch": 0.11086142322097378,
"grad_norm": 14.584527015686035,
"learning_rate": 9.777777777777779e-06,
"loss": 0.3289,
"step": 37
},
{
"epoch": 0.11385767790262172,
"grad_norm": 36.990325927734375,
"learning_rate": 9.771771771771774e-06,
"loss": 0.3665,
"step": 38
},
{
"epoch": 0.11685393258426967,
"grad_norm": 20.206125259399414,
"learning_rate": 9.765765765765767e-06,
"loss": 0.4474,
"step": 39
},
{
"epoch": 0.1198501872659176,
"grad_norm": 10.573413848876953,
"learning_rate": 9.75975975975976e-06,
"loss": 0.3983,
"step": 40
},
{
"epoch": 0.12284644194756554,
"grad_norm": 14.455694198608398,
"learning_rate": 9.753753753753753e-06,
"loss": 0.2375,
"step": 41
},
{
"epoch": 0.1258426966292135,
"grad_norm": 20.108503341674805,
"learning_rate": 9.747747747747748e-06,
"loss": 0.3686,
"step": 42
},
{
"epoch": 0.12883895131086143,
"grad_norm": 23.18209457397461,
"learning_rate": 9.741741741741743e-06,
"loss": 0.3195,
"step": 43
},
{
"epoch": 0.13183520599250936,
"grad_norm": 12.491003036499023,
"learning_rate": 9.735735735735737e-06,
"loss": 0.3443,
"step": 44
},
{
"epoch": 0.1348314606741573,
"grad_norm": 14.124627113342285,
"learning_rate": 9.729729729729732e-06,
"loss": 0.3179,
"step": 45
},
{
"epoch": 0.13782771535580524,
"grad_norm": 20.555522918701172,
"learning_rate": 9.723723723723725e-06,
"loss": 0.2215,
"step": 46
},
{
"epoch": 0.14082397003745317,
"grad_norm": 11.338384628295898,
"learning_rate": 9.717717717717718e-06,
"loss": 0.267,
"step": 47
},
{
"epoch": 0.14382022471910114,
"grad_norm": 13.290757179260254,
"learning_rate": 9.711711711711711e-06,
"loss": 0.1393,
"step": 48
},
{
"epoch": 0.14681647940074907,
"grad_norm": 12.90415096282959,
"learning_rate": 9.705705705705706e-06,
"loss": 0.2614,
"step": 49
},
{
"epoch": 0.149812734082397,
"grad_norm": 20.324295043945312,
"learning_rate": 9.699699699699701e-06,
"loss": 0.3713,
"step": 50
},
{
"epoch": 0.15280898876404495,
"grad_norm": 51.13927459716797,
"learning_rate": 9.693693693693694e-06,
"loss": 0.4306,
"step": 51
},
{
"epoch": 0.15580524344569288,
"grad_norm": 6.610884189605713,
"learning_rate": 9.687687687687688e-06,
"loss": 0.2065,
"step": 52
},
{
"epoch": 0.15880149812734082,
"grad_norm": 38.602378845214844,
"learning_rate": 9.681681681681683e-06,
"loss": 0.3805,
"step": 53
},
{
"epoch": 0.16179775280898875,
"grad_norm": 15.254252433776855,
"learning_rate": 9.675675675675676e-06,
"loss": 0.2872,
"step": 54
},
{
"epoch": 0.1647940074906367,
"grad_norm": 25.657941818237305,
"learning_rate": 9.669669669669671e-06,
"loss": 0.1818,
"step": 55
},
{
"epoch": 0.16779026217228465,
"grad_norm": 115.34745788574219,
"learning_rate": 9.663663663663664e-06,
"loss": 0.2394,
"step": 56
},
{
"epoch": 0.1707865168539326,
"grad_norm": 17.47480583190918,
"learning_rate": 9.65765765765766e-06,
"loss": 0.2356,
"step": 57
},
{
"epoch": 0.17378277153558053,
"grad_norm": 25.645423889160156,
"learning_rate": 9.651651651651652e-06,
"loss": 0.2226,
"step": 58
},
{
"epoch": 0.17677902621722846,
"grad_norm": 11.551637649536133,
"learning_rate": 9.645645645645646e-06,
"loss": 0.2643,
"step": 59
},
{
"epoch": 0.1797752808988764,
"grad_norm": 9.87855052947998,
"learning_rate": 9.63963963963964e-06,
"loss": 0.1593,
"step": 60
},
{
"epoch": 0.18277153558052434,
"grad_norm": 28.706981658935547,
"learning_rate": 9.633633633633634e-06,
"loss": 0.1934,
"step": 61
},
{
"epoch": 0.18576779026217227,
"grad_norm": 18.82908058166504,
"learning_rate": 9.627627627627629e-06,
"loss": 0.1193,
"step": 62
},
{
"epoch": 0.18876404494382024,
"grad_norm": 25.83577537536621,
"learning_rate": 9.621621621621622e-06,
"loss": 0.3402,
"step": 63
},
{
"epoch": 0.19176029962546817,
"grad_norm": 23.55269432067871,
"learning_rate": 9.615615615615617e-06,
"loss": 0.1416,
"step": 64
},
{
"epoch": 0.1947565543071161,
"grad_norm": 19.169841766357422,
"learning_rate": 9.60960960960961e-06,
"loss": 0.2281,
"step": 65
},
{
"epoch": 0.19775280898876405,
"grad_norm": 39.77779006958008,
"learning_rate": 9.603603603603604e-06,
"loss": 0.2663,
"step": 66
},
{
"epoch": 0.20074906367041198,
"grad_norm": 25.331012725830078,
"learning_rate": 9.597597597597599e-06,
"loss": 0.2053,
"step": 67
},
{
"epoch": 0.20374531835205992,
"grad_norm": 11.722957611083984,
"learning_rate": 9.591591591591592e-06,
"loss": 0.0716,
"step": 68
},
{
"epoch": 0.20674157303370785,
"grad_norm": 9.155919075012207,
"learning_rate": 9.585585585585587e-06,
"loss": 0.0777,
"step": 69
},
{
"epoch": 0.20973782771535582,
"grad_norm": 21.291912078857422,
"learning_rate": 9.57957957957958e-06,
"loss": 0.1228,
"step": 70
},
{
"epoch": 0.21273408239700375,
"grad_norm": 13.857783317565918,
"learning_rate": 9.573573573573575e-06,
"loss": 0.1877,
"step": 71
},
{
"epoch": 0.2157303370786517,
"grad_norm": 62.58148956298828,
"learning_rate": 9.567567567567568e-06,
"loss": 0.1719,
"step": 72
},
{
"epoch": 0.21872659176029963,
"grad_norm": 29.480480194091797,
"learning_rate": 9.561561561561562e-06,
"loss": 0.2129,
"step": 73
},
{
"epoch": 0.22172284644194756,
"grad_norm": 24.820606231689453,
"learning_rate": 9.555555555555556e-06,
"loss": 0.2905,
"step": 74
},
{
"epoch": 0.2247191011235955,
"grad_norm": 11.047852516174316,
"learning_rate": 9.54954954954955e-06,
"loss": 0.3314,
"step": 75
},
{
"epoch": 0.22771535580524344,
"grad_norm": 7.338944435119629,
"learning_rate": 9.543543543543545e-06,
"loss": 0.0453,
"step": 76
},
{
"epoch": 0.23071161048689137,
"grad_norm": 32.50682067871094,
"learning_rate": 9.537537537537538e-06,
"loss": 0.1425,
"step": 77
},
{
"epoch": 0.23370786516853934,
"grad_norm": 1.6429194211959839,
"learning_rate": 9.531531531531533e-06,
"loss": 0.0247,
"step": 78
},
{
"epoch": 0.23670411985018727,
"grad_norm": 41.79351806640625,
"learning_rate": 9.525525525525526e-06,
"loss": 0.184,
"step": 79
},
{
"epoch": 0.2397003745318352,
"grad_norm": 13.814924240112305,
"learning_rate": 9.51951951951952e-06,
"loss": 0.1562,
"step": 80
},
{
"epoch": 0.24269662921348314,
"grad_norm": 56.38475799560547,
"learning_rate": 9.513513513513514e-06,
"loss": 0.3752,
"step": 81
},
{
"epoch": 0.24569288389513108,
"grad_norm": 44.72063446044922,
"learning_rate": 9.507507507507508e-06,
"loss": 0.3172,
"step": 82
},
{
"epoch": 0.24868913857677902,
"grad_norm": 18.458343505859375,
"learning_rate": 9.501501501501503e-06,
"loss": 0.1315,
"step": 83
},
{
"epoch": 0.251685393258427,
"grad_norm": 17.870281219482422,
"learning_rate": 9.495495495495496e-06,
"loss": 0.2761,
"step": 84
},
{
"epoch": 0.2546816479400749,
"grad_norm": 7.055416584014893,
"learning_rate": 9.489489489489491e-06,
"loss": 0.1989,
"step": 85
},
{
"epoch": 0.25767790262172285,
"grad_norm": 12.610936164855957,
"learning_rate": 9.483483483483484e-06,
"loss": 0.1071,
"step": 86
},
{
"epoch": 0.2606741573033708,
"grad_norm": 17.32162094116211,
"learning_rate": 9.477477477477477e-06,
"loss": 0.1906,
"step": 87
},
{
"epoch": 0.2636704119850187,
"grad_norm": 42.67469787597656,
"learning_rate": 9.471471471471472e-06,
"loss": 0.1468,
"step": 88
},
{
"epoch": 0.26666666666666666,
"grad_norm": 35.54982376098633,
"learning_rate": 9.465465465465467e-06,
"loss": 0.2978,
"step": 89
},
{
"epoch": 0.2696629213483146,
"grad_norm": 15.13595199584961,
"learning_rate": 9.45945945945946e-06,
"loss": 0.2086,
"step": 90
},
{
"epoch": 0.27265917602996254,
"grad_norm": 31.031539916992188,
"learning_rate": 9.453453453453454e-06,
"loss": 0.2054,
"step": 91
},
{
"epoch": 0.27565543071161047,
"grad_norm": 31.638675689697266,
"learning_rate": 9.447447447447449e-06,
"loss": 0.2501,
"step": 92
},
{
"epoch": 0.2786516853932584,
"grad_norm": 15.932416915893555,
"learning_rate": 9.441441441441442e-06,
"loss": 0.3287,
"step": 93
},
{
"epoch": 0.28164794007490634,
"grad_norm": 8.57020378112793,
"learning_rate": 9.435435435435435e-06,
"loss": 0.2637,
"step": 94
},
{
"epoch": 0.2846441947565543,
"grad_norm": 8.692102432250977,
"learning_rate": 9.42942942942943e-06,
"loss": 0.0893,
"step": 95
},
{
"epoch": 0.2876404494382023,
"grad_norm": 5.30082368850708,
"learning_rate": 9.423423423423425e-06,
"loss": 0.0564,
"step": 96
},
{
"epoch": 0.2906367041198502,
"grad_norm": 21.0859432220459,
"learning_rate": 9.417417417417418e-06,
"loss": 0.1392,
"step": 97
},
{
"epoch": 0.29363295880149815,
"grad_norm": 6.980339527130127,
"learning_rate": 9.411411411411412e-06,
"loss": 0.2184,
"step": 98
},
{
"epoch": 0.2966292134831461,
"grad_norm": 7.253073692321777,
"learning_rate": 9.405405405405407e-06,
"loss": 0.1224,
"step": 99
},
{
"epoch": 0.299625468164794,
"grad_norm": 3.25008487701416,
"learning_rate": 9.3993993993994e-06,
"loss": 0.0364,
"step": 100
},
{
"epoch": 0.30262172284644195,
"grad_norm": 13.387791633605957,
"learning_rate": 9.393393393393393e-06,
"loss": 0.1356,
"step": 101
},
{
"epoch": 0.3056179775280899,
"grad_norm": 2.4467055797576904,
"learning_rate": 9.387387387387388e-06,
"loss": 0.0245,
"step": 102
},
{
"epoch": 0.3086142322097378,
"grad_norm": 11.636857986450195,
"learning_rate": 9.381381381381383e-06,
"loss": 0.2406,
"step": 103
},
{
"epoch": 0.31161048689138576,
"grad_norm": 26.644214630126953,
"learning_rate": 9.375375375375376e-06,
"loss": 0.1765,
"step": 104
},
{
"epoch": 0.3146067415730337,
"grad_norm": 7.455554008483887,
"learning_rate": 9.36936936936937e-06,
"loss": 0.0346,
"step": 105
},
{
"epoch": 0.31760299625468164,
"grad_norm": 7.504270076751709,
"learning_rate": 9.363363363363363e-06,
"loss": 0.1419,
"step": 106
},
{
"epoch": 0.32059925093632957,
"grad_norm": 84.28512573242188,
"learning_rate": 9.357357357357358e-06,
"loss": 0.1742,
"step": 107
},
{
"epoch": 0.3235955056179775,
"grad_norm": 31.72047996520996,
"learning_rate": 9.351351351351353e-06,
"loss": 0.3746,
"step": 108
},
{
"epoch": 0.32659176029962544,
"grad_norm": 18.567527770996094,
"learning_rate": 9.345345345345346e-06,
"loss": 0.1845,
"step": 109
},
{
"epoch": 0.3295880149812734,
"grad_norm": 45.910888671875,
"learning_rate": 9.339339339339341e-06,
"loss": 0.496,
"step": 110
},
{
"epoch": 0.3325842696629214,
"grad_norm": 31.651735305786133,
"learning_rate": 9.333333333333334e-06,
"loss": 0.3987,
"step": 111
},
{
"epoch": 0.3355805243445693,
"grad_norm": 13.848103523254395,
"learning_rate": 9.327327327327328e-06,
"loss": 0.0471,
"step": 112
},
{
"epoch": 0.33857677902621724,
"grad_norm": 22.99566078186035,
"learning_rate": 9.321321321321321e-06,
"loss": 0.0891,
"step": 113
},
{
"epoch": 0.3415730337078652,
"grad_norm": 47.94853973388672,
"learning_rate": 9.315315315315316e-06,
"loss": 0.2814,
"step": 114
},
{
"epoch": 0.3445692883895131,
"grad_norm": 17.474658966064453,
"learning_rate": 9.30930930930931e-06,
"loss": 0.1467,
"step": 115
},
{
"epoch": 0.34756554307116105,
"grad_norm": 20.27558135986328,
"learning_rate": 9.303303303303304e-06,
"loss": 0.0898,
"step": 116
},
{
"epoch": 0.350561797752809,
"grad_norm": 7.167867660522461,
"learning_rate": 9.297297297297299e-06,
"loss": 0.1388,
"step": 117
},
{
"epoch": 0.3535580524344569,
"grad_norm": 21.503070831298828,
"learning_rate": 9.291291291291292e-06,
"loss": 0.1127,
"step": 118
},
{
"epoch": 0.35655430711610486,
"grad_norm": 10.975125312805176,
"learning_rate": 9.285285285285286e-06,
"loss": 0.0986,
"step": 119
},
{
"epoch": 0.3595505617977528,
"grad_norm": 11.209259986877441,
"learning_rate": 9.27927927927928e-06,
"loss": 0.3344,
"step": 120
},
{
"epoch": 0.36254681647940074,
"grad_norm": 15.64116096496582,
"learning_rate": 9.273273273273274e-06,
"loss": 0.0601,
"step": 121
},
{
"epoch": 0.36554307116104867,
"grad_norm": 30.275739669799805,
"learning_rate": 9.267267267267269e-06,
"loss": 0.2309,
"step": 122
},
{
"epoch": 0.3685393258426966,
"grad_norm": 18.100448608398438,
"learning_rate": 9.261261261261262e-06,
"loss": 0.0974,
"step": 123
},
{
"epoch": 0.37153558052434454,
"grad_norm": 15.199792861938477,
"learning_rate": 9.255255255255255e-06,
"loss": 0.1668,
"step": 124
},
{
"epoch": 0.37453183520599254,
"grad_norm": 37.463462829589844,
"learning_rate": 9.24924924924925e-06,
"loss": 0.2607,
"step": 125
},
{
"epoch": 0.3775280898876405,
"grad_norm": 3.9865729808807373,
"learning_rate": 9.243243243243243e-06,
"loss": 0.0319,
"step": 126
},
{
"epoch": 0.3805243445692884,
"grad_norm": 7.066198825836182,
"learning_rate": 9.237237237237238e-06,
"loss": 0.1569,
"step": 127
},
{
"epoch": 0.38352059925093634,
"grad_norm": 40.82876205444336,
"learning_rate": 9.231231231231232e-06,
"loss": 0.2919,
"step": 128
},
{
"epoch": 0.3865168539325843,
"grad_norm": 10.111753463745117,
"learning_rate": 9.225225225225227e-06,
"loss": 0.0578,
"step": 129
},
{
"epoch": 0.3895131086142322,
"grad_norm": 22.75676727294922,
"learning_rate": 9.21921921921922e-06,
"loss": 0.111,
"step": 130
},
{
"epoch": 0.39250936329588015,
"grad_norm": 35.19755554199219,
"learning_rate": 9.213213213213213e-06,
"loss": 0.3111,
"step": 131
},
{
"epoch": 0.3955056179775281,
"grad_norm": 42.04582977294922,
"learning_rate": 9.207207207207208e-06,
"loss": 0.6253,
"step": 132
},
{
"epoch": 0.398501872659176,
"grad_norm": 6.762131214141846,
"learning_rate": 9.201201201201201e-06,
"loss": 0.1322,
"step": 133
},
{
"epoch": 0.40149812734082396,
"grad_norm": 15.04703426361084,
"learning_rate": 9.195195195195196e-06,
"loss": 0.264,
"step": 134
},
{
"epoch": 0.4044943820224719,
"grad_norm": 7.994354724884033,
"learning_rate": 9.189189189189191e-06,
"loss": 0.1069,
"step": 135
},
{
"epoch": 0.40749063670411984,
"grad_norm": 16.184736251831055,
"learning_rate": 9.183183183183185e-06,
"loss": 0.196,
"step": 136
},
{
"epoch": 0.41048689138576777,
"grad_norm": 19.40424919128418,
"learning_rate": 9.177177177177178e-06,
"loss": 0.335,
"step": 137
},
{
"epoch": 0.4134831460674157,
"grad_norm": 27.46769142150879,
"learning_rate": 9.171171171171171e-06,
"loss": 0.1453,
"step": 138
},
{
"epoch": 0.41647940074906364,
"grad_norm": 9.040424346923828,
"learning_rate": 9.165165165165166e-06,
"loss": 0.1096,
"step": 139
},
{
"epoch": 0.41947565543071164,
"grad_norm": 16.803510665893555,
"learning_rate": 9.15915915915916e-06,
"loss": 0.2407,
"step": 140
},
{
"epoch": 0.42247191011235957,
"grad_norm": 7.690004825592041,
"learning_rate": 9.153153153153154e-06,
"loss": 0.0914,
"step": 141
},
{
"epoch": 0.4254681647940075,
"grad_norm": 20.977710723876953,
"learning_rate": 9.147147147147147e-06,
"loss": 0.2522,
"step": 142
},
{
"epoch": 0.42846441947565544,
"grad_norm": 5.203715801239014,
"learning_rate": 9.141141141141142e-06,
"loss": 0.0493,
"step": 143
},
{
"epoch": 0.4314606741573034,
"grad_norm": 4.435147762298584,
"learning_rate": 9.135135135135136e-06,
"loss": 0.038,
"step": 144
},
{
"epoch": 0.4344569288389513,
"grad_norm": 12.076634407043457,
"learning_rate": 9.129129129129129e-06,
"loss": 0.2559,
"step": 145
},
{
"epoch": 0.43745318352059925,
"grad_norm": 10.864753723144531,
"learning_rate": 9.123123123123124e-06,
"loss": 0.1158,
"step": 146
},
{
"epoch": 0.4404494382022472,
"grad_norm": 10.240983009338379,
"learning_rate": 9.117117117117117e-06,
"loss": 0.1289,
"step": 147
},
{
"epoch": 0.4434456928838951,
"grad_norm": 10.733732223510742,
"learning_rate": 9.111111111111112e-06,
"loss": 0.0875,
"step": 148
},
{
"epoch": 0.44644194756554306,
"grad_norm": 15.851396560668945,
"learning_rate": 9.105105105105105e-06,
"loss": 0.1594,
"step": 149
},
{
"epoch": 0.449438202247191,
"grad_norm": 30.241884231567383,
"learning_rate": 9.0990990990991e-06,
"loss": 0.1735,
"step": 150
},
{
"epoch": 0.45243445692883894,
"grad_norm": 3.5728492736816406,
"learning_rate": 9.093093093093094e-06,
"loss": 0.023,
"step": 151
},
{
"epoch": 0.45543071161048687,
"grad_norm": 20.475706100463867,
"learning_rate": 9.087087087087087e-06,
"loss": 0.1483,
"step": 152
},
{
"epoch": 0.4584269662921348,
"grad_norm": 7.573640823364258,
"learning_rate": 9.081081081081082e-06,
"loss": 0.1161,
"step": 153
},
{
"epoch": 0.46142322097378274,
"grad_norm": 12.72209644317627,
"learning_rate": 9.075075075075077e-06,
"loss": 0.2493,
"step": 154
},
{
"epoch": 0.46441947565543074,
"grad_norm": 61.5605354309082,
"learning_rate": 9.06906906906907e-06,
"loss": 0.3252,
"step": 155
},
{
"epoch": 0.46741573033707867,
"grad_norm": 20.318923950195312,
"learning_rate": 9.063063063063063e-06,
"loss": 0.2094,
"step": 156
},
{
"epoch": 0.4704119850187266,
"grad_norm": 7.439236164093018,
"learning_rate": 9.057057057057058e-06,
"loss": 0.2375,
"step": 157
},
{
"epoch": 0.47340823970037454,
"grad_norm": 5.565708160400391,
"learning_rate": 9.051051051051052e-06,
"loss": 0.1719,
"step": 158
},
{
"epoch": 0.4764044943820225,
"grad_norm": 15.711952209472656,
"learning_rate": 9.045045045045045e-06,
"loss": 0.3358,
"step": 159
},
{
"epoch": 0.4794007490636704,
"grad_norm": 22.145919799804688,
"learning_rate": 9.03903903903904e-06,
"loss": 0.1433,
"step": 160
},
{
"epoch": 0.48239700374531835,
"grad_norm": 13.894994735717773,
"learning_rate": 9.033033033033035e-06,
"loss": 0.3261,
"step": 161
},
{
"epoch": 0.4853932584269663,
"grad_norm": 16.804580688476562,
"learning_rate": 9.027027027027028e-06,
"loss": 0.0538,
"step": 162
},
{
"epoch": 0.4883895131086142,
"grad_norm": 12.285701751708984,
"learning_rate": 9.021021021021021e-06,
"loss": 0.0615,
"step": 163
},
{
"epoch": 0.49138576779026216,
"grad_norm": 16.515159606933594,
"learning_rate": 9.015015015015016e-06,
"loss": 0.2548,
"step": 164
},
{
"epoch": 0.4943820224719101,
"grad_norm": 8.001544952392578,
"learning_rate": 9.00900900900901e-06,
"loss": 0.0656,
"step": 165
},
{
"epoch": 0.49737827715355803,
"grad_norm": 41.51729965209961,
"learning_rate": 9.003003003003003e-06,
"loss": 0.2541,
"step": 166
},
{
"epoch": 0.500374531835206,
"grad_norm": 21.762195587158203,
"learning_rate": 8.996996996996998e-06,
"loss": 0.1823,
"step": 167
},
{
"epoch": 0.503370786516854,
"grad_norm": 9.834053993225098,
"learning_rate": 8.990990990990993e-06,
"loss": 0.1474,
"step": 168
},
{
"epoch": 0.5063670411985018,
"grad_norm": 10.880240440368652,
"learning_rate": 8.984984984984986e-06,
"loss": 0.0919,
"step": 169
},
{
"epoch": 0.5093632958801498,
"grad_norm": 3.719357490539551,
"learning_rate": 8.97897897897898e-06,
"loss": 0.0335,
"step": 170
},
{
"epoch": 0.5123595505617977,
"grad_norm": 9.114514350891113,
"learning_rate": 8.972972972972974e-06,
"loss": 0.0635,
"step": 171
},
{
"epoch": 0.5153558052434457,
"grad_norm": 32.97814178466797,
"learning_rate": 8.966966966966967e-06,
"loss": 0.2202,
"step": 172
},
{
"epoch": 0.5183520599250936,
"grad_norm": 31.934534072875977,
"learning_rate": 8.960960960960962e-06,
"loss": 0.1925,
"step": 173
},
{
"epoch": 0.5213483146067416,
"grad_norm": 25.543624877929688,
"learning_rate": 8.954954954954956e-06,
"loss": 0.1673,
"step": 174
},
{
"epoch": 0.5243445692883895,
"grad_norm": 21.611799240112305,
"learning_rate": 8.94894894894895e-06,
"loss": 0.1972,
"step": 175
},
{
"epoch": 0.5273408239700375,
"grad_norm": 40.040523529052734,
"learning_rate": 8.942942942942944e-06,
"loss": 0.0953,
"step": 176
},
{
"epoch": 0.5303370786516854,
"grad_norm": 11.244500160217285,
"learning_rate": 8.936936936936937e-06,
"loss": 0.2063,
"step": 177
},
{
"epoch": 0.5333333333333333,
"grad_norm": 2.2858376502990723,
"learning_rate": 8.93093093093093e-06,
"loss": 0.0222,
"step": 178
},
{
"epoch": 0.5363295880149813,
"grad_norm": 19.4034366607666,
"learning_rate": 8.924924924924925e-06,
"loss": 0.2228,
"step": 179
},
{
"epoch": 0.5393258426966292,
"grad_norm": 10.577714920043945,
"learning_rate": 8.91891891891892e-06,
"loss": 0.393,
"step": 180
},
{
"epoch": 0.5423220973782772,
"grad_norm": 41.52705764770508,
"learning_rate": 8.912912912912914e-06,
"loss": 0.0604,
"step": 181
},
{
"epoch": 0.5453183520599251,
"grad_norm": 14.04199504852295,
"learning_rate": 8.906906906906909e-06,
"loss": 0.2713,
"step": 182
},
{
"epoch": 0.5483146067415731,
"grad_norm": 13.048672676086426,
"learning_rate": 8.900900900900902e-06,
"loss": 0.1217,
"step": 183
},
{
"epoch": 0.5513108614232209,
"grad_norm": 24.741180419921875,
"learning_rate": 8.894894894894895e-06,
"loss": 0.1374,
"step": 184
},
{
"epoch": 0.5543071161048689,
"grad_norm": 88.56827545166016,
"learning_rate": 8.888888888888888e-06,
"loss": 0.2562,
"step": 185
},
{
"epoch": 0.5573033707865168,
"grad_norm": 19.29568099975586,
"learning_rate": 8.882882882882883e-06,
"loss": 0.2243,
"step": 186
},
{
"epoch": 0.5602996254681648,
"grad_norm": 25.720760345458984,
"learning_rate": 8.876876876876878e-06,
"loss": 0.2908,
"step": 187
},
{
"epoch": 0.5632958801498127,
"grad_norm": 3.2698912620544434,
"learning_rate": 8.870870870870871e-06,
"loss": 0.0457,
"step": 188
},
{
"epoch": 0.5662921348314607,
"grad_norm": 8.583268165588379,
"learning_rate": 8.864864864864866e-06,
"loss": 0.2066,
"step": 189
},
{
"epoch": 0.5692883895131086,
"grad_norm": 17.988779067993164,
"learning_rate": 8.85885885885886e-06,
"loss": 0.1369,
"step": 190
},
{
"epoch": 0.5722846441947566,
"grad_norm": 12.211359024047852,
"learning_rate": 8.852852852852853e-06,
"loss": 0.1654,
"step": 191
},
{
"epoch": 0.5752808988764045,
"grad_norm": 2.5672764778137207,
"learning_rate": 8.846846846846848e-06,
"loss": 0.0272,
"step": 192
},
{
"epoch": 0.5782771535580524,
"grad_norm": 10.646368026733398,
"learning_rate": 8.840840840840841e-06,
"loss": 0.2832,
"step": 193
},
{
"epoch": 0.5812734082397004,
"grad_norm": 26.280115127563477,
"learning_rate": 8.834834834834836e-06,
"loss": 0.2149,
"step": 194
},
{
"epoch": 0.5842696629213483,
"grad_norm": 3.840461254119873,
"learning_rate": 8.82882882882883e-06,
"loss": 0.0419,
"step": 195
},
{
"epoch": 0.5872659176029963,
"grad_norm": 17.350587844848633,
"learning_rate": 8.822822822822824e-06,
"loss": 0.2152,
"step": 196
},
{
"epoch": 0.5902621722846442,
"grad_norm": 12.285447120666504,
"learning_rate": 8.816816816816818e-06,
"loss": 0.2658,
"step": 197
},
{
"epoch": 0.5932584269662922,
"grad_norm": 32.076202392578125,
"learning_rate": 8.810810810810811e-06,
"loss": 0.3356,
"step": 198
},
{
"epoch": 0.59625468164794,
"grad_norm": 2.1881797313690186,
"learning_rate": 8.804804804804806e-06,
"loss": 0.0237,
"step": 199
},
{
"epoch": 0.599250936329588,
"grad_norm": 11.07189655303955,
"learning_rate": 8.798798798798799e-06,
"loss": 0.1987,
"step": 200
},
{
"epoch": 0.6022471910112359,
"grad_norm": 5.4934983253479,
"learning_rate": 8.792792792792794e-06,
"loss": 0.0278,
"step": 201
},
{
"epoch": 0.6052434456928839,
"grad_norm": 13.686779022216797,
"learning_rate": 8.786786786786787e-06,
"loss": 0.2953,
"step": 202
},
{
"epoch": 0.6082397003745318,
"grad_norm": 2.1537270545959473,
"learning_rate": 8.78078078078078e-06,
"loss": 0.0217,
"step": 203
},
{
"epoch": 0.6112359550561798,
"grad_norm": 22.37981605529785,
"learning_rate": 8.774774774774776e-06,
"loss": 0.0577,
"step": 204
},
{
"epoch": 0.6142322097378277,
"grad_norm": 9.048362731933594,
"learning_rate": 8.768768768768769e-06,
"loss": 0.1542,
"step": 205
},
{
"epoch": 0.6172284644194757,
"grad_norm": 40.02397155761719,
"learning_rate": 8.762762762762764e-06,
"loss": 0.4363,
"step": 206
},
{
"epoch": 0.6202247191011236,
"grad_norm": 92.53890991210938,
"learning_rate": 8.756756756756759e-06,
"loss": 0.1394,
"step": 207
},
{
"epoch": 0.6232209737827715,
"grad_norm": 10.263978958129883,
"learning_rate": 8.750750750750752e-06,
"loss": 0.0705,
"step": 208
},
{
"epoch": 0.6262172284644195,
"grad_norm": 12.193284034729004,
"learning_rate": 8.744744744744745e-06,
"loss": 0.1682,
"step": 209
},
{
"epoch": 0.6292134831460674,
"grad_norm": 20.703235626220703,
"learning_rate": 8.738738738738739e-06,
"loss": 0.1674,
"step": 210
},
{
"epoch": 0.6322097378277154,
"grad_norm": 12.266006469726562,
"learning_rate": 8.732732732732733e-06,
"loss": 0.0886,
"step": 211
},
{
"epoch": 0.6352059925093633,
"grad_norm": 10.525797843933105,
"learning_rate": 8.726726726726727e-06,
"loss": 0.1492,
"step": 212
},
{
"epoch": 0.6382022471910113,
"grad_norm": 34.268333435058594,
"learning_rate": 8.720720720720722e-06,
"loss": 0.1805,
"step": 213
},
{
"epoch": 0.6411985018726591,
"grad_norm": 34.239017486572266,
"learning_rate": 8.714714714714717e-06,
"loss": 0.2837,
"step": 214
},
{
"epoch": 0.6441947565543071,
"grad_norm": 28.905134201049805,
"learning_rate": 8.70870870870871e-06,
"loss": 0.0671,
"step": 215
},
{
"epoch": 0.647191011235955,
"grad_norm": 12.800658226013184,
"learning_rate": 8.702702702702703e-06,
"loss": 0.0734,
"step": 216
},
{
"epoch": 0.650187265917603,
"grad_norm": 16.8038272857666,
"learning_rate": 8.696696696696696e-06,
"loss": 0.076,
"step": 217
},
{
"epoch": 0.6531835205992509,
"grad_norm": 35.847591400146484,
"learning_rate": 8.690690690690691e-06,
"loss": 0.23,
"step": 218
},
{
"epoch": 0.6561797752808989,
"grad_norm": 29.712432861328125,
"learning_rate": 8.684684684684686e-06,
"loss": 0.4,
"step": 219
},
{
"epoch": 0.6591760299625468,
"grad_norm": 4.912635803222656,
"learning_rate": 8.67867867867868e-06,
"loss": 0.0355,
"step": 220
},
{
"epoch": 0.6621722846441948,
"grad_norm": 4.8858184814453125,
"learning_rate": 8.672672672672673e-06,
"loss": 0.0877,
"step": 221
},
{
"epoch": 0.6651685393258427,
"grad_norm": 13.752501487731934,
"learning_rate": 8.666666666666668e-06,
"loss": 0.1442,
"step": 222
},
{
"epoch": 0.6681647940074906,
"grad_norm": 24.75152587890625,
"learning_rate": 8.660660660660661e-06,
"loss": 0.102,
"step": 223
},
{
"epoch": 0.6711610486891386,
"grad_norm": 10.803027153015137,
"learning_rate": 8.654654654654654e-06,
"loss": 0.0781,
"step": 224
},
{
"epoch": 0.6741573033707865,
"grad_norm": 12.187631607055664,
"learning_rate": 8.64864864864865e-06,
"loss": 0.1672,
"step": 225
},
{
"epoch": 0.6771535580524345,
"grad_norm": 14.320629119873047,
"learning_rate": 8.642642642642644e-06,
"loss": 0.0654,
"step": 226
},
{
"epoch": 0.6801498127340824,
"grad_norm": 17.489456176757812,
"learning_rate": 8.636636636636638e-06,
"loss": 0.2263,
"step": 227
},
{
"epoch": 0.6831460674157304,
"grad_norm": 20.89166831970215,
"learning_rate": 8.63063063063063e-06,
"loss": 0.2127,
"step": 228
},
{
"epoch": 0.6861423220973782,
"grad_norm": 16.09478759765625,
"learning_rate": 8.624624624624626e-06,
"loss": 0.5156,
"step": 229
},
{
"epoch": 0.6891385767790262,
"grad_norm": 12.102310180664062,
"learning_rate": 8.618618618618619e-06,
"loss": 0.2197,
"step": 230
},
{
"epoch": 0.6921348314606741,
"grad_norm": 10.38111686706543,
"learning_rate": 8.612612612612612e-06,
"loss": 0.1235,
"step": 231
},
{
"epoch": 0.6951310861423221,
"grad_norm": 16.827526092529297,
"learning_rate": 8.606606606606607e-06,
"loss": 0.1645,
"step": 232
},
{
"epoch": 0.69812734082397,
"grad_norm": 6.083835601806641,
"learning_rate": 8.600600600600602e-06,
"loss": 0.0408,
"step": 233
},
{
"epoch": 0.701123595505618,
"grad_norm": 11.81071662902832,
"learning_rate": 8.594594594594595e-06,
"loss": 0.1148,
"step": 234
},
{
"epoch": 0.704119850187266,
"grad_norm": 19.527591705322266,
"learning_rate": 8.588588588588589e-06,
"loss": 0.1755,
"step": 235
},
{
"epoch": 0.7071161048689139,
"grad_norm": 29.794513702392578,
"learning_rate": 8.582582582582584e-06,
"loss": 0.276,
"step": 236
},
{
"epoch": 0.7101123595505618,
"grad_norm": 4.606306076049805,
"learning_rate": 8.576576576576577e-06,
"loss": 0.0844,
"step": 237
},
{
"epoch": 0.7131086142322097,
"grad_norm": 7.751136302947998,
"learning_rate": 8.570570570570572e-06,
"loss": 0.0422,
"step": 238
},
{
"epoch": 0.7161048689138577,
"grad_norm": 25.384265899658203,
"learning_rate": 8.564564564564565e-06,
"loss": 0.2254,
"step": 239
},
{
"epoch": 0.7191011235955056,
"grad_norm": 17.381261825561523,
"learning_rate": 8.55855855855856e-06,
"loss": 0.2021,
"step": 240
},
{
"epoch": 0.7220973782771536,
"grad_norm": 6.3656721115112305,
"learning_rate": 8.552552552552553e-06,
"loss": 0.0418,
"step": 241
},
{
"epoch": 0.7250936329588015,
"grad_norm": 17.37508773803711,
"learning_rate": 8.546546546546547e-06,
"loss": 0.2149,
"step": 242
},
{
"epoch": 0.7280898876404495,
"grad_norm": 15.539769172668457,
"learning_rate": 8.540540540540542e-06,
"loss": 0.1091,
"step": 243
},
{
"epoch": 0.7310861423220973,
"grad_norm": 20.707759857177734,
"learning_rate": 8.534534534534535e-06,
"loss": 0.193,
"step": 244
},
{
"epoch": 0.7340823970037453,
"grad_norm": 58.69212341308594,
"learning_rate": 8.52852852852853e-06,
"loss": 0.3135,
"step": 245
},
{
"epoch": 0.7370786516853932,
"grad_norm": 7.689098834991455,
"learning_rate": 8.522522522522523e-06,
"loss": 0.1342,
"step": 246
},
{
"epoch": 0.7400749063670412,
"grad_norm": 4.829769134521484,
"learning_rate": 8.516516516516518e-06,
"loss": 0.0578,
"step": 247
},
{
"epoch": 0.7430711610486891,
"grad_norm": 22.263853073120117,
"learning_rate": 8.510510510510511e-06,
"loss": 0.1343,
"step": 248
},
{
"epoch": 0.7460674157303371,
"grad_norm": 7.045224666595459,
"learning_rate": 8.504504504504505e-06,
"loss": 0.2689,
"step": 249
},
{
"epoch": 0.7490636704119851,
"grad_norm": 6.928183555603027,
"learning_rate": 8.4984984984985e-06,
"loss": 0.1844,
"step": 250
},
{
"epoch": 0.752059925093633,
"grad_norm": 28.74045753479004,
"learning_rate": 8.492492492492493e-06,
"loss": 0.0597,
"step": 251
},
{
"epoch": 0.755056179775281,
"grad_norm": 13.25463581085205,
"learning_rate": 8.486486486486488e-06,
"loss": 0.1473,
"step": 252
},
{
"epoch": 0.7580524344569288,
"grad_norm": 15.956781387329102,
"learning_rate": 8.480480480480481e-06,
"loss": 0.3126,
"step": 253
},
{
"epoch": 0.7610486891385768,
"grad_norm": 2.675602674484253,
"learning_rate": 8.474474474474476e-06,
"loss": 0.0166,
"step": 254
},
{
"epoch": 0.7640449438202247,
"grad_norm": 37.20961380004883,
"learning_rate": 8.46846846846847e-06,
"loss": 0.1657,
"step": 255
},
{
"epoch": 0.7670411985018727,
"grad_norm": 15.512018203735352,
"learning_rate": 8.462462462462462e-06,
"loss": 0.0386,
"step": 256
},
{
"epoch": 0.7700374531835206,
"grad_norm": 32.963443756103516,
"learning_rate": 8.456456456456457e-06,
"loss": 0.2774,
"step": 257
},
{
"epoch": 0.7730337078651686,
"grad_norm": 16.815654754638672,
"learning_rate": 8.45045045045045e-06,
"loss": 0.129,
"step": 258
},
{
"epoch": 0.7760299625468164,
"grad_norm": 6.724097728729248,
"learning_rate": 8.444444444444446e-06,
"loss": 0.1062,
"step": 259
},
{
"epoch": 0.7790262172284644,
"grad_norm": 29.041170120239258,
"learning_rate": 8.438438438438439e-06,
"loss": 0.2664,
"step": 260
},
{
"epoch": 0.7820224719101123,
"grad_norm": 12.00438117980957,
"learning_rate": 8.432432432432434e-06,
"loss": 0.1892,
"step": 261
},
{
"epoch": 0.7850187265917603,
"grad_norm": 11.682869911193848,
"learning_rate": 8.426426426426427e-06,
"loss": 0.1562,
"step": 262
},
{
"epoch": 0.7880149812734082,
"grad_norm": 8.800082206726074,
"learning_rate": 8.42042042042042e-06,
"loss": 0.1216,
"step": 263
},
{
"epoch": 0.7910112359550562,
"grad_norm": 7.602773666381836,
"learning_rate": 8.414414414414415e-06,
"loss": 0.1247,
"step": 264
},
{
"epoch": 0.7940074906367042,
"grad_norm": 27.003143310546875,
"learning_rate": 8.408408408408409e-06,
"loss": 0.1152,
"step": 265
},
{
"epoch": 0.797003745318352,
"grad_norm": 8.360997200012207,
"learning_rate": 8.402402402402404e-06,
"loss": 0.0904,
"step": 266
},
{
"epoch": 0.8,
"grad_norm": 19.083864212036133,
"learning_rate": 8.396396396396397e-06,
"loss": 0.2671,
"step": 267
},
{
"epoch": 0.8029962546816479,
"grad_norm": 8.989147186279297,
"learning_rate": 8.390390390390392e-06,
"loss": 0.1077,
"step": 268
},
{
"epoch": 0.8059925093632959,
"grad_norm": 8.945937156677246,
"learning_rate": 8.384384384384385e-06,
"loss": 0.0694,
"step": 269
},
{
"epoch": 0.8089887640449438,
"grad_norm": 24.782485961914062,
"learning_rate": 8.378378378378378e-06,
"loss": 0.229,
"step": 270
},
{
"epoch": 0.8119850187265918,
"grad_norm": 13.788922309875488,
"learning_rate": 8.372372372372373e-06,
"loss": 0.288,
"step": 271
},
{
"epoch": 0.8149812734082397,
"grad_norm": 2.8008830547332764,
"learning_rate": 8.366366366366368e-06,
"loss": 0.0261,
"step": 272
},
{
"epoch": 0.8179775280898877,
"grad_norm": 8.207698822021484,
"learning_rate": 8.360360360360362e-06,
"loss": 0.0412,
"step": 273
},
{
"epoch": 0.8209737827715355,
"grad_norm": 22.58644676208496,
"learning_rate": 8.354354354354355e-06,
"loss": 0.1497,
"step": 274
},
{
"epoch": 0.8239700374531835,
"grad_norm": 45.544166564941406,
"learning_rate": 8.348348348348348e-06,
"loss": 0.5003,
"step": 275
},
{
"epoch": 0.8269662921348314,
"grad_norm": 21.0858097076416,
"learning_rate": 8.342342342342343e-06,
"loss": 0.134,
"step": 276
},
{
"epoch": 0.8299625468164794,
"grad_norm": 9.136311531066895,
"learning_rate": 8.336336336336336e-06,
"loss": 0.1608,
"step": 277
},
{
"epoch": 0.8329588014981273,
"grad_norm": 34.26778030395508,
"learning_rate": 8.330330330330331e-06,
"loss": 0.1974,
"step": 278
},
{
"epoch": 0.8359550561797753,
"grad_norm": 18.104585647583008,
"learning_rate": 8.324324324324326e-06,
"loss": 0.2067,
"step": 279
},
{
"epoch": 0.8389513108614233,
"grad_norm": 5.197758674621582,
"learning_rate": 8.31831831831832e-06,
"loss": 0.0304,
"step": 280
},
{
"epoch": 0.8419475655430712,
"grad_norm": 17.64066505432129,
"learning_rate": 8.312312312312313e-06,
"loss": 0.1255,
"step": 281
},
{
"epoch": 0.8449438202247191,
"grad_norm": 7.24787712097168,
"learning_rate": 8.306306306306306e-06,
"loss": 0.0542,
"step": 282
},
{
"epoch": 0.847940074906367,
"grad_norm": 22.338401794433594,
"learning_rate": 8.300300300300301e-06,
"loss": 0.2414,
"step": 283
},
{
"epoch": 0.850936329588015,
"grad_norm": 2.4518821239471436,
"learning_rate": 8.294294294294294e-06,
"loss": 0.0225,
"step": 284
},
{
"epoch": 0.8539325842696629,
"grad_norm": 10.993781089782715,
"learning_rate": 8.288288288288289e-06,
"loss": 0.1272,
"step": 285
},
{
"epoch": 0.8569288389513109,
"grad_norm": 20.767696380615234,
"learning_rate": 8.282282282282284e-06,
"loss": 0.2292,
"step": 286
},
{
"epoch": 0.8599250936329588,
"grad_norm": 8.080395698547363,
"learning_rate": 8.276276276276277e-06,
"loss": 0.0724,
"step": 287
},
{
"epoch": 0.8629213483146068,
"grad_norm": 12.226408004760742,
"learning_rate": 8.27027027027027e-06,
"loss": 0.0517,
"step": 288
},
{
"epoch": 0.8659176029962546,
"grad_norm": 19.302963256835938,
"learning_rate": 8.264264264264264e-06,
"loss": 0.2312,
"step": 289
},
{
"epoch": 0.8689138576779026,
"grad_norm": 12.43891716003418,
"learning_rate": 8.258258258258259e-06,
"loss": 0.1691,
"step": 290
},
{
"epoch": 0.8719101123595505,
"grad_norm": 22.69147300720215,
"learning_rate": 8.252252252252254e-06,
"loss": 0.4963,
"step": 291
},
{
"epoch": 0.8749063670411985,
"grad_norm": 28.830711364746094,
"learning_rate": 8.246246246246247e-06,
"loss": 0.1232,
"step": 292
},
{
"epoch": 0.8779026217228464,
"grad_norm": 18.377243041992188,
"learning_rate": 8.24024024024024e-06,
"loss": 0.3857,
"step": 293
},
{
"epoch": 0.8808988764044944,
"grad_norm": 15.206390380859375,
"learning_rate": 8.234234234234235e-06,
"loss": 0.3878,
"step": 294
},
{
"epoch": 0.8838951310861424,
"grad_norm": 8.833643913269043,
"learning_rate": 8.228228228228229e-06,
"loss": 0.0323,
"step": 295
},
{
"epoch": 0.8868913857677903,
"grad_norm": 23.821523666381836,
"learning_rate": 8.222222222222222e-06,
"loss": 0.0463,
"step": 296
},
{
"epoch": 0.8898876404494382,
"grad_norm": 6.677974700927734,
"learning_rate": 8.216216216216217e-06,
"loss": 0.0694,
"step": 297
},
{
"epoch": 0.8928838951310861,
"grad_norm": 9.413395881652832,
"learning_rate": 8.210210210210212e-06,
"loss": 0.1021,
"step": 298
},
{
"epoch": 0.8958801498127341,
"grad_norm": 29.241336822509766,
"learning_rate": 8.204204204204205e-06,
"loss": 0.3005,
"step": 299
},
{
"epoch": 0.898876404494382,
"grad_norm": 36.27946090698242,
"learning_rate": 8.198198198198198e-06,
"loss": 0.1948,
"step": 300
},
{
"epoch": 0.90187265917603,
"grad_norm": 28.775794982910156,
"learning_rate": 8.192192192192193e-06,
"loss": 0.2941,
"step": 301
},
{
"epoch": 0.9048689138576779,
"grad_norm": 17.307905197143555,
"learning_rate": 8.186186186186186e-06,
"loss": 0.2186,
"step": 302
},
{
"epoch": 0.9078651685393259,
"grad_norm": 10.828490257263184,
"learning_rate": 8.18018018018018e-06,
"loss": 0.1976,
"step": 303
},
{
"epoch": 0.9108614232209737,
"grad_norm": 19.335052490234375,
"learning_rate": 8.174174174174175e-06,
"loss": 0.1453,
"step": 304
},
{
"epoch": 0.9138576779026217,
"grad_norm": 3.647679328918457,
"learning_rate": 8.16816816816817e-06,
"loss": 0.0316,
"step": 305
},
{
"epoch": 0.9168539325842696,
"grad_norm": 6.035706043243408,
"learning_rate": 8.162162162162163e-06,
"loss": 0.1271,
"step": 306
},
{
"epoch": 0.9198501872659176,
"grad_norm": 1.434196949005127,
"learning_rate": 8.156156156156156e-06,
"loss": 0.0144,
"step": 307
},
{
"epoch": 0.9228464419475655,
"grad_norm": 9.214011192321777,
"learning_rate": 8.150150150150151e-06,
"loss": 0.0604,
"step": 308
},
{
"epoch": 0.9258426966292135,
"grad_norm": 12.949444770812988,
"learning_rate": 8.144144144144144e-06,
"loss": 0.1786,
"step": 309
},
{
"epoch": 0.9288389513108615,
"grad_norm": 23.851194381713867,
"learning_rate": 8.13813813813814e-06,
"loss": 0.1708,
"step": 310
},
{
"epoch": 0.9318352059925094,
"grad_norm": 12.829343795776367,
"learning_rate": 8.132132132132133e-06,
"loss": 0.1446,
"step": 311
},
{
"epoch": 0.9348314606741573,
"grad_norm": 15.807905197143555,
"learning_rate": 8.126126126126128e-06,
"loss": 0.1975,
"step": 312
},
{
"epoch": 0.9378277153558052,
"grad_norm": 8.796603202819824,
"learning_rate": 8.12012012012012e-06,
"loss": 0.187,
"step": 313
},
{
"epoch": 0.9408239700374532,
"grad_norm": 10.466839790344238,
"learning_rate": 8.114114114114114e-06,
"loss": 0.2681,
"step": 314
},
{
"epoch": 0.9438202247191011,
"grad_norm": 7.512752056121826,
"learning_rate": 8.108108108108109e-06,
"loss": 0.1491,
"step": 315
},
{
"epoch": 0.9468164794007491,
"grad_norm": 15.471590995788574,
"learning_rate": 8.102102102102102e-06,
"loss": 0.1303,
"step": 316
},
{
"epoch": 0.949812734082397,
"grad_norm": 12.372467994689941,
"learning_rate": 8.096096096096097e-06,
"loss": 0.1722,
"step": 317
},
{
"epoch": 0.952808988764045,
"grad_norm": 35.21929931640625,
"learning_rate": 8.09009009009009e-06,
"loss": 0.1994,
"step": 318
},
{
"epoch": 0.9558052434456928,
"grad_norm": 10.004145622253418,
"learning_rate": 8.084084084084085e-06,
"loss": 0.2364,
"step": 319
},
{
"epoch": 0.9588014981273408,
"grad_norm": 18.422414779663086,
"learning_rate": 8.078078078078079e-06,
"loss": 0.174,
"step": 320
},
{
"epoch": 0.9617977528089887,
"grad_norm": 29.870635986328125,
"learning_rate": 8.072072072072072e-06,
"loss": 0.2527,
"step": 321
},
{
"epoch": 0.9647940074906367,
"grad_norm": 22.32419776916504,
"learning_rate": 8.066066066066067e-06,
"loss": 0.1479,
"step": 322
},
{
"epoch": 0.9677902621722847,
"grad_norm": 4.837985515594482,
"learning_rate": 8.06006006006006e-06,
"loss": 0.0536,
"step": 323
},
{
"epoch": 0.9707865168539326,
"grad_norm": 13.291611671447754,
"learning_rate": 8.054054054054055e-06,
"loss": 0.1632,
"step": 324
},
{
"epoch": 0.9737827715355806,
"grad_norm": 15.556182861328125,
"learning_rate": 8.048048048048048e-06,
"loss": 0.2155,
"step": 325
},
{
"epoch": 0.9767790262172285,
"grad_norm": 7.722557067871094,
"learning_rate": 8.042042042042043e-06,
"loss": 0.1771,
"step": 326
},
{
"epoch": 0.9797752808988764,
"grad_norm": 7.998552322387695,
"learning_rate": 8.036036036036037e-06,
"loss": 0.1449,
"step": 327
},
{
"epoch": 0.9827715355805243,
"grad_norm": 6.057829856872559,
"learning_rate": 8.03003003003003e-06,
"loss": 0.0371,
"step": 328
},
{
"epoch": 0.9857677902621723,
"grad_norm": 21.65921401977539,
"learning_rate": 8.024024024024025e-06,
"loss": 0.1521,
"step": 329
},
{
"epoch": 0.9887640449438202,
"grad_norm": 7.719748497009277,
"learning_rate": 8.018018018018018e-06,
"loss": 0.0284,
"step": 330
},
{
"epoch": 0.9917602996254682,
"grad_norm": 10.670609474182129,
"learning_rate": 8.012012012012013e-06,
"loss": 0.1507,
"step": 331
},
{
"epoch": 0.9947565543071161,
"grad_norm": 7.674363136291504,
"learning_rate": 8.006006006006006e-06,
"loss": 0.1219,
"step": 332
},
{
"epoch": 0.9977528089887641,
"grad_norm": 17.153457641601562,
"learning_rate": 8.000000000000001e-06,
"loss": 0.0474,
"step": 333
},
{
"epoch": 1.0,
"grad_norm": 29.845190048217773,
"learning_rate": 7.993993993993995e-06,
"loss": 0.1987,
"step": 334
},
{
"epoch": 1.0,
"eval_accuracy": 0.9591474245115453,
"eval_loss": 0.11946738511323929,
"eval_runtime": 21.1179,
"eval_samples_per_second": 26.66,
"eval_steps_per_second": 3.362,
"step": 334
},
{
"epoch": 1.002996254681648,
"grad_norm": 8.704512596130371,
"learning_rate": 7.987987987987988e-06,
"loss": 0.0673,
"step": 335
},
{
"epoch": 1.005992509363296,
"grad_norm": 2.9829344749450684,
"learning_rate": 7.981981981981983e-06,
"loss": 0.0514,
"step": 336
},
{
"epoch": 1.0089887640449438,
"grad_norm": 8.938423156738281,
"learning_rate": 7.975975975975978e-06,
"loss": 0.0306,
"step": 337
},
{
"epoch": 1.0119850187265917,
"grad_norm": 12.165358543395996,
"learning_rate": 7.969969969969971e-06,
"loss": 0.0555,
"step": 338
},
{
"epoch": 1.0149812734082397,
"grad_norm": 8.754467964172363,
"learning_rate": 7.963963963963964e-06,
"loss": 0.1657,
"step": 339
},
{
"epoch": 1.0179775280898877,
"grad_norm": 37.51422882080078,
"learning_rate": 7.95795795795796e-06,
"loss": 0.3651,
"step": 340
},
{
"epoch": 1.0209737827715355,
"grad_norm": 20.821075439453125,
"learning_rate": 7.951951951951953e-06,
"loss": 0.2785,
"step": 341
},
{
"epoch": 1.0239700374531835,
"grad_norm": 28.079492568969727,
"learning_rate": 7.945945945945946e-06,
"loss": 0.0578,
"step": 342
},
{
"epoch": 1.0269662921348315,
"grad_norm": 7.524195194244385,
"learning_rate": 7.93993993993994e-06,
"loss": 0.1938,
"step": 343
},
{
"epoch": 1.0299625468164795,
"grad_norm": 12.441058158874512,
"learning_rate": 7.933933933933936e-06,
"loss": 0.135,
"step": 344
},
{
"epoch": 1.0329588014981272,
"grad_norm": 31.68756675720215,
"learning_rate": 7.927927927927929e-06,
"loss": 0.2056,
"step": 345
},
{
"epoch": 1.0359550561797752,
"grad_norm": 13.770333290100098,
"learning_rate": 7.921921921921922e-06,
"loss": 0.2414,
"step": 346
},
{
"epoch": 1.0389513108614232,
"grad_norm": 8.553635597229004,
"learning_rate": 7.915915915915915e-06,
"loss": 0.0912,
"step": 347
},
{
"epoch": 1.0419475655430712,
"grad_norm": 7.250548839569092,
"learning_rate": 7.90990990990991e-06,
"loss": 0.1249,
"step": 348
},
{
"epoch": 1.0449438202247192,
"grad_norm": 5.451450824737549,
"learning_rate": 7.903903903903904e-06,
"loss": 0.0234,
"step": 349
},
{
"epoch": 1.047940074906367,
"grad_norm": 86.4170913696289,
"learning_rate": 7.897897897897899e-06,
"loss": 0.2227,
"step": 350
},
{
"epoch": 1.050936329588015,
"grad_norm": 20.818153381347656,
"learning_rate": 7.891891891891894e-06,
"loss": 0.139,
"step": 351
},
{
"epoch": 1.053932584269663,
"grad_norm": 2.33195161819458,
"learning_rate": 7.885885885885887e-06,
"loss": 0.0168,
"step": 352
},
{
"epoch": 1.056928838951311,
"grad_norm": 4.339377403259277,
"learning_rate": 7.87987987987988e-06,
"loss": 0.0861,
"step": 353
},
{
"epoch": 1.0599250936329587,
"grad_norm": 2.396423101425171,
"learning_rate": 7.873873873873873e-06,
"loss": 0.021,
"step": 354
},
{
"epoch": 1.0629213483146067,
"grad_norm": 11.536689758300781,
"learning_rate": 7.867867867867868e-06,
"loss": 0.0589,
"step": 355
},
{
"epoch": 1.0659176029962547,
"grad_norm": 12.232866287231445,
"learning_rate": 7.861861861861863e-06,
"loss": 0.2006,
"step": 356
},
{
"epoch": 1.0689138576779027,
"grad_norm": 1.4649150371551514,
"learning_rate": 7.855855855855857e-06,
"loss": 0.0155,
"step": 357
},
{
"epoch": 1.0719101123595505,
"grad_norm": 10.53365421295166,
"learning_rate": 7.849849849849852e-06,
"loss": 0.101,
"step": 358
},
{
"epoch": 1.0749063670411985,
"grad_norm": 22.114564895629883,
"learning_rate": 7.843843843843845e-06,
"loss": 0.0534,
"step": 359
},
{
"epoch": 1.0779026217228465,
"grad_norm": 21.454910278320312,
"learning_rate": 7.837837837837838e-06,
"loss": 0.0903,
"step": 360
},
{
"epoch": 1.0808988764044944,
"grad_norm": 1.3954575061798096,
"learning_rate": 7.831831831831831e-06,
"loss": 0.0103,
"step": 361
},
{
"epoch": 1.0838951310861424,
"grad_norm": 14.362143516540527,
"learning_rate": 7.825825825825826e-06,
"loss": 0.1225,
"step": 362
},
{
"epoch": 1.0868913857677902,
"grad_norm": 14.380382537841797,
"learning_rate": 7.819819819819821e-06,
"loss": 0.0574,
"step": 363
},
{
"epoch": 1.0898876404494382,
"grad_norm": 29.942373275756836,
"learning_rate": 7.813813813813815e-06,
"loss": 0.0988,
"step": 364
},
{
"epoch": 1.0928838951310862,
"grad_norm": 2.7837705612182617,
"learning_rate": 7.807807807807808e-06,
"loss": 0.0099,
"step": 365
},
{
"epoch": 1.0958801498127342,
"grad_norm": 0.6954925656318665,
"learning_rate": 7.801801801801803e-06,
"loss": 0.0065,
"step": 366
},
{
"epoch": 1.098876404494382,
"grad_norm": 14.443358421325684,
"learning_rate": 7.795795795795796e-06,
"loss": 0.112,
"step": 367
},
{
"epoch": 1.10187265917603,
"grad_norm": 5.893136024475098,
"learning_rate": 7.78978978978979e-06,
"loss": 0.1652,
"step": 368
},
{
"epoch": 1.104868913857678,
"grad_norm": 6.0014824867248535,
"learning_rate": 7.783783783783784e-06,
"loss": 0.021,
"step": 369
},
{
"epoch": 1.107865168539326,
"grad_norm": 8.743016242980957,
"learning_rate": 7.77777777777778e-06,
"loss": 0.1724,
"step": 370
},
{
"epoch": 1.1108614232209737,
"grad_norm": 0.37825778126716614,
"learning_rate": 7.771771771771772e-06,
"loss": 0.0046,
"step": 371
},
{
"epoch": 1.1138576779026217,
"grad_norm": 8.950119972229004,
"learning_rate": 7.765765765765766e-06,
"loss": 0.1493,
"step": 372
},
{
"epoch": 1.1168539325842697,
"grad_norm": 14.684659004211426,
"learning_rate": 7.75975975975976e-06,
"loss": 0.4266,
"step": 373
},
{
"epoch": 1.1198501872659177,
"grad_norm": 4.281994342803955,
"learning_rate": 7.753753753753754e-06,
"loss": 0.0188,
"step": 374
},
{
"epoch": 1.1228464419475654,
"grad_norm": 17.326810836791992,
"learning_rate": 7.747747747747749e-06,
"loss": 0.1295,
"step": 375
},
{
"epoch": 1.1258426966292134,
"grad_norm": 7.717726707458496,
"learning_rate": 7.741741741741742e-06,
"loss": 0.1392,
"step": 376
},
{
"epoch": 1.1288389513108614,
"grad_norm": 9.157553672790527,
"learning_rate": 7.735735735735737e-06,
"loss": 0.0142,
"step": 377
},
{
"epoch": 1.1318352059925094,
"grad_norm": 12.936677932739258,
"learning_rate": 7.72972972972973e-06,
"loss": 0.0543,
"step": 378
},
{
"epoch": 1.1348314606741572,
"grad_norm": 79.90240478515625,
"learning_rate": 7.723723723723724e-06,
"loss": 0.1265,
"step": 379
},
{
"epoch": 1.1378277153558052,
"grad_norm": 39.708614349365234,
"learning_rate": 7.717717717717719e-06,
"loss": 0.3713,
"step": 380
},
{
"epoch": 1.1408239700374532,
"grad_norm": 35.54008102416992,
"learning_rate": 7.711711711711712e-06,
"loss": 0.1263,
"step": 381
},
{
"epoch": 1.1438202247191012,
"grad_norm": 0.7054191827774048,
"learning_rate": 7.705705705705707e-06,
"loss": 0.0061,
"step": 382
},
{
"epoch": 1.1468164794007492,
"grad_norm": 8.379510879516602,
"learning_rate": 7.6996996996997e-06,
"loss": 0.1285,
"step": 383
},
{
"epoch": 1.149812734082397,
"grad_norm": 1.627866506576538,
"learning_rate": 7.693693693693695e-06,
"loss": 0.0154,
"step": 384
},
{
"epoch": 1.152808988764045,
"grad_norm": 4.262239933013916,
"learning_rate": 7.687687687687688e-06,
"loss": 0.0106,
"step": 385
},
{
"epoch": 1.155805243445693,
"grad_norm": 10.119087219238281,
"learning_rate": 7.681681681681682e-06,
"loss": 0.1108,
"step": 386
},
{
"epoch": 1.158801498127341,
"grad_norm": 0.4903702139854431,
"learning_rate": 7.675675675675676e-06,
"loss": 0.0048,
"step": 387
},
{
"epoch": 1.1617977528089887,
"grad_norm": 14.855613708496094,
"learning_rate": 7.66966966966967e-06,
"loss": 0.2215,
"step": 388
},
{
"epoch": 1.1647940074906367,
"grad_norm": 0.7505202293395996,
"learning_rate": 7.663663663663665e-06,
"loss": 0.0071,
"step": 389
},
{
"epoch": 1.1677902621722847,
"grad_norm": 16.078067779541016,
"learning_rate": 7.657657657657658e-06,
"loss": 0.0508,
"step": 390
},
{
"epoch": 1.1707865168539326,
"grad_norm": 3.9604501724243164,
"learning_rate": 7.651651651651653e-06,
"loss": 0.2049,
"step": 391
},
{
"epoch": 1.1737827715355804,
"grad_norm": 4.397295951843262,
"learning_rate": 7.645645645645646e-06,
"loss": 0.0075,
"step": 392
},
{
"epoch": 1.1767790262172284,
"grad_norm": 7.140230178833008,
"learning_rate": 7.63963963963964e-06,
"loss": 0.1843,
"step": 393
},
{
"epoch": 1.1797752808988764,
"grad_norm": 1.3169139623641968,
"learning_rate": 7.633633633633634e-06,
"loss": 0.0049,
"step": 394
},
{
"epoch": 1.1827715355805244,
"grad_norm": 19.79604721069336,
"learning_rate": 7.6276276276276285e-06,
"loss": 0.2981,
"step": 395
},
{
"epoch": 1.1857677902621724,
"grad_norm": 23.419668197631836,
"learning_rate": 7.621621621621622e-06,
"loss": 0.0854,
"step": 396
},
{
"epoch": 1.1887640449438202,
"grad_norm": 11.932912826538086,
"learning_rate": 7.615615615615616e-06,
"loss": 0.2126,
"step": 397
},
{
"epoch": 1.1917602996254681,
"grad_norm": 2.394584894180298,
"learning_rate": 7.609609609609611e-06,
"loss": 0.0109,
"step": 398
},
{
"epoch": 1.1947565543071161,
"grad_norm": 37.54982376098633,
"learning_rate": 7.603603603603604e-06,
"loss": 0.2556,
"step": 399
},
{
"epoch": 1.1977528089887641,
"grad_norm": 16.414182662963867,
"learning_rate": 7.597597597597598e-06,
"loss": 0.3071,
"step": 400
},
{
"epoch": 1.200749063670412,
"grad_norm": 8.574248313903809,
"learning_rate": 7.591591591591592e-06,
"loss": 0.1423,
"step": 401
},
{
"epoch": 1.20374531835206,
"grad_norm": 5.446165084838867,
"learning_rate": 7.5855855855855865e-06,
"loss": 0.13,
"step": 402
},
{
"epoch": 1.2067415730337079,
"grad_norm": 4.059351444244385,
"learning_rate": 7.57957957957958e-06,
"loss": 0.0203,
"step": 403
},
{
"epoch": 1.2097378277153559,
"grad_norm": 8.453432083129883,
"learning_rate": 7.573573573573574e-06,
"loss": 0.1575,
"step": 404
},
{
"epoch": 1.2127340823970036,
"grad_norm": 15.920672416687012,
"learning_rate": 7.567567567567569e-06,
"loss": 0.1366,
"step": 405
},
{
"epoch": 1.2157303370786516,
"grad_norm": 32.0057258605957,
"learning_rate": 7.561561561561562e-06,
"loss": 0.1107,
"step": 406
},
{
"epoch": 1.2187265917602996,
"grad_norm": 13.950464248657227,
"learning_rate": 7.555555555555556e-06,
"loss": 0.0376,
"step": 407
},
{
"epoch": 1.2217228464419476,
"grad_norm": 28.328393936157227,
"learning_rate": 7.549549549549549e-06,
"loss": 0.2798,
"step": 408
},
{
"epoch": 1.2247191011235956,
"grad_norm": 27.61097526550293,
"learning_rate": 7.543543543543544e-06,
"loss": 0.0284,
"step": 409
},
{
"epoch": 1.2277153558052434,
"grad_norm": 3.6726973056793213,
"learning_rate": 7.5375375375375385e-06,
"loss": 0.0174,
"step": 410
},
{
"epoch": 1.2307116104868914,
"grad_norm": 6.744252681732178,
"learning_rate": 7.531531531531532e-06,
"loss": 0.1594,
"step": 411
},
{
"epoch": 1.2337078651685394,
"grad_norm": 4.745820045471191,
"learning_rate": 7.525525525525527e-06,
"loss": 0.0222,
"step": 412
},
{
"epoch": 1.2367041198501874,
"grad_norm": 10.849603652954102,
"learning_rate": 7.51951951951952e-06,
"loss": 0.2271,
"step": 413
},
{
"epoch": 1.2397003745318351,
"grad_norm": 8.78612995147705,
"learning_rate": 7.513513513513514e-06,
"loss": 0.109,
"step": 414
},
{
"epoch": 1.2426966292134831,
"grad_norm": 1.3302922248840332,
"learning_rate": 7.507507507507507e-06,
"loss": 0.0111,
"step": 415
},
{
"epoch": 1.245692883895131,
"grad_norm": 8.550155639648438,
"learning_rate": 7.501501501501502e-06,
"loss": 0.0296,
"step": 416
},
{
"epoch": 1.248689138576779,
"grad_norm": 25.59956169128418,
"learning_rate": 7.495495495495496e-06,
"loss": 0.2662,
"step": 417
},
{
"epoch": 1.2516853932584269,
"grad_norm": 26.77549934387207,
"learning_rate": 7.48948948948949e-06,
"loss": 0.1273,
"step": 418
},
{
"epoch": 1.2546816479400749,
"grad_norm": 2.1836867332458496,
"learning_rate": 7.483483483483485e-06,
"loss": 0.0123,
"step": 419
},
{
"epoch": 1.2576779026217229,
"grad_norm": 14.294233322143555,
"learning_rate": 7.477477477477479e-06,
"loss": 0.0355,
"step": 420
},
{
"epoch": 1.2606741573033708,
"grad_norm": 15.087601661682129,
"learning_rate": 7.471471471471472e-06,
"loss": 0.2491,
"step": 421
},
{
"epoch": 1.2636704119850188,
"grad_norm": 10.146037101745605,
"learning_rate": 7.465465465465466e-06,
"loss": 0.0924,
"step": 422
},
{
"epoch": 1.2666666666666666,
"grad_norm": 18.425317764282227,
"learning_rate": 7.45945945945946e-06,
"loss": 0.2718,
"step": 423
},
{
"epoch": 1.2696629213483146,
"grad_norm": 24.406692504882812,
"learning_rate": 7.453453453453454e-06,
"loss": 0.1304,
"step": 424
},
{
"epoch": 1.2726591760299626,
"grad_norm": 13.614672660827637,
"learning_rate": 7.447447447447448e-06,
"loss": 0.2901,
"step": 425
},
{
"epoch": 1.2756554307116104,
"grad_norm": 1.9076930284500122,
"learning_rate": 7.441441441441442e-06,
"loss": 0.0173,
"step": 426
},
{
"epoch": 1.2786516853932584,
"grad_norm": 20.179975509643555,
"learning_rate": 7.435435435435437e-06,
"loss": 0.0924,
"step": 427
},
{
"epoch": 1.2816479400749063,
"grad_norm": 16.551319122314453,
"learning_rate": 7.42942942942943e-06,
"loss": 0.0628,
"step": 428
},
{
"epoch": 1.2846441947565543,
"grad_norm": 6.466628551483154,
"learning_rate": 7.423423423423424e-06,
"loss": 0.0284,
"step": 429
},
{
"epoch": 1.2876404494382023,
"grad_norm": 47.095340728759766,
"learning_rate": 7.417417417417418e-06,
"loss": 0.0875,
"step": 430
},
{
"epoch": 1.29063670411985,
"grad_norm": 10.281209945678711,
"learning_rate": 7.411411411411412e-06,
"loss": 0.0202,
"step": 431
},
{
"epoch": 1.293632958801498,
"grad_norm": 12.131975173950195,
"learning_rate": 7.4054054054054055e-06,
"loss": 0.1233,
"step": 432
},
{
"epoch": 1.296629213483146,
"grad_norm": 3.8274757862091064,
"learning_rate": 7.3993993993994e-06,
"loss": 0.0253,
"step": 433
},
{
"epoch": 1.299625468164794,
"grad_norm": 13.201749801635742,
"learning_rate": 7.393393393393395e-06,
"loss": 0.0396,
"step": 434
},
{
"epoch": 1.302621722846442,
"grad_norm": 4.812632083892822,
"learning_rate": 7.387387387387388e-06,
"loss": 0.104,
"step": 435
},
{
"epoch": 1.3056179775280898,
"grad_norm": 20.616060256958008,
"learning_rate": 7.381381381381382e-06,
"loss": 0.2587,
"step": 436
},
{
"epoch": 1.3086142322097378,
"grad_norm": 16.196264266967773,
"learning_rate": 7.375375375375377e-06,
"loss": 0.1342,
"step": 437
},
{
"epoch": 1.3116104868913858,
"grad_norm": 14.186867713928223,
"learning_rate": 7.36936936936937e-06,
"loss": 0.2195,
"step": 438
},
{
"epoch": 1.3146067415730336,
"grad_norm": 1.1713473796844482,
"learning_rate": 7.363363363363364e-06,
"loss": 0.013,
"step": 439
},
{
"epoch": 1.3176029962546816,
"grad_norm": 17.85066032409668,
"learning_rate": 7.3573573573573575e-06,
"loss": 0.2871,
"step": 440
},
{
"epoch": 1.3205992509363296,
"grad_norm": 6.779399394989014,
"learning_rate": 7.3513513513513525e-06,
"loss": 0.1663,
"step": 441
},
{
"epoch": 1.3235955056179776,
"grad_norm": 41.116676330566406,
"learning_rate": 7.345345345345346e-06,
"loss": 0.0885,
"step": 442
},
{
"epoch": 1.3265917602996256,
"grad_norm": 2.9757792949676514,
"learning_rate": 7.33933933933934e-06,
"loss": 0.0165,
"step": 443
},
{
"epoch": 1.3295880149812733,
"grad_norm": 2.5939130783081055,
"learning_rate": 7.333333333333333e-06,
"loss": 0.0141,
"step": 444
},
{
"epoch": 1.3325842696629213,
"grad_norm": 18.92623519897461,
"learning_rate": 7.327327327327328e-06,
"loss": 0.0533,
"step": 445
},
{
"epoch": 1.3355805243445693,
"grad_norm": 22.663772583007812,
"learning_rate": 7.321321321321322e-06,
"loss": 0.1222,
"step": 446
},
{
"epoch": 1.3385767790262173,
"grad_norm": 1.3068466186523438,
"learning_rate": 7.3153153153153155e-06,
"loss": 0.0125,
"step": 447
},
{
"epoch": 1.3415730337078653,
"grad_norm": 4.3140130043029785,
"learning_rate": 7.3093093093093104e-06,
"loss": 0.1783,
"step": 448
},
{
"epoch": 1.344569288389513,
"grad_norm": 7.359616756439209,
"learning_rate": 7.303303303303304e-06,
"loss": 0.3706,
"step": 449
},
{
"epoch": 1.347565543071161,
"grad_norm": 0.9203206896781921,
"learning_rate": 7.297297297297298e-06,
"loss": 0.0072,
"step": 450
},
{
"epoch": 1.350561797752809,
"grad_norm": 15.939404487609863,
"learning_rate": 7.291291291291291e-06,
"loss": 0.1554,
"step": 451
},
{
"epoch": 1.3535580524344568,
"grad_norm": 5.646368026733398,
"learning_rate": 7.285285285285286e-06,
"loss": 0.0907,
"step": 452
},
{
"epoch": 1.3565543071161048,
"grad_norm": 0.47080573439598083,
"learning_rate": 7.27927927927928e-06,
"loss": 0.0055,
"step": 453
},
{
"epoch": 1.3595505617977528,
"grad_norm": 47.65138244628906,
"learning_rate": 7.273273273273273e-06,
"loss": 0.1361,
"step": 454
},
{
"epoch": 1.3625468164794008,
"grad_norm": 8.572495460510254,
"learning_rate": 7.267267267267268e-06,
"loss": 0.2409,
"step": 455
},
{
"epoch": 1.3655430711610488,
"grad_norm": 12.817127227783203,
"learning_rate": 7.2612612612612625e-06,
"loss": 0.024,
"step": 456
},
{
"epoch": 1.3685393258426966,
"grad_norm": 37.24135208129883,
"learning_rate": 7.255255255255256e-06,
"loss": 0.1954,
"step": 457
},
{
"epoch": 1.3715355805243445,
"grad_norm": 64.09664154052734,
"learning_rate": 7.24924924924925e-06,
"loss": 0.0542,
"step": 458
},
{
"epoch": 1.3745318352059925,
"grad_norm": 0.8731926679611206,
"learning_rate": 7.243243243243244e-06,
"loss": 0.0054,
"step": 459
},
{
"epoch": 1.3775280898876405,
"grad_norm": 13.14030933380127,
"learning_rate": 7.237237237237238e-06,
"loss": 0.1353,
"step": 460
},
{
"epoch": 1.3805243445692885,
"grad_norm": 5.747009754180908,
"learning_rate": 7.231231231231231e-06,
"loss": 0.1059,
"step": 461
},
{
"epoch": 1.3835205992509363,
"grad_norm": 8.522408485412598,
"learning_rate": 7.2252252252252254e-06,
"loss": 0.1198,
"step": 462
},
{
"epoch": 1.3865168539325843,
"grad_norm": 2.3080382347106934,
"learning_rate": 7.21921921921922e-06,
"loss": 0.0152,
"step": 463
},
{
"epoch": 1.3895131086142323,
"grad_norm": 9.45796012878418,
"learning_rate": 7.213213213213214e-06,
"loss": 0.1234,
"step": 464
},
{
"epoch": 1.39250936329588,
"grad_norm": 6.192313194274902,
"learning_rate": 7.207207207207208e-06,
"loss": 0.1422,
"step": 465
},
{
"epoch": 1.395505617977528,
"grad_norm": 1.42098069190979,
"learning_rate": 7.201201201201202e-06,
"loss": 0.0107,
"step": 466
},
{
"epoch": 1.398501872659176,
"grad_norm": 10.835694313049316,
"learning_rate": 7.195195195195196e-06,
"loss": 0.1974,
"step": 467
},
{
"epoch": 1.401498127340824,
"grad_norm": 22.209035873413086,
"learning_rate": 7.189189189189189e-06,
"loss": 0.0638,
"step": 468
},
{
"epoch": 1.404494382022472,
"grad_norm": 34.53959655761719,
"learning_rate": 7.183183183183183e-06,
"loss": 0.0737,
"step": 469
},
{
"epoch": 1.4074906367041198,
"grad_norm": 5.806796550750732,
"learning_rate": 7.177177177177178e-06,
"loss": 0.1531,
"step": 470
},
{
"epoch": 1.4104868913857678,
"grad_norm": 10.332561492919922,
"learning_rate": 7.1711711711711716e-06,
"loss": 0.0959,
"step": 471
},
{
"epoch": 1.4134831460674158,
"grad_norm": 4.447213649749756,
"learning_rate": 7.165165165165166e-06,
"loss": 0.0605,
"step": 472
},
{
"epoch": 1.4164794007490635,
"grad_norm": 3.727720260620117,
"learning_rate": 7.159159159159161e-06,
"loss": 0.0743,
"step": 473
},
{
"epoch": 1.4194756554307117,
"grad_norm": 11.457908630371094,
"learning_rate": 7.153153153153154e-06,
"loss": 0.0506,
"step": 474
},
{
"epoch": 1.4224719101123595,
"grad_norm": 43.636070251464844,
"learning_rate": 7.147147147147148e-06,
"loss": 0.2081,
"step": 475
},
{
"epoch": 1.4254681647940075,
"grad_norm": 15.501445770263672,
"learning_rate": 7.141141141141141e-06,
"loss": 0.0602,
"step": 476
},
{
"epoch": 1.4284644194756555,
"grad_norm": 4.321882724761963,
"learning_rate": 7.135135135135136e-06,
"loss": 0.0219,
"step": 477
},
{
"epoch": 1.4314606741573033,
"grad_norm": 17.598709106445312,
"learning_rate": 7.1291291291291295e-06,
"loss": 0.2692,
"step": 478
},
{
"epoch": 1.4344569288389513,
"grad_norm": 16.832195281982422,
"learning_rate": 7.123123123123124e-06,
"loss": 0.0655,
"step": 479
},
{
"epoch": 1.4374531835205993,
"grad_norm": 2.132268190383911,
"learning_rate": 7.117117117117117e-06,
"loss": 0.017,
"step": 480
},
{
"epoch": 1.4404494382022472,
"grad_norm": 21.201562881469727,
"learning_rate": 7.111111111111112e-06,
"loss": 0.3869,
"step": 481
},
{
"epoch": 1.4434456928838952,
"grad_norm": 4.5265069007873535,
"learning_rate": 7.105105105105106e-06,
"loss": 0.0298,
"step": 482
},
{
"epoch": 1.446441947565543,
"grad_norm": 2.2739415168762207,
"learning_rate": 7.099099099099099e-06,
"loss": 0.0131,
"step": 483
},
{
"epoch": 1.449438202247191,
"grad_norm": 7.272430419921875,
"learning_rate": 7.093093093093094e-06,
"loss": 0.1106,
"step": 484
},
{
"epoch": 1.452434456928839,
"grad_norm": 55.692535400390625,
"learning_rate": 7.087087087087087e-06,
"loss": 0.0646,
"step": 485
},
{
"epoch": 1.4554307116104868,
"grad_norm": 35.61100769042969,
"learning_rate": 7.0810810810810815e-06,
"loss": 0.0924,
"step": 486
},
{
"epoch": 1.4584269662921348,
"grad_norm": 0.891657292842865,
"learning_rate": 7.075075075075075e-06,
"loss": 0.0077,
"step": 487
},
{
"epoch": 1.4614232209737827,
"grad_norm": 25.506799697875977,
"learning_rate": 7.06906906906907e-06,
"loss": 0.255,
"step": 488
},
{
"epoch": 1.4644194756554307,
"grad_norm": 28.113414764404297,
"learning_rate": 7.063063063063064e-06,
"loss": 0.1042,
"step": 489
},
{
"epoch": 1.4674157303370787,
"grad_norm": 5.091357231140137,
"learning_rate": 7.057057057057057e-06,
"loss": 0.0782,
"step": 490
},
{
"epoch": 1.4704119850187265,
"grad_norm": 15.681722640991211,
"learning_rate": 7.051051051051052e-06,
"loss": 0.3396,
"step": 491
},
{
"epoch": 1.4734082397003745,
"grad_norm": 19.764301300048828,
"learning_rate": 7.045045045045046e-06,
"loss": 0.4275,
"step": 492
},
{
"epoch": 1.4764044943820225,
"grad_norm": 10.320667266845703,
"learning_rate": 7.0390390390390395e-06,
"loss": 0.3144,
"step": 493
},
{
"epoch": 1.4794007490636705,
"grad_norm": 0.8163776993751526,
"learning_rate": 7.0330330330330336e-06,
"loss": 0.0076,
"step": 494
},
{
"epoch": 1.4823970037453185,
"grad_norm": 12.22108268737793,
"learning_rate": 7.027027027027028e-06,
"loss": 0.0954,
"step": 495
},
{
"epoch": 1.4853932584269662,
"grad_norm": 23.646915435791016,
"learning_rate": 7.021021021021022e-06,
"loss": 0.1453,
"step": 496
},
{
"epoch": 1.4883895131086142,
"grad_norm": 22.996501922607422,
"learning_rate": 7.015015015015015e-06,
"loss": 0.1926,
"step": 497
},
{
"epoch": 1.4913857677902622,
"grad_norm": 6.45268440246582,
"learning_rate": 7.009009009009009e-06,
"loss": 0.0603,
"step": 498
},
{
"epoch": 1.49438202247191,
"grad_norm": 18.419157028198242,
"learning_rate": 7.003003003003004e-06,
"loss": 0.0509,
"step": 499
},
{
"epoch": 1.497378277153558,
"grad_norm": 30.36125373840332,
"learning_rate": 6.996996996996997e-06,
"loss": 0.239,
"step": 500
},
{
"epoch": 1.500374531835206,
"grad_norm": 12.767863273620605,
"learning_rate": 6.9909909909909915e-06,
"loss": 0.0604,
"step": 501
},
{
"epoch": 1.503370786516854,
"grad_norm": 19.505027770996094,
"learning_rate": 6.984984984984986e-06,
"loss": 0.4936,
"step": 502
},
{
"epoch": 1.506367041198502,
"grad_norm": 9.562569618225098,
"learning_rate": 6.97897897897898e-06,
"loss": 0.1404,
"step": 503
},
{
"epoch": 1.5093632958801497,
"grad_norm": 10.234158515930176,
"learning_rate": 6.972972972972973e-06,
"loss": 0.0465,
"step": 504
},
{
"epoch": 1.5123595505617977,
"grad_norm": 7.085693359375,
"learning_rate": 6.966966966966967e-06,
"loss": 0.0108,
"step": 505
},
{
"epoch": 1.5153558052434457,
"grad_norm": 7.290648460388184,
"learning_rate": 6.960960960960962e-06,
"loss": 0.2304,
"step": 506
},
{
"epoch": 1.5183520599250935,
"grad_norm": 43.409732818603516,
"learning_rate": 6.954954954954955e-06,
"loss": 0.1585,
"step": 507
},
{
"epoch": 1.5213483146067417,
"grad_norm": 4.9887309074401855,
"learning_rate": 6.948948948948949e-06,
"loss": 0.0938,
"step": 508
},
{
"epoch": 1.5243445692883895,
"grad_norm": 1.0400062799453735,
"learning_rate": 6.942942942942944e-06,
"loss": 0.0101,
"step": 509
},
{
"epoch": 1.5273408239700375,
"grad_norm": 5.245956897735596,
"learning_rate": 6.936936936936938e-06,
"loss": 0.0337,
"step": 510
},
{
"epoch": 1.5303370786516854,
"grad_norm": 9.969259262084961,
"learning_rate": 6.930930930930932e-06,
"loss": 0.1278,
"step": 511
},
{
"epoch": 1.5333333333333332,
"grad_norm": 8.882756233215332,
"learning_rate": 6.924924924924925e-06,
"loss": 0.3435,
"step": 512
},
{
"epoch": 1.5363295880149814,
"grad_norm": 5.988860607147217,
"learning_rate": 6.91891891891892e-06,
"loss": 0.0529,
"step": 513
},
{
"epoch": 1.5393258426966292,
"grad_norm": 2.451871156692505,
"learning_rate": 6.912912912912913e-06,
"loss": 0.0198,
"step": 514
},
{
"epoch": 1.5423220973782772,
"grad_norm": 2.6480090618133545,
"learning_rate": 6.906906906906907e-06,
"loss": 0.0213,
"step": 515
},
{
"epoch": 1.5453183520599252,
"grad_norm": 7.050961017608643,
"learning_rate": 6.900900900900901e-06,
"loss": 0.0405,
"step": 516
},
{
"epoch": 1.548314606741573,
"grad_norm": 15.774571418762207,
"learning_rate": 6.8948948948948955e-06,
"loss": 0.2177,
"step": 517
},
{
"epoch": 1.551310861423221,
"grad_norm": 34.69354248046875,
"learning_rate": 6.88888888888889e-06,
"loss": 0.1795,
"step": 518
},
{
"epoch": 1.554307116104869,
"grad_norm": 29.670717239379883,
"learning_rate": 6.882882882882883e-06,
"loss": 0.1604,
"step": 519
},
{
"epoch": 1.5573033707865167,
"grad_norm": 1.4310669898986816,
"learning_rate": 6.876876876876878e-06,
"loss": 0.0117,
"step": 520
},
{
"epoch": 1.560299625468165,
"grad_norm": 38.75698471069336,
"learning_rate": 6.870870870870872e-06,
"loss": 0.0964,
"step": 521
},
{
"epoch": 1.5632958801498127,
"grad_norm": 26.66681671142578,
"learning_rate": 6.864864864864865e-06,
"loss": 0.1959,
"step": 522
},
{
"epoch": 1.5662921348314607,
"grad_norm": 1.6589252948760986,
"learning_rate": 6.8588588588588585e-06,
"loss": 0.0081,
"step": 523
},
{
"epoch": 1.5692883895131087,
"grad_norm": 89.270263671875,
"learning_rate": 6.8528528528528535e-06,
"loss": 0.2365,
"step": 524
},
{
"epoch": 1.5722846441947564,
"grad_norm": 3.4999654293060303,
"learning_rate": 6.846846846846848e-06,
"loss": 0.0102,
"step": 525
},
{
"epoch": 1.5752808988764047,
"grad_norm": 8.316466331481934,
"learning_rate": 6.840840840840841e-06,
"loss": 0.1912,
"step": 526
},
{
"epoch": 1.5782771535580524,
"grad_norm": 24.64727783203125,
"learning_rate": 6.834834834834836e-06,
"loss": 0.4597,
"step": 527
},
{
"epoch": 1.5812734082397004,
"grad_norm": 11.571331024169922,
"learning_rate": 6.82882882882883e-06,
"loss": 0.2734,
"step": 528
},
{
"epoch": 1.5842696629213484,
"grad_norm": 80.09754180908203,
"learning_rate": 6.822822822822823e-06,
"loss": 0.1977,
"step": 529
},
{
"epoch": 1.5872659176029962,
"grad_norm": 1.2138384580612183,
"learning_rate": 6.816816816816817e-06,
"loss": 0.0044,
"step": 530
},
{
"epoch": 1.5902621722846442,
"grad_norm": 3.866417169570923,
"learning_rate": 6.810810810810811e-06,
"loss": 0.0172,
"step": 531
},
{
"epoch": 1.5932584269662922,
"grad_norm": 33.33708953857422,
"learning_rate": 6.8048048048048055e-06,
"loss": 0.2992,
"step": 532
},
{
"epoch": 1.59625468164794,
"grad_norm": 5.784562110900879,
"learning_rate": 6.798798798798799e-06,
"loss": 0.0195,
"step": 533
},
{
"epoch": 1.5992509363295881,
"grad_norm": 49.21910858154297,
"learning_rate": 6.792792792792793e-06,
"loss": 0.1381,
"step": 534
},
{
"epoch": 1.602247191011236,
"grad_norm": 49.31438064575195,
"learning_rate": 6.786786786786788e-06,
"loss": 0.2871,
"step": 535
},
{
"epoch": 1.605243445692884,
"grad_norm": 35.16240310668945,
"learning_rate": 6.780780780780781e-06,
"loss": 0.4596,
"step": 536
},
{
"epoch": 1.608239700374532,
"grad_norm": 7.544735908508301,
"learning_rate": 6.774774774774775e-06,
"loss": 0.0482,
"step": 537
},
{
"epoch": 1.6112359550561797,
"grad_norm": 9.65584659576416,
"learning_rate": 6.76876876876877e-06,
"loss": 0.2395,
"step": 538
},
{
"epoch": 1.6142322097378277,
"grad_norm": 21.089454650878906,
"learning_rate": 6.7627627627627634e-06,
"loss": 0.3148,
"step": 539
},
{
"epoch": 1.6172284644194757,
"grad_norm": 1.6967628002166748,
"learning_rate": 6.7567567567567575e-06,
"loss": 0.0106,
"step": 540
},
{
"epoch": 1.6202247191011236,
"grad_norm": 13.342013359069824,
"learning_rate": 6.750750750750751e-06,
"loss": 0.1322,
"step": 541
},
{
"epoch": 1.6232209737827716,
"grad_norm": 16.040504455566406,
"learning_rate": 6.744744744744746e-06,
"loss": 0.1228,
"step": 542
},
{
"epoch": 1.6262172284644194,
"grad_norm": 20.384510040283203,
"learning_rate": 6.738738738738739e-06,
"loss": 0.1223,
"step": 543
},
{
"epoch": 1.6292134831460674,
"grad_norm": 28.042816162109375,
"learning_rate": 6.732732732732733e-06,
"loss": 0.222,
"step": 544
},
{
"epoch": 1.6322097378277154,
"grad_norm": 15.969573974609375,
"learning_rate": 6.726726726726728e-06,
"loss": 0.1599,
"step": 545
},
{
"epoch": 1.6352059925093632,
"grad_norm": 18.415395736694336,
"learning_rate": 6.720720720720721e-06,
"loss": 0.2701,
"step": 546
},
{
"epoch": 1.6382022471910114,
"grad_norm": 6.37872838973999,
"learning_rate": 6.7147147147147155e-06,
"loss": 0.0233,
"step": 547
},
{
"epoch": 1.6411985018726591,
"grad_norm": 16.17081642150879,
"learning_rate": 6.708708708708709e-06,
"loss": 0.1433,
"step": 548
},
{
"epoch": 1.6441947565543071,
"grad_norm": 17.56867790222168,
"learning_rate": 6.702702702702704e-06,
"loss": 0.2109,
"step": 549
},
{
"epoch": 1.6471910112359551,
"grad_norm": 1.078957200050354,
"learning_rate": 6.696696696696697e-06,
"loss": 0.0067,
"step": 550
},
{
"epoch": 1.650187265917603,
"grad_norm": 11.603385925292969,
"learning_rate": 6.690690690690691e-06,
"loss": 0.0667,
"step": 551
},
{
"epoch": 1.6531835205992509,
"grad_norm": 12.611420631408691,
"learning_rate": 6.684684684684684e-06,
"loss": 0.1252,
"step": 552
},
{
"epoch": 1.6561797752808989,
"grad_norm": 27.921640396118164,
"learning_rate": 6.678678678678679e-06,
"loss": 0.1692,
"step": 553
},
{
"epoch": 1.6591760299625467,
"grad_norm": 26.80571937561035,
"learning_rate": 6.672672672672673e-06,
"loss": 0.1332,
"step": 554
},
{
"epoch": 1.6621722846441949,
"grad_norm": 11.543664932250977,
"learning_rate": 6.666666666666667e-06,
"loss": 0.3633,
"step": 555
},
{
"epoch": 1.6651685393258426,
"grad_norm": 5.858215808868408,
"learning_rate": 6.660660660660662e-06,
"loss": 0.0784,
"step": 556
},
{
"epoch": 1.6681647940074906,
"grad_norm": 7.0159101486206055,
"learning_rate": 6.654654654654656e-06,
"loss": 0.2038,
"step": 557
},
{
"epoch": 1.6711610486891386,
"grad_norm": 1.6647109985351562,
"learning_rate": 6.648648648648649e-06,
"loss": 0.0147,
"step": 558
},
{
"epoch": 1.6741573033707864,
"grad_norm": 13.881926536560059,
"learning_rate": 6.642642642642643e-06,
"loss": 0.2186,
"step": 559
},
{
"epoch": 1.6771535580524346,
"grad_norm": 11.37131404876709,
"learning_rate": 6.636636636636637e-06,
"loss": 0.2356,
"step": 560
},
{
"epoch": 1.6801498127340824,
"grad_norm": 10.474662780761719,
"learning_rate": 6.630630630630631e-06,
"loss": 0.0665,
"step": 561
},
{
"epoch": 1.6831460674157304,
"grad_norm": 6.607825756072998,
"learning_rate": 6.6246246246246246e-06,
"loss": 0.0881,
"step": 562
},
{
"epoch": 1.6861423220973784,
"grad_norm": 6.6188483238220215,
"learning_rate": 6.6186186186186195e-06,
"loss": 0.0392,
"step": 563
},
{
"epoch": 1.6891385767790261,
"grad_norm": 5.466091632843018,
"learning_rate": 6.612612612612614e-06,
"loss": 0.0838,
"step": 564
},
{
"epoch": 1.6921348314606741,
"grad_norm": 9.057148933410645,
"learning_rate": 6.606606606606607e-06,
"loss": 0.1985,
"step": 565
},
{
"epoch": 1.695131086142322,
"grad_norm": 6.330142498016357,
"learning_rate": 6.600600600600601e-06,
"loss": 0.1581,
"step": 566
},
{
"epoch": 1.6981273408239699,
"grad_norm": 11.276079177856445,
"learning_rate": 6.594594594594595e-06,
"loss": 0.1456,
"step": 567
},
{
"epoch": 1.701123595505618,
"grad_norm": 44.33750915527344,
"learning_rate": 6.588588588588589e-06,
"loss": 0.0747,
"step": 568
},
{
"epoch": 1.7041198501872659,
"grad_norm": 2.712155818939209,
"learning_rate": 6.5825825825825825e-06,
"loss": 0.0127,
"step": 569
},
{
"epoch": 1.7071161048689139,
"grad_norm": 9.837244033813477,
"learning_rate": 6.5765765765765775e-06,
"loss": 0.0491,
"step": 570
},
{
"epoch": 1.7101123595505618,
"grad_norm": 15.660017967224121,
"learning_rate": 6.5705705705705716e-06,
"loss": 0.0321,
"step": 571
},
{
"epoch": 1.7131086142322096,
"grad_norm": 3.569031238555908,
"learning_rate": 6.564564564564565e-06,
"loss": 0.0171,
"step": 572
},
{
"epoch": 1.7161048689138578,
"grad_norm": 7.967754364013672,
"learning_rate": 6.558558558558559e-06,
"loss": 0.0196,
"step": 573
},
{
"epoch": 1.7191011235955056,
"grad_norm": 38.77738952636719,
"learning_rate": 6.552552552552554e-06,
"loss": 0.1644,
"step": 574
},
{
"epoch": 1.7220973782771536,
"grad_norm": 15.910918235778809,
"learning_rate": 6.546546546546547e-06,
"loss": 0.0418,
"step": 575
},
{
"epoch": 1.7250936329588016,
"grad_norm": 1.2709795236587524,
"learning_rate": 6.540540540540541e-06,
"loss": 0.0092,
"step": 576
},
{
"epoch": 1.7280898876404494,
"grad_norm": 5.7393083572387695,
"learning_rate": 6.5345345345345345e-06,
"loss": 0.0153,
"step": 577
},
{
"epoch": 1.7310861423220973,
"grad_norm": 9.134088516235352,
"learning_rate": 6.5285285285285295e-06,
"loss": 0.1522,
"step": 578
},
{
"epoch": 1.7340823970037453,
"grad_norm": 1.5715709924697876,
"learning_rate": 6.522522522522523e-06,
"loss": 0.0086,
"step": 579
},
{
"epoch": 1.737078651685393,
"grad_norm": 14.366767883300781,
"learning_rate": 6.516516516516517e-06,
"loss": 0.1476,
"step": 580
},
{
"epoch": 1.7400749063670413,
"grad_norm": 86.79978942871094,
"learning_rate": 6.510510510510512e-06,
"loss": 0.3148,
"step": 581
},
{
"epoch": 1.743071161048689,
"grad_norm": 25.74228858947754,
"learning_rate": 6.504504504504505e-06,
"loss": 0.0823,
"step": 582
},
{
"epoch": 1.746067415730337,
"grad_norm": 2.1786141395568848,
"learning_rate": 6.498498498498499e-06,
"loss": 0.0407,
"step": 583
},
{
"epoch": 1.749063670411985,
"grad_norm": 11.25271224975586,
"learning_rate": 6.4924924924924924e-06,
"loss": 0.1132,
"step": 584
},
{
"epoch": 1.7520599250936328,
"grad_norm": 6.763452053070068,
"learning_rate": 6.486486486486487e-06,
"loss": 0.1273,
"step": 585
},
{
"epoch": 1.755056179775281,
"grad_norm": 11.457433700561523,
"learning_rate": 6.480480480480481e-06,
"loss": 0.1285,
"step": 586
},
{
"epoch": 1.7580524344569288,
"grad_norm": 13.343894004821777,
"learning_rate": 6.474474474474475e-06,
"loss": 0.1073,
"step": 587
},
{
"epoch": 1.7610486891385768,
"grad_norm": 2.7385926246643066,
"learning_rate": 6.46846846846847e-06,
"loss": 0.0106,
"step": 588
},
{
"epoch": 1.7640449438202248,
"grad_norm": 10.37146282196045,
"learning_rate": 6.462462462462463e-06,
"loss": 0.1138,
"step": 589
},
{
"epoch": 1.7670411985018726,
"grad_norm": 12.535324096679688,
"learning_rate": 6.456456456456457e-06,
"loss": 0.3249,
"step": 590
},
{
"epoch": 1.7700374531835206,
"grad_norm": 16.183120727539062,
"learning_rate": 6.45045045045045e-06,
"loss": 0.146,
"step": 591
},
{
"epoch": 1.7730337078651686,
"grad_norm": 1.8901420831680298,
"learning_rate": 6.444444444444445e-06,
"loss": 0.0093,
"step": 592
},
{
"epoch": 1.7760299625468163,
"grad_norm": 6.514512062072754,
"learning_rate": 6.4384384384384394e-06,
"loss": 0.0634,
"step": 593
},
{
"epoch": 1.7790262172284645,
"grad_norm": 0.9441383481025696,
"learning_rate": 6.432432432432433e-06,
"loss": 0.0057,
"step": 594
},
{
"epoch": 1.7820224719101123,
"grad_norm": 12.353267669677734,
"learning_rate": 6.426426426426427e-06,
"loss": 0.1619,
"step": 595
},
{
"epoch": 1.7850187265917603,
"grad_norm": 43.51881408691406,
"learning_rate": 6.420420420420421e-06,
"loss": 0.0852,
"step": 596
},
{
"epoch": 1.7880149812734083,
"grad_norm": 6.979588985443115,
"learning_rate": 6.414414414414415e-06,
"loss": 0.0682,
"step": 597
},
{
"epoch": 1.791011235955056,
"grad_norm": 89.94390106201172,
"learning_rate": 6.408408408408408e-06,
"loss": 0.0609,
"step": 598
},
{
"epoch": 1.7940074906367043,
"grad_norm": 0.7167533040046692,
"learning_rate": 6.402402402402403e-06,
"loss": 0.0063,
"step": 599
},
{
"epoch": 1.797003745318352,
"grad_norm": 9.951140403747559,
"learning_rate": 6.396396396396397e-06,
"loss": 0.0798,
"step": 600
},
{
"epoch": 1.8,
"grad_norm": 16.502790451049805,
"learning_rate": 6.390390390390391e-06,
"loss": 0.0604,
"step": 601
},
{
"epoch": 1.802996254681648,
"grad_norm": 0.9641748070716858,
"learning_rate": 6.384384384384385e-06,
"loss": 0.0077,
"step": 602
},
{
"epoch": 1.8059925093632958,
"grad_norm": 7.25760555267334,
"learning_rate": 6.378378378378379e-06,
"loss": 0.2486,
"step": 603
},
{
"epoch": 1.8089887640449438,
"grad_norm": 3.220059633255005,
"learning_rate": 6.372372372372373e-06,
"loss": 0.0191,
"step": 604
},
{
"epoch": 1.8119850187265918,
"grad_norm": 16.940893173217773,
"learning_rate": 6.366366366366366e-06,
"loss": 0.3028,
"step": 605
},
{
"epoch": 1.8149812734082396,
"grad_norm": 18.268726348876953,
"learning_rate": 6.360360360360361e-06,
"loss": 0.0792,
"step": 606
},
{
"epoch": 1.8179775280898878,
"grad_norm": 10.651285171508789,
"learning_rate": 6.354354354354355e-06,
"loss": 0.2412,
"step": 607
},
{
"epoch": 1.8209737827715355,
"grad_norm": 6.449036121368408,
"learning_rate": 6.3483483483483485e-06,
"loss": 0.1626,
"step": 608
},
{
"epoch": 1.8239700374531835,
"grad_norm": 8.067935943603516,
"learning_rate": 6.342342342342343e-06,
"loss": 0.2036,
"step": 609
},
{
"epoch": 1.8269662921348315,
"grad_norm": 20.39588165283203,
"learning_rate": 6.336336336336338e-06,
"loss": 0.0415,
"step": 610
},
{
"epoch": 1.8299625468164793,
"grad_norm": 2.1398937702178955,
"learning_rate": 6.330330330330331e-06,
"loss": 0.0161,
"step": 611
},
{
"epoch": 1.8329588014981273,
"grad_norm": 10.125085830688477,
"learning_rate": 6.324324324324325e-06,
"loss": 0.0958,
"step": 612
},
{
"epoch": 1.8359550561797753,
"grad_norm": 6.839081764221191,
"learning_rate": 6.318318318318318e-06,
"loss": 0.033,
"step": 613
},
{
"epoch": 1.8389513108614233,
"grad_norm": 14.015419006347656,
"learning_rate": 6.312312312312313e-06,
"loss": 0.0511,
"step": 614
},
{
"epoch": 1.8419475655430713,
"grad_norm": 1.7917118072509766,
"learning_rate": 6.3063063063063065e-06,
"loss": 0.0129,
"step": 615
},
{
"epoch": 1.844943820224719,
"grad_norm": 42.57280349731445,
"learning_rate": 6.300300300300301e-06,
"loss": 0.5527,
"step": 616
},
{
"epoch": 1.847940074906367,
"grad_norm": 8.393253326416016,
"learning_rate": 6.2942942942942955e-06,
"loss": 0.1261,
"step": 617
},
{
"epoch": 1.850936329588015,
"grad_norm": 11.131349563598633,
"learning_rate": 6.288288288288289e-06,
"loss": 0.1046,
"step": 618
},
{
"epoch": 1.8539325842696628,
"grad_norm": 3.5293877124786377,
"learning_rate": 6.282282282282283e-06,
"loss": 0.019,
"step": 619
},
{
"epoch": 1.856928838951311,
"grad_norm": 17.706586837768555,
"learning_rate": 6.276276276276276e-06,
"loss": 0.1971,
"step": 620
},
{
"epoch": 1.8599250936329588,
"grad_norm": 6.9695820808410645,
"learning_rate": 6.270270270270271e-06,
"loss": 0.0962,
"step": 621
},
{
"epoch": 1.8629213483146068,
"grad_norm": 27.412729263305664,
"learning_rate": 6.264264264264264e-06,
"loss": 0.1069,
"step": 622
},
{
"epoch": 1.8659176029962548,
"grad_norm": 0.4648933708667755,
"learning_rate": 6.2582582582582585e-06,
"loss": 0.0042,
"step": 623
},
{
"epoch": 1.8689138576779025,
"grad_norm": 0.4726819694042206,
"learning_rate": 6.2522522522522535e-06,
"loss": 0.0046,
"step": 624
},
{
"epoch": 1.8719101123595505,
"grad_norm": 1.470893383026123,
"learning_rate": 6.246246246246247e-06,
"loss": 0.0077,
"step": 625
},
{
"epoch": 1.8749063670411985,
"grad_norm": 3.6832358837127686,
"learning_rate": 6.240240240240241e-06,
"loss": 0.0115,
"step": 626
},
{
"epoch": 1.8779026217228463,
"grad_norm": 18.701379776000977,
"learning_rate": 6.234234234234234e-06,
"loss": 0.0779,
"step": 627
},
{
"epoch": 1.8808988764044945,
"grad_norm": 30.573083877563477,
"learning_rate": 6.228228228228229e-06,
"loss": 0.1874,
"step": 628
},
{
"epoch": 1.8838951310861423,
"grad_norm": 9.952974319458008,
"learning_rate": 6.222222222222223e-06,
"loss": 0.0224,
"step": 629
},
{
"epoch": 1.8868913857677903,
"grad_norm": 8.398797035217285,
"learning_rate": 6.2162162162162164e-06,
"loss": 0.023,
"step": 630
},
{
"epoch": 1.8898876404494382,
"grad_norm": 57.172767639160156,
"learning_rate": 6.2102102102102105e-06,
"loss": 0.309,
"step": 631
},
{
"epoch": 1.892883895131086,
"grad_norm": 13.952315330505371,
"learning_rate": 6.204204204204205e-06,
"loss": 0.2085,
"step": 632
},
{
"epoch": 1.8958801498127342,
"grad_norm": 2.267925262451172,
"learning_rate": 6.198198198198199e-06,
"loss": 0.0089,
"step": 633
},
{
"epoch": 1.898876404494382,
"grad_norm": 22.36163902282715,
"learning_rate": 6.192192192192192e-06,
"loss": 0.0828,
"step": 634
},
{
"epoch": 1.90187265917603,
"grad_norm": 27.486753463745117,
"learning_rate": 6.186186186186187e-06,
"loss": 0.2391,
"step": 635
},
{
"epoch": 1.904868913857678,
"grad_norm": 19.787206649780273,
"learning_rate": 6.180180180180181e-06,
"loss": 0.1322,
"step": 636
},
{
"epoch": 1.9078651685393258,
"grad_norm": 25.75490951538086,
"learning_rate": 6.174174174174174e-06,
"loss": 0.3078,
"step": 637
},
{
"epoch": 1.9108614232209737,
"grad_norm": 1.3198350667953491,
"learning_rate": 6.1681681681681685e-06,
"loss": 0.0095,
"step": 638
},
{
"epoch": 1.9138576779026217,
"grad_norm": 22.31900978088379,
"learning_rate": 6.162162162162163e-06,
"loss": 0.0727,
"step": 639
},
{
"epoch": 1.9168539325842695,
"grad_norm": 1.1945544481277466,
"learning_rate": 6.156156156156157e-06,
"loss": 0.0063,
"step": 640
},
{
"epoch": 1.9198501872659177,
"grad_norm": 27.085445404052734,
"learning_rate": 6.150150150150151e-06,
"loss": 0.2401,
"step": 641
},
{
"epoch": 1.9228464419475655,
"grad_norm": 16.98576545715332,
"learning_rate": 6.144144144144145e-06,
"loss": 0.1808,
"step": 642
},
{
"epoch": 1.9258426966292135,
"grad_norm": 49.55059814453125,
"learning_rate": 6.138138138138139e-06,
"loss": 0.0933,
"step": 643
},
{
"epoch": 1.9288389513108615,
"grad_norm": 28.487606048583984,
"learning_rate": 6.132132132132132e-06,
"loss": 0.178,
"step": 644
},
{
"epoch": 1.9318352059925092,
"grad_norm": 20.86649513244629,
"learning_rate": 6.126126126126126e-06,
"loss": 0.264,
"step": 645
},
{
"epoch": 1.9348314606741575,
"grad_norm": 19.417577743530273,
"learning_rate": 6.120120120120121e-06,
"loss": 0.1081,
"step": 646
},
{
"epoch": 1.9378277153558052,
"grad_norm": 53.47108459472656,
"learning_rate": 6.114114114114115e-06,
"loss": 0.1331,
"step": 647
},
{
"epoch": 1.9408239700374532,
"grad_norm": 49.08430480957031,
"learning_rate": 6.108108108108109e-06,
"loss": 0.0763,
"step": 648
},
{
"epoch": 1.9438202247191012,
"grad_norm": 0.42823904752731323,
"learning_rate": 6.102102102102102e-06,
"loss": 0.0037,
"step": 649
},
{
"epoch": 1.946816479400749,
"grad_norm": 18.32347297668457,
"learning_rate": 6.096096096096097e-06,
"loss": 0.1231,
"step": 650
},
{
"epoch": 1.949812734082397,
"grad_norm": 20.45136260986328,
"learning_rate": 6.09009009009009e-06,
"loss": 0.0879,
"step": 651
},
{
"epoch": 1.952808988764045,
"grad_norm": 1.428236722946167,
"learning_rate": 6.084084084084084e-06,
"loss": 0.0037,
"step": 652
},
{
"epoch": 1.9558052434456927,
"grad_norm": 11.381040573120117,
"learning_rate": 6.078078078078079e-06,
"loss": 0.1399,
"step": 653
},
{
"epoch": 1.958801498127341,
"grad_norm": 15.425947189331055,
"learning_rate": 6.0720720720720725e-06,
"loss": 0.1007,
"step": 654
},
{
"epoch": 1.9617977528089887,
"grad_norm": 107.60688018798828,
"learning_rate": 6.066066066066067e-06,
"loss": 0.2099,
"step": 655
},
{
"epoch": 1.9647940074906367,
"grad_norm": 67.12344360351562,
"learning_rate": 6.06006006006006e-06,
"loss": 0.2502,
"step": 656
},
{
"epoch": 1.9677902621722847,
"grad_norm": 28.404937744140625,
"learning_rate": 6.054054054054055e-06,
"loss": 0.0673,
"step": 657
},
{
"epoch": 1.9707865168539325,
"grad_norm": 9.660676002502441,
"learning_rate": 6.048048048048049e-06,
"loss": 0.0241,
"step": 658
},
{
"epoch": 1.9737827715355807,
"grad_norm": 31.48454475402832,
"learning_rate": 6.042042042042042e-06,
"loss": 0.0621,
"step": 659
},
{
"epoch": 1.9767790262172285,
"grad_norm": 1.8083164691925049,
"learning_rate": 6.036036036036037e-06,
"loss": 0.0069,
"step": 660
},
{
"epoch": 1.9797752808988764,
"grad_norm": 0.15951628983020782,
"learning_rate": 6.0300300300300304e-06,
"loss": 0.0022,
"step": 661
},
{
"epoch": 1.9827715355805244,
"grad_norm": 40.35248565673828,
"learning_rate": 6.0240240240240246e-06,
"loss": 0.4294,
"step": 662
},
{
"epoch": 1.9857677902621722,
"grad_norm": 1.4828264713287354,
"learning_rate": 6.018018018018018e-06,
"loss": 0.0224,
"step": 663
},
{
"epoch": 1.9887640449438202,
"grad_norm": 1.8770043849945068,
"learning_rate": 6.012012012012013e-06,
"loss": 0.0061,
"step": 664
},
{
"epoch": 1.9917602996254682,
"grad_norm": 1.222030758857727,
"learning_rate": 6.006006006006007e-06,
"loss": 0.0056,
"step": 665
},
{
"epoch": 1.994756554307116,
"grad_norm": 8.759161949157715,
"learning_rate": 6e-06,
"loss": 0.028,
"step": 666
},
{
"epoch": 1.9977528089887642,
"grad_norm": 7.441140174865723,
"learning_rate": 5.993993993993994e-06,
"loss": 0.1808,
"step": 667
},
{
"epoch": 2.0,
"grad_norm": 0.9321421384811401,
"learning_rate": 5.987987987987988e-06,
"loss": 0.0064,
"step": 668
},
{
"epoch": 2.0,
"eval_accuracy": 0.9680284191829485,
"eval_loss": 0.13619664311408997,
"eval_runtime": 21.2748,
"eval_samples_per_second": 26.463,
"eval_steps_per_second": 3.337,
"step": 668
},
{
"epoch": 2.0029962546816478,
"grad_norm": 0.5909117460250854,
"learning_rate": 5.9819819819819825e-06,
"loss": 0.0045,
"step": 669
},
{
"epoch": 2.005992509363296,
"grad_norm": 2.8122856616973877,
"learning_rate": 5.975975975975976e-06,
"loss": 0.0159,
"step": 670
},
{
"epoch": 2.0089887640449438,
"grad_norm": 6.904968738555908,
"learning_rate": 5.969969969969971e-06,
"loss": 0.1124,
"step": 671
},
{
"epoch": 2.011985018726592,
"grad_norm": 2.411268472671509,
"learning_rate": 5.963963963963965e-06,
"loss": 0.0125,
"step": 672
},
{
"epoch": 2.0149812734082397,
"grad_norm": 5.258452415466309,
"learning_rate": 5.957957957957958e-06,
"loss": 0.0101,
"step": 673
},
{
"epoch": 2.0179775280898875,
"grad_norm": 1.4426406621932983,
"learning_rate": 5.951951951951952e-06,
"loss": 0.0046,
"step": 674
},
{
"epoch": 2.0209737827715357,
"grad_norm": 20.04401969909668,
"learning_rate": 5.945945945945947e-06,
"loss": 0.1996,
"step": 675
},
{
"epoch": 2.0239700374531835,
"grad_norm": 14.030840873718262,
"learning_rate": 5.93993993993994e-06,
"loss": 0.136,
"step": 676
},
{
"epoch": 2.0269662921348313,
"grad_norm": 28.96600914001465,
"learning_rate": 5.9339339339339345e-06,
"loss": 0.5518,
"step": 677
},
{
"epoch": 2.0299625468164795,
"grad_norm": 0.33994242548942566,
"learning_rate": 5.927927927927929e-06,
"loss": 0.0029,
"step": 678
},
{
"epoch": 2.0329588014981272,
"grad_norm": 27.4215030670166,
"learning_rate": 5.921921921921923e-06,
"loss": 0.0131,
"step": 679
},
{
"epoch": 2.0359550561797755,
"grad_norm": 0.48077064752578735,
"learning_rate": 5.915915915915916e-06,
"loss": 0.0027,
"step": 680
},
{
"epoch": 2.0389513108614232,
"grad_norm": 23.7380428314209,
"learning_rate": 5.90990990990991e-06,
"loss": 0.097,
"step": 681
},
{
"epoch": 2.041947565543071,
"grad_norm": 9.703587532043457,
"learning_rate": 5.903903903903905e-06,
"loss": 0.0113,
"step": 682
},
{
"epoch": 2.044943820224719,
"grad_norm": 17.765623092651367,
"learning_rate": 5.897897897897898e-06,
"loss": 0.0674,
"step": 683
},
{
"epoch": 2.047940074906367,
"grad_norm": 9.970433235168457,
"learning_rate": 5.8918918918918924e-06,
"loss": 0.1364,
"step": 684
},
{
"epoch": 2.050936329588015,
"grad_norm": 6.063045978546143,
"learning_rate": 5.885885885885886e-06,
"loss": 0.0695,
"step": 685
},
{
"epoch": 2.053932584269663,
"grad_norm": 0.3433529734611511,
"learning_rate": 5.879879879879881e-06,
"loss": 0.003,
"step": 686
},
{
"epoch": 2.0569288389513107,
"grad_norm": 16.1558780670166,
"learning_rate": 5.873873873873874e-06,
"loss": 0.14,
"step": 687
},
{
"epoch": 2.059925093632959,
"grad_norm": 9.964441299438477,
"learning_rate": 5.867867867867868e-06,
"loss": 0.3482,
"step": 688
},
{
"epoch": 2.0629213483146067,
"grad_norm": 53.629356384277344,
"learning_rate": 5.861861861861863e-06,
"loss": 0.0792,
"step": 689
},
{
"epoch": 2.0659176029962545,
"grad_norm": 9.418923377990723,
"learning_rate": 5.855855855855856e-06,
"loss": 0.1174,
"step": 690
},
{
"epoch": 2.0689138576779027,
"grad_norm": 0.4195489287376404,
"learning_rate": 5.84984984984985e-06,
"loss": 0.0039,
"step": 691
},
{
"epoch": 2.0719101123595505,
"grad_norm": 0.2967524528503418,
"learning_rate": 5.843843843843844e-06,
"loss": 0.0028,
"step": 692
},
{
"epoch": 2.0749063670411987,
"grad_norm": 12.891443252563477,
"learning_rate": 5.837837837837839e-06,
"loss": 0.1519,
"step": 693
},
{
"epoch": 2.0779026217228465,
"grad_norm": 1.4585908651351929,
"learning_rate": 5.831831831831833e-06,
"loss": 0.0041,
"step": 694
},
{
"epoch": 2.0808988764044942,
"grad_norm": 0.25973114371299744,
"learning_rate": 5.825825825825826e-06,
"loss": 0.002,
"step": 695
},
{
"epoch": 2.0838951310861424,
"grad_norm": 0.362197607755661,
"learning_rate": 5.819819819819821e-06,
"loss": 0.0033,
"step": 696
},
{
"epoch": 2.08689138576779,
"grad_norm": 12.347769737243652,
"learning_rate": 5.813813813813814e-06,
"loss": 0.1076,
"step": 697
},
{
"epoch": 2.0898876404494384,
"grad_norm": 8.291030883789062,
"learning_rate": 5.807807807807808e-06,
"loss": 0.1345,
"step": 698
},
{
"epoch": 2.092883895131086,
"grad_norm": 7.131426811218262,
"learning_rate": 5.8018018018018015e-06,
"loss": 0.1503,
"step": 699
},
{
"epoch": 2.095880149812734,
"grad_norm": 53.37166213989258,
"learning_rate": 5.7957957957957965e-06,
"loss": 0.1402,
"step": 700
},
{
"epoch": 2.098876404494382,
"grad_norm": 33.311214447021484,
"learning_rate": 5.789789789789791e-06,
"loss": 0.2643,
"step": 701
},
{
"epoch": 2.10187265917603,
"grad_norm": 21.769689559936523,
"learning_rate": 5.783783783783784e-06,
"loss": 0.0221,
"step": 702
},
{
"epoch": 2.1048689138576777,
"grad_norm": 0.9422563314437866,
"learning_rate": 5.777777777777778e-06,
"loss": 0.0071,
"step": 703
},
{
"epoch": 2.107865168539326,
"grad_norm": 114.99813079833984,
"learning_rate": 5.771771771771772e-06,
"loss": 0.0244,
"step": 704
},
{
"epoch": 2.1108614232209737,
"grad_norm": 7.310803413391113,
"learning_rate": 5.765765765765766e-06,
"loss": 0.1388,
"step": 705
},
{
"epoch": 2.113857677902622,
"grad_norm": 0.3826982378959656,
"learning_rate": 5.7597597597597595e-06,
"loss": 0.0031,
"step": 706
},
{
"epoch": 2.1168539325842697,
"grad_norm": 18.026264190673828,
"learning_rate": 5.7537537537537544e-06,
"loss": 0.106,
"step": 707
},
{
"epoch": 2.1198501872659175,
"grad_norm": 3.6052420139312744,
"learning_rate": 5.7477477477477485e-06,
"loss": 0.0363,
"step": 708
},
{
"epoch": 2.1228464419475657,
"grad_norm": 24.524499893188477,
"learning_rate": 5.741741741741742e-06,
"loss": 0.1066,
"step": 709
},
{
"epoch": 2.1258426966292134,
"grad_norm": 1.308187484741211,
"learning_rate": 5.735735735735736e-06,
"loss": 0.0179,
"step": 710
},
{
"epoch": 2.128838951310861,
"grad_norm": 29.05257797241211,
"learning_rate": 5.729729729729731e-06,
"loss": 0.2827,
"step": 711
},
{
"epoch": 2.1318352059925094,
"grad_norm": 7.687588691711426,
"learning_rate": 5.723723723723724e-06,
"loss": 0.0988,
"step": 712
},
{
"epoch": 2.134831460674157,
"grad_norm": 0.3173719346523285,
"learning_rate": 5.717717717717718e-06,
"loss": 0.0026,
"step": 713
},
{
"epoch": 2.1378277153558054,
"grad_norm": 1.266904592514038,
"learning_rate": 5.711711711711712e-06,
"loss": 0.0063,
"step": 714
},
{
"epoch": 2.140823970037453,
"grad_norm": 6.052237510681152,
"learning_rate": 5.7057057057057065e-06,
"loss": 0.1987,
"step": 715
},
{
"epoch": 2.143820224719101,
"grad_norm": 1.3977775573730469,
"learning_rate": 5.6996996996997e-06,
"loss": 0.0103,
"step": 716
},
{
"epoch": 2.146816479400749,
"grad_norm": 9.335174560546875,
"learning_rate": 5.693693693693694e-06,
"loss": 0.1433,
"step": 717
},
{
"epoch": 2.149812734082397,
"grad_norm": 29.303775787353516,
"learning_rate": 5.687687687687689e-06,
"loss": 0.1656,
"step": 718
},
{
"epoch": 2.152808988764045,
"grad_norm": 9.021591186523438,
"learning_rate": 5.681681681681682e-06,
"loss": 0.0573,
"step": 719
},
{
"epoch": 2.155805243445693,
"grad_norm": 44.36116409301758,
"learning_rate": 5.675675675675676e-06,
"loss": 0.1947,
"step": 720
},
{
"epoch": 2.1588014981273407,
"grad_norm": 9.322854995727539,
"learning_rate": 5.6696696696696694e-06,
"loss": 0.0363,
"step": 721
},
{
"epoch": 2.161797752808989,
"grad_norm": 49.801944732666016,
"learning_rate": 5.663663663663664e-06,
"loss": 0.0502,
"step": 722
},
{
"epoch": 2.1647940074906367,
"grad_norm": 3.590524435043335,
"learning_rate": 5.657657657657658e-06,
"loss": 0.0074,
"step": 723
},
{
"epoch": 2.167790262172285,
"grad_norm": 0.7181494235992432,
"learning_rate": 5.651651651651652e-06,
"loss": 0.0034,
"step": 724
},
{
"epoch": 2.1707865168539326,
"grad_norm": 3.66660213470459,
"learning_rate": 5.645645645645647e-06,
"loss": 0.0144,
"step": 725
},
{
"epoch": 2.1737827715355804,
"grad_norm": 1.0453420877456665,
"learning_rate": 5.63963963963964e-06,
"loss": 0.0047,
"step": 726
},
{
"epoch": 2.1767790262172286,
"grad_norm": 22.018798828125,
"learning_rate": 5.633633633633634e-06,
"loss": 0.0382,
"step": 727
},
{
"epoch": 2.1797752808988764,
"grad_norm": 0.10004288703203201,
"learning_rate": 5.627627627627627e-06,
"loss": 0.0015,
"step": 728
},
{
"epoch": 2.182771535580524,
"grad_norm": 0.48340409994125366,
"learning_rate": 5.621621621621622e-06,
"loss": 0.0027,
"step": 729
},
{
"epoch": 2.1857677902621724,
"grad_norm": 0.16499565541744232,
"learning_rate": 5.615615615615616e-06,
"loss": 0.0019,
"step": 730
},
{
"epoch": 2.18876404494382,
"grad_norm": 13.637986183166504,
"learning_rate": 5.60960960960961e-06,
"loss": 0.1302,
"step": 731
},
{
"epoch": 2.1917602996254684,
"grad_norm": 0.1634170562028885,
"learning_rate": 5.603603603603605e-06,
"loss": 0.0018,
"step": 732
},
{
"epoch": 2.194756554307116,
"grad_norm": 0.20700804889202118,
"learning_rate": 5.597597597597598e-06,
"loss": 0.0019,
"step": 733
},
{
"epoch": 2.197752808988764,
"grad_norm": 8.509710311889648,
"learning_rate": 5.591591591591592e-06,
"loss": 0.0384,
"step": 734
},
{
"epoch": 2.200749063670412,
"grad_norm": 53.0845947265625,
"learning_rate": 5.585585585585585e-06,
"loss": 0.3973,
"step": 735
},
{
"epoch": 2.20374531835206,
"grad_norm": 2.7898035049438477,
"learning_rate": 5.57957957957958e-06,
"loss": 0.0293,
"step": 736
},
{
"epoch": 2.2067415730337077,
"grad_norm": 2.3603007793426514,
"learning_rate": 5.573573573573574e-06,
"loss": 0.0074,
"step": 737
},
{
"epoch": 2.209737827715356,
"grad_norm": 1.4276190996170044,
"learning_rate": 5.567567567567568e-06,
"loss": 0.0041,
"step": 738
},
{
"epoch": 2.2127340823970036,
"grad_norm": 11.371862411499023,
"learning_rate": 5.561561561561562e-06,
"loss": 0.1776,
"step": 739
},
{
"epoch": 2.215730337078652,
"grad_norm": 2.8665268421173096,
"learning_rate": 5.555555555555557e-06,
"loss": 0.2261,
"step": 740
},
{
"epoch": 2.2187265917602996,
"grad_norm": 10.104736328125,
"learning_rate": 5.54954954954955e-06,
"loss": 0.1393,
"step": 741
},
{
"epoch": 2.2217228464419474,
"grad_norm": 15.026305198669434,
"learning_rate": 5.543543543543544e-06,
"loss": 0.0207,
"step": 742
},
{
"epoch": 2.2247191011235956,
"grad_norm": 0.11509612202644348,
"learning_rate": 5.537537537537538e-06,
"loss": 0.0012,
"step": 743
},
{
"epoch": 2.2277153558052434,
"grad_norm": 1.4837394952774048,
"learning_rate": 5.531531531531532e-06,
"loss": 0.0027,
"step": 744
},
{
"epoch": 2.2307116104868916,
"grad_norm": 36.25615692138672,
"learning_rate": 5.5255255255255255e-06,
"loss": 0.123,
"step": 745
},
{
"epoch": 2.2337078651685394,
"grad_norm": 22.573949813842773,
"learning_rate": 5.51951951951952e-06,
"loss": 0.1777,
"step": 746
},
{
"epoch": 2.236704119850187,
"grad_norm": 18.751577377319336,
"learning_rate": 5.513513513513515e-06,
"loss": 0.0642,
"step": 747
},
{
"epoch": 2.2397003745318353,
"grad_norm": 0.2828042507171631,
"learning_rate": 5.507507507507508e-06,
"loss": 0.0018,
"step": 748
},
{
"epoch": 2.242696629213483,
"grad_norm": 0.3677152991294861,
"learning_rate": 5.501501501501502e-06,
"loss": 0.0018,
"step": 749
},
{
"epoch": 2.245692883895131,
"grad_norm": 0.7993999719619751,
"learning_rate": 5.495495495495496e-06,
"loss": 0.0029,
"step": 750
},
{
"epoch": 2.248689138576779,
"grad_norm": 0.25123536586761475,
"learning_rate": 5.48948948948949e-06,
"loss": 0.0019,
"step": 751
},
{
"epoch": 2.251685393258427,
"grad_norm": 27.455381393432617,
"learning_rate": 5.4834834834834834e-06,
"loss": 0.2809,
"step": 752
},
{
"epoch": 2.254681647940075,
"grad_norm": 18.18854331970215,
"learning_rate": 5.4774774774774776e-06,
"loss": 0.0799,
"step": 753
},
{
"epoch": 2.257677902621723,
"grad_norm": 11.329863548278809,
"learning_rate": 5.4714714714714725e-06,
"loss": 0.3422,
"step": 754
},
{
"epoch": 2.2606741573033706,
"grad_norm": 0.5013195872306824,
"learning_rate": 5.465465465465466e-06,
"loss": 0.0024,
"step": 755
},
{
"epoch": 2.263670411985019,
"grad_norm": 0.2460617870092392,
"learning_rate": 5.45945945945946e-06,
"loss": 0.0017,
"step": 756
},
{
"epoch": 2.2666666666666666,
"grad_norm": 11.609503746032715,
"learning_rate": 5.453453453453455e-06,
"loss": 0.2084,
"step": 757
},
{
"epoch": 2.2696629213483144,
"grad_norm": 13.621426582336426,
"learning_rate": 5.447447447447448e-06,
"loss": 0.1697,
"step": 758
},
{
"epoch": 2.2726591760299626,
"grad_norm": 54.862449645996094,
"learning_rate": 5.441441441441442e-06,
"loss": 0.2744,
"step": 759
},
{
"epoch": 2.2756554307116104,
"grad_norm": 20.107084274291992,
"learning_rate": 5.4354354354354355e-06,
"loss": 0.0514,
"step": 760
},
{
"epoch": 2.2786516853932586,
"grad_norm": 19.26397705078125,
"learning_rate": 5.4294294294294304e-06,
"loss": 0.115,
"step": 761
},
{
"epoch": 2.2816479400749063,
"grad_norm": 10.051045417785645,
"learning_rate": 5.423423423423424e-06,
"loss": 0.0323,
"step": 762
},
{
"epoch": 2.284644194756554,
"grad_norm": 21.69451904296875,
"learning_rate": 5.417417417417418e-06,
"loss": 0.1779,
"step": 763
},
{
"epoch": 2.2876404494382023,
"grad_norm": 4.480892181396484,
"learning_rate": 5.411411411411411e-06,
"loss": 0.0154,
"step": 764
},
{
"epoch": 2.29063670411985,
"grad_norm": 32.57766342163086,
"learning_rate": 5.405405405405406e-06,
"loss": 0.1731,
"step": 765
},
{
"epoch": 2.2936329588014983,
"grad_norm": 4.176045894622803,
"learning_rate": 5.3993993993994e-06,
"loss": 0.0297,
"step": 766
},
{
"epoch": 2.296629213483146,
"grad_norm": 0.535074770450592,
"learning_rate": 5.393393393393393e-06,
"loss": 0.002,
"step": 767
},
{
"epoch": 2.299625468164794,
"grad_norm": 2.382888078689575,
"learning_rate": 5.387387387387388e-06,
"loss": 0.0125,
"step": 768
},
{
"epoch": 2.302621722846442,
"grad_norm": 3.4220054149627686,
"learning_rate": 5.381381381381382e-06,
"loss": 0.0209,
"step": 769
},
{
"epoch": 2.30561797752809,
"grad_norm": 0.531780481338501,
"learning_rate": 5.375375375375376e-06,
"loss": 0.0029,
"step": 770
},
{
"epoch": 2.308614232209738,
"grad_norm": 1.319352149963379,
"learning_rate": 5.369369369369369e-06,
"loss": 0.0051,
"step": 771
},
{
"epoch": 2.311610486891386,
"grad_norm": 15.450730323791504,
"learning_rate": 5.363363363363364e-06,
"loss": 0.0317,
"step": 772
},
{
"epoch": 2.3146067415730336,
"grad_norm": 0.8007951974868774,
"learning_rate": 5.357357357357358e-06,
"loss": 0.0087,
"step": 773
},
{
"epoch": 2.317602996254682,
"grad_norm": 11.054052352905273,
"learning_rate": 5.351351351351351e-06,
"loss": 0.1622,
"step": 774
},
{
"epoch": 2.3205992509363296,
"grad_norm": 12.34724235534668,
"learning_rate": 5.345345345345346e-06,
"loss": 0.1531,
"step": 775
},
{
"epoch": 2.3235955056179773,
"grad_norm": 5.4940714836120605,
"learning_rate": 5.33933933933934e-06,
"loss": 0.0071,
"step": 776
},
{
"epoch": 2.3265917602996256,
"grad_norm": 0.43366900086402893,
"learning_rate": 5.333333333333334e-06,
"loss": 0.0024,
"step": 777
},
{
"epoch": 2.3295880149812733,
"grad_norm": 0.31749555468559265,
"learning_rate": 5.327327327327328e-06,
"loss": 0.003,
"step": 778
},
{
"epoch": 2.3325842696629215,
"grad_norm": 41.52970504760742,
"learning_rate": 5.321321321321322e-06,
"loss": 0.1501,
"step": 779
},
{
"epoch": 2.3355805243445693,
"grad_norm": 56.170494079589844,
"learning_rate": 5.315315315315316e-06,
"loss": 0.2017,
"step": 780
},
{
"epoch": 2.338576779026217,
"grad_norm": 27.163827896118164,
"learning_rate": 5.309309309309309e-06,
"loss": 0.4548,
"step": 781
},
{
"epoch": 2.3415730337078653,
"grad_norm": 0.11042162775993347,
"learning_rate": 5.303303303303303e-06,
"loss": 0.0012,
"step": 782
},
{
"epoch": 2.344569288389513,
"grad_norm": 8.128288269042969,
"learning_rate": 5.297297297297298e-06,
"loss": 0.1435,
"step": 783
},
{
"epoch": 2.347565543071161,
"grad_norm": 18.697711944580078,
"learning_rate": 5.291291291291292e-06,
"loss": 0.0559,
"step": 784
},
{
"epoch": 2.350561797752809,
"grad_norm": 11.155041694641113,
"learning_rate": 5.285285285285286e-06,
"loss": 0.1482,
"step": 785
},
{
"epoch": 2.353558052434457,
"grad_norm": 11.597505569458008,
"learning_rate": 5.27927927927928e-06,
"loss": 0.1998,
"step": 786
},
{
"epoch": 2.356554307116105,
"grad_norm": 0.6385789513587952,
"learning_rate": 5.273273273273274e-06,
"loss": 0.0036,
"step": 787
},
{
"epoch": 2.359550561797753,
"grad_norm": 14.634647369384766,
"learning_rate": 5.267267267267267e-06,
"loss": 0.0066,
"step": 788
},
{
"epoch": 2.3625468164794006,
"grad_norm": 3.1089072227478027,
"learning_rate": 5.261261261261261e-06,
"loss": 0.0051,
"step": 789
},
{
"epoch": 2.365543071161049,
"grad_norm": 10.165287971496582,
"learning_rate": 5.255255255255256e-06,
"loss": 0.1874,
"step": 790
},
{
"epoch": 2.3685393258426966,
"grad_norm": 18.6910400390625,
"learning_rate": 5.2492492492492495e-06,
"loss": 0.0269,
"step": 791
},
{
"epoch": 2.3715355805243448,
"grad_norm": 15.441289901733398,
"learning_rate": 5.243243243243244e-06,
"loss": 0.1545,
"step": 792
},
{
"epoch": 2.3745318352059925,
"grad_norm": 1.1983261108398438,
"learning_rate": 5.2372372372372386e-06,
"loss": 0.0066,
"step": 793
},
{
"epoch": 2.3775280898876403,
"grad_norm": 9.271286010742188,
"learning_rate": 5.231231231231232e-06,
"loss": 0.0771,
"step": 794
},
{
"epoch": 2.3805243445692885,
"grad_norm": 0.5765110850334167,
"learning_rate": 5.225225225225226e-06,
"loss": 0.0033,
"step": 795
},
{
"epoch": 2.3835205992509363,
"grad_norm": 22.070837020874023,
"learning_rate": 5.219219219219219e-06,
"loss": 0.2031,
"step": 796
},
{
"epoch": 2.3865168539325845,
"grad_norm": 0.5393401980400085,
"learning_rate": 5.213213213213214e-06,
"loss": 0.0034,
"step": 797
},
{
"epoch": 2.3895131086142323,
"grad_norm": 1.5076879262924194,
"learning_rate": 5.2072072072072074e-06,
"loss": 0.0075,
"step": 798
},
{
"epoch": 2.39250936329588,
"grad_norm": 36.4506950378418,
"learning_rate": 5.2012012012012015e-06,
"loss": 0.0209,
"step": 799
},
{
"epoch": 2.3955056179775283,
"grad_norm": 47.675262451171875,
"learning_rate": 5.195195195195195e-06,
"loss": 0.1072,
"step": 800
},
{
"epoch": 2.398501872659176,
"grad_norm": 0.2616981565952301,
"learning_rate": 5.18918918918919e-06,
"loss": 0.0016,
"step": 801
},
{
"epoch": 2.401498127340824,
"grad_norm": 0.1060936376452446,
"learning_rate": 5.183183183183184e-06,
"loss": 0.0013,
"step": 802
},
{
"epoch": 2.404494382022472,
"grad_norm": 0.11718081682920456,
"learning_rate": 5.177177177177177e-06,
"loss": 0.0014,
"step": 803
},
{
"epoch": 2.40749063670412,
"grad_norm": 3.4686341285705566,
"learning_rate": 5.171171171171172e-06,
"loss": 0.0075,
"step": 804
},
{
"epoch": 2.4104868913857675,
"grad_norm": 26.21915054321289,
"learning_rate": 5.165165165165165e-06,
"loss": 0.0425,
"step": 805
},
{
"epoch": 2.4134831460674158,
"grad_norm": 1.9707993268966675,
"learning_rate": 5.1591591591591595e-06,
"loss": 0.0034,
"step": 806
},
{
"epoch": 2.4164794007490635,
"grad_norm": 0.2612178921699524,
"learning_rate": 5.153153153153153e-06,
"loss": 0.0021,
"step": 807
},
{
"epoch": 2.4194756554307117,
"grad_norm": 3.325965642929077,
"learning_rate": 5.147147147147148e-06,
"loss": 0.2177,
"step": 808
},
{
"epoch": 2.4224719101123595,
"grad_norm": 0.3344706892967224,
"learning_rate": 5.141141141141142e-06,
"loss": 0.0018,
"step": 809
},
{
"epoch": 2.4254681647940073,
"grad_norm": 3.4124205112457275,
"learning_rate": 5.135135135135135e-06,
"loss": 0.0077,
"step": 810
},
{
"epoch": 2.4284644194756555,
"grad_norm": 0.11304173618555069,
"learning_rate": 5.12912912912913e-06,
"loss": 0.0013,
"step": 811
},
{
"epoch": 2.4314606741573033,
"grad_norm": 20.09816551208496,
"learning_rate": 5.123123123123124e-06,
"loss": 0.1898,
"step": 812
},
{
"epoch": 2.4344569288389515,
"grad_norm": 20.79599380493164,
"learning_rate": 5.117117117117117e-06,
"loss": 0.1413,
"step": 813
},
{
"epoch": 2.4374531835205993,
"grad_norm": 38.39596939086914,
"learning_rate": 5.1111111111111115e-06,
"loss": 0.4453,
"step": 814
},
{
"epoch": 2.440449438202247,
"grad_norm": 3.7794981002807617,
"learning_rate": 5.105105105105106e-06,
"loss": 0.2574,
"step": 815
},
{
"epoch": 2.4434456928838952,
"grad_norm": 9.27757453918457,
"learning_rate": 5.0990990990991e-06,
"loss": 0.3929,
"step": 816
},
{
"epoch": 2.446441947565543,
"grad_norm": 11.782530784606934,
"learning_rate": 5.093093093093093e-06,
"loss": 0.2448,
"step": 817
},
{
"epoch": 2.449438202247191,
"grad_norm": 21.72043800354004,
"learning_rate": 5.087087087087087e-06,
"loss": 0.0636,
"step": 818
},
{
"epoch": 2.452434456928839,
"grad_norm": 0.24695485830307007,
"learning_rate": 5.081081081081082e-06,
"loss": 0.0023,
"step": 819
},
{
"epoch": 2.4554307116104868,
"grad_norm": 0.15776212513446808,
"learning_rate": 5.075075075075075e-06,
"loss": 0.0016,
"step": 820
},
{
"epoch": 2.458426966292135,
"grad_norm": 0.12474209070205688,
"learning_rate": 5.069069069069069e-06,
"loss": 0.0014,
"step": 821
},
{
"epoch": 2.4614232209737827,
"grad_norm": 0.823479950428009,
"learning_rate": 5.0630630630630635e-06,
"loss": 0.0033,
"step": 822
},
{
"epoch": 2.464419475655431,
"grad_norm": 1.9294389486312866,
"learning_rate": 5.057057057057058e-06,
"loss": 0.0051,
"step": 823
},
{
"epoch": 2.4674157303370787,
"grad_norm": 9.133439064025879,
"learning_rate": 5.051051051051051e-06,
"loss": 0.1372,
"step": 824
},
{
"epoch": 2.4704119850187265,
"grad_norm": 0.1794380247592926,
"learning_rate": 5.045045045045045e-06,
"loss": 0.0018,
"step": 825
},
{
"epoch": 2.4734082397003747,
"grad_norm": 40.159278869628906,
"learning_rate": 5.03903903903904e-06,
"loss": 0.0716,
"step": 826
},
{
"epoch": 2.4764044943820225,
"grad_norm": 0.2856251895427704,
"learning_rate": 5.033033033033033e-06,
"loss": 0.0021,
"step": 827
},
{
"epoch": 2.4794007490636703,
"grad_norm": 29.261083602905273,
"learning_rate": 5.027027027027027e-06,
"loss": 0.3959,
"step": 828
},
{
"epoch": 2.4823970037453185,
"grad_norm": 15.03356647491455,
"learning_rate": 5.021021021021022e-06,
"loss": 0.3091,
"step": 829
},
{
"epoch": 2.4853932584269662,
"grad_norm": 0.49418002367019653,
"learning_rate": 5.0150150150150156e-06,
"loss": 0.0027,
"step": 830
},
{
"epoch": 2.488389513108614,
"grad_norm": 0.5899301171302795,
"learning_rate": 5.00900900900901e-06,
"loss": 0.0027,
"step": 831
},
{
"epoch": 2.491385767790262,
"grad_norm": 17.68525505065918,
"learning_rate": 5.003003003003003e-06,
"loss": 0.1421,
"step": 832
},
{
"epoch": 2.49438202247191,
"grad_norm": 0.6795111298561096,
"learning_rate": 4.996996996996997e-06,
"loss": 0.0042,
"step": 833
},
{
"epoch": 2.497378277153558,
"grad_norm": 11.507546424865723,
"learning_rate": 4.990990990990991e-06,
"loss": 0.1846,
"step": 834
},
{
"epoch": 2.500374531835206,
"grad_norm": 27.412553787231445,
"learning_rate": 4.984984984984985e-06,
"loss": 0.0149,
"step": 835
},
{
"epoch": 2.5033707865168537,
"grad_norm": 13.921358108520508,
"learning_rate": 4.978978978978979e-06,
"loss": 0.0957,
"step": 836
},
{
"epoch": 2.506367041198502,
"grad_norm": 0.35653361678123474,
"learning_rate": 4.9729729729729735e-06,
"loss": 0.0023,
"step": 837
},
{
"epoch": 2.5093632958801497,
"grad_norm": 0.6814135909080505,
"learning_rate": 4.966966966966968e-06,
"loss": 0.0038,
"step": 838
},
{
"epoch": 2.512359550561798,
"grad_norm": 0.20339810848236084,
"learning_rate": 4.960960960960962e-06,
"loss": 0.002,
"step": 839
},
{
"epoch": 2.5153558052434457,
"grad_norm": 9.292779922485352,
"learning_rate": 4.954954954954955e-06,
"loss": 0.0212,
"step": 840
},
{
"epoch": 2.5183520599250935,
"grad_norm": 55.11958694458008,
"learning_rate": 4.94894894894895e-06,
"loss": 0.2677,
"step": 841
},
{
"epoch": 2.5213483146067417,
"grad_norm": 0.5604162812232971,
"learning_rate": 4.942942942942943e-06,
"loss": 0.0026,
"step": 842
},
{
"epoch": 2.5243445692883895,
"grad_norm": 43.203163146972656,
"learning_rate": 4.936936936936937e-06,
"loss": 0.0814,
"step": 843
},
{
"epoch": 2.5273408239700377,
"grad_norm": 7.531084060668945,
"learning_rate": 4.930930930930931e-06,
"loss": 0.0413,
"step": 844
},
{
"epoch": 2.5303370786516854,
"grad_norm": 6.6757612228393555,
"learning_rate": 4.9249249249249255e-06,
"loss": 0.2019,
"step": 845
},
{
"epoch": 2.533333333333333,
"grad_norm": 23.998159408569336,
"learning_rate": 4.91891891891892e-06,
"loss": 0.6601,
"step": 846
},
{
"epoch": 2.5363295880149814,
"grad_norm": 0.18272723257541656,
"learning_rate": 4.912912912912913e-06,
"loss": 0.0018,
"step": 847
},
{
"epoch": 2.539325842696629,
"grad_norm": 0.9130871295928955,
"learning_rate": 4.906906906906908e-06,
"loss": 0.0035,
"step": 848
},
{
"epoch": 2.5423220973782774,
"grad_norm": 0.418591171503067,
"learning_rate": 4.900900900900901e-06,
"loss": 0.0017,
"step": 849
},
{
"epoch": 2.545318352059925,
"grad_norm": 40.452972412109375,
"learning_rate": 4.894894894894895e-06,
"loss": 0.0702,
"step": 850
},
{
"epoch": 2.548314606741573,
"grad_norm": 2.587827682495117,
"learning_rate": 4.888888888888889e-06,
"loss": 0.0032,
"step": 851
},
{
"epoch": 2.5513108614232207,
"grad_norm": 20.748512268066406,
"learning_rate": 4.8828828828828834e-06,
"loss": 0.2098,
"step": 852
},
{
"epoch": 2.554307116104869,
"grad_norm": 0.8581587076187134,
"learning_rate": 4.876876876876877e-06,
"loss": 0.0045,
"step": 853
},
{
"epoch": 2.5573033707865167,
"grad_norm": 9.334095001220703,
"learning_rate": 4.870870870870872e-06,
"loss": 0.1712,
"step": 854
},
{
"epoch": 2.560299625468165,
"grad_norm": 0.5096704363822937,
"learning_rate": 4.864864864864866e-06,
"loss": 0.0035,
"step": 855
},
{
"epoch": 2.5632958801498127,
"grad_norm": 0.1468096524477005,
"learning_rate": 4.858858858858859e-06,
"loss": 0.0014,
"step": 856
},
{
"epoch": 2.5662921348314605,
"grad_norm": 2.789402723312378,
"learning_rate": 4.852852852852853e-06,
"loss": 0.0046,
"step": 857
},
{
"epoch": 2.5692883895131087,
"grad_norm": 6.312567234039307,
"learning_rate": 4.846846846846847e-06,
"loss": 0.1153,
"step": 858
},
{
"epoch": 2.5722846441947564,
"grad_norm": 46.265533447265625,
"learning_rate": 4.840840840840841e-06,
"loss": 0.1256,
"step": 859
},
{
"epoch": 2.5752808988764047,
"grad_norm": 3.9658403396606445,
"learning_rate": 4.8348348348348355e-06,
"loss": 0.0808,
"step": 860
},
{
"epoch": 2.5782771535580524,
"grad_norm": 0.1250729262828827,
"learning_rate": 4.82882882882883e-06,
"loss": 0.0012,
"step": 861
},
{
"epoch": 2.5812734082397,
"grad_norm": 123.43838500976562,
"learning_rate": 4.822822822822823e-06,
"loss": 0.3392,
"step": 862
},
{
"epoch": 2.5842696629213484,
"grad_norm": 0.9565964937210083,
"learning_rate": 4.816816816816817e-06,
"loss": 0.0029,
"step": 863
},
{
"epoch": 2.587265917602996,
"grad_norm": 0.3829185962677002,
"learning_rate": 4.810810810810811e-06,
"loss": 0.0029,
"step": 864
},
{
"epoch": 2.5902621722846444,
"grad_norm": 0.17929819226264954,
"learning_rate": 4.804804804804805e-06,
"loss": 0.0014,
"step": 865
},
{
"epoch": 2.593258426966292,
"grad_norm": 9.648361206054688,
"learning_rate": 4.798798798798799e-06,
"loss": 0.1846,
"step": 866
},
{
"epoch": 2.59625468164794,
"grad_norm": 0.1267472803592682,
"learning_rate": 4.792792792792793e-06,
"loss": 0.0014,
"step": 867
},
{
"epoch": 2.599250936329588,
"grad_norm": 25.92026710510254,
"learning_rate": 4.7867867867867875e-06,
"loss": 0.3681,
"step": 868
},
{
"epoch": 2.602247191011236,
"grad_norm": 0.7075502872467041,
"learning_rate": 4.780780780780781e-06,
"loss": 0.0049,
"step": 869
},
{
"epoch": 2.605243445692884,
"grad_norm": 11.952223777770996,
"learning_rate": 4.774774774774775e-06,
"loss": 0.0262,
"step": 870
},
{
"epoch": 2.608239700374532,
"grad_norm": 0.10766910016536713,
"learning_rate": 4.768768768768769e-06,
"loss": 0.0012,
"step": 871
},
{
"epoch": 2.6112359550561797,
"grad_norm": 0.3023926615715027,
"learning_rate": 4.762762762762763e-06,
"loss": 0.0015,
"step": 872
},
{
"epoch": 2.6142322097378274,
"grad_norm": 0.7597098350524902,
"learning_rate": 4.756756756756757e-06,
"loss": 0.0043,
"step": 873
},
{
"epoch": 2.6172284644194757,
"grad_norm": 1.2177233695983887,
"learning_rate": 4.750750750750751e-06,
"loss": 0.0032,
"step": 874
},
{
"epoch": 2.620224719101124,
"grad_norm": 25.467269897460938,
"learning_rate": 4.7447447447447454e-06,
"loss": 0.1228,
"step": 875
},
{
"epoch": 2.6232209737827716,
"grad_norm": 442.0234375,
"learning_rate": 4.738738738738739e-06,
"loss": 0.0861,
"step": 876
},
{
"epoch": 2.6262172284644194,
"grad_norm": 26.847919464111328,
"learning_rate": 4.732732732732734e-06,
"loss": 0.0681,
"step": 877
},
{
"epoch": 2.629213483146067,
"grad_norm": 1.892886996269226,
"learning_rate": 4.726726726726727e-06,
"loss": 0.0031,
"step": 878
},
{
"epoch": 2.6322097378277154,
"grad_norm": 66.78556823730469,
"learning_rate": 4.720720720720721e-06,
"loss": 0.0983,
"step": 879
},
{
"epoch": 2.635205992509363,
"grad_norm": 8.591114044189453,
"learning_rate": 4.714714714714715e-06,
"loss": 0.1965,
"step": 880
},
{
"epoch": 2.6382022471910114,
"grad_norm": 0.11544928699731827,
"learning_rate": 4.708708708708709e-06,
"loss": 0.0013,
"step": 881
},
{
"epoch": 2.641198501872659,
"grad_norm": 37.577293395996094,
"learning_rate": 4.702702702702703e-06,
"loss": 0.017,
"step": 882
},
{
"epoch": 2.644194756554307,
"grad_norm": 28.23415184020996,
"learning_rate": 4.696696696696697e-06,
"loss": 0.0643,
"step": 883
},
{
"epoch": 2.647191011235955,
"grad_norm": 18.36665916442871,
"learning_rate": 4.6906906906906916e-06,
"loss": 0.0892,
"step": 884
},
{
"epoch": 2.650187265917603,
"grad_norm": 41.964542388916016,
"learning_rate": 4.684684684684685e-06,
"loss": 0.2094,
"step": 885
},
{
"epoch": 2.653183520599251,
"grad_norm": 0.08253569155931473,
"learning_rate": 4.678678678678679e-06,
"loss": 0.0012,
"step": 886
},
{
"epoch": 2.656179775280899,
"grad_norm": 6.9803667068481445,
"learning_rate": 4.672672672672673e-06,
"loss": 0.1977,
"step": 887
},
{
"epoch": 2.6591760299625467,
"grad_norm": 18.16313934326172,
"learning_rate": 4.666666666666667e-06,
"loss": 0.1378,
"step": 888
},
{
"epoch": 2.662172284644195,
"grad_norm": 0.06998983025550842,
"learning_rate": 4.6606606606606604e-06,
"loss": 0.0011,
"step": 889
},
{
"epoch": 2.6651685393258426,
"grad_norm": 7.2744550704956055,
"learning_rate": 4.654654654654655e-06,
"loss": 0.1735,
"step": 890
},
{
"epoch": 2.668164794007491,
"grad_norm": 30.42450523376465,
"learning_rate": 4.6486486486486495e-06,
"loss": 0.0356,
"step": 891
},
{
"epoch": 2.6711610486891386,
"grad_norm": 44.49040222167969,
"learning_rate": 4.642642642642643e-06,
"loss": 0.1461,
"step": 892
},
{
"epoch": 2.6741573033707864,
"grad_norm": 1.59342360496521,
"learning_rate": 4.636636636636637e-06,
"loss": 0.0036,
"step": 893
},
{
"epoch": 2.6771535580524346,
"grad_norm": 29.2288875579834,
"learning_rate": 4.630630630630631e-06,
"loss": 0.0307,
"step": 894
},
{
"epoch": 2.6801498127340824,
"grad_norm": 11.434829711914062,
"learning_rate": 4.624624624624625e-06,
"loss": 0.1934,
"step": 895
},
{
"epoch": 2.6831460674157306,
"grad_norm": 2.359086751937866,
"learning_rate": 4.618618618618619e-06,
"loss": 0.0037,
"step": 896
},
{
"epoch": 2.6861423220973784,
"grad_norm": 60.97223663330078,
"learning_rate": 4.612612612612613e-06,
"loss": 0.0463,
"step": 897
},
{
"epoch": 2.689138576779026,
"grad_norm": 11.485381126403809,
"learning_rate": 4.6066066066066066e-06,
"loss": 0.3903,
"step": 898
},
{
"epoch": 2.692134831460674,
"grad_norm": 6.2090606689453125,
"learning_rate": 4.600600600600601e-06,
"loss": 0.0086,
"step": 899
},
{
"epoch": 2.695131086142322,
"grad_norm": 9.591205596923828,
"learning_rate": 4.594594594594596e-06,
"loss": 0.1575,
"step": 900
},
{
"epoch": 2.69812734082397,
"grad_norm": 6.412684440612793,
"learning_rate": 4.588588588588589e-06,
"loss": 0.1917,
"step": 901
},
{
"epoch": 2.701123595505618,
"grad_norm": 22.770511627197266,
"learning_rate": 4.582582582582583e-06,
"loss": 0.0435,
"step": 902
},
{
"epoch": 2.704119850187266,
"grad_norm": 2.934674024581909,
"learning_rate": 4.576576576576577e-06,
"loss": 0.008,
"step": 903
},
{
"epoch": 2.7071161048689136,
"grad_norm": 17.256736755371094,
"learning_rate": 4.570570570570571e-06,
"loss": 0.1505,
"step": 904
},
{
"epoch": 2.710112359550562,
"grad_norm": 62.274959564208984,
"learning_rate": 4.5645645645645645e-06,
"loss": 0.0212,
"step": 905
},
{
"epoch": 2.7131086142322096,
"grad_norm": 12.418325424194336,
"learning_rate": 4.558558558558559e-06,
"loss": 0.2201,
"step": 906
},
{
"epoch": 2.716104868913858,
"grad_norm": 8.286155700683594,
"learning_rate": 4.552552552552553e-06,
"loss": 0.0192,
"step": 907
},
{
"epoch": 2.7191011235955056,
"grad_norm": 27.978260040283203,
"learning_rate": 4.546546546546547e-06,
"loss": 0.0299,
"step": 908
},
{
"epoch": 2.7220973782771534,
"grad_norm": 42.434749603271484,
"learning_rate": 4.540540540540541e-06,
"loss": 0.1445,
"step": 909
},
{
"epoch": 2.7250936329588016,
"grad_norm": 27.940105438232422,
"learning_rate": 4.534534534534535e-06,
"loss": 0.1832,
"step": 910
},
{
"epoch": 2.7280898876404494,
"grad_norm": 1.3547303676605225,
"learning_rate": 4.528528528528529e-06,
"loss": 0.0098,
"step": 911
},
{
"epoch": 2.7310861423220976,
"grad_norm": 16.92689323425293,
"learning_rate": 4.522522522522522e-06,
"loss": 0.2324,
"step": 912
},
{
"epoch": 2.7340823970037453,
"grad_norm": 12.57506275177002,
"learning_rate": 4.516516516516517e-06,
"loss": 0.0518,
"step": 913
},
{
"epoch": 2.737078651685393,
"grad_norm": 2.679365873336792,
"learning_rate": 4.510510510510511e-06,
"loss": 0.0101,
"step": 914
},
{
"epoch": 2.7400749063670413,
"grad_norm": 0.8655032515525818,
"learning_rate": 4.504504504504505e-06,
"loss": 0.0065,
"step": 915
},
{
"epoch": 2.743071161048689,
"grad_norm": 3.3864078521728516,
"learning_rate": 4.498498498498499e-06,
"loss": 0.013,
"step": 916
},
{
"epoch": 2.7460674157303373,
"grad_norm": 1.2563741207122803,
"learning_rate": 4.492492492492493e-06,
"loss": 0.011,
"step": 917
},
{
"epoch": 2.749063670411985,
"grad_norm": 0.49878114461898804,
"learning_rate": 4.486486486486487e-06,
"loss": 0.0039,
"step": 918
},
{
"epoch": 2.752059925093633,
"grad_norm": 11.697792053222656,
"learning_rate": 4.480480480480481e-06,
"loss": 0.045,
"step": 919
},
{
"epoch": 2.755056179775281,
"grad_norm": 0.36841416358947754,
"learning_rate": 4.474474474474475e-06,
"loss": 0.0031,
"step": 920
},
{
"epoch": 2.758052434456929,
"grad_norm": 41.0269660949707,
"learning_rate": 4.4684684684684686e-06,
"loss": 0.0636,
"step": 921
},
{
"epoch": 2.761048689138577,
"grad_norm": 101.55135345458984,
"learning_rate": 4.462462462462463e-06,
"loss": 0.1721,
"step": 922
},
{
"epoch": 2.764044943820225,
"grad_norm": 0.4171302020549774,
"learning_rate": 4.456456456456457e-06,
"loss": 0.0036,
"step": 923
},
{
"epoch": 2.7670411985018726,
"grad_norm": 0.4484114646911621,
"learning_rate": 4.450450450450451e-06,
"loss": 0.0028,
"step": 924
},
{
"epoch": 2.7700374531835203,
"grad_norm": 5.420213222503662,
"learning_rate": 4.444444444444444e-06,
"loss": 0.071,
"step": 925
},
{
"epoch": 2.7730337078651686,
"grad_norm": 0.597194492816925,
"learning_rate": 4.438438438438439e-06,
"loss": 0.0033,
"step": 926
},
{
"epoch": 2.7760299625468163,
"grad_norm": 4.167457580566406,
"learning_rate": 4.432432432432433e-06,
"loss": 0.0647,
"step": 927
},
{
"epoch": 2.7790262172284645,
"grad_norm": 1.0770858526229858,
"learning_rate": 4.4264264264264265e-06,
"loss": 0.0051,
"step": 928
},
{
"epoch": 2.7820224719101123,
"grad_norm": 37.39751052856445,
"learning_rate": 4.420420420420421e-06,
"loss": 0.2673,
"step": 929
},
{
"epoch": 2.78501872659176,
"grad_norm": 11.69890308380127,
"learning_rate": 4.414414414414415e-06,
"loss": 0.1224,
"step": 930
},
{
"epoch": 2.7880149812734083,
"grad_norm": 3.142592191696167,
"learning_rate": 4.408408408408409e-06,
"loss": 0.0518,
"step": 931
},
{
"epoch": 2.791011235955056,
"grad_norm": 0.5001957416534424,
"learning_rate": 4.402402402402403e-06,
"loss": 0.0022,
"step": 932
},
{
"epoch": 2.7940074906367043,
"grad_norm": 5.0716681480407715,
"learning_rate": 4.396396396396397e-06,
"loss": 0.004,
"step": 933
},
{
"epoch": 2.797003745318352,
"grad_norm": 0.25839465856552124,
"learning_rate": 4.39039039039039e-06,
"loss": 0.0017,
"step": 934
},
{
"epoch": 2.8,
"grad_norm": 0.493880033493042,
"learning_rate": 4.384384384384384e-06,
"loss": 0.0022,
"step": 935
},
{
"epoch": 2.802996254681648,
"grad_norm": 29.217018127441406,
"learning_rate": 4.378378378378379e-06,
"loss": 0.264,
"step": 936
},
{
"epoch": 2.805992509363296,
"grad_norm": 0.13375498354434967,
"learning_rate": 4.372372372372373e-06,
"loss": 0.0012,
"step": 937
},
{
"epoch": 2.808988764044944,
"grad_norm": 2.6857287883758545,
"learning_rate": 4.366366366366367e-06,
"loss": 0.2199,
"step": 938
},
{
"epoch": 2.811985018726592,
"grad_norm": 24.061279296875,
"learning_rate": 4.360360360360361e-06,
"loss": 0.0098,
"step": 939
},
{
"epoch": 2.8149812734082396,
"grad_norm": 1.165687918663025,
"learning_rate": 4.354354354354355e-06,
"loss": 0.0028,
"step": 940
},
{
"epoch": 2.8179775280898878,
"grad_norm": 26.728168487548828,
"learning_rate": 4.348348348348348e-06,
"loss": 0.0122,
"step": 941
},
{
"epoch": 2.8209737827715355,
"grad_norm": 0.08892860263586044,
"learning_rate": 4.342342342342343e-06,
"loss": 0.001,
"step": 942
},
{
"epoch": 2.8239700374531838,
"grad_norm": 0.7470696568489075,
"learning_rate": 4.3363363363363364e-06,
"loss": 0.0019,
"step": 943
},
{
"epoch": 2.8269662921348315,
"grad_norm": 3.205700159072876,
"learning_rate": 4.3303303303303305e-06,
"loss": 0.1981,
"step": 944
},
{
"epoch": 2.8299625468164793,
"grad_norm": 23.71075439453125,
"learning_rate": 4.324324324324325e-06,
"loss": 0.0421,
"step": 945
},
{
"epoch": 2.832958801498127,
"grad_norm": 64.29513549804688,
"learning_rate": 4.318318318318319e-06,
"loss": 0.1661,
"step": 946
},
{
"epoch": 2.8359550561797753,
"grad_norm": 6.911731243133545,
"learning_rate": 4.312312312312313e-06,
"loss": 0.0765,
"step": 947
},
{
"epoch": 2.8389513108614235,
"grad_norm": 27.351591110229492,
"learning_rate": 4.306306306306306e-06,
"loss": 0.1383,
"step": 948
},
{
"epoch": 2.8419475655430713,
"grad_norm": 1.3769259452819824,
"learning_rate": 4.300300300300301e-06,
"loss": 0.0116,
"step": 949
},
{
"epoch": 2.844943820224719,
"grad_norm": 8.797236442565918,
"learning_rate": 4.294294294294294e-06,
"loss": 0.0105,
"step": 950
},
{
"epoch": 2.847940074906367,
"grad_norm": 0.6100442409515381,
"learning_rate": 4.2882882882882885e-06,
"loss": 0.0031,
"step": 951
},
{
"epoch": 2.850936329588015,
"grad_norm": 0.11245360225439072,
"learning_rate": 4.282282282282283e-06,
"loss": 0.0015,
"step": 952
},
{
"epoch": 2.853932584269663,
"grad_norm": 13.6712064743042,
"learning_rate": 4.276276276276277e-06,
"loss": 0.1337,
"step": 953
},
{
"epoch": 2.856928838951311,
"grad_norm": 0.09966795146465302,
"learning_rate": 4.270270270270271e-06,
"loss": 0.0014,
"step": 954
},
{
"epoch": 2.8599250936329588,
"grad_norm": 9.589887619018555,
"learning_rate": 4.264264264264265e-06,
"loss": 0.2213,
"step": 955
},
{
"epoch": 2.8629213483146065,
"grad_norm": 0.7842001914978027,
"learning_rate": 4.258258258258259e-06,
"loss": 0.0027,
"step": 956
},
{
"epoch": 2.8659176029962548,
"grad_norm": 0.18834924697875977,
"learning_rate": 4.252252252252252e-06,
"loss": 0.0017,
"step": 957
},
{
"epoch": 2.8689138576779025,
"grad_norm": 12.313905715942383,
"learning_rate": 4.246246246246246e-06,
"loss": 0.1867,
"step": 958
},
{
"epoch": 2.8719101123595507,
"grad_norm": 4.269379138946533,
"learning_rate": 4.2402402402402405e-06,
"loss": 0.1962,
"step": 959
},
{
"epoch": 2.8749063670411985,
"grad_norm": 7.370109558105469,
"learning_rate": 4.234234234234235e-06,
"loss": 0.0208,
"step": 960
},
{
"epoch": 2.8779026217228463,
"grad_norm": 204.92921447753906,
"learning_rate": 4.228228228228229e-06,
"loss": 0.174,
"step": 961
},
{
"epoch": 2.8808988764044945,
"grad_norm": 0.2679640054702759,
"learning_rate": 4.222222222222223e-06,
"loss": 0.002,
"step": 962
},
{
"epoch": 2.8838951310861423,
"grad_norm": 0.22722303867340088,
"learning_rate": 4.216216216216217e-06,
"loss": 0.0021,
"step": 963
},
{
"epoch": 2.8868913857677905,
"grad_norm": 7.144226551055908,
"learning_rate": 4.21021021021021e-06,
"loss": 0.036,
"step": 964
},
{
"epoch": 2.8898876404494382,
"grad_norm": 1.2280908823013306,
"learning_rate": 4.204204204204204e-06,
"loss": 0.0038,
"step": 965
},
{
"epoch": 2.892883895131086,
"grad_norm": 132.9735107421875,
"learning_rate": 4.1981981981981984e-06,
"loss": 0.108,
"step": 966
},
{
"epoch": 2.8958801498127342,
"grad_norm": 59.87287139892578,
"learning_rate": 4.1921921921921925e-06,
"loss": 0.0791,
"step": 967
},
{
"epoch": 2.898876404494382,
"grad_norm": 27.576318740844727,
"learning_rate": 4.186186186186187e-06,
"loss": 0.1638,
"step": 968
},
{
"epoch": 2.90187265917603,
"grad_norm": 5.267337799072266,
"learning_rate": 4.180180180180181e-06,
"loss": 0.0397,
"step": 969
},
{
"epoch": 2.904868913857678,
"grad_norm": 0.24482539296150208,
"learning_rate": 4.174174174174174e-06,
"loss": 0.0021,
"step": 970
},
{
"epoch": 2.9078651685393258,
"grad_norm": 10.936861038208008,
"learning_rate": 4.168168168168168e-06,
"loss": 0.078,
"step": 971
},
{
"epoch": 2.9108614232209735,
"grad_norm": 1.8542410135269165,
"learning_rate": 4.162162162162163e-06,
"loss": 0.0051,
"step": 972
},
{
"epoch": 2.9138576779026217,
"grad_norm": 0.12661808729171753,
"learning_rate": 4.156156156156156e-06,
"loss": 0.0014,
"step": 973
},
{
"epoch": 2.9168539325842695,
"grad_norm": 30.025484085083008,
"learning_rate": 4.1501501501501505e-06,
"loss": 0.0946,
"step": 974
},
{
"epoch": 2.9198501872659177,
"grad_norm": 68.63817596435547,
"learning_rate": 4.1441441441441446e-06,
"loss": 0.1685,
"step": 975
},
{
"epoch": 2.9228464419475655,
"grad_norm": 48.791038513183594,
"learning_rate": 4.138138138138139e-06,
"loss": 0.0473,
"step": 976
},
{
"epoch": 2.9258426966292133,
"grad_norm": 15.335400581359863,
"learning_rate": 4.132132132132132e-06,
"loss": 0.1922,
"step": 977
},
{
"epoch": 2.9288389513108615,
"grad_norm": 77.87091827392578,
"learning_rate": 4.126126126126127e-06,
"loss": 0.1616,
"step": 978
},
{
"epoch": 2.9318352059925092,
"grad_norm": 0.6916086077690125,
"learning_rate": 4.12012012012012e-06,
"loss": 0.0024,
"step": 979
},
{
"epoch": 2.9348314606741575,
"grad_norm": 149.41226196289062,
"learning_rate": 4.114114114114114e-06,
"loss": 0.034,
"step": 980
},
{
"epoch": 2.9378277153558052,
"grad_norm": 54.30559158325195,
"learning_rate": 4.108108108108108e-06,
"loss": 0.0713,
"step": 981
},
{
"epoch": 2.940823970037453,
"grad_norm": 1.0688893795013428,
"learning_rate": 4.1021021021021025e-06,
"loss": 0.0098,
"step": 982
},
{
"epoch": 2.943820224719101,
"grad_norm": 21.895240783691406,
"learning_rate": 4.096096096096097e-06,
"loss": 0.1052,
"step": 983
},
{
"epoch": 2.946816479400749,
"grad_norm": 0.12046666443347931,
"learning_rate": 4.09009009009009e-06,
"loss": 0.0015,
"step": 984
},
{
"epoch": 2.949812734082397,
"grad_norm": 9.820867538452148,
"learning_rate": 4.084084084084085e-06,
"loss": 0.2097,
"step": 985
},
{
"epoch": 2.952808988764045,
"grad_norm": 1.024439811706543,
"learning_rate": 4.078078078078078e-06,
"loss": 0.0032,
"step": 986
},
{
"epoch": 2.9558052434456927,
"grad_norm": 30.047760009765625,
"learning_rate": 4.072072072072072e-06,
"loss": 0.2213,
"step": 987
},
{
"epoch": 2.958801498127341,
"grad_norm": 0.10897261649370193,
"learning_rate": 4.066066066066066e-06,
"loss": 0.0013,
"step": 988
},
{
"epoch": 2.9617977528089887,
"grad_norm": 46.3768196105957,
"learning_rate": 4.06006006006006e-06,
"loss": 0.1911,
"step": 989
},
{
"epoch": 2.964794007490637,
"grad_norm": 0.3740360140800476,
"learning_rate": 4.0540540540540545e-06,
"loss": 0.002,
"step": 990
},
{
"epoch": 2.9677902621722847,
"grad_norm": 0.14058686792850494,
"learning_rate": 4.048048048048049e-06,
"loss": 0.0014,
"step": 991
},
{
"epoch": 2.9707865168539325,
"grad_norm": 13.293370246887207,
"learning_rate": 4.042042042042043e-06,
"loss": 0.1753,
"step": 992
},
{
"epoch": 2.9737827715355807,
"grad_norm": 0.35173657536506653,
"learning_rate": 4.036036036036036e-06,
"loss": 0.0021,
"step": 993
},
{
"epoch": 2.9767790262172285,
"grad_norm": 0.5589150786399841,
"learning_rate": 4.03003003003003e-06,
"loss": 0.0023,
"step": 994
},
{
"epoch": 2.9797752808988767,
"grad_norm": 0.4406314492225647,
"learning_rate": 4.024024024024024e-06,
"loss": 0.0022,
"step": 995
},
{
"epoch": 2.9827715355805244,
"grad_norm": 2.573115348815918,
"learning_rate": 4.018018018018018e-06,
"loss": 0.0415,
"step": 996
},
{
"epoch": 2.985767790262172,
"grad_norm": 15.697823524475098,
"learning_rate": 4.0120120120120124e-06,
"loss": 0.0301,
"step": 997
},
{
"epoch": 2.98876404494382,
"grad_norm": 15.628379821777344,
"learning_rate": 4.0060060060060066e-06,
"loss": 0.1527,
"step": 998
},
{
"epoch": 2.991760299625468,
"grad_norm": 0.258297860622406,
"learning_rate": 4.000000000000001e-06,
"loss": 0.0015,
"step": 999
},
{
"epoch": 2.994756554307116,
"grad_norm": 25.658567428588867,
"learning_rate": 3.993993993993994e-06,
"loss": 0.2117,
"step": 1000
},
{
"epoch": 2.997752808988764,
"grad_norm": 0.44404181838035583,
"learning_rate": 3.987987987987989e-06,
"loss": 0.0017,
"step": 1001
},
{
"epoch": 3.0,
"grad_norm": 0.25101685523986816,
"learning_rate": 3.981981981981982e-06,
"loss": 0.0012,
"step": 1002
},
{
"epoch": 3.0,
"eval_accuracy": 0.9626998223801065,
"eval_loss": 0.1771629899740219,
"eval_runtime": 21.4693,
"eval_samples_per_second": 26.224,
"eval_steps_per_second": 3.307,
"step": 1002
},
{
"epoch": 3.0029962546816478,
"grad_norm": 28.166404724121094,
"learning_rate": 3.975975975975976e-06,
"loss": 0.1058,
"step": 1003
},
{
"epoch": 3.005992509363296,
"grad_norm": 6.128506183624268,
"learning_rate": 3.96996996996997e-06,
"loss": 0.0118,
"step": 1004
},
{
"epoch": 3.0089887640449438,
"grad_norm": 0.18079061806201935,
"learning_rate": 3.9639639639639645e-06,
"loss": 0.0017,
"step": 1005
},
{
"epoch": 3.011985018726592,
"grad_norm": 3.7501842975616455,
"learning_rate": 3.957957957957958e-06,
"loss": 0.209,
"step": 1006
},
{
"epoch": 3.0149812734082397,
"grad_norm": 17.950193405151367,
"learning_rate": 3.951951951951952e-06,
"loss": 0.0178,
"step": 1007
},
{
"epoch": 3.0179775280898875,
"grad_norm": 4.206531047821045,
"learning_rate": 3.945945945945947e-06,
"loss": 0.007,
"step": 1008
},
{
"epoch": 3.0209737827715357,
"grad_norm": 3.0722286701202393,
"learning_rate": 3.93993993993994e-06,
"loss": 0.0099,
"step": 1009
},
{
"epoch": 3.0239700374531835,
"grad_norm": 5.700033187866211,
"learning_rate": 3.933933933933934e-06,
"loss": 0.0056,
"step": 1010
},
{
"epoch": 3.0269662921348313,
"grad_norm": 0.17911633849143982,
"learning_rate": 3.927927927927928e-06,
"loss": 0.0016,
"step": 1011
},
{
"epoch": 3.0299625468164795,
"grad_norm": 22.76339340209961,
"learning_rate": 3.921921921921922e-06,
"loss": 0.166,
"step": 1012
},
{
"epoch": 3.0329588014981272,
"grad_norm": 82.44022369384766,
"learning_rate": 3.915915915915916e-06,
"loss": 0.0263,
"step": 1013
},
{
"epoch": 3.0359550561797755,
"grad_norm": 20.501956939697266,
"learning_rate": 3.909909909909911e-06,
"loss": 0.155,
"step": 1014
},
{
"epoch": 3.0389513108614232,
"grad_norm": 1.1340712308883667,
"learning_rate": 3.903903903903904e-06,
"loss": 0.0099,
"step": 1015
},
{
"epoch": 3.041947565543071,
"grad_norm": 17.413236618041992,
"learning_rate": 3.897897897897898e-06,
"loss": 0.6069,
"step": 1016
},
{
"epoch": 3.044943820224719,
"grad_norm": 0.5072112083435059,
"learning_rate": 3.891891891891892e-06,
"loss": 0.0014,
"step": 1017
},
{
"epoch": 3.047940074906367,
"grad_norm": 82.4122314453125,
"learning_rate": 3.885885885885886e-06,
"loss": 0.0481,
"step": 1018
},
{
"epoch": 3.050936329588015,
"grad_norm": 0.10695996880531311,
"learning_rate": 3.87987987987988e-06,
"loss": 0.0011,
"step": 1019
},
{
"epoch": 3.053932584269663,
"grad_norm": 0.18931803107261658,
"learning_rate": 3.8738738738738744e-06,
"loss": 0.0014,
"step": 1020
},
{
"epoch": 3.0569288389513107,
"grad_norm": 12.460945129394531,
"learning_rate": 3.8678678678678685e-06,
"loss": 0.1022,
"step": 1021
},
{
"epoch": 3.059925093632959,
"grad_norm": 0.30106574296951294,
"learning_rate": 3.861861861861862e-06,
"loss": 0.0026,
"step": 1022
},
{
"epoch": 3.0629213483146067,
"grad_norm": 35.0296630859375,
"learning_rate": 3.855855855855856e-06,
"loss": 0.1508,
"step": 1023
},
{
"epoch": 3.0659176029962545,
"grad_norm": 0.24383395910263062,
"learning_rate": 3.84984984984985e-06,
"loss": 0.0021,
"step": 1024
},
{
"epoch": 3.0689138576779027,
"grad_norm": 13.463871002197266,
"learning_rate": 3.843843843843844e-06,
"loss": 0.0238,
"step": 1025
},
{
"epoch": 3.0719101123595505,
"grad_norm": 0.19878296554088593,
"learning_rate": 3.837837837837838e-06,
"loss": 0.0019,
"step": 1026
},
{
"epoch": 3.0749063670411987,
"grad_norm": 0.2445116937160492,
"learning_rate": 3.831831831831832e-06,
"loss": 0.002,
"step": 1027
},
{
"epoch": 3.0779026217228465,
"grad_norm": 16.643571853637695,
"learning_rate": 3.8258258258258265e-06,
"loss": 0.0283,
"step": 1028
},
{
"epoch": 3.0808988764044942,
"grad_norm": 22.042783737182617,
"learning_rate": 3.81981981981982e-06,
"loss": 0.1676,
"step": 1029
},
{
"epoch": 3.0838951310861424,
"grad_norm": 1.406181812286377,
"learning_rate": 3.8138138138138143e-06,
"loss": 0.0074,
"step": 1030
},
{
"epoch": 3.08689138576779,
"grad_norm": 2.1913270950317383,
"learning_rate": 3.807807807807808e-06,
"loss": 0.003,
"step": 1031
},
{
"epoch": 3.0898876404494384,
"grad_norm": 169.27557373046875,
"learning_rate": 3.801801801801802e-06,
"loss": 0.09,
"step": 1032
},
{
"epoch": 3.092883895131086,
"grad_norm": 0.09239841997623444,
"learning_rate": 3.795795795795796e-06,
"loss": 0.0011,
"step": 1033
},
{
"epoch": 3.095880149812734,
"grad_norm": 0.07967542111873627,
"learning_rate": 3.78978978978979e-06,
"loss": 0.001,
"step": 1034
},
{
"epoch": 3.098876404494382,
"grad_norm": 40.274864196777344,
"learning_rate": 3.7837837837837844e-06,
"loss": 0.1974,
"step": 1035
},
{
"epoch": 3.10187265917603,
"grad_norm": 0.18797746300697327,
"learning_rate": 3.777777777777778e-06,
"loss": 0.0015,
"step": 1036
},
{
"epoch": 3.1048689138576777,
"grad_norm": 15.524847030639648,
"learning_rate": 3.771771771771772e-06,
"loss": 0.0983,
"step": 1037
},
{
"epoch": 3.107865168539326,
"grad_norm": 2.8484082221984863,
"learning_rate": 3.765765765765766e-06,
"loss": 0.0037,
"step": 1038
},
{
"epoch": 3.1108614232209737,
"grad_norm": 5.206206798553467,
"learning_rate": 3.75975975975976e-06,
"loss": 0.1621,
"step": 1039
},
{
"epoch": 3.113857677902622,
"grad_norm": 0.14623048901557922,
"learning_rate": 3.7537537537537537e-06,
"loss": 0.0014,
"step": 1040
},
{
"epoch": 3.1168539325842697,
"grad_norm": 0.13604526221752167,
"learning_rate": 3.747747747747748e-06,
"loss": 0.0011,
"step": 1041
},
{
"epoch": 3.1198501872659175,
"grad_norm": 0.1345546543598175,
"learning_rate": 3.7417417417417423e-06,
"loss": 0.0012,
"step": 1042
},
{
"epoch": 3.1228464419475657,
"grad_norm": 0.31141021847724915,
"learning_rate": 3.735735735735736e-06,
"loss": 0.0013,
"step": 1043
},
{
"epoch": 3.1258426966292134,
"grad_norm": 3.8004722595214844,
"learning_rate": 3.72972972972973e-06,
"loss": 0.0573,
"step": 1044
},
{
"epoch": 3.128838951310861,
"grad_norm": 0.34975168108940125,
"learning_rate": 3.723723723723724e-06,
"loss": 0.0022,
"step": 1045
},
{
"epoch": 3.1318352059925094,
"grad_norm": 0.836336076259613,
"learning_rate": 3.7177177177177183e-06,
"loss": 0.0073,
"step": 1046
},
{
"epoch": 3.134831460674157,
"grad_norm": 1.127227544784546,
"learning_rate": 3.711711711711712e-06,
"loss": 0.0037,
"step": 1047
},
{
"epoch": 3.1378277153558054,
"grad_norm": 1.775098204612732,
"learning_rate": 3.705705705705706e-06,
"loss": 0.0091,
"step": 1048
},
{
"epoch": 3.140823970037453,
"grad_norm": 27.621877670288086,
"learning_rate": 3.6996996996997e-06,
"loss": 0.0827,
"step": 1049
},
{
"epoch": 3.143820224719101,
"grad_norm": 0.06538525223731995,
"learning_rate": 3.693693693693694e-06,
"loss": 0.0009,
"step": 1050
},
{
"epoch": 3.146816479400749,
"grad_norm": 0.06390548497438431,
"learning_rate": 3.6876876876876885e-06,
"loss": 0.0008,
"step": 1051
},
{
"epoch": 3.149812734082397,
"grad_norm": 2.483194351196289,
"learning_rate": 3.681681681681682e-06,
"loss": 0.0175,
"step": 1052
},
{
"epoch": 3.152808988764045,
"grad_norm": 39.44814682006836,
"learning_rate": 3.6756756756756763e-06,
"loss": 0.0202,
"step": 1053
},
{
"epoch": 3.155805243445693,
"grad_norm": 19.42127227783203,
"learning_rate": 3.66966966966967e-06,
"loss": 0.4345,
"step": 1054
},
{
"epoch": 3.1588014981273407,
"grad_norm": 0.10200697928667068,
"learning_rate": 3.663663663663664e-06,
"loss": 0.0009,
"step": 1055
},
{
"epoch": 3.161797752808989,
"grad_norm": 0.7569524049758911,
"learning_rate": 3.6576576576576577e-06,
"loss": 0.0135,
"step": 1056
},
{
"epoch": 3.1647940074906367,
"grad_norm": 44.45731735229492,
"learning_rate": 3.651651651651652e-06,
"loss": 0.1243,
"step": 1057
},
{
"epoch": 3.167790262172285,
"grad_norm": 0.7283141016960144,
"learning_rate": 3.6456456456456455e-06,
"loss": 0.0012,
"step": 1058
},
{
"epoch": 3.1707865168539326,
"grad_norm": 45.335819244384766,
"learning_rate": 3.63963963963964e-06,
"loss": 0.024,
"step": 1059
},
{
"epoch": 3.1737827715355804,
"grad_norm": 2.0355594158172607,
"learning_rate": 3.633633633633634e-06,
"loss": 0.013,
"step": 1060
},
{
"epoch": 3.1767790262172286,
"grad_norm": 83.11621856689453,
"learning_rate": 3.627627627627628e-06,
"loss": 0.1168,
"step": 1061
},
{
"epoch": 3.1797752808988764,
"grad_norm": 0.09632370620965958,
"learning_rate": 3.621621621621622e-06,
"loss": 0.0008,
"step": 1062
},
{
"epoch": 3.182771535580524,
"grad_norm": 6.644118309020996,
"learning_rate": 3.6156156156156157e-06,
"loss": 0.2081,
"step": 1063
},
{
"epoch": 3.1857677902621724,
"grad_norm": 0.16423919796943665,
"learning_rate": 3.60960960960961e-06,
"loss": 0.001,
"step": 1064
},
{
"epoch": 3.18876404494382,
"grad_norm": 87.73657989501953,
"learning_rate": 3.603603603603604e-06,
"loss": 0.1015,
"step": 1065
},
{
"epoch": 3.1917602996254684,
"grad_norm": 15.627496719360352,
"learning_rate": 3.597597597597598e-06,
"loss": 0.1539,
"step": 1066
},
{
"epoch": 3.194756554307116,
"grad_norm": 0.07909096032381058,
"learning_rate": 3.5915915915915917e-06,
"loss": 0.0009,
"step": 1067
},
{
"epoch": 3.197752808988764,
"grad_norm": 0.9036718010902405,
"learning_rate": 3.5855855855855858e-06,
"loss": 0.0017,
"step": 1068
},
{
"epoch": 3.200749063670412,
"grad_norm": 2.1594274044036865,
"learning_rate": 3.5795795795795803e-06,
"loss": 0.0037,
"step": 1069
},
{
"epoch": 3.20374531835206,
"grad_norm": 0.0753878653049469,
"learning_rate": 3.573573573573574e-06,
"loss": 0.0009,
"step": 1070
},
{
"epoch": 3.2067415730337077,
"grad_norm": 3.1338107585906982,
"learning_rate": 3.567567567567568e-06,
"loss": 0.0111,
"step": 1071
},
{
"epoch": 3.209737827715356,
"grad_norm": 0.14338922500610352,
"learning_rate": 3.561561561561562e-06,
"loss": 0.0013,
"step": 1072
},
{
"epoch": 3.2127340823970036,
"grad_norm": 12.572104454040527,
"learning_rate": 3.555555555555556e-06,
"loss": 0.2825,
"step": 1073
},
{
"epoch": 3.215730337078652,
"grad_norm": 20.024112701416016,
"learning_rate": 3.5495495495495496e-06,
"loss": 0.1248,
"step": 1074
},
{
"epoch": 3.2187265917602996,
"grad_norm": 11.643209457397461,
"learning_rate": 3.5435435435435437e-06,
"loss": 0.085,
"step": 1075
},
{
"epoch": 3.2217228464419474,
"grad_norm": 0.15489691495895386,
"learning_rate": 3.5375375375375374e-06,
"loss": 0.0014,
"step": 1076
},
{
"epoch": 3.2247191011235956,
"grad_norm": 77.08690643310547,
"learning_rate": 3.531531531531532e-06,
"loss": 0.0644,
"step": 1077
},
{
"epoch": 3.2277153558052434,
"grad_norm": 0.11881520599126816,
"learning_rate": 3.525525525525526e-06,
"loss": 0.0011,
"step": 1078
},
{
"epoch": 3.2307116104868916,
"grad_norm": 14.568486213684082,
"learning_rate": 3.5195195195195197e-06,
"loss": 0.1417,
"step": 1079
},
{
"epoch": 3.2337078651685394,
"grad_norm": 29.129297256469727,
"learning_rate": 3.513513513513514e-06,
"loss": 0.2043,
"step": 1080
},
{
"epoch": 3.236704119850187,
"grad_norm": 2.8419158458709717,
"learning_rate": 3.5075075075075075e-06,
"loss": 0.009,
"step": 1081
},
{
"epoch": 3.2397003745318353,
"grad_norm": 0.82688969373703,
"learning_rate": 3.501501501501502e-06,
"loss": 0.0024,
"step": 1082
},
{
"epoch": 3.242696629213483,
"grad_norm": 40.012550354003906,
"learning_rate": 3.4954954954954957e-06,
"loss": 0.1276,
"step": 1083
},
{
"epoch": 3.245692883895131,
"grad_norm": 0.16128391027450562,
"learning_rate": 3.48948948948949e-06,
"loss": 0.0012,
"step": 1084
},
{
"epoch": 3.248689138576779,
"grad_norm": 0.23443275690078735,
"learning_rate": 3.4834834834834835e-06,
"loss": 0.0016,
"step": 1085
},
{
"epoch": 3.251685393258427,
"grad_norm": 9.920722007751465,
"learning_rate": 3.4774774774774776e-06,
"loss": 0.099,
"step": 1086
},
{
"epoch": 3.254681647940075,
"grad_norm": 0.3433252274990082,
"learning_rate": 3.471471471471472e-06,
"loss": 0.0026,
"step": 1087
},
{
"epoch": 3.257677902621723,
"grad_norm": 514.9418334960938,
"learning_rate": 3.465465465465466e-06,
"loss": 0.071,
"step": 1088
},
{
"epoch": 3.2606741573033706,
"grad_norm": 1.8422298431396484,
"learning_rate": 3.45945945945946e-06,
"loss": 0.0092,
"step": 1089
},
{
"epoch": 3.263670411985019,
"grad_norm": 0.14507924020290375,
"learning_rate": 3.4534534534534537e-06,
"loss": 0.0008,
"step": 1090
},
{
"epoch": 3.2666666666666666,
"grad_norm": 1.0720672607421875,
"learning_rate": 3.4474474474474478e-06,
"loss": 0.0079,
"step": 1091
},
{
"epoch": 3.2696629213483144,
"grad_norm": 0.17518070340156555,
"learning_rate": 3.4414414414414415e-06,
"loss": 0.0013,
"step": 1092
},
{
"epoch": 3.2726591760299626,
"grad_norm": 0.053091756999492645,
"learning_rate": 3.435435435435436e-06,
"loss": 0.0008,
"step": 1093
},
{
"epoch": 3.2756554307116104,
"grad_norm": 0.7648254632949829,
"learning_rate": 3.4294294294294293e-06,
"loss": 0.0059,
"step": 1094
},
{
"epoch": 3.2786516853932586,
"grad_norm": 3.3494112491607666,
"learning_rate": 3.423423423423424e-06,
"loss": 0.0508,
"step": 1095
},
{
"epoch": 3.2816479400749063,
"grad_norm": 0.06449177116155624,
"learning_rate": 3.417417417417418e-06,
"loss": 0.0008,
"step": 1096
},
{
"epoch": 3.284644194756554,
"grad_norm": 13.884971618652344,
"learning_rate": 3.4114114114114116e-06,
"loss": 0.2007,
"step": 1097
},
{
"epoch": 3.2876404494382023,
"grad_norm": 6.7031402587890625,
"learning_rate": 3.4054054054054057e-06,
"loss": 0.2305,
"step": 1098
},
{
"epoch": 3.29063670411985,
"grad_norm": 0.06286942958831787,
"learning_rate": 3.3993993993993994e-06,
"loss": 0.0008,
"step": 1099
},
{
"epoch": 3.2936329588014983,
"grad_norm": 0.16238509118556976,
"learning_rate": 3.393393393393394e-06,
"loss": 0.0012,
"step": 1100
},
{
"epoch": 3.296629213483146,
"grad_norm": 14.298246383666992,
"learning_rate": 3.3873873873873876e-06,
"loss": 0.119,
"step": 1101
},
{
"epoch": 3.299625468164794,
"grad_norm": 9.177990913391113,
"learning_rate": 3.3813813813813817e-06,
"loss": 0.282,
"step": 1102
},
{
"epoch": 3.302621722846442,
"grad_norm": 8.57481861114502,
"learning_rate": 3.3753753753753754e-06,
"loss": 0.1291,
"step": 1103
},
{
"epoch": 3.30561797752809,
"grad_norm": 28.68311882019043,
"learning_rate": 3.3693693693693695e-06,
"loss": 0.0891,
"step": 1104
},
{
"epoch": 3.308614232209738,
"grad_norm": 10.132109642028809,
"learning_rate": 3.363363363363364e-06,
"loss": 0.0121,
"step": 1105
},
{
"epoch": 3.311610486891386,
"grad_norm": 8.56514835357666,
"learning_rate": 3.3573573573573577e-06,
"loss": 0.1897,
"step": 1106
},
{
"epoch": 3.3146067415730336,
"grad_norm": 6.790737628936768,
"learning_rate": 3.351351351351352e-06,
"loss": 0.1676,
"step": 1107
},
{
"epoch": 3.317602996254682,
"grad_norm": 3.557913303375244,
"learning_rate": 3.3453453453453455e-06,
"loss": 0.2126,
"step": 1108
},
{
"epoch": 3.3205992509363296,
"grad_norm": 0.2737390398979187,
"learning_rate": 3.3393393393393396e-06,
"loss": 0.0018,
"step": 1109
},
{
"epoch": 3.3235955056179773,
"grad_norm": 23.446863174438477,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.0165,
"step": 1110
},
{
"epoch": 3.3265917602996256,
"grad_norm": 0.42158132791519165,
"learning_rate": 3.327327327327328e-06,
"loss": 0.0026,
"step": 1111
},
{
"epoch": 3.3295880149812733,
"grad_norm": 0.5871883034706116,
"learning_rate": 3.3213213213213215e-06,
"loss": 0.0019,
"step": 1112
},
{
"epoch": 3.3325842696629215,
"grad_norm": 0.16152256727218628,
"learning_rate": 3.3153153153153157e-06,
"loss": 0.0013,
"step": 1113
},
{
"epoch": 3.3355805243445693,
"grad_norm": 0.23422731459140778,
"learning_rate": 3.3093093093093098e-06,
"loss": 0.002,
"step": 1114
},
{
"epoch": 3.338576779026217,
"grad_norm": 13.004279136657715,
"learning_rate": 3.3033033033033035e-06,
"loss": 0.079,
"step": 1115
},
{
"epoch": 3.3415730337078653,
"grad_norm": 0.5757007002830505,
"learning_rate": 3.2972972972972976e-06,
"loss": 0.0035,
"step": 1116
},
{
"epoch": 3.344569288389513,
"grad_norm": 26.546422958374023,
"learning_rate": 3.2912912912912912e-06,
"loss": 0.1743,
"step": 1117
},
{
"epoch": 3.347565543071161,
"grad_norm": 7.220470905303955,
"learning_rate": 3.2852852852852858e-06,
"loss": 0.017,
"step": 1118
},
{
"epoch": 3.350561797752809,
"grad_norm": 140.53004455566406,
"learning_rate": 3.2792792792792795e-06,
"loss": 0.0916,
"step": 1119
},
{
"epoch": 3.353558052434457,
"grad_norm": 0.8136566281318665,
"learning_rate": 3.2732732732732736e-06,
"loss": 0.0046,
"step": 1120
},
{
"epoch": 3.356554307116105,
"grad_norm": 17.27185821533203,
"learning_rate": 3.2672672672672673e-06,
"loss": 0.0256,
"step": 1121
},
{
"epoch": 3.359550561797753,
"grad_norm": 10.133230209350586,
"learning_rate": 3.2612612612612614e-06,
"loss": 0.1214,
"step": 1122
},
{
"epoch": 3.3625468164794006,
"grad_norm": 0.5158509612083435,
"learning_rate": 3.255255255255256e-06,
"loss": 0.0027,
"step": 1123
},
{
"epoch": 3.365543071161049,
"grad_norm": 0.46048080921173096,
"learning_rate": 3.2492492492492496e-06,
"loss": 0.0022,
"step": 1124
},
{
"epoch": 3.3685393258426966,
"grad_norm": 3.597918748855591,
"learning_rate": 3.2432432432432437e-06,
"loss": 0.0051,
"step": 1125
},
{
"epoch": 3.3715355805243448,
"grad_norm": 0.5563721060752869,
"learning_rate": 3.2372372372372374e-06,
"loss": 0.0035,
"step": 1126
},
{
"epoch": 3.3745318352059925,
"grad_norm": 5.093927383422852,
"learning_rate": 3.2312312312312315e-06,
"loss": 0.0049,
"step": 1127
},
{
"epoch": 3.3775280898876403,
"grad_norm": 1.1499890089035034,
"learning_rate": 3.225225225225225e-06,
"loss": 0.0118,
"step": 1128
},
{
"epoch": 3.3805243445692885,
"grad_norm": 10.400493621826172,
"learning_rate": 3.2192192192192197e-06,
"loss": 0.0791,
"step": 1129
},
{
"epoch": 3.3835205992509363,
"grad_norm": 21.45985984802246,
"learning_rate": 3.2132132132132134e-06,
"loss": 0.0098,
"step": 1130
},
{
"epoch": 3.3865168539325845,
"grad_norm": 0.8989500403404236,
"learning_rate": 3.2072072072072075e-06,
"loss": 0.002,
"step": 1131
},
{
"epoch": 3.3895131086142323,
"grad_norm": 12.515076637268066,
"learning_rate": 3.2012012012012016e-06,
"loss": 0.1325,
"step": 1132
},
{
"epoch": 3.39250936329588,
"grad_norm": 0.42851266264915466,
"learning_rate": 3.1951951951951953e-06,
"loss": 0.0037,
"step": 1133
},
{
"epoch": 3.3955056179775283,
"grad_norm": 0.10597572475671768,
"learning_rate": 3.1891891891891894e-06,
"loss": 0.0011,
"step": 1134
},
{
"epoch": 3.398501872659176,
"grad_norm": 14.891583442687988,
"learning_rate": 3.183183183183183e-06,
"loss": 0.3976,
"step": 1135
},
{
"epoch": 3.401498127340824,
"grad_norm": 90.40242004394531,
"learning_rate": 3.1771771771771776e-06,
"loss": 0.1723,
"step": 1136
},
{
"epoch": 3.404494382022472,
"grad_norm": 0.5071372985839844,
"learning_rate": 3.1711711711711713e-06,
"loss": 0.0036,
"step": 1137
},
{
"epoch": 3.40749063670412,
"grad_norm": 0.10828603804111481,
"learning_rate": 3.1651651651651654e-06,
"loss": 0.0008,
"step": 1138
},
{
"epoch": 3.4104868913857675,
"grad_norm": 0.1566867232322693,
"learning_rate": 3.159159159159159e-06,
"loss": 0.001,
"step": 1139
},
{
"epoch": 3.4134831460674158,
"grad_norm": 28.552108764648438,
"learning_rate": 3.1531531531531532e-06,
"loss": 0.1246,
"step": 1140
},
{
"epoch": 3.4164794007490635,
"grad_norm": 5.321919918060303,
"learning_rate": 3.1471471471471478e-06,
"loss": 0.0049,
"step": 1141
},
{
"epoch": 3.4194756554307117,
"grad_norm": 5.983062744140625,
"learning_rate": 3.1411411411411415e-06,
"loss": 0.1989,
"step": 1142
},
{
"epoch": 3.4224719101123595,
"grad_norm": 0.12518221139907837,
"learning_rate": 3.1351351351351356e-06,
"loss": 0.0009,
"step": 1143
},
{
"epoch": 3.4254681647940073,
"grad_norm": 0.0917065292596817,
"learning_rate": 3.1291291291291293e-06,
"loss": 0.0009,
"step": 1144
},
{
"epoch": 3.4284644194756555,
"grad_norm": 21.737844467163086,
"learning_rate": 3.1231231231231234e-06,
"loss": 0.1554,
"step": 1145
},
{
"epoch": 3.4314606741573033,
"grad_norm": 0.10727402567863464,
"learning_rate": 3.117117117117117e-06,
"loss": 0.0013,
"step": 1146
},
{
"epoch": 3.4344569288389515,
"grad_norm": 14.968389511108398,
"learning_rate": 3.1111111111111116e-06,
"loss": 0.0216,
"step": 1147
},
{
"epoch": 3.4374531835205993,
"grad_norm": 0.12470707297325134,
"learning_rate": 3.1051051051051053e-06,
"loss": 0.0009,
"step": 1148
},
{
"epoch": 3.440449438202247,
"grad_norm": 6.339749336242676,
"learning_rate": 3.0990990990990994e-06,
"loss": 0.1745,
"step": 1149
},
{
"epoch": 3.4434456928838952,
"grad_norm": 2.9502055644989014,
"learning_rate": 3.0930930930930935e-06,
"loss": 0.014,
"step": 1150
},
{
"epoch": 3.446441947565543,
"grad_norm": 0.4890146553516388,
"learning_rate": 3.087087087087087e-06,
"loss": 0.0017,
"step": 1151
},
{
"epoch": 3.449438202247191,
"grad_norm": 5.794363021850586,
"learning_rate": 3.0810810810810817e-06,
"loss": 0.0158,
"step": 1152
},
{
"epoch": 3.452434456928839,
"grad_norm": 0.17519600689411163,
"learning_rate": 3.0750750750750754e-06,
"loss": 0.0014,
"step": 1153
},
{
"epoch": 3.4554307116104868,
"grad_norm": 0.11729874461889267,
"learning_rate": 3.0690690690690695e-06,
"loss": 0.0012,
"step": 1154
},
{
"epoch": 3.458426966292135,
"grad_norm": 0.20305968821048737,
"learning_rate": 3.063063063063063e-06,
"loss": 0.0016,
"step": 1155
},
{
"epoch": 3.4614232209737827,
"grad_norm": 13.6810941696167,
"learning_rate": 3.0570570570570573e-06,
"loss": 0.1434,
"step": 1156
},
{
"epoch": 3.464419475655431,
"grad_norm": 0.112913116812706,
"learning_rate": 3.051051051051051e-06,
"loss": 0.0012,
"step": 1157
},
{
"epoch": 3.4674157303370787,
"grad_norm": 0.40983065962791443,
"learning_rate": 3.045045045045045e-06,
"loss": 0.0021,
"step": 1158
},
{
"epoch": 3.4704119850187265,
"grad_norm": 0.36891645193099976,
"learning_rate": 3.0390390390390396e-06,
"loss": 0.0019,
"step": 1159
},
{
"epoch": 3.4734082397003747,
"grad_norm": 37.0242805480957,
"learning_rate": 3.0330330330330333e-06,
"loss": 0.0663,
"step": 1160
},
{
"epoch": 3.4764044943820225,
"grad_norm": 0.4011613130569458,
"learning_rate": 3.0270270270270274e-06,
"loss": 0.0022,
"step": 1161
},
{
"epoch": 3.4794007490636703,
"grad_norm": 0.0622095912694931,
"learning_rate": 3.021021021021021e-06,
"loss": 0.0007,
"step": 1162
},
{
"epoch": 3.4823970037453185,
"grad_norm": 0.2718900442123413,
"learning_rate": 3.0150150150150152e-06,
"loss": 0.0011,
"step": 1163
},
{
"epoch": 3.4853932584269662,
"grad_norm": 10.158575057983398,
"learning_rate": 3.009009009009009e-06,
"loss": 0.0113,
"step": 1164
},
{
"epoch": 3.488389513108614,
"grad_norm": 34.16524124145508,
"learning_rate": 3.0030030030030034e-06,
"loss": 0.1656,
"step": 1165
},
{
"epoch": 3.491385767790262,
"grad_norm": 21.019149780273438,
"learning_rate": 2.996996996996997e-06,
"loss": 0.0156,
"step": 1166
},
{
"epoch": 3.49438202247191,
"grad_norm": 8.08090877532959,
"learning_rate": 2.9909909909909912e-06,
"loss": 0.1539,
"step": 1167
},
{
"epoch": 3.497378277153558,
"grad_norm": 0.28100255131721497,
"learning_rate": 2.9849849849849854e-06,
"loss": 0.0013,
"step": 1168
},
{
"epoch": 3.500374531835206,
"grad_norm": 11.16100788116455,
"learning_rate": 2.978978978978979e-06,
"loss": 0.0329,
"step": 1169
},
{
"epoch": 3.5033707865168537,
"grad_norm": 13.391993522644043,
"learning_rate": 2.9729729729729736e-06,
"loss": 0.1595,
"step": 1170
},
{
"epoch": 3.506367041198502,
"grad_norm": 199.873046875,
"learning_rate": 2.9669669669669673e-06,
"loss": 0.0179,
"step": 1171
},
{
"epoch": 3.5093632958801497,
"grad_norm": 0.5481918454170227,
"learning_rate": 2.9609609609609614e-06,
"loss": 0.0027,
"step": 1172
},
{
"epoch": 3.512359550561798,
"grad_norm": 0.3004973828792572,
"learning_rate": 2.954954954954955e-06,
"loss": 0.0015,
"step": 1173
},
{
"epoch": 3.5153558052434457,
"grad_norm": 0.07423954457044601,
"learning_rate": 2.948948948948949e-06,
"loss": 0.0007,
"step": 1174
},
{
"epoch": 3.5183520599250935,
"grad_norm": 0.15803757309913635,
"learning_rate": 2.942942942942943e-06,
"loss": 0.0012,
"step": 1175
},
{
"epoch": 3.5213483146067417,
"grad_norm": 0.08874224126338959,
"learning_rate": 2.936936936936937e-06,
"loss": 0.0008,
"step": 1176
},
{
"epoch": 3.5243445692883895,
"grad_norm": 2.6903014183044434,
"learning_rate": 2.9309309309309315e-06,
"loss": 0.0025,
"step": 1177
},
{
"epoch": 3.5273408239700377,
"grad_norm": 18.532550811767578,
"learning_rate": 2.924924924924925e-06,
"loss": 0.1481,
"step": 1178
},
{
"epoch": 3.5303370786516854,
"grad_norm": 1.5022221803665161,
"learning_rate": 2.9189189189189193e-06,
"loss": 0.0079,
"step": 1179
},
{
"epoch": 3.533333333333333,
"grad_norm": 4.695957183837891,
"learning_rate": 2.912912912912913e-06,
"loss": 0.0722,
"step": 1180
},
{
"epoch": 3.5363295880149814,
"grad_norm": 90.71923828125,
"learning_rate": 2.906906906906907e-06,
"loss": 0.1585,
"step": 1181
},
{
"epoch": 3.539325842696629,
"grad_norm": 0.32794615626335144,
"learning_rate": 2.9009009009009008e-06,
"loss": 0.0016,
"step": 1182
},
{
"epoch": 3.5423220973782774,
"grad_norm": 13.338924407958984,
"learning_rate": 2.8948948948948953e-06,
"loss": 0.1505,
"step": 1183
},
{
"epoch": 3.545318352059925,
"grad_norm": 4.8918585777282715,
"learning_rate": 2.888888888888889e-06,
"loss": 0.003,
"step": 1184
},
{
"epoch": 3.548314606741573,
"grad_norm": 0.2953185737133026,
"learning_rate": 2.882882882882883e-06,
"loss": 0.0014,
"step": 1185
},
{
"epoch": 3.5513108614232207,
"grad_norm": 15.525997161865234,
"learning_rate": 2.8768768768768772e-06,
"loss": 0.0163,
"step": 1186
},
{
"epoch": 3.554307116104869,
"grad_norm": 18.109050750732422,
"learning_rate": 2.870870870870871e-06,
"loss": 0.1467,
"step": 1187
},
{
"epoch": 3.5573033707865167,
"grad_norm": 25.510013580322266,
"learning_rate": 2.8648648648648654e-06,
"loss": 0.2551,
"step": 1188
},
{
"epoch": 3.560299625468165,
"grad_norm": 6.449516773223877,
"learning_rate": 2.858858858858859e-06,
"loss": 0.0027,
"step": 1189
},
{
"epoch": 3.5632958801498127,
"grad_norm": 13.221821784973145,
"learning_rate": 2.8528528528528532e-06,
"loss": 0.3049,
"step": 1190
},
{
"epoch": 3.5662921348314605,
"grad_norm": 0.14982223510742188,
"learning_rate": 2.846846846846847e-06,
"loss": 0.0011,
"step": 1191
},
{
"epoch": 3.5692883895131087,
"grad_norm": 28.092092514038086,
"learning_rate": 2.840840840840841e-06,
"loss": 0.2558,
"step": 1192
},
{
"epoch": 3.5722846441947564,
"grad_norm": 43.792537689208984,
"learning_rate": 2.8348348348348347e-06,
"loss": 0.1863,
"step": 1193
},
{
"epoch": 3.5752808988764047,
"grad_norm": 0.21688073873519897,
"learning_rate": 2.828828828828829e-06,
"loss": 0.0014,
"step": 1194
},
{
"epoch": 3.5782771535580524,
"grad_norm": 0.12059924006462097,
"learning_rate": 2.8228228228228234e-06,
"loss": 0.001,
"step": 1195
},
{
"epoch": 3.5812734082397,
"grad_norm": 0.19299505650997162,
"learning_rate": 2.816816816816817e-06,
"loss": 0.0011,
"step": 1196
},
{
"epoch": 3.5842696629213484,
"grad_norm": 12.536391258239746,
"learning_rate": 2.810810810810811e-06,
"loss": 0.0137,
"step": 1197
},
{
"epoch": 3.587265917602996,
"grad_norm": 0.033527225255966187,
"learning_rate": 2.804804804804805e-06,
"loss": 0.0006,
"step": 1198
},
{
"epoch": 3.5902621722846444,
"grad_norm": 0.07377205044031143,
"learning_rate": 2.798798798798799e-06,
"loss": 0.0009,
"step": 1199
},
{
"epoch": 3.593258426966292,
"grad_norm": 1.6257638931274414,
"learning_rate": 2.7927927927927926e-06,
"loss": 0.0017,
"step": 1200
},
{
"epoch": 3.59625468164794,
"grad_norm": 0.12966857850551605,
"learning_rate": 2.786786786786787e-06,
"loss": 0.0011,
"step": 1201
},
{
"epoch": 3.599250936329588,
"grad_norm": 12.588142395019531,
"learning_rate": 2.780780780780781e-06,
"loss": 0.1614,
"step": 1202
},
{
"epoch": 3.602247191011236,
"grad_norm": 0.5331131219863892,
"learning_rate": 2.774774774774775e-06,
"loss": 0.0035,
"step": 1203
},
{
"epoch": 3.605243445692884,
"grad_norm": 0.12895500659942627,
"learning_rate": 2.768768768768769e-06,
"loss": 0.0012,
"step": 1204
},
{
"epoch": 3.608239700374532,
"grad_norm": 0.1572701781988144,
"learning_rate": 2.7627627627627628e-06,
"loss": 0.0012,
"step": 1205
},
{
"epoch": 3.6112359550561797,
"grad_norm": 0.09796961396932602,
"learning_rate": 2.7567567567567573e-06,
"loss": 0.0009,
"step": 1206
},
{
"epoch": 3.6142322097378274,
"grad_norm": 0.7374432682991028,
"learning_rate": 2.750750750750751e-06,
"loss": 0.0026,
"step": 1207
},
{
"epoch": 3.6172284644194757,
"grad_norm": 0.14040535688400269,
"learning_rate": 2.744744744744745e-06,
"loss": 0.0006,
"step": 1208
},
{
"epoch": 3.620224719101124,
"grad_norm": 0.3786684274673462,
"learning_rate": 2.7387387387387388e-06,
"loss": 0.0016,
"step": 1209
},
{
"epoch": 3.6232209737827716,
"grad_norm": 7.29923677444458,
"learning_rate": 2.732732732732733e-06,
"loss": 0.2036,
"step": 1210
},
{
"epoch": 3.6262172284644194,
"grad_norm": 3.8888790607452393,
"learning_rate": 2.7267267267267274e-06,
"loss": 0.0053,
"step": 1211
},
{
"epoch": 3.629213483146067,
"grad_norm": 0.05211644619703293,
"learning_rate": 2.720720720720721e-06,
"loss": 0.0007,
"step": 1212
},
{
"epoch": 3.6322097378277154,
"grad_norm": 36.055294036865234,
"learning_rate": 2.7147147147147152e-06,
"loss": 0.3021,
"step": 1213
},
{
"epoch": 3.635205992509363,
"grad_norm": 7.528492450714111,
"learning_rate": 2.708708708708709e-06,
"loss": 0.0677,
"step": 1214
},
{
"epoch": 3.6382022471910114,
"grad_norm": 1.370915174484253,
"learning_rate": 2.702702702702703e-06,
"loss": 0.003,
"step": 1215
},
{
"epoch": 3.641198501872659,
"grad_norm": 0.09668157994747162,
"learning_rate": 2.6966966966966967e-06,
"loss": 0.0009,
"step": 1216
},
{
"epoch": 3.644194756554307,
"grad_norm": 0.2665283679962158,
"learning_rate": 2.690690690690691e-06,
"loss": 0.0017,
"step": 1217
},
{
"epoch": 3.647191011235955,
"grad_norm": 5.99869441986084,
"learning_rate": 2.6846846846846845e-06,
"loss": 0.0947,
"step": 1218
},
{
"epoch": 3.650187265917603,
"grad_norm": 0.8665609955787659,
"learning_rate": 2.678678678678679e-06,
"loss": 0.0017,
"step": 1219
},
{
"epoch": 3.653183520599251,
"grad_norm": 78.24776458740234,
"learning_rate": 2.672672672672673e-06,
"loss": 0.0377,
"step": 1220
},
{
"epoch": 3.656179775280899,
"grad_norm": 59.363826751708984,
"learning_rate": 2.666666666666667e-06,
"loss": 0.2971,
"step": 1221
},
{
"epoch": 3.6591760299625467,
"grad_norm": 25.839784622192383,
"learning_rate": 2.660660660660661e-06,
"loss": 0.3201,
"step": 1222
},
{
"epoch": 3.662172284644195,
"grad_norm": 8.989326477050781,
"learning_rate": 2.6546546546546546e-06,
"loss": 0.2436,
"step": 1223
},
{
"epoch": 3.6651685393258426,
"grad_norm": 25.314912796020508,
"learning_rate": 2.648648648648649e-06,
"loss": 0.0178,
"step": 1224
},
{
"epoch": 3.668164794007491,
"grad_norm": 0.3571781516075134,
"learning_rate": 2.642642642642643e-06,
"loss": 0.0017,
"step": 1225
},
{
"epoch": 3.6711610486891386,
"grad_norm": 0.15881821513175964,
"learning_rate": 2.636636636636637e-06,
"loss": 0.001,
"step": 1226
},
{
"epoch": 3.6741573033707864,
"grad_norm": 0.06995400786399841,
"learning_rate": 2.6306306306306306e-06,
"loss": 0.0008,
"step": 1227
},
{
"epoch": 3.6771535580524346,
"grad_norm": 1.3104853630065918,
"learning_rate": 2.6246246246246248e-06,
"loss": 0.0025,
"step": 1228
},
{
"epoch": 3.6801498127340824,
"grad_norm": 0.11513541638851166,
"learning_rate": 2.6186186186186193e-06,
"loss": 0.0009,
"step": 1229
},
{
"epoch": 3.6831460674157306,
"grad_norm": 0.1014426201581955,
"learning_rate": 2.612612612612613e-06,
"loss": 0.0008,
"step": 1230
},
{
"epoch": 3.6861423220973784,
"grad_norm": 5.059184551239014,
"learning_rate": 2.606606606606607e-06,
"loss": 0.0538,
"step": 1231
},
{
"epoch": 3.689138576779026,
"grad_norm": 0.059635184705257416,
"learning_rate": 2.6006006006006008e-06,
"loss": 0.0007,
"step": 1232
},
{
"epoch": 3.692134831460674,
"grad_norm": 0.3208077549934387,
"learning_rate": 2.594594594594595e-06,
"loss": 0.0019,
"step": 1233
},
{
"epoch": 3.695131086142322,
"grad_norm": 10.169331550598145,
"learning_rate": 2.5885885885885886e-06,
"loss": 0.3249,
"step": 1234
},
{
"epoch": 3.69812734082397,
"grad_norm": 0.4259358048439026,
"learning_rate": 2.5825825825825827e-06,
"loss": 0.0011,
"step": 1235
},
{
"epoch": 3.701123595505618,
"grad_norm": 9.01215648651123,
"learning_rate": 2.5765765765765764e-06,
"loss": 0.1755,
"step": 1236
},
{
"epoch": 3.704119850187266,
"grad_norm": 24.161672592163086,
"learning_rate": 2.570570570570571e-06,
"loss": 0.1799,
"step": 1237
},
{
"epoch": 3.7071161048689136,
"grad_norm": 0.2945901155471802,
"learning_rate": 2.564564564564565e-06,
"loss": 0.0011,
"step": 1238
},
{
"epoch": 3.710112359550562,
"grad_norm": 0.33426448702812195,
"learning_rate": 2.5585585585585587e-06,
"loss": 0.0015,
"step": 1239
},
{
"epoch": 3.7131086142322096,
"grad_norm": 52.860809326171875,
"learning_rate": 2.552552552552553e-06,
"loss": 0.0502,
"step": 1240
},
{
"epoch": 3.716104868913858,
"grad_norm": 0.05555523559451103,
"learning_rate": 2.5465465465465465e-06,
"loss": 0.0006,
"step": 1241
},
{
"epoch": 3.7191011235955056,
"grad_norm": 0.3306257128715515,
"learning_rate": 2.540540540540541e-06,
"loss": 0.0015,
"step": 1242
},
{
"epoch": 3.7220973782771534,
"grad_norm": 7.327410697937012,
"learning_rate": 2.5345345345345347e-06,
"loss": 0.0058,
"step": 1243
},
{
"epoch": 3.7250936329588016,
"grad_norm": 18.794662475585938,
"learning_rate": 2.528528528528529e-06,
"loss": 0.0064,
"step": 1244
},
{
"epoch": 3.7280898876404494,
"grad_norm": 32.82272720336914,
"learning_rate": 2.5225225225225225e-06,
"loss": 0.2091,
"step": 1245
},
{
"epoch": 3.7310861423220976,
"grad_norm": 0.1005844995379448,
"learning_rate": 2.5165165165165166e-06,
"loss": 0.001,
"step": 1246
},
{
"epoch": 3.7340823970037453,
"grad_norm": 0.7663953900337219,
"learning_rate": 2.510510510510511e-06,
"loss": 0.0052,
"step": 1247
},
{
"epoch": 3.737078651685393,
"grad_norm": 0.6252509355545044,
"learning_rate": 2.504504504504505e-06,
"loss": 0.0013,
"step": 1248
},
{
"epoch": 3.7400749063670413,
"grad_norm": 0.9177179932594299,
"learning_rate": 2.4984984984984985e-06,
"loss": 0.013,
"step": 1249
},
{
"epoch": 3.743071161048689,
"grad_norm": 32.366127014160156,
"learning_rate": 2.4924924924924926e-06,
"loss": 0.2137,
"step": 1250
},
{
"epoch": 3.7460674157303373,
"grad_norm": 0.07178188860416412,
"learning_rate": 2.4864864864864867e-06,
"loss": 0.0007,
"step": 1251
},
{
"epoch": 3.749063670411985,
"grad_norm": 7.637673854827881,
"learning_rate": 2.480480480480481e-06,
"loss": 0.0474,
"step": 1252
},
{
"epoch": 3.752059925093633,
"grad_norm": 0.0994248017668724,
"learning_rate": 2.474474474474475e-06,
"loss": 0.0009,
"step": 1253
},
{
"epoch": 3.755056179775281,
"grad_norm": 0.1058611124753952,
"learning_rate": 2.4684684684684686e-06,
"loss": 0.0008,
"step": 1254
},
{
"epoch": 3.758052434456929,
"grad_norm": 0.055875182151794434,
"learning_rate": 2.4624624624624628e-06,
"loss": 0.0006,
"step": 1255
},
{
"epoch": 3.761048689138577,
"grad_norm": 0.04026207700371742,
"learning_rate": 2.4564564564564564e-06,
"loss": 0.0006,
"step": 1256
},
{
"epoch": 3.764044943820225,
"grad_norm": 20.685853958129883,
"learning_rate": 2.4504504504504506e-06,
"loss": 0.1561,
"step": 1257
},
{
"epoch": 3.7670411985018726,
"grad_norm": 0.8278095722198486,
"learning_rate": 2.4444444444444447e-06,
"loss": 0.0032,
"step": 1258
},
{
"epoch": 3.7700374531835203,
"grad_norm": 0.13282153010368347,
"learning_rate": 2.4384384384384383e-06,
"loss": 0.0008,
"step": 1259
},
{
"epoch": 3.7730337078651686,
"grad_norm": 0.24338458478450775,
"learning_rate": 2.432432432432433e-06,
"loss": 0.0016,
"step": 1260
},
{
"epoch": 3.7760299625468163,
"grad_norm": 0.05848381295800209,
"learning_rate": 2.4264264264264266e-06,
"loss": 0.0005,
"step": 1261
},
{
"epoch": 3.7790262172284645,
"grad_norm": 0.749457061290741,
"learning_rate": 2.4204204204204207e-06,
"loss": 0.0028,
"step": 1262
},
{
"epoch": 3.7820224719101123,
"grad_norm": 46.25499725341797,
"learning_rate": 2.414414414414415e-06,
"loss": 0.0224,
"step": 1263
},
{
"epoch": 3.78501872659176,
"grad_norm": 0.604515552520752,
"learning_rate": 2.4084084084084085e-06,
"loss": 0.0016,
"step": 1264
},
{
"epoch": 3.7880149812734083,
"grad_norm": 0.13347892463207245,
"learning_rate": 2.4024024024024026e-06,
"loss": 0.0011,
"step": 1265
},
{
"epoch": 3.791011235955056,
"grad_norm": 102.36054229736328,
"learning_rate": 2.3963963963963967e-06,
"loss": 0.0152,
"step": 1266
},
{
"epoch": 3.7940074906367043,
"grad_norm": 13.67664623260498,
"learning_rate": 2.3903903903903904e-06,
"loss": 0.2729,
"step": 1267
},
{
"epoch": 3.797003745318352,
"grad_norm": 0.3461918234825134,
"learning_rate": 2.3843843843843845e-06,
"loss": 0.0014,
"step": 1268
},
{
"epoch": 3.8,
"grad_norm": 0.16493171453475952,
"learning_rate": 2.3783783783783786e-06,
"loss": 0.001,
"step": 1269
},
{
"epoch": 3.802996254681648,
"grad_norm": 0.06612278521060944,
"learning_rate": 2.3723723723723727e-06,
"loss": 0.0008,
"step": 1270
},
{
"epoch": 3.805992509363296,
"grad_norm": 24.139698028564453,
"learning_rate": 2.366366366366367e-06,
"loss": 0.0803,
"step": 1271
},
{
"epoch": 3.808988764044944,
"grad_norm": 1.3810616731643677,
"learning_rate": 2.3603603603603605e-06,
"loss": 0.0016,
"step": 1272
},
{
"epoch": 3.811985018726592,
"grad_norm": 0.07393044978380203,
"learning_rate": 2.3543543543543546e-06,
"loss": 0.0008,
"step": 1273
},
{
"epoch": 3.8149812734082396,
"grad_norm": 0.26108217239379883,
"learning_rate": 2.3483483483483483e-06,
"loss": 0.0013,
"step": 1274
},
{
"epoch": 3.8179775280898878,
"grad_norm": 1.6452735662460327,
"learning_rate": 2.3423423423423424e-06,
"loss": 0.0011,
"step": 1275
},
{
"epoch": 3.8209737827715355,
"grad_norm": 14.377902030944824,
"learning_rate": 2.3363363363363365e-06,
"loss": 0.0184,
"step": 1276
},
{
"epoch": 3.8239700374531838,
"grad_norm": 12.000560760498047,
"learning_rate": 2.3303303303303302e-06,
"loss": 0.2862,
"step": 1277
},
{
"epoch": 3.8269662921348315,
"grad_norm": 79.2288818359375,
"learning_rate": 2.3243243243243247e-06,
"loss": 0.2978,
"step": 1278
},
{
"epoch": 3.8299625468164793,
"grad_norm": 0.21543991565704346,
"learning_rate": 2.3183183183183184e-06,
"loss": 0.0007,
"step": 1279
},
{
"epoch": 3.832958801498127,
"grad_norm": 283.70477294921875,
"learning_rate": 2.3123123123123125e-06,
"loss": 0.1213,
"step": 1280
},
{
"epoch": 3.8359550561797753,
"grad_norm": 10.034940719604492,
"learning_rate": 2.3063063063063067e-06,
"loss": 0.1786,
"step": 1281
},
{
"epoch": 3.8389513108614235,
"grad_norm": 0.6665639877319336,
"learning_rate": 2.3003003003003003e-06,
"loss": 0.0014,
"step": 1282
},
{
"epoch": 3.8419475655430713,
"grad_norm": 0.2588105797767639,
"learning_rate": 2.2942942942942944e-06,
"loss": 0.0009,
"step": 1283
},
{
"epoch": 3.844943820224719,
"grad_norm": 0.08212500810623169,
"learning_rate": 2.2882882882882886e-06,
"loss": 0.0007,
"step": 1284
},
{
"epoch": 3.847940074906367,
"grad_norm": 0.02389441430568695,
"learning_rate": 2.2822822822822822e-06,
"loss": 0.0004,
"step": 1285
},
{
"epoch": 3.850936329588015,
"grad_norm": 0.049602072685956955,
"learning_rate": 2.2762762762762764e-06,
"loss": 0.0006,
"step": 1286
},
{
"epoch": 3.853932584269663,
"grad_norm": 0.19609133899211884,
"learning_rate": 2.2702702702702705e-06,
"loss": 0.0012,
"step": 1287
},
{
"epoch": 3.856928838951311,
"grad_norm": 0.08416257798671722,
"learning_rate": 2.2642642642642646e-06,
"loss": 0.0008,
"step": 1288
},
{
"epoch": 3.8599250936329588,
"grad_norm": 0.19125483930110931,
"learning_rate": 2.2582582582582587e-06,
"loss": 0.0009,
"step": 1289
},
{
"epoch": 3.8629213483146065,
"grad_norm": 0.041926681995391846,
"learning_rate": 2.2522522522522524e-06,
"loss": 0.0005,
"step": 1290
},
{
"epoch": 3.8659176029962548,
"grad_norm": 20.296092987060547,
"learning_rate": 2.2462462462462465e-06,
"loss": 0.2466,
"step": 1291
},
{
"epoch": 3.8689138576779025,
"grad_norm": 0.6759893894195557,
"learning_rate": 2.2402402402402406e-06,
"loss": 0.0015,
"step": 1292
},
{
"epoch": 3.8719101123595507,
"grad_norm": 23.822509765625,
"learning_rate": 2.2342342342342343e-06,
"loss": 0.1403,
"step": 1293
},
{
"epoch": 3.8749063670411985,
"grad_norm": 8.756211280822754,
"learning_rate": 2.2282282282282284e-06,
"loss": 0.1108,
"step": 1294
},
{
"epoch": 3.8779026217228463,
"grad_norm": 0.20383726060390472,
"learning_rate": 2.222222222222222e-06,
"loss": 0.0007,
"step": 1295
},
{
"epoch": 3.8808988764044945,
"grad_norm": 14.42196273803711,
"learning_rate": 2.2162162162162166e-06,
"loss": 0.0044,
"step": 1296
},
{
"epoch": 3.8838951310861423,
"grad_norm": 0.06926296651363373,
"learning_rate": 2.2102102102102103e-06,
"loss": 0.0006,
"step": 1297
},
{
"epoch": 3.8868913857677905,
"grad_norm": 0.06584805250167847,
"learning_rate": 2.2042042042042044e-06,
"loss": 0.0006,
"step": 1298
},
{
"epoch": 3.8898876404494382,
"grad_norm": 12.623332977294922,
"learning_rate": 2.1981981981981985e-06,
"loss": 0.0212,
"step": 1299
},
{
"epoch": 3.892883895131086,
"grad_norm": 9.110832214355469,
"learning_rate": 2.192192192192192e-06,
"loss": 0.1787,
"step": 1300
},
{
"epoch": 3.8958801498127342,
"grad_norm": 0.860475480556488,
"learning_rate": 2.1861861861861863e-06,
"loss": 0.0029,
"step": 1301
},
{
"epoch": 3.898876404494382,
"grad_norm": 0.27404239773750305,
"learning_rate": 2.1801801801801804e-06,
"loss": 0.0007,
"step": 1302
},
{
"epoch": 3.90187265917603,
"grad_norm": 9.200570106506348,
"learning_rate": 2.174174174174174e-06,
"loss": 0.0071,
"step": 1303
},
{
"epoch": 3.904868913857678,
"grad_norm": 0.8322509527206421,
"learning_rate": 2.1681681681681682e-06,
"loss": 0.0015,
"step": 1304
},
{
"epoch": 3.9078651685393258,
"grad_norm": 1.297377586364746,
"learning_rate": 2.1621621621621623e-06,
"loss": 0.0059,
"step": 1305
},
{
"epoch": 3.9108614232209735,
"grad_norm": 16.307981491088867,
"learning_rate": 2.1561561561561564e-06,
"loss": 0.1893,
"step": 1306
},
{
"epoch": 3.9138576779026217,
"grad_norm": 14.265915870666504,
"learning_rate": 2.1501501501501505e-06,
"loss": 0.089,
"step": 1307
},
{
"epoch": 3.9168539325842695,
"grad_norm": 0.03952779248356819,
"learning_rate": 2.1441441441441442e-06,
"loss": 0.0005,
"step": 1308
},
{
"epoch": 3.9198501872659177,
"grad_norm": 27.469043731689453,
"learning_rate": 2.1381381381381383e-06,
"loss": 0.112,
"step": 1309
},
{
"epoch": 3.9228464419475655,
"grad_norm": 13.053864479064941,
"learning_rate": 2.1321321321321325e-06,
"loss": 0.003,
"step": 1310
},
{
"epoch": 3.9258426966292133,
"grad_norm": 0.07893949002027512,
"learning_rate": 2.126126126126126e-06,
"loss": 0.0006,
"step": 1311
},
{
"epoch": 3.9288389513108615,
"grad_norm": 89.36550903320312,
"learning_rate": 2.1201201201201203e-06,
"loss": 0.0763,
"step": 1312
},
{
"epoch": 3.9318352059925092,
"grad_norm": 20.414094924926758,
"learning_rate": 2.1141141141141144e-06,
"loss": 0.1954,
"step": 1313
},
{
"epoch": 3.9348314606741575,
"grad_norm": 0.20142106711864471,
"learning_rate": 2.1081081081081085e-06,
"loss": 0.0008,
"step": 1314
},
{
"epoch": 3.9378277153558052,
"grad_norm": 9.736869812011719,
"learning_rate": 2.102102102102102e-06,
"loss": 0.0663,
"step": 1315
},
{
"epoch": 3.940823970037453,
"grad_norm": 176.92698669433594,
"learning_rate": 2.0960960960960963e-06,
"loss": 0.082,
"step": 1316
},
{
"epoch": 3.943820224719101,
"grad_norm": 11.484830856323242,
"learning_rate": 2.0900900900900904e-06,
"loss": 0.2074,
"step": 1317
},
{
"epoch": 3.946816479400749,
"grad_norm": 18.220382690429688,
"learning_rate": 2.084084084084084e-06,
"loss": 0.043,
"step": 1318
},
{
"epoch": 3.949812734082397,
"grad_norm": 183.95484924316406,
"learning_rate": 2.078078078078078e-06,
"loss": 0.0971,
"step": 1319
},
{
"epoch": 3.952808988764045,
"grad_norm": 2.3032569885253906,
"learning_rate": 2.0720720720720723e-06,
"loss": 0.0062,
"step": 1320
},
{
"epoch": 3.9558052434456927,
"grad_norm": 0.24705463647842407,
"learning_rate": 2.066066066066066e-06,
"loss": 0.0011,
"step": 1321
},
{
"epoch": 3.958801498127341,
"grad_norm": 11.232757568359375,
"learning_rate": 2.06006006006006e-06,
"loss": 0.0855,
"step": 1322
},
{
"epoch": 3.9617977528089887,
"grad_norm": 0.07256613671779633,
"learning_rate": 2.054054054054054e-06,
"loss": 0.0007,
"step": 1323
},
{
"epoch": 3.964794007490637,
"grad_norm": 0.05835743993520737,
"learning_rate": 2.0480480480480483e-06,
"loss": 0.0006,
"step": 1324
},
{
"epoch": 3.9677902621722847,
"grad_norm": 8.296839714050293,
"learning_rate": 2.0420420420420424e-06,
"loss": 0.1819,
"step": 1325
},
{
"epoch": 3.9707865168539325,
"grad_norm": 0.13280335068702698,
"learning_rate": 2.036036036036036e-06,
"loss": 0.001,
"step": 1326
},
{
"epoch": 3.9737827715355807,
"grad_norm": 0.15633881092071533,
"learning_rate": 2.03003003003003e-06,
"loss": 0.0007,
"step": 1327
},
{
"epoch": 3.9767790262172285,
"grad_norm": 1.779038906097412,
"learning_rate": 2.0240240240240243e-06,
"loss": 0.0027,
"step": 1328
},
{
"epoch": 3.9797752808988767,
"grad_norm": 0.23506179451942444,
"learning_rate": 2.018018018018018e-06,
"loss": 0.001,
"step": 1329
},
{
"epoch": 3.9827715355805244,
"grad_norm": 0.13996227085590363,
"learning_rate": 2.012012012012012e-06,
"loss": 0.0007,
"step": 1330
},
{
"epoch": 3.985767790262172,
"grad_norm": 0.7493147850036621,
"learning_rate": 2.0060060060060062e-06,
"loss": 0.0015,
"step": 1331
},
{
"epoch": 3.98876404494382,
"grad_norm": 1.240777611732483,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.0014,
"step": 1332
},
{
"epoch": 3.991760299625468,
"grad_norm": 0.06861300766468048,
"learning_rate": 1.9939939939939944e-06,
"loss": 0.0006,
"step": 1333
},
{
"epoch": 3.994756554307116,
"grad_norm": 19.111255645751953,
"learning_rate": 1.987987987987988e-06,
"loss": 0.1151,
"step": 1334
},
{
"epoch": 3.997752808988764,
"grad_norm": 39.03422164916992,
"learning_rate": 1.9819819819819822e-06,
"loss": 0.1441,
"step": 1335
},
{
"epoch": 4.0,
"grad_norm": 39.581024169921875,
"learning_rate": 1.975975975975976e-06,
"loss": 0.0454,
"step": 1336
},
{
"epoch": 4.0,
"eval_accuracy": 0.9609236234458259,
"eval_loss": 0.20387248694896698,
"eval_runtime": 21.4264,
"eval_samples_per_second": 26.276,
"eval_steps_per_second": 3.314,
"step": 1336
},
{
"epoch": 4.002996254681648,
"grad_norm": 1.1857681274414062,
"learning_rate": 1.96996996996997e-06,
"loss": 0.0034,
"step": 1337
},
{
"epoch": 4.0059925093632955,
"grad_norm": 6.235864639282227,
"learning_rate": 1.963963963963964e-06,
"loss": 0.2273,
"step": 1338
},
{
"epoch": 4.008988764044944,
"grad_norm": 0.11324425041675568,
"learning_rate": 1.957957957957958e-06,
"loss": 0.0009,
"step": 1339
},
{
"epoch": 4.011985018726592,
"grad_norm": 0.08317345380783081,
"learning_rate": 1.951951951951952e-06,
"loss": 0.0008,
"step": 1340
},
{
"epoch": 4.01498127340824,
"grad_norm": 0.13736970722675323,
"learning_rate": 1.945945945945946e-06,
"loss": 0.0008,
"step": 1341
},
{
"epoch": 4.0179775280898875,
"grad_norm": 27.349716186523438,
"learning_rate": 1.93993993993994e-06,
"loss": 0.1527,
"step": 1342
},
{
"epoch": 4.020973782771535,
"grad_norm": 8.773882865905762,
"learning_rate": 1.9339339339339343e-06,
"loss": 0.2085,
"step": 1343
},
{
"epoch": 4.023970037453184,
"grad_norm": 8.670924186706543,
"learning_rate": 1.927927927927928e-06,
"loss": 0.1434,
"step": 1344
},
{
"epoch": 4.026966292134832,
"grad_norm": 24.5421142578125,
"learning_rate": 1.921921921921922e-06,
"loss": 0.0524,
"step": 1345
},
{
"epoch": 4.0299625468164795,
"grad_norm": 90.18608856201172,
"learning_rate": 1.915915915915916e-06,
"loss": 0.0207,
"step": 1346
},
{
"epoch": 4.032958801498127,
"grad_norm": 0.14902891218662262,
"learning_rate": 1.90990990990991e-06,
"loss": 0.0011,
"step": 1347
},
{
"epoch": 4.035955056179775,
"grad_norm": 123.59732818603516,
"learning_rate": 1.903903903903904e-06,
"loss": 0.1631,
"step": 1348
},
{
"epoch": 4.038951310861423,
"grad_norm": 0.2144223004579544,
"learning_rate": 1.897897897897898e-06,
"loss": 0.0013,
"step": 1349
},
{
"epoch": 4.041947565543071,
"grad_norm": 0.5613626837730408,
"learning_rate": 1.8918918918918922e-06,
"loss": 0.0093,
"step": 1350
},
{
"epoch": 4.044943820224719,
"grad_norm": 58.52296447753906,
"learning_rate": 1.885885885885886e-06,
"loss": 0.1351,
"step": 1351
},
{
"epoch": 4.047940074906367,
"grad_norm": 0.1171746775507927,
"learning_rate": 1.87987987987988e-06,
"loss": 0.001,
"step": 1352
},
{
"epoch": 4.050936329588015,
"grad_norm": 1.8539955615997314,
"learning_rate": 1.873873873873874e-06,
"loss": 0.0068,
"step": 1353
},
{
"epoch": 4.0539325842696625,
"grad_norm": 0.4277438521385193,
"learning_rate": 1.867867867867868e-06,
"loss": 0.0032,
"step": 1354
},
{
"epoch": 4.056928838951311,
"grad_norm": 0.16552048921585083,
"learning_rate": 1.861861861861862e-06,
"loss": 0.0011,
"step": 1355
},
{
"epoch": 4.059925093632959,
"grad_norm": 40.794246673583984,
"learning_rate": 1.855855855855856e-06,
"loss": 0.1617,
"step": 1356
},
{
"epoch": 4.062921348314607,
"grad_norm": 0.38631919026374817,
"learning_rate": 1.84984984984985e-06,
"loss": 0.0012,
"step": 1357
},
{
"epoch": 4.0659176029962545,
"grad_norm": 0.1515495777130127,
"learning_rate": 1.8438438438438442e-06,
"loss": 0.0011,
"step": 1358
},
{
"epoch": 4.068913857677902,
"grad_norm": 0.19438177347183228,
"learning_rate": 1.8378378378378381e-06,
"loss": 0.0012,
"step": 1359
},
{
"epoch": 4.071910112359551,
"grad_norm": 7.856164455413818,
"learning_rate": 1.831831831831832e-06,
"loss": 0.2061,
"step": 1360
},
{
"epoch": 4.074906367041199,
"grad_norm": 2.1598777770996094,
"learning_rate": 1.825825825825826e-06,
"loss": 0.0045,
"step": 1361
},
{
"epoch": 4.0779026217228465,
"grad_norm": 0.08982834219932556,
"learning_rate": 1.81981981981982e-06,
"loss": 0.0007,
"step": 1362
},
{
"epoch": 4.080898876404494,
"grad_norm": 0.24650844931602478,
"learning_rate": 1.813813813813814e-06,
"loss": 0.0014,
"step": 1363
},
{
"epoch": 4.083895131086142,
"grad_norm": 0.15062983334064484,
"learning_rate": 1.8078078078078078e-06,
"loss": 0.0008,
"step": 1364
},
{
"epoch": 4.086891385767791,
"grad_norm": 0.030074920505285263,
"learning_rate": 1.801801801801802e-06,
"loss": 0.0004,
"step": 1365
},
{
"epoch": 4.089887640449438,
"grad_norm": 2.6458821296691895,
"learning_rate": 1.7957957957957958e-06,
"loss": 0.0532,
"step": 1366
},
{
"epoch": 4.092883895131086,
"grad_norm": 0.07172573357820511,
"learning_rate": 1.7897897897897902e-06,
"loss": 0.0007,
"step": 1367
},
{
"epoch": 4.095880149812734,
"grad_norm": 0.08971745520830154,
"learning_rate": 1.783783783783784e-06,
"loss": 0.0008,
"step": 1368
},
{
"epoch": 4.098876404494382,
"grad_norm": 0.06516806036233902,
"learning_rate": 1.777777777777778e-06,
"loss": 0.0006,
"step": 1369
},
{
"epoch": 4.10187265917603,
"grad_norm": 0.22777901589870453,
"learning_rate": 1.7717717717717719e-06,
"loss": 0.001,
"step": 1370
},
{
"epoch": 4.104868913857678,
"grad_norm": 0.10793136805295944,
"learning_rate": 1.765765765765766e-06,
"loss": 0.0008,
"step": 1371
},
{
"epoch": 4.107865168539326,
"grad_norm": 0.40473106503486633,
"learning_rate": 1.7597597597597599e-06,
"loss": 0.0027,
"step": 1372
},
{
"epoch": 4.110861423220974,
"grad_norm": 0.08503331989049911,
"learning_rate": 1.7537537537537538e-06,
"loss": 0.0006,
"step": 1373
},
{
"epoch": 4.1138576779026215,
"grad_norm": 1.3830671310424805,
"learning_rate": 1.7477477477477479e-06,
"loss": 0.0025,
"step": 1374
},
{
"epoch": 4.116853932584269,
"grad_norm": 0.08276256918907166,
"learning_rate": 1.7417417417417418e-06,
"loss": 0.0006,
"step": 1375
},
{
"epoch": 4.119850187265918,
"grad_norm": 0.0544453039765358,
"learning_rate": 1.735735735735736e-06,
"loss": 0.0006,
"step": 1376
},
{
"epoch": 4.122846441947566,
"grad_norm": 6.3252105712890625,
"learning_rate": 1.72972972972973e-06,
"loss": 0.0136,
"step": 1377
},
{
"epoch": 4.125842696629213,
"grad_norm": 0.47332167625427246,
"learning_rate": 1.7237237237237239e-06,
"loss": 0.004,
"step": 1378
},
{
"epoch": 4.128838951310861,
"grad_norm": 2.1658825874328613,
"learning_rate": 1.717717717717718e-06,
"loss": 0.0504,
"step": 1379
},
{
"epoch": 4.131835205992509,
"grad_norm": 8.90479564666748,
"learning_rate": 1.711711711711712e-06,
"loss": 0.0515,
"step": 1380
},
{
"epoch": 4.134831460674158,
"grad_norm": 0.4345548450946808,
"learning_rate": 1.7057057057057058e-06,
"loss": 0.003,
"step": 1381
},
{
"epoch": 4.137827715355805,
"grad_norm": 0.08130639791488647,
"learning_rate": 1.6996996996996997e-06,
"loss": 0.0008,
"step": 1382
},
{
"epoch": 4.140823970037453,
"grad_norm": 0.20989352464675903,
"learning_rate": 1.6936936936936938e-06,
"loss": 0.0008,
"step": 1383
},
{
"epoch": 4.143820224719101,
"grad_norm": 0.3344493508338928,
"learning_rate": 1.6876876876876877e-06,
"loss": 0.0014,
"step": 1384
},
{
"epoch": 4.146816479400749,
"grad_norm": 0.13763809204101562,
"learning_rate": 1.681681681681682e-06,
"loss": 0.0007,
"step": 1385
},
{
"epoch": 4.149812734082397,
"grad_norm": 0.17194509506225586,
"learning_rate": 1.675675675675676e-06,
"loss": 0.0007,
"step": 1386
},
{
"epoch": 4.152808988764045,
"grad_norm": 0.049271054565906525,
"learning_rate": 1.6696696696696698e-06,
"loss": 0.0005,
"step": 1387
},
{
"epoch": 4.155805243445693,
"grad_norm": 0.05265004560351372,
"learning_rate": 1.663663663663664e-06,
"loss": 0.0006,
"step": 1388
},
{
"epoch": 4.158801498127341,
"grad_norm": 0.3376707136631012,
"learning_rate": 1.6576576576576578e-06,
"loss": 0.0008,
"step": 1389
},
{
"epoch": 4.1617977528089884,
"grad_norm": 0.038651950657367706,
"learning_rate": 1.6516516516516517e-06,
"loss": 0.0005,
"step": 1390
},
{
"epoch": 4.164794007490637,
"grad_norm": 1.4642596244812012,
"learning_rate": 1.6456456456456456e-06,
"loss": 0.002,
"step": 1391
},
{
"epoch": 4.167790262172285,
"grad_norm": 17.094865798950195,
"learning_rate": 1.6396396396396397e-06,
"loss": 0.1198,
"step": 1392
},
{
"epoch": 4.170786516853933,
"grad_norm": 10.820752143859863,
"learning_rate": 1.6336336336336336e-06,
"loss": 0.224,
"step": 1393
},
{
"epoch": 4.17378277153558,
"grad_norm": 0.9100580811500549,
"learning_rate": 1.627627627627628e-06,
"loss": 0.0075,
"step": 1394
},
{
"epoch": 4.176779026217228,
"grad_norm": 11.763751029968262,
"learning_rate": 1.6216216216216219e-06,
"loss": 0.1434,
"step": 1395
},
{
"epoch": 4.179775280898877,
"grad_norm": 0.9637830257415771,
"learning_rate": 1.6156156156156157e-06,
"loss": 0.0014,
"step": 1396
},
{
"epoch": 4.182771535580525,
"grad_norm": 39.45403289794922,
"learning_rate": 1.6096096096096099e-06,
"loss": 0.4169,
"step": 1397
},
{
"epoch": 4.185767790262172,
"grad_norm": 0.4272950291633606,
"learning_rate": 1.6036036036036038e-06,
"loss": 0.0015,
"step": 1398
},
{
"epoch": 4.18876404494382,
"grad_norm": 0.11011972278356552,
"learning_rate": 1.5975975975975977e-06,
"loss": 0.0008,
"step": 1399
},
{
"epoch": 4.191760299625468,
"grad_norm": 6.655617713928223,
"learning_rate": 1.5915915915915916e-06,
"loss": 0.1092,
"step": 1400
},
{
"epoch": 4.194756554307116,
"grad_norm": 0.489435076713562,
"learning_rate": 1.5855855855855857e-06,
"loss": 0.0034,
"step": 1401
},
{
"epoch": 4.197752808988764,
"grad_norm": 42.1137809753418,
"learning_rate": 1.5795795795795796e-06,
"loss": 0.3307,
"step": 1402
},
{
"epoch": 4.200749063670412,
"grad_norm": 0.39769285917282104,
"learning_rate": 1.5735735735735739e-06,
"loss": 0.0016,
"step": 1403
},
{
"epoch": 4.20374531835206,
"grad_norm": 0.7645822763442993,
"learning_rate": 1.5675675675675678e-06,
"loss": 0.0048,
"step": 1404
},
{
"epoch": 4.206741573033708,
"grad_norm": 0.0968170017004013,
"learning_rate": 1.5615615615615617e-06,
"loss": 0.0009,
"step": 1405
},
{
"epoch": 4.209737827715355,
"grad_norm": 11.000382423400879,
"learning_rate": 1.5555555555555558e-06,
"loss": 0.1335,
"step": 1406
},
{
"epoch": 4.212734082397004,
"grad_norm": 0.23898589611053467,
"learning_rate": 1.5495495495495497e-06,
"loss": 0.001,
"step": 1407
},
{
"epoch": 4.215730337078652,
"grad_norm": 5.47330379486084,
"learning_rate": 1.5435435435435436e-06,
"loss": 0.0855,
"step": 1408
},
{
"epoch": 4.2187265917603,
"grad_norm": 0.07906647026538849,
"learning_rate": 1.5375375375375377e-06,
"loss": 0.0008,
"step": 1409
},
{
"epoch": 4.221722846441947,
"grad_norm": 5.534304141998291,
"learning_rate": 1.5315315315315316e-06,
"loss": 0.0049,
"step": 1410
},
{
"epoch": 4.224719101123595,
"grad_norm": 0.08531952649354935,
"learning_rate": 1.5255255255255255e-06,
"loss": 0.0006,
"step": 1411
},
{
"epoch": 4.227715355805244,
"grad_norm": 0.13503497838974,
"learning_rate": 1.5195195195195198e-06,
"loss": 0.0007,
"step": 1412
},
{
"epoch": 4.230711610486892,
"grad_norm": 0.20444563031196594,
"learning_rate": 1.5135135135135137e-06,
"loss": 0.0014,
"step": 1413
},
{
"epoch": 4.233707865168539,
"grad_norm": 0.20145131647586823,
"learning_rate": 1.5075075075075076e-06,
"loss": 0.0014,
"step": 1414
},
{
"epoch": 4.236704119850187,
"grad_norm": 0.07926031947135925,
"learning_rate": 1.5015015015015017e-06,
"loss": 0.0007,
"step": 1415
},
{
"epoch": 4.239700374531835,
"grad_norm": 0.05385823920369148,
"learning_rate": 1.4954954954954956e-06,
"loss": 0.0006,
"step": 1416
},
{
"epoch": 4.242696629213484,
"grad_norm": 4.707670211791992,
"learning_rate": 1.4894894894894895e-06,
"loss": 0.0808,
"step": 1417
},
{
"epoch": 4.245692883895131,
"grad_norm": 0.35531431436538696,
"learning_rate": 1.4834834834834836e-06,
"loss": 0.0017,
"step": 1418
},
{
"epoch": 4.248689138576779,
"grad_norm": 1.0318095684051514,
"learning_rate": 1.4774774774774775e-06,
"loss": 0.0017,
"step": 1419
},
{
"epoch": 4.251685393258427,
"grad_norm": 0.06952908635139465,
"learning_rate": 1.4714714714714714e-06,
"loss": 0.0007,
"step": 1420
},
{
"epoch": 4.254681647940075,
"grad_norm": 0.26402905583381653,
"learning_rate": 1.4654654654654657e-06,
"loss": 0.0017,
"step": 1421
},
{
"epoch": 4.257677902621722,
"grad_norm": 1.758280873298645,
"learning_rate": 1.4594594594594596e-06,
"loss": 0.0047,
"step": 1422
},
{
"epoch": 4.260674157303371,
"grad_norm": 4.719300746917725,
"learning_rate": 1.4534534534534535e-06,
"loss": 0.0755,
"step": 1423
},
{
"epoch": 4.263670411985019,
"grad_norm": 0.03589978441596031,
"learning_rate": 1.4474474474474477e-06,
"loss": 0.0005,
"step": 1424
},
{
"epoch": 4.266666666666667,
"grad_norm": 0.03675052523612976,
"learning_rate": 1.4414414414414416e-06,
"loss": 0.0005,
"step": 1425
},
{
"epoch": 4.269662921348314,
"grad_norm": 0.05167229846119881,
"learning_rate": 1.4354354354354354e-06,
"loss": 0.0004,
"step": 1426
},
{
"epoch": 4.272659176029962,
"grad_norm": 0.07443572580814362,
"learning_rate": 1.4294294294294296e-06,
"loss": 0.0005,
"step": 1427
},
{
"epoch": 4.275655430711611,
"grad_norm": 0.13271695375442505,
"learning_rate": 1.4234234234234235e-06,
"loss": 0.0009,
"step": 1428
},
{
"epoch": 4.278651685393259,
"grad_norm": 1.346874475479126,
"learning_rate": 1.4174174174174174e-06,
"loss": 0.0082,
"step": 1429
},
{
"epoch": 4.281647940074906,
"grad_norm": 0.051524121314287186,
"learning_rate": 1.4114114114114117e-06,
"loss": 0.0005,
"step": 1430
},
{
"epoch": 4.284644194756554,
"grad_norm": 0.3209790587425232,
"learning_rate": 1.4054054054054056e-06,
"loss": 0.0019,
"step": 1431
},
{
"epoch": 4.287640449438202,
"grad_norm": 8.087427139282227,
"learning_rate": 1.3993993993993995e-06,
"loss": 0.2492,
"step": 1432
},
{
"epoch": 4.2906367041198505,
"grad_norm": 37.19725799560547,
"learning_rate": 1.3933933933933936e-06,
"loss": 0.1393,
"step": 1433
},
{
"epoch": 4.293632958801498,
"grad_norm": 0.0404701828956604,
"learning_rate": 1.3873873873873875e-06,
"loss": 0.0005,
"step": 1434
},
{
"epoch": 4.296629213483146,
"grad_norm": 0.08821617066860199,
"learning_rate": 1.3813813813813814e-06,
"loss": 0.0007,
"step": 1435
},
{
"epoch": 4.299625468164794,
"grad_norm": 0.10452385246753693,
"learning_rate": 1.3753753753753755e-06,
"loss": 0.0007,
"step": 1436
},
{
"epoch": 4.302621722846442,
"grad_norm": 8.277549743652344,
"learning_rate": 1.3693693693693694e-06,
"loss": 0.1589,
"step": 1437
},
{
"epoch": 4.30561797752809,
"grad_norm": 24.767230987548828,
"learning_rate": 1.3633633633633637e-06,
"loss": 0.1681,
"step": 1438
},
{
"epoch": 4.308614232209738,
"grad_norm": 19.996173858642578,
"learning_rate": 1.3573573573573576e-06,
"loss": 0.0549,
"step": 1439
},
{
"epoch": 4.311610486891386,
"grad_norm": 0.043132588267326355,
"learning_rate": 1.3513513513513515e-06,
"loss": 0.0005,
"step": 1440
},
{
"epoch": 4.314606741573034,
"grad_norm": 11.817047119140625,
"learning_rate": 1.3453453453453454e-06,
"loss": 0.1785,
"step": 1441
},
{
"epoch": 4.317602996254681,
"grad_norm": 1.5767825841903687,
"learning_rate": 1.3393393393393395e-06,
"loss": 0.0034,
"step": 1442
},
{
"epoch": 4.320599250936329,
"grad_norm": 0.0721612200140953,
"learning_rate": 1.3333333333333334e-06,
"loss": 0.0006,
"step": 1443
},
{
"epoch": 4.323595505617978,
"grad_norm": 0.033141907304525375,
"learning_rate": 1.3273273273273273e-06,
"loss": 0.0004,
"step": 1444
},
{
"epoch": 4.3265917602996256,
"grad_norm": 0.2832183241844177,
"learning_rate": 1.3213213213213214e-06,
"loss": 0.0007,
"step": 1445
},
{
"epoch": 4.329588014981273,
"grad_norm": 0.037515345960855484,
"learning_rate": 1.3153153153153153e-06,
"loss": 0.0004,
"step": 1446
},
{
"epoch": 4.332584269662921,
"grad_norm": 0.13072574138641357,
"learning_rate": 1.3093093093093096e-06,
"loss": 0.0007,
"step": 1447
},
{
"epoch": 4.33558052434457,
"grad_norm": 0.17389430105686188,
"learning_rate": 1.3033033033033035e-06,
"loss": 0.0014,
"step": 1448
},
{
"epoch": 4.3385767790262175,
"grad_norm": 0.27292880415916443,
"learning_rate": 1.2972972972972974e-06,
"loss": 0.0015,
"step": 1449
},
{
"epoch": 4.341573033707865,
"grad_norm": 0.057960934937000275,
"learning_rate": 1.2912912912912913e-06,
"loss": 0.0006,
"step": 1450
},
{
"epoch": 4.344569288389513,
"grad_norm": 9.254718780517578,
"learning_rate": 1.2852852852852854e-06,
"loss": 0.179,
"step": 1451
},
{
"epoch": 4.347565543071161,
"grad_norm": 11.239828109741211,
"learning_rate": 1.2792792792792793e-06,
"loss": 0.204,
"step": 1452
},
{
"epoch": 4.350561797752809,
"grad_norm": 25.04166030883789,
"learning_rate": 1.2732732732732732e-06,
"loss": 0.1516,
"step": 1453
},
{
"epoch": 4.353558052434457,
"grad_norm": 4.792388916015625,
"learning_rate": 1.2672672672672674e-06,
"loss": 0.0527,
"step": 1454
},
{
"epoch": 4.356554307116105,
"grad_norm": 0.6955904960632324,
"learning_rate": 1.2612612612612613e-06,
"loss": 0.0028,
"step": 1455
},
{
"epoch": 4.359550561797753,
"grad_norm": 0.9539737701416016,
"learning_rate": 1.2552552552552556e-06,
"loss": 0.0018,
"step": 1456
},
{
"epoch": 4.362546816479401,
"grad_norm": 11.421793937683105,
"learning_rate": 1.2492492492492493e-06,
"loss": 0.0464,
"step": 1457
},
{
"epoch": 4.365543071161048,
"grad_norm": 0.18076151609420776,
"learning_rate": 1.2432432432432434e-06,
"loss": 0.0011,
"step": 1458
},
{
"epoch": 4.368539325842697,
"grad_norm": 4.3292436599731445,
"learning_rate": 1.2372372372372375e-06,
"loss": 0.0656,
"step": 1459
},
{
"epoch": 4.371535580524345,
"grad_norm": 21.925264358520508,
"learning_rate": 1.2312312312312314e-06,
"loss": 0.0919,
"step": 1460
},
{
"epoch": 4.3745318352059925,
"grad_norm": 0.10040438920259476,
"learning_rate": 1.2252252252252253e-06,
"loss": 0.0007,
"step": 1461
},
{
"epoch": 4.37752808988764,
"grad_norm": 0.13610941171646118,
"learning_rate": 1.2192192192192192e-06,
"loss": 0.0007,
"step": 1462
},
{
"epoch": 4.380524344569288,
"grad_norm": 0.09223781526088715,
"learning_rate": 1.2132132132132133e-06,
"loss": 0.0006,
"step": 1463
},
{
"epoch": 4.383520599250937,
"grad_norm": 5.989538669586182,
"learning_rate": 1.2072072072072074e-06,
"loss": 0.034,
"step": 1464
},
{
"epoch": 4.3865168539325845,
"grad_norm": 40.965965270996094,
"learning_rate": 1.2012012012012013e-06,
"loss": 0.1816,
"step": 1465
},
{
"epoch": 4.389513108614232,
"grad_norm": 0.09305235743522644,
"learning_rate": 1.1951951951951952e-06,
"loss": 0.0006,
"step": 1466
},
{
"epoch": 4.39250936329588,
"grad_norm": 5.541313171386719,
"learning_rate": 1.1891891891891893e-06,
"loss": 0.0143,
"step": 1467
},
{
"epoch": 4.395505617977528,
"grad_norm": 3.9298205375671387,
"learning_rate": 1.1831831831831834e-06,
"loss": 0.0435,
"step": 1468
},
{
"epoch": 4.3985018726591765,
"grad_norm": 8.565621376037598,
"learning_rate": 1.1771771771771773e-06,
"loss": 0.2326,
"step": 1469
},
{
"epoch": 4.401498127340824,
"grad_norm": 0.10688000172376633,
"learning_rate": 1.1711711711711712e-06,
"loss": 0.0007,
"step": 1470
},
{
"epoch": 4.404494382022472,
"grad_norm": 20.08437156677246,
"learning_rate": 1.1651651651651651e-06,
"loss": 0.0434,
"step": 1471
},
{
"epoch": 4.40749063670412,
"grad_norm": 0.11514133960008621,
"learning_rate": 1.1591591591591592e-06,
"loss": 0.0008,
"step": 1472
},
{
"epoch": 4.4104868913857675,
"grad_norm": 1.3515924215316772,
"learning_rate": 1.1531531531531533e-06,
"loss": 0.0109,
"step": 1473
},
{
"epoch": 4.413483146067415,
"grad_norm": 0.2089802622795105,
"learning_rate": 1.1471471471471472e-06,
"loss": 0.001,
"step": 1474
},
{
"epoch": 4.416479400749064,
"grad_norm": 0.05701066181063652,
"learning_rate": 1.1411411411411411e-06,
"loss": 0.0005,
"step": 1475
},
{
"epoch": 4.419475655430712,
"grad_norm": 12.600296020507812,
"learning_rate": 1.1351351351351352e-06,
"loss": 0.179,
"step": 1476
},
{
"epoch": 4.4224719101123595,
"grad_norm": 0.1585162729024887,
"learning_rate": 1.1291291291291293e-06,
"loss": 0.0009,
"step": 1477
},
{
"epoch": 4.425468164794007,
"grad_norm": 0.7637054324150085,
"learning_rate": 1.1231231231231232e-06,
"loss": 0.0018,
"step": 1478
},
{
"epoch": 4.428464419475655,
"grad_norm": 22.581663131713867,
"learning_rate": 1.1171171171171171e-06,
"loss": 0.0429,
"step": 1479
},
{
"epoch": 4.431460674157304,
"grad_norm": 0.14158785343170166,
"learning_rate": 1.111111111111111e-06,
"loss": 0.0011,
"step": 1480
},
{
"epoch": 4.4344569288389515,
"grad_norm": 21.828317642211914,
"learning_rate": 1.1051051051051051e-06,
"loss": 0.1492,
"step": 1481
},
{
"epoch": 4.437453183520599,
"grad_norm": 292.28729248046875,
"learning_rate": 1.0990990990990993e-06,
"loss": 0.0653,
"step": 1482
},
{
"epoch": 4.440449438202247,
"grad_norm": 15.90126895904541,
"learning_rate": 1.0930930930930932e-06,
"loss": 0.1522,
"step": 1483
},
{
"epoch": 4.443445692883895,
"grad_norm": 0.898688018321991,
"learning_rate": 1.087087087087087e-06,
"loss": 0.0023,
"step": 1484
},
{
"epoch": 4.4464419475655435,
"grad_norm": 0.09900274872779846,
"learning_rate": 1.0810810810810812e-06,
"loss": 0.0007,
"step": 1485
},
{
"epoch": 4.449438202247191,
"grad_norm": 1.6458231210708618,
"learning_rate": 1.0750750750750753e-06,
"loss": 0.001,
"step": 1486
},
{
"epoch": 4.452434456928839,
"grad_norm": 0.09181365370750427,
"learning_rate": 1.0690690690690692e-06,
"loss": 0.0007,
"step": 1487
},
{
"epoch": 4.455430711610487,
"grad_norm": 0.4765637516975403,
"learning_rate": 1.063063063063063e-06,
"loss": 0.0009,
"step": 1488
},
{
"epoch": 4.4584269662921345,
"grad_norm": 0.10589593648910522,
"learning_rate": 1.0570570570570572e-06,
"loss": 0.0007,
"step": 1489
},
{
"epoch": 4.461423220973783,
"grad_norm": 0.12367613613605499,
"learning_rate": 1.051051051051051e-06,
"loss": 0.0008,
"step": 1490
},
{
"epoch": 4.464419475655431,
"grad_norm": 0.08688446879386902,
"learning_rate": 1.0450450450450452e-06,
"loss": 0.0006,
"step": 1491
},
{
"epoch": 4.467415730337079,
"grad_norm": 0.260585218667984,
"learning_rate": 1.039039039039039e-06,
"loss": 0.0009,
"step": 1492
},
{
"epoch": 4.4704119850187265,
"grad_norm": 0.08936718106269836,
"learning_rate": 1.033033033033033e-06,
"loss": 0.0006,
"step": 1493
},
{
"epoch": 4.473408239700374,
"grad_norm": 0.25395792722702026,
"learning_rate": 1.027027027027027e-06,
"loss": 0.001,
"step": 1494
},
{
"epoch": 4.476404494382022,
"grad_norm": 0.14581674337387085,
"learning_rate": 1.0210210210210212e-06,
"loss": 0.0007,
"step": 1495
},
{
"epoch": 4.479400749063671,
"grad_norm": 0.8298152685165405,
"learning_rate": 1.015015015015015e-06,
"loss": 0.0046,
"step": 1496
},
{
"epoch": 4.4823970037453185,
"grad_norm": 3.017728328704834,
"learning_rate": 1.009009009009009e-06,
"loss": 0.001,
"step": 1497
},
{
"epoch": 4.485393258426966,
"grad_norm": 8.733826637268066,
"learning_rate": 1.0030030030030031e-06,
"loss": 0.1548,
"step": 1498
},
{
"epoch": 4.488389513108614,
"grad_norm": 0.06368312984704971,
"learning_rate": 9.969969969969972e-07,
"loss": 0.0005,
"step": 1499
},
{
"epoch": 4.491385767790262,
"grad_norm": 2.6524369716644287,
"learning_rate": 9.909909909909911e-07,
"loss": 0.044,
"step": 1500
},
{
"epoch": 4.49438202247191,
"grad_norm": 0.05978919938206673,
"learning_rate": 9.84984984984985e-07,
"loss": 0.0005,
"step": 1501
},
{
"epoch": 4.497378277153558,
"grad_norm": 0.6960961222648621,
"learning_rate": 9.78978978978979e-07,
"loss": 0.001,
"step": 1502
},
{
"epoch": 4.500374531835206,
"grad_norm": 0.06039566174149513,
"learning_rate": 9.72972972972973e-07,
"loss": 0.0005,
"step": 1503
},
{
"epoch": 4.503370786516854,
"grad_norm": 17.048011779785156,
"learning_rate": 9.669669669669671e-07,
"loss": 0.1489,
"step": 1504
},
{
"epoch": 4.5063670411985015,
"grad_norm": 0.0968472957611084,
"learning_rate": 9.60960960960961e-07,
"loss": 0.0007,
"step": 1505
},
{
"epoch": 4.50936329588015,
"grad_norm": 0.051448822021484375,
"learning_rate": 9.54954954954955e-07,
"loss": 0.0005,
"step": 1506
},
{
"epoch": 4.512359550561798,
"grad_norm": 0.08097102493047714,
"learning_rate": 9.48948948948949e-07,
"loss": 0.0006,
"step": 1507
},
{
"epoch": 4.515355805243446,
"grad_norm": 1.6621593236923218,
"learning_rate": 9.42942942942943e-07,
"loss": 0.0097,
"step": 1508
},
{
"epoch": 4.5183520599250935,
"grad_norm": 0.05666017532348633,
"learning_rate": 9.36936936936937e-07,
"loss": 0.0005,
"step": 1509
},
{
"epoch": 4.521348314606741,
"grad_norm": 0.046145763248205185,
"learning_rate": 9.30930930930931e-07,
"loss": 0.0004,
"step": 1510
},
{
"epoch": 4.52434456928839,
"grad_norm": 18.9801082611084,
"learning_rate": 9.24924924924925e-07,
"loss": 0.0118,
"step": 1511
},
{
"epoch": 4.527340823970038,
"grad_norm": 142.37619018554688,
"learning_rate": 9.189189189189191e-07,
"loss": 0.1597,
"step": 1512
},
{
"epoch": 4.5303370786516854,
"grad_norm": 0.4038632810115814,
"learning_rate": 9.12912912912913e-07,
"loss": 0.001,
"step": 1513
},
{
"epoch": 4.533333333333333,
"grad_norm": 0.10070820897817612,
"learning_rate": 9.06906906906907e-07,
"loss": 0.0007,
"step": 1514
},
{
"epoch": 4.536329588014981,
"grad_norm": 0.08822520077228546,
"learning_rate": 9.00900900900901e-07,
"loss": 0.0007,
"step": 1515
},
{
"epoch": 4.539325842696629,
"grad_norm": 0.07798154652118683,
"learning_rate": 8.948948948948951e-07,
"loss": 0.0007,
"step": 1516
},
{
"epoch": 4.542322097378277,
"grad_norm": 0.06188088282942772,
"learning_rate": 8.88888888888889e-07,
"loss": 0.0006,
"step": 1517
},
{
"epoch": 4.545318352059925,
"grad_norm": 0.06798405200242996,
"learning_rate": 8.82882882882883e-07,
"loss": 0.0006,
"step": 1518
},
{
"epoch": 4.548314606741573,
"grad_norm": 0.06251358240842819,
"learning_rate": 8.768768768768769e-07,
"loss": 0.0004,
"step": 1519
},
{
"epoch": 4.551310861423221,
"grad_norm": 0.11483646929264069,
"learning_rate": 8.708708708708709e-07,
"loss": 0.0008,
"step": 1520
},
{
"epoch": 4.554307116104869,
"grad_norm": 0.5001565217971802,
"learning_rate": 8.64864864864865e-07,
"loss": 0.0082,
"step": 1521
},
{
"epoch": 4.557303370786517,
"grad_norm": 11.353193283081055,
"learning_rate": 8.58858858858859e-07,
"loss": 0.1925,
"step": 1522
},
{
"epoch": 4.560299625468165,
"grad_norm": 0.09219124168157578,
"learning_rate": 8.528528528528529e-07,
"loss": 0.0006,
"step": 1523
},
{
"epoch": 4.563295880149813,
"grad_norm": 0.044498857110738754,
"learning_rate": 8.468468468468469e-07,
"loss": 0.0004,
"step": 1524
},
{
"epoch": 4.5662921348314605,
"grad_norm": 0.09277030825614929,
"learning_rate": 8.40840840840841e-07,
"loss": 0.0007,
"step": 1525
},
{
"epoch": 4.569288389513108,
"grad_norm": 11.887762069702148,
"learning_rate": 8.348348348348349e-07,
"loss": 0.1228,
"step": 1526
},
{
"epoch": 4.572284644194757,
"grad_norm": 3.9825165271759033,
"learning_rate": 8.288288288288289e-07,
"loss": 0.0391,
"step": 1527
},
{
"epoch": 4.575280898876405,
"grad_norm": 0.074421726167202,
"learning_rate": 8.228228228228228e-07,
"loss": 0.0005,
"step": 1528
},
{
"epoch": 4.578277153558052,
"grad_norm": 2.651801109313965,
"learning_rate": 8.168168168168168e-07,
"loss": 0.0024,
"step": 1529
},
{
"epoch": 4.5812734082397,
"grad_norm": 1.7624642848968506,
"learning_rate": 8.108108108108109e-07,
"loss": 0.0097,
"step": 1530
},
{
"epoch": 4.584269662921348,
"grad_norm": 0.4752442538738251,
"learning_rate": 8.048048048048049e-07,
"loss": 0.0008,
"step": 1531
},
{
"epoch": 4.587265917602997,
"grad_norm": 1.1370651721954346,
"learning_rate": 7.987987987987988e-07,
"loss": 0.0075,
"step": 1532
},
{
"epoch": 4.590262172284644,
"grad_norm": 0.04361056536436081,
"learning_rate": 7.927927927927928e-07,
"loss": 0.0004,
"step": 1533
},
{
"epoch": 4.593258426966292,
"grad_norm": 0.08027056604623795,
"learning_rate": 7.867867867867869e-07,
"loss": 0.0005,
"step": 1534
},
{
"epoch": 4.59625468164794,
"grad_norm": 0.10883516818284988,
"learning_rate": 7.807807807807808e-07,
"loss": 0.0008,
"step": 1535
},
{
"epoch": 4.599250936329588,
"grad_norm": 0.027541639283299446,
"learning_rate": 7.747747747747748e-07,
"loss": 0.0004,
"step": 1536
},
{
"epoch": 4.6022471910112355,
"grad_norm": 23.273799896240234,
"learning_rate": 7.687687687687688e-07,
"loss": 0.0141,
"step": 1537
},
{
"epoch": 4.605243445692884,
"grad_norm": 0.040743179619312286,
"learning_rate": 7.627627627627627e-07,
"loss": 0.0004,
"step": 1538
},
{
"epoch": 4.608239700374532,
"grad_norm": 0.8939301371574402,
"learning_rate": 7.567567567567569e-07,
"loss": 0.0024,
"step": 1539
},
{
"epoch": 4.61123595505618,
"grad_norm": 0.07921671867370605,
"learning_rate": 7.507507507507509e-07,
"loss": 0.0007,
"step": 1540
},
{
"epoch": 4.614232209737827,
"grad_norm": 0.08782128989696503,
"learning_rate": 7.447447447447448e-07,
"loss": 0.0005,
"step": 1541
},
{
"epoch": 4.617228464419476,
"grad_norm": 0.052603669464588165,
"learning_rate": 7.387387387387388e-07,
"loss": 0.0006,
"step": 1542
},
{
"epoch": 4.620224719101124,
"grad_norm": 0.06112679839134216,
"learning_rate": 7.327327327327329e-07,
"loss": 0.0005,
"step": 1543
},
{
"epoch": 4.623220973782772,
"grad_norm": 1.0430890321731567,
"learning_rate": 7.267267267267268e-07,
"loss": 0.0045,
"step": 1544
},
{
"epoch": 4.626217228464419,
"grad_norm": 8.98888111114502,
"learning_rate": 7.207207207207208e-07,
"loss": 0.2234,
"step": 1545
},
{
"epoch": 4.629213483146067,
"grad_norm": 1.58928382396698,
"learning_rate": 7.147147147147148e-07,
"loss": 0.0075,
"step": 1546
},
{
"epoch": 4.632209737827715,
"grad_norm": 0.9272421598434448,
"learning_rate": 7.087087087087087e-07,
"loss": 0.0015,
"step": 1547
},
{
"epoch": 4.635205992509364,
"grad_norm": 0.05971525236964226,
"learning_rate": 7.027027027027028e-07,
"loss": 0.0005,
"step": 1548
},
{
"epoch": 4.638202247191011,
"grad_norm": 0.05799536034464836,
"learning_rate": 6.966966966966968e-07,
"loss": 0.0004,
"step": 1549
},
{
"epoch": 4.641198501872659,
"grad_norm": 0.04386816546320915,
"learning_rate": 6.906906906906907e-07,
"loss": 0.0005,
"step": 1550
},
{
"epoch": 4.644194756554307,
"grad_norm": 70.71249389648438,
"learning_rate": 6.846846846846847e-07,
"loss": 0.0763,
"step": 1551
},
{
"epoch": 4.647191011235955,
"grad_norm": 6.4223408699035645,
"learning_rate": 6.786786786786788e-07,
"loss": 0.0508,
"step": 1552
},
{
"epoch": 4.650187265917603,
"grad_norm": 0.06315252929925919,
"learning_rate": 6.726726726726727e-07,
"loss": 0.0005,
"step": 1553
},
{
"epoch": 4.653183520599251,
"grad_norm": 0.07991476356983185,
"learning_rate": 6.666666666666667e-07,
"loss": 0.0005,
"step": 1554
},
{
"epoch": 4.656179775280899,
"grad_norm": 7.226046562194824,
"learning_rate": 6.606606606606607e-07,
"loss": 0.0488,
"step": 1555
},
{
"epoch": 4.659176029962547,
"grad_norm": 0.03190997242927551,
"learning_rate": 6.546546546546548e-07,
"loss": 0.0004,
"step": 1556
},
{
"epoch": 4.662172284644194,
"grad_norm": 0.0839492604136467,
"learning_rate": 6.486486486486487e-07,
"loss": 0.0007,
"step": 1557
},
{
"epoch": 4.665168539325843,
"grad_norm": 19.33370018005371,
"learning_rate": 6.426426426426427e-07,
"loss": 0.2913,
"step": 1558
},
{
"epoch": 4.668164794007491,
"grad_norm": 0.050525616854429245,
"learning_rate": 6.366366366366366e-07,
"loss": 0.0005,
"step": 1559
},
{
"epoch": 4.671161048689139,
"grad_norm": 0.2810042202472687,
"learning_rate": 6.306306306306306e-07,
"loss": 0.0016,
"step": 1560
},
{
"epoch": 4.674157303370786,
"grad_norm": 3.1624624729156494,
"learning_rate": 6.246246246246246e-07,
"loss": 0.0491,
"step": 1561
},
{
"epoch": 4.677153558052434,
"grad_norm": 0.9163456559181213,
"learning_rate": 6.186186186186187e-07,
"loss": 0.0065,
"step": 1562
},
{
"epoch": 4.680149812734083,
"grad_norm": 60.5709228515625,
"learning_rate": 6.126126126126126e-07,
"loss": 0.171,
"step": 1563
},
{
"epoch": 4.683146067415731,
"grad_norm": 20.184545516967773,
"learning_rate": 6.066066066066066e-07,
"loss": 0.0389,
"step": 1564
},
{
"epoch": 4.686142322097378,
"grad_norm": 0.021519027650356293,
"learning_rate": 6.006006006006006e-07,
"loss": 0.0003,
"step": 1565
},
{
"epoch": 4.689138576779026,
"grad_norm": 6.1736273765563965,
"learning_rate": 5.945945945945947e-07,
"loss": 0.0405,
"step": 1566
},
{
"epoch": 4.692134831460674,
"grad_norm": 0.04734325036406517,
"learning_rate": 5.885885885885887e-07,
"loss": 0.0004,
"step": 1567
},
{
"epoch": 4.695131086142322,
"grad_norm": 0.016578013077378273,
"learning_rate": 5.825825825825826e-07,
"loss": 0.0003,
"step": 1568
},
{
"epoch": 4.69812734082397,
"grad_norm": 12.243136405944824,
"learning_rate": 5.765765765765767e-07,
"loss": 0.1434,
"step": 1569
},
{
"epoch": 4.701123595505618,
"grad_norm": 21.25614356994629,
"learning_rate": 5.705705705705706e-07,
"loss": 0.0662,
"step": 1570
},
{
"epoch": 4.704119850187266,
"grad_norm": 0.03454562649130821,
"learning_rate": 5.645645645645647e-07,
"loss": 0.0004,
"step": 1571
},
{
"epoch": 4.707116104868914,
"grad_norm": 0.08108682930469513,
"learning_rate": 5.585585585585586e-07,
"loss": 0.0006,
"step": 1572
},
{
"epoch": 4.710112359550562,
"grad_norm": 22.30251693725586,
"learning_rate": 5.525525525525526e-07,
"loss": 0.2068,
"step": 1573
},
{
"epoch": 4.71310861423221,
"grad_norm": 0.02572021074593067,
"learning_rate": 5.465465465465466e-07,
"loss": 0.0004,
"step": 1574
},
{
"epoch": 4.716104868913858,
"grad_norm": 12.543827056884766,
"learning_rate": 5.405405405405406e-07,
"loss": 0.1658,
"step": 1575
},
{
"epoch": 4.719101123595506,
"grad_norm": 0.2886572778224945,
"learning_rate": 5.345345345345346e-07,
"loss": 0.002,
"step": 1576
},
{
"epoch": 4.722097378277153,
"grad_norm": 0.37491849064826965,
"learning_rate": 5.285285285285286e-07,
"loss": 0.0008,
"step": 1577
},
{
"epoch": 4.725093632958801,
"grad_norm": 4.240332126617432,
"learning_rate": 5.225225225225226e-07,
"loss": 0.0491,
"step": 1578
},
{
"epoch": 4.72808988764045,
"grad_norm": 2.1326327323913574,
"learning_rate": 5.165165165165165e-07,
"loss": 0.0039,
"step": 1579
},
{
"epoch": 4.731086142322098,
"grad_norm": 0.03157065063714981,
"learning_rate": 5.105105105105106e-07,
"loss": 0.0003,
"step": 1580
},
{
"epoch": 4.734082397003745,
"grad_norm": 11.946758270263672,
"learning_rate": 5.045045045045045e-07,
"loss": 0.199,
"step": 1581
},
{
"epoch": 4.737078651685393,
"grad_norm": 0.07092578709125519,
"learning_rate": 4.984984984984986e-07,
"loss": 0.0005,
"step": 1582
},
{
"epoch": 4.740074906367041,
"grad_norm": 0.13529425859451294,
"learning_rate": 4.924924924924925e-07,
"loss": 0.0007,
"step": 1583
},
{
"epoch": 4.7430711610486895,
"grad_norm": 0.05387164652347565,
"learning_rate": 4.864864864864865e-07,
"loss": 0.0005,
"step": 1584
},
{
"epoch": 4.746067415730337,
"grad_norm": 0.0637047290802002,
"learning_rate": 4.804804804804805e-07,
"loss": 0.0006,
"step": 1585
},
{
"epoch": 4.749063670411985,
"grad_norm": 2.3102166652679443,
"learning_rate": 4.744744744744745e-07,
"loss": 0.0522,
"step": 1586
},
{
"epoch": 4.752059925093633,
"grad_norm": 0.16145378351211548,
"learning_rate": 4.684684684684685e-07,
"loss": 0.0007,
"step": 1587
},
{
"epoch": 4.755056179775281,
"grad_norm": 0.15398642420768738,
"learning_rate": 4.624624624624625e-07,
"loss": 0.0008,
"step": 1588
},
{
"epoch": 4.758052434456928,
"grad_norm": 23.60244369506836,
"learning_rate": 4.564564564564565e-07,
"loss": 0.1416,
"step": 1589
},
{
"epoch": 4.761048689138577,
"grad_norm": 5.4439592361450195,
"learning_rate": 4.504504504504505e-07,
"loss": 0.2506,
"step": 1590
},
{
"epoch": 4.764044943820225,
"grad_norm": 28.074872970581055,
"learning_rate": 4.444444444444445e-07,
"loss": 0.2312,
"step": 1591
},
{
"epoch": 4.767041198501873,
"grad_norm": 0.30833926796913147,
"learning_rate": 4.3843843843843844e-07,
"loss": 0.0009,
"step": 1592
},
{
"epoch": 4.77003745318352,
"grad_norm": 0.06983055919408798,
"learning_rate": 4.324324324324325e-07,
"loss": 0.0006,
"step": 1593
},
{
"epoch": 4.773033707865169,
"grad_norm": 0.04961135983467102,
"learning_rate": 4.2642642642642645e-07,
"loss": 0.0004,
"step": 1594
},
{
"epoch": 4.776029962546817,
"grad_norm": 0.3669885993003845,
"learning_rate": 4.204204204204205e-07,
"loss": 0.0018,
"step": 1595
},
{
"epoch": 4.7790262172284645,
"grad_norm": 0.03794948756694794,
"learning_rate": 4.1441441441441446e-07,
"loss": 0.0004,
"step": 1596
},
{
"epoch": 4.782022471910112,
"grad_norm": 1.9154077768325806,
"learning_rate": 4.084084084084084e-07,
"loss": 0.0397,
"step": 1597
},
{
"epoch": 4.78501872659176,
"grad_norm": 0.4471288025379181,
"learning_rate": 4.0240240240240246e-07,
"loss": 0.0042,
"step": 1598
},
{
"epoch": 4.788014981273408,
"grad_norm": 43.2827033996582,
"learning_rate": 3.963963963963964e-07,
"loss": 0.1564,
"step": 1599
},
{
"epoch": 4.7910112359550565,
"grad_norm": 10.669339179992676,
"learning_rate": 3.903903903903904e-07,
"loss": 0.0663,
"step": 1600
},
{
"epoch": 4.794007490636704,
"grad_norm": 0.08896947652101517,
"learning_rate": 3.843843843843844e-07,
"loss": 0.0005,
"step": 1601
},
{
"epoch": 4.797003745318352,
"grad_norm": 25.632553100585938,
"learning_rate": 3.7837837837837843e-07,
"loss": 0.028,
"step": 1602
},
{
"epoch": 4.8,
"grad_norm": 0.10706740617752075,
"learning_rate": 3.723723723723724e-07,
"loss": 0.0008,
"step": 1603
},
{
"epoch": 4.802996254681648,
"grad_norm": 2.436537981033325,
"learning_rate": 3.6636636636636644e-07,
"loss": 0.006,
"step": 1604
},
{
"epoch": 4.805992509363296,
"grad_norm": 7.362716197967529,
"learning_rate": 3.603603603603604e-07,
"loss": 0.0199,
"step": 1605
},
{
"epoch": 4.808988764044944,
"grad_norm": 0.13014669716358185,
"learning_rate": 3.5435435435435434e-07,
"loss": 0.0007,
"step": 1606
},
{
"epoch": 4.811985018726592,
"grad_norm": 0.06362096965312958,
"learning_rate": 3.483483483483484e-07,
"loss": 0.0006,
"step": 1607
},
{
"epoch": 4.81498127340824,
"grad_norm": 13.201666831970215,
"learning_rate": 3.4234234234234235e-07,
"loss": 0.161,
"step": 1608
},
{
"epoch": 4.817977528089887,
"grad_norm": 0.024186957627534866,
"learning_rate": 3.3633633633633635e-07,
"loss": 0.0003,
"step": 1609
},
{
"epoch": 4.820973782771535,
"grad_norm": 0.07626641541719437,
"learning_rate": 3.3033033033033036e-07,
"loss": 0.0005,
"step": 1610
},
{
"epoch": 4.823970037453184,
"grad_norm": 0.08752576261758804,
"learning_rate": 3.2432432432432436e-07,
"loss": 0.0006,
"step": 1611
},
{
"epoch": 4.8269662921348315,
"grad_norm": 5.421555042266846,
"learning_rate": 3.183183183183183e-07,
"loss": 0.0076,
"step": 1612
},
{
"epoch": 4.829962546816479,
"grad_norm": 17.79936981201172,
"learning_rate": 3.123123123123123e-07,
"loss": 0.173,
"step": 1613
},
{
"epoch": 4.832958801498127,
"grad_norm": 1.3809971809387207,
"learning_rate": 3.063063063063063e-07,
"loss": 0.003,
"step": 1614
},
{
"epoch": 4.835955056179776,
"grad_norm": 0.04778476431965828,
"learning_rate": 3.003003003003003e-07,
"loss": 0.0005,
"step": 1615
},
{
"epoch": 4.8389513108614235,
"grad_norm": 2.6461575031280518,
"learning_rate": 2.9429429429429433e-07,
"loss": 0.0033,
"step": 1616
},
{
"epoch": 4.841947565543071,
"grad_norm": 0.057272329926490784,
"learning_rate": 2.8828828828828833e-07,
"loss": 0.0005,
"step": 1617
},
{
"epoch": 4.844943820224719,
"grad_norm": 0.9794812202453613,
"learning_rate": 2.8228228228228234e-07,
"loss": 0.0017,
"step": 1618
},
{
"epoch": 4.847940074906367,
"grad_norm": 14.971752166748047,
"learning_rate": 2.762762762762763e-07,
"loss": 0.1581,
"step": 1619
},
{
"epoch": 4.850936329588015,
"grad_norm": 0.6323643922805786,
"learning_rate": 2.702702702702703e-07,
"loss": 0.0044,
"step": 1620
},
{
"epoch": 4.853932584269663,
"grad_norm": 12.307567596435547,
"learning_rate": 2.642642642642643e-07,
"loss": 0.1463,
"step": 1621
},
{
"epoch": 4.856928838951311,
"grad_norm": 0.22150635719299316,
"learning_rate": 2.5825825825825825e-07,
"loss": 0.0007,
"step": 1622
},
{
"epoch": 4.859925093632959,
"grad_norm": 0.7072132229804993,
"learning_rate": 2.5225225225225225e-07,
"loss": 0.0033,
"step": 1623
},
{
"epoch": 4.8629213483146065,
"grad_norm": 0.1475359946489334,
"learning_rate": 2.4624624624624625e-07,
"loss": 0.0005,
"step": 1624
},
{
"epoch": 4.865917602996254,
"grad_norm": 0.7722948789596558,
"learning_rate": 2.4024024024024026e-07,
"loss": 0.009,
"step": 1625
},
{
"epoch": 4.868913857677903,
"grad_norm": 0.24325625598430634,
"learning_rate": 2.3423423423423426e-07,
"loss": 0.0008,
"step": 1626
},
{
"epoch": 4.871910112359551,
"grad_norm": 0.06290898472070694,
"learning_rate": 2.2822822822822824e-07,
"loss": 0.0005,
"step": 1627
},
{
"epoch": 4.8749063670411985,
"grad_norm": 0.05529392883181572,
"learning_rate": 2.2222222222222224e-07,
"loss": 0.0005,
"step": 1628
},
{
"epoch": 4.877902621722846,
"grad_norm": 0.12186314165592194,
"learning_rate": 2.1621621621621625e-07,
"loss": 0.0008,
"step": 1629
},
{
"epoch": 4.880898876404494,
"grad_norm": 0.26348569989204407,
"learning_rate": 2.1021021021021025e-07,
"loss": 0.0011,
"step": 1630
},
{
"epoch": 4.883895131086143,
"grad_norm": 5.453798294067383,
"learning_rate": 2.042042042042042e-07,
"loss": 0.0578,
"step": 1631
},
{
"epoch": 4.8868913857677905,
"grad_norm": 0.07181008160114288,
"learning_rate": 1.981981981981982e-07,
"loss": 0.0007,
"step": 1632
},
{
"epoch": 4.889887640449438,
"grad_norm": 0.05274451896548271,
"learning_rate": 1.921921921921922e-07,
"loss": 0.0005,
"step": 1633
},
{
"epoch": 4.892883895131086,
"grad_norm": 0.036867499351501465,
"learning_rate": 1.861861861861862e-07,
"loss": 0.0005,
"step": 1634
},
{
"epoch": 4.895880149812734,
"grad_norm": 0.24336205422878265,
"learning_rate": 1.801801801801802e-07,
"loss": 0.001,
"step": 1635
},
{
"epoch": 4.898876404494382,
"grad_norm": 1.4599392414093018,
"learning_rate": 1.741741741741742e-07,
"loss": 0.0071,
"step": 1636
},
{
"epoch": 4.90187265917603,
"grad_norm": 0.09582925587892532,
"learning_rate": 1.6816816816816818e-07,
"loss": 0.0005,
"step": 1637
},
{
"epoch": 4.904868913857678,
"grad_norm": 0.7571275234222412,
"learning_rate": 1.6216216216216218e-07,
"loss": 0.0117,
"step": 1638
},
{
"epoch": 4.907865168539326,
"grad_norm": 1.4339208602905273,
"learning_rate": 1.5615615615615616e-07,
"loss": 0.0016,
"step": 1639
},
{
"epoch": 4.9108614232209735,
"grad_norm": 14.157052993774414,
"learning_rate": 1.5015015015015016e-07,
"loss": 0.0134,
"step": 1640
},
{
"epoch": 4.913857677902621,
"grad_norm": 0.026463618502020836,
"learning_rate": 1.4414414414414417e-07,
"loss": 0.0004,
"step": 1641
},
{
"epoch": 4.91685393258427,
"grad_norm": 7.497573375701904,
"learning_rate": 1.3813813813813814e-07,
"loss": 0.0236,
"step": 1642
},
{
"epoch": 4.919850187265918,
"grad_norm": 0.03946645185351372,
"learning_rate": 1.3213213213213215e-07,
"loss": 0.0004,
"step": 1643
},
{
"epoch": 4.9228464419475655,
"grad_norm": 12.166119575500488,
"learning_rate": 1.2612612612612613e-07,
"loss": 0.0254,
"step": 1644
},
{
"epoch": 4.925842696629213,
"grad_norm": 0.10254757106304169,
"learning_rate": 1.2012012012012013e-07,
"loss": 0.0006,
"step": 1645
},
{
"epoch": 4.928838951310862,
"grad_norm": 0.023163557052612305,
"learning_rate": 1.1411411411411412e-07,
"loss": 0.0003,
"step": 1646
},
{
"epoch": 4.93183520599251,
"grad_norm": 0.23772574961185455,
"learning_rate": 1.0810810810810812e-07,
"loss": 0.0014,
"step": 1647
},
{
"epoch": 4.9348314606741575,
"grad_norm": 0.046774666756391525,
"learning_rate": 1.021021021021021e-07,
"loss": 0.0005,
"step": 1648
},
{
"epoch": 4.937827715355805,
"grad_norm": 14.182709693908691,
"learning_rate": 9.60960960960961e-08,
"loss": 0.1443,
"step": 1649
},
{
"epoch": 4.940823970037453,
"grad_norm": 5.33063268661499,
"learning_rate": 9.00900900900901e-08,
"loss": 0.0456,
"step": 1650
},
{
"epoch": 4.943820224719101,
"grad_norm": 2.2647621631622314,
"learning_rate": 8.408408408408409e-08,
"loss": 0.0025,
"step": 1651
},
{
"epoch": 4.946816479400749,
"grad_norm": 4.9730448722839355,
"learning_rate": 7.807807807807808e-08,
"loss": 0.0142,
"step": 1652
},
{
"epoch": 4.949812734082397,
"grad_norm": 6.086433410644531,
"learning_rate": 7.207207207207208e-08,
"loss": 0.0814,
"step": 1653
},
{
"epoch": 4.952808988764045,
"grad_norm": 1.8743048906326294,
"learning_rate": 6.606606606606607e-08,
"loss": 0.0462,
"step": 1654
},
{
"epoch": 4.955805243445693,
"grad_norm": 0.5199087858200073,
"learning_rate": 6.006006006006006e-08,
"loss": 0.0011,
"step": 1655
},
{
"epoch": 4.9588014981273405,
"grad_norm": 0.11391440033912659,
"learning_rate": 5.405405405405406e-08,
"loss": 0.0006,
"step": 1656
},
{
"epoch": 4.961797752808989,
"grad_norm": 0.0856897309422493,
"learning_rate": 4.804804804804805e-08,
"loss": 0.0006,
"step": 1657
},
{
"epoch": 4.964794007490637,
"grad_norm": 0.08336736261844635,
"learning_rate": 4.2042042042042044e-08,
"loss": 0.0006,
"step": 1658
},
{
"epoch": 4.967790262172285,
"grad_norm": 92.152099609375,
"learning_rate": 3.603603603603604e-08,
"loss": 0.1152,
"step": 1659
},
{
"epoch": 4.9707865168539325,
"grad_norm": 0.9015845656394958,
"learning_rate": 3.003003003003003e-08,
"loss": 0.0013,
"step": 1660
},
{
"epoch": 4.97378277153558,
"grad_norm": 2.5232093334198,
"learning_rate": 2.4024024024024027e-08,
"loss": 0.0463,
"step": 1661
},
{
"epoch": 4.976779026217228,
"grad_norm": 5.621363639831543,
"learning_rate": 1.801801801801802e-08,
"loss": 0.0573,
"step": 1662
},
{
"epoch": 4.979775280898877,
"grad_norm": 14.393972396850586,
"learning_rate": 1.2012012012012013e-08,
"loss": 0.0637,
"step": 1663
},
{
"epoch": 4.982771535580524,
"grad_norm": 2.26298451423645,
"learning_rate": 6.006006006006007e-09,
"loss": 0.0035,
"step": 1664
},
{
"epoch": 4.985767790262172,
"grad_norm": 1.5570151805877686,
"learning_rate": 0.0,
"loss": 0.009,
"step": 1665
},
{
"epoch": 4.985767790262172,
"eval_accuracy": 0.9591474245115453,
"eval_loss": 0.2113608717918396,
"eval_runtime": 21.1251,
"eval_samples_per_second": 26.651,
"eval_steps_per_second": 3.361,
"step": 1665
}
],
"logging_steps": 1,
"max_steps": 1665,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2253970492858368.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}
{
"best_global_step": 334,
"best_metric": 0.11946738511323929,
"best_model_checkpoint": "./albert-large-v2-checkpoints\\checkpoint-334",
"epoch": 4.985767790262172,
"eval_steps": 500,
"global_step": 1665,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00299625468164794,
"grad_norm": 14.905317306518555,
"learning_rate": 9.993993993993994e-06,
"loss": 0.7434,
"step": 1
},
{
"epoch": 0.00599250936329588,
"grad_norm": 14.136712074279785,
"learning_rate": 9.987987987987989e-06,
"loss": 0.6318,
"step": 2
},
{
"epoch": 0.008988764044943821,
"grad_norm": 14.850204467773438,
"learning_rate": 9.981981981981982e-06,
"loss": 0.6954,
"step": 3
},
{
"epoch": 0.01198501872659176,
"grad_norm": 14.774368286132812,
"learning_rate": 9.975975975975977e-06,
"loss": 0.6352,
"step": 4
},
{
"epoch": 0.0149812734082397,
"grad_norm": 14.21596908569336,
"learning_rate": 9.96996996996997e-06,
"loss": 0.5841,
"step": 5
},
{
"epoch": 0.017977528089887642,
"grad_norm": 11.977082252502441,
"learning_rate": 9.963963963963965e-06,
"loss": 0.5649,
"step": 6
},
{
"epoch": 0.020973782771535582,
"grad_norm": 15.396467208862305,
"learning_rate": 9.957957957957959e-06,
"loss": 0.6076,
"step": 7
},
{
"epoch": 0.02397003745318352,
"grad_norm": 14.53940486907959,
"learning_rate": 9.951951951951952e-06,
"loss": 0.5034,
"step": 8
},
{
"epoch": 0.02696629213483146,
"grad_norm": 11.599512100219727,
"learning_rate": 9.945945945945947e-06,
"loss": 0.4279,
"step": 9
},
{
"epoch": 0.0299625468164794,
"grad_norm": 32.86847686767578,
"learning_rate": 9.93993993993994e-06,
"loss": 0.467,
"step": 10
},
{
"epoch": 0.03295880149812734,
"grad_norm": 28.4532527923584,
"learning_rate": 9.933933933933935e-06,
"loss": 0.524,
"step": 11
},
{
"epoch": 0.035955056179775284,
"grad_norm": 44.46889877319336,
"learning_rate": 9.927927927927928e-06,
"loss": 0.5606,
"step": 12
},
{
"epoch": 0.03895131086142322,
"grad_norm": 21.121667861938477,
"learning_rate": 9.921921921921923e-06,
"loss": 0.5904,