COBOL-Coder-7B-Instruct / trainer_state.json
vananh0905's picture
Upload 15 files
890d301 verified
Raw
History Blame Contribute Delete
142 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 810,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0037157454714352067,
"grad_norm": 1.211179494857788,
"learning_rate": 3.7037037037037036e-07,
"loss": 0.65,
"step": 1
},
{
"epoch": 0.0074314909428704135,
"grad_norm": 1.2230112552642822,
"learning_rate": 7.407407407407407e-07,
"loss": 0.6472,
"step": 2
},
{
"epoch": 0.01114723641430562,
"grad_norm": 1.2186671495437622,
"learning_rate": 1.111111111111111e-06,
"loss": 0.6446,
"step": 3
},
{
"epoch": 0.014862981885740827,
"grad_norm": 1.2264225482940674,
"learning_rate": 1.4814814814814815e-06,
"loss": 0.6526,
"step": 4
},
{
"epoch": 0.018578727357176035,
"grad_norm": 1.1856335401535034,
"learning_rate": 1.8518518518518519e-06,
"loss": 0.6489,
"step": 5
},
{
"epoch": 0.02229447282861124,
"grad_norm": 1.1635355949401855,
"learning_rate": 2.222222222222222e-06,
"loss": 0.6521,
"step": 6
},
{
"epoch": 0.026010218300046448,
"grad_norm": 1.0501785278320312,
"learning_rate": 2.5925925925925925e-06,
"loss": 0.6281,
"step": 7
},
{
"epoch": 0.029725963771481654,
"grad_norm": 0.8198655247688293,
"learning_rate": 2.962962962962963e-06,
"loss": 0.6169,
"step": 8
},
{
"epoch": 0.03344170924291686,
"grad_norm": 0.7379683256149292,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.6042,
"step": 9
},
{
"epoch": 0.03715745471435207,
"grad_norm": 0.7059187889099121,
"learning_rate": 3.7037037037037037e-06,
"loss": 0.6045,
"step": 10
},
{
"epoch": 0.04087320018578727,
"grad_norm": 0.7011018395423889,
"learning_rate": 4.074074074074074e-06,
"loss": 0.5657,
"step": 11
},
{
"epoch": 0.04458894565722248,
"grad_norm": 0.7294642329216003,
"learning_rate": 4.444444444444444e-06,
"loss": 0.5744,
"step": 12
},
{
"epoch": 0.048304691128657685,
"grad_norm": 0.6476544737815857,
"learning_rate": 4.814814814814815e-06,
"loss": 0.5587,
"step": 13
},
{
"epoch": 0.052020436600092895,
"grad_norm": 0.5719444751739502,
"learning_rate": 5.185185185185185e-06,
"loss": 0.5629,
"step": 14
},
{
"epoch": 0.0557361820715281,
"grad_norm": 0.4757300019264221,
"learning_rate": 5.555555555555557e-06,
"loss": 0.5338,
"step": 15
},
{
"epoch": 0.05945192754296331,
"grad_norm": 0.4802500009536743,
"learning_rate": 5.925925925925926e-06,
"loss": 0.5262,
"step": 16
},
{
"epoch": 0.06316767301439852,
"grad_norm": 0.4405558705329895,
"learning_rate": 6.296296296296297e-06,
"loss": 0.5204,
"step": 17
},
{
"epoch": 0.06688341848583372,
"grad_norm": 0.3984779119491577,
"learning_rate": 6.666666666666667e-06,
"loss": 0.5152,
"step": 18
},
{
"epoch": 0.07059916395726892,
"grad_norm": 0.3693501651287079,
"learning_rate": 7.0370370370370375e-06,
"loss": 0.5167,
"step": 19
},
{
"epoch": 0.07431490942870414,
"grad_norm": 0.344971239566803,
"learning_rate": 7.4074074074074075e-06,
"loss": 0.5002,
"step": 20
},
{
"epoch": 0.07803065490013934,
"grad_norm": 0.35149791836738586,
"learning_rate": 7.77777777777778e-06,
"loss": 0.4958,
"step": 21
},
{
"epoch": 0.08174640037157455,
"grad_norm": 0.33035123348236084,
"learning_rate": 8.148148148148148e-06,
"loss": 0.4889,
"step": 22
},
{
"epoch": 0.08546214584300975,
"grad_norm": 0.2947096824645996,
"learning_rate": 8.518518518518519e-06,
"loss": 0.4769,
"step": 23
},
{
"epoch": 0.08917789131444497,
"grad_norm": 0.26928022503852844,
"learning_rate": 8.888888888888888e-06,
"loss": 0.4694,
"step": 24
},
{
"epoch": 0.09289363678588017,
"grad_norm": 0.25431105494499207,
"learning_rate": 9.25925925925926e-06,
"loss": 0.4827,
"step": 25
},
{
"epoch": 0.09660938225731537,
"grad_norm": 0.25912928581237793,
"learning_rate": 9.62962962962963e-06,
"loss": 0.4677,
"step": 26
},
{
"epoch": 0.10032512772875057,
"grad_norm": 0.23261705040931702,
"learning_rate": 1e-05,
"loss": 0.466,
"step": 27
},
{
"epoch": 0.10404087320018579,
"grad_norm": 0.21991899609565735,
"learning_rate": 1.037037037037037e-05,
"loss": 0.4602,
"step": 28
},
{
"epoch": 0.107756618671621,
"grad_norm": 0.20101113617420197,
"learning_rate": 1.0740740740740742e-05,
"loss": 0.4551,
"step": 29
},
{
"epoch": 0.1114723641430562,
"grad_norm": 0.19477994740009308,
"learning_rate": 1.1111111111111113e-05,
"loss": 0.4517,
"step": 30
},
{
"epoch": 0.11518810961449141,
"grad_norm": 0.19734297692775726,
"learning_rate": 1.1481481481481482e-05,
"loss": 0.4457,
"step": 31
},
{
"epoch": 0.11890385508592662,
"grad_norm": 0.19136592745780945,
"learning_rate": 1.1851851851851852e-05,
"loss": 0.4487,
"step": 32
},
{
"epoch": 0.12261960055736182,
"grad_norm": 0.19089332222938538,
"learning_rate": 1.2222222222222224e-05,
"loss": 0.4482,
"step": 33
},
{
"epoch": 0.12633534602879704,
"grad_norm": 0.18238893151283264,
"learning_rate": 1.2592592592592593e-05,
"loss": 0.4342,
"step": 34
},
{
"epoch": 0.13005109150023222,
"grad_norm": 0.16636449098587036,
"learning_rate": 1.2962962962962964e-05,
"loss": 0.4432,
"step": 35
},
{
"epoch": 0.13376683697166744,
"grad_norm": 0.15438637137413025,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.4462,
"step": 36
},
{
"epoch": 0.13748258244310266,
"grad_norm": 0.1491541564464569,
"learning_rate": 1.3703703703703706e-05,
"loss": 0.4394,
"step": 37
},
{
"epoch": 0.14119832791453785,
"grad_norm": 0.15826033055782318,
"learning_rate": 1.4074074074074075e-05,
"loss": 0.4327,
"step": 38
},
{
"epoch": 0.14491407338597306,
"grad_norm": 0.14914660155773163,
"learning_rate": 1.4444444444444446e-05,
"loss": 0.4236,
"step": 39
},
{
"epoch": 0.14862981885740828,
"grad_norm": 0.13645023107528687,
"learning_rate": 1.4814814814814815e-05,
"loss": 0.4274,
"step": 40
},
{
"epoch": 0.15234556432884347,
"grad_norm": 0.13193705677986145,
"learning_rate": 1.5185185185185187e-05,
"loss": 0.4308,
"step": 41
},
{
"epoch": 0.15606130980027869,
"grad_norm": 0.1360337734222412,
"learning_rate": 1.555555555555556e-05,
"loss": 0.4276,
"step": 42
},
{
"epoch": 0.15977705527171387,
"grad_norm": 0.13185858726501465,
"learning_rate": 1.5925925925925926e-05,
"loss": 0.427,
"step": 43
},
{
"epoch": 0.1634928007431491,
"grad_norm": 0.12475940585136414,
"learning_rate": 1.6296296296296297e-05,
"loss": 0.4239,
"step": 44
},
{
"epoch": 0.1672085462145843,
"grad_norm": 0.12000305205583572,
"learning_rate": 1.6666666666666667e-05,
"loss": 0.4084,
"step": 45
},
{
"epoch": 0.1709242916860195,
"grad_norm": 0.12439227849245071,
"learning_rate": 1.7037037037037038e-05,
"loss": 0.4205,
"step": 46
},
{
"epoch": 0.17464003715745471,
"grad_norm": 0.11812030524015427,
"learning_rate": 1.740740740740741e-05,
"loss": 0.4175,
"step": 47
},
{
"epoch": 0.17835578262888993,
"grad_norm": 0.11951964348554611,
"learning_rate": 1.7777777777777777e-05,
"loss": 0.4174,
"step": 48
},
{
"epoch": 0.18207152810032512,
"grad_norm": 0.11201340705156326,
"learning_rate": 1.814814814814815e-05,
"loss": 0.4164,
"step": 49
},
{
"epoch": 0.18578727357176034,
"grad_norm": 0.10615763068199158,
"learning_rate": 1.851851851851852e-05,
"loss": 0.405,
"step": 50
},
{
"epoch": 0.18950301904319555,
"grad_norm": 0.1137438714504242,
"learning_rate": 1.888888888888889e-05,
"loss": 0.4153,
"step": 51
},
{
"epoch": 0.19321876451463074,
"grad_norm": 0.11462141573429108,
"learning_rate": 1.925925925925926e-05,
"loss": 0.4036,
"step": 52
},
{
"epoch": 0.19693450998606596,
"grad_norm": 0.11151576787233353,
"learning_rate": 1.962962962962963e-05,
"loss": 0.4053,
"step": 53
},
{
"epoch": 0.20065025545750115,
"grad_norm": 0.11205246299505234,
"learning_rate": 2e-05,
"loss": 0.3984,
"step": 54
},
{
"epoch": 0.20436600092893636,
"grad_norm": 0.10914111137390137,
"learning_rate": 1.999995275372036e-05,
"loss": 0.3992,
"step": 55
},
{
"epoch": 0.20808174640037158,
"grad_norm": 0.10943379253149033,
"learning_rate": 1.9999811015327872e-05,
"loss": 0.4047,
"step": 56
},
{
"epoch": 0.21179749187180677,
"grad_norm": 0.10846959799528122,
"learning_rate": 1.9999574786161863e-05,
"loss": 0.3999,
"step": 57
},
{
"epoch": 0.215513237343242,
"grad_norm": 0.11512856185436249,
"learning_rate": 1.999924406845452e-05,
"loss": 0.4053,
"step": 58
},
{
"epoch": 0.2192289828146772,
"grad_norm": 0.09854571521282196,
"learning_rate": 1.9998818865330883e-05,
"loss": 0.3971,
"step": 59
},
{
"epoch": 0.2229447282861124,
"grad_norm": 0.10970833152532578,
"learning_rate": 1.99982991808088e-05,
"loss": 0.3962,
"step": 60
},
{
"epoch": 0.2266604737575476,
"grad_norm": 0.11050631105899811,
"learning_rate": 1.9997685019798913e-05,
"loss": 0.4065,
"step": 61
},
{
"epoch": 0.23037621922898283,
"grad_norm": 0.10118336975574493,
"learning_rate": 1.9996976388104573e-05,
"loss": 0.4011,
"step": 62
},
{
"epoch": 0.23409196470041801,
"grad_norm": 0.11157616972923279,
"learning_rate": 1.999617329242183e-05,
"loss": 0.3991,
"step": 63
},
{
"epoch": 0.23780771017185323,
"grad_norm": 0.10527008026838303,
"learning_rate": 1.9995275740339345e-05,
"loss": 0.3911,
"step": 64
},
{
"epoch": 0.24152345564328845,
"grad_norm": 0.10907680541276932,
"learning_rate": 1.999428374033831e-05,
"loss": 0.3958,
"step": 65
},
{
"epoch": 0.24523920111472364,
"grad_norm": 0.10424818843603134,
"learning_rate": 1.9993197301792383e-05,
"loss": 0.4019,
"step": 66
},
{
"epoch": 0.24895494658615885,
"grad_norm": 0.09746669232845306,
"learning_rate": 1.9992016434967612e-05,
"loss": 0.3886,
"step": 67
},
{
"epoch": 0.25267069205759407,
"grad_norm": 0.1136242151260376,
"learning_rate": 1.9990741151022302e-05,
"loss": 0.3934,
"step": 68
},
{
"epoch": 0.2563864375290293,
"grad_norm": 0.10507749021053314,
"learning_rate": 1.9989371462006938e-05,
"loss": 0.3962,
"step": 69
},
{
"epoch": 0.26010218300046445,
"grad_norm": 0.10580289363861084,
"learning_rate": 1.9987907380864063e-05,
"loss": 0.3934,
"step": 70
},
{
"epoch": 0.26381792847189967,
"grad_norm": 0.11454343050718307,
"learning_rate": 1.998634892142816e-05,
"loss": 0.4021,
"step": 71
},
{
"epoch": 0.2675336739433349,
"grad_norm": 0.11967692524194717,
"learning_rate": 1.99846960984255e-05,
"loss": 0.389,
"step": 72
},
{
"epoch": 0.2712494194147701,
"grad_norm": 0.10581792145967484,
"learning_rate": 1.9982948927474036e-05,
"loss": 0.3892,
"step": 73
},
{
"epoch": 0.2749651648862053,
"grad_norm": 0.12725721299648285,
"learning_rate": 1.9981107425083234e-05,
"loss": 0.3906,
"step": 74
},
{
"epoch": 0.2786809103576405,
"grad_norm": 0.12701508402824402,
"learning_rate": 1.9979171608653926e-05,
"loss": 0.3958,
"step": 75
},
{
"epoch": 0.2823966558290757,
"grad_norm": 0.09920006990432739,
"learning_rate": 1.9977141496478124e-05,
"loss": 0.3971,
"step": 76
},
{
"epoch": 0.2861124013005109,
"grad_norm": 0.11396557092666626,
"learning_rate": 1.997501710773889e-05,
"loss": 0.3882,
"step": 77
},
{
"epoch": 0.2898281467719461,
"grad_norm": 0.09563567489385605,
"learning_rate": 1.9972798462510114e-05,
"loss": 0.3881,
"step": 78
},
{
"epoch": 0.29354389224338134,
"grad_norm": 0.11256393790245056,
"learning_rate": 1.9970485581756337e-05,
"loss": 0.3837,
"step": 79
},
{
"epoch": 0.29725963771481656,
"grad_norm": 0.12789539992809296,
"learning_rate": 1.996807848733257e-05,
"loss": 0.3846,
"step": 80
},
{
"epoch": 0.3009753831862517,
"grad_norm": 0.10919850319623947,
"learning_rate": 1.9965577201984054e-05,
"loss": 0.3856,
"step": 81
},
{
"epoch": 0.30469112865768694,
"grad_norm": 0.10867376625537872,
"learning_rate": 1.996298174934608e-05,
"loss": 0.3907,
"step": 82
},
{
"epoch": 0.30840687412912215,
"grad_norm": 0.12707559764385223,
"learning_rate": 1.9960292153943746e-05,
"loss": 0.3881,
"step": 83
},
{
"epoch": 0.31212261960055737,
"grad_norm": 0.10862530022859573,
"learning_rate": 1.9957508441191722e-05,
"loss": 0.3869,
"step": 84
},
{
"epoch": 0.3158383650719926,
"grad_norm": 0.10388710349798203,
"learning_rate": 1.9954630637394027e-05,
"loss": 0.3905,
"step": 85
},
{
"epoch": 0.31955411054342775,
"grad_norm": 0.12069334834814072,
"learning_rate": 1.995165876974377e-05,
"loss": 0.3844,
"step": 86
},
{
"epoch": 0.32326985601486297,
"grad_norm": 0.1042049452662468,
"learning_rate": 1.9948592866322875e-05,
"loss": 0.3824,
"step": 87
},
{
"epoch": 0.3269856014862982,
"grad_norm": 0.12331200391054153,
"learning_rate": 1.994543295610186e-05,
"loss": 0.3838,
"step": 88
},
{
"epoch": 0.3307013469577334,
"grad_norm": 0.12420689314603806,
"learning_rate": 1.994217906893952e-05,
"loss": 0.3901,
"step": 89
},
{
"epoch": 0.3344170924291686,
"grad_norm": 0.11597050726413727,
"learning_rate": 1.9938831235582673e-05,
"loss": 0.3882,
"step": 90
},
{
"epoch": 0.33813283790060383,
"grad_norm": 0.12018215656280518,
"learning_rate": 1.9935389487665845e-05,
"loss": 0.3781,
"step": 91
},
{
"epoch": 0.341848583372039,
"grad_norm": 0.11860138922929764,
"learning_rate": 1.9931853857710996e-05,
"loss": 0.3816,
"step": 92
},
{
"epoch": 0.3455643288434742,
"grad_norm": 0.10522197932004929,
"learning_rate": 1.9928224379127203e-05,
"loss": 0.3857,
"step": 93
},
{
"epoch": 0.34928007431490943,
"grad_norm": 0.11816974729299545,
"learning_rate": 1.9924501086210334e-05,
"loss": 0.3811,
"step": 94
},
{
"epoch": 0.35299581978634464,
"grad_norm": 0.10380753129720688,
"learning_rate": 1.9920684014142736e-05,
"loss": 0.3921,
"step": 95
},
{
"epoch": 0.35671156525777986,
"grad_norm": 0.1059751808643341,
"learning_rate": 1.99167731989929e-05,
"loss": 0.3757,
"step": 96
},
{
"epoch": 0.360427310729215,
"grad_norm": 0.10807336866855621,
"learning_rate": 1.9912768677715123e-05,
"loss": 0.3851,
"step": 97
},
{
"epoch": 0.36414305620065024,
"grad_norm": 0.10463322699069977,
"learning_rate": 1.990867048814915e-05,
"loss": 0.3852,
"step": 98
},
{
"epoch": 0.36785880167208546,
"grad_norm": 0.09851374477148056,
"learning_rate": 1.9904478669019817e-05,
"loss": 0.3886,
"step": 99
},
{
"epoch": 0.3715745471435207,
"grad_norm": 0.09652000665664673,
"learning_rate": 1.9900193259936706e-05,
"loss": 0.3849,
"step": 100
},
{
"epoch": 0.3752902926149559,
"grad_norm": 0.10542068630456924,
"learning_rate": 1.9895814301393735e-05,
"loss": 0.3785,
"step": 101
},
{
"epoch": 0.3790060380863911,
"grad_norm": 0.09634078294038773,
"learning_rate": 1.9891341834768807e-05,
"loss": 0.3843,
"step": 102
},
{
"epoch": 0.38272178355782627,
"grad_norm": 0.1124754548072815,
"learning_rate": 1.9886775902323405e-05,
"loss": 0.3843,
"step": 103
},
{
"epoch": 0.3864375290292615,
"grad_norm": 0.10203322023153305,
"learning_rate": 1.9882116547202194e-05,
"loss": 0.379,
"step": 104
},
{
"epoch": 0.3901532745006967,
"grad_norm": 0.10191947966814041,
"learning_rate": 1.987736381343261e-05,
"loss": 0.3749,
"step": 105
},
{
"epoch": 0.3938690199721319,
"grad_norm": 0.10634920746088028,
"learning_rate": 1.9872517745924457e-05,
"loss": 0.3825,
"step": 106
},
{
"epoch": 0.39758476544356713,
"grad_norm": 0.10025389492511749,
"learning_rate": 1.986757839046946e-05,
"loss": 0.3804,
"step": 107
},
{
"epoch": 0.4013005109150023,
"grad_norm": 0.1094253808259964,
"learning_rate": 1.9862545793740854e-05,
"loss": 0.3723,
"step": 108
},
{
"epoch": 0.4050162563864375,
"grad_norm": 0.10391519218683243,
"learning_rate": 1.9857420003292934e-05,
"loss": 0.3795,
"step": 109
},
{
"epoch": 0.40873200185787273,
"grad_norm": 0.11390256136655807,
"learning_rate": 1.9852201067560607e-05,
"loss": 0.3879,
"step": 110
},
{
"epoch": 0.41244774732930795,
"grad_norm": 0.10451369732618332,
"learning_rate": 1.9846889035858934e-05,
"loss": 0.3787,
"step": 111
},
{
"epoch": 0.41616349280074316,
"grad_norm": 0.10160938650369644,
"learning_rate": 1.984148395838266e-05,
"loss": 0.3763,
"step": 112
},
{
"epoch": 0.4198792382721784,
"grad_norm": 0.10861197859048843,
"learning_rate": 1.9835985886205744e-05,
"loss": 0.3743,
"step": 113
},
{
"epoch": 0.42359498374361354,
"grad_norm": 0.10781834274530411,
"learning_rate": 1.983039487128088e-05,
"loss": 0.3795,
"step": 114
},
{
"epoch": 0.42731072921504876,
"grad_norm": 0.1119246706366539,
"learning_rate": 1.9824710966438996e-05,
"loss": 0.3833,
"step": 115
},
{
"epoch": 0.431026474686484,
"grad_norm": 0.10627570003271103,
"learning_rate": 1.981893422538877e-05,
"loss": 0.3779,
"step": 116
},
{
"epoch": 0.4347422201579192,
"grad_norm": 0.10912427306175232,
"learning_rate": 1.9813064702716094e-05,
"loss": 0.3796,
"step": 117
},
{
"epoch": 0.4384579656293544,
"grad_norm": 0.10746470093727112,
"learning_rate": 1.9807102453883602e-05,
"loss": 0.3756,
"step": 118
},
{
"epoch": 0.44217371110078957,
"grad_norm": 0.11847472190856934,
"learning_rate": 1.9801047535230105e-05,
"loss": 0.3778,
"step": 119
},
{
"epoch": 0.4458894565722248,
"grad_norm": 0.11306995898485184,
"learning_rate": 1.9794900003970076e-05,
"loss": 0.3788,
"step": 120
},
{
"epoch": 0.44960520204366,
"grad_norm": 0.11839523911476135,
"learning_rate": 1.9788659918193117e-05,
"loss": 0.3753,
"step": 121
},
{
"epoch": 0.4533209475150952,
"grad_norm": 0.11891651153564453,
"learning_rate": 1.9782327336863393e-05,
"loss": 0.3828,
"step": 122
},
{
"epoch": 0.45703669298653044,
"grad_norm": 0.11552026867866516,
"learning_rate": 1.9775902319819082e-05,
"loss": 0.3692,
"step": 123
},
{
"epoch": 0.46075243845796565,
"grad_norm": 0.10442527383565903,
"learning_rate": 1.976938492777182e-05,
"loss": 0.3692,
"step": 124
},
{
"epoch": 0.4644681839294008,
"grad_norm": 0.1254386007785797,
"learning_rate": 1.9762775222306106e-05,
"loss": 0.3804,
"step": 125
},
{
"epoch": 0.46818392940083603,
"grad_norm": 0.11650244146585464,
"learning_rate": 1.9756073265878748e-05,
"loss": 0.3801,
"step": 126
},
{
"epoch": 0.47189967487227125,
"grad_norm": 0.12222098559141159,
"learning_rate": 1.9749279121818235e-05,
"loss": 0.3655,
"step": 127
},
{
"epoch": 0.47561542034370646,
"grad_norm": 0.09941545128822327,
"learning_rate": 1.9742392854324187e-05,
"loss": 0.3744,
"step": 128
},
{
"epoch": 0.4793311658151417,
"grad_norm": 0.1074577048420906,
"learning_rate": 1.9735414528466698e-05,
"loss": 0.3758,
"step": 129
},
{
"epoch": 0.4830469112865769,
"grad_norm": 0.11292421817779541,
"learning_rate": 1.972834421018576e-05,
"loss": 0.3779,
"step": 130
},
{
"epoch": 0.48676265675801206,
"grad_norm": 0.10404345393180847,
"learning_rate": 1.9721181966290614e-05,
"loss": 0.3673,
"step": 131
},
{
"epoch": 0.4904784022294473,
"grad_norm": 0.12067055702209473,
"learning_rate": 1.971392786445914e-05,
"loss": 0.3682,
"step": 132
},
{
"epoch": 0.4941941477008825,
"grad_norm": 0.11403202265501022,
"learning_rate": 1.9706581973237207e-05,
"loss": 0.38,
"step": 133
},
{
"epoch": 0.4979098931723177,
"grad_norm": 0.12584474682807922,
"learning_rate": 1.9699144362038015e-05,
"loss": 0.3818,
"step": 134
},
{
"epoch": 0.5016256386437529,
"grad_norm": 0.1250823587179184,
"learning_rate": 1.9691615101141454e-05,
"loss": 0.3708,
"step": 135
},
{
"epoch": 0.5053413841151881,
"grad_norm": 0.1091843917965889,
"learning_rate": 1.968399426169344e-05,
"loss": 0.3658,
"step": 136
},
{
"epoch": 0.5090571295866233,
"grad_norm": 0.12343791872262955,
"learning_rate": 1.967628191570524e-05,
"loss": 0.3749,
"step": 137
},
{
"epoch": 0.5127728750580586,
"grad_norm": 0.10410905629396439,
"learning_rate": 1.9668478136052776e-05,
"loss": 0.3638,
"step": 138
},
{
"epoch": 0.5164886205294937,
"grad_norm": 0.1207507774233818,
"learning_rate": 1.9660582996475966e-05,
"loss": 0.3656,
"step": 139
},
{
"epoch": 0.5202043660009289,
"grad_norm": 0.11468645930290222,
"learning_rate": 1.9652596571578003e-05,
"loss": 0.3686,
"step": 140
},
{
"epoch": 0.5239201114723642,
"grad_norm": 0.10709523409605026,
"learning_rate": 1.9644518936824658e-05,
"loss": 0.3639,
"step": 141
},
{
"epoch": 0.5276358569437993,
"grad_norm": 0.10620342195034027,
"learning_rate": 1.9636350168543572e-05,
"loss": 0.3742,
"step": 142
},
{
"epoch": 0.5313516024152346,
"grad_norm": 0.10497118532657623,
"learning_rate": 1.9628090343923524e-05,
"loss": 0.3778,
"step": 143
},
{
"epoch": 0.5350673478866698,
"grad_norm": 0.10633561760187149,
"learning_rate": 1.961973954101371e-05,
"loss": 0.3723,
"step": 144
},
{
"epoch": 0.5387830933581049,
"grad_norm": 0.10530222207307816,
"learning_rate": 1.961129783872301e-05,
"loss": 0.3715,
"step": 145
},
{
"epoch": 0.5424988388295402,
"grad_norm": 0.11706908047199249,
"learning_rate": 1.9602765316819222e-05,
"loss": 0.3723,
"step": 146
},
{
"epoch": 0.5462145843009754,
"grad_norm": 0.11296215653419495,
"learning_rate": 1.9594142055928333e-05,
"loss": 0.3699,
"step": 147
},
{
"epoch": 0.5499303297724106,
"grad_norm": 0.10609306395053864,
"learning_rate": 1.9585428137533743e-05,
"loss": 0.3696,
"step": 148
},
{
"epoch": 0.5536460752438458,
"grad_norm": 0.1202760711312294,
"learning_rate": 1.9576623643975495e-05,
"loss": 0.3683,
"step": 149
},
{
"epoch": 0.557361820715281,
"grad_norm": 0.10370510816574097,
"learning_rate": 1.9567728658449504e-05,
"loss": 0.3681,
"step": 150
},
{
"epoch": 0.5610775661867162,
"grad_norm": 0.11221983283758163,
"learning_rate": 1.9558743265006764e-05,
"loss": 0.3791,
"step": 151
},
{
"epoch": 0.5647933116581514,
"grad_norm": 0.10173673182725906,
"learning_rate": 1.9549667548552557e-05,
"loss": 0.3749,
"step": 152
},
{
"epoch": 0.5685090571295867,
"grad_norm": 0.1054166704416275,
"learning_rate": 1.9540501594845645e-05,
"loss": 0.3741,
"step": 153
},
{
"epoch": 0.5722248026010218,
"grad_norm": 0.10377232730388641,
"learning_rate": 1.9531245490497478e-05,
"loss": 0.3731,
"step": 154
},
{
"epoch": 0.575940548072457,
"grad_norm": 0.10552498698234558,
"learning_rate": 1.9521899322971353e-05,
"loss": 0.3653,
"step": 155
},
{
"epoch": 0.5796562935438923,
"grad_norm": 0.10727725178003311,
"learning_rate": 1.95124631805816e-05,
"loss": 0.367,
"step": 156
},
{
"epoch": 0.5833720390153274,
"grad_norm": 0.12488004565238953,
"learning_rate": 1.950293715249274e-05,
"loss": 0.3711,
"step": 157
},
{
"epoch": 0.5870877844867627,
"grad_norm": 0.09955500811338425,
"learning_rate": 1.9493321328718653e-05,
"loss": 0.3661,
"step": 158
},
{
"epoch": 0.5908035299581978,
"grad_norm": 0.12957961857318878,
"learning_rate": 1.9483615800121717e-05,
"loss": 0.3683,
"step": 159
},
{
"epoch": 0.5945192754296331,
"grad_norm": 0.14880992472171783,
"learning_rate": 1.947382065841196e-05,
"loss": 0.3751,
"step": 160
},
{
"epoch": 0.5982350209010683,
"grad_norm": 0.13926413655281067,
"learning_rate": 1.9463935996146173e-05,
"loss": 0.3733,
"step": 161
},
{
"epoch": 0.6019507663725034,
"grad_norm": 0.10962341725826263,
"learning_rate": 1.9453961906727073e-05,
"loss": 0.3691,
"step": 162
},
{
"epoch": 0.6056665118439387,
"grad_norm": 0.11052726209163666,
"learning_rate": 1.944389848440237e-05,
"loss": 0.368,
"step": 163
},
{
"epoch": 0.6093822573153739,
"grad_norm": 0.12043924629688263,
"learning_rate": 1.943374582426393e-05,
"loss": 0.3661,
"step": 164
},
{
"epoch": 0.6130980027868091,
"grad_norm": 0.12165231257677078,
"learning_rate": 1.9423504022246825e-05,
"loss": 0.3608,
"step": 165
},
{
"epoch": 0.6168137482582443,
"grad_norm": 0.11346674710512161,
"learning_rate": 1.9413173175128472e-05,
"loss": 0.3694,
"step": 166
},
{
"epoch": 0.6205294937296795,
"grad_norm": 0.11015301942825317,
"learning_rate": 1.940275338052769e-05,
"loss": 0.367,
"step": 167
},
{
"epoch": 0.6242452392011147,
"grad_norm": 0.11212456971406937,
"learning_rate": 1.9392244736903774e-05,
"loss": 0.3658,
"step": 168
},
{
"epoch": 0.6279609846725499,
"grad_norm": 0.12165389209985733,
"learning_rate": 1.9381647343555597e-05,
"loss": 0.3733,
"step": 169
},
{
"epoch": 0.6316767301439852,
"grad_norm": 0.11079975217580795,
"learning_rate": 1.9370961300620637e-05,
"loss": 0.3713,
"step": 170
},
{
"epoch": 0.6353924756154203,
"grad_norm": 0.12430576235055923,
"learning_rate": 1.9360186709074055e-05,
"loss": 0.3662,
"step": 171
},
{
"epoch": 0.6391082210868555,
"grad_norm": 0.10625098645687103,
"learning_rate": 1.934932367072772e-05,
"loss": 0.3718,
"step": 172
},
{
"epoch": 0.6428239665582908,
"grad_norm": 0.11880301684141159,
"learning_rate": 1.9338372288229253e-05,
"loss": 0.3731,
"step": 173
},
{
"epoch": 0.6465397120297259,
"grad_norm": 0.10719984769821167,
"learning_rate": 1.9327332665061083e-05,
"loss": 0.3737,
"step": 174
},
{
"epoch": 0.6502554575011612,
"grad_norm": 0.10688327997922897,
"learning_rate": 1.9316204905539425e-05,
"loss": 0.3677,
"step": 175
},
{
"epoch": 0.6539712029725964,
"grad_norm": 0.10859767347574234,
"learning_rate": 1.9304989114813333e-05,
"loss": 0.3622,
"step": 176
},
{
"epoch": 0.6576869484440316,
"grad_norm": 0.10489044338464737,
"learning_rate": 1.9293685398863685e-05,
"loss": 0.3645,
"step": 177
},
{
"epoch": 0.6614026939154668,
"grad_norm": 0.10492727905511856,
"learning_rate": 1.9282293864502176e-05,
"loss": 0.3635,
"step": 178
},
{
"epoch": 0.665118439386902,
"grad_norm": 0.1101205050945282,
"learning_rate": 1.9270814619370342e-05,
"loss": 0.3638,
"step": 179
},
{
"epoch": 0.6688341848583372,
"grad_norm": 0.10615579038858414,
"learning_rate": 1.92592477719385e-05,
"loss": 0.367,
"step": 180
},
{
"epoch": 0.6725499303297724,
"grad_norm": 0.11769100278615952,
"learning_rate": 1.9247593431504756e-05,
"loss": 0.3617,
"step": 181
},
{
"epoch": 0.6762656758012077,
"grad_norm": 0.10983094573020935,
"learning_rate": 1.923585170819395e-05,
"loss": 0.3632,
"step": 182
},
{
"epoch": 0.6799814212726428,
"grad_norm": 0.10805229097604752,
"learning_rate": 1.922402271295664e-05,
"loss": 0.3582,
"step": 183
},
{
"epoch": 0.683697166744078,
"grad_norm": 0.10664580762386322,
"learning_rate": 1.9212106557568018e-05,
"loss": 0.3689,
"step": 184
},
{
"epoch": 0.6874129122155133,
"grad_norm": 0.11054529249668121,
"learning_rate": 1.9200103354626892e-05,
"loss": 0.358,
"step": 185
},
{
"epoch": 0.6911286576869484,
"grad_norm": 0.10236749053001404,
"learning_rate": 1.91880132175546e-05,
"loss": 0.3687,
"step": 186
},
{
"epoch": 0.6948444031583837,
"grad_norm": 0.10728856176137924,
"learning_rate": 1.9175836260593937e-05,
"loss": 0.3665,
"step": 187
},
{
"epoch": 0.6985601486298189,
"grad_norm": 0.10645803809165955,
"learning_rate": 1.9163572598808093e-05,
"loss": 0.3632,
"step": 188
},
{
"epoch": 0.702275894101254,
"grad_norm": 0.1138240322470665,
"learning_rate": 1.915122234807954e-05,
"loss": 0.3691,
"step": 189
},
{
"epoch": 0.7059916395726893,
"grad_norm": 0.11220771074295044,
"learning_rate": 1.913878562510896e-05,
"loss": 0.3645,
"step": 190
},
{
"epoch": 0.7097073850441245,
"grad_norm": 0.11357756704092026,
"learning_rate": 1.912626254741413e-05,
"loss": 0.362,
"step": 191
},
{
"epoch": 0.7134231305155597,
"grad_norm": 0.10898102819919586,
"learning_rate": 1.9113653233328816e-05,
"loss": 0.3681,
"step": 192
},
{
"epoch": 0.7171388759869949,
"grad_norm": 0.11341891437768936,
"learning_rate": 1.9100957802001655e-05,
"loss": 0.3668,
"step": 193
},
{
"epoch": 0.72085462145843,
"grad_norm": 0.11596344411373138,
"learning_rate": 1.908817637339503e-05,
"loss": 0.3691,
"step": 194
},
{
"epoch": 0.7245703669298653,
"grad_norm": 0.11261321604251862,
"learning_rate": 1.907530906828393e-05,
"loss": 0.3654,
"step": 195
},
{
"epoch": 0.7282861124013005,
"grad_norm": 0.1231698989868164,
"learning_rate": 1.906235600825481e-05,
"loss": 0.3656,
"step": 196
},
{
"epoch": 0.7320018578727358,
"grad_norm": 0.11808507144451141,
"learning_rate": 1.9049317315704447e-05,
"loss": 0.3622,
"step": 197
},
{
"epoch": 0.7357176033441709,
"grad_norm": 0.12203074991703033,
"learning_rate": 1.9036193113838792e-05,
"loss": 0.3603,
"step": 198
},
{
"epoch": 0.7394333488156062,
"grad_norm": 0.11299382895231247,
"learning_rate": 1.9022983526671786e-05,
"loss": 0.3596,
"step": 199
},
{
"epoch": 0.7431490942870413,
"grad_norm": 0.1264580339193344,
"learning_rate": 1.900968867902419e-05,
"loss": 0.3633,
"step": 200
},
{
"epoch": 0.7468648397584765,
"grad_norm": 0.14948038756847382,
"learning_rate": 1.8996308696522435e-05,
"loss": 0.3656,
"step": 201
},
{
"epoch": 0.7505805852299118,
"grad_norm": 0.11759644001722336,
"learning_rate": 1.898284370559739e-05,
"loss": 0.3569,
"step": 202
},
{
"epoch": 0.7542963307013469,
"grad_norm": 0.1067824587225914,
"learning_rate": 1.8969293833483204e-05,
"loss": 0.3706,
"step": 203
},
{
"epoch": 0.7580120761727822,
"grad_norm": 0.12216956168413162,
"learning_rate": 1.8955659208216087e-05,
"loss": 0.3688,
"step": 204
},
{
"epoch": 0.7617278216442174,
"grad_norm": 0.12257716059684753,
"learning_rate": 1.89419399586331e-05,
"loss": 0.359,
"step": 205
},
{
"epoch": 0.7654435671156525,
"grad_norm": 0.11582986265420914,
"learning_rate": 1.892813621437095e-05,
"loss": 0.3617,
"step": 206
},
{
"epoch": 0.7691593125870878,
"grad_norm": 0.11466754972934723,
"learning_rate": 1.891424810586474e-05,
"loss": 0.3593,
"step": 207
},
{
"epoch": 0.772875058058523,
"grad_norm": 0.11693622171878815,
"learning_rate": 1.890027576434677e-05,
"loss": 0.3593,
"step": 208
},
{
"epoch": 0.7765908035299582,
"grad_norm": 0.11666169762611389,
"learning_rate": 1.8886219321845263e-05,
"loss": 0.3689,
"step": 209
},
{
"epoch": 0.7803065490013934,
"grad_norm": 0.11272235959768295,
"learning_rate": 1.887207891118315e-05,
"loss": 0.3633,
"step": 210
},
{
"epoch": 0.7840222944728286,
"grad_norm": 0.11142394691705704,
"learning_rate": 1.885785466597678e-05,
"loss": 0.3658,
"step": 211
},
{
"epoch": 0.7877380399442638,
"grad_norm": 0.11649249494075775,
"learning_rate": 1.8843546720634695e-05,
"loss": 0.3629,
"step": 212
},
{
"epoch": 0.791453785415699,
"grad_norm": 0.1062132716178894,
"learning_rate": 1.8829155210356332e-05,
"loss": 0.3649,
"step": 213
},
{
"epoch": 0.7951695308871343,
"grad_norm": 0.11438262462615967,
"learning_rate": 1.881468027113075e-05,
"loss": 0.3565,
"step": 214
},
{
"epoch": 0.7988852763585694,
"grad_norm": 0.10582791268825531,
"learning_rate": 1.880012203973536e-05,
"loss": 0.3617,
"step": 215
},
{
"epoch": 0.8026010218300046,
"grad_norm": 0.10940583795309067,
"learning_rate": 1.8785480653734608e-05,
"loss": 0.3573,
"step": 216
},
{
"epoch": 0.8063167673014399,
"grad_norm": 0.10012167692184448,
"learning_rate": 1.8770756251478705e-05,
"loss": 0.358,
"step": 217
},
{
"epoch": 0.810032512772875,
"grad_norm": 0.09781812876462936,
"learning_rate": 1.8755948972102297e-05,
"loss": 0.3631,
"step": 218
},
{
"epoch": 0.8137482582443103,
"grad_norm": 0.11387172341346741,
"learning_rate": 1.8741058955523147e-05,
"loss": 0.3579,
"step": 219
},
{
"epoch": 0.8174640037157455,
"grad_norm": 0.11608204245567322,
"learning_rate": 1.8726086342440847e-05,
"loss": 0.3555,
"step": 220
},
{
"epoch": 0.8211797491871807,
"grad_norm": 0.11751598864793777,
"learning_rate": 1.8711031274335438e-05,
"loss": 0.3636,
"step": 221
},
{
"epoch": 0.8248954946586159,
"grad_norm": 0.10483776032924652,
"learning_rate": 1.869589389346611e-05,
"loss": 0.3573,
"step": 222
},
{
"epoch": 0.828611240130051,
"grad_norm": 0.10811923444271088,
"learning_rate": 1.868067434286986e-05,
"loss": 0.3621,
"step": 223
},
{
"epoch": 0.8323269856014863,
"grad_norm": 0.10653597861528397,
"learning_rate": 1.866537276636011e-05,
"loss": 0.3534,
"step": 224
},
{
"epoch": 0.8360427310729215,
"grad_norm": 0.10610423982143402,
"learning_rate": 1.8649989308525374e-05,
"loss": 0.3598,
"step": 225
},
{
"epoch": 0.8397584765443568,
"grad_norm": 0.10406066477298737,
"learning_rate": 1.863452411472788e-05,
"loss": 0.3632,
"step": 226
},
{
"epoch": 0.8434742220157919,
"grad_norm": 0.1097031980752945,
"learning_rate": 1.8618977331102207e-05,
"loss": 0.3643,
"step": 227
},
{
"epoch": 0.8471899674872271,
"grad_norm": 0.09758152812719345,
"learning_rate": 1.8603349104553884e-05,
"loss": 0.3616,
"step": 228
},
{
"epoch": 0.8509057129586624,
"grad_norm": 0.10443262755870819,
"learning_rate": 1.8587639582758032e-05,
"loss": 0.3625,
"step": 229
},
{
"epoch": 0.8546214584300975,
"grad_norm": 0.10417632013559341,
"learning_rate": 1.857184891415794e-05,
"loss": 0.355,
"step": 230
},
{
"epoch": 0.8583372039015328,
"grad_norm": 0.1136803850531578,
"learning_rate": 1.8555977247963674e-05,
"loss": 0.3621,
"step": 231
},
{
"epoch": 0.862052949372968,
"grad_norm": 0.10396552085876465,
"learning_rate": 1.8540024734150673e-05,
"loss": 0.3476,
"step": 232
},
{
"epoch": 0.8657686948444031,
"grad_norm": 0.12654678523540497,
"learning_rate": 1.852399152345832e-05,
"loss": 0.362,
"step": 233
},
{
"epoch": 0.8694844403158384,
"grad_norm": 0.10213786363601685,
"learning_rate": 1.8507877767388533e-05,
"loss": 0.3664,
"step": 234
},
{
"epoch": 0.8732001857872735,
"grad_norm": 0.10471751540899277,
"learning_rate": 1.849168361820431e-05,
"loss": 0.3645,
"step": 235
},
{
"epoch": 0.8769159312587088,
"grad_norm": 0.11185527592897415,
"learning_rate": 1.8475409228928314e-05,
"loss": 0.3514,
"step": 236
},
{
"epoch": 0.880631676730144,
"grad_norm": 0.10277125984430313,
"learning_rate": 1.845905475334141e-05,
"loss": 0.3547,
"step": 237
},
{
"epoch": 0.8843474222015791,
"grad_norm": 0.10685877501964569,
"learning_rate": 1.8442620345981232e-05,
"loss": 0.359,
"step": 238
},
{
"epoch": 0.8880631676730144,
"grad_norm": 0.10827268660068512,
"learning_rate": 1.8426106162140693e-05,
"loss": 0.3655,
"step": 239
},
{
"epoch": 0.8917789131444496,
"grad_norm": 0.11376289278268814,
"learning_rate": 1.8409512357866548e-05,
"loss": 0.3615,
"step": 240
},
{
"epoch": 0.8954946586158848,
"grad_norm": 0.11439840495586395,
"learning_rate": 1.83928390899579e-05,
"loss": 0.3665,
"step": 241
},
{
"epoch": 0.89921040408732,
"grad_norm": 0.11564568430185318,
"learning_rate": 1.837608651596472e-05,
"loss": 0.3675,
"step": 242
},
{
"epoch": 0.9029261495587553,
"grad_norm": 0.13850247859954834,
"learning_rate": 1.8359254794186368e-05,
"loss": 0.358,
"step": 243
},
{
"epoch": 0.9066418950301904,
"grad_norm": 0.12600892782211304,
"learning_rate": 1.8342344083670097e-05,
"loss": 0.3476,
"step": 244
},
{
"epoch": 0.9103576405016256,
"grad_norm": 0.11614871025085449,
"learning_rate": 1.8325354544209534e-05,
"loss": 0.3566,
"step": 245
},
{
"epoch": 0.9140733859730609,
"grad_norm": 0.13191549479961395,
"learning_rate": 1.8308286336343187e-05,
"loss": 0.3556,
"step": 246
},
{
"epoch": 0.917789131444496,
"grad_norm": 0.1293947845697403,
"learning_rate": 1.8291139621352916e-05,
"loss": 0.3736,
"step": 247
},
{
"epoch": 0.9215048769159313,
"grad_norm": 0.12305119633674622,
"learning_rate": 1.8273914561262425e-05,
"loss": 0.355,
"step": 248
},
{
"epoch": 0.9252206223873665,
"grad_norm": 0.10830698162317276,
"learning_rate": 1.825661131883571e-05,
"loss": 0.3664,
"step": 249
},
{
"epoch": 0.9289363678588016,
"grad_norm": 0.12654940783977509,
"learning_rate": 1.8239230057575542e-05,
"loss": 0.3614,
"step": 250
},
{
"epoch": 0.9326521133302369,
"grad_norm": 0.1320241093635559,
"learning_rate": 1.8221770941721907e-05,
"loss": 0.3576,
"step": 251
},
{
"epoch": 0.9363678588016721,
"grad_norm": 0.12282443791627884,
"learning_rate": 1.8204234136250454e-05,
"loss": 0.3537,
"step": 252
},
{
"epoch": 0.9400836042731073,
"grad_norm": 0.11634375900030136,
"learning_rate": 1.818661980687095e-05,
"loss": 0.3521,
"step": 253
},
{
"epoch": 0.9437993497445425,
"grad_norm": 0.12420698255300522,
"learning_rate": 1.8168928120025698e-05,
"loss": 0.3568,
"step": 254
},
{
"epoch": 0.9475150952159777,
"grad_norm": 0.11452219635248184,
"learning_rate": 1.8151159242887983e-05,
"loss": 0.3597,
"step": 255
},
{
"epoch": 0.9512308406874129,
"grad_norm": 0.1083439439535141,
"learning_rate": 1.8133313343360462e-05,
"loss": 0.358,
"step": 256
},
{
"epoch": 0.9549465861588481,
"grad_norm": 0.12687116861343384,
"learning_rate": 1.8115390590073612e-05,
"loss": 0.3675,
"step": 257
},
{
"epoch": 0.9586623316302834,
"grad_norm": 0.10974804311990738,
"learning_rate": 1.8097391152384117e-05,
"loss": 0.3567,
"step": 258
},
{
"epoch": 0.9623780771017185,
"grad_norm": 0.12404585629701614,
"learning_rate": 1.8079315200373266e-05,
"loss": 0.3593,
"step": 259
},
{
"epoch": 0.9660938225731538,
"grad_norm": 0.10702972114086151,
"learning_rate": 1.806116290484536e-05,
"loss": 0.3674,
"step": 260
},
{
"epoch": 0.969809568044589,
"grad_norm": 0.12391137331724167,
"learning_rate": 1.8042934437326085e-05,
"loss": 0.3546,
"step": 261
},
{
"epoch": 0.9735253135160241,
"grad_norm": 0.1159466803073883,
"learning_rate": 1.802462997006089e-05,
"loss": 0.3578,
"step": 262
},
{
"epoch": 0.9772410589874594,
"grad_norm": 0.13155779242515564,
"learning_rate": 1.8006249676013378e-05,
"loss": 0.352,
"step": 263
},
{
"epoch": 0.9809568044588945,
"grad_norm": 0.11537192761898041,
"learning_rate": 1.798779372886365e-05,
"loss": 0.3558,
"step": 264
},
{
"epoch": 0.9846725499303298,
"grad_norm": 0.12012824416160583,
"learning_rate": 1.796926230300667e-05,
"loss": 0.3492,
"step": 265
},
{
"epoch": 0.988388295401765,
"grad_norm": 0.10498210787773132,
"learning_rate": 1.795065557355063e-05,
"loss": 0.3556,
"step": 266
},
{
"epoch": 0.9921040408732001,
"grad_norm": 0.12285551428794861,
"learning_rate": 1.7931973716315272e-05,
"loss": 0.3536,
"step": 267
},
{
"epoch": 0.9958197863446354,
"grad_norm": 0.10645238310098648,
"learning_rate": 1.7913216907830252e-05,
"loss": 0.3543,
"step": 268
},
{
"epoch": 0.9995355318160706,
"grad_norm": 0.11486140638589859,
"learning_rate": 1.7894385325333447e-05,
"loss": 0.355,
"step": 269
},
{
"epoch": 1.0,
"grad_norm": 0.11486140638589859,
"learning_rate": 1.7875479146769306e-05,
"loss": 0.3608,
"step": 270
},
{
"epoch": 1.0037157454714352,
"grad_norm": 0.284243643283844,
"learning_rate": 1.7856498550787144e-05,
"loss": 0.3421,
"step": 271
},
{
"epoch": 1.0074314909428703,
"grad_norm": 0.1473366618156433,
"learning_rate": 1.7837443716739477e-05,
"loss": 0.3389,
"step": 272
},
{
"epoch": 1.0111472364143057,
"grad_norm": 0.14558644592761993,
"learning_rate": 1.78183148246803e-05,
"loss": 0.3379,
"step": 273
},
{
"epoch": 1.0148629818857409,
"grad_norm": 0.13411030173301697,
"learning_rate": 1.7799112055363415e-05,
"loss": 0.3415,
"step": 274
},
{
"epoch": 1.018578727357176,
"grad_norm": 0.12966971099376678,
"learning_rate": 1.77798355902407e-05,
"loss": 0.3424,
"step": 275
},
{
"epoch": 1.0222944728286112,
"grad_norm": 0.1261550337076187,
"learning_rate": 1.7760485611460416e-05,
"loss": 0.3374,
"step": 276
},
{
"epoch": 1.0260102183000464,
"grad_norm": 0.13321344554424286,
"learning_rate": 1.7741062301865454e-05,
"loss": 0.3385,
"step": 277
},
{
"epoch": 1.0297259637714817,
"grad_norm": 0.1240149587392807,
"learning_rate": 1.7721565844991643e-05,
"loss": 0.3398,
"step": 278
},
{
"epoch": 1.033441709242917,
"grad_norm": 0.13695065677165985,
"learning_rate": 1.7701996425065995e-05,
"loss": 0.3356,
"step": 279
},
{
"epoch": 1.037157454714352,
"grad_norm": 0.11709772050380707,
"learning_rate": 1.768235422700496e-05,
"loss": 0.3448,
"step": 280
},
{
"epoch": 1.0408732001857872,
"grad_norm": 0.12430210411548615,
"learning_rate": 1.7662639436412705e-05,
"loss": 0.3375,
"step": 281
},
{
"epoch": 1.0445889456572224,
"grad_norm": 0.1151481345295906,
"learning_rate": 1.7642852239579324e-05,
"loss": 0.3377,
"step": 282
},
{
"epoch": 1.0483046911286578,
"grad_norm": 0.12525425851345062,
"learning_rate": 1.7622992823479104e-05,
"loss": 0.3509,
"step": 283
},
{
"epoch": 1.052020436600093,
"grad_norm": 0.11774883419275284,
"learning_rate": 1.7603061375768756e-05,
"loss": 0.3424,
"step": 284
},
{
"epoch": 1.055736182071528,
"grad_norm": 0.12612788379192352,
"learning_rate": 1.7583058084785626e-05,
"loss": 0.3369,
"step": 285
},
{
"epoch": 1.0594519275429632,
"grad_norm": 0.11893551051616669,
"learning_rate": 1.756298313954593e-05,
"loss": 0.3383,
"step": 286
},
{
"epoch": 1.0631676730143984,
"grad_norm": 0.1180834174156189,
"learning_rate": 1.7542836729742966e-05,
"loss": 0.3409,
"step": 287
},
{
"epoch": 1.0668834184858338,
"grad_norm": 0.11312295496463776,
"learning_rate": 1.7522619045745315e-05,
"loss": 0.3408,
"step": 288
},
{
"epoch": 1.070599163957269,
"grad_norm": 0.12109804898500443,
"learning_rate": 1.7502330278595044e-05,
"loss": 0.3437,
"step": 289
},
{
"epoch": 1.0743149094287041,
"grad_norm": 0.11416348069906235,
"learning_rate": 1.7481970620005913e-05,
"loss": 0.3438,
"step": 290
},
{
"epoch": 1.0780306549001393,
"grad_norm": 0.11594592779874802,
"learning_rate": 1.7461540262361542e-05,
"loss": 0.3337,
"step": 291
},
{
"epoch": 1.0817464003715744,
"grad_norm": 0.11707190424203873,
"learning_rate": 1.744103939871361e-05,
"loss": 0.3394,
"step": 292
},
{
"epoch": 1.0854621458430098,
"grad_norm": 0.11415137350559235,
"learning_rate": 1.742046822278002e-05,
"loss": 0.3402,
"step": 293
},
{
"epoch": 1.089177891314445,
"grad_norm": 0.10898122191429138,
"learning_rate": 1.7399826928943087e-05,
"loss": 0.3441,
"step": 294
},
{
"epoch": 1.0928936367858801,
"grad_norm": 0.11431027948856354,
"learning_rate": 1.7379115712247676e-05,
"loss": 0.3373,
"step": 295
},
{
"epoch": 1.0966093822573153,
"grad_norm": 0.11516235768795013,
"learning_rate": 1.735833476839937e-05,
"loss": 0.3323,
"step": 296
},
{
"epoch": 1.1003251277287505,
"grad_norm": 0.1106986552476883,
"learning_rate": 1.733748429376263e-05,
"loss": 0.3458,
"step": 297
},
{
"epoch": 1.1040408732001858,
"grad_norm": 0.11559002101421356,
"learning_rate": 1.7316564485358923e-05,
"loss": 0.3356,
"step": 298
},
{
"epoch": 1.107756618671621,
"grad_norm": 0.11351880431175232,
"learning_rate": 1.7295575540864878e-05,
"loss": 0.3374,
"step": 299
},
{
"epoch": 1.1114723641430562,
"grad_norm": 0.11151888221502304,
"learning_rate": 1.72745176586104e-05,
"loss": 0.3376,
"step": 300
},
{
"epoch": 1.1151881096144913,
"grad_norm": 0.11173103004693985,
"learning_rate": 1.7253391037576806e-05,
"loss": 0.3441,
"step": 301
},
{
"epoch": 1.1189038550859267,
"grad_norm": 0.11951515078544617,
"learning_rate": 1.7232195877394952e-05,
"loss": 0.3366,
"step": 302
},
{
"epoch": 1.1226196005573619,
"grad_norm": 0.10734188556671143,
"learning_rate": 1.721093237834332e-05,
"loss": 0.3374,
"step": 303
},
{
"epoch": 1.126335346028797,
"grad_norm": 0.11176922172307968,
"learning_rate": 1.7189600741346165e-05,
"loss": 0.3367,
"step": 304
},
{
"epoch": 1.1300510915002322,
"grad_norm": 0.11130116879940033,
"learning_rate": 1.7168201167971582e-05,
"loss": 0.3407,
"step": 305
},
{
"epoch": 1.1337668369716674,
"grad_norm": 0.10894190520048141,
"learning_rate": 1.7146733860429614e-05,
"loss": 0.3377,
"step": 306
},
{
"epoch": 1.1374825824431027,
"grad_norm": 0.1069476529955864,
"learning_rate": 1.7125199021570342e-05,
"loss": 0.3416,
"step": 307
},
{
"epoch": 1.141198327914538,
"grad_norm": 0.10981668531894684,
"learning_rate": 1.710359685488197e-05,
"loss": 0.3374,
"step": 308
},
{
"epoch": 1.144914073385973,
"grad_norm": 0.11587081104516983,
"learning_rate": 1.708192756448891e-05,
"loss": 0.3358,
"step": 309
},
{
"epoch": 1.1486298188574082,
"grad_norm": 0.11336593329906464,
"learning_rate": 1.706019135514982e-05,
"loss": 0.342,
"step": 310
},
{
"epoch": 1.1523455643288434,
"grad_norm": 0.10128185898065567,
"learning_rate": 1.703838843225571e-05,
"loss": 0.3336,
"step": 311
},
{
"epoch": 1.1560613098002788,
"grad_norm": 0.12036030739545822,
"learning_rate": 1.701651900182798e-05,
"loss": 0.3352,
"step": 312
},
{
"epoch": 1.159777055271714,
"grad_norm": 0.11904885619878769,
"learning_rate": 1.699458327051647e-05,
"loss": 0.3352,
"step": 313
},
{
"epoch": 1.163492800743149,
"grad_norm": 0.116232730448246,
"learning_rate": 1.697258144559753e-05,
"loss": 0.3352,
"step": 314
},
{
"epoch": 1.1672085462145843,
"grad_norm": 0.11403744667768478,
"learning_rate": 1.6950513734972022e-05,
"loss": 0.3386,
"step": 315
},
{
"epoch": 1.1709242916860194,
"grad_norm": 0.12414857000112534,
"learning_rate": 1.6928380347163396e-05,
"loss": 0.3385,
"step": 316
},
{
"epoch": 1.1746400371574548,
"grad_norm": 0.1062345877289772,
"learning_rate": 1.69061814913157e-05,
"loss": 0.3286,
"step": 317
},
{
"epoch": 1.17835578262889,
"grad_norm": 0.11717105656862259,
"learning_rate": 1.6883917377191604e-05,
"loss": 0.3426,
"step": 318
},
{
"epoch": 1.1820715281003251,
"grad_norm": 0.11142522841691971,
"learning_rate": 1.6861588215170416e-05,
"loss": 0.3374,
"step": 319
},
{
"epoch": 1.1857872735717603,
"grad_norm": 0.12603576481342316,
"learning_rate": 1.683919421624611e-05,
"loss": 0.3468,
"step": 320
},
{
"epoch": 1.1895030190431957,
"grad_norm": 0.11312548816204071,
"learning_rate": 1.6816735592025303e-05,
"loss": 0.3352,
"step": 321
},
{
"epoch": 1.1932187645146308,
"grad_norm": 0.12090343981981277,
"learning_rate": 1.6794212554725294e-05,
"loss": 0.3355,
"step": 322
},
{
"epoch": 1.196934509986066,
"grad_norm": 0.1098797544836998,
"learning_rate": 1.677162531717202e-05,
"loss": 0.3385,
"step": 323
},
{
"epoch": 1.2006502554575011,
"grad_norm": 0.11589297652244568,
"learning_rate": 1.674897409279807e-05,
"loss": 0.3299,
"step": 324
},
{
"epoch": 1.2043660009289363,
"grad_norm": 0.12202473729848862,
"learning_rate": 1.6726259095640666e-05,
"loss": 0.3354,
"step": 325
},
{
"epoch": 1.2080817464003717,
"grad_norm": 0.11492861062288284,
"learning_rate": 1.670348054033962e-05,
"loss": 0.3436,
"step": 326
},
{
"epoch": 1.2117974918718069,
"grad_norm": 0.1032748594880104,
"learning_rate": 1.6680638642135337e-05,
"loss": 0.3329,
"step": 327
},
{
"epoch": 1.215513237343242,
"grad_norm": 0.11737876385450363,
"learning_rate": 1.6657733616866755e-05,
"loss": 0.3398,
"step": 328
},
{
"epoch": 1.2192289828146772,
"grad_norm": 0.11444680392742157,
"learning_rate": 1.6634765680969325e-05,
"loss": 0.3378,
"step": 329
},
{
"epoch": 1.2229447282861123,
"grad_norm": 0.11126520484685898,
"learning_rate": 1.661173505147295e-05,
"loss": 0.334,
"step": 330
},
{
"epoch": 1.2266604737575477,
"grad_norm": 0.13582204282283783,
"learning_rate": 1.658864194599994e-05,
"loss": 0.3345,
"step": 331
},
{
"epoch": 1.2303762192289829,
"grad_norm": 0.11902488023042679,
"learning_rate": 1.656548658276296e-05,
"loss": 0.3401,
"step": 332
},
{
"epoch": 1.234091964700418,
"grad_norm": 0.11862975358963013,
"learning_rate": 1.6542269180562963e-05,
"loss": 0.339,
"step": 333
},
{
"epoch": 1.2378077101718532,
"grad_norm": 0.11739562451839447,
"learning_rate": 1.6518989958787126e-05,
"loss": 0.3348,
"step": 334
},
{
"epoch": 1.2415234556432884,
"grad_norm": 0.12765569984912872,
"learning_rate": 1.649564913740677e-05,
"loss": 0.3352,
"step": 335
},
{
"epoch": 1.2452392011147237,
"grad_norm": 0.11639034003019333,
"learning_rate": 1.6472246936975297e-05,
"loss": 0.3432,
"step": 336
},
{
"epoch": 1.248954946586159,
"grad_norm": 0.10757054388523102,
"learning_rate": 1.6448783578626076e-05,
"loss": 0.3328,
"step": 337
},
{
"epoch": 1.252670692057594,
"grad_norm": 0.117894247174263,
"learning_rate": 1.64252592840704e-05,
"loss": 0.3371,
"step": 338
},
{
"epoch": 1.2563864375290292,
"grad_norm": 0.10041515529155731,
"learning_rate": 1.640167427559533e-05,
"loss": 0.3421,
"step": 339
},
{
"epoch": 1.2601021830004644,
"grad_norm": 0.11587150394916534,
"learning_rate": 1.6378028776061667e-05,
"loss": 0.3396,
"step": 340
},
{
"epoch": 1.2638179284718998,
"grad_norm": 0.11105579137802124,
"learning_rate": 1.6354323008901774e-05,
"loss": 0.3317,
"step": 341
},
{
"epoch": 1.267533673943335,
"grad_norm": 0.11884349584579468,
"learning_rate": 1.633055719811752e-05,
"loss": 0.3361,
"step": 342
},
{
"epoch": 1.27124941941477,
"grad_norm": 0.10920147597789764,
"learning_rate": 1.630673156827813e-05,
"loss": 0.3386,
"step": 343
},
{
"epoch": 1.2749651648862053,
"grad_norm": 0.11774492263793945,
"learning_rate": 1.6282846344518076e-05,
"loss": 0.3411,
"step": 344
},
{
"epoch": 1.2786809103576404,
"grad_norm": 0.10830100625753403,
"learning_rate": 1.625890175253495e-05,
"loss": 0.3341,
"step": 345
},
{
"epoch": 1.2823966558290758,
"grad_norm": 0.1259647011756897,
"learning_rate": 1.6234898018587336e-05,
"loss": 0.3387,
"step": 346
},
{
"epoch": 1.286112401300511,
"grad_norm": 0.1034000962972641,
"learning_rate": 1.6210835369492654e-05,
"loss": 0.3378,
"step": 347
},
{
"epoch": 1.2898281467719461,
"grad_norm": 0.12871000170707703,
"learning_rate": 1.6186714032625036e-05,
"loss": 0.3372,
"step": 348
},
{
"epoch": 1.2935438922433813,
"grad_norm": 0.1100439801812172,
"learning_rate": 1.6162534235913163e-05,
"loss": 0.3384,
"step": 349
},
{
"epoch": 1.2972596377148164,
"grad_norm": 0.12109148502349854,
"learning_rate": 1.6138296207838128e-05,
"loss": 0.3387,
"step": 350
},
{
"epoch": 1.3009753831862518,
"grad_norm": 0.1027180626988411,
"learning_rate": 1.6114000177431263e-05,
"loss": 0.3401,
"step": 351
},
{
"epoch": 1.304691128657687,
"grad_norm": 0.12273887544870377,
"learning_rate": 1.608964637427197e-05,
"loss": 0.3298,
"step": 352
},
{
"epoch": 1.3084068741291222,
"grad_norm": 0.11235188692808151,
"learning_rate": 1.6065235028485568e-05,
"loss": 0.3384,
"step": 353
},
{
"epoch": 1.3121226196005573,
"grad_norm": 0.12132605910301208,
"learning_rate": 1.6040766370741118e-05,
"loss": 0.3389,
"step": 354
},
{
"epoch": 1.3158383650719925,
"grad_norm": 0.11250706017017365,
"learning_rate": 1.6016240632249224e-05,
"loss": 0.3335,
"step": 355
},
{
"epoch": 1.3195541105434279,
"grad_norm": 0.10493439435958862,
"learning_rate": 1.5991658044759874e-05,
"loss": 0.3402,
"step": 356
},
{
"epoch": 1.323269856014863,
"grad_norm": 0.1091570034623146,
"learning_rate": 1.5967018840560214e-05,
"loss": 0.3315,
"step": 357
},
{
"epoch": 1.3269856014862982,
"grad_norm": 0.10770201683044434,
"learning_rate": 1.59423232524724e-05,
"loss": 0.3393,
"step": 358
},
{
"epoch": 1.3307013469577333,
"grad_norm": 0.11085475981235504,
"learning_rate": 1.5917571513851367e-05,
"loss": 0.3347,
"step": 359
},
{
"epoch": 1.3344170924291685,
"grad_norm": 0.09745285660028458,
"learning_rate": 1.589276385858262e-05,
"loss": 0.3306,
"step": 360
},
{
"epoch": 1.3381328379006039,
"grad_norm": 0.10505405813455582,
"learning_rate": 1.5867900521080047e-05,
"loss": 0.3398,
"step": 361
},
{
"epoch": 1.341848583372039,
"grad_norm": 0.1016383096575737,
"learning_rate": 1.5842981736283686e-05,
"loss": 0.3375,
"step": 362
},
{
"epoch": 1.3455643288434742,
"grad_norm": 0.10377795994281769,
"learning_rate": 1.5818007739657514e-05,
"loss": 0.3321,
"step": 363
},
{
"epoch": 1.3492800743149094,
"grad_norm": 0.10698731988668442,
"learning_rate": 1.5792978767187213e-05,
"loss": 0.3378,
"step": 364
},
{
"epoch": 1.3529958197863445,
"grad_norm": 0.10323892533779144,
"learning_rate": 1.5767895055377952e-05,
"loss": 0.331,
"step": 365
},
{
"epoch": 1.35671156525778,
"grad_norm": 0.09851153939962387,
"learning_rate": 1.5742756841252145e-05,
"loss": 0.3223,
"step": 366
},
{
"epoch": 1.360427310729215,
"grad_norm": 0.10611561685800552,
"learning_rate": 1.5717564362347205e-05,
"loss": 0.3321,
"step": 367
},
{
"epoch": 1.3641430562006502,
"grad_norm": 0.11031083762645721,
"learning_rate": 1.569231785671332e-05,
"loss": 0.334,
"step": 368
},
{
"epoch": 1.3678588016720854,
"grad_norm": 0.12395249307155609,
"learning_rate": 1.566701756291118e-05,
"loss": 0.3389,
"step": 369
},
{
"epoch": 1.3715745471435206,
"grad_norm": 0.10423075407743454,
"learning_rate": 1.5641663720009733e-05,
"loss": 0.3343,
"step": 370
},
{
"epoch": 1.375290292614956,
"grad_norm": 0.13267524540424347,
"learning_rate": 1.5616256567583935e-05,
"loss": 0.329,
"step": 371
},
{
"epoch": 1.379006038086391,
"grad_norm": 0.1115020290017128,
"learning_rate": 1.5590796345712468e-05,
"loss": 0.3409,
"step": 372
},
{
"epoch": 1.3827217835578263,
"grad_norm": 0.10712186247110367,
"learning_rate": 1.5565283294975482e-05,
"loss": 0.3289,
"step": 373
},
{
"epoch": 1.3864375290292614,
"grad_norm": 0.11980023980140686,
"learning_rate": 1.553971765645233e-05,
"loss": 0.3369,
"step": 374
},
{
"epoch": 1.3901532745006966,
"grad_norm": 0.10561339557170868,
"learning_rate": 1.551409967171927e-05,
"loss": 0.3319,
"step": 375
},
{
"epoch": 1.393869019972132,
"grad_norm": 0.11171392351388931,
"learning_rate": 1.5488429582847194e-05,
"loss": 0.3262,
"step": 376
},
{
"epoch": 1.3975847654435671,
"grad_norm": 0.11271104216575623,
"learning_rate": 1.5462707632399345e-05,
"loss": 0.3389,
"step": 377
},
{
"epoch": 1.4013005109150023,
"grad_norm": 0.11088487505912781,
"learning_rate": 1.5436934063429013e-05,
"loss": 0.331,
"step": 378
},
{
"epoch": 1.4050162563864375,
"grad_norm": 0.10341040790081024,
"learning_rate": 1.5411109119477247e-05,
"loss": 0.3373,
"step": 379
},
{
"epoch": 1.4087320018578726,
"grad_norm": 0.11629636585712433,
"learning_rate": 1.5385233044570555e-05,
"loss": 0.3389,
"step": 380
},
{
"epoch": 1.412447747329308,
"grad_norm": 0.10395944863557816,
"learning_rate": 1.5359306083218592e-05,
"loss": 0.3337,
"step": 381
},
{
"epoch": 1.4161634928007432,
"grad_norm": 0.12571927905082703,
"learning_rate": 1.5333328480411844e-05,
"loss": 0.3269,
"step": 382
},
{
"epoch": 1.4198792382721783,
"grad_norm": 0.10714583098888397,
"learning_rate": 1.5307300481619334e-05,
"loss": 0.3434,
"step": 383
},
{
"epoch": 1.4235949837436135,
"grad_norm": 0.10667157173156738,
"learning_rate": 1.5281222332786284e-05,
"loss": 0.335,
"step": 384
},
{
"epoch": 1.4273107292150486,
"grad_norm": 0.10444430261850357,
"learning_rate": 1.5255094280331797e-05,
"loss": 0.3353,
"step": 385
},
{
"epoch": 1.431026474686484,
"grad_norm": 0.10662422329187393,
"learning_rate": 1.5228916571146523e-05,
"loss": 0.3307,
"step": 386
},
{
"epoch": 1.4347422201579192,
"grad_norm": 0.10255824029445648,
"learning_rate": 1.520268945259034e-05,
"loss": 0.3374,
"step": 387
},
{
"epoch": 1.4384579656293544,
"grad_norm": 0.11058612167835236,
"learning_rate": 1.5176413172490003e-05,
"loss": 0.3354,
"step": 388
},
{
"epoch": 1.4421737111007895,
"grad_norm": 0.11343454569578171,
"learning_rate": 1.5150087979136802e-05,
"loss": 0.3337,
"step": 389
},
{
"epoch": 1.4458894565722247,
"grad_norm": 0.10026858747005463,
"learning_rate": 1.512371412128424e-05,
"loss": 0.3354,
"step": 390
},
{
"epoch": 1.44960520204366,
"grad_norm": 0.11643780767917633,
"learning_rate": 1.5097291848145632e-05,
"loss": 0.3367,
"step": 391
},
{
"epoch": 1.4533209475150952,
"grad_norm": 0.101609006524086,
"learning_rate": 1.5070821409391814e-05,
"loss": 0.3336,
"step": 392
},
{
"epoch": 1.4570366929865304,
"grad_norm": 0.10110028088092804,
"learning_rate": 1.5044303055148724e-05,
"loss": 0.3318,
"step": 393
},
{
"epoch": 1.4607524384579658,
"grad_norm": 0.10438964515924454,
"learning_rate": 1.5017737035995088e-05,
"loss": 0.3319,
"step": 394
},
{
"epoch": 1.4644681839294007,
"grad_norm": 0.10237043350934982,
"learning_rate": 1.4991123602960018e-05,
"loss": 0.3293,
"step": 395
},
{
"epoch": 1.468183929400836,
"grad_norm": 0.10324394702911377,
"learning_rate": 1.4964463007520649e-05,
"loss": 0.333,
"step": 396
},
{
"epoch": 1.4718996748722712,
"grad_norm": 0.1022440642118454,
"learning_rate": 1.4937755501599774e-05,
"loss": 0.3371,
"step": 397
},
{
"epoch": 1.4756154203437064,
"grad_norm": 0.09690335392951965,
"learning_rate": 1.491100133756345e-05,
"loss": 0.3319,
"step": 398
},
{
"epoch": 1.4793311658151418,
"grad_norm": 0.09800193458795547,
"learning_rate": 1.4884200768218626e-05,
"loss": 0.3349,
"step": 399
},
{
"epoch": 1.483046911286577,
"grad_norm": 0.0996604934334755,
"learning_rate": 1.4857354046810732e-05,
"loss": 0.3355,
"step": 400
},
{
"epoch": 1.4867626567580121,
"grad_norm": 0.10491257160902023,
"learning_rate": 1.4830461427021314e-05,
"loss": 0.3378,
"step": 401
},
{
"epoch": 1.4904784022294473,
"grad_norm": 0.09817502647638321,
"learning_rate": 1.480352316296562e-05,
"loss": 0.3384,
"step": 402
},
{
"epoch": 1.4941941477008824,
"grad_norm": 0.11181376874446869,
"learning_rate": 1.47765395091902e-05,
"loss": 0.3376,
"step": 403
},
{
"epoch": 1.4979098931723178,
"grad_norm": 0.10846973955631256,
"learning_rate": 1.4749510720670506e-05,
"loss": 0.3349,
"step": 404
},
{
"epoch": 1.5016256386437528,
"grad_norm": 0.09587910026311874,
"learning_rate": 1.4722437052808473e-05,
"loss": 0.331,
"step": 405
},
{
"epoch": 1.5053413841151881,
"grad_norm": 0.1171889379620552,
"learning_rate": 1.4695318761430121e-05,
"loss": 0.3319,
"step": 406
},
{
"epoch": 1.5090571295866233,
"grad_norm": 0.09778203070163727,
"learning_rate": 1.4668156102783126e-05,
"loss": 0.3359,
"step": 407
},
{
"epoch": 1.5127728750580585,
"grad_norm": 0.10606710612773895,
"learning_rate": 1.4640949333534402e-05,
"loss": 0.332,
"step": 408
},
{
"epoch": 1.5164886205294938,
"grad_norm": 0.11034330725669861,
"learning_rate": 1.4613698710767677e-05,
"loss": 0.3376,
"step": 409
},
{
"epoch": 1.5202043660009288,
"grad_norm": 0.10495816916227341,
"learning_rate": 1.4586404491981051e-05,
"loss": 0.3311,
"step": 410
},
{
"epoch": 1.5239201114723642,
"grad_norm": 0.11816810071468353,
"learning_rate": 1.455906693508459e-05,
"loss": 0.3409,
"step": 411
},
{
"epoch": 1.5276358569437993,
"grad_norm": 0.10562555491924286,
"learning_rate": 1.4531686298397863e-05,
"loss": 0.326,
"step": 412
},
{
"epoch": 1.5313516024152345,
"grad_norm": 0.1073359027504921,
"learning_rate": 1.4504262840647514e-05,
"loss": 0.3317,
"step": 413
},
{
"epoch": 1.5350673478866699,
"grad_norm": 0.12166634202003479,
"learning_rate": 1.4476796820964816e-05,
"loss": 0.323,
"step": 414
},
{
"epoch": 1.5387830933581048,
"grad_norm": 0.10392868518829346,
"learning_rate": 1.4449288498883212e-05,
"loss": 0.3262,
"step": 415
},
{
"epoch": 1.5424988388295402,
"grad_norm": 0.11974501609802246,
"learning_rate": 1.4421738134335883e-05,
"loss": 0.3373,
"step": 416
},
{
"epoch": 1.5462145843009754,
"grad_norm": 0.115992471575737,
"learning_rate": 1.4394145987653274e-05,
"loss": 0.3354,
"step": 417
},
{
"epoch": 1.5499303297724105,
"grad_norm": 0.12289142608642578,
"learning_rate": 1.4366512319560642e-05,
"loss": 0.3272,
"step": 418
},
{
"epoch": 1.553646075243846,
"grad_norm": 0.12877260148525238,
"learning_rate": 1.4338837391175582e-05,
"loss": 0.3441,
"step": 419
},
{
"epoch": 1.5573618207152808,
"grad_norm": 0.11049580574035645,
"learning_rate": 1.4311121464005584e-05,
"loss": 0.3327,
"step": 420
},
{
"epoch": 1.5610775661867162,
"grad_norm": 0.12512782216072083,
"learning_rate": 1.428336479994553e-05,
"loss": 0.3355,
"step": 421
},
{
"epoch": 1.5647933116581514,
"grad_norm": 0.11800656467676163,
"learning_rate": 1.425556766127525e-05,
"loss": 0.334,
"step": 422
},
{
"epoch": 1.5685090571295865,
"grad_norm": 0.1239934042096138,
"learning_rate": 1.4227730310657012e-05,
"loss": 0.3344,
"step": 423
},
{
"epoch": 1.572224802601022,
"grad_norm": 0.11580350250005722,
"learning_rate": 1.4199853011133073e-05,
"loss": 0.3433,
"step": 424
},
{
"epoch": 1.5759405480724569,
"grad_norm": 0.11455916613340378,
"learning_rate": 1.417193602612317e-05,
"loss": 0.3289,
"step": 425
},
{
"epoch": 1.5796562935438923,
"grad_norm": 0.11868759989738464,
"learning_rate": 1.4143979619422036e-05,
"loss": 0.3342,
"step": 426
},
{
"epoch": 1.5833720390153274,
"grad_norm": 0.10682390630245209,
"learning_rate": 1.4115984055196919e-05,
"loss": 0.332,
"step": 427
},
{
"epoch": 1.5870877844867626,
"grad_norm": 0.13644340634346008,
"learning_rate": 1.4087949597985064e-05,
"loss": 0.3343,
"step": 428
},
{
"epoch": 1.590803529958198,
"grad_norm": 0.10938290506601334,
"learning_rate": 1.4059876512691233e-05,
"loss": 0.3316,
"step": 429
},
{
"epoch": 1.5945192754296331,
"grad_norm": 0.11587871611118317,
"learning_rate": 1.4031765064585196e-05,
"loss": 0.3349,
"step": 430
},
{
"epoch": 1.5982350209010683,
"grad_norm": 0.1172982007265091,
"learning_rate": 1.4003615519299219e-05,
"loss": 0.3274,
"step": 431
},
{
"epoch": 1.6019507663725034,
"grad_norm": 0.10714416205883026,
"learning_rate": 1.3975428142825562e-05,
"loss": 0.33,
"step": 432
},
{
"epoch": 1.6056665118439386,
"grad_norm": 0.11143547296524048,
"learning_rate": 1.3947203201513955e-05,
"loss": 0.329,
"step": 433
},
{
"epoch": 1.609382257315374,
"grad_norm": 0.11266589909791946,
"learning_rate": 1.3918940962069094e-05,
"loss": 0.3318,
"step": 434
},
{
"epoch": 1.6130980027868091,
"grad_norm": 0.10318125039339066,
"learning_rate": 1.3890641691548114e-05,
"loss": 0.3384,
"step": 435
},
{
"epoch": 1.6168137482582443,
"grad_norm": 0.10905440151691437,
"learning_rate": 1.3862305657358065e-05,
"loss": 0.3318,
"step": 436
},
{
"epoch": 1.6205294937296795,
"grad_norm": 0.10952167212963104,
"learning_rate": 1.3833933127253385e-05,
"loss": 0.3302,
"step": 437
},
{
"epoch": 1.6242452392011146,
"grad_norm": 0.10449928790330887,
"learning_rate": 1.3805524369333374e-05,
"loss": 0.3318,
"step": 438
},
{
"epoch": 1.62796098467255,
"grad_norm": 0.10911315679550171,
"learning_rate": 1.3777079652039649e-05,
"loss": 0.3307,
"step": 439
},
{
"epoch": 1.6316767301439852,
"grad_norm": 0.10771377384662628,
"learning_rate": 1.3748599244153633e-05,
"loss": 0.3364,
"step": 440
},
{
"epoch": 1.6353924756154203,
"grad_norm": 0.1044982373714447,
"learning_rate": 1.3720083414793985e-05,
"loss": 0.337,
"step": 441
},
{
"epoch": 1.6391082210868555,
"grad_norm": 0.1033010333776474,
"learning_rate": 1.3691532433414074e-05,
"loss": 0.3328,
"step": 442
},
{
"epoch": 1.6428239665582907,
"grad_norm": 0.10454479604959488,
"learning_rate": 1.366294656979943e-05,
"loss": 0.3311,
"step": 443
},
{
"epoch": 1.646539712029726,
"grad_norm": 0.10470984876155853,
"learning_rate": 1.3634326094065196e-05,
"loss": 0.3321,
"step": 444
},
{
"epoch": 1.6502554575011612,
"grad_norm": 0.10136619955301285,
"learning_rate": 1.3605671276653568e-05,
"loss": 0.3357,
"step": 445
},
{
"epoch": 1.6539712029725964,
"grad_norm": 0.11278172582387924,
"learning_rate": 1.3576982388331258e-05,
"loss": 0.3331,
"step": 446
},
{
"epoch": 1.6576869484440317,
"grad_norm": 0.10161139816045761,
"learning_rate": 1.3548259700186902e-05,
"loss": 0.3248,
"step": 447
},
{
"epoch": 1.6614026939154667,
"grad_norm": 0.10559763014316559,
"learning_rate": 1.3519503483628542e-05,
"loss": 0.3284,
"step": 448
},
{
"epoch": 1.665118439386902,
"grad_norm": 0.10481450706720352,
"learning_rate": 1.349071401038102e-05,
"loss": 0.3332,
"step": 449
},
{
"epoch": 1.6688341848583372,
"grad_norm": 0.10347817838191986,
"learning_rate": 1.3461891552483443e-05,
"loss": 0.3303,
"step": 450
},
{
"epoch": 1.6725499303297724,
"grad_norm": 0.10819240659475327,
"learning_rate": 1.3433036382286591e-05,
"loss": 0.3337,
"step": 451
},
{
"epoch": 1.6762656758012078,
"grad_norm": 0.10216568410396576,
"learning_rate": 1.340414877245035e-05,
"loss": 0.3313,
"step": 452
},
{
"epoch": 1.6799814212726427,
"grad_norm": 0.10621066391468048,
"learning_rate": 1.3375228995941135e-05,
"loss": 0.3287,
"step": 453
},
{
"epoch": 1.683697166744078,
"grad_norm": 0.10264114290475845,
"learning_rate": 1.334627732602932e-05,
"loss": 0.3304,
"step": 454
},
{
"epoch": 1.6874129122155133,
"grad_norm": 0.10609594732522964,
"learning_rate": 1.3317294036286645e-05,
"loss": 0.3245,
"step": 455
},
{
"epoch": 1.6911286576869484,
"grad_norm": 0.10905460268259048,
"learning_rate": 1.3288279400583633e-05,
"loss": 0.3348,
"step": 456
},
{
"epoch": 1.6948444031583838,
"grad_norm": 0.1078936755657196,
"learning_rate": 1.3259233693086995e-05,
"loss": 0.338,
"step": 457
},
{
"epoch": 1.6985601486298187,
"grad_norm": 0.10438873618841171,
"learning_rate": 1.3230157188257062e-05,
"loss": 0.3319,
"step": 458
},
{
"epoch": 1.7022758941012541,
"grad_norm": 0.10727696865797043,
"learning_rate": 1.3201050160845166e-05,
"loss": 0.3251,
"step": 459
},
{
"epoch": 1.7059916395726893,
"grad_norm": 0.09836070984601974,
"learning_rate": 1.3171912885891063e-05,
"loss": 0.3343,
"step": 460
},
{
"epoch": 1.7097073850441245,
"grad_norm": 0.10552895069122314,
"learning_rate": 1.3142745638720315e-05,
"loss": 0.3354,
"step": 461
},
{
"epoch": 1.7134231305155598,
"grad_norm": 0.10016204416751862,
"learning_rate": 1.3113548694941709e-05,
"loss": 0.3332,
"step": 462
},
{
"epoch": 1.7171388759869948,
"grad_norm": 0.10548803955316544,
"learning_rate": 1.3084322330444637e-05,
"loss": 0.3326,
"step": 463
},
{
"epoch": 1.7208546214584302,
"grad_norm": 0.09914383292198181,
"learning_rate": 1.30550668213965e-05,
"loss": 0.3272,
"step": 464
},
{
"epoch": 1.7245703669298653,
"grad_norm": 0.10793361067771912,
"learning_rate": 1.3025782444240087e-05,
"loss": 0.3332,
"step": 465
},
{
"epoch": 1.7282861124013005,
"grad_norm": 0.10292205214500427,
"learning_rate": 1.2996469475690976e-05,
"loss": 0.3354,
"step": 466
},
{
"epoch": 1.7320018578727359,
"grad_norm": 0.10405939817428589,
"learning_rate": 1.2967128192734903e-05,
"loss": 0.3235,
"step": 467
},
{
"epoch": 1.7357176033441708,
"grad_norm": 0.1032538041472435,
"learning_rate": 1.2937758872625169e-05,
"loss": 0.3284,
"step": 468
},
{
"epoch": 1.7394333488156062,
"grad_norm": 0.10621588677167892,
"learning_rate": 1.2908361792879986e-05,
"loss": 0.3409,
"step": 469
},
{
"epoch": 1.7431490942870413,
"grad_norm": 0.10607681423425674,
"learning_rate": 1.2878937231279894e-05,
"loss": 0.3279,
"step": 470
},
{
"epoch": 1.7468648397584765,
"grad_norm": 0.09700904041528702,
"learning_rate": 1.2849485465865094e-05,
"loss": 0.3303,
"step": 471
},
{
"epoch": 1.750580585229912,
"grad_norm": 0.10592197626829147,
"learning_rate": 1.2820006774932866e-05,
"loss": 0.3292,
"step": 472
},
{
"epoch": 1.7542963307013468,
"grad_norm": 0.10124707967042923,
"learning_rate": 1.2790501437034901e-05,
"loss": 0.3231,
"step": 473
},
{
"epoch": 1.7580120761727822,
"grad_norm": 0.10446885228157043,
"learning_rate": 1.2760969730974692e-05,
"loss": 0.3286,
"step": 474
},
{
"epoch": 1.7617278216442174,
"grad_norm": 0.11013483256101608,
"learning_rate": 1.2731411935804878e-05,
"loss": 0.3317,
"step": 475
},
{
"epoch": 1.7654435671156525,
"grad_norm": 0.1071079894900322,
"learning_rate": 1.270182833082464e-05,
"loss": 0.3357,
"step": 476
},
{
"epoch": 1.769159312587088,
"grad_norm": 0.10007348656654358,
"learning_rate": 1.2672219195577025e-05,
"loss": 0.3293,
"step": 477
},
{
"epoch": 1.7728750580585229,
"grad_norm": 0.10183950513601303,
"learning_rate": 1.2642584809846335e-05,
"loss": 0.3324,
"step": 478
},
{
"epoch": 1.7765908035299582,
"grad_norm": 0.10295970737934113,
"learning_rate": 1.2612925453655465e-05,
"loss": 0.3306,
"step": 479
},
{
"epoch": 1.7803065490013934,
"grad_norm": 0.09775254130363464,
"learning_rate": 1.258324140726326e-05,
"loss": 0.3246,
"step": 480
},
{
"epoch": 1.7840222944728286,
"grad_norm": 0.09783802926540375,
"learning_rate": 1.255353295116187e-05,
"loss": 0.3253,
"step": 481
},
{
"epoch": 1.787738039944264,
"grad_norm": 0.09948322921991348,
"learning_rate": 1.2523800366074106e-05,
"loss": 0.3275,
"step": 482
},
{
"epoch": 1.7914537854156989,
"grad_norm": 0.09534478187561035,
"learning_rate": 1.249404393295077e-05,
"loss": 0.3397,
"step": 483
},
{
"epoch": 1.7951695308871343,
"grad_norm": 0.10330142825841904,
"learning_rate": 1.2464263932968013e-05,
"loss": 0.3278,
"step": 484
},
{
"epoch": 1.7988852763585694,
"grad_norm": 0.0956583321094513,
"learning_rate": 1.2434460647524675e-05,
"loss": 0.3339,
"step": 485
},
{
"epoch": 1.8026010218300046,
"grad_norm": 0.1020829975605011,
"learning_rate": 1.2404634358239633e-05,
"loss": 0.3289,
"step": 486
},
{
"epoch": 1.80631676730144,
"grad_norm": 0.1022854745388031,
"learning_rate": 1.2374785346949125e-05,
"loss": 0.3397,
"step": 487
},
{
"epoch": 1.810032512772875,
"grad_norm": 0.09987523406744003,
"learning_rate": 1.2344913895704099e-05,
"loss": 0.3356,
"step": 488
},
{
"epoch": 1.8137482582443103,
"grad_norm": 0.10145365446805954,
"learning_rate": 1.231502028676754e-05,
"loss": 0.3262,
"step": 489
},
{
"epoch": 1.8174640037157455,
"grad_norm": 0.10481409728527069,
"learning_rate": 1.2285104802611813e-05,
"loss": 0.3276,
"step": 490
},
{
"epoch": 1.8211797491871806,
"grad_norm": 0.10363679379224777,
"learning_rate": 1.2255167725915981e-05,
"loss": 0.3319,
"step": 491
},
{
"epoch": 1.824895494658616,
"grad_norm": 0.11416313797235489,
"learning_rate": 1.2225209339563144e-05,
"loss": 0.3321,
"step": 492
},
{
"epoch": 1.828611240130051,
"grad_norm": 0.11109332740306854,
"learning_rate": 1.2195229926637767e-05,
"loss": 0.3349,
"step": 493
},
{
"epoch": 1.8323269856014863,
"grad_norm": 0.12219825387001038,
"learning_rate": 1.2165229770422989e-05,
"loss": 0.3307,
"step": 494
},
{
"epoch": 1.8360427310729215,
"grad_norm": 0.11590065062046051,
"learning_rate": 1.2135209154397962e-05,
"loss": 0.3342,
"step": 495
},
{
"epoch": 1.8397584765443566,
"grad_norm": 0.10380223393440247,
"learning_rate": 1.2105168362235178e-05,
"loss": 0.3286,
"step": 496
},
{
"epoch": 1.843474222015792,
"grad_norm": 0.10542934387922287,
"learning_rate": 1.2075107677797763e-05,
"loss": 0.3242,
"step": 497
},
{
"epoch": 1.847189967487227,
"grad_norm": 0.10022494941949844,
"learning_rate": 1.2045027385136824e-05,
"loss": 0.3331,
"step": 498
},
{
"epoch": 1.8509057129586624,
"grad_norm": 0.10723895579576492,
"learning_rate": 1.201492776848874e-05,
"loss": 0.3329,
"step": 499
},
{
"epoch": 1.8546214584300975,
"grad_norm": 0.100627101957798,
"learning_rate": 1.1984809112272495e-05,
"loss": 0.3366,
"step": 500
},
{
"epoch": 1.8583372039015327,
"grad_norm": 0.11326257139444351,
"learning_rate": 1.1954671701086977e-05,
"loss": 0.327,
"step": 501
},
{
"epoch": 1.862052949372968,
"grad_norm": 0.0987544134259224,
"learning_rate": 1.19245158197083e-05,
"loss": 0.33,
"step": 502
},
{
"epoch": 1.865768694844403,
"grad_norm": 0.11274933815002441,
"learning_rate": 1.1894341753087106e-05,
"loss": 0.3303,
"step": 503
},
{
"epoch": 1.8694844403158384,
"grad_norm": 0.10070671886205673,
"learning_rate": 1.1864149786345869e-05,
"loss": 0.3285,
"step": 504
},
{
"epoch": 1.8732001857872735,
"grad_norm": 0.11528735607862473,
"learning_rate": 1.1833940204776208e-05,
"loss": 0.3323,
"step": 505
},
{
"epoch": 1.8769159312587087,
"grad_norm": 0.10606225579977036,
"learning_rate": 1.18037132938362e-05,
"loss": 0.3323,
"step": 506
},
{
"epoch": 1.880631676730144,
"grad_norm": 0.10966872423887253,
"learning_rate": 1.1773469339147654e-05,
"loss": 0.3313,
"step": 507
},
{
"epoch": 1.884347422201579,
"grad_norm": 0.11551006883382797,
"learning_rate": 1.174320862649344e-05,
"loss": 0.3327,
"step": 508
},
{
"epoch": 1.8880631676730144,
"grad_norm": 0.09956555813550949,
"learning_rate": 1.1712931441814776e-05,
"loss": 0.3227,
"step": 509
},
{
"epoch": 1.8917789131444496,
"grad_norm": 0.12450045347213745,
"learning_rate": 1.1682638071208535e-05,
"loss": 0.3332,
"step": 510
},
{
"epoch": 1.8954946586158847,
"grad_norm": 0.10522767901420593,
"learning_rate": 1.1652328800924517e-05,
"loss": 0.3394,
"step": 511
},
{
"epoch": 1.8992104040873201,
"grad_norm": 0.11313209682703018,
"learning_rate": 1.1622003917362789e-05,
"loss": 0.3274,
"step": 512
},
{
"epoch": 1.9029261495587553,
"grad_norm": 0.10978467017412186,
"learning_rate": 1.1591663707070922e-05,
"loss": 0.3271,
"step": 513
},
{
"epoch": 1.9066418950301904,
"grad_norm": 0.1085636168718338,
"learning_rate": 1.1561308456741338e-05,
"loss": 0.334,
"step": 514
},
{
"epoch": 1.9103576405016256,
"grad_norm": 0.11466681212186813,
"learning_rate": 1.153093845320856e-05,
"loss": 0.3376,
"step": 515
},
{
"epoch": 1.9140733859730608,
"grad_norm": 0.10606027394533157,
"learning_rate": 1.1500553983446527e-05,
"loss": 0.3278,
"step": 516
},
{
"epoch": 1.9177891314444961,
"grad_norm": 0.11401081830263138,
"learning_rate": 1.147015533456587e-05,
"loss": 0.3405,
"step": 517
},
{
"epoch": 1.9215048769159313,
"grad_norm": 0.10391564667224884,
"learning_rate": 1.1439742793811206e-05,
"loss": 0.3329,
"step": 518
},
{
"epoch": 1.9252206223873665,
"grad_norm": 0.1010107472538948,
"learning_rate": 1.1409316648558406e-05,
"loss": 0.3206,
"step": 519
},
{
"epoch": 1.9289363678588016,
"grad_norm": 0.09453730285167694,
"learning_rate": 1.1378877186311911e-05,
"loss": 0.3282,
"step": 520
},
{
"epoch": 1.9326521133302368,
"grad_norm": 0.1002352312207222,
"learning_rate": 1.1348424694701992e-05,
"loss": 0.3327,
"step": 521
},
{
"epoch": 1.9363678588016722,
"grad_norm": 0.09964974969625473,
"learning_rate": 1.1317959461482029e-05,
"loss": 0.3306,
"step": 522
},
{
"epoch": 1.9400836042731073,
"grad_norm": 0.09215464442968369,
"learning_rate": 1.128748177452581e-05,
"loss": 0.3297,
"step": 523
},
{
"epoch": 1.9437993497445425,
"grad_norm": 0.09653985500335693,
"learning_rate": 1.12569919218248e-05,
"loss": 0.3286,
"step": 524
},
{
"epoch": 1.9475150952159777,
"grad_norm": 0.09677861630916595,
"learning_rate": 1.1226490191485422e-05,
"loss": 0.3319,
"step": 525
},
{
"epoch": 1.9512308406874128,
"grad_norm": 0.09930659830570221,
"learning_rate": 1.1195976871726333e-05,
"loss": 0.3276,
"step": 526
},
{
"epoch": 1.9549465861588482,
"grad_norm": 0.0958540067076683,
"learning_rate": 1.1165452250875693e-05,
"loss": 0.3302,
"step": 527
},
{
"epoch": 1.9586623316302834,
"grad_norm": 0.10563844442367554,
"learning_rate": 1.1134916617368466e-05,
"loss": 0.329,
"step": 528
},
{
"epoch": 1.9623780771017185,
"grad_norm": 0.10037673264741898,
"learning_rate": 1.1104370259743659e-05,
"loss": 0.3366,
"step": 529
},
{
"epoch": 1.966093822573154,
"grad_norm": 0.09517338871955872,
"learning_rate": 1.1073813466641633e-05,
"loss": 0.3233,
"step": 530
},
{
"epoch": 1.9698095680445888,
"grad_norm": 0.09956544637680054,
"learning_rate": 1.1043246526801339e-05,
"loss": 0.3244,
"step": 531
},
{
"epoch": 1.9735253135160242,
"grad_norm": 0.09636662900447845,
"learning_rate": 1.1012669729057616e-05,
"loss": 0.3243,
"step": 532
},
{
"epoch": 1.9772410589874594,
"grad_norm": 0.094456747174263,
"learning_rate": 1.0982083362338451e-05,
"loss": 0.3275,
"step": 533
},
{
"epoch": 1.9809568044588945,
"grad_norm": 0.09380447119474411,
"learning_rate": 1.0951487715662255e-05,
"loss": 0.3265,
"step": 534
},
{
"epoch": 1.98467254993033,
"grad_norm": 0.09866400808095932,
"learning_rate": 1.0920883078135118e-05,
"loss": 0.3259,
"step": 535
},
{
"epoch": 1.9883882954017649,
"grad_norm": 0.09099185466766357,
"learning_rate": 1.0890269738948099e-05,
"loss": 0.3247,
"step": 536
},
{
"epoch": 1.9921040408732003,
"grad_norm": 0.09784712642431259,
"learning_rate": 1.0859647987374467e-05,
"loss": 0.3307,
"step": 537
},
{
"epoch": 1.9958197863446354,
"grad_norm": 0.09866229444742203,
"learning_rate": 1.0829018112766996e-05,
"loss": 0.3271,
"step": 538
},
{
"epoch": 1.9995355318160706,
"grad_norm": 0.09773606806993484,
"learning_rate": 1.0798380404555204e-05,
"loss": 0.3278,
"step": 539
},
{
"epoch": 2.0,
"grad_norm": 0.09773606806993484,
"learning_rate": 1.0767735152242649e-05,
"loss": 0.3111,
"step": 540
},
{
"epoch": 2.0037157454714354,
"grad_norm": 0.2571142315864563,
"learning_rate": 1.0737082645404149e-05,
"loss": 0.2987,
"step": 541
},
{
"epoch": 2.0074314909428703,
"grad_norm": 0.11478757858276367,
"learning_rate": 1.0706423173683094e-05,
"loss": 0.3092,
"step": 542
},
{
"epoch": 2.0111472364143057,
"grad_norm": 0.1191437616944313,
"learning_rate": 1.0675757026788673e-05,
"loss": 0.3069,
"step": 543
},
{
"epoch": 2.0148629818857406,
"grad_norm": 0.1135106086730957,
"learning_rate": 1.0645084494493166e-05,
"loss": 0.3157,
"step": 544
},
{
"epoch": 2.018578727357176,
"grad_norm": 0.1076584905385971,
"learning_rate": 1.0614405866629171e-05,
"loss": 0.3092,
"step": 545
},
{
"epoch": 2.0222944728286114,
"grad_norm": 0.10705482959747314,
"learning_rate": 1.05837214330869e-05,
"loss": 0.3124,
"step": 546
},
{
"epoch": 2.0260102183000464,
"grad_norm": 0.1070384830236435,
"learning_rate": 1.0553031483811416e-05,
"loss": 0.31,
"step": 547
},
{
"epoch": 2.0297259637714817,
"grad_norm": 0.11147139221429825,
"learning_rate": 1.0522336308799906e-05,
"loss": 0.3061,
"step": 548
},
{
"epoch": 2.0334417092429167,
"grad_norm": 0.10954634100198746,
"learning_rate": 1.0491636198098934e-05,
"loss": 0.3091,
"step": 549
},
{
"epoch": 2.037157454714352,
"grad_norm": 0.10510841012001038,
"learning_rate": 1.04609314418017e-05,
"loss": 0.3077,
"step": 550
},
{
"epoch": 2.0408732001857874,
"grad_norm": 0.10965508222579956,
"learning_rate": 1.0430222330045306e-05,
"loss": 0.3051,
"step": 551
},
{
"epoch": 2.0445889456572224,
"grad_norm": 0.10617036372423172,
"learning_rate": 1.0399509153008012e-05,
"loss": 0.3091,
"step": 552
},
{
"epoch": 2.0483046911286578,
"grad_norm": 0.11374052613973618,
"learning_rate": 1.0368792200906484e-05,
"loss": 0.3065,
"step": 553
},
{
"epoch": 2.0520204366000927,
"grad_norm": 0.10345732420682907,
"learning_rate": 1.0338071763993067e-05,
"loss": 0.3106,
"step": 554
},
{
"epoch": 2.055736182071528,
"grad_norm": 0.11452768743038177,
"learning_rate": 1.0307348132553026e-05,
"loss": 0.3143,
"step": 555
},
{
"epoch": 2.0594519275429635,
"grad_norm": 0.11430029571056366,
"learning_rate": 1.0276621596901824e-05,
"loss": 0.3075,
"step": 556
},
{
"epoch": 2.0631676730143984,
"grad_norm": 0.10091587901115417,
"learning_rate": 1.0245892447382355e-05,
"loss": 0.3075,
"step": 557
},
{
"epoch": 2.066883418485834,
"grad_norm": 0.11518266052007675,
"learning_rate": 1.0215160974362224e-05,
"loss": 0.3031,
"step": 558
},
{
"epoch": 2.0705991639572687,
"grad_norm": 0.10245639830827713,
"learning_rate": 1.0184427468230978e-05,
"loss": 0.3061,
"step": 559
},
{
"epoch": 2.074314909428704,
"grad_norm": 0.10168883204460144,
"learning_rate": 1.0153692219397387e-05,
"loss": 0.311,
"step": 560
},
{
"epoch": 2.0780306549001395,
"grad_norm": 0.10546957701444626,
"learning_rate": 1.0122955518286675e-05,
"loss": 0.304,
"step": 561
},
{
"epoch": 2.0817464003715744,
"grad_norm": 0.1051434800028801,
"learning_rate": 1.0092217655337805e-05,
"loss": 0.3007,
"step": 562
},
{
"epoch": 2.08546214584301,
"grad_norm": 0.09844998270273209,
"learning_rate": 1.0061478921000711e-05,
"loss": 0.3114,
"step": 563
},
{
"epoch": 2.0891778913144448,
"grad_norm": 0.11394447088241577,
"learning_rate": 1.003073960573356e-05,
"loss": 0.3069,
"step": 564
},
{
"epoch": 2.09289363678588,
"grad_norm": 0.10783050209283829,
"learning_rate": 1e-05,
"loss": 0.3042,
"step": 565
},
{
"epoch": 2.0966093822573155,
"grad_norm": 0.10865683108568192,
"learning_rate": 9.969260394266447e-06,
"loss": 0.3129,
"step": 566
},
{
"epoch": 2.1003251277287505,
"grad_norm": 0.1120116114616394,
"learning_rate": 9.93852107899929e-06,
"loss": 0.3126,
"step": 567
},
{
"epoch": 2.104040873200186,
"grad_norm": 0.11556120961904526,
"learning_rate": 9.907782344662197e-06,
"loss": 0.3111,
"step": 568
},
{
"epoch": 2.107756618671621,
"grad_norm": 0.11345527321100235,
"learning_rate": 9.877044481713328e-06,
"loss": 0.3056,
"step": 569
},
{
"epoch": 2.111472364143056,
"grad_norm": 0.10415022820234299,
"learning_rate": 9.84630778060262e-06,
"loss": 0.3108,
"step": 570
},
{
"epoch": 2.1151881096144916,
"grad_norm": 0.1070273369550705,
"learning_rate": 9.815572531769022e-06,
"loss": 0.3049,
"step": 571
},
{
"epoch": 2.1189038550859265,
"grad_norm": 0.1060386449098587,
"learning_rate": 9.78483902563778e-06,
"loss": 0.3016,
"step": 572
},
{
"epoch": 2.122619600557362,
"grad_norm": 0.10964947938919067,
"learning_rate": 9.754107552617647e-06,
"loss": 0.3078,
"step": 573
},
{
"epoch": 2.126335346028797,
"grad_norm": 0.10297326743602753,
"learning_rate": 9.723378403098181e-06,
"loss": 0.3104,
"step": 574
},
{
"epoch": 2.130051091500232,
"grad_norm": 0.1009993776679039,
"learning_rate": 9.692651867446974e-06,
"loss": 0.3118,
"step": 575
},
{
"epoch": 2.1337668369716676,
"grad_norm": 0.10273797065019608,
"learning_rate": 9.661928236006936e-06,
"loss": 0.3134,
"step": 576
},
{
"epoch": 2.1374825824431025,
"grad_norm": 0.10296160727739334,
"learning_rate": 9.631207799093521e-06,
"loss": 0.311,
"step": 577
},
{
"epoch": 2.141198327914538,
"grad_norm": 0.10543222725391388,
"learning_rate": 9.600490846991993e-06,
"loss": 0.312,
"step": 578
},
{
"epoch": 2.144914073385973,
"grad_norm": 0.1131412461400032,
"learning_rate": 9.569777669954694e-06,
"loss": 0.3135,
"step": 579
},
{
"epoch": 2.1486298188574082,
"grad_norm": 0.10779815912246704,
"learning_rate": 9.539068558198303e-06,
"loss": 0.3113,
"step": 580
},
{
"epoch": 2.1523455643288436,
"grad_norm": 0.11109422147274017,
"learning_rate": 9.508363801901071e-06,
"loss": 0.3074,
"step": 581
},
{
"epoch": 2.1560613098002785,
"grad_norm": 0.10456559807062149,
"learning_rate": 9.4776636912001e-06,
"loss": 0.3065,
"step": 582
},
{
"epoch": 2.159777055271714,
"grad_norm": 0.1100071594119072,
"learning_rate": 9.446968516188585e-06,
"loss": 0.3089,
"step": 583
},
{
"epoch": 2.163492800743149,
"grad_norm": 0.10959737747907639,
"learning_rate": 9.416278566913102e-06,
"loss": 0.308,
"step": 584
},
{
"epoch": 2.1672085462145843,
"grad_norm": 0.1048310399055481,
"learning_rate": 9.38559413337083e-06,
"loss": 0.3058,
"step": 585
},
{
"epoch": 2.1709242916860196,
"grad_norm": 0.10062061250209808,
"learning_rate": 9.354915505506839e-06,
"loss": 0.3066,
"step": 586
},
{
"epoch": 2.1746400371574546,
"grad_norm": 0.10441904515028,
"learning_rate": 9.324242973211327e-06,
"loss": 0.3079,
"step": 587
},
{
"epoch": 2.17835578262889,
"grad_norm": 0.10219687968492508,
"learning_rate": 9.29357682631691e-06,
"loss": 0.304,
"step": 588
},
{
"epoch": 2.182071528100325,
"grad_norm": 0.09982629865407944,
"learning_rate": 9.262917354595855e-06,
"loss": 0.301,
"step": 589
},
{
"epoch": 2.1857872735717603,
"grad_norm": 0.10712262243032455,
"learning_rate": 9.232264847757358e-06,
"loss": 0.3076,
"step": 590
},
{
"epoch": 2.1895030190431957,
"grad_norm": 0.09682407230138779,
"learning_rate": 9.201619595444796e-06,
"loss": 0.3049,
"step": 591
},
{
"epoch": 2.1932187645146306,
"grad_norm": 0.10327095538377762,
"learning_rate": 9.170981887233008e-06,
"loss": 0.3112,
"step": 592
},
{
"epoch": 2.196934509986066,
"grad_norm": 0.10311618447303772,
"learning_rate": 9.140352012625538e-06,
"loss": 0.3087,
"step": 593
},
{
"epoch": 2.200650255457501,
"grad_norm": 0.10312650352716446,
"learning_rate": 9.109730261051906e-06,
"loss": 0.3113,
"step": 594
},
{
"epoch": 2.2043660009289363,
"grad_norm": 0.10070528835058212,
"learning_rate": 9.079116921864884e-06,
"loss": 0.3049,
"step": 595
},
{
"epoch": 2.2080817464003717,
"grad_norm": 0.09913690388202667,
"learning_rate": 9.048512284337748e-06,
"loss": 0.3113,
"step": 596
},
{
"epoch": 2.2117974918718066,
"grad_norm": 0.10673213005065918,
"learning_rate": 9.017916637661552e-06,
"loss": 0.3089,
"step": 597
},
{
"epoch": 2.215513237343242,
"grad_norm": 0.10604621469974518,
"learning_rate": 8.987330270942389e-06,
"loss": 0.3075,
"step": 598
},
{
"epoch": 2.2192289828146774,
"grad_norm": 0.10301215201616287,
"learning_rate": 8.956753473198663e-06,
"loss": 0.3029,
"step": 599
},
{
"epoch": 2.2229447282861123,
"grad_norm": 0.10370362550020218,
"learning_rate": 8.92618653335837e-06,
"loss": 0.3106,
"step": 600
},
{
"epoch": 2.2266604737575477,
"grad_norm": 0.1033041924238205,
"learning_rate": 8.895629740256344e-06,
"loss": 0.3079,
"step": 601
},
{
"epoch": 2.2303762192289827,
"grad_norm": 0.10055390745401382,
"learning_rate": 8.865083382631541e-06,
"loss": 0.3058,
"step": 602
},
{
"epoch": 2.234091964700418,
"grad_norm": 0.10223246365785599,
"learning_rate": 8.834547749124307e-06,
"loss": 0.3136,
"step": 603
},
{
"epoch": 2.2378077101718534,
"grad_norm": 0.10344142466783524,
"learning_rate": 8.80402312827367e-06,
"loss": 0.3057,
"step": 604
},
{
"epoch": 2.2415234556432884,
"grad_norm": 0.10563880205154419,
"learning_rate": 8.773509808514583e-06,
"loss": 0.3001,
"step": 605
},
{
"epoch": 2.2452392011147237,
"grad_norm": 0.09688053280115128,
"learning_rate": 8.743008078175204e-06,
"loss": 0.3103,
"step": 606
},
{
"epoch": 2.2489549465861587,
"grad_norm": 0.11425190418958664,
"learning_rate": 8.712518225474191e-06,
"loss": 0.3103,
"step": 607
},
{
"epoch": 2.252670692057594,
"grad_norm": 0.10146244615316391,
"learning_rate": 8.682040538517975e-06,
"loss": 0.3129,
"step": 608
},
{
"epoch": 2.2563864375290295,
"grad_norm": 0.11020337045192719,
"learning_rate": 8.651575305298012e-06,
"loss": 0.3087,
"step": 609
},
{
"epoch": 2.2601021830004644,
"grad_norm": 0.10513182729482651,
"learning_rate": 8.62112281368809e-06,
"loss": 0.3014,
"step": 610
},
{
"epoch": 2.2638179284718998,
"grad_norm": 0.1040269061923027,
"learning_rate": 8.590683351441594e-06,
"loss": 0.3085,
"step": 611
},
{
"epoch": 2.2675336739433347,
"grad_norm": 0.10419883579015732,
"learning_rate": 8.560257206188798e-06,
"loss": 0.3056,
"step": 612
},
{
"epoch": 2.27124941941477,
"grad_norm": 0.10336934030056,
"learning_rate": 8.529844665434131e-06,
"loss": 0.3148,
"step": 613
},
{
"epoch": 2.2749651648862055,
"grad_norm": 0.10142672061920166,
"learning_rate": 8.499446016553475e-06,
"loss": 0.3172,
"step": 614
},
{
"epoch": 2.2786809103576404,
"grad_norm": 0.10555445402860641,
"learning_rate": 8.469061546791441e-06,
"loss": 0.3073,
"step": 615
},
{
"epoch": 2.282396655829076,
"grad_norm": 0.10317610204219818,
"learning_rate": 8.438691543258667e-06,
"loss": 0.3058,
"step": 616
},
{
"epoch": 2.2861124013005107,
"grad_norm": 0.11002574115991592,
"learning_rate": 8.40833629292908e-06,
"loss": 0.3118,
"step": 617
},
{
"epoch": 2.289828146771946,
"grad_norm": 0.10648485273122787,
"learning_rate": 8.377996082637216e-06,
"loss": 0.3111,
"step": 618
},
{
"epoch": 2.2935438922433815,
"grad_norm": 0.09864701330661774,
"learning_rate": 8.347671199075482e-06,
"loss": 0.3063,
"step": 619
},
{
"epoch": 2.2972596377148164,
"grad_norm": 0.12031319737434387,
"learning_rate": 8.317361928791469e-06,
"loss": 0.3126,
"step": 620
},
{
"epoch": 2.300975383186252,
"grad_norm": 0.10228867083787918,
"learning_rate": 8.287068558185225e-06,
"loss": 0.3071,
"step": 621
},
{
"epoch": 2.3046911286576868,
"grad_norm": 0.11017017066478729,
"learning_rate": 8.256791373506565e-06,
"loss": 0.3008,
"step": 622
},
{
"epoch": 2.308406874129122,
"grad_norm": 0.1076800525188446,
"learning_rate": 8.22653066085235e-06,
"loss": 0.3064,
"step": 623
},
{
"epoch": 2.3121226196005575,
"grad_norm": 0.09784411638975143,
"learning_rate": 8.196286706163804e-06,
"loss": 0.3019,
"step": 624
},
{
"epoch": 2.3158383650719925,
"grad_norm": 0.11791478097438812,
"learning_rate": 8.166059795223793e-06,
"loss": 0.3066,
"step": 625
},
{
"epoch": 2.319554110543428,
"grad_norm": 0.10086996853351593,
"learning_rate": 8.135850213654136e-06,
"loss": 0.3026,
"step": 626
},
{
"epoch": 2.323269856014863,
"grad_norm": 0.10620040446519852,
"learning_rate": 8.105658246912895e-06,
"loss": 0.3033,
"step": 627
},
{
"epoch": 2.326985601486298,
"grad_norm": 0.11432202160358429,
"learning_rate": 8.075484180291702e-06,
"loss": 0.304,
"step": 628
},
{
"epoch": 2.3307013469577336,
"grad_norm": 0.10481267422437668,
"learning_rate": 8.045328298913025e-06,
"loss": 0.3116,
"step": 629
},
{
"epoch": 2.3344170924291685,
"grad_norm": 0.11234014481306076,
"learning_rate": 8.01519088772751e-06,
"loss": 0.3015,
"step": 630
},
{
"epoch": 2.338132837900604,
"grad_norm": 0.09836047887802124,
"learning_rate": 7.985072231511261e-06,
"loss": 0.3066,
"step": 631
},
{
"epoch": 2.341848583372039,
"grad_norm": 0.1030438095331192,
"learning_rate": 7.954972614863177e-06,
"loss": 0.3105,
"step": 632
},
{
"epoch": 2.345564328843474,
"grad_norm": 0.10446859896183014,
"learning_rate": 7.924892322202238e-06,
"loss": 0.3084,
"step": 633
},
{
"epoch": 2.3492800743149096,
"grad_norm": 0.0999365970492363,
"learning_rate": 7.894831637764829e-06,
"loss": 0.3041,
"step": 634
},
{
"epoch": 2.3529958197863445,
"grad_norm": 0.09953846782445908,
"learning_rate": 7.86479084560204e-06,
"loss": 0.3125,
"step": 635
},
{
"epoch": 2.35671156525778,
"grad_norm": 0.09793154895305634,
"learning_rate": 7.834770229577016e-06,
"loss": 0.3006,
"step": 636
},
{
"epoch": 2.360427310729215,
"grad_norm": 0.10001702606678009,
"learning_rate": 7.804770073362237e-06,
"loss": 0.3077,
"step": 637
},
{
"epoch": 2.3641430562006502,
"grad_norm": 0.09829147905111313,
"learning_rate": 7.774790660436857e-06,
"loss": 0.3107,
"step": 638
},
{
"epoch": 2.3678588016720856,
"grad_norm": 0.09844832867383957,
"learning_rate": 7.74483227408402e-06,
"loss": 0.3077,
"step": 639
},
{
"epoch": 2.3715745471435206,
"grad_norm": 0.10406388342380524,
"learning_rate": 7.71489519738819e-06,
"loss": 0.3045,
"step": 640
},
{
"epoch": 2.375290292614956,
"grad_norm": 0.10157857090234756,
"learning_rate": 7.684979713232463e-06,
"loss": 0.3111,
"step": 641
},
{
"epoch": 2.3790060380863913,
"grad_norm": 0.10471557080745697,
"learning_rate": 7.655086104295904e-06,
"loss": 0.3085,
"step": 642
},
{
"epoch": 2.3827217835578263,
"grad_norm": 0.09915372729301453,
"learning_rate": 7.625214653050875e-06,
"loss": 0.3001,
"step": 643
},
{
"epoch": 2.3864375290292617,
"grad_norm": 0.11579260975122452,
"learning_rate": 7.595365641760368e-06,
"loss": 0.3108,
"step": 644
},
{
"epoch": 2.3901532745006966,
"grad_norm": 0.10019627213478088,
"learning_rate": 7.565539352475326e-06,
"loss": 0.3129,
"step": 645
},
{
"epoch": 2.393869019972132,
"grad_norm": 0.10966377705335617,
"learning_rate": 7.535736067031991e-06,
"loss": 0.3087,
"step": 646
},
{
"epoch": 2.3975847654435674,
"grad_norm": 0.09704477339982986,
"learning_rate": 7.505956067049233e-06,
"loss": 0.3153,
"step": 647
},
{
"epoch": 2.4013005109150023,
"grad_norm": 0.10267549008131027,
"learning_rate": 7.4761996339258955e-06,
"loss": 0.3027,
"step": 648
},
{
"epoch": 2.4050162563864377,
"grad_norm": 0.10359707474708557,
"learning_rate": 7.446467048838131e-06,
"loss": 0.307,
"step": 649
},
{
"epoch": 2.4087320018578726,
"grad_norm": 0.09659229218959808,
"learning_rate": 7.416758592736744e-06,
"loss": 0.3044,
"step": 650
},
{
"epoch": 2.412447747329308,
"grad_norm": 0.1045624315738678,
"learning_rate": 7.387074546344537e-06,
"loss": 0.3107,
"step": 651
},
{
"epoch": 2.4161634928007434,
"grad_norm": 0.09614600241184235,
"learning_rate": 7.357415190153667e-06,
"loss": 0.3094,
"step": 652
},
{
"epoch": 2.4198792382721783,
"grad_norm": 0.10104124993085861,
"learning_rate": 7.327780804422978e-06,
"loss": 0.3077,
"step": 653
},
{
"epoch": 2.4235949837436137,
"grad_norm": 0.10080961883068085,
"learning_rate": 7.298171669175366e-06,
"loss": 0.3054,
"step": 654
},
{
"epoch": 2.4273107292150486,
"grad_norm": 0.09801700711250305,
"learning_rate": 7.268588064195123e-06,
"loss": 0.3043,
"step": 655
},
{
"epoch": 2.431026474686484,
"grad_norm": 0.09909406304359436,
"learning_rate": 7.239030269025311e-06,
"loss": 0.317,
"step": 656
},
{
"epoch": 2.4347422201579194,
"grad_norm": 0.09725164622068405,
"learning_rate": 7.2094985629651015e-06,
"loss": 0.3095,
"step": 657
},
{
"epoch": 2.4384579656293544,
"grad_norm": 0.09465891867876053,
"learning_rate": 7.179993225067137e-06,
"loss": 0.3041,
"step": 658
},
{
"epoch": 2.4421737111007897,
"grad_norm": 0.0986727625131607,
"learning_rate": 7.150514534134906e-06,
"loss": 0.3059,
"step": 659
},
{
"epoch": 2.4458894565722247,
"grad_norm": 0.09608076512813568,
"learning_rate": 7.121062768720111e-06,
"loss": 0.3038,
"step": 660
},
{
"epoch": 2.44960520204366,
"grad_norm": 0.10091105103492737,
"learning_rate": 7.0916382071200155e-06,
"loss": 0.3033,
"step": 661
},
{
"epoch": 2.4533209475150954,
"grad_norm": 0.09536115825176239,
"learning_rate": 7.062241127374838e-06,
"loss": 0.3138,
"step": 662
},
{
"epoch": 2.4570366929865304,
"grad_norm": 0.09733925014734268,
"learning_rate": 7.032871807265097e-06,
"loss": 0.3049,
"step": 663
},
{
"epoch": 2.4607524384579658,
"grad_norm": 0.09790763258934021,
"learning_rate": 7.0035305243090266e-06,
"loss": 0.3119,
"step": 664
},
{
"epoch": 2.4644681839294007,
"grad_norm": 0.10021320730447769,
"learning_rate": 6.974217555759914e-06,
"loss": 0.3077,
"step": 665
},
{
"epoch": 2.468183929400836,
"grad_norm": 0.09405742585659027,
"learning_rate": 6.944933178603503e-06,
"loss": 0.2992,
"step": 666
},
{
"epoch": 2.4718996748722715,
"grad_norm": 0.09465889632701874,
"learning_rate": 6.915677669555364e-06,
"loss": 0.3105,
"step": 667
},
{
"epoch": 2.4756154203437064,
"grad_norm": 0.09789913147687912,
"learning_rate": 6.8864513050582936e-06,
"loss": 0.3057,
"step": 668
},
{
"epoch": 2.479331165815142,
"grad_norm": 0.09432370960712433,
"learning_rate": 6.857254361279688e-06,
"loss": 0.3069,
"step": 669
},
{
"epoch": 2.4830469112865767,
"grad_norm": 0.09729275107383728,
"learning_rate": 6.8280871141089415e-06,
"loss": 0.3156,
"step": 670
},
{
"epoch": 2.486762656758012,
"grad_norm": 0.09846030920743942,
"learning_rate": 6.7989498391548345e-06,
"loss": 0.3095,
"step": 671
},
{
"epoch": 2.4904784022294475,
"grad_norm": 0.09744524210691452,
"learning_rate": 6.7698428117429414e-06,
"loss": 0.3081,
"step": 672
},
{
"epoch": 2.4941941477008824,
"grad_norm": 0.09955080598592758,
"learning_rate": 6.740766306913008e-06,
"loss": 0.3099,
"step": 673
},
{
"epoch": 2.497909893172318,
"grad_norm": 0.09705106914043427,
"learning_rate": 6.711720599416374e-06,
"loss": 0.3041,
"step": 674
},
{
"epoch": 2.5016256386437528,
"grad_norm": 0.10080622881650925,
"learning_rate": 6.682705963713356e-06,
"loss": 0.3088,
"step": 675
},
{
"epoch": 2.505341384115188,
"grad_norm": 0.10080254822969437,
"learning_rate": 6.653722673970682e-06,
"loss": 0.3168,
"step": 676
},
{
"epoch": 2.5090571295866235,
"grad_norm": 0.0976005494594574,
"learning_rate": 6.624771004058869e-06,
"loss": 0.3115,
"step": 677
},
{
"epoch": 2.5127728750580585,
"grad_norm": 0.09836669266223907,
"learning_rate": 6.595851227549657e-06,
"loss": 0.3016,
"step": 678
},
{
"epoch": 2.516488620529494,
"grad_norm": 0.09733550995588303,
"learning_rate": 6.566963617713412e-06,
"loss": 0.3093,
"step": 679
},
{
"epoch": 2.520204366000929,
"grad_norm": 0.10233534127473831,
"learning_rate": 6.538108447516558e-06,
"loss": 0.3067,
"step": 680
},
{
"epoch": 2.523920111472364,
"grad_norm": 0.09999525547027588,
"learning_rate": 6.509285989618981e-06,
"loss": 0.3078,
"step": 681
},
{
"epoch": 2.5276358569437996,
"grad_norm": 0.09615099430084229,
"learning_rate": 6.480496516371462e-06,
"loss": 0.3029,
"step": 682
},
{
"epoch": 2.5313516024152345,
"grad_norm": 0.10191499441862106,
"learning_rate": 6.451740299813098e-06,
"loss": 0.3077,
"step": 683
},
{
"epoch": 2.53506734788667,
"grad_norm": 0.10270554572343826,
"learning_rate": 6.423017611668745e-06,
"loss": 0.3098,
"step": 684
},
{
"epoch": 2.538783093358105,
"grad_norm": 0.10041619092226028,
"learning_rate": 6.394328723346435e-06,
"loss": 0.3093,
"step": 685
},
{
"epoch": 2.54249883882954,
"grad_norm": 0.10207582265138626,
"learning_rate": 6.36567390593481e-06,
"loss": 0.3047,
"step": 686
},
{
"epoch": 2.5462145843009756,
"grad_norm": 0.10021024942398071,
"learning_rate": 6.337053430200571e-06,
"loss": 0.3159,
"step": 687
},
{
"epoch": 2.5499303297724105,
"grad_norm": 0.10303162038326263,
"learning_rate": 6.308467566585928e-06,
"loss": 0.306,
"step": 688
},
{
"epoch": 2.553646075243846,
"grad_norm": 0.10272860527038574,
"learning_rate": 6.279916585206018e-06,
"loss": 0.3094,
"step": 689
},
{
"epoch": 2.557361820715281,
"grad_norm": 0.10093838721513748,
"learning_rate": 6.251400755846372e-06,
"loss": 0.3025,
"step": 690
},
{
"epoch": 2.5610775661867162,
"grad_norm": 0.10561376065015793,
"learning_rate": 6.22292034796035e-06,
"loss": 0.2985,
"step": 691
},
{
"epoch": 2.5647933116581516,
"grad_norm": 0.09984979778528214,
"learning_rate": 6.19447563066663e-06,
"loss": 0.3071,
"step": 692
},
{
"epoch": 2.5685090571295865,
"grad_norm": 0.10431116819381714,
"learning_rate": 6.166066872746617e-06,
"loss": 0.3126,
"step": 693
},
{
"epoch": 2.572224802601022,
"grad_norm": 0.10080066323280334,
"learning_rate": 6.137694342641938e-06,
"loss": 0.3041,
"step": 694
},
{
"epoch": 2.575940548072457,
"grad_norm": 0.0997147485613823,
"learning_rate": 6.109358308451886e-06,
"loss": 0.3056,
"step": 695
},
{
"epoch": 2.5796562935438923,
"grad_norm": 0.0985570177435875,
"learning_rate": 6.0810590379309075e-06,
"loss": 0.305,
"step": 696
},
{
"epoch": 2.5833720390153276,
"grad_norm": 0.09698834270238876,
"learning_rate": 6.052796798486049e-06,
"loss": 0.3109,
"step": 697
},
{
"epoch": 2.5870877844867626,
"grad_norm": 0.10580847412347794,
"learning_rate": 6.024571857174443e-06,
"loss": 0.2999,
"step": 698
},
{
"epoch": 2.590803529958198,
"grad_norm": 0.09868834167718887,
"learning_rate": 5.996384480700783e-06,
"loss": 0.3099,
"step": 699
},
{
"epoch": 2.594519275429633,
"grad_norm": 0.10117078572511673,
"learning_rate": 5.968234935414807e-06,
"loss": 0.3054,
"step": 700
},
{
"epoch": 2.5982350209010683,
"grad_norm": 0.09644345194101334,
"learning_rate": 5.940123487308771e-06,
"loss": 0.3053,
"step": 701
},
{
"epoch": 2.6019507663725037,
"grad_norm": 0.09700403362512589,
"learning_rate": 5.912050402014942e-06,
"loss": 0.3082,
"step": 702
},
{
"epoch": 2.6056665118439386,
"grad_norm": 0.09916125982999802,
"learning_rate": 5.884015944803084e-06,
"loss": 0.3117,
"step": 703
},
{
"epoch": 2.609382257315374,
"grad_norm": 0.09514093399047852,
"learning_rate": 5.856020380577965e-06,
"loss": 0.3125,
"step": 704
},
{
"epoch": 2.613098002786809,
"grad_norm": 0.09283678978681564,
"learning_rate": 5.828063973876834e-06,
"loss": 0.3009,
"step": 705
},
{
"epoch": 2.6168137482582443,
"grad_norm": 0.09611599147319794,
"learning_rate": 5.800146988866928e-06,
"loss": 0.3055,
"step": 706
},
{
"epoch": 2.6205294937296797,
"grad_norm": 0.09528901427984238,
"learning_rate": 5.7722696893429885e-06,
"loss": 0.3082,
"step": 707
},
{
"epoch": 2.6242452392011146,
"grad_norm": 0.09725724160671234,
"learning_rate": 5.744432338724755e-06,
"loss": 0.3027,
"step": 708
},
{
"epoch": 2.62796098467255,
"grad_norm": 0.09872622042894363,
"learning_rate": 5.71663520005447e-06,
"loss": 0.3046,
"step": 709
},
{
"epoch": 2.631676730143985,
"grad_norm": 0.0979916974902153,
"learning_rate": 5.688878535994421e-06,
"loss": 0.3043,
"step": 710
},
{
"epoch": 2.6353924756154203,
"grad_norm": 0.10130661725997925,
"learning_rate": 5.66116260882442e-06,
"loss": 0.3118,
"step": 711
},
{
"epoch": 2.6391082210868557,
"grad_norm": 0.09516625851392746,
"learning_rate": 5.633487680439362e-06,
"loss": 0.3065,
"step": 712
},
{
"epoch": 2.6428239665582907,
"grad_norm": 0.09672761708498001,
"learning_rate": 5.60585401234673e-06,
"loss": 0.3055,
"step": 713
},
{
"epoch": 2.646539712029726,
"grad_norm": 0.10369584709405899,
"learning_rate": 5.578261865664118e-06,
"loss": 0.3074,
"step": 714
},
{
"epoch": 2.650255457501161,
"grad_norm": 0.09531301259994507,
"learning_rate": 5.550711501116789e-06,
"loss": 0.3048,
"step": 715
},
{
"epoch": 2.6539712029725964,
"grad_norm": 0.09632929414510727,
"learning_rate": 5.5232031790351905e-06,
"loss": 0.3035,
"step": 716
},
{
"epoch": 2.6576869484440317,
"grad_norm": 0.10002360492944717,
"learning_rate": 5.495737159352488e-06,
"loss": 0.3065,
"step": 717
},
{
"epoch": 2.6614026939154667,
"grad_norm": 0.09608901292085648,
"learning_rate": 5.468313701602142e-06,
"loss": 0.3049,
"step": 718
},
{
"epoch": 2.665118439386902,
"grad_norm": 0.09645000100135803,
"learning_rate": 5.440933064915414e-06,
"loss": 0.3029,
"step": 719
},
{
"epoch": 2.668834184858337,
"grad_norm": 0.09075741469860077,
"learning_rate": 5.413595508018952e-06,
"loss": 0.3056,
"step": 720
},
{
"epoch": 2.6725499303297724,
"grad_norm": 0.09812219440937042,
"learning_rate": 5.38630128923233e-06,
"loss": 0.3109,
"step": 721
},
{
"epoch": 2.6762656758012078,
"grad_norm": 0.09795837104320526,
"learning_rate": 5.359050666465599e-06,
"loss": 0.3015,
"step": 722
},
{
"epoch": 2.6799814212726427,
"grad_norm": 0.09233727306127548,
"learning_rate": 5.331843897216873e-06,
"loss": 0.3021,
"step": 723
},
{
"epoch": 2.683697166744078,
"grad_norm": 0.09850447624921799,
"learning_rate": 5.304681238569878e-06,
"loss": 0.3122,
"step": 724
},
{
"epoch": 2.687412912215513,
"grad_norm": 0.09807287156581879,
"learning_rate": 5.27756294719153e-06,
"loss": 0.3033,
"step": 725
},
{
"epoch": 2.6911286576869484,
"grad_norm": 0.09560731798410416,
"learning_rate": 5.250489279329501e-06,
"loss": 0.3195,
"step": 726
},
{
"epoch": 2.694844403158384,
"grad_norm": 0.09743351489305496,
"learning_rate": 5.2234604908098e-06,
"loss": 0.3104,
"step": 727
},
{
"epoch": 2.6985601486298187,
"grad_norm": 0.0972006618976593,
"learning_rate": 5.196476837034381e-06,
"loss": 0.3059,
"step": 728
},
{
"epoch": 2.702275894101254,
"grad_norm": 0.09495711326599121,
"learning_rate": 5.169538572978685e-06,
"loss": 0.3058,
"step": 729
},
{
"epoch": 2.705991639572689,
"grad_norm": 0.09505169093608856,
"learning_rate": 5.142645953189271e-06,
"loss": 0.3059,
"step": 730
},
{
"epoch": 2.7097073850441245,
"grad_norm": 0.09332065284252167,
"learning_rate": 5.115799231781377e-06,
"loss": 0.3079,
"step": 731
},
{
"epoch": 2.71342313051556,
"grad_norm": 0.09825999289751053,
"learning_rate": 5.088998662436548e-06,
"loss": 0.3037,
"step": 732
},
{
"epoch": 2.7171388759869948,
"grad_norm": 0.10823912918567657,
"learning_rate": 5.062244498400228e-06,
"loss": 0.3016,
"step": 733
},
{
"epoch": 2.72085462145843,
"grad_norm": 0.09551139175891876,
"learning_rate": 5.035536992479353e-06,
"loss": 0.3007,
"step": 734
},
{
"epoch": 2.724570366929865,
"grad_norm": 0.10059615224599838,
"learning_rate": 5.008876397039984e-06,
"loss": 0.3061,
"step": 735
},
{
"epoch": 2.7282861124013005,
"grad_norm": 0.09422159940004349,
"learning_rate": 4.982262964004913e-06,
"loss": 0.3075,
"step": 736
},
{
"epoch": 2.732001857872736,
"grad_norm": 0.09418869763612747,
"learning_rate": 4.955696944851276e-06,
"loss": 0.3118,
"step": 737
},
{
"epoch": 2.735717603344171,
"grad_norm": 0.09603273868560791,
"learning_rate": 4.929178590608192e-06,
"loss": 0.3048,
"step": 738
},
{
"epoch": 2.739433348815606,
"grad_norm": 0.09696154296398163,
"learning_rate": 4.902708151854371e-06,
"loss": 0.3053,
"step": 739
},
{
"epoch": 2.743149094287041,
"grad_norm": 0.09474286437034607,
"learning_rate": 4.876285878715764e-06,
"loss": 0.3067,
"step": 740
},
{
"epoch": 2.7468648397584765,
"grad_norm": 0.09602458029985428,
"learning_rate": 4.8499120208631985e-06,
"loss": 0.3049,
"step": 741
},
{
"epoch": 2.750580585229912,
"grad_norm": 0.09145466238260269,
"learning_rate": 4.823586827510001e-06,
"loss": 0.3014,
"step": 742
},
{
"epoch": 2.754296330701347,
"grad_norm": 0.09181688725948334,
"learning_rate": 4.797310547409662e-06,
"loss": 0.3027,
"step": 743
},
{
"epoch": 2.758012076172782,
"grad_norm": 0.0925426259636879,
"learning_rate": 4.77108342885348e-06,
"loss": 0.3125,
"step": 744
},
{
"epoch": 2.761727821644217,
"grad_norm": 0.09156447649002075,
"learning_rate": 4.744905719668207e-06,
"loss": 0.3096,
"step": 745
},
{
"epoch": 2.7654435671156525,
"grad_norm": 0.0934828370809555,
"learning_rate": 4.71877766721372e-06,
"loss": 0.3091,
"step": 746
},
{
"epoch": 2.769159312587088,
"grad_norm": 0.09138301014900208,
"learning_rate": 4.692699518380664e-06,
"loss": 0.3078,
"step": 747
},
{
"epoch": 2.772875058058523,
"grad_norm": 0.09329181164503098,
"learning_rate": 4.666671519588159e-06,
"loss": 0.314,
"step": 748
},
{
"epoch": 2.7765908035299582,
"grad_norm": 0.09307435899972916,
"learning_rate": 4.640693916781415e-06,
"loss": 0.3114,
"step": 749
},
{
"epoch": 2.780306549001393,
"grad_norm": 0.09326577186584473,
"learning_rate": 4.614766955429447e-06,
"loss": 0.3091,
"step": 750
},
{
"epoch": 2.7840222944728286,
"grad_norm": 0.09264156967401505,
"learning_rate": 4.588890880522754e-06,
"loss": 0.31,
"step": 751
},
{
"epoch": 2.787738039944264,
"grad_norm": 0.09267882257699966,
"learning_rate": 4.563065936570989e-06,
"loss": 0.3011,
"step": 752
},
{
"epoch": 2.791453785415699,
"grad_norm": 0.09082110971212387,
"learning_rate": 4.537292367600659e-06,
"loss": 0.2981,
"step": 753
},
{
"epoch": 2.7951695308871343,
"grad_norm": 0.09388922899961472,
"learning_rate": 4.5115704171528105e-06,
"loss": 0.3058,
"step": 754
},
{
"epoch": 2.798885276358569,
"grad_norm": 0.09500233083963394,
"learning_rate": 4.485900328280731e-06,
"loss": 0.3025,
"step": 755
},
{
"epoch": 2.8026010218300046,
"grad_norm": 0.0962151437997818,
"learning_rate": 4.460282343547672e-06,
"loss": 0.3075,
"step": 756
},
{
"epoch": 2.80631676730144,
"grad_norm": 0.09406083822250366,
"learning_rate": 4.434716705024519e-06,
"loss": 0.308,
"step": 757
},
{
"epoch": 2.810032512772875,
"grad_norm": 0.09220906347036362,
"learning_rate": 4.409203654287538e-06,
"loss": 0.3056,
"step": 758
},
{
"epoch": 2.8137482582443103,
"grad_norm": 0.10043519735336304,
"learning_rate": 4.3837434324160686e-06,
"loss": 0.3089,
"step": 759
},
{
"epoch": 2.8174640037157452,
"grad_norm": 0.09357497841119766,
"learning_rate": 4.358336279990268e-06,
"loss": 0.3076,
"step": 760
},
{
"epoch": 2.8211797491871806,
"grad_norm": 0.09373392909765244,
"learning_rate": 4.332982437088825e-06,
"loss": 0.307,
"step": 761
},
{
"epoch": 2.824895494658616,
"grad_norm": 0.09675802290439606,
"learning_rate": 4.307682143286683e-06,
"loss": 0.3123,
"step": 762
},
{
"epoch": 2.828611240130051,
"grad_norm": 0.09059885889291763,
"learning_rate": 4.282435637652795e-06,
"loss": 0.3037,
"step": 763
},
{
"epoch": 2.8323269856014863,
"grad_norm": 0.09335997700691223,
"learning_rate": 4.25724315874786e-06,
"loss": 0.3072,
"step": 764
},
{
"epoch": 2.8360427310729213,
"grad_norm": 0.09414941072463989,
"learning_rate": 4.232104944622051e-06,
"loss": 0.3013,
"step": 765
},
{
"epoch": 2.8397584765443566,
"grad_norm": 0.09102322906255722,
"learning_rate": 4.207021232812792e-06,
"loss": 0.3064,
"step": 766
},
{
"epoch": 2.843474222015792,
"grad_norm": 0.09293723106384277,
"learning_rate": 4.18199226034249e-06,
"loss": 0.3011,
"step": 767
},
{
"epoch": 2.847189967487227,
"grad_norm": 0.09253998100757599,
"learning_rate": 4.1570182637163155e-06,
"loss": 0.3009,
"step": 768
},
{
"epoch": 2.8509057129586624,
"grad_norm": 0.09481462091207504,
"learning_rate": 4.132099478919957e-06,
"loss": 0.3039,
"step": 769
},
{
"epoch": 2.8546214584300973,
"grad_norm": 0.09245918691158295,
"learning_rate": 4.107236141417382e-06,
"loss": 0.3061,
"step": 770
},
{
"epoch": 2.8583372039015327,
"grad_norm": 0.09828555583953857,
"learning_rate": 4.082428486148635e-06,
"loss": 0.3037,
"step": 771
},
{
"epoch": 2.862052949372968,
"grad_norm": 0.09272627532482147,
"learning_rate": 4.057676747527601e-06,
"loss": 0.3034,
"step": 772
},
{
"epoch": 2.865768694844403,
"grad_norm": 0.09524228423833847,
"learning_rate": 4.032981159439787e-06,
"loss": 0.3054,
"step": 773
},
{
"epoch": 2.8694844403158384,
"grad_norm": 0.0950593426823616,
"learning_rate": 4.008341955240132e-06,
"loss": 0.2976,
"step": 774
},
{
"epoch": 2.8732001857872733,
"grad_norm": 0.09009818732738495,
"learning_rate": 3.983759367750772e-06,
"loss": 0.2996,
"step": 775
},
{
"epoch": 2.8769159312587087,
"grad_norm": 0.09884744882583618,
"learning_rate": 3.959233629258883e-06,
"loss": 0.308,
"step": 776
},
{
"epoch": 2.880631676730144,
"grad_norm": 0.09184954315423965,
"learning_rate": 3.9347649715144345e-06,
"loss": 0.3047,
"step": 777
},
{
"epoch": 2.884347422201579,
"grad_norm": 0.09278717637062073,
"learning_rate": 3.910353625728035e-06,
"loss": 0.3061,
"step": 778
},
{
"epoch": 2.8880631676730144,
"grad_norm": 0.134787455201149,
"learning_rate": 3.88599982256874e-06,
"loss": 0.3083,
"step": 779
},
{
"epoch": 2.8917789131444493,
"grad_norm": 0.09245725721120834,
"learning_rate": 3.8617037921618705e-06,
"loss": 0.3061,
"step": 780
},
{
"epoch": 2.8954946586158847,
"grad_norm": 0.09511572867631912,
"learning_rate": 3.837465764086837e-06,
"loss": 0.3084,
"step": 781
},
{
"epoch": 2.89921040408732,
"grad_norm": 0.09454077482223511,
"learning_rate": 3.8132859673749688e-06,
"loss": 0.306,
"step": 782
},
{
"epoch": 2.9029261495587555,
"grad_norm": 0.09116362035274506,
"learning_rate": 3.789164630507346e-06,
"loss": 0.2998,
"step": 783
},
{
"epoch": 2.9066418950301904,
"grad_norm": 0.09073933959007263,
"learning_rate": 3.7651019814126656e-06,
"loss": 0.3089,
"step": 784
},
{
"epoch": 2.9103576405016254,
"grad_norm": 0.09033652395009995,
"learning_rate": 3.741098247465049e-06,
"loss": 0.3071,
"step": 785
},
{
"epoch": 2.9140733859730608,
"grad_norm": 0.09258072823286057,
"learning_rate": 3.7171536554819277e-06,
"loss": 0.3059,
"step": 786
},
{
"epoch": 2.917789131444496,
"grad_norm": 0.09817518293857574,
"learning_rate": 3.6932684317218735e-06,
"loss": 0.3101,
"step": 787
},
{
"epoch": 2.9215048769159315,
"grad_norm": 0.0915360152721405,
"learning_rate": 3.66944280188248e-06,
"loss": 0.3111,
"step": 788
},
{
"epoch": 2.9252206223873665,
"grad_norm": 0.09389065951108932,
"learning_rate": 3.645676991098227e-06,
"loss": 0.3018,
"step": 789
},
{
"epoch": 2.9289363678588014,
"grad_norm": 0.09066713601350784,
"learning_rate": 3.6219712239383343e-06,
"loss": 0.3039,
"step": 790
},
{
"epoch": 2.932652113330237,
"grad_norm": 0.09362225234508514,
"learning_rate": 3.5983257244046675e-06,
"loss": 0.3133,
"step": 791
},
{
"epoch": 2.936367858801672,
"grad_norm": 0.09071122854948044,
"learning_rate": 3.5747407159296066e-06,
"loss": 0.3091,
"step": 792
},
{
"epoch": 2.9400836042731076,
"grad_norm": 0.09518720954656601,
"learning_rate": 3.551216421373924e-06,
"loss": 0.3034,
"step": 793
},
{
"epoch": 2.9437993497445425,
"grad_norm": 0.0891517847776413,
"learning_rate": 3.527753063024708e-06,
"loss": 0.3048,
"step": 794
},
{
"epoch": 2.9475150952159774,
"grad_norm": 0.09428103268146515,
"learning_rate": 3.504350862593231e-06,
"loss": 0.3085,
"step": 795
},
{
"epoch": 2.951230840687413,
"grad_norm": 0.09328701347112656,
"learning_rate": 3.4810100412128743e-06,
"loss": 0.3017,
"step": 796
},
{
"epoch": 2.954946586158848,
"grad_norm": 0.09581801295280457,
"learning_rate": 3.4577308194370385e-06,
"loss": 0.3042,
"step": 797
},
{
"epoch": 2.9586623316302836,
"grad_norm": 0.08901771903038025,
"learning_rate": 3.434513417237041e-06,
"loss": 0.302,
"step": 798
},
{
"epoch": 2.9623780771017185,
"grad_norm": 0.08823176473379135,
"learning_rate": 3.4113580540000603e-06,
"loss": 0.3071,
"step": 799
},
{
"epoch": 2.966093822573154,
"grad_norm": 0.09045370668172836,
"learning_rate": 3.3882649485270524e-06,
"loss": 0.3062,
"step": 800
},
{
"epoch": 2.969809568044589,
"grad_norm": 0.09323316067457199,
"learning_rate": 3.365234319030676e-06,
"loss": 0.3114,
"step": 801
},
{
"epoch": 2.9735253135160242,
"grad_norm": 0.09227319061756134,
"learning_rate": 3.342266383133248e-06,
"loss": 0.3046,
"step": 802
},
{
"epoch": 2.9772410589874596,
"grad_norm": 0.0941251888871193,
"learning_rate": 3.3193613578646633e-06,
"loss": 0.3101,
"step": 803
},
{
"epoch": 2.9809568044588945,
"grad_norm": 0.0914355143904686,
"learning_rate": 3.2965194596603834e-06,
"loss": 0.3089,
"step": 804
},
{
"epoch": 2.98467254993033,
"grad_norm": 0.09086716175079346,
"learning_rate": 3.2737409043593404e-06,
"loss": 0.3022,
"step": 805
},
{
"epoch": 2.988388295401765,
"grad_norm": 0.09587659686803818,
"learning_rate": 3.251025907201932e-06,
"loss": 0.3154,
"step": 806
},
{
"epoch": 2.9921040408732003,
"grad_norm": 0.09052026271820068,
"learning_rate": 3.228374682827983e-06,
"loss": 0.3057,
"step": 807
},
{
"epoch": 2.9958197863446356,
"grad_norm": 0.09436668455600739,
"learning_rate": 3.205787445274707e-06,
"loss": 0.3062,
"step": 808
},
{
"epoch": 2.9995355318160706,
"grad_norm": 0.09400416910648346,
"learning_rate": 3.183264407974699e-06,
"loss": 0.3022,
"step": 809
},
{
"epoch": 3.0,
"grad_norm": 0.09400416910648346,
"learning_rate": 3.1608057837538976e-06,
"loss": 0.2979,
"step": 810
}
],
"logging_steps": 1.0,
"max_steps": 1076,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3.145004700717495e+19,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}