Thiyaga158's picture
Upload 95 files
19d50d2 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.5380710659898478,
"eval_steps": 500,
"global_step": 3000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.008460236886632826,
"grad_norm": 2.4233109951019287,
"learning_rate": 1.9949238578680207e-05,
"loss": 1.3216,
"step": 10
},
{
"epoch": 0.01692047377326565,
"grad_norm": 1.1721662282943726,
"learning_rate": 1.9892836999435986e-05,
"loss": 0.5312,
"step": 20
},
{
"epoch": 0.025380710659898477,
"grad_norm": 0.899889349937439,
"learning_rate": 1.9836435420191765e-05,
"loss": 0.2969,
"step": 30
},
{
"epoch": 0.0338409475465313,
"grad_norm": 0.4674186408519745,
"learning_rate": 1.9780033840947548e-05,
"loss": 0.1564,
"step": 40
},
{
"epoch": 0.04230118443316413,
"grad_norm": 0.9013752937316895,
"learning_rate": 1.972363226170333e-05,
"loss": 0.0966,
"step": 50
},
{
"epoch": 0.050761421319796954,
"grad_norm": 0.364164799451828,
"learning_rate": 1.966723068245911e-05,
"loss": 0.0516,
"step": 60
},
{
"epoch": 0.05922165820642978,
"grad_norm": 0.22717136144638062,
"learning_rate": 1.9610829103214893e-05,
"loss": 0.0307,
"step": 70
},
{
"epoch": 0.0676818950930626,
"grad_norm": 0.09149003773927689,
"learning_rate": 1.9554427523970672e-05,
"loss": 0.0182,
"step": 80
},
{
"epoch": 0.07614213197969544,
"grad_norm": 0.12942063808441162,
"learning_rate": 1.949802594472645e-05,
"loss": 0.017,
"step": 90
},
{
"epoch": 0.08460236886632826,
"grad_norm": 0.27440059185028076,
"learning_rate": 1.9441624365482234e-05,
"loss": 0.0113,
"step": 100
},
{
"epoch": 0.09306260575296109,
"grad_norm": 0.5005059242248535,
"learning_rate": 1.9385222786238017e-05,
"loss": 0.0082,
"step": 110
},
{
"epoch": 0.10152284263959391,
"grad_norm": 0.19820384681224823,
"learning_rate": 1.9328821206993796e-05,
"loss": 0.0057,
"step": 120
},
{
"epoch": 0.10998307952622674,
"grad_norm": 0.03431914001703262,
"learning_rate": 1.927241962774958e-05,
"loss": 0.0055,
"step": 130
},
{
"epoch": 0.11844331641285956,
"grad_norm": 0.4414117634296417,
"learning_rate": 1.921601804850536e-05,
"loss": 0.0063,
"step": 140
},
{
"epoch": 0.12690355329949238,
"grad_norm": 0.9055352807044983,
"learning_rate": 1.915961646926114e-05,
"loss": 0.0048,
"step": 150
},
{
"epoch": 0.1353637901861252,
"grad_norm": 0.036724768579006195,
"learning_rate": 1.910321489001692e-05,
"loss": 0.0035,
"step": 160
},
{
"epoch": 0.14382402707275804,
"grad_norm": 0.0888664722442627,
"learning_rate": 1.9046813310772703e-05,
"loss": 0.0033,
"step": 170
},
{
"epoch": 0.15228426395939088,
"grad_norm": 0.019098607823252678,
"learning_rate": 1.8990411731528486e-05,
"loss": 0.0029,
"step": 180
},
{
"epoch": 0.16074450084602368,
"grad_norm": 0.0171552412211895,
"learning_rate": 1.8934010152284265e-05,
"loss": 0.0029,
"step": 190
},
{
"epoch": 0.1692047377326565,
"grad_norm": 0.01885647512972355,
"learning_rate": 1.8877608573040048e-05,
"loss": 0.0028,
"step": 200
},
{
"epoch": 0.17766497461928935,
"grad_norm": 0.01599389687180519,
"learning_rate": 1.8821206993795827e-05,
"loss": 0.0022,
"step": 210
},
{
"epoch": 0.18612521150592218,
"grad_norm": 0.017242038622498512,
"learning_rate": 1.876480541455161e-05,
"loss": 0.0023,
"step": 220
},
{
"epoch": 0.19458544839255498,
"grad_norm": 0.5941590666770935,
"learning_rate": 1.870840383530739e-05,
"loss": 0.0031,
"step": 230
},
{
"epoch": 0.20304568527918782,
"grad_norm": 0.4872148931026459,
"learning_rate": 1.8652002256063172e-05,
"loss": 0.0027,
"step": 240
},
{
"epoch": 0.21150592216582065,
"grad_norm": 0.014173777773976326,
"learning_rate": 1.859560067681895e-05,
"loss": 0.0021,
"step": 250
},
{
"epoch": 0.21996615905245348,
"grad_norm": 0.024433018639683723,
"learning_rate": 1.8539199097574734e-05,
"loss": 0.0017,
"step": 260
},
{
"epoch": 0.22842639593908629,
"grad_norm": 0.017857005819678307,
"learning_rate": 1.8482797518330516e-05,
"loss": 0.0017,
"step": 270
},
{
"epoch": 0.23688663282571912,
"grad_norm": 0.014704135246574879,
"learning_rate": 1.8426395939086296e-05,
"loss": 0.0021,
"step": 280
},
{
"epoch": 0.24534686971235195,
"grad_norm": 0.022361995652318,
"learning_rate": 1.8369994359842075e-05,
"loss": 0.0016,
"step": 290
},
{
"epoch": 0.25380710659898476,
"grad_norm": 0.01112055778503418,
"learning_rate": 1.8313592780597858e-05,
"loss": 0.0015,
"step": 300
},
{
"epoch": 0.2622673434856176,
"grad_norm": 0.010048450902104378,
"learning_rate": 1.825719120135364e-05,
"loss": 0.0015,
"step": 310
},
{
"epoch": 0.2707275803722504,
"grad_norm": 0.012935510836541653,
"learning_rate": 1.820078962210942e-05,
"loss": 0.0014,
"step": 320
},
{
"epoch": 0.27918781725888325,
"grad_norm": 0.010541570372879505,
"learning_rate": 1.8144388042865203e-05,
"loss": 0.0013,
"step": 330
},
{
"epoch": 0.2876480541455161,
"grad_norm": 0.009258048608899117,
"learning_rate": 1.8087986463620982e-05,
"loss": 0.0012,
"step": 340
},
{
"epoch": 0.2961082910321489,
"grad_norm": 0.009454768151044846,
"learning_rate": 1.8031584884376765e-05,
"loss": 0.0015,
"step": 350
},
{
"epoch": 0.30456852791878175,
"grad_norm": 0.018366724252700806,
"learning_rate": 1.7975183305132544e-05,
"loss": 0.0012,
"step": 360
},
{
"epoch": 0.3130287648054145,
"grad_norm": 0.009252658113837242,
"learning_rate": 1.7918781725888327e-05,
"loss": 0.0011,
"step": 370
},
{
"epoch": 0.32148900169204736,
"grad_norm": 0.1430797576904297,
"learning_rate": 1.7862380146644106e-05,
"loss": 0.0018,
"step": 380
},
{
"epoch": 0.3299492385786802,
"grad_norm": 0.009182159788906574,
"learning_rate": 1.780597856739989e-05,
"loss": 0.0018,
"step": 390
},
{
"epoch": 0.338409475465313,
"grad_norm": 0.00843009352684021,
"learning_rate": 1.774957698815567e-05,
"loss": 0.0017,
"step": 400
},
{
"epoch": 0.34686971235194586,
"grad_norm": 0.008176930248737335,
"learning_rate": 1.769317540891145e-05,
"loss": 0.0013,
"step": 410
},
{
"epoch": 0.3553299492385787,
"grad_norm": 0.008539380505681038,
"learning_rate": 1.7636773829667233e-05,
"loss": 0.001,
"step": 420
},
{
"epoch": 0.3637901861252115,
"grad_norm": 0.008693977259099483,
"learning_rate": 1.7580372250423013e-05,
"loss": 0.0014,
"step": 430
},
{
"epoch": 0.37225042301184436,
"grad_norm": 0.009141940623521805,
"learning_rate": 1.7523970671178795e-05,
"loss": 0.0012,
"step": 440
},
{
"epoch": 0.38071065989847713,
"grad_norm": 0.009133282117545605,
"learning_rate": 1.7467569091934575e-05,
"loss": 0.0015,
"step": 450
},
{
"epoch": 0.38917089678510997,
"grad_norm": 0.5799979567527771,
"learning_rate": 1.7411167512690357e-05,
"loss": 0.0013,
"step": 460
},
{
"epoch": 0.3976311336717428,
"grad_norm": 0.04120843857526779,
"learning_rate": 1.7354765933446137e-05,
"loss": 0.0011,
"step": 470
},
{
"epoch": 0.40609137055837563,
"grad_norm": 0.01276948768645525,
"learning_rate": 1.729836435420192e-05,
"loss": 0.0012,
"step": 480
},
{
"epoch": 0.41455160744500846,
"grad_norm": 0.0076572224497795105,
"learning_rate": 1.7241962774957702e-05,
"loss": 0.0014,
"step": 490
},
{
"epoch": 0.4230118443316413,
"grad_norm": 0.008293437771499157,
"learning_rate": 1.718556119571348e-05,
"loss": 0.0009,
"step": 500
},
{
"epoch": 0.43147208121827413,
"grad_norm": 0.0068985214456915855,
"learning_rate": 1.712915961646926e-05,
"loss": 0.0009,
"step": 510
},
{
"epoch": 0.43993231810490696,
"grad_norm": 0.029666345566511154,
"learning_rate": 1.7072758037225043e-05,
"loss": 0.0009,
"step": 520
},
{
"epoch": 0.44839255499153974,
"grad_norm": 0.014065027236938477,
"learning_rate": 1.7016356457980826e-05,
"loss": 0.0008,
"step": 530
},
{
"epoch": 0.45685279187817257,
"grad_norm": 0.00650499714538455,
"learning_rate": 1.6959954878736606e-05,
"loss": 0.0008,
"step": 540
},
{
"epoch": 0.4653130287648054,
"grad_norm": 0.0065047661773860455,
"learning_rate": 1.6903553299492388e-05,
"loss": 0.0008,
"step": 550
},
{
"epoch": 0.47377326565143824,
"grad_norm": 0.0071571446023881435,
"learning_rate": 1.684715172024817e-05,
"loss": 0.0008,
"step": 560
},
{
"epoch": 0.48223350253807107,
"grad_norm": 0.005875805858522654,
"learning_rate": 1.679075014100395e-05,
"loss": 0.0008,
"step": 570
},
{
"epoch": 0.4906937394247039,
"grad_norm": 0.005848041269928217,
"learning_rate": 1.673434856175973e-05,
"loss": 0.0007,
"step": 580
},
{
"epoch": 0.49915397631133673,
"grad_norm": 0.006032396107912064,
"learning_rate": 1.6677946982515512e-05,
"loss": 0.0007,
"step": 590
},
{
"epoch": 0.5076142131979695,
"grad_norm": 0.005858593620359898,
"learning_rate": 1.662154540327129e-05,
"loss": 0.0007,
"step": 600
},
{
"epoch": 0.5160744500846024,
"grad_norm": 0.006055026315152645,
"learning_rate": 1.6565143824027074e-05,
"loss": 0.0007,
"step": 610
},
{
"epoch": 0.5245346869712352,
"grad_norm": 0.0055991546250879765,
"learning_rate": 1.6508742244782857e-05,
"loss": 0.0039,
"step": 620
},
{
"epoch": 0.5329949238578681,
"grad_norm": 0.006043154280632734,
"learning_rate": 1.6452340665538636e-05,
"loss": 0.0009,
"step": 630
},
{
"epoch": 0.5414551607445008,
"grad_norm": 0.006406570319086313,
"learning_rate": 1.6395939086294416e-05,
"loss": 0.0008,
"step": 640
},
{
"epoch": 0.5499153976311336,
"grad_norm": 0.006317532621324062,
"learning_rate": 1.63395375070502e-05,
"loss": 0.0007,
"step": 650
},
{
"epoch": 0.5583756345177665,
"grad_norm": 0.0260737594217062,
"learning_rate": 1.628313592780598e-05,
"loss": 0.0007,
"step": 660
},
{
"epoch": 0.5668358714043993,
"grad_norm": 0.005516720470041037,
"learning_rate": 1.622673434856176e-05,
"loss": 0.0007,
"step": 670
},
{
"epoch": 0.5752961082910322,
"grad_norm": 0.005611425265669823,
"learning_rate": 1.6170332769317543e-05,
"loss": 0.0006,
"step": 680
},
{
"epoch": 0.583756345177665,
"grad_norm": 0.005698953289538622,
"learning_rate": 1.6113931190073326e-05,
"loss": 0.0006,
"step": 690
},
{
"epoch": 0.5922165820642978,
"grad_norm": 0.005045793950557709,
"learning_rate": 1.6057529610829105e-05,
"loss": 0.0006,
"step": 700
},
{
"epoch": 0.6006768189509306,
"grad_norm": 0.005404220428317785,
"learning_rate": 1.6001128031584884e-05,
"loss": 0.0007,
"step": 710
},
{
"epoch": 0.6091370558375635,
"grad_norm": 0.005528539884835482,
"learning_rate": 1.5944726452340667e-05,
"loss": 0.0007,
"step": 720
},
{
"epoch": 0.6175972927241963,
"grad_norm": 0.0052613625302910805,
"learning_rate": 1.5888324873096446e-05,
"loss": 0.0006,
"step": 730
},
{
"epoch": 0.626057529610829,
"grad_norm": 0.0050328257493674755,
"learning_rate": 1.583192329385223e-05,
"loss": 0.0006,
"step": 740
},
{
"epoch": 0.6345177664974619,
"grad_norm": 0.005106599070131779,
"learning_rate": 1.5775521714608012e-05,
"loss": 0.0006,
"step": 750
},
{
"epoch": 0.6429780033840947,
"grad_norm": 0.004836504813283682,
"learning_rate": 1.571912013536379e-05,
"loss": 0.0006,
"step": 760
},
{
"epoch": 0.6514382402707276,
"grad_norm": 0.004986999090760946,
"learning_rate": 1.566271855611957e-05,
"loss": 0.0006,
"step": 770
},
{
"epoch": 0.6598984771573604,
"grad_norm": 0.005235906690359116,
"learning_rate": 1.5606316976875353e-05,
"loss": 0.0005,
"step": 780
},
{
"epoch": 0.6683587140439933,
"grad_norm": 0.4625732898712158,
"learning_rate": 1.5549915397631136e-05,
"loss": 0.0009,
"step": 790
},
{
"epoch": 0.676818950930626,
"grad_norm": 0.004936546087265015,
"learning_rate": 1.5493513818386915e-05,
"loss": 0.0006,
"step": 800
},
{
"epoch": 0.6852791878172588,
"grad_norm": 0.004956043791025877,
"learning_rate": 1.5437112239142698e-05,
"loss": 0.0005,
"step": 810
},
{
"epoch": 0.6937394247038917,
"grad_norm": 0.005084918346256018,
"learning_rate": 1.538071065989848e-05,
"loss": 0.0005,
"step": 820
},
{
"epoch": 0.7021996615905245,
"grad_norm": 0.006050592288374901,
"learning_rate": 1.532430908065426e-05,
"loss": 0.0005,
"step": 830
},
{
"epoch": 0.7106598984771574,
"grad_norm": 0.0045923274010419846,
"learning_rate": 1.526790750141004e-05,
"loss": 0.0005,
"step": 840
},
{
"epoch": 0.7191201353637902,
"grad_norm": 0.004406214691698551,
"learning_rate": 1.5211505922165822e-05,
"loss": 0.0005,
"step": 850
},
{
"epoch": 0.727580372250423,
"grad_norm": 0.00449965288862586,
"learning_rate": 1.5155104342921603e-05,
"loss": 0.0005,
"step": 860
},
{
"epoch": 0.7360406091370558,
"grad_norm": 0.00427517294883728,
"learning_rate": 1.5098702763677384e-05,
"loss": 0.0006,
"step": 870
},
{
"epoch": 0.7445008460236887,
"grad_norm": 0.006013456266373396,
"learning_rate": 1.5042301184433165e-05,
"loss": 0.0005,
"step": 880
},
{
"epoch": 0.7529610829103215,
"grad_norm": 0.004456107504665852,
"learning_rate": 1.4985899605188948e-05,
"loss": 0.0005,
"step": 890
},
{
"epoch": 0.7614213197969543,
"grad_norm": 0.004223175812512636,
"learning_rate": 1.4929498025944729e-05,
"loss": 0.0005,
"step": 900
},
{
"epoch": 0.7698815566835872,
"grad_norm": 0.004154487047344446,
"learning_rate": 1.4873096446700508e-05,
"loss": 0.0005,
"step": 910
},
{
"epoch": 0.7783417935702199,
"grad_norm": 0.007055574096739292,
"learning_rate": 1.4816694867456289e-05,
"loss": 0.0005,
"step": 920
},
{
"epoch": 0.7868020304568528,
"grad_norm": 0.005269153974950314,
"learning_rate": 1.476029328821207e-05,
"loss": 0.0005,
"step": 930
},
{
"epoch": 0.7952622673434856,
"grad_norm": 0.004041856154799461,
"learning_rate": 1.4703891708967853e-05,
"loss": 0.0005,
"step": 940
},
{
"epoch": 0.8037225042301185,
"grad_norm": 0.003965588286519051,
"learning_rate": 1.4647490129723634e-05,
"loss": 0.0005,
"step": 950
},
{
"epoch": 0.8121827411167513,
"grad_norm": 0.005661304574459791,
"learning_rate": 1.4591088550479415e-05,
"loss": 0.0005,
"step": 960
},
{
"epoch": 0.8206429780033841,
"grad_norm": 0.0039055896922945976,
"learning_rate": 1.4534686971235198e-05,
"loss": 0.0011,
"step": 970
},
{
"epoch": 0.8291032148900169,
"grad_norm": 1.8310741186141968,
"learning_rate": 1.4478285391990977e-05,
"loss": 0.001,
"step": 980
},
{
"epoch": 0.8375634517766497,
"grad_norm": 0.005107107572257519,
"learning_rate": 1.4421883812746758e-05,
"loss": 0.0008,
"step": 990
},
{
"epoch": 0.8460236886632826,
"grad_norm": 0.005387498531490564,
"learning_rate": 1.4365482233502539e-05,
"loss": 0.0005,
"step": 1000
},
{
"epoch": 0.8544839255499154,
"grad_norm": 0.0051966467872262,
"learning_rate": 1.430908065425832e-05,
"loss": 0.0005,
"step": 1010
},
{
"epoch": 0.8629441624365483,
"grad_norm": 0.004603747744113207,
"learning_rate": 1.4252679075014103e-05,
"loss": 0.0005,
"step": 1020
},
{
"epoch": 0.871404399323181,
"grad_norm": 0.0113060986623168,
"learning_rate": 1.4196277495769884e-05,
"loss": 0.0005,
"step": 1030
},
{
"epoch": 0.8798646362098139,
"grad_norm": 0.004159980919212103,
"learning_rate": 1.4139875916525665e-05,
"loss": 0.0004,
"step": 1040
},
{
"epoch": 0.8883248730964467,
"grad_norm": 0.004094019532203674,
"learning_rate": 1.4083474337281444e-05,
"loss": 0.0004,
"step": 1050
},
{
"epoch": 0.8967851099830795,
"grad_norm": 0.00435257563367486,
"learning_rate": 1.4027072758037225e-05,
"loss": 0.0004,
"step": 1060
},
{
"epoch": 0.9052453468697124,
"grad_norm": 0.003843541955575347,
"learning_rate": 1.3970671178793008e-05,
"loss": 0.0004,
"step": 1070
},
{
"epoch": 0.9137055837563451,
"grad_norm": 0.16885782778263092,
"learning_rate": 1.3914269599548789e-05,
"loss": 0.0005,
"step": 1080
},
{
"epoch": 0.922165820642978,
"grad_norm": 0.0037802334409207106,
"learning_rate": 1.385786802030457e-05,
"loss": 0.0004,
"step": 1090
},
{
"epoch": 0.9306260575296108,
"grad_norm": 0.0037738329265266657,
"learning_rate": 1.3801466441060352e-05,
"loss": 0.0005,
"step": 1100
},
{
"epoch": 0.9390862944162437,
"grad_norm": 0.012493060901761055,
"learning_rate": 1.374506486181613e-05,
"loss": 0.0004,
"step": 1110
},
{
"epoch": 0.9475465313028765,
"grad_norm": 0.0037303464487195015,
"learning_rate": 1.3688663282571913e-05,
"loss": 0.0004,
"step": 1120
},
{
"epoch": 0.9560067681895094,
"grad_norm": 0.0035704418551176786,
"learning_rate": 1.3632261703327694e-05,
"loss": 0.0007,
"step": 1130
},
{
"epoch": 0.9644670050761421,
"grad_norm": 0.0038276014383882284,
"learning_rate": 1.3575860124083475e-05,
"loss": 0.0004,
"step": 1140
},
{
"epoch": 0.9729272419627749,
"grad_norm": 0.004172508604824543,
"learning_rate": 1.3519458544839257e-05,
"loss": 0.0004,
"step": 1150
},
{
"epoch": 0.9813874788494078,
"grad_norm": 0.20282192528247833,
"learning_rate": 1.3463056965595038e-05,
"loss": 0.0004,
"step": 1160
},
{
"epoch": 0.9898477157360406,
"grad_norm": 0.0035703997127711773,
"learning_rate": 1.340665538635082e-05,
"loss": 0.0004,
"step": 1170
},
{
"epoch": 0.9983079526226735,
"grad_norm": 0.003505186177790165,
"learning_rate": 1.3350253807106599e-05,
"loss": 0.0004,
"step": 1180
},
{
"epoch": 1.0067681895093064,
"grad_norm": 0.003632462117820978,
"learning_rate": 1.329385222786238e-05,
"loss": 0.0004,
"step": 1190
},
{
"epoch": 1.015228426395939,
"grad_norm": 0.0034904363565146923,
"learning_rate": 1.3237450648618163e-05,
"loss": 0.0004,
"step": 1200
},
{
"epoch": 1.023688663282572,
"grad_norm": 0.003400213085114956,
"learning_rate": 1.3181049069373944e-05,
"loss": 0.0004,
"step": 1210
},
{
"epoch": 1.0321489001692048,
"grad_norm": 0.02098667249083519,
"learning_rate": 1.3124647490129725e-05,
"loss": 0.0004,
"step": 1220
},
{
"epoch": 1.0406091370558375,
"grad_norm": 0.0033113211393356323,
"learning_rate": 1.3068245910885506e-05,
"loss": 0.0004,
"step": 1230
},
{
"epoch": 1.0490693739424704,
"grad_norm": 0.0075438846834003925,
"learning_rate": 1.3011844331641288e-05,
"loss": 0.001,
"step": 1240
},
{
"epoch": 1.0575296108291032,
"grad_norm": 0.003400578862056136,
"learning_rate": 1.2955442752397068e-05,
"loss": 0.0006,
"step": 1250
},
{
"epoch": 1.0659898477157361,
"grad_norm": 0.003571244655176997,
"learning_rate": 1.2899041173152849e-05,
"loss": 0.0004,
"step": 1260
},
{
"epoch": 1.0744500846023688,
"grad_norm": 0.0032794324215501547,
"learning_rate": 1.284263959390863e-05,
"loss": 0.0004,
"step": 1270
},
{
"epoch": 1.0829103214890017,
"grad_norm": 0.0032312178518623114,
"learning_rate": 1.2786238014664412e-05,
"loss": 0.0011,
"step": 1280
},
{
"epoch": 1.0913705583756346,
"grad_norm": 0.003224333981052041,
"learning_rate": 1.2729836435420193e-05,
"loss": 0.0003,
"step": 1290
},
{
"epoch": 1.0998307952622675,
"grad_norm": 0.0033167279325425625,
"learning_rate": 1.2673434856175974e-05,
"loss": 0.0004,
"step": 1300
},
{
"epoch": 1.1082910321489001,
"grad_norm": 0.003647689940407872,
"learning_rate": 1.2617033276931755e-05,
"loss": 0.0009,
"step": 1310
},
{
"epoch": 1.116751269035533,
"grad_norm": 0.005791259463876486,
"learning_rate": 1.2560631697687535e-05,
"loss": 0.0004,
"step": 1320
},
{
"epoch": 1.125211505922166,
"grad_norm": 0.0036105727776885033,
"learning_rate": 1.2504230118443317e-05,
"loss": 0.0005,
"step": 1330
},
{
"epoch": 1.1336717428087986,
"grad_norm": 0.0033262548968195915,
"learning_rate": 1.2447828539199098e-05,
"loss": 0.0003,
"step": 1340
},
{
"epoch": 1.1421319796954315,
"grad_norm": 0.003255200106650591,
"learning_rate": 1.239142695995488e-05,
"loss": 0.0003,
"step": 1350
},
{
"epoch": 1.1505922165820643,
"grad_norm": 0.0031492102425545454,
"learning_rate": 1.233502538071066e-05,
"loss": 0.0003,
"step": 1360
},
{
"epoch": 1.1590524534686972,
"grad_norm": 0.0031071833800524473,
"learning_rate": 1.2278623801466443e-05,
"loss": 0.0006,
"step": 1370
},
{
"epoch": 1.16751269035533,
"grad_norm": 0.003029242157936096,
"learning_rate": 1.2222222222222224e-05,
"loss": 0.0004,
"step": 1380
},
{
"epoch": 1.1759729272419628,
"grad_norm": 0.003054770641028881,
"learning_rate": 1.2165820642978003e-05,
"loss": 0.0004,
"step": 1390
},
{
"epoch": 1.1844331641285957,
"grad_norm": 0.002985539613291621,
"learning_rate": 1.2109419063733784e-05,
"loss": 0.0003,
"step": 1400
},
{
"epoch": 1.1928934010152283,
"grad_norm": 0.0029861750081181526,
"learning_rate": 1.2053017484489567e-05,
"loss": 0.0003,
"step": 1410
},
{
"epoch": 1.2013536379018612,
"grad_norm": 0.0031786304898560047,
"learning_rate": 1.1996615905245348e-05,
"loss": 0.0003,
"step": 1420
},
{
"epoch": 1.2098138747884941,
"grad_norm": 0.0029660591389983892,
"learning_rate": 1.194021432600113e-05,
"loss": 0.0003,
"step": 1430
},
{
"epoch": 1.218274111675127,
"grad_norm": 0.0029080845415592194,
"learning_rate": 1.188381274675691e-05,
"loss": 0.0003,
"step": 1440
},
{
"epoch": 1.2267343485617597,
"grad_norm": 0.003032978158444166,
"learning_rate": 1.1827411167512693e-05,
"loss": 0.0003,
"step": 1450
},
{
"epoch": 1.2351945854483926,
"grad_norm": 0.0028783201705664396,
"learning_rate": 1.1771009588268472e-05,
"loss": 0.0003,
"step": 1460
},
{
"epoch": 1.2436548223350254,
"grad_norm": 0.002919497899711132,
"learning_rate": 1.1714608009024253e-05,
"loss": 0.0003,
"step": 1470
},
{
"epoch": 1.252115059221658,
"grad_norm": 0.002829916076734662,
"learning_rate": 1.1658206429780034e-05,
"loss": 0.0003,
"step": 1480
},
{
"epoch": 1.260575296108291,
"grad_norm": 0.0028766861651092768,
"learning_rate": 1.1601804850535815e-05,
"loss": 0.0003,
"step": 1490
},
{
"epoch": 1.2690355329949239,
"grad_norm": 0.0027290356811136007,
"learning_rate": 1.1545403271291598e-05,
"loss": 0.0003,
"step": 1500
},
{
"epoch": 1.2774957698815568,
"grad_norm": 0.0027461503632366657,
"learning_rate": 1.1489001692047379e-05,
"loss": 0.0003,
"step": 1510
},
{
"epoch": 1.2859560067681894,
"grad_norm": 0.0027623216155916452,
"learning_rate": 1.143260011280316e-05,
"loss": 0.0003,
"step": 1520
},
{
"epoch": 1.2944162436548223,
"grad_norm": 0.0028906099032610655,
"learning_rate": 1.137619853355894e-05,
"loss": 0.0003,
"step": 1530
},
{
"epoch": 1.3028764805414552,
"grad_norm": 0.0027789094019681215,
"learning_rate": 1.1319796954314722e-05,
"loss": 0.0003,
"step": 1540
},
{
"epoch": 1.3113367174280879,
"grad_norm": 0.00271011283621192,
"learning_rate": 1.1263395375070503e-05,
"loss": 0.0003,
"step": 1550
},
{
"epoch": 1.3197969543147208,
"grad_norm": 0.002672335831448436,
"learning_rate": 1.1206993795826284e-05,
"loss": 0.0003,
"step": 1560
},
{
"epoch": 1.3282571912013537,
"grad_norm": 0.002650258131325245,
"learning_rate": 1.1150592216582065e-05,
"loss": 0.0003,
"step": 1570
},
{
"epoch": 1.3367174280879865,
"grad_norm": 0.002780508017167449,
"learning_rate": 1.1094190637337848e-05,
"loss": 0.0003,
"step": 1580
},
{
"epoch": 1.3451776649746192,
"grad_norm": 0.0026137123350054026,
"learning_rate": 1.1037789058093627e-05,
"loss": 0.0003,
"step": 1590
},
{
"epoch": 1.353637901861252,
"grad_norm": 0.0026269517838954926,
"learning_rate": 1.0981387478849408e-05,
"loss": 0.0003,
"step": 1600
},
{
"epoch": 1.362098138747885,
"grad_norm": 0.0025848057121038437,
"learning_rate": 1.092498589960519e-05,
"loss": 0.0003,
"step": 1610
},
{
"epoch": 1.3705583756345177,
"grad_norm": 0.0025470599066466093,
"learning_rate": 1.086858432036097e-05,
"loss": 0.0003,
"step": 1620
},
{
"epoch": 1.3790186125211505,
"grad_norm": 0.0027439745608717203,
"learning_rate": 1.0812182741116753e-05,
"loss": 0.0003,
"step": 1630
},
{
"epoch": 1.3874788494077834,
"grad_norm": 0.002537579508498311,
"learning_rate": 1.0755781161872534e-05,
"loss": 0.0003,
"step": 1640
},
{
"epoch": 1.3959390862944163,
"grad_norm": 0.0025688863825052977,
"learning_rate": 1.0699379582628315e-05,
"loss": 0.0003,
"step": 1650
},
{
"epoch": 1.404399323181049,
"grad_norm": 0.0026179850101470947,
"learning_rate": 1.0642978003384094e-05,
"loss": 0.0003,
"step": 1660
},
{
"epoch": 1.4128595600676819,
"grad_norm": 0.002496235305443406,
"learning_rate": 1.0586576424139875e-05,
"loss": 0.0003,
"step": 1670
},
{
"epoch": 1.4213197969543148,
"grad_norm": 0.0026708145160228014,
"learning_rate": 1.0530174844895658e-05,
"loss": 0.0003,
"step": 1680
},
{
"epoch": 1.4297800338409474,
"grad_norm": 0.0025831579696387053,
"learning_rate": 1.0473773265651439e-05,
"loss": 0.0003,
"step": 1690
},
{
"epoch": 1.4382402707275803,
"grad_norm": 0.0024914774112403393,
"learning_rate": 1.041737168640722e-05,
"loss": 0.0003,
"step": 1700
},
{
"epoch": 1.4467005076142132,
"grad_norm": 0.002529192017391324,
"learning_rate": 1.0360970107163003e-05,
"loss": 0.0003,
"step": 1710
},
{
"epoch": 1.455160744500846,
"grad_norm": 0.0024393699131906033,
"learning_rate": 1.0304568527918784e-05,
"loss": 0.0011,
"step": 1720
},
{
"epoch": 1.463620981387479,
"grad_norm": 0.0043233660981059074,
"learning_rate": 1.0248166948674563e-05,
"loss": 0.0004,
"step": 1730
},
{
"epoch": 1.4720812182741116,
"grad_norm": 0.002547438722103834,
"learning_rate": 1.0191765369430344e-05,
"loss": 0.0005,
"step": 1740
},
{
"epoch": 1.4805414551607445,
"grad_norm": 0.0025178396608680487,
"learning_rate": 1.0135363790186125e-05,
"loss": 0.0003,
"step": 1750
},
{
"epoch": 1.4890016920473772,
"grad_norm": 0.003602534532546997,
"learning_rate": 1.0078962210941908e-05,
"loss": 0.0003,
"step": 1760
},
{
"epoch": 1.49746192893401,
"grad_norm": 0.002990358741953969,
"learning_rate": 1.0022560631697689e-05,
"loss": 0.0003,
"step": 1770
},
{
"epoch": 1.505922165820643,
"grad_norm": 0.0024538026191294193,
"learning_rate": 9.96615905245347e-06,
"loss": 0.0003,
"step": 1780
},
{
"epoch": 1.5143824027072759,
"grad_norm": 0.0024382509291172028,
"learning_rate": 9.90975747320925e-06,
"loss": 0.0003,
"step": 1790
},
{
"epoch": 1.5228426395939088,
"grad_norm": 0.0026386724784970284,
"learning_rate": 9.853355893965032e-06,
"loss": 0.0003,
"step": 1800
},
{
"epoch": 1.5313028764805414,
"grad_norm": 0.0023787450045347214,
"learning_rate": 9.796954314720813e-06,
"loss": 0.0003,
"step": 1810
},
{
"epoch": 1.5397631133671743,
"grad_norm": 0.002435258822515607,
"learning_rate": 9.740552735476594e-06,
"loss": 0.0003,
"step": 1820
},
{
"epoch": 1.548223350253807,
"grad_norm": 0.32252246141433716,
"learning_rate": 9.684151156232375e-06,
"loss": 0.0003,
"step": 1830
},
{
"epoch": 1.5566835871404399,
"grad_norm": 0.0023758108727633953,
"learning_rate": 9.627749576988158e-06,
"loss": 0.0003,
"step": 1840
},
{
"epoch": 1.5651438240270727,
"grad_norm": 0.0023211485240608454,
"learning_rate": 9.571347997743937e-06,
"loss": 0.0003,
"step": 1850
},
{
"epoch": 1.5736040609137056,
"grad_norm": 0.002314375713467598,
"learning_rate": 9.514946418499718e-06,
"loss": 0.0003,
"step": 1860
},
{
"epoch": 1.5820642978003385,
"grad_norm": 0.0023043151013553143,
"learning_rate": 9.4585448392555e-06,
"loss": 0.0003,
"step": 1870
},
{
"epoch": 1.5905245346869712,
"grad_norm": 0.0022787386551499367,
"learning_rate": 9.402143260011282e-06,
"loss": 0.0002,
"step": 1880
},
{
"epoch": 1.598984771573604,
"grad_norm": 0.0022805132903158665,
"learning_rate": 9.345741680767063e-06,
"loss": 0.0003,
"step": 1890
},
{
"epoch": 1.6074450084602367,
"grad_norm": 0.0022434191778302193,
"learning_rate": 9.289340101522844e-06,
"loss": 0.0003,
"step": 1900
},
{
"epoch": 1.6159052453468696,
"grad_norm": 0.0022070412524044514,
"learning_rate": 9.232938522278625e-06,
"loss": 0.0002,
"step": 1910
},
{
"epoch": 1.6243654822335025,
"grad_norm": 0.002298792591318488,
"learning_rate": 9.176536943034406e-06,
"loss": 0.0002,
"step": 1920
},
{
"epoch": 1.6328257191201354,
"grad_norm": 0.002347590634599328,
"learning_rate": 9.120135363790187e-06,
"loss": 0.0002,
"step": 1930
},
{
"epoch": 1.6412859560067683,
"grad_norm": 0.0022417185828089714,
"learning_rate": 9.063733784545968e-06,
"loss": 0.0002,
"step": 1940
},
{
"epoch": 1.649746192893401,
"grad_norm": 0.002178118098527193,
"learning_rate": 9.007332205301749e-06,
"loss": 0.0002,
"step": 1950
},
{
"epoch": 1.6582064297800339,
"grad_norm": 0.002204876858741045,
"learning_rate": 8.95093062605753e-06,
"loss": 0.0002,
"step": 1960
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.0021794794593006372,
"learning_rate": 8.894529046813312e-06,
"loss": 0.0002,
"step": 1970
},
{
"epoch": 1.6751269035532994,
"grad_norm": 0.002448797458782792,
"learning_rate": 8.838127467569093e-06,
"loss": 0.0002,
"step": 1980
},
{
"epoch": 1.6835871404399323,
"grad_norm": 0.00216574571095407,
"learning_rate": 8.781725888324873e-06,
"loss": 0.0002,
"step": 1990
},
{
"epoch": 1.6920473773265652,
"grad_norm": 0.0022357029374688864,
"learning_rate": 8.725324309080655e-06,
"loss": 0.0002,
"step": 2000
},
{
"epoch": 1.700507614213198,
"grad_norm": 0.0021086037158966064,
"learning_rate": 8.668922729836436e-06,
"loss": 0.0002,
"step": 2010
},
{
"epoch": 1.708967851099831,
"grad_norm": 0.002162441611289978,
"learning_rate": 8.612521150592217e-06,
"loss": 0.0002,
"step": 2020
},
{
"epoch": 1.7174280879864636,
"grad_norm": 0.0021641228813678026,
"learning_rate": 8.556119571347998e-06,
"loss": 0.0011,
"step": 2030
},
{
"epoch": 1.7258883248730963,
"grad_norm": 0.0021566457580775023,
"learning_rate": 8.49971799210378e-06,
"loss": 0.0002,
"step": 2040
},
{
"epoch": 1.7343485617597292,
"grad_norm": 0.0021678991615772247,
"learning_rate": 8.44331641285956e-06,
"loss": 0.0002,
"step": 2050
},
{
"epoch": 1.742808798646362,
"grad_norm": 0.002119549084454775,
"learning_rate": 8.386914833615342e-06,
"loss": 0.0002,
"step": 2060
},
{
"epoch": 1.751269035532995,
"grad_norm": 0.002093096962198615,
"learning_rate": 8.330513254371123e-06,
"loss": 0.0002,
"step": 2070
},
{
"epoch": 1.7597292724196278,
"grad_norm": 0.002090982161462307,
"learning_rate": 8.274111675126905e-06,
"loss": 0.0002,
"step": 2080
},
{
"epoch": 1.7681895093062607,
"grad_norm": 0.004076903220266104,
"learning_rate": 8.217710095882685e-06,
"loss": 0.0002,
"step": 2090
},
{
"epoch": 1.7766497461928934,
"grad_norm": 0.0020308117382228374,
"learning_rate": 8.161308516638467e-06,
"loss": 0.0002,
"step": 2100
},
{
"epoch": 1.785109983079526,
"grad_norm": 0.0021146961953490973,
"learning_rate": 8.104906937394248e-06,
"loss": 0.0002,
"step": 2110
},
{
"epoch": 1.793570219966159,
"grad_norm": 0.0020670590456575155,
"learning_rate": 8.048505358150028e-06,
"loss": 0.0002,
"step": 2120
},
{
"epoch": 1.8020304568527918,
"grad_norm": 0.0024381864350289106,
"learning_rate": 7.99210377890581e-06,
"loss": 0.0002,
"step": 2130
},
{
"epoch": 1.8104906937394247,
"grad_norm": 0.002149208215996623,
"learning_rate": 7.935702199661591e-06,
"loss": 0.0008,
"step": 2140
},
{
"epoch": 1.8189509306260576,
"grad_norm": 0.004185215570032597,
"learning_rate": 7.879300620417372e-06,
"loss": 0.0002,
"step": 2150
},
{
"epoch": 1.8274111675126905,
"grad_norm": 0.002098744735121727,
"learning_rate": 7.822899041173153e-06,
"loss": 0.0002,
"step": 2160
},
{
"epoch": 1.8358714043993232,
"grad_norm": 0.0019923567306250334,
"learning_rate": 7.766497461928934e-06,
"loss": 0.0004,
"step": 2170
},
{
"epoch": 1.844331641285956,
"grad_norm": 0.002045057248324156,
"learning_rate": 7.710095882684715e-06,
"loss": 0.0002,
"step": 2180
},
{
"epoch": 1.8527918781725887,
"grad_norm": 0.002222336595878005,
"learning_rate": 7.653694303440496e-06,
"loss": 0.0002,
"step": 2190
},
{
"epoch": 1.8612521150592216,
"grad_norm": 0.0021004409063607454,
"learning_rate": 7.597292724196278e-06,
"loss": 0.0002,
"step": 2200
},
{
"epoch": 1.8697123519458545,
"grad_norm": 0.001979230437427759,
"learning_rate": 7.540891144952059e-06,
"loss": 0.0002,
"step": 2210
},
{
"epoch": 1.8781725888324874,
"grad_norm": 0.0020105023868381977,
"learning_rate": 7.484489565707841e-06,
"loss": 0.0002,
"step": 2220
},
{
"epoch": 1.8866328257191203,
"grad_norm": 0.0020976634696125984,
"learning_rate": 7.428087986463621e-06,
"loss": 0.0002,
"step": 2230
},
{
"epoch": 1.895093062605753,
"grad_norm": 0.0020316061563789845,
"learning_rate": 7.371686407219403e-06,
"loss": 0.0002,
"step": 2240
},
{
"epoch": 1.9035532994923858,
"grad_norm": 0.00200808048248291,
"learning_rate": 7.315284827975184e-06,
"loss": 0.0002,
"step": 2250
},
{
"epoch": 1.9120135363790185,
"grad_norm": 0.002250692341476679,
"learning_rate": 7.258883248730964e-06,
"loss": 0.0002,
"step": 2260
},
{
"epoch": 1.9204737732656514,
"grad_norm": 0.0022177433129400015,
"learning_rate": 7.202481669486746e-06,
"loss": 0.0007,
"step": 2270
},
{
"epoch": 1.9289340101522843,
"grad_norm": 0.0018741836538538337,
"learning_rate": 7.146080090242527e-06,
"loss": 0.0002,
"step": 2280
},
{
"epoch": 1.9373942470389172,
"grad_norm": 0.002023757202550769,
"learning_rate": 7.089678510998309e-06,
"loss": 0.0002,
"step": 2290
},
{
"epoch": 1.94585448392555,
"grad_norm": 0.0021521553862839937,
"learning_rate": 7.033276931754089e-06,
"loss": 0.0002,
"step": 2300
},
{
"epoch": 1.9543147208121827,
"grad_norm": 0.0019342171726748347,
"learning_rate": 6.976875352509871e-06,
"loss": 0.0008,
"step": 2310
},
{
"epoch": 1.9627749576988156,
"grad_norm": 0.0019481374183669686,
"learning_rate": 6.920473773265652e-06,
"loss": 0.0007,
"step": 2320
},
{
"epoch": 1.9712351945854483,
"grad_norm": 0.002077361335977912,
"learning_rate": 6.864072194021433e-06,
"loss": 0.0011,
"step": 2330
},
{
"epoch": 1.9796954314720812,
"grad_norm": 0.0020031894091516733,
"learning_rate": 6.807670614777214e-06,
"loss": 0.0003,
"step": 2340
},
{
"epoch": 1.988155668358714,
"grad_norm": 0.00219974503852427,
"learning_rate": 6.751269035532996e-06,
"loss": 0.0002,
"step": 2350
},
{
"epoch": 1.996615905245347,
"grad_norm": 0.002148480387404561,
"learning_rate": 6.694867456288776e-06,
"loss": 0.0002,
"step": 2360
},
{
"epoch": 2.00507614213198,
"grad_norm": 0.0018948954530060291,
"learning_rate": 6.638465877044558e-06,
"loss": 0.0002,
"step": 2370
},
{
"epoch": 2.0135363790186127,
"grad_norm": 0.007534192409366369,
"learning_rate": 6.582064297800339e-06,
"loss": 0.0002,
"step": 2380
},
{
"epoch": 2.021996615905245,
"grad_norm": 0.0018520756857469678,
"learning_rate": 6.525662718556121e-06,
"loss": 0.0002,
"step": 2390
},
{
"epoch": 2.030456852791878,
"grad_norm": 0.0020236349664628506,
"learning_rate": 6.469261139311901e-06,
"loss": 0.0002,
"step": 2400
},
{
"epoch": 2.038917089678511,
"grad_norm": 0.0018428469775244594,
"learning_rate": 6.412859560067682e-06,
"loss": 0.0002,
"step": 2410
},
{
"epoch": 2.047377326565144,
"grad_norm": 0.0018592760898172855,
"learning_rate": 6.356457980823464e-06,
"loss": 0.0002,
"step": 2420
},
{
"epoch": 2.0558375634517767,
"grad_norm": 0.001955309882760048,
"learning_rate": 6.300056401579244e-06,
"loss": 0.0002,
"step": 2430
},
{
"epoch": 2.0642978003384096,
"grad_norm": 0.001861874246969819,
"learning_rate": 6.243654822335026e-06,
"loss": 0.0002,
"step": 2440
},
{
"epoch": 2.0727580372250425,
"grad_norm": 0.0018448787741363049,
"learning_rate": 6.187253243090807e-06,
"loss": 0.0002,
"step": 2450
},
{
"epoch": 2.081218274111675,
"grad_norm": 0.0018381918780505657,
"learning_rate": 6.130851663846589e-06,
"loss": 0.0002,
"step": 2460
},
{
"epoch": 2.089678510998308,
"grad_norm": 0.0018074287800118327,
"learning_rate": 6.074450084602369e-06,
"loss": 0.0002,
"step": 2470
},
{
"epoch": 2.0981387478849407,
"grad_norm": 0.0018625473603606224,
"learning_rate": 6.018048505358151e-06,
"loss": 0.0002,
"step": 2480
},
{
"epoch": 2.1065989847715736,
"grad_norm": 0.0017794481245800853,
"learning_rate": 5.961646926113932e-06,
"loss": 0.0002,
"step": 2490
},
{
"epoch": 2.1150592216582065,
"grad_norm": 0.0017721441108733416,
"learning_rate": 5.905245346869712e-06,
"loss": 0.0002,
"step": 2500
},
{
"epoch": 2.1235194585448394,
"grad_norm": 0.005279191303998232,
"learning_rate": 5.848843767625494e-06,
"loss": 0.0002,
"step": 2510
},
{
"epoch": 2.1319796954314723,
"grad_norm": 0.0017379368655383587,
"learning_rate": 5.792442188381276e-06,
"loss": 0.0002,
"step": 2520
},
{
"epoch": 2.1404399323181047,
"grad_norm": 0.001751760020852089,
"learning_rate": 5.736040609137057e-06,
"loss": 0.0002,
"step": 2530
},
{
"epoch": 2.1489001692047376,
"grad_norm": 0.001752077485434711,
"learning_rate": 5.679639029892837e-06,
"loss": 0.0002,
"step": 2540
},
{
"epoch": 2.1573604060913705,
"grad_norm": 0.0017231889069080353,
"learning_rate": 5.623237450648619e-06,
"loss": 0.0002,
"step": 2550
},
{
"epoch": 2.1658206429780034,
"grad_norm": 0.0017523319693282247,
"learning_rate": 5.5668358714044e-06,
"loss": 0.0002,
"step": 2560
},
{
"epoch": 2.1742808798646363,
"grad_norm": 0.0017014428740367293,
"learning_rate": 5.510434292160181e-06,
"loss": 0.0002,
"step": 2570
},
{
"epoch": 2.182741116751269,
"grad_norm": 0.0017529787728562951,
"learning_rate": 5.454032712915962e-06,
"loss": 0.0002,
"step": 2580
},
{
"epoch": 2.191201353637902,
"grad_norm": 0.0017768230754882097,
"learning_rate": 5.397631133671744e-06,
"loss": 0.0002,
"step": 2590
},
{
"epoch": 2.199661590524535,
"grad_norm": 0.0017734098946675658,
"learning_rate": 5.341229554427524e-06,
"loss": 0.0002,
"step": 2600
},
{
"epoch": 2.2081218274111674,
"grad_norm": 0.00168097042478621,
"learning_rate": 5.284827975183306e-06,
"loss": 0.0002,
"step": 2610
},
{
"epoch": 2.2165820642978002,
"grad_norm": 0.0017450157320126891,
"learning_rate": 5.228426395939087e-06,
"loss": 0.0002,
"step": 2620
},
{
"epoch": 2.225042301184433,
"grad_norm": 0.0016924841329455376,
"learning_rate": 5.1720248166948686e-06,
"loss": 0.0002,
"step": 2630
},
{
"epoch": 2.233502538071066,
"grad_norm": 0.0017653257818892598,
"learning_rate": 5.115623237450649e-06,
"loss": 0.0002,
"step": 2640
},
{
"epoch": 2.241962774957699,
"grad_norm": 0.00172147445846349,
"learning_rate": 5.059221658206431e-06,
"loss": 0.0002,
"step": 2650
},
{
"epoch": 2.250423011844332,
"grad_norm": 0.0017959652468562126,
"learning_rate": 5.002820078962212e-06,
"loss": 0.0002,
"step": 2660
},
{
"epoch": 2.2588832487309647,
"grad_norm": 0.0016738857375457883,
"learning_rate": 4.946418499717993e-06,
"loss": 0.0002,
"step": 2670
},
{
"epoch": 2.267343485617597,
"grad_norm": 0.0016783688915893435,
"learning_rate": 4.890016920473774e-06,
"loss": 0.0002,
"step": 2680
},
{
"epoch": 2.27580372250423,
"grad_norm": 0.001678799162618816,
"learning_rate": 4.833615341229555e-06,
"loss": 0.0002,
"step": 2690
},
{
"epoch": 2.284263959390863,
"grad_norm": 0.0017365720123052597,
"learning_rate": 4.777213761985336e-06,
"loss": 0.0002,
"step": 2700
},
{
"epoch": 2.292724196277496,
"grad_norm": 0.00171087181661278,
"learning_rate": 4.7208121827411175e-06,
"loss": 0.0002,
"step": 2710
},
{
"epoch": 2.3011844331641287,
"grad_norm": 0.001663476345129311,
"learning_rate": 4.6644106034968985e-06,
"loss": 0.0002,
"step": 2720
},
{
"epoch": 2.3096446700507616,
"grad_norm": 0.0016244918806478381,
"learning_rate": 4.6080090242526796e-06,
"loss": 0.0002,
"step": 2730
},
{
"epoch": 2.3181049069373945,
"grad_norm": 0.0016214584466069937,
"learning_rate": 4.5516074450084606e-06,
"loss": 0.0002,
"step": 2740
},
{
"epoch": 2.326565143824027,
"grad_norm": 0.0016812806716188788,
"learning_rate": 4.4952058657642416e-06,
"loss": 0.0002,
"step": 2750
},
{
"epoch": 2.33502538071066,
"grad_norm": 0.0016772805247455835,
"learning_rate": 4.4388042865200234e-06,
"loss": 0.0002,
"step": 2760
},
{
"epoch": 2.3434856175972927,
"grad_norm": 0.0016064831288531423,
"learning_rate": 4.3824027072758044e-06,
"loss": 0.0002,
"step": 2770
},
{
"epoch": 2.3519458544839256,
"grad_norm": 0.0016283520963042974,
"learning_rate": 4.326001128031585e-06,
"loss": 0.0002,
"step": 2780
},
{
"epoch": 2.3604060913705585,
"grad_norm": 0.0017511517507955432,
"learning_rate": 4.2695995487873665e-06,
"loss": 0.0002,
"step": 2790
},
{
"epoch": 2.3688663282571913,
"grad_norm": 0.001622919924557209,
"learning_rate": 4.2131979695431475e-06,
"loss": 0.0002,
"step": 2800
},
{
"epoch": 2.3773265651438242,
"grad_norm": 0.0017815829487517476,
"learning_rate": 4.1567963902989285e-06,
"loss": 0.0004,
"step": 2810
},
{
"epoch": 2.3857868020304567,
"grad_norm": 0.001768069458194077,
"learning_rate": 4.1003948110547095e-06,
"loss": 0.0002,
"step": 2820
},
{
"epoch": 2.3942470389170896,
"grad_norm": 0.001824942184612155,
"learning_rate": 4.0439932318104905e-06,
"loss": 0.0002,
"step": 2830
},
{
"epoch": 2.4027072758037225,
"grad_norm": 0.0016344970790669322,
"learning_rate": 3.987591652566272e-06,
"loss": 0.0005,
"step": 2840
},
{
"epoch": 2.4111675126903553,
"grad_norm": 0.0016487556276842952,
"learning_rate": 3.931190073322053e-06,
"loss": 0.0002,
"step": 2850
},
{
"epoch": 2.4196277495769882,
"grad_norm": 0.0016864290228113532,
"learning_rate": 3.874788494077834e-06,
"loss": 0.0002,
"step": 2860
},
{
"epoch": 2.428087986463621,
"grad_norm": 0.0017508761957287788,
"learning_rate": 3.8183869148336154e-06,
"loss": 0.0002,
"step": 2870
},
{
"epoch": 2.436548223350254,
"grad_norm": 0.0016393917612731457,
"learning_rate": 3.761985335589397e-06,
"loss": 0.0002,
"step": 2880
},
{
"epoch": 2.4450084602368864,
"grad_norm": 0.001650340622290969,
"learning_rate": 3.705583756345178e-06,
"loss": 0.0002,
"step": 2890
},
{
"epoch": 2.4534686971235193,
"grad_norm": 0.0016186356078833342,
"learning_rate": 3.649182177100959e-06,
"loss": 0.0002,
"step": 2900
},
{
"epoch": 2.4619289340101522,
"grad_norm": 0.001632276806049049,
"learning_rate": 3.5927805978567403e-06,
"loss": 0.0002,
"step": 2910
},
{
"epoch": 2.470389170896785,
"grad_norm": 0.0017024771077558398,
"learning_rate": 3.5363790186125213e-06,
"loss": 0.0002,
"step": 2920
},
{
"epoch": 2.478849407783418,
"grad_norm": 0.0015938036376610398,
"learning_rate": 3.4799774393683028e-06,
"loss": 0.0002,
"step": 2930
},
{
"epoch": 2.487309644670051,
"grad_norm": 0.001573930960148573,
"learning_rate": 3.4235758601240838e-06,
"loss": 0.0002,
"step": 2940
},
{
"epoch": 2.495769881556684,
"grad_norm": 0.0015760001260787249,
"learning_rate": 3.367174280879865e-06,
"loss": 0.0002,
"step": 2950
},
{
"epoch": 2.504230118443316,
"grad_norm": 0.0016050884732976556,
"learning_rate": 3.3107727016356462e-06,
"loss": 0.0002,
"step": 2960
},
{
"epoch": 2.512690355329949,
"grad_norm": 0.0017449154984205961,
"learning_rate": 3.2543711223914273e-06,
"loss": 0.0002,
"step": 2970
},
{
"epoch": 2.521150592216582,
"grad_norm": 0.0015464244643226266,
"learning_rate": 3.1979695431472087e-06,
"loss": 0.0002,
"step": 2980
},
{
"epoch": 2.529610829103215,
"grad_norm": 0.0016948716947808862,
"learning_rate": 3.1415679639029893e-06,
"loss": 0.0002,
"step": 2990
},
{
"epoch": 2.5380710659898478,
"grad_norm": 0.0015749187441542745,
"learning_rate": 3.085166384658771e-06,
"loss": 0.0002,
"step": 3000
}
],
"logging_steps": 10,
"max_steps": 3546,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 895710243680028.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}