Llama-3.2-3B-SFT-RM-bm / trainer_state.json
rshwndsz's picture
Upload folder using huggingface_hub
b5b2e70 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9997593647228684,
"eval_steps": 500,
"global_step": 3116,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.003208470361755033,
"grad_norm": 6.442734116834532,
"learning_rate": 1.9942233632862646e-05,
"loss": 0.646,
"step": 10
},
{
"epoch": 0.006416940723510066,
"grad_norm": 16.670848007363354,
"learning_rate": 1.9878048780487806e-05,
"loss": 0.5538,
"step": 20
},
{
"epoch": 0.0096254110852651,
"grad_norm": 5.007229537641644,
"learning_rate": 1.9813863928112966e-05,
"loss": 0.5532,
"step": 30
},
{
"epoch": 0.012833881447020133,
"grad_norm": 6.381096528206163,
"learning_rate": 1.974967907573813e-05,
"loss": 0.5403,
"step": 40
},
{
"epoch": 0.016042351808775165,
"grad_norm": 3.202143132985174,
"learning_rate": 1.968549422336329e-05,
"loss": 0.5484,
"step": 50
},
{
"epoch": 0.0192508221705302,
"grad_norm": 5.974056462183794,
"learning_rate": 1.962130937098845e-05,
"loss": 0.5738,
"step": 60
},
{
"epoch": 0.022459292532285232,
"grad_norm": 8.235054035652611,
"learning_rate": 1.955712451861361e-05,
"loss": 0.5311,
"step": 70
},
{
"epoch": 0.025667762894040266,
"grad_norm": 8.918082235487985,
"learning_rate": 1.9492939666238767e-05,
"loss": 0.5128,
"step": 80
},
{
"epoch": 0.0288762332557953,
"grad_norm": 7.299533718731339,
"learning_rate": 1.942875481386393e-05,
"loss": 0.5444,
"step": 90
},
{
"epoch": 0.03208470361755033,
"grad_norm": 6.793642690075093,
"learning_rate": 1.936456996148909e-05,
"loss": 0.4792,
"step": 100
},
{
"epoch": 0.03529317397930536,
"grad_norm": 5.35608624427688,
"learning_rate": 1.930038510911425e-05,
"loss": 0.4838,
"step": 110
},
{
"epoch": 0.0385016443410604,
"grad_norm": 8.214012839502532,
"learning_rate": 1.923620025673941e-05,
"loss": 0.4947,
"step": 120
},
{
"epoch": 0.04171011470281543,
"grad_norm": 17.30363857344939,
"learning_rate": 1.9172015404364574e-05,
"loss": 0.539,
"step": 130
},
{
"epoch": 0.044918585064570464,
"grad_norm": 5.2411969812270405,
"learning_rate": 1.910783055198973e-05,
"loss": 0.4976,
"step": 140
},
{
"epoch": 0.0481270554263255,
"grad_norm": 5.825912105537252,
"learning_rate": 1.904364569961489e-05,
"loss": 0.5374,
"step": 150
},
{
"epoch": 0.05133552578808053,
"grad_norm": 7.038630553849513,
"learning_rate": 1.897946084724005e-05,
"loss": 0.4682,
"step": 160
},
{
"epoch": 0.054543996149835565,
"grad_norm": 5.686461374890298,
"learning_rate": 1.8915275994865214e-05,
"loss": 0.5205,
"step": 170
},
{
"epoch": 0.0577524665115906,
"grad_norm": 5.179564550909498,
"learning_rate": 1.8851091142490375e-05,
"loss": 0.491,
"step": 180
},
{
"epoch": 0.06096093687334563,
"grad_norm": 8.463629137447528,
"learning_rate": 1.8786906290115535e-05,
"loss": 0.5228,
"step": 190
},
{
"epoch": 0.06416940723510066,
"grad_norm": 4.011565557460287,
"learning_rate": 1.8722721437740695e-05,
"loss": 0.479,
"step": 200
},
{
"epoch": 0.06737787759685569,
"grad_norm": 7.7055557079808805,
"learning_rate": 1.8658536585365855e-05,
"loss": 0.4687,
"step": 210
},
{
"epoch": 0.07058634795861073,
"grad_norm": 6.724744753360089,
"learning_rate": 1.8594351732991015e-05,
"loss": 0.4818,
"step": 220
},
{
"epoch": 0.07379481832036576,
"grad_norm": 4.678677086919036,
"learning_rate": 1.8530166880616175e-05,
"loss": 0.5139,
"step": 230
},
{
"epoch": 0.0770032886821208,
"grad_norm": 9.311365908507138,
"learning_rate": 1.8465982028241335e-05,
"loss": 0.4747,
"step": 240
},
{
"epoch": 0.08021175904387583,
"grad_norm": 4.443842779885137,
"learning_rate": 1.8401797175866495e-05,
"loss": 0.4891,
"step": 250
},
{
"epoch": 0.08342022940563086,
"grad_norm": 8.33835692877245,
"learning_rate": 1.833761232349166e-05,
"loss": 0.47,
"step": 260
},
{
"epoch": 0.0866286997673859,
"grad_norm": 6.389873022971522,
"learning_rate": 1.827342747111682e-05,
"loss": 0.4484,
"step": 270
},
{
"epoch": 0.08983717012914093,
"grad_norm": 7.274203538153994,
"learning_rate": 1.820924261874198e-05,
"loss": 0.4517,
"step": 280
},
{
"epoch": 0.09304564049089596,
"grad_norm": 3.978695084684557,
"learning_rate": 1.814505776636714e-05,
"loss": 0.4807,
"step": 290
},
{
"epoch": 0.096254110852651,
"grad_norm": 5.846527325287561,
"learning_rate": 1.80808729139923e-05,
"loss": 0.4572,
"step": 300
},
{
"epoch": 0.09946258121440603,
"grad_norm": 4.899874671814943,
"learning_rate": 1.801668806161746e-05,
"loss": 0.4337,
"step": 310
},
{
"epoch": 0.10267105157616106,
"grad_norm": 7.441397827942358,
"learning_rate": 1.795250320924262e-05,
"loss": 0.4794,
"step": 320
},
{
"epoch": 0.1058795219379161,
"grad_norm": 4.8167626402693,
"learning_rate": 1.788831835686778e-05,
"loss": 0.4547,
"step": 330
},
{
"epoch": 0.10908799229967113,
"grad_norm": 4.830906300715028,
"learning_rate": 1.7824133504492943e-05,
"loss": 0.4155,
"step": 340
},
{
"epoch": 0.11229646266142616,
"grad_norm": 2.783779241391255,
"learning_rate": 1.7759948652118103e-05,
"loss": 0.4254,
"step": 350
},
{
"epoch": 0.1155049330231812,
"grad_norm": 4.482561979814794,
"learning_rate": 1.7695763799743263e-05,
"loss": 0.5048,
"step": 360
},
{
"epoch": 0.11871340338493623,
"grad_norm": 4.736082727818665,
"learning_rate": 1.763157894736842e-05,
"loss": 0.4088,
"step": 370
},
{
"epoch": 0.12192187374669126,
"grad_norm": 3.540097005993698,
"learning_rate": 1.7567394094993584e-05,
"loss": 0.5214,
"step": 380
},
{
"epoch": 0.1251303441084463,
"grad_norm": 4.695488227787937,
"learning_rate": 1.7503209242618744e-05,
"loss": 0.4434,
"step": 390
},
{
"epoch": 0.12833881447020132,
"grad_norm": 3.7024498981669662,
"learning_rate": 1.7439024390243904e-05,
"loss": 0.5021,
"step": 400
},
{
"epoch": 0.13154728483195635,
"grad_norm": 2.707105021471608,
"learning_rate": 1.7374839537869064e-05,
"loss": 0.4844,
"step": 410
},
{
"epoch": 0.13475575519371139,
"grad_norm": 5.70948950157765,
"learning_rate": 1.7310654685494224e-05,
"loss": 0.4903,
"step": 420
},
{
"epoch": 0.13796422555546642,
"grad_norm": 3.696922556933147,
"learning_rate": 1.7246469833119384e-05,
"loss": 0.4441,
"step": 430
},
{
"epoch": 0.14117269591722145,
"grad_norm": 4.483329616797758,
"learning_rate": 1.7182284980744544e-05,
"loss": 0.5387,
"step": 440
},
{
"epoch": 0.14438116627897649,
"grad_norm": 3.394450035525867,
"learning_rate": 1.7118100128369705e-05,
"loss": 0.4662,
"step": 450
},
{
"epoch": 0.14758963664073152,
"grad_norm": 8.726490331816862,
"learning_rate": 1.7053915275994865e-05,
"loss": 0.4507,
"step": 460
},
{
"epoch": 0.15079810700248655,
"grad_norm": 4.353194363650416,
"learning_rate": 1.6989730423620028e-05,
"loss": 0.4526,
"step": 470
},
{
"epoch": 0.1540065773642416,
"grad_norm": 3.950021174504244,
"learning_rate": 1.692554557124519e-05,
"loss": 0.4821,
"step": 480
},
{
"epoch": 0.15721504772599662,
"grad_norm": 3.5715866632711832,
"learning_rate": 1.686136071887035e-05,
"loss": 0.4288,
"step": 490
},
{
"epoch": 0.16042351808775165,
"grad_norm": 3.8300844137879917,
"learning_rate": 1.679717586649551e-05,
"loss": 0.5043,
"step": 500
},
{
"epoch": 0.1636319884495067,
"grad_norm": 3.6133285758576235,
"learning_rate": 1.673299101412067e-05,
"loss": 0.4682,
"step": 510
},
{
"epoch": 0.16684045881126172,
"grad_norm": 6.472583456901733,
"learning_rate": 1.666880616174583e-05,
"loss": 0.4002,
"step": 520
},
{
"epoch": 0.17004892917301676,
"grad_norm": 3.701841978979134,
"learning_rate": 1.660462130937099e-05,
"loss": 0.5022,
"step": 530
},
{
"epoch": 0.1732573995347718,
"grad_norm": 4.773360444080344,
"learning_rate": 1.654043645699615e-05,
"loss": 0.5066,
"step": 540
},
{
"epoch": 0.17646586989652682,
"grad_norm": 13.276543301213977,
"learning_rate": 1.6476251604621313e-05,
"loss": 0.4648,
"step": 550
},
{
"epoch": 0.17967434025828186,
"grad_norm": 8.185811344239216,
"learning_rate": 1.6412066752246473e-05,
"loss": 0.4597,
"step": 560
},
{
"epoch": 0.1828828106200369,
"grad_norm": 3.870551639158725,
"learning_rate": 1.6347881899871633e-05,
"loss": 0.4644,
"step": 570
},
{
"epoch": 0.18609128098179192,
"grad_norm": 2.9435165160063628,
"learning_rate": 1.6283697047496793e-05,
"loss": 0.4996,
"step": 580
},
{
"epoch": 0.18929975134354696,
"grad_norm": 5.242572609854114,
"learning_rate": 1.6219512195121953e-05,
"loss": 0.4465,
"step": 590
},
{
"epoch": 0.192508221705302,
"grad_norm": 3.296578668491081,
"learning_rate": 1.6155327342747113e-05,
"loss": 0.404,
"step": 600
},
{
"epoch": 0.19571669206705702,
"grad_norm": 5.090117348527818,
"learning_rate": 1.6091142490372273e-05,
"loss": 0.4277,
"step": 610
},
{
"epoch": 0.19892516242881206,
"grad_norm": 2.888476032144689,
"learning_rate": 1.6026957637997433e-05,
"loss": 0.4783,
"step": 620
},
{
"epoch": 0.2021336327905671,
"grad_norm": 4.135431165940949,
"learning_rate": 1.5962772785622594e-05,
"loss": 0.4236,
"step": 630
},
{
"epoch": 0.20534210315232213,
"grad_norm": 4.833487869831366,
"learning_rate": 1.5898587933247757e-05,
"loss": 0.504,
"step": 640
},
{
"epoch": 0.20855057351407716,
"grad_norm": 4.282352689606544,
"learning_rate": 1.5834403080872917e-05,
"loss": 0.4277,
"step": 650
},
{
"epoch": 0.2117590438758322,
"grad_norm": 6.760216642975651,
"learning_rate": 1.5770218228498074e-05,
"loss": 0.4916,
"step": 660
},
{
"epoch": 0.21496751423758723,
"grad_norm": 4.349450853536778,
"learning_rate": 1.5706033376123234e-05,
"loss": 0.4863,
"step": 670
},
{
"epoch": 0.21817598459934226,
"grad_norm": 8.907091580804641,
"learning_rate": 1.5641848523748397e-05,
"loss": 0.4454,
"step": 680
},
{
"epoch": 0.2213844549610973,
"grad_norm": 12.495791509691397,
"learning_rate": 1.5577663671373558e-05,
"loss": 0.4374,
"step": 690
},
{
"epoch": 0.22459292532285233,
"grad_norm": 4.162195469033971,
"learning_rate": 1.5513478818998718e-05,
"loss": 0.398,
"step": 700
},
{
"epoch": 0.22780139568460736,
"grad_norm": 4.9440877367494585,
"learning_rate": 1.5449293966623878e-05,
"loss": 0.4386,
"step": 710
},
{
"epoch": 0.2310098660463624,
"grad_norm": 3.3193418621579194,
"learning_rate": 1.5385109114249038e-05,
"loss": 0.3993,
"step": 720
},
{
"epoch": 0.23421833640811743,
"grad_norm": 3.388886050375809,
"learning_rate": 1.5320924261874198e-05,
"loss": 0.3915,
"step": 730
},
{
"epoch": 0.23742680676987246,
"grad_norm": 3.758493220998932,
"learning_rate": 1.525673940949936e-05,
"loss": 0.4452,
"step": 740
},
{
"epoch": 0.2406352771316275,
"grad_norm": 5.132628959411491,
"learning_rate": 1.519255455712452e-05,
"loss": 0.3703,
"step": 750
},
{
"epoch": 0.24384374749338253,
"grad_norm": 6.2691579411030025,
"learning_rate": 1.5128369704749682e-05,
"loss": 0.4158,
"step": 760
},
{
"epoch": 0.24705221785513756,
"grad_norm": 4.103316818662405,
"learning_rate": 1.5064184852374842e-05,
"loss": 0.4233,
"step": 770
},
{
"epoch": 0.2502606882168926,
"grad_norm": 6.0422539237827975,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.4278,
"step": 780
},
{
"epoch": 0.2534691585786476,
"grad_norm": 2.7653877942568954,
"learning_rate": 1.493581514762516e-05,
"loss": 0.4095,
"step": 790
},
{
"epoch": 0.25667762894040264,
"grad_norm": 4.114902146610447,
"learning_rate": 1.487163029525032e-05,
"loss": 0.4584,
"step": 800
},
{
"epoch": 0.25988609930215767,
"grad_norm": 5.28663521958676,
"learning_rate": 1.4807445442875482e-05,
"loss": 0.4368,
"step": 810
},
{
"epoch": 0.2630945696639127,
"grad_norm": 3.8311121578946303,
"learning_rate": 1.4743260590500643e-05,
"loss": 0.4107,
"step": 820
},
{
"epoch": 0.26630304002566774,
"grad_norm": 3.318670161966376,
"learning_rate": 1.4679075738125803e-05,
"loss": 0.4245,
"step": 830
},
{
"epoch": 0.26951151038742277,
"grad_norm": 4.755364543588735,
"learning_rate": 1.4614890885750963e-05,
"loss": 0.4787,
"step": 840
},
{
"epoch": 0.2727199807491778,
"grad_norm": 2.786479482223897,
"learning_rate": 1.4550706033376125e-05,
"loss": 0.4336,
"step": 850
},
{
"epoch": 0.27592845111093284,
"grad_norm": 4.070784873347869,
"learning_rate": 1.4486521181001285e-05,
"loss": 0.4123,
"step": 860
},
{
"epoch": 0.27913692147268787,
"grad_norm": 5.342176013467221,
"learning_rate": 1.4422336328626445e-05,
"loss": 0.4497,
"step": 870
},
{
"epoch": 0.2823453918344429,
"grad_norm": 2.881520241461501,
"learning_rate": 1.4358151476251605e-05,
"loss": 0.4872,
"step": 880
},
{
"epoch": 0.28555386219619794,
"grad_norm": 4.478805078448799,
"learning_rate": 1.4293966623876767e-05,
"loss": 0.3984,
"step": 890
},
{
"epoch": 0.28876233255795297,
"grad_norm": 6.101899594897867,
"learning_rate": 1.4229781771501927e-05,
"loss": 0.4628,
"step": 900
},
{
"epoch": 0.291970802919708,
"grad_norm": 3.0826453501277715,
"learning_rate": 1.4165596919127087e-05,
"loss": 0.4432,
"step": 910
},
{
"epoch": 0.29517927328146304,
"grad_norm": 2.4770273632891953,
"learning_rate": 1.4101412066752247e-05,
"loss": 0.506,
"step": 920
},
{
"epoch": 0.2983877436432181,
"grad_norm": 3.422572341295809,
"learning_rate": 1.4037227214377409e-05,
"loss": 0.3826,
"step": 930
},
{
"epoch": 0.3015962140049731,
"grad_norm": 6.410925492330843,
"learning_rate": 1.3973042362002569e-05,
"loss": 0.4976,
"step": 940
},
{
"epoch": 0.30480468436672814,
"grad_norm": 3.702470810813249,
"learning_rate": 1.390885750962773e-05,
"loss": 0.3977,
"step": 950
},
{
"epoch": 0.3080131547284832,
"grad_norm": 4.496582483074578,
"learning_rate": 1.384467265725289e-05,
"loss": 0.3427,
"step": 960
},
{
"epoch": 0.3112216250902382,
"grad_norm": 5.771712327672755,
"learning_rate": 1.378048780487805e-05,
"loss": 0.4011,
"step": 970
},
{
"epoch": 0.31443009545199324,
"grad_norm": 6.116944263448949,
"learning_rate": 1.3716302952503211e-05,
"loss": 0.3832,
"step": 980
},
{
"epoch": 0.3176385658137483,
"grad_norm": 5.217590349147242,
"learning_rate": 1.3652118100128371e-05,
"loss": 0.4484,
"step": 990
},
{
"epoch": 0.3208470361755033,
"grad_norm": 6.326859919098317,
"learning_rate": 1.3587933247753531e-05,
"loss": 0.4238,
"step": 1000
},
{
"epoch": 0.32405550653725834,
"grad_norm": 3.6100683655298518,
"learning_rate": 1.3523748395378692e-05,
"loss": 0.4521,
"step": 1010
},
{
"epoch": 0.3272639768990134,
"grad_norm": 8.35582852568852,
"learning_rate": 1.3459563543003853e-05,
"loss": 0.4277,
"step": 1020
},
{
"epoch": 0.3304724472607684,
"grad_norm": 6.127881803928465,
"learning_rate": 1.3395378690629014e-05,
"loss": 0.3845,
"step": 1030
},
{
"epoch": 0.33368091762252344,
"grad_norm": 5.346412016792554,
"learning_rate": 1.3331193838254172e-05,
"loss": 0.3598,
"step": 1040
},
{
"epoch": 0.3368893879842785,
"grad_norm": 5.518712653284187,
"learning_rate": 1.3267008985879332e-05,
"loss": 0.4684,
"step": 1050
},
{
"epoch": 0.3400978583460335,
"grad_norm": 3.0171159430683647,
"learning_rate": 1.3202824133504496e-05,
"loss": 0.402,
"step": 1060
},
{
"epoch": 0.34330632870778854,
"grad_norm": 3.4525267211589585,
"learning_rate": 1.3138639281129654e-05,
"loss": 0.4587,
"step": 1070
},
{
"epoch": 0.3465147990695436,
"grad_norm": 4.923842642383016,
"learning_rate": 1.3074454428754814e-05,
"loss": 0.4231,
"step": 1080
},
{
"epoch": 0.3497232694312986,
"grad_norm": 3.982906316656253,
"learning_rate": 1.3010269576379974e-05,
"loss": 0.419,
"step": 1090
},
{
"epoch": 0.35293173979305364,
"grad_norm": 3.9801555228114998,
"learning_rate": 1.2946084724005136e-05,
"loss": 0.3668,
"step": 1100
},
{
"epoch": 0.3561402101548087,
"grad_norm": 4.030778915107762,
"learning_rate": 1.2881899871630296e-05,
"loss": 0.3739,
"step": 1110
},
{
"epoch": 0.3593486805165637,
"grad_norm": 6.386525169017585,
"learning_rate": 1.2817715019255456e-05,
"loss": 0.4972,
"step": 1120
},
{
"epoch": 0.36255715087831875,
"grad_norm": 4.20862662435033,
"learning_rate": 1.2753530166880616e-05,
"loss": 0.4154,
"step": 1130
},
{
"epoch": 0.3657656212400738,
"grad_norm": 4.858924338028923,
"learning_rate": 1.2689345314505778e-05,
"loss": 0.4478,
"step": 1140
},
{
"epoch": 0.3689740916018288,
"grad_norm": 5.608430455111102,
"learning_rate": 1.2625160462130938e-05,
"loss": 0.4599,
"step": 1150
},
{
"epoch": 0.37218256196358385,
"grad_norm": 5.911587578207529,
"learning_rate": 1.2560975609756098e-05,
"loss": 0.3869,
"step": 1160
},
{
"epoch": 0.3753910323253389,
"grad_norm": 4.283524740830092,
"learning_rate": 1.2496790757381259e-05,
"loss": 0.3837,
"step": 1170
},
{
"epoch": 0.3785995026870939,
"grad_norm": 6.821302089832709,
"learning_rate": 1.2432605905006419e-05,
"loss": 0.442,
"step": 1180
},
{
"epoch": 0.38180797304884895,
"grad_norm": 5.085358904961751,
"learning_rate": 1.236842105263158e-05,
"loss": 0.3264,
"step": 1190
},
{
"epoch": 0.385016443410604,
"grad_norm": 4.605248058627308,
"learning_rate": 1.230423620025674e-05,
"loss": 0.3895,
"step": 1200
},
{
"epoch": 0.388224913772359,
"grad_norm": 3.477144602759826,
"learning_rate": 1.22400513478819e-05,
"loss": 0.4405,
"step": 1210
},
{
"epoch": 0.39143338413411405,
"grad_norm": 4.706693999402434,
"learning_rate": 1.2175866495507061e-05,
"loss": 0.4102,
"step": 1220
},
{
"epoch": 0.3946418544958691,
"grad_norm": 3.204192427462165,
"learning_rate": 1.2111681643132223e-05,
"loss": 0.4244,
"step": 1230
},
{
"epoch": 0.3978503248576241,
"grad_norm": 5.415400158823864,
"learning_rate": 1.2047496790757383e-05,
"loss": 0.389,
"step": 1240
},
{
"epoch": 0.40105879521937915,
"grad_norm": 4.1016299695920555,
"learning_rate": 1.1983311938382543e-05,
"loss": 0.4348,
"step": 1250
},
{
"epoch": 0.4042672655811342,
"grad_norm": 5.0119293140447265,
"learning_rate": 1.1919127086007703e-05,
"loss": 0.4076,
"step": 1260
},
{
"epoch": 0.4074757359428892,
"grad_norm": 3.1843597604004144,
"learning_rate": 1.1854942233632865e-05,
"loss": 0.4093,
"step": 1270
},
{
"epoch": 0.41068420630464425,
"grad_norm": 3.243310369644707,
"learning_rate": 1.1790757381258025e-05,
"loss": 0.3996,
"step": 1280
},
{
"epoch": 0.4138926766663993,
"grad_norm": 6.5805185922112654,
"learning_rate": 1.1726572528883185e-05,
"loss": 0.4443,
"step": 1290
},
{
"epoch": 0.4171011470281543,
"grad_norm": 3.5901965588799833,
"learning_rate": 1.1662387676508344e-05,
"loss": 0.4222,
"step": 1300
},
{
"epoch": 0.42030961738990935,
"grad_norm": 3.5975503587085513,
"learning_rate": 1.1598202824133507e-05,
"loss": 0.3803,
"step": 1310
},
{
"epoch": 0.4235180877516644,
"grad_norm": 3.8370137218295612,
"learning_rate": 1.1534017971758667e-05,
"loss": 0.377,
"step": 1320
},
{
"epoch": 0.4267265581134194,
"grad_norm": 6.055162453845242,
"learning_rate": 1.1469833119383826e-05,
"loss": 0.4371,
"step": 1330
},
{
"epoch": 0.42993502847517445,
"grad_norm": 4.303475483832273,
"learning_rate": 1.1405648267008986e-05,
"loss": 0.3991,
"step": 1340
},
{
"epoch": 0.4331434988369295,
"grad_norm": 3.481300920221893,
"learning_rate": 1.1341463414634146e-05,
"loss": 0.408,
"step": 1350
},
{
"epoch": 0.4363519691986845,
"grad_norm": 2.70794769803311,
"learning_rate": 1.1277278562259308e-05,
"loss": 0.4105,
"step": 1360
},
{
"epoch": 0.43956043956043955,
"grad_norm": 4.540637646839311,
"learning_rate": 1.1213093709884468e-05,
"loss": 0.3762,
"step": 1370
},
{
"epoch": 0.4427689099221946,
"grad_norm": 3.240923093729635,
"learning_rate": 1.1148908857509628e-05,
"loss": 0.3198,
"step": 1380
},
{
"epoch": 0.4459773802839496,
"grad_norm": 4.079516203900489,
"learning_rate": 1.1084724005134788e-05,
"loss": 0.4255,
"step": 1390
},
{
"epoch": 0.44918585064570465,
"grad_norm": 4.424913833083467,
"learning_rate": 1.102053915275995e-05,
"loss": 0.422,
"step": 1400
},
{
"epoch": 0.4523943210074597,
"grad_norm": 4.215968513158678,
"learning_rate": 1.095635430038511e-05,
"loss": 0.3717,
"step": 1410
},
{
"epoch": 0.4556027913692147,
"grad_norm": 7.080680051130726,
"learning_rate": 1.089216944801027e-05,
"loss": 0.4167,
"step": 1420
},
{
"epoch": 0.45881126173096975,
"grad_norm": 54.87405885391659,
"learning_rate": 1.082798459563543e-05,
"loss": 0.3701,
"step": 1430
},
{
"epoch": 0.4620197320927248,
"grad_norm": 4.745370324575341,
"learning_rate": 1.0763799743260592e-05,
"loss": 0.4192,
"step": 1440
},
{
"epoch": 0.4652282024544798,
"grad_norm": 4.072739876459792,
"learning_rate": 1.0699614890885752e-05,
"loss": 0.3648,
"step": 1450
},
{
"epoch": 0.46843667281623486,
"grad_norm": 5.361505931287111,
"learning_rate": 1.0635430038510912e-05,
"loss": 0.3908,
"step": 1460
},
{
"epoch": 0.4716451431779899,
"grad_norm": 6.609039672159613,
"learning_rate": 1.0571245186136072e-05,
"loss": 0.3927,
"step": 1470
},
{
"epoch": 0.4748536135397449,
"grad_norm": 5.257109634802722,
"learning_rate": 1.0507060333761234e-05,
"loss": 0.3901,
"step": 1480
},
{
"epoch": 0.47806208390149996,
"grad_norm": 2.959492453957149,
"learning_rate": 1.0442875481386394e-05,
"loss": 0.3616,
"step": 1490
},
{
"epoch": 0.481270554263255,
"grad_norm": 4.53017762648938,
"learning_rate": 1.0378690629011554e-05,
"loss": 0.4099,
"step": 1500
},
{
"epoch": 0.48447902462501,
"grad_norm": 5.632280418736274,
"learning_rate": 1.0314505776636715e-05,
"loss": 0.4176,
"step": 1510
},
{
"epoch": 0.48768749498676506,
"grad_norm": 3.4384598745003627,
"learning_rate": 1.0250320924261875e-05,
"loss": 0.3586,
"step": 1520
},
{
"epoch": 0.4908959653485201,
"grad_norm": 3.024437029724823,
"learning_rate": 1.0186136071887036e-05,
"loss": 0.4255,
"step": 1530
},
{
"epoch": 0.4941044357102751,
"grad_norm": 2.378291269172342,
"learning_rate": 1.0121951219512197e-05,
"loss": 0.3958,
"step": 1540
},
{
"epoch": 0.49731290607203016,
"grad_norm": 4.62474422398616,
"learning_rate": 1.0057766367137357e-05,
"loss": 0.4006,
"step": 1550
},
{
"epoch": 0.5005213764337852,
"grad_norm": 3.6745219934625144,
"learning_rate": 9.993581514762517e-06,
"loss": 0.3682,
"step": 1560
},
{
"epoch": 0.5037298467955402,
"grad_norm": 4.448848307155705,
"learning_rate": 9.929396662387677e-06,
"loss": 0.4272,
"step": 1570
},
{
"epoch": 0.5069383171572952,
"grad_norm": 3.8894278011874235,
"learning_rate": 9.865211810012839e-06,
"loss": 0.3569,
"step": 1580
},
{
"epoch": 0.5101467875190503,
"grad_norm": 4.774303748597002,
"learning_rate": 9.801026957637997e-06,
"loss": 0.3933,
"step": 1590
},
{
"epoch": 0.5133552578808053,
"grad_norm": 5.045270496853991,
"learning_rate": 9.736842105263159e-06,
"loss": 0.3719,
"step": 1600
},
{
"epoch": 0.5165637282425604,
"grad_norm": 4.740584901110403,
"learning_rate": 9.672657252888319e-06,
"loss": 0.4254,
"step": 1610
},
{
"epoch": 0.5197721986043153,
"grad_norm": 5.884473992854898,
"learning_rate": 9.60847240051348e-06,
"loss": 0.4251,
"step": 1620
},
{
"epoch": 0.5229806689660704,
"grad_norm": 3.4560812191487233,
"learning_rate": 9.54428754813864e-06,
"loss": 0.3884,
"step": 1630
},
{
"epoch": 0.5261891393278254,
"grad_norm": 4.436488938480811,
"learning_rate": 9.480102695763801e-06,
"loss": 0.3791,
"step": 1640
},
{
"epoch": 0.5293976096895805,
"grad_norm": 5.122741798151655,
"learning_rate": 9.415917843388961e-06,
"loss": 0.4109,
"step": 1650
},
{
"epoch": 0.5326060800513355,
"grad_norm": 2.8268137365317694,
"learning_rate": 9.351732991014121e-06,
"loss": 0.4002,
"step": 1660
},
{
"epoch": 0.5358145504130906,
"grad_norm": 4.961028503814302,
"learning_rate": 9.287548138639282e-06,
"loss": 0.4117,
"step": 1670
},
{
"epoch": 0.5390230207748455,
"grad_norm": 5.913940405056516,
"learning_rate": 9.223363286264443e-06,
"loss": 0.3837,
"step": 1680
},
{
"epoch": 0.5422314911366006,
"grad_norm": 4.981120773191797,
"learning_rate": 9.159178433889603e-06,
"loss": 0.4258,
"step": 1690
},
{
"epoch": 0.5454399614983556,
"grad_norm": 5.444655853775516,
"learning_rate": 9.094993581514764e-06,
"loss": 0.4301,
"step": 1700
},
{
"epoch": 0.5486484318601107,
"grad_norm": 3.848013181214462,
"learning_rate": 9.030808729139924e-06,
"loss": 0.4004,
"step": 1710
},
{
"epoch": 0.5518569022218657,
"grad_norm": 2.7256377363961692,
"learning_rate": 8.966623876765084e-06,
"loss": 0.4199,
"step": 1720
},
{
"epoch": 0.5550653725836208,
"grad_norm": 3.9591345977979278,
"learning_rate": 8.902439024390244e-06,
"loss": 0.3923,
"step": 1730
},
{
"epoch": 0.5582738429453757,
"grad_norm": 3.694097683677857,
"learning_rate": 8.838254172015404e-06,
"loss": 0.4245,
"step": 1740
},
{
"epoch": 0.5614823133071308,
"grad_norm": 4.752754487463968,
"learning_rate": 8.774069319640566e-06,
"loss": 0.3623,
"step": 1750
},
{
"epoch": 0.5646907836688858,
"grad_norm": 4.933398001346964,
"learning_rate": 8.709884467265726e-06,
"loss": 0.3529,
"step": 1760
},
{
"epoch": 0.5678992540306409,
"grad_norm": 4.10267822290099,
"learning_rate": 8.645699614890886e-06,
"loss": 0.3903,
"step": 1770
},
{
"epoch": 0.5711077243923959,
"grad_norm": 5.948577818513,
"learning_rate": 8.581514762516046e-06,
"loss": 0.3923,
"step": 1780
},
{
"epoch": 0.574316194754151,
"grad_norm": 4.32270415025434,
"learning_rate": 8.517329910141208e-06,
"loss": 0.4162,
"step": 1790
},
{
"epoch": 0.5775246651159059,
"grad_norm": 4.186201731866516,
"learning_rate": 8.453145057766368e-06,
"loss": 0.4006,
"step": 1800
},
{
"epoch": 0.580733135477661,
"grad_norm": 4.818472247869851,
"learning_rate": 8.388960205391528e-06,
"loss": 0.3841,
"step": 1810
},
{
"epoch": 0.583941605839416,
"grad_norm": 3.722377675945526,
"learning_rate": 8.324775353016688e-06,
"loss": 0.4015,
"step": 1820
},
{
"epoch": 0.5871500762011711,
"grad_norm": 3.581092339067193,
"learning_rate": 8.26059050064185e-06,
"loss": 0.4194,
"step": 1830
},
{
"epoch": 0.5903585465629261,
"grad_norm": 3.8711833078388134,
"learning_rate": 8.19640564826701e-06,
"loss": 0.5039,
"step": 1840
},
{
"epoch": 0.5935670169246812,
"grad_norm": 5.312406861505809,
"learning_rate": 8.13222079589217e-06,
"loss": 0.3954,
"step": 1850
},
{
"epoch": 0.5967754872864361,
"grad_norm": 5.128909473296003,
"learning_rate": 8.06803594351733e-06,
"loss": 0.3476,
"step": 1860
},
{
"epoch": 0.5999839576481912,
"grad_norm": 6.225968269360909,
"learning_rate": 8.00385109114249e-06,
"loss": 0.3918,
"step": 1870
},
{
"epoch": 0.6031924280099462,
"grad_norm": 4.660989653248528,
"learning_rate": 7.93966623876765e-06,
"loss": 0.4013,
"step": 1880
},
{
"epoch": 0.6064008983717013,
"grad_norm": 4.299519878675749,
"learning_rate": 7.875481386392811e-06,
"loss": 0.3775,
"step": 1890
},
{
"epoch": 0.6096093687334563,
"grad_norm": 4.680629637314835,
"learning_rate": 7.811296534017973e-06,
"loss": 0.3612,
"step": 1900
},
{
"epoch": 0.6128178390952114,
"grad_norm": 4.152954271805758,
"learning_rate": 7.747111681643133e-06,
"loss": 0.3806,
"step": 1910
},
{
"epoch": 0.6160263094569663,
"grad_norm": 4.6801179801552415,
"learning_rate": 7.682926829268293e-06,
"loss": 0.3506,
"step": 1920
},
{
"epoch": 0.6192347798187214,
"grad_norm": 4.251504275426107,
"learning_rate": 7.618741976893453e-06,
"loss": 0.3589,
"step": 1930
},
{
"epoch": 0.6224432501804764,
"grad_norm": 6.009730313403801,
"learning_rate": 7.554557124518614e-06,
"loss": 0.3968,
"step": 1940
},
{
"epoch": 0.6256517205422315,
"grad_norm": 3.2723481829787464,
"learning_rate": 7.490372272143774e-06,
"loss": 0.3936,
"step": 1950
},
{
"epoch": 0.6288601909039865,
"grad_norm": 4.097990240581078,
"learning_rate": 7.426187419768935e-06,
"loss": 0.4144,
"step": 1960
},
{
"epoch": 0.6320686612657416,
"grad_norm": 2.7191742984555995,
"learning_rate": 7.362002567394095e-06,
"loss": 0.409,
"step": 1970
},
{
"epoch": 0.6352771316274965,
"grad_norm": 4.40050340710905,
"learning_rate": 7.297817715019256e-06,
"loss": 0.3526,
"step": 1980
},
{
"epoch": 0.6384856019892516,
"grad_norm": 3.5711401894801194,
"learning_rate": 7.233632862644416e-06,
"loss": 0.3865,
"step": 1990
},
{
"epoch": 0.6416940723510066,
"grad_norm": 5.155189378251862,
"learning_rate": 7.169448010269577e-06,
"loss": 0.4037,
"step": 2000
},
{
"epoch": 0.6449025427127617,
"grad_norm": 5.764606055822457,
"learning_rate": 7.1052631578947375e-06,
"loss": 0.3806,
"step": 2010
},
{
"epoch": 0.6481110130745167,
"grad_norm": 3.3930437034797585,
"learning_rate": 7.0410783055198984e-06,
"loss": 0.3951,
"step": 2020
},
{
"epoch": 0.6513194834362718,
"grad_norm": 4.922727726376936,
"learning_rate": 6.9768934531450586e-06,
"loss": 0.3847,
"step": 2030
},
{
"epoch": 0.6545279537980268,
"grad_norm": 2.7890174522963367,
"learning_rate": 6.9127086007702195e-06,
"loss": 0.3479,
"step": 2040
},
{
"epoch": 0.6577364241597818,
"grad_norm": 4.412823863289807,
"learning_rate": 6.84852374839538e-06,
"loss": 0.4173,
"step": 2050
},
{
"epoch": 0.6609448945215368,
"grad_norm": 5.417820507568683,
"learning_rate": 6.784338896020539e-06,
"loss": 0.381,
"step": 2060
},
{
"epoch": 0.6641533648832919,
"grad_norm": 3.8506255680933528,
"learning_rate": 6.7201540436457e-06,
"loss": 0.3693,
"step": 2070
},
{
"epoch": 0.6673618352450469,
"grad_norm": 3.5416828656888786,
"learning_rate": 6.65596919127086e-06,
"loss": 0.358,
"step": 2080
},
{
"epoch": 0.670570305606802,
"grad_norm": 3.5323170162835775,
"learning_rate": 6.591784338896021e-06,
"loss": 0.3237,
"step": 2090
},
{
"epoch": 0.673778775968557,
"grad_norm": 6.128726604158042,
"learning_rate": 6.527599486521181e-06,
"loss": 0.3859,
"step": 2100
},
{
"epoch": 0.676987246330312,
"grad_norm": 4.891887980181221,
"learning_rate": 6.463414634146342e-06,
"loss": 0.3223,
"step": 2110
},
{
"epoch": 0.680195716692067,
"grad_norm": 3.565557213786112,
"learning_rate": 6.399229781771502e-06,
"loss": 0.363,
"step": 2120
},
{
"epoch": 0.6834041870538221,
"grad_norm": 3.4294258572573337,
"learning_rate": 6.335044929396663e-06,
"loss": 0.3781,
"step": 2130
},
{
"epoch": 0.6866126574155771,
"grad_norm": 4.518476500237087,
"learning_rate": 6.270860077021823e-06,
"loss": 0.4325,
"step": 2140
},
{
"epoch": 0.6898211277773322,
"grad_norm": 4.743599740115472,
"learning_rate": 6.206675224646984e-06,
"loss": 0.411,
"step": 2150
},
{
"epoch": 0.6930295981390872,
"grad_norm": 2.5630593956213015,
"learning_rate": 6.142490372272144e-06,
"loss": 0.3717,
"step": 2160
},
{
"epoch": 0.6962380685008422,
"grad_norm": 6.2586555195558,
"learning_rate": 6.078305519897305e-06,
"loss": 0.3561,
"step": 2170
},
{
"epoch": 0.6994465388625972,
"grad_norm": 4.894715427924892,
"learning_rate": 6.0141206675224654e-06,
"loss": 0.4164,
"step": 2180
},
{
"epoch": 0.7026550092243523,
"grad_norm": 3.5252148379990573,
"learning_rate": 5.949935815147626e-06,
"loss": 0.3209,
"step": 2190
},
{
"epoch": 0.7058634795861073,
"grad_norm": 3.256470353538883,
"learning_rate": 5.885750962772786e-06,
"loss": 0.3893,
"step": 2200
},
{
"epoch": 0.7090719499478624,
"grad_norm": 3.4671867719647573,
"learning_rate": 5.821566110397947e-06,
"loss": 0.3836,
"step": 2210
},
{
"epoch": 0.7122804203096174,
"grad_norm": 3.6300469654014806,
"learning_rate": 5.757381258023107e-06,
"loss": 0.3604,
"step": 2220
},
{
"epoch": 0.7154888906713724,
"grad_norm": 4.8222353267885545,
"learning_rate": 5.693196405648267e-06,
"loss": 0.3402,
"step": 2230
},
{
"epoch": 0.7186973610331274,
"grad_norm": 3.088736996240411,
"learning_rate": 5.629011553273428e-06,
"loss": 0.3571,
"step": 2240
},
{
"epoch": 0.7219058313948825,
"grad_norm": 6.468377008869226,
"learning_rate": 5.564826700898588e-06,
"loss": 0.3857,
"step": 2250
},
{
"epoch": 0.7251143017566375,
"grad_norm": 4.976006702523603,
"learning_rate": 5.500641848523749e-06,
"loss": 0.328,
"step": 2260
},
{
"epoch": 0.7283227721183926,
"grad_norm": 7.44768864343581,
"learning_rate": 5.436456996148909e-06,
"loss": 0.3415,
"step": 2270
},
{
"epoch": 0.7315312424801476,
"grad_norm": 6.342940977120239,
"learning_rate": 5.37227214377407e-06,
"loss": 0.3452,
"step": 2280
},
{
"epoch": 0.7347397128419026,
"grad_norm": 4.281955030745344,
"learning_rate": 5.30808729139923e-06,
"loss": 0.4647,
"step": 2290
},
{
"epoch": 0.7379481832036576,
"grad_norm": 3.678790025410135,
"learning_rate": 5.243902439024391e-06,
"loss": 0.3742,
"step": 2300
},
{
"epoch": 0.7411566535654127,
"grad_norm": 3.3547281464794096,
"learning_rate": 5.179717586649551e-06,
"loss": 0.3835,
"step": 2310
},
{
"epoch": 0.7443651239271677,
"grad_norm": 5.13222360440861,
"learning_rate": 5.115532734274712e-06,
"loss": 0.3859,
"step": 2320
},
{
"epoch": 0.7475735942889228,
"grad_norm": 3.972710625172556,
"learning_rate": 5.0513478818998715e-06,
"loss": 0.3537,
"step": 2330
},
{
"epoch": 0.7507820646506778,
"grad_norm": 3.2744384458301687,
"learning_rate": 4.9871630295250324e-06,
"loss": 0.3793,
"step": 2340
},
{
"epoch": 0.7539905350124328,
"grad_norm": 3.8532031308587396,
"learning_rate": 4.9229781771501926e-06,
"loss": 0.3245,
"step": 2350
},
{
"epoch": 0.7571990053741878,
"grad_norm": 6.097220741982411,
"learning_rate": 4.8587933247753535e-06,
"loss": 0.4381,
"step": 2360
},
{
"epoch": 0.7604074757359429,
"grad_norm": 3.6161720053467743,
"learning_rate": 4.794608472400514e-06,
"loss": 0.3904,
"step": 2370
},
{
"epoch": 0.7636159460976979,
"grad_norm": 3.2126570713513196,
"learning_rate": 4.730423620025675e-06,
"loss": 0.3698,
"step": 2380
},
{
"epoch": 0.766824416459453,
"grad_norm": 4.440937942880446,
"learning_rate": 4.666238767650835e-06,
"loss": 0.3672,
"step": 2390
},
{
"epoch": 0.770032886821208,
"grad_norm": 4.034010642616632,
"learning_rate": 4.602053915275995e-06,
"loss": 0.3453,
"step": 2400
},
{
"epoch": 0.773241357182963,
"grad_norm": 4.320673481292839,
"learning_rate": 4.537869062901156e-06,
"loss": 0.3874,
"step": 2410
},
{
"epoch": 0.776449827544718,
"grad_norm": 5.721110452311635,
"learning_rate": 4.473684210526316e-06,
"loss": 0.3361,
"step": 2420
},
{
"epoch": 0.7796582979064731,
"grad_norm": 5.553593669289626,
"learning_rate": 4.409499358151477e-06,
"loss": 0.3761,
"step": 2430
},
{
"epoch": 0.7828667682682281,
"grad_norm": 3.5931839661915506,
"learning_rate": 4.345314505776637e-06,
"loss": 0.3303,
"step": 2440
},
{
"epoch": 0.7860752386299832,
"grad_norm": 3.191917593220325,
"learning_rate": 4.281129653401798e-06,
"loss": 0.4581,
"step": 2450
},
{
"epoch": 0.7892837089917382,
"grad_norm": 2.071662511958508,
"learning_rate": 4.216944801026958e-06,
"loss": 0.3602,
"step": 2460
},
{
"epoch": 0.7924921793534933,
"grad_norm": 4.1098396579418015,
"learning_rate": 4.152759948652118e-06,
"loss": 0.3658,
"step": 2470
},
{
"epoch": 0.7957006497152482,
"grad_norm": 3.342343894793557,
"learning_rate": 4.088575096277279e-06,
"loss": 0.3533,
"step": 2480
},
{
"epoch": 0.7989091200770033,
"grad_norm": 3.501852515883717,
"learning_rate": 4.024390243902439e-06,
"loss": 0.3086,
"step": 2490
},
{
"epoch": 0.8021175904387583,
"grad_norm": 5.6740246031343275,
"learning_rate": 3.9602053915276e-06,
"loss": 0.3619,
"step": 2500
},
{
"epoch": 0.8053260608005134,
"grad_norm": 3.6625993713157867,
"learning_rate": 3.89602053915276e-06,
"loss": 0.3536,
"step": 2510
},
{
"epoch": 0.8085345311622684,
"grad_norm": 4.12817611506417,
"learning_rate": 3.8318356867779205e-06,
"loss": 0.3552,
"step": 2520
},
{
"epoch": 0.8117430015240235,
"grad_norm": 6.752665333536269,
"learning_rate": 3.767650834403081e-06,
"loss": 0.3906,
"step": 2530
},
{
"epoch": 0.8149514718857784,
"grad_norm": 4.0698833515045765,
"learning_rate": 3.7034659820282416e-06,
"loss": 0.369,
"step": 2540
},
{
"epoch": 0.8181599422475335,
"grad_norm": 3.8061893249692167,
"learning_rate": 3.639281129653402e-06,
"loss": 0.391,
"step": 2550
},
{
"epoch": 0.8213684126092885,
"grad_norm": 4.2054055524032075,
"learning_rate": 3.5750962772785623e-06,
"loss": 0.3781,
"step": 2560
},
{
"epoch": 0.8245768829710436,
"grad_norm": 3.1509974167473294,
"learning_rate": 3.510911424903723e-06,
"loss": 0.4246,
"step": 2570
},
{
"epoch": 0.8277853533327986,
"grad_norm": 3.344075353049933,
"learning_rate": 3.4467265725288834e-06,
"loss": 0.3839,
"step": 2580
},
{
"epoch": 0.8309938236945537,
"grad_norm": 4.3862748314587625,
"learning_rate": 3.382541720154044e-06,
"loss": 0.3243,
"step": 2590
},
{
"epoch": 0.8342022940563086,
"grad_norm": 4.430571971071977,
"learning_rate": 3.3183568677792044e-06,
"loss": 0.4219,
"step": 2600
},
{
"epoch": 0.8374107644180637,
"grad_norm": 4.216572691360644,
"learning_rate": 3.254172015404365e-06,
"loss": 0.3453,
"step": 2610
},
{
"epoch": 0.8406192347798187,
"grad_norm": 5.730663867597084,
"learning_rate": 3.1899871630295255e-06,
"loss": 0.3652,
"step": 2620
},
{
"epoch": 0.8438277051415738,
"grad_norm": 3.152373466591917,
"learning_rate": 3.125802310654686e-06,
"loss": 0.3761,
"step": 2630
},
{
"epoch": 0.8470361755033288,
"grad_norm": 4.049903812572876,
"learning_rate": 3.061617458279846e-06,
"loss": 0.4274,
"step": 2640
},
{
"epoch": 0.8502446458650839,
"grad_norm": 3.7109620908295002,
"learning_rate": 2.9974326059050067e-06,
"loss": 0.3575,
"step": 2650
},
{
"epoch": 0.8534531162268388,
"grad_norm": 6.151962412156622,
"learning_rate": 2.9332477535301673e-06,
"loss": 0.3935,
"step": 2660
},
{
"epoch": 0.8566615865885939,
"grad_norm": 2.8483008748019474,
"learning_rate": 2.869062901155328e-06,
"loss": 0.3327,
"step": 2670
},
{
"epoch": 0.8598700569503489,
"grad_norm": 3.884150141973327,
"learning_rate": 2.8048780487804884e-06,
"loss": 0.3222,
"step": 2680
},
{
"epoch": 0.863078527312104,
"grad_norm": 5.091883011304281,
"learning_rate": 2.740693196405648e-06,
"loss": 0.4241,
"step": 2690
},
{
"epoch": 0.866286997673859,
"grad_norm": 6.256322006655992,
"learning_rate": 2.6765083440308086e-06,
"loss": 0.3896,
"step": 2700
},
{
"epoch": 0.8694954680356141,
"grad_norm": 3.185423190731702,
"learning_rate": 2.612323491655969e-06,
"loss": 0.3503,
"step": 2710
},
{
"epoch": 0.872703938397369,
"grad_norm": 2.501677043437045,
"learning_rate": 2.5481386392811297e-06,
"loss": 0.308,
"step": 2720
},
{
"epoch": 0.8759124087591241,
"grad_norm": 3.994307293016134,
"learning_rate": 2.4839537869062902e-06,
"loss": 0.3146,
"step": 2730
},
{
"epoch": 0.8791208791208791,
"grad_norm": 6.3218538833338185,
"learning_rate": 2.4197689345314508e-06,
"loss": 0.3861,
"step": 2740
},
{
"epoch": 0.8823293494826342,
"grad_norm": 4.2929895131508875,
"learning_rate": 2.3555840821566113e-06,
"loss": 0.3514,
"step": 2750
},
{
"epoch": 0.8855378198443892,
"grad_norm": 5.013441973517175,
"learning_rate": 2.2913992297817714e-06,
"loss": 0.3394,
"step": 2760
},
{
"epoch": 0.8887462902061443,
"grad_norm": 4.942588457664286,
"learning_rate": 2.227214377406932e-06,
"loss": 0.3916,
"step": 2770
},
{
"epoch": 0.8919547605678992,
"grad_norm": 4.093044885114057,
"learning_rate": 2.1630295250320925e-06,
"loss": 0.3316,
"step": 2780
},
{
"epoch": 0.8951632309296543,
"grad_norm": 3.6241454294578044,
"learning_rate": 2.098844672657253e-06,
"loss": 0.328,
"step": 2790
},
{
"epoch": 0.8983717012914093,
"grad_norm": 3.7585099242216105,
"learning_rate": 2.0346598202824136e-06,
"loss": 0.313,
"step": 2800
},
{
"epoch": 0.9015801716531644,
"grad_norm": 3.092191111039683,
"learning_rate": 1.970474967907574e-06,
"loss": 0.3207,
"step": 2810
},
{
"epoch": 0.9047886420149194,
"grad_norm": 7.126196821569378,
"learning_rate": 1.9062901155327343e-06,
"loss": 0.339,
"step": 2820
},
{
"epoch": 0.9079971123766745,
"grad_norm": 3.98517835436473,
"learning_rate": 1.8421052631578948e-06,
"loss": 0.3119,
"step": 2830
},
{
"epoch": 0.9112055827384294,
"grad_norm": 5.155996828725616,
"learning_rate": 1.7779204107830554e-06,
"loss": 0.3524,
"step": 2840
},
{
"epoch": 0.9144140531001845,
"grad_norm": 6.2418873185093116,
"learning_rate": 1.7137355584082157e-06,
"loss": 0.3697,
"step": 2850
},
{
"epoch": 0.9176225234619395,
"grad_norm": 4.93642870905445,
"learning_rate": 1.6495507060333762e-06,
"loss": 0.3567,
"step": 2860
},
{
"epoch": 0.9208309938236946,
"grad_norm": 3.4661449590899815,
"learning_rate": 1.5853658536585368e-06,
"loss": 0.3049,
"step": 2870
},
{
"epoch": 0.9240394641854496,
"grad_norm": 3.9858026943112046,
"learning_rate": 1.5211810012836973e-06,
"loss": 0.342,
"step": 2880
},
{
"epoch": 0.9272479345472047,
"grad_norm": 5.747224506777868,
"learning_rate": 1.4569961489088577e-06,
"loss": 0.3786,
"step": 2890
},
{
"epoch": 0.9304564049089596,
"grad_norm": 4.2756510193950135,
"learning_rate": 1.3928112965340182e-06,
"loss": 0.345,
"step": 2900
},
{
"epoch": 0.9336648752707147,
"grad_norm": 5.737301674162469,
"learning_rate": 1.3286264441591783e-06,
"loss": 0.3048,
"step": 2910
},
{
"epoch": 0.9368733456324697,
"grad_norm": 4.5368921290576125,
"learning_rate": 1.2644415917843389e-06,
"loss": 0.3373,
"step": 2920
},
{
"epoch": 0.9400818159942248,
"grad_norm": 6.281727114562743,
"learning_rate": 1.2002567394094996e-06,
"loss": 0.3713,
"step": 2930
},
{
"epoch": 0.9432902863559798,
"grad_norm": 4.613863643534097,
"learning_rate": 1.13607188703466e-06,
"loss": 0.3115,
"step": 2940
},
{
"epoch": 0.9464987567177349,
"grad_norm": 4.4955128133503415,
"learning_rate": 1.0718870346598203e-06,
"loss": 0.3544,
"step": 2950
},
{
"epoch": 0.9497072270794898,
"grad_norm": 6.144202111803078,
"learning_rate": 1.0077021822849808e-06,
"loss": 0.3419,
"step": 2960
},
{
"epoch": 0.9529156974412449,
"grad_norm": 3.416734695378003,
"learning_rate": 9.435173299101414e-07,
"loss": 0.3866,
"step": 2970
},
{
"epoch": 0.9561241678029999,
"grad_norm": 2.9130051363163507,
"learning_rate": 8.793324775353017e-07,
"loss": 0.3294,
"step": 2980
},
{
"epoch": 0.959332638164755,
"grad_norm": 4.1544972260670345,
"learning_rate": 8.151476251604621e-07,
"loss": 0.337,
"step": 2990
},
{
"epoch": 0.96254110852651,
"grad_norm": 5.333751729311107,
"learning_rate": 7.509627727856227e-07,
"loss": 0.3707,
"step": 3000
},
{
"epoch": 0.9657495788882651,
"grad_norm": 4.224378334244017,
"learning_rate": 6.867779204107831e-07,
"loss": 0.3266,
"step": 3010
},
{
"epoch": 0.96895804925002,
"grad_norm": 4.68085662929714,
"learning_rate": 6.225930680359436e-07,
"loss": 0.3228,
"step": 3020
},
{
"epoch": 0.9721665196117751,
"grad_norm": 2.44350959307893,
"learning_rate": 5.584082156611041e-07,
"loss": 0.3038,
"step": 3030
},
{
"epoch": 0.9753749899735301,
"grad_norm": 5.5263577593599225,
"learning_rate": 4.942233632862644e-07,
"loss": 0.3073,
"step": 3040
},
{
"epoch": 0.9785834603352852,
"grad_norm": 5.517219216862406,
"learning_rate": 4.30038510911425e-07,
"loss": 0.4016,
"step": 3050
},
{
"epoch": 0.9817919306970402,
"grad_norm": 3.559703237499543,
"learning_rate": 3.6585365853658536e-07,
"loss": 0.4128,
"step": 3060
},
{
"epoch": 0.9850004010587953,
"grad_norm": 3.2825888661809057,
"learning_rate": 3.0166880616174585e-07,
"loss": 0.2873,
"step": 3070
},
{
"epoch": 0.9882088714205502,
"grad_norm": 3.755613548892917,
"learning_rate": 2.374839537869063e-07,
"loss": 0.3854,
"step": 3080
},
{
"epoch": 0.9914173417823053,
"grad_norm": 4.172412303358024,
"learning_rate": 1.7329910141206678e-07,
"loss": 0.328,
"step": 3090
},
{
"epoch": 0.9946258121440603,
"grad_norm": 6.547343932529505,
"learning_rate": 1.0911424903722721e-07,
"loss": 0.3947,
"step": 3100
},
{
"epoch": 0.9978342825058154,
"grad_norm": 4.624930868093504,
"learning_rate": 4.4929396662387685e-08,
"loss": 0.3312,
"step": 3110
}
],
"logging_steps": 10,
"max_steps": 3116,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}