Llama-3.2-3B-SFT-RM-bi / trainer_state.json
rshwndsz's picture
Upload folder using huggingface_hub
4a8b3c8 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": 28.962624668372126,
"learning_rate": 1.9942233632862646e-05,
"loss": 0.7127,
"step": 10
},
{
"epoch": 0.006416940723510066,
"grad_norm": 40.83190702360499,
"learning_rate": 1.9878048780487806e-05,
"loss": 0.6527,
"step": 20
},
{
"epoch": 0.0096254110852651,
"grad_norm": 3.5904467097134516,
"learning_rate": 1.9813863928112966e-05,
"loss": 0.7869,
"step": 30
},
{
"epoch": 0.012833881447020133,
"grad_norm": 4.538031215636058,
"learning_rate": 1.974967907573813e-05,
"loss": 0.6023,
"step": 40
},
{
"epoch": 0.016042351808775165,
"grad_norm": 7.2561694515252135,
"learning_rate": 1.968549422336329e-05,
"loss": 0.6164,
"step": 50
},
{
"epoch": 0.0192508221705302,
"grad_norm": 8.11462162647704,
"learning_rate": 1.962130937098845e-05,
"loss": 0.5159,
"step": 60
},
{
"epoch": 0.022459292532285232,
"grad_norm": 3.178981153593806,
"learning_rate": 1.955712451861361e-05,
"loss": 0.587,
"step": 70
},
{
"epoch": 0.025667762894040266,
"grad_norm": 2.850821452950468,
"learning_rate": 1.9492939666238767e-05,
"loss": 0.6162,
"step": 80
},
{
"epoch": 0.0288762332557953,
"grad_norm": 13.775135056345045,
"learning_rate": 1.942875481386393e-05,
"loss": 0.5739,
"step": 90
},
{
"epoch": 0.03208470361755033,
"grad_norm": 5.79356600659686,
"learning_rate": 1.936456996148909e-05,
"loss": 0.5711,
"step": 100
},
{
"epoch": 0.03529317397930536,
"grad_norm": 11.420684946729882,
"learning_rate": 1.930038510911425e-05,
"loss": 0.5451,
"step": 110
},
{
"epoch": 0.0385016443410604,
"grad_norm": 7.2576002708630885,
"learning_rate": 1.923620025673941e-05,
"loss": 0.5589,
"step": 120
},
{
"epoch": 0.04171011470281543,
"grad_norm": 19.723218198122705,
"learning_rate": 1.9172015404364574e-05,
"loss": 0.5755,
"step": 130
},
{
"epoch": 0.044918585064570464,
"grad_norm": 9.49503737660675,
"learning_rate": 1.910783055198973e-05,
"loss": 0.5659,
"step": 140
},
{
"epoch": 0.0481270554263255,
"grad_norm": 69.08236589580909,
"learning_rate": 1.904364569961489e-05,
"loss": 0.6462,
"step": 150
},
{
"epoch": 0.05133552578808053,
"grad_norm": 5.336246081074227,
"learning_rate": 1.897946084724005e-05,
"loss": 0.496,
"step": 160
},
{
"epoch": 0.054543996149835565,
"grad_norm": 3.33109537703745,
"learning_rate": 1.8915275994865214e-05,
"loss": 0.65,
"step": 170
},
{
"epoch": 0.0577524665115906,
"grad_norm": 2.6278156207881733,
"learning_rate": 1.8851091142490375e-05,
"loss": 0.5478,
"step": 180
},
{
"epoch": 0.06096093687334563,
"grad_norm": 4.599277071245572,
"learning_rate": 1.8786906290115535e-05,
"loss": 0.5982,
"step": 190
},
{
"epoch": 0.06416940723510066,
"grad_norm": 4.683411304137455,
"learning_rate": 1.8722721437740695e-05,
"loss": 0.5286,
"step": 200
},
{
"epoch": 0.06737787759685569,
"grad_norm": 5.723525300882751,
"learning_rate": 1.8658536585365855e-05,
"loss": 0.5801,
"step": 210
},
{
"epoch": 0.07058634795861073,
"grad_norm": 3.3681712914871267,
"learning_rate": 1.8594351732991015e-05,
"loss": 0.488,
"step": 220
},
{
"epoch": 0.07379481832036576,
"grad_norm": 5.149531756228593,
"learning_rate": 1.8530166880616175e-05,
"loss": 0.5275,
"step": 230
},
{
"epoch": 0.0770032886821208,
"grad_norm": 2.36276348719608,
"learning_rate": 1.8465982028241335e-05,
"loss": 0.5999,
"step": 240
},
{
"epoch": 0.08021175904387583,
"grad_norm": 4.1742494520943385,
"learning_rate": 1.8401797175866495e-05,
"loss": 0.4599,
"step": 250
},
{
"epoch": 0.08342022940563086,
"grad_norm": 5.863113033224418,
"learning_rate": 1.833761232349166e-05,
"loss": 0.5478,
"step": 260
},
{
"epoch": 0.0866286997673859,
"grad_norm": 3.362154435694139,
"learning_rate": 1.827342747111682e-05,
"loss": 0.5701,
"step": 270
},
{
"epoch": 0.08983717012914093,
"grad_norm": 2.214965696532549,
"learning_rate": 1.820924261874198e-05,
"loss": 0.5405,
"step": 280
},
{
"epoch": 0.09304564049089596,
"grad_norm": 2.4403160493361002,
"learning_rate": 1.814505776636714e-05,
"loss": 0.606,
"step": 290
},
{
"epoch": 0.096254110852651,
"grad_norm": 5.230913771058336,
"learning_rate": 1.80808729139923e-05,
"loss": 0.5164,
"step": 300
},
{
"epoch": 0.09946258121440603,
"grad_norm": 8.676841652114097,
"learning_rate": 1.801668806161746e-05,
"loss": 0.4947,
"step": 310
},
{
"epoch": 0.10267105157616106,
"grad_norm": 4.229891956660176,
"learning_rate": 1.795250320924262e-05,
"loss": 0.5498,
"step": 320
},
{
"epoch": 0.1058795219379161,
"grad_norm": 5.729777414238761,
"learning_rate": 1.788831835686778e-05,
"loss": 0.5233,
"step": 330
},
{
"epoch": 0.10908799229967113,
"grad_norm": 5.0150304103119066,
"learning_rate": 1.7824133504492943e-05,
"loss": 0.5724,
"step": 340
},
{
"epoch": 0.11229646266142616,
"grad_norm": 5.102515509807658,
"learning_rate": 1.7759948652118103e-05,
"loss": 0.5058,
"step": 350
},
{
"epoch": 0.1155049330231812,
"grad_norm": 6.548980463569489,
"learning_rate": 1.7695763799743263e-05,
"loss": 0.5197,
"step": 360
},
{
"epoch": 0.11871340338493623,
"grad_norm": 6.889762448941459,
"learning_rate": 1.763157894736842e-05,
"loss": 0.4994,
"step": 370
},
{
"epoch": 0.12192187374669126,
"grad_norm": 4.161313910880235,
"learning_rate": 1.7567394094993584e-05,
"loss": 0.5149,
"step": 380
},
{
"epoch": 0.1251303441084463,
"grad_norm": 3.1957418525794594,
"learning_rate": 1.7503209242618744e-05,
"loss": 0.5204,
"step": 390
},
{
"epoch": 0.12833881447020132,
"grad_norm": 2.8111005635758737,
"learning_rate": 1.7439024390243904e-05,
"loss": 0.5322,
"step": 400
},
{
"epoch": 0.13154728483195635,
"grad_norm": 2.5609366797797235,
"learning_rate": 1.7374839537869064e-05,
"loss": 0.5293,
"step": 410
},
{
"epoch": 0.13475575519371139,
"grad_norm": 2.6988610146254453,
"learning_rate": 1.7310654685494224e-05,
"loss": 0.5418,
"step": 420
},
{
"epoch": 0.13796422555546642,
"grad_norm": 2.5790644514696073,
"learning_rate": 1.7246469833119384e-05,
"loss": 0.5364,
"step": 430
},
{
"epoch": 0.14117269591722145,
"grad_norm": 2.2548548111368363,
"learning_rate": 1.7182284980744544e-05,
"loss": 0.531,
"step": 440
},
{
"epoch": 0.14438116627897649,
"grad_norm": 3.876505734407925,
"learning_rate": 1.7118100128369705e-05,
"loss": 0.5468,
"step": 450
},
{
"epoch": 0.14758963664073152,
"grad_norm": 2.3322492801350507,
"learning_rate": 1.7053915275994865e-05,
"loss": 0.5518,
"step": 460
},
{
"epoch": 0.15079810700248655,
"grad_norm": 2.9859767898263083,
"learning_rate": 1.6989730423620028e-05,
"loss": 0.4452,
"step": 470
},
{
"epoch": 0.1540065773642416,
"grad_norm": 3.467011650018448,
"learning_rate": 1.692554557124519e-05,
"loss": 0.5655,
"step": 480
},
{
"epoch": 0.15721504772599662,
"grad_norm": 1.683532622712415,
"learning_rate": 1.686136071887035e-05,
"loss": 0.5001,
"step": 490
},
{
"epoch": 0.16042351808775165,
"grad_norm": 4.116712390224751,
"learning_rate": 1.679717586649551e-05,
"loss": 0.4813,
"step": 500
},
{
"epoch": 0.1636319884495067,
"grad_norm": 2.103559867663371,
"learning_rate": 1.673299101412067e-05,
"loss": 0.5645,
"step": 510
},
{
"epoch": 0.16684045881126172,
"grad_norm": 2.4694939323846548,
"learning_rate": 1.666880616174583e-05,
"loss": 0.5125,
"step": 520
},
{
"epoch": 0.17004892917301676,
"grad_norm": 1.8947360245697789,
"learning_rate": 1.660462130937099e-05,
"loss": 0.5141,
"step": 530
},
{
"epoch": 0.1732573995347718,
"grad_norm": 3.238607664236306,
"learning_rate": 1.654043645699615e-05,
"loss": 0.4956,
"step": 540
},
{
"epoch": 0.17646586989652682,
"grad_norm": 2.353465408424258,
"learning_rate": 1.6476251604621313e-05,
"loss": 0.5519,
"step": 550
},
{
"epoch": 0.17967434025828186,
"grad_norm": 2.3759882226075173,
"learning_rate": 1.6412066752246473e-05,
"loss": 0.5024,
"step": 560
},
{
"epoch": 0.1828828106200369,
"grad_norm": 2.8012970950830813,
"learning_rate": 1.6347881899871633e-05,
"loss": 0.5538,
"step": 570
},
{
"epoch": 0.18609128098179192,
"grad_norm": 2.022401128854332,
"learning_rate": 1.6283697047496793e-05,
"loss": 0.4805,
"step": 580
},
{
"epoch": 0.18929975134354696,
"grad_norm": 4.470521851999792,
"learning_rate": 1.6219512195121953e-05,
"loss": 0.5247,
"step": 590
},
{
"epoch": 0.192508221705302,
"grad_norm": 3.2544380093576573,
"learning_rate": 1.6155327342747113e-05,
"loss": 0.5103,
"step": 600
},
{
"epoch": 0.19571669206705702,
"grad_norm": 1.9968945151022015,
"learning_rate": 1.6091142490372273e-05,
"loss": 0.4856,
"step": 610
},
{
"epoch": 0.19892516242881206,
"grad_norm": 2.291451765421134,
"learning_rate": 1.6026957637997433e-05,
"loss": 0.5562,
"step": 620
},
{
"epoch": 0.2021336327905671,
"grad_norm": 2.7623193687984404,
"learning_rate": 1.5962772785622594e-05,
"loss": 0.5227,
"step": 630
},
{
"epoch": 0.20534210315232213,
"grad_norm": 3.1301915722545623,
"learning_rate": 1.5898587933247757e-05,
"loss": 0.5062,
"step": 640
},
{
"epoch": 0.20855057351407716,
"grad_norm": 3.4265676844281336,
"learning_rate": 1.5834403080872917e-05,
"loss": 0.5372,
"step": 650
},
{
"epoch": 0.2117590438758322,
"grad_norm": 2.622419582355671,
"learning_rate": 1.5770218228498074e-05,
"loss": 0.4891,
"step": 660
},
{
"epoch": 0.21496751423758723,
"grad_norm": 3.619394589579225,
"learning_rate": 1.5706033376123234e-05,
"loss": 0.4631,
"step": 670
},
{
"epoch": 0.21817598459934226,
"grad_norm": 3.163462566837581,
"learning_rate": 1.5641848523748397e-05,
"loss": 0.5311,
"step": 680
},
{
"epoch": 0.2213844549610973,
"grad_norm": 2.25896546636362,
"learning_rate": 1.5577663671373558e-05,
"loss": 0.556,
"step": 690
},
{
"epoch": 0.22459292532285233,
"grad_norm": 2.621602161152921,
"learning_rate": 1.5513478818998718e-05,
"loss": 0.5044,
"step": 700
},
{
"epoch": 0.22780139568460736,
"grad_norm": 2.384987921495825,
"learning_rate": 1.5449293966623878e-05,
"loss": 0.5233,
"step": 710
},
{
"epoch": 0.2310098660463624,
"grad_norm": 2.9900524475871086,
"learning_rate": 1.5385109114249038e-05,
"loss": 0.4475,
"step": 720
},
{
"epoch": 0.23421833640811743,
"grad_norm": 8.182093983069542,
"learning_rate": 1.5320924261874198e-05,
"loss": 0.573,
"step": 730
},
{
"epoch": 0.23742680676987246,
"grad_norm": 2.472013578988132,
"learning_rate": 1.525673940949936e-05,
"loss": 0.4952,
"step": 740
},
{
"epoch": 0.2406352771316275,
"grad_norm": 1.605749825961759,
"learning_rate": 1.519255455712452e-05,
"loss": 0.5739,
"step": 750
},
{
"epoch": 0.24384374749338253,
"grad_norm": 3.9316988027215363,
"learning_rate": 1.5128369704749682e-05,
"loss": 0.4913,
"step": 760
},
{
"epoch": 0.24705221785513756,
"grad_norm": 2.9837644539540817,
"learning_rate": 1.5064184852374842e-05,
"loss": 0.4709,
"step": 770
},
{
"epoch": 0.2502606882168926,
"grad_norm": 2.648880843504189,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.5545,
"step": 780
},
{
"epoch": 0.2534691585786476,
"grad_norm": 1.9189980549728858,
"learning_rate": 1.493581514762516e-05,
"loss": 0.5249,
"step": 790
},
{
"epoch": 0.25667762894040264,
"grad_norm": 2.840046662447408,
"learning_rate": 1.487163029525032e-05,
"loss": 0.4957,
"step": 800
},
{
"epoch": 0.25988609930215767,
"grad_norm": 3.951112516970372,
"learning_rate": 1.4807445442875482e-05,
"loss": 0.4858,
"step": 810
},
{
"epoch": 0.2630945696639127,
"grad_norm": 2.0113103957025924,
"learning_rate": 1.4743260590500643e-05,
"loss": 0.5611,
"step": 820
},
{
"epoch": 0.26630304002566774,
"grad_norm": 2.1903304602185867,
"learning_rate": 1.4679075738125803e-05,
"loss": 0.4777,
"step": 830
},
{
"epoch": 0.26951151038742277,
"grad_norm": 2.3710284722009853,
"learning_rate": 1.4614890885750963e-05,
"loss": 0.4642,
"step": 840
},
{
"epoch": 0.2727199807491778,
"grad_norm": 3.9101691341234637,
"learning_rate": 1.4550706033376125e-05,
"loss": 0.4889,
"step": 850
},
{
"epoch": 0.27592845111093284,
"grad_norm": 2.952363378779562,
"learning_rate": 1.4486521181001285e-05,
"loss": 0.4777,
"step": 860
},
{
"epoch": 0.27913692147268787,
"grad_norm": 2.7290967023139974,
"learning_rate": 1.4422336328626445e-05,
"loss": 0.4876,
"step": 870
},
{
"epoch": 0.2823453918344429,
"grad_norm": 3.8061885678039107,
"learning_rate": 1.4358151476251605e-05,
"loss": 0.474,
"step": 880
},
{
"epoch": 0.28555386219619794,
"grad_norm": 3.2311031015838463,
"learning_rate": 1.4293966623876767e-05,
"loss": 0.4986,
"step": 890
},
{
"epoch": 0.28876233255795297,
"grad_norm": 3.5264525429966325,
"learning_rate": 1.4229781771501927e-05,
"loss": 0.4735,
"step": 900
},
{
"epoch": 0.291970802919708,
"grad_norm": 4.018761523963046,
"learning_rate": 1.4165596919127087e-05,
"loss": 0.4252,
"step": 910
},
{
"epoch": 0.29517927328146304,
"grad_norm": 8.852718468433057,
"learning_rate": 1.4101412066752247e-05,
"loss": 0.4928,
"step": 920
},
{
"epoch": 0.2983877436432181,
"grad_norm": 2.7603431611657063,
"learning_rate": 1.4037227214377409e-05,
"loss": 0.4511,
"step": 930
},
{
"epoch": 0.3015962140049731,
"grad_norm": 3.0352627378285235,
"learning_rate": 1.3973042362002569e-05,
"loss": 0.4901,
"step": 940
},
{
"epoch": 0.30480468436672814,
"grad_norm": 2.9470105935081428,
"learning_rate": 1.390885750962773e-05,
"loss": 0.5307,
"step": 950
},
{
"epoch": 0.3080131547284832,
"grad_norm": 2.955919829305562,
"learning_rate": 1.384467265725289e-05,
"loss": 0.479,
"step": 960
},
{
"epoch": 0.3112216250902382,
"grad_norm": 6.644848963046123,
"learning_rate": 1.378048780487805e-05,
"loss": 0.5007,
"step": 970
},
{
"epoch": 0.31443009545199324,
"grad_norm": 3.5199239244904272,
"learning_rate": 1.3716302952503211e-05,
"loss": 0.5033,
"step": 980
},
{
"epoch": 0.3176385658137483,
"grad_norm": 2.065870849140256,
"learning_rate": 1.3652118100128371e-05,
"loss": 0.5392,
"step": 990
},
{
"epoch": 0.3208470361755033,
"grad_norm": 2.413778269125468,
"learning_rate": 1.3587933247753531e-05,
"loss": 0.5288,
"step": 1000
},
{
"epoch": 0.32405550653725834,
"grad_norm": 3.2146909935356502,
"learning_rate": 1.3523748395378692e-05,
"loss": 0.4934,
"step": 1010
},
{
"epoch": 0.3272639768990134,
"grad_norm": 2.480925294107376,
"learning_rate": 1.3459563543003853e-05,
"loss": 0.4863,
"step": 1020
},
{
"epoch": 0.3304724472607684,
"grad_norm": 2.9783703488356146,
"learning_rate": 1.3395378690629014e-05,
"loss": 0.4817,
"step": 1030
},
{
"epoch": 0.33368091762252344,
"grad_norm": 3.639386122561336,
"learning_rate": 1.3331193838254172e-05,
"loss": 0.4593,
"step": 1040
},
{
"epoch": 0.3368893879842785,
"grad_norm": 1.9228374918291111,
"learning_rate": 1.3267008985879332e-05,
"loss": 0.52,
"step": 1050
},
{
"epoch": 0.3400978583460335,
"grad_norm": 3.669574333733756,
"learning_rate": 1.3202824133504496e-05,
"loss": 0.4615,
"step": 1060
},
{
"epoch": 0.34330632870778854,
"grad_norm": 2.8023771884080957,
"learning_rate": 1.3138639281129654e-05,
"loss": 0.4425,
"step": 1070
},
{
"epoch": 0.3465147990695436,
"grad_norm": 2.6520853045574406,
"learning_rate": 1.3074454428754814e-05,
"loss": 0.4855,
"step": 1080
},
{
"epoch": 0.3497232694312986,
"grad_norm": 2.8800252087526816,
"learning_rate": 1.3010269576379974e-05,
"loss": 0.4513,
"step": 1090
},
{
"epoch": 0.35293173979305364,
"grad_norm": 4.041897289359244,
"learning_rate": 1.2946084724005136e-05,
"loss": 0.4452,
"step": 1100
},
{
"epoch": 0.3561402101548087,
"grad_norm": 4.26233252928846,
"learning_rate": 1.2881899871630296e-05,
"loss": 0.434,
"step": 1110
},
{
"epoch": 0.3593486805165637,
"grad_norm": 3.4927249634501876,
"learning_rate": 1.2817715019255456e-05,
"loss": 0.452,
"step": 1120
},
{
"epoch": 0.36255715087831875,
"grad_norm": 2.6582633853498523,
"learning_rate": 1.2753530166880616e-05,
"loss": 0.5514,
"step": 1130
},
{
"epoch": 0.3657656212400738,
"grad_norm": 3.1556020622384304,
"learning_rate": 1.2689345314505778e-05,
"loss": 0.4795,
"step": 1140
},
{
"epoch": 0.3689740916018288,
"grad_norm": 3.9929879724837347,
"learning_rate": 1.2625160462130938e-05,
"loss": 0.4787,
"step": 1150
},
{
"epoch": 0.37218256196358385,
"grad_norm": 2.9260030419548104,
"learning_rate": 1.2560975609756098e-05,
"loss": 0.4774,
"step": 1160
},
{
"epoch": 0.3753910323253389,
"grad_norm": 2.9983398150479084,
"learning_rate": 1.2496790757381259e-05,
"loss": 0.4896,
"step": 1170
},
{
"epoch": 0.3785995026870939,
"grad_norm": 2.209211943075685,
"learning_rate": 1.2432605905006419e-05,
"loss": 0.4717,
"step": 1180
},
{
"epoch": 0.38180797304884895,
"grad_norm": 4.374188573483919,
"learning_rate": 1.236842105263158e-05,
"loss": 0.4637,
"step": 1190
},
{
"epoch": 0.385016443410604,
"grad_norm": 4.244625220521253,
"learning_rate": 1.230423620025674e-05,
"loss": 0.5024,
"step": 1200
},
{
"epoch": 0.388224913772359,
"grad_norm": 3.5167143804742507,
"learning_rate": 1.22400513478819e-05,
"loss": 0.508,
"step": 1210
},
{
"epoch": 0.39143338413411405,
"grad_norm": 2.400382007637674,
"learning_rate": 1.2175866495507061e-05,
"loss": 0.5444,
"step": 1220
},
{
"epoch": 0.3946418544958691,
"grad_norm": 2.6699722334400087,
"learning_rate": 1.2111681643132223e-05,
"loss": 0.4427,
"step": 1230
},
{
"epoch": 0.3978503248576241,
"grad_norm": 3.1363599950404297,
"learning_rate": 1.2047496790757383e-05,
"loss": 0.5484,
"step": 1240
},
{
"epoch": 0.40105879521937915,
"grad_norm": 2.5247313838937835,
"learning_rate": 1.1983311938382543e-05,
"loss": 0.4482,
"step": 1250
},
{
"epoch": 0.4042672655811342,
"grad_norm": 4.325865942489473,
"learning_rate": 1.1919127086007703e-05,
"loss": 0.4634,
"step": 1260
},
{
"epoch": 0.4074757359428892,
"grad_norm": 5.342742581507827,
"learning_rate": 1.1854942233632865e-05,
"loss": 0.4621,
"step": 1270
},
{
"epoch": 0.41068420630464425,
"grad_norm": 3.652667902032412,
"learning_rate": 1.1790757381258025e-05,
"loss": 0.4521,
"step": 1280
},
{
"epoch": 0.4138926766663993,
"grad_norm": 4.912991499737618,
"learning_rate": 1.1726572528883185e-05,
"loss": 0.4585,
"step": 1290
},
{
"epoch": 0.4171011470281543,
"grad_norm": 3.5942743788211367,
"learning_rate": 1.1662387676508344e-05,
"loss": 0.5203,
"step": 1300
},
{
"epoch": 0.42030961738990935,
"grad_norm": 3.5948972213370496,
"learning_rate": 1.1598202824133507e-05,
"loss": 0.5324,
"step": 1310
},
{
"epoch": 0.4235180877516644,
"grad_norm": 3.2423123257702144,
"learning_rate": 1.1534017971758667e-05,
"loss": 0.4511,
"step": 1320
},
{
"epoch": 0.4267265581134194,
"grad_norm": 3.1372521267300737,
"learning_rate": 1.1469833119383826e-05,
"loss": 0.4607,
"step": 1330
},
{
"epoch": 0.42993502847517445,
"grad_norm": 2.7989457017474386,
"learning_rate": 1.1405648267008986e-05,
"loss": 0.4939,
"step": 1340
},
{
"epoch": 0.4331434988369295,
"grad_norm": 2.873878909867159,
"learning_rate": 1.1341463414634146e-05,
"loss": 0.4614,
"step": 1350
},
{
"epoch": 0.4363519691986845,
"grad_norm": 5.508444120708098,
"learning_rate": 1.1277278562259308e-05,
"loss": 0.4999,
"step": 1360
},
{
"epoch": 0.43956043956043955,
"grad_norm": 4.61907829831726,
"learning_rate": 1.1213093709884468e-05,
"loss": 0.4332,
"step": 1370
},
{
"epoch": 0.4427689099221946,
"grad_norm": 4.032088995782747,
"learning_rate": 1.1148908857509628e-05,
"loss": 0.417,
"step": 1380
},
{
"epoch": 0.4459773802839496,
"grad_norm": 3.179640697268296,
"learning_rate": 1.1084724005134788e-05,
"loss": 0.5368,
"step": 1390
},
{
"epoch": 0.44918585064570465,
"grad_norm": 2.5857678925367686,
"learning_rate": 1.102053915275995e-05,
"loss": 0.4427,
"step": 1400
},
{
"epoch": 0.4523943210074597,
"grad_norm": 3.119333721193506,
"learning_rate": 1.095635430038511e-05,
"loss": 0.5349,
"step": 1410
},
{
"epoch": 0.4556027913692147,
"grad_norm": 2.535312149882907,
"learning_rate": 1.089216944801027e-05,
"loss": 0.4494,
"step": 1420
},
{
"epoch": 0.45881126173096975,
"grad_norm": 4.307386342890277,
"learning_rate": 1.082798459563543e-05,
"loss": 0.4424,
"step": 1430
},
{
"epoch": 0.4620197320927248,
"grad_norm": 4.636318948357751,
"learning_rate": 1.0763799743260592e-05,
"loss": 0.5043,
"step": 1440
},
{
"epoch": 0.4652282024544798,
"grad_norm": 3.284720353393845,
"learning_rate": 1.0699614890885752e-05,
"loss": 0.4412,
"step": 1450
},
{
"epoch": 0.46843667281623486,
"grad_norm": 4.531770132295603,
"learning_rate": 1.0635430038510912e-05,
"loss": 0.477,
"step": 1460
},
{
"epoch": 0.4716451431779899,
"grad_norm": 5.337340216101621,
"learning_rate": 1.0571245186136072e-05,
"loss": 0.4647,
"step": 1470
},
{
"epoch": 0.4748536135397449,
"grad_norm": 3.990612631287861,
"learning_rate": 1.0507060333761234e-05,
"loss": 0.459,
"step": 1480
},
{
"epoch": 0.47806208390149996,
"grad_norm": 3.715722505194795,
"learning_rate": 1.0442875481386394e-05,
"loss": 0.4346,
"step": 1490
},
{
"epoch": 0.481270554263255,
"grad_norm": 4.810153027732736,
"learning_rate": 1.0378690629011554e-05,
"loss": 0.5225,
"step": 1500
},
{
"epoch": 0.48447902462501,
"grad_norm": 4.079307241955873,
"learning_rate": 1.0314505776636715e-05,
"loss": 0.4518,
"step": 1510
},
{
"epoch": 0.48768749498676506,
"grad_norm": 3.68729326730919,
"learning_rate": 1.0250320924261875e-05,
"loss": 0.4269,
"step": 1520
},
{
"epoch": 0.4908959653485201,
"grad_norm": 4.254306350469639,
"learning_rate": 1.0186136071887036e-05,
"loss": 0.3975,
"step": 1530
},
{
"epoch": 0.4941044357102751,
"grad_norm": 3.0833625373993083,
"learning_rate": 1.0121951219512197e-05,
"loss": 0.5438,
"step": 1540
},
{
"epoch": 0.49731290607203016,
"grad_norm": 3.59089189260731,
"learning_rate": 1.0057766367137357e-05,
"loss": 0.4417,
"step": 1550
},
{
"epoch": 0.5005213764337852,
"grad_norm": 3.237058915356381,
"learning_rate": 9.993581514762517e-06,
"loss": 0.4557,
"step": 1560
},
{
"epoch": 0.5037298467955402,
"grad_norm": 5.297271530811518,
"learning_rate": 9.929396662387677e-06,
"loss": 0.4145,
"step": 1570
},
{
"epoch": 0.5069383171572952,
"grad_norm": 3.610819876445011,
"learning_rate": 9.865211810012839e-06,
"loss": 0.46,
"step": 1580
},
{
"epoch": 0.5101467875190503,
"grad_norm": 3.7494000521584,
"learning_rate": 9.801026957637997e-06,
"loss": 0.432,
"step": 1590
},
{
"epoch": 0.5133552578808053,
"grad_norm": 3.499306405822366,
"learning_rate": 9.736842105263159e-06,
"loss": 0.4619,
"step": 1600
},
{
"epoch": 0.5165637282425604,
"grad_norm": 3.1804132252621105,
"learning_rate": 9.672657252888319e-06,
"loss": 0.463,
"step": 1610
},
{
"epoch": 0.5197721986043153,
"grad_norm": 2.9096715202099217,
"learning_rate": 9.60847240051348e-06,
"loss": 0.441,
"step": 1620
},
{
"epoch": 0.5229806689660704,
"grad_norm": 5.100772321999976,
"learning_rate": 9.54428754813864e-06,
"loss": 0.4719,
"step": 1630
},
{
"epoch": 0.5261891393278254,
"grad_norm": 4.551318414118453,
"learning_rate": 9.480102695763801e-06,
"loss": 0.4592,
"step": 1640
},
{
"epoch": 0.5293976096895805,
"grad_norm": 3.2591870164503396,
"learning_rate": 9.415917843388961e-06,
"loss": 0.3951,
"step": 1650
},
{
"epoch": 0.5326060800513355,
"grad_norm": 3.9139265084170685,
"learning_rate": 9.351732991014121e-06,
"loss": 0.4514,
"step": 1660
},
{
"epoch": 0.5358145504130906,
"grad_norm": 5.612004065083535,
"learning_rate": 9.287548138639282e-06,
"loss": 0.4091,
"step": 1670
},
{
"epoch": 0.5390230207748455,
"grad_norm": 4.535440967836521,
"learning_rate": 9.223363286264443e-06,
"loss": 0.4142,
"step": 1680
},
{
"epoch": 0.5422314911366006,
"grad_norm": 4.47578230841436,
"learning_rate": 9.159178433889603e-06,
"loss": 0.452,
"step": 1690
},
{
"epoch": 0.5454399614983556,
"grad_norm": 3.904566068173329,
"learning_rate": 9.094993581514764e-06,
"loss": 0.4614,
"step": 1700
},
{
"epoch": 0.5486484318601107,
"grad_norm": 5.1973431860771235,
"learning_rate": 9.030808729139924e-06,
"loss": 0.4461,
"step": 1710
},
{
"epoch": 0.5518569022218657,
"grad_norm": 2.640914950150599,
"learning_rate": 8.966623876765084e-06,
"loss": 0.415,
"step": 1720
},
{
"epoch": 0.5550653725836208,
"grad_norm": 3.4701402796437413,
"learning_rate": 8.902439024390244e-06,
"loss": 0.4601,
"step": 1730
},
{
"epoch": 0.5582738429453757,
"grad_norm": 2.465609370132583,
"learning_rate": 8.838254172015404e-06,
"loss": 0.5022,
"step": 1740
},
{
"epoch": 0.5614823133071308,
"grad_norm": 3.708119324189301,
"learning_rate": 8.774069319640566e-06,
"loss": 0.4482,
"step": 1750
},
{
"epoch": 0.5646907836688858,
"grad_norm": 2.787335621305455,
"learning_rate": 8.709884467265726e-06,
"loss": 0.4565,
"step": 1760
},
{
"epoch": 0.5678992540306409,
"grad_norm": 3.12311139793224,
"learning_rate": 8.645699614890886e-06,
"loss": 0.4438,
"step": 1770
},
{
"epoch": 0.5711077243923959,
"grad_norm": 3.7334357424674676,
"learning_rate": 8.581514762516046e-06,
"loss": 0.527,
"step": 1780
},
{
"epoch": 0.574316194754151,
"grad_norm": 3.028289664673344,
"learning_rate": 8.517329910141208e-06,
"loss": 0.4604,
"step": 1790
},
{
"epoch": 0.5775246651159059,
"grad_norm": 3.4462505055852395,
"learning_rate": 8.453145057766368e-06,
"loss": 0.4505,
"step": 1800
},
{
"epoch": 0.580733135477661,
"grad_norm": 2.9231124271855027,
"learning_rate": 8.388960205391528e-06,
"loss": 0.477,
"step": 1810
},
{
"epoch": 0.583941605839416,
"grad_norm": 3.380678577680448,
"learning_rate": 8.324775353016688e-06,
"loss": 0.4637,
"step": 1820
},
{
"epoch": 0.5871500762011711,
"grad_norm": 2.1604791729799393,
"learning_rate": 8.26059050064185e-06,
"loss": 0.4247,
"step": 1830
},
{
"epoch": 0.5903585465629261,
"grad_norm": 7.324804377048017,
"learning_rate": 8.19640564826701e-06,
"loss": 0.481,
"step": 1840
},
{
"epoch": 0.5935670169246812,
"grad_norm": 3.56229093488035,
"learning_rate": 8.13222079589217e-06,
"loss": 0.4524,
"step": 1850
},
{
"epoch": 0.5967754872864361,
"grad_norm": 2.9851487649386264,
"learning_rate": 8.06803594351733e-06,
"loss": 0.4711,
"step": 1860
},
{
"epoch": 0.5999839576481912,
"grad_norm": 3.4086079773035958,
"learning_rate": 8.00385109114249e-06,
"loss": 0.5267,
"step": 1870
},
{
"epoch": 0.6031924280099462,
"grad_norm": 3.9347002620116207,
"learning_rate": 7.93966623876765e-06,
"loss": 0.4605,
"step": 1880
},
{
"epoch": 0.6064008983717013,
"grad_norm": 4.119014906734088,
"learning_rate": 7.875481386392811e-06,
"loss": 0.4254,
"step": 1890
},
{
"epoch": 0.6096093687334563,
"grad_norm": 4.938854924688182,
"learning_rate": 7.811296534017973e-06,
"loss": 0.4422,
"step": 1900
},
{
"epoch": 0.6128178390952114,
"grad_norm": 4.186657480383815,
"learning_rate": 7.747111681643133e-06,
"loss": 0.4198,
"step": 1910
},
{
"epoch": 0.6160263094569663,
"grad_norm": 3.89892582957843,
"learning_rate": 7.682926829268293e-06,
"loss": 0.4412,
"step": 1920
},
{
"epoch": 0.6192347798187214,
"grad_norm": 3.585561416303363,
"learning_rate": 7.618741976893453e-06,
"loss": 0.3946,
"step": 1930
},
{
"epoch": 0.6224432501804764,
"grad_norm": 4.78580164859429,
"learning_rate": 7.554557124518614e-06,
"loss": 0.5507,
"step": 1940
},
{
"epoch": 0.6256517205422315,
"grad_norm": 2.914386305013241,
"learning_rate": 7.490372272143774e-06,
"loss": 0.4968,
"step": 1950
},
{
"epoch": 0.6288601909039865,
"grad_norm": 3.32458898490946,
"learning_rate": 7.426187419768935e-06,
"loss": 0.3965,
"step": 1960
},
{
"epoch": 0.6320686612657416,
"grad_norm": 4.929703501803635,
"learning_rate": 7.362002567394095e-06,
"loss": 0.4861,
"step": 1970
},
{
"epoch": 0.6352771316274965,
"grad_norm": 3.598051864515239,
"learning_rate": 7.297817715019256e-06,
"loss": 0.4207,
"step": 1980
},
{
"epoch": 0.6384856019892516,
"grad_norm": 4.096289572909773,
"learning_rate": 7.233632862644416e-06,
"loss": 0.4957,
"step": 1990
},
{
"epoch": 0.6416940723510066,
"grad_norm": 3.3893589556689245,
"learning_rate": 7.169448010269577e-06,
"loss": 0.4624,
"step": 2000
},
{
"epoch": 0.6449025427127617,
"grad_norm": 4.907884833709519,
"learning_rate": 7.1052631578947375e-06,
"loss": 0.4058,
"step": 2010
},
{
"epoch": 0.6481110130745167,
"grad_norm": 3.973310201701845,
"learning_rate": 7.0410783055198984e-06,
"loss": 0.3997,
"step": 2020
},
{
"epoch": 0.6513194834362718,
"grad_norm": 4.54948810422854,
"learning_rate": 6.9768934531450586e-06,
"loss": 0.4204,
"step": 2030
},
{
"epoch": 0.6545279537980268,
"grad_norm": 6.197502093297683,
"learning_rate": 6.9127086007702195e-06,
"loss": 0.3986,
"step": 2040
},
{
"epoch": 0.6577364241597818,
"grad_norm": 5.619769890699808,
"learning_rate": 6.84852374839538e-06,
"loss": 0.4481,
"step": 2050
},
{
"epoch": 0.6609448945215368,
"grad_norm": 4.127234906064621,
"learning_rate": 6.784338896020539e-06,
"loss": 0.4834,
"step": 2060
},
{
"epoch": 0.6641533648832919,
"grad_norm": 2.8888742751433503,
"learning_rate": 6.7201540436457e-06,
"loss": 0.4473,
"step": 2070
},
{
"epoch": 0.6673618352450469,
"grad_norm": 4.168159373138179,
"learning_rate": 6.65596919127086e-06,
"loss": 0.4435,
"step": 2080
},
{
"epoch": 0.670570305606802,
"grad_norm": 3.4090037884789797,
"learning_rate": 6.591784338896021e-06,
"loss": 0.3963,
"step": 2090
},
{
"epoch": 0.673778775968557,
"grad_norm": 5.6886431016692525,
"learning_rate": 6.527599486521181e-06,
"loss": 0.4251,
"step": 2100
},
{
"epoch": 0.676987246330312,
"grad_norm": 4.113610869063736,
"learning_rate": 6.463414634146342e-06,
"loss": 0.4259,
"step": 2110
},
{
"epoch": 0.680195716692067,
"grad_norm": 3.7739117521255614,
"learning_rate": 6.399229781771502e-06,
"loss": 0.4629,
"step": 2120
},
{
"epoch": 0.6834041870538221,
"grad_norm": 3.481942659600282,
"learning_rate": 6.335044929396663e-06,
"loss": 0.4105,
"step": 2130
},
{
"epoch": 0.6866126574155771,
"grad_norm": 4.710993726687856,
"learning_rate": 6.270860077021823e-06,
"loss": 0.4713,
"step": 2140
},
{
"epoch": 0.6898211277773322,
"grad_norm": 4.412137149590203,
"learning_rate": 6.206675224646984e-06,
"loss": 0.4512,
"step": 2150
},
{
"epoch": 0.6930295981390872,
"grad_norm": 3.379903359603239,
"learning_rate": 6.142490372272144e-06,
"loss": 0.406,
"step": 2160
},
{
"epoch": 0.6962380685008422,
"grad_norm": 4.8974119039822765,
"learning_rate": 6.078305519897305e-06,
"loss": 0.4401,
"step": 2170
},
{
"epoch": 0.6994465388625972,
"grad_norm": 5.162061128021679,
"learning_rate": 6.0141206675224654e-06,
"loss": 0.4585,
"step": 2180
},
{
"epoch": 0.7026550092243523,
"grad_norm": 3.003309052194279,
"learning_rate": 5.949935815147626e-06,
"loss": 0.4072,
"step": 2190
},
{
"epoch": 0.7058634795861073,
"grad_norm": 3.0294439563791804,
"learning_rate": 5.885750962772786e-06,
"loss": 0.4465,
"step": 2200
},
{
"epoch": 0.7090719499478624,
"grad_norm": 3.3642147201952644,
"learning_rate": 5.821566110397947e-06,
"loss": 0.4557,
"step": 2210
},
{
"epoch": 0.7122804203096174,
"grad_norm": 5.568881453270445,
"learning_rate": 5.757381258023107e-06,
"loss": 0.4147,
"step": 2220
},
{
"epoch": 0.7154888906713724,
"grad_norm": 6.767424606475986,
"learning_rate": 5.693196405648267e-06,
"loss": 0.4794,
"step": 2230
},
{
"epoch": 0.7186973610331274,
"grad_norm": 3.4691493448860036,
"learning_rate": 5.629011553273428e-06,
"loss": 0.3804,
"step": 2240
},
{
"epoch": 0.7219058313948825,
"grad_norm": 4.389174505130007,
"learning_rate": 5.564826700898588e-06,
"loss": 0.4314,
"step": 2250
},
{
"epoch": 0.7251143017566375,
"grad_norm": 5.1763397195942495,
"learning_rate": 5.500641848523749e-06,
"loss": 0.4516,
"step": 2260
},
{
"epoch": 0.7283227721183926,
"grad_norm": 6.453395088205449,
"learning_rate": 5.436456996148909e-06,
"loss": 0.3971,
"step": 2270
},
{
"epoch": 0.7315312424801476,
"grad_norm": 4.859129046226108,
"learning_rate": 5.37227214377407e-06,
"loss": 0.4209,
"step": 2280
},
{
"epoch": 0.7347397128419026,
"grad_norm": 6.20390961892639,
"learning_rate": 5.30808729139923e-06,
"loss": 0.4405,
"step": 2290
},
{
"epoch": 0.7379481832036576,
"grad_norm": 6.554464041755333,
"learning_rate": 5.243902439024391e-06,
"loss": 0.4829,
"step": 2300
},
{
"epoch": 0.7411566535654127,
"grad_norm": 4.0876830410935865,
"learning_rate": 5.179717586649551e-06,
"loss": 0.4512,
"step": 2310
},
{
"epoch": 0.7443651239271677,
"grad_norm": 5.396362200394774,
"learning_rate": 5.115532734274712e-06,
"loss": 0.4915,
"step": 2320
},
{
"epoch": 0.7475735942889228,
"grad_norm": 4.072600137297783,
"learning_rate": 5.0513478818998715e-06,
"loss": 0.4366,
"step": 2330
},
{
"epoch": 0.7507820646506778,
"grad_norm": 3.9870063050444107,
"learning_rate": 4.9871630295250324e-06,
"loss": 0.3927,
"step": 2340
},
{
"epoch": 0.7539905350124328,
"grad_norm": 5.0372805923715145,
"learning_rate": 4.9229781771501926e-06,
"loss": 0.4174,
"step": 2350
},
{
"epoch": 0.7571990053741878,
"grad_norm": 5.331045879340357,
"learning_rate": 4.8587933247753535e-06,
"loss": 0.4141,
"step": 2360
},
{
"epoch": 0.7604074757359429,
"grad_norm": 6.769471319170081,
"learning_rate": 4.794608472400514e-06,
"loss": 0.47,
"step": 2370
},
{
"epoch": 0.7636159460976979,
"grad_norm": 3.585314593462169,
"learning_rate": 4.730423620025675e-06,
"loss": 0.4423,
"step": 2380
},
{
"epoch": 0.766824416459453,
"grad_norm": 5.427931476960088,
"learning_rate": 4.666238767650835e-06,
"loss": 0.4721,
"step": 2390
},
{
"epoch": 0.770032886821208,
"grad_norm": 4.5004756513383635,
"learning_rate": 4.602053915275995e-06,
"loss": 0.43,
"step": 2400
},
{
"epoch": 0.773241357182963,
"grad_norm": 4.409847168633205,
"learning_rate": 4.537869062901156e-06,
"loss": 0.4584,
"step": 2410
},
{
"epoch": 0.776449827544718,
"grad_norm": 7.0363734110991665,
"learning_rate": 4.473684210526316e-06,
"loss": 0.4765,
"step": 2420
},
{
"epoch": 0.7796582979064731,
"grad_norm": 5.141242089016943,
"learning_rate": 4.409499358151477e-06,
"loss": 0.4248,
"step": 2430
},
{
"epoch": 0.7828667682682281,
"grad_norm": 5.033719326184196,
"learning_rate": 4.345314505776637e-06,
"loss": 0.419,
"step": 2440
},
{
"epoch": 0.7860752386299832,
"grad_norm": 7.676121465498259,
"learning_rate": 4.281129653401798e-06,
"loss": 0.4561,
"step": 2450
},
{
"epoch": 0.7892837089917382,
"grad_norm": 5.491299992756061,
"learning_rate": 4.216944801026958e-06,
"loss": 0.389,
"step": 2460
},
{
"epoch": 0.7924921793534933,
"grad_norm": 4.7733994239458095,
"learning_rate": 4.152759948652118e-06,
"loss": 0.4261,
"step": 2470
},
{
"epoch": 0.7957006497152482,
"grad_norm": 5.037783180233028,
"learning_rate": 4.088575096277279e-06,
"loss": 0.4903,
"step": 2480
},
{
"epoch": 0.7989091200770033,
"grad_norm": 4.176851759512309,
"learning_rate": 4.024390243902439e-06,
"loss": 0.4026,
"step": 2490
},
{
"epoch": 0.8021175904387583,
"grad_norm": 4.864155304277976,
"learning_rate": 3.9602053915276e-06,
"loss": 0.4178,
"step": 2500
},
{
"epoch": 0.8053260608005134,
"grad_norm": 3.245115627201169,
"learning_rate": 3.89602053915276e-06,
"loss": 0.4091,
"step": 2510
},
{
"epoch": 0.8085345311622684,
"grad_norm": 4.572366748536081,
"learning_rate": 3.8318356867779205e-06,
"loss": 0.3905,
"step": 2520
},
{
"epoch": 0.8117430015240235,
"grad_norm": 3.241814776447211,
"learning_rate": 3.767650834403081e-06,
"loss": 0.377,
"step": 2530
},
{
"epoch": 0.8149514718857784,
"grad_norm": 5.503948157222115,
"learning_rate": 3.7034659820282416e-06,
"loss": 0.4008,
"step": 2540
},
{
"epoch": 0.8181599422475335,
"grad_norm": 5.857007420290123,
"learning_rate": 3.639281129653402e-06,
"loss": 0.4473,
"step": 2550
},
{
"epoch": 0.8213684126092885,
"grad_norm": 6.504205015228026,
"learning_rate": 3.5750962772785623e-06,
"loss": 0.3961,
"step": 2560
},
{
"epoch": 0.8245768829710436,
"grad_norm": 4.825768379913821,
"learning_rate": 3.510911424903723e-06,
"loss": 0.3894,
"step": 2570
},
{
"epoch": 0.8277853533327986,
"grad_norm": 3.9433924818002155,
"learning_rate": 3.4467265725288834e-06,
"loss": 0.419,
"step": 2580
},
{
"epoch": 0.8309938236945537,
"grad_norm": 5.060947152981452,
"learning_rate": 3.382541720154044e-06,
"loss": 0.4674,
"step": 2590
},
{
"epoch": 0.8342022940563086,
"grad_norm": 5.270918878343617,
"learning_rate": 3.3183568677792044e-06,
"loss": 0.4186,
"step": 2600
},
{
"epoch": 0.8374107644180637,
"grad_norm": 4.222147972608217,
"learning_rate": 3.254172015404365e-06,
"loss": 0.3555,
"step": 2610
},
{
"epoch": 0.8406192347798187,
"grad_norm": 3.610938014717987,
"learning_rate": 3.1899871630295255e-06,
"loss": 0.4082,
"step": 2620
},
{
"epoch": 0.8438277051415738,
"grad_norm": 5.165321090869274,
"learning_rate": 3.125802310654686e-06,
"loss": 0.4528,
"step": 2630
},
{
"epoch": 0.8470361755033288,
"grad_norm": 4.659717678208829,
"learning_rate": 3.061617458279846e-06,
"loss": 0.3743,
"step": 2640
},
{
"epoch": 0.8502446458650839,
"grad_norm": 3.9834909458830254,
"learning_rate": 2.9974326059050067e-06,
"loss": 0.4089,
"step": 2650
},
{
"epoch": 0.8534531162268388,
"grad_norm": 4.487091524427801,
"learning_rate": 2.9332477535301673e-06,
"loss": 0.4179,
"step": 2660
},
{
"epoch": 0.8566615865885939,
"grad_norm": 5.689083151350334,
"learning_rate": 2.869062901155328e-06,
"loss": 0.3458,
"step": 2670
},
{
"epoch": 0.8598700569503489,
"grad_norm": 4.958805228171173,
"learning_rate": 2.8048780487804884e-06,
"loss": 0.4503,
"step": 2680
},
{
"epoch": 0.863078527312104,
"grad_norm": 3.617994608871077,
"learning_rate": 2.740693196405648e-06,
"loss": 0.4462,
"step": 2690
},
{
"epoch": 0.866286997673859,
"grad_norm": 3.9580985359611245,
"learning_rate": 2.6765083440308086e-06,
"loss": 0.4561,
"step": 2700
},
{
"epoch": 0.8694954680356141,
"grad_norm": 3.863305244723279,
"learning_rate": 2.612323491655969e-06,
"loss": 0.38,
"step": 2710
},
{
"epoch": 0.872703938397369,
"grad_norm": 4.959804377866047,
"learning_rate": 2.5481386392811297e-06,
"loss": 0.3674,
"step": 2720
},
{
"epoch": 0.8759124087591241,
"grad_norm": 5.683599292818139,
"learning_rate": 2.4839537869062902e-06,
"loss": 0.433,
"step": 2730
},
{
"epoch": 0.8791208791208791,
"grad_norm": 5.107983778132738,
"learning_rate": 2.4197689345314508e-06,
"loss": 0.4292,
"step": 2740
},
{
"epoch": 0.8823293494826342,
"grad_norm": 4.173671184279824,
"learning_rate": 2.3555840821566113e-06,
"loss": 0.4009,
"step": 2750
},
{
"epoch": 0.8855378198443892,
"grad_norm": 4.235948185899856,
"learning_rate": 2.2913992297817714e-06,
"loss": 0.4275,
"step": 2760
},
{
"epoch": 0.8887462902061443,
"grad_norm": 5.541986982618706,
"learning_rate": 2.227214377406932e-06,
"loss": 0.4273,
"step": 2770
},
{
"epoch": 0.8919547605678992,
"grad_norm": 4.863277904592314,
"learning_rate": 2.1630295250320925e-06,
"loss": 0.4146,
"step": 2780
},
{
"epoch": 0.8951632309296543,
"grad_norm": 5.264055996210179,
"learning_rate": 2.098844672657253e-06,
"loss": 0.4107,
"step": 2790
},
{
"epoch": 0.8983717012914093,
"grad_norm": 3.825897290555289,
"learning_rate": 2.0346598202824136e-06,
"loss": 0.4036,
"step": 2800
},
{
"epoch": 0.9015801716531644,
"grad_norm": 4.922428436093768,
"learning_rate": 1.970474967907574e-06,
"loss": 0.3853,
"step": 2810
},
{
"epoch": 0.9047886420149194,
"grad_norm": 5.000879897373935,
"learning_rate": 1.9062901155327343e-06,
"loss": 0.4738,
"step": 2820
},
{
"epoch": 0.9079971123766745,
"grad_norm": 5.434069165890149,
"learning_rate": 1.8421052631578948e-06,
"loss": 0.452,
"step": 2830
},
{
"epoch": 0.9112055827384294,
"grad_norm": 4.954943186888116,
"learning_rate": 1.7779204107830554e-06,
"loss": 0.4303,
"step": 2840
},
{
"epoch": 0.9144140531001845,
"grad_norm": 4.274128582433497,
"learning_rate": 1.7137355584082157e-06,
"loss": 0.3942,
"step": 2850
},
{
"epoch": 0.9176225234619395,
"grad_norm": 5.733692944520431,
"learning_rate": 1.6495507060333762e-06,
"loss": 0.4164,
"step": 2860
},
{
"epoch": 0.9208309938236946,
"grad_norm": 5.336504400357715,
"learning_rate": 1.5853658536585368e-06,
"loss": 0.433,
"step": 2870
},
{
"epoch": 0.9240394641854496,
"grad_norm": 2.908475531628647,
"learning_rate": 1.5211810012836973e-06,
"loss": 0.4041,
"step": 2880
},
{
"epoch": 0.9272479345472047,
"grad_norm": 5.0009529548940295,
"learning_rate": 1.4569961489088577e-06,
"loss": 0.4364,
"step": 2890
},
{
"epoch": 0.9304564049089596,
"grad_norm": 4.956683680522817,
"learning_rate": 1.3928112965340182e-06,
"loss": 0.4521,
"step": 2900
},
{
"epoch": 0.9336648752707147,
"grad_norm": 4.462915922757512,
"learning_rate": 1.3286264441591783e-06,
"loss": 0.4406,
"step": 2910
},
{
"epoch": 0.9368733456324697,
"grad_norm": 5.190537589927558,
"learning_rate": 1.2644415917843389e-06,
"loss": 0.476,
"step": 2920
},
{
"epoch": 0.9400818159942248,
"grad_norm": 6.232818723236986,
"learning_rate": 1.2002567394094996e-06,
"loss": 0.3868,
"step": 2930
},
{
"epoch": 0.9432902863559798,
"grad_norm": 4.205260171634237,
"learning_rate": 1.13607188703466e-06,
"loss": 0.4728,
"step": 2940
},
{
"epoch": 0.9464987567177349,
"grad_norm": 3.671867784386402,
"learning_rate": 1.0718870346598203e-06,
"loss": 0.4779,
"step": 2950
},
{
"epoch": 0.9497072270794898,
"grad_norm": 3.758934283201711,
"learning_rate": 1.0077021822849808e-06,
"loss": 0.3564,
"step": 2960
},
{
"epoch": 0.9529156974412449,
"grad_norm": 4.27082338644512,
"learning_rate": 9.435173299101414e-07,
"loss": 0.3839,
"step": 2970
},
{
"epoch": 0.9561241678029999,
"grad_norm": 3.981941932574833,
"learning_rate": 8.793324775353017e-07,
"loss": 0.4346,
"step": 2980
},
{
"epoch": 0.959332638164755,
"grad_norm": 4.991690419408721,
"learning_rate": 8.151476251604621e-07,
"loss": 0.4237,
"step": 2990
},
{
"epoch": 0.96254110852651,
"grad_norm": 4.501522520126696,
"learning_rate": 7.509627727856227e-07,
"loss": 0.4185,
"step": 3000
},
{
"epoch": 0.9657495788882651,
"grad_norm": 5.223742371073297,
"learning_rate": 6.867779204107831e-07,
"loss": 0.3976,
"step": 3010
},
{
"epoch": 0.96895804925002,
"grad_norm": 7.427879323616558,
"learning_rate": 6.225930680359436e-07,
"loss": 0.4295,
"step": 3020
},
{
"epoch": 0.9721665196117751,
"grad_norm": 5.531122878697305,
"learning_rate": 5.584082156611041e-07,
"loss": 0.3633,
"step": 3030
},
{
"epoch": 0.9753749899735301,
"grad_norm": 3.900734513917918,
"learning_rate": 4.942233632862644e-07,
"loss": 0.435,
"step": 3040
},
{
"epoch": 0.9785834603352852,
"grad_norm": 5.534251744807489,
"learning_rate": 4.30038510911425e-07,
"loss": 0.4325,
"step": 3050
},
{
"epoch": 0.9817919306970402,
"grad_norm": 4.06665836174948,
"learning_rate": 3.6585365853658536e-07,
"loss": 0.3905,
"step": 3060
},
{
"epoch": 0.9850004010587953,
"grad_norm": 6.27706107511146,
"learning_rate": 3.0166880616174585e-07,
"loss": 0.4282,
"step": 3070
},
{
"epoch": 0.9882088714205502,
"grad_norm": 5.28526780087084,
"learning_rate": 2.374839537869063e-07,
"loss": 0.4614,
"step": 3080
},
{
"epoch": 0.9914173417823053,
"grad_norm": 4.5376678523257805,
"learning_rate": 1.7329910141206678e-07,
"loss": 0.4754,
"step": 3090
},
{
"epoch": 0.9946258121440603,
"grad_norm": 6.811875062252693,
"learning_rate": 1.0911424903722721e-07,
"loss": 0.4395,
"step": 3100
},
{
"epoch": 0.9978342825058154,
"grad_norm": 4.813868859792431,
"learning_rate": 4.4929396662387685e-08,
"loss": 0.3769,
"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
}