OpenR1_7B / trainer_state.json
AugustJW's picture
Add files using upload-large-folder tool
63e9bcf verified
Raw History Blame Contribute Delete
113 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9978976874562018,
"eval_steps": 500,
"global_step": 3210,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00467180565288484,
"grad_norm": 6.5849092047230275,
"learning_rate": 7.788161993769471e-07,
"loss": 0.8384,
"step": 5
},
{
"epoch": 0.00934361130576968,
"grad_norm": 4.61094411262779,
"learning_rate": 1.5576323987538942e-06,
"loss": 0.7947,
"step": 10
},
{
"epoch": 0.01401541695865452,
"grad_norm": 1.8435119819726173,
"learning_rate": 2.3364485981308413e-06,
"loss": 0.7214,
"step": 15
},
{
"epoch": 0.01868722261153936,
"grad_norm": 3.0021873815913644,
"learning_rate": 3.1152647975077884e-06,
"loss": 0.6818,
"step": 20
},
{
"epoch": 0.023359028264424198,
"grad_norm": 1.3458398248940706,
"learning_rate": 3.894080996884735e-06,
"loss": 0.6502,
"step": 25
},
{
"epoch": 0.02803083391730904,
"grad_norm": 0.8995789891488724,
"learning_rate": 4.6728971962616825e-06,
"loss": 0.5971,
"step": 30
},
{
"epoch": 0.03270263957019388,
"grad_norm": 0.7821045004164564,
"learning_rate": 5.451713395638629e-06,
"loss": 0.5851,
"step": 35
},
{
"epoch": 0.03737444522307872,
"grad_norm": 0.5507380709997592,
"learning_rate": 6.230529595015577e-06,
"loss": 0.5601,
"step": 40
},
{
"epoch": 0.04204625087596356,
"grad_norm": 0.5543300767479522,
"learning_rate": 7.009345794392523e-06,
"loss": 0.5459,
"step": 45
},
{
"epoch": 0.046718056528848396,
"grad_norm": 0.5494576085846301,
"learning_rate": 7.78816199376947e-06,
"loss": 0.5234,
"step": 50
},
{
"epoch": 0.05138986218173324,
"grad_norm": 0.6799687466591195,
"learning_rate": 8.566978193146417e-06,
"loss": 0.5267,
"step": 55
},
{
"epoch": 0.05606166783461808,
"grad_norm": 0.5224001156008252,
"learning_rate": 9.345794392523365e-06,
"loss": 0.5121,
"step": 60
},
{
"epoch": 0.06073347348750292,
"grad_norm": 0.6119075251302423,
"learning_rate": 1.0124610591900312e-05,
"loss": 0.5077,
"step": 65
},
{
"epoch": 0.06540527914038775,
"grad_norm": 0.538707841022484,
"learning_rate": 1.0903426791277258e-05,
"loss": 0.498,
"step": 70
},
{
"epoch": 0.0700770847932726,
"grad_norm": 0.5115435427700418,
"learning_rate": 1.1682242990654207e-05,
"loss": 0.5022,
"step": 75
},
{
"epoch": 0.07474889044615744,
"grad_norm": 0.4730973816090308,
"learning_rate": 1.2461059190031153e-05,
"loss": 0.5027,
"step": 80
},
{
"epoch": 0.07942069609904227,
"grad_norm": 0.6381076811686469,
"learning_rate": 1.3239875389408098e-05,
"loss": 0.4951,
"step": 85
},
{
"epoch": 0.08409250175192712,
"grad_norm": 0.47825027480292387,
"learning_rate": 1.4018691588785047e-05,
"loss": 0.4896,
"step": 90
},
{
"epoch": 0.08876430740481196,
"grad_norm": 0.5213800100146638,
"learning_rate": 1.4797507788161993e-05,
"loss": 0.4985,
"step": 95
},
{
"epoch": 0.09343611305769679,
"grad_norm": 0.49407391890354213,
"learning_rate": 1.557632398753894e-05,
"loss": 0.4824,
"step": 100
},
{
"epoch": 0.09810791871058164,
"grad_norm": 0.5365478505954182,
"learning_rate": 1.6355140186915887e-05,
"loss": 0.4953,
"step": 105
},
{
"epoch": 0.10277972436346648,
"grad_norm": 0.6078111407272674,
"learning_rate": 1.7133956386292833e-05,
"loss": 0.4834,
"step": 110
},
{
"epoch": 0.10745153001635133,
"grad_norm": 0.5195064753667656,
"learning_rate": 1.7912772585669783e-05,
"loss": 0.476,
"step": 115
},
{
"epoch": 0.11212333566923616,
"grad_norm": 0.6593547359372591,
"learning_rate": 1.869158878504673e-05,
"loss": 0.486,
"step": 120
},
{
"epoch": 0.116795141322121,
"grad_norm": 0.5674834512319046,
"learning_rate": 1.9470404984423677e-05,
"loss": 0.4716,
"step": 125
},
{
"epoch": 0.12146694697500585,
"grad_norm": 0.6754791789379834,
"learning_rate": 2.0249221183800623e-05,
"loss": 0.4793,
"step": 130
},
{
"epoch": 0.12613875262789068,
"grad_norm": 0.5386668487948113,
"learning_rate": 2.102803738317757e-05,
"loss": 0.4697,
"step": 135
},
{
"epoch": 0.1308105582807755,
"grad_norm": 0.7375849000862479,
"learning_rate": 2.1806853582554517e-05,
"loss": 0.4829,
"step": 140
},
{
"epoch": 0.13548236393366037,
"grad_norm": 0.6065483172084838,
"learning_rate": 2.2585669781931463e-05,
"loss": 0.4703,
"step": 145
},
{
"epoch": 0.1401541695865452,
"grad_norm": 0.6469842870299393,
"learning_rate": 2.3364485981308414e-05,
"loss": 0.4785,
"step": 150
},
{
"epoch": 0.14482597523943003,
"grad_norm": 0.6204616827515984,
"learning_rate": 2.414330218068536e-05,
"loss": 0.4705,
"step": 155
},
{
"epoch": 0.14949778089231489,
"grad_norm": 0.5681499667531437,
"learning_rate": 2.4922118380062307e-05,
"loss": 0.4672,
"step": 160
},
{
"epoch": 0.15416958654519972,
"grad_norm": 0.7024348214235372,
"learning_rate": 2.570093457943925e-05,
"loss": 0.4816,
"step": 165
},
{
"epoch": 0.15884139219808455,
"grad_norm": 0.5368975336292129,
"learning_rate": 2.6479750778816197e-05,
"loss": 0.4793,
"step": 170
},
{
"epoch": 0.1635131978509694,
"grad_norm": 0.6286411582851216,
"learning_rate": 2.7258566978193147e-05,
"loss": 0.4695,
"step": 175
},
{
"epoch": 0.16818500350385424,
"grad_norm": 0.5619692031588104,
"learning_rate": 2.8037383177570094e-05,
"loss": 0.4701,
"step": 180
},
{
"epoch": 0.17285680915673907,
"grad_norm": 0.5376059374373086,
"learning_rate": 2.881619937694704e-05,
"loss": 0.46,
"step": 185
},
{
"epoch": 0.17752861480962392,
"grad_norm": 0.5293184664587077,
"learning_rate": 2.9595015576323987e-05,
"loss": 0.4666,
"step": 190
},
{
"epoch": 0.18220042046250876,
"grad_norm": 0.538553970845718,
"learning_rate": 3.0373831775700934e-05,
"loss": 0.4693,
"step": 195
},
{
"epoch": 0.18687222611539359,
"grad_norm": 0.5754372071909925,
"learning_rate": 3.115264797507788e-05,
"loss": 0.4604,
"step": 200
},
{
"epoch": 0.19154403176827844,
"grad_norm": 1.5287938871222497,
"learning_rate": 3.193146417445483e-05,
"loss": 0.4675,
"step": 205
},
{
"epoch": 0.19621583742116327,
"grad_norm": 0.6867290552733082,
"learning_rate": 3.2710280373831774e-05,
"loss": 0.4639,
"step": 210
},
{
"epoch": 0.20088764307404813,
"grad_norm": 0.6570231144514197,
"learning_rate": 3.3489096573208724e-05,
"loss": 0.4652,
"step": 215
},
{
"epoch": 0.20555944872693296,
"grad_norm": 0.5825313840322547,
"learning_rate": 3.426791277258567e-05,
"loss": 0.4567,
"step": 220
},
{
"epoch": 0.2102312543798178,
"grad_norm": 0.5488208879435486,
"learning_rate": 3.504672897196262e-05,
"loss": 0.4548,
"step": 225
},
{
"epoch": 0.21490306003270265,
"grad_norm": 0.7284840010175765,
"learning_rate": 3.582554517133957e-05,
"loss": 0.4787,
"step": 230
},
{
"epoch": 0.21957486568558748,
"grad_norm": 0.5553400756934265,
"learning_rate": 3.660436137071651e-05,
"loss": 0.4687,
"step": 235
},
{
"epoch": 0.22424667133847231,
"grad_norm": 0.5927210460920687,
"learning_rate": 3.738317757009346e-05,
"loss": 0.4575,
"step": 240
},
{
"epoch": 0.22891847699135717,
"grad_norm": 0.5903305676459951,
"learning_rate": 3.8161993769470404e-05,
"loss": 0.4603,
"step": 245
},
{
"epoch": 0.233590282644242,
"grad_norm": 0.5150698254021607,
"learning_rate": 3.8940809968847354e-05,
"loss": 0.4659,
"step": 250
},
{
"epoch": 0.23826208829712683,
"grad_norm": 0.743620295674892,
"learning_rate": 3.97196261682243e-05,
"loss": 0.467,
"step": 255
},
{
"epoch": 0.2429338939500117,
"grad_norm": 0.42402540228407937,
"learning_rate": 4.049844236760125e-05,
"loss": 0.4624,
"step": 260
},
{
"epoch": 0.24760569960289652,
"grad_norm": 0.554982941861264,
"learning_rate": 4.12772585669782e-05,
"loss": 0.4713,
"step": 265
},
{
"epoch": 0.25227750525578135,
"grad_norm": 0.5497043901037699,
"learning_rate": 4.205607476635514e-05,
"loss": 0.4594,
"step": 270
},
{
"epoch": 0.2569493109086662,
"grad_norm": 0.5739585074447527,
"learning_rate": 4.283489096573209e-05,
"loss": 0.4738,
"step": 275
},
{
"epoch": 0.261621116561551,
"grad_norm": 0.5617901363811946,
"learning_rate": 4.3613707165109034e-05,
"loss": 0.4572,
"step": 280
},
{
"epoch": 0.2662929222144359,
"grad_norm": 0.7578271463984279,
"learning_rate": 4.4392523364485984e-05,
"loss": 0.459,
"step": 285
},
{
"epoch": 0.27096472786732073,
"grad_norm": 0.5529809237446733,
"learning_rate": 4.517133956386293e-05,
"loss": 0.4718,
"step": 290
},
{
"epoch": 0.27563653352020556,
"grad_norm": 0.6706441535453462,
"learning_rate": 4.595015576323988e-05,
"loss": 0.4766,
"step": 295
},
{
"epoch": 0.2803083391730904,
"grad_norm": 0.5378625100867989,
"learning_rate": 4.672897196261683e-05,
"loss": 0.4654,
"step": 300
},
{
"epoch": 0.2849801448259752,
"grad_norm": 0.46882980303541366,
"learning_rate": 4.750778816199377e-05,
"loss": 0.46,
"step": 305
},
{
"epoch": 0.28965195047886005,
"grad_norm": 0.5813716070993428,
"learning_rate": 4.828660436137072e-05,
"loss": 0.4708,
"step": 310
},
{
"epoch": 0.29432375613174494,
"grad_norm": 0.46127903996367153,
"learning_rate": 4.9065420560747664e-05,
"loss": 0.4533,
"step": 315
},
{
"epoch": 0.29899556178462977,
"grad_norm": 0.5808765073618293,
"learning_rate": 4.9844236760124614e-05,
"loss": 0.4685,
"step": 320
},
{
"epoch": 0.3036673674375146,
"grad_norm": 0.46270266996163867,
"learning_rate": 4.9930771893388716e-05,
"loss": 0.461,
"step": 325
},
{
"epoch": 0.30833917309039943,
"grad_norm": 0.5229103628074003,
"learning_rate": 4.9844236760124614e-05,
"loss": 0.4639,
"step": 330
},
{
"epoch": 0.31301097874328426,
"grad_norm": 0.6141983061786421,
"learning_rate": 4.975770162686051e-05,
"loss": 0.4691,
"step": 335
},
{
"epoch": 0.3176827843961691,
"grad_norm": 0.5125828484735886,
"learning_rate": 4.96711664935964e-05,
"loss": 0.4618,
"step": 340
},
{
"epoch": 0.322354590049054,
"grad_norm": 0.38513926523537817,
"learning_rate": 4.958463136033229e-05,
"loss": 0.4526,
"step": 345
},
{
"epoch": 0.3270263957019388,
"grad_norm": 0.5158037081526002,
"learning_rate": 4.949809622706819e-05,
"loss": 0.4748,
"step": 350
},
{
"epoch": 0.33169820135482364,
"grad_norm": 0.44580024906396737,
"learning_rate": 4.941156109380409e-05,
"loss": 0.4561,
"step": 355
},
{
"epoch": 0.33637000700770847,
"grad_norm": 0.47240803357103267,
"learning_rate": 4.9325025960539985e-05,
"loss": 0.4603,
"step": 360
},
{
"epoch": 0.3410418126605933,
"grad_norm": 0.4171975765665295,
"learning_rate": 4.9238490827275876e-05,
"loss": 0.4644,
"step": 365
},
{
"epoch": 0.34571361831347813,
"grad_norm": 0.5149246027911816,
"learning_rate": 4.915195569401177e-05,
"loss": 0.4599,
"step": 370
},
{
"epoch": 0.350385423966363,
"grad_norm": 0.44805158398160827,
"learning_rate": 4.9065420560747664e-05,
"loss": 0.4485,
"step": 375
},
{
"epoch": 0.35505722961924785,
"grad_norm": 0.46060651419384396,
"learning_rate": 4.897888542748356e-05,
"loss": 0.4755,
"step": 380
},
{
"epoch": 0.3597290352721327,
"grad_norm": 0.4101897284418394,
"learning_rate": 4.889235029421946e-05,
"loss": 0.4553,
"step": 385
},
{
"epoch": 0.3644008409250175,
"grad_norm": 0.4263502492874191,
"learning_rate": 4.880581516095535e-05,
"loss": 0.4613,
"step": 390
},
{
"epoch": 0.36907264657790234,
"grad_norm": 0.4384675170506129,
"learning_rate": 4.871928002769125e-05,
"loss": 0.4571,
"step": 395
},
{
"epoch": 0.37374445223078717,
"grad_norm": 0.5314449846878417,
"learning_rate": 4.863274489442714e-05,
"loss": 0.4505,
"step": 400
},
{
"epoch": 0.37841625788367206,
"grad_norm": 0.45885464617859106,
"learning_rate": 4.8546209761163035e-05,
"loss": 0.4581,
"step": 405
},
{
"epoch": 0.3830880635365569,
"grad_norm": 0.41981023264382344,
"learning_rate": 4.8459674627898926e-05,
"loss": 0.4619,
"step": 410
},
{
"epoch": 0.3877598691894417,
"grad_norm": 0.5001990411511336,
"learning_rate": 4.837313949463482e-05,
"loss": 0.4479,
"step": 415
},
{
"epoch": 0.39243167484232655,
"grad_norm": 0.47006354204867734,
"learning_rate": 4.828660436137072e-05,
"loss": 0.4588,
"step": 420
},
{
"epoch": 0.3971034804952114,
"grad_norm": 0.362931702172058,
"learning_rate": 4.820006922810662e-05,
"loss": 0.4524,
"step": 425
},
{
"epoch": 0.40177528614809627,
"grad_norm": 0.37965880434504107,
"learning_rate": 4.811353409484251e-05,
"loss": 0.4513,
"step": 430
},
{
"epoch": 0.4064470918009811,
"grad_norm": 0.3996713765953527,
"learning_rate": 4.80269989615784e-05,
"loss": 0.4496,
"step": 435
},
{
"epoch": 0.41111889745386593,
"grad_norm": 0.393265542392205,
"learning_rate": 4.79404638283143e-05,
"loss": 0.4454,
"step": 440
},
{
"epoch": 0.41579070310675076,
"grad_norm": 0.4197075446586915,
"learning_rate": 4.7853928695050194e-05,
"loss": 0.4529,
"step": 445
},
{
"epoch": 0.4204625087596356,
"grad_norm": 0.3912153941911287,
"learning_rate": 4.776739356178609e-05,
"loss": 0.4409,
"step": 450
},
{
"epoch": 0.4251343144125204,
"grad_norm": 0.5031957618744557,
"learning_rate": 4.768085842852198e-05,
"loss": 0.4581,
"step": 455
},
{
"epoch": 0.4298061200654053,
"grad_norm": 0.44049560387779857,
"learning_rate": 4.759432329525787e-05,
"loss": 0.4448,
"step": 460
},
{
"epoch": 0.43447792571829014,
"grad_norm": 0.3876310557886793,
"learning_rate": 4.750778816199377e-05,
"loss": 0.4465,
"step": 465
},
{
"epoch": 0.43914973137117497,
"grad_norm": 0.4096764260352836,
"learning_rate": 4.742125302872967e-05,
"loss": 0.452,
"step": 470
},
{
"epoch": 0.4438215370240598,
"grad_norm": 0.46488136608448666,
"learning_rate": 4.7334717895465565e-05,
"loss": 0.4567,
"step": 475
},
{
"epoch": 0.44849334267694463,
"grad_norm": 0.38268740712661503,
"learning_rate": 4.7248182762201456e-05,
"loss": 0.4513,
"step": 480
},
{
"epoch": 0.45316514832982946,
"grad_norm": 0.4088767439205309,
"learning_rate": 4.7161647628937347e-05,
"loss": 0.4539,
"step": 485
},
{
"epoch": 0.45783695398271435,
"grad_norm": 0.37909015395074475,
"learning_rate": 4.7075112495673244e-05,
"loss": 0.4517,
"step": 490
},
{
"epoch": 0.4625087596355992,
"grad_norm": 0.3869805606590519,
"learning_rate": 4.698857736240914e-05,
"loss": 0.4453,
"step": 495
},
{
"epoch": 0.467180565288484,
"grad_norm": 0.39533855079014657,
"learning_rate": 4.690204222914504e-05,
"loss": 0.4473,
"step": 500
},
{
"epoch": 0.47185237094136884,
"grad_norm": 0.38986872211354673,
"learning_rate": 4.681550709588093e-05,
"loss": 0.4326,
"step": 505
},
{
"epoch": 0.47652417659425367,
"grad_norm": 0.41457009436391673,
"learning_rate": 4.672897196261683e-05,
"loss": 0.4348,
"step": 510
},
{
"epoch": 0.4811959822471385,
"grad_norm": 0.37668853659830476,
"learning_rate": 4.664243682935272e-05,
"loss": 0.4479,
"step": 515
},
{
"epoch": 0.4858677879000234,
"grad_norm": 0.3988473145902496,
"learning_rate": 4.6555901696088615e-05,
"loss": 0.4498,
"step": 520
},
{
"epoch": 0.4905395935529082,
"grad_norm": 0.40575686742656814,
"learning_rate": 4.6469366562824506e-05,
"loss": 0.4276,
"step": 525
},
{
"epoch": 0.49521139920579305,
"grad_norm": 0.3988250386349381,
"learning_rate": 4.63828314295604e-05,
"loss": 0.4543,
"step": 530
},
{
"epoch": 0.4998832048586779,
"grad_norm": 0.34185207238603266,
"learning_rate": 4.62962962962963e-05,
"loss": 0.4503,
"step": 535
},
{
"epoch": 0.5045550105115627,
"grad_norm": 0.3569274370144077,
"learning_rate": 4.620976116303219e-05,
"loss": 0.4552,
"step": 540
},
{
"epoch": 0.5092268161644475,
"grad_norm": 0.3782018081919045,
"learning_rate": 4.612322602976809e-05,
"loss": 0.4451,
"step": 545
},
{
"epoch": 0.5138986218173324,
"grad_norm": 0.4702354991539984,
"learning_rate": 4.603669089650398e-05,
"loss": 0.4421,
"step": 550
},
{
"epoch": 0.5185704274702172,
"grad_norm": 0.43688198880703416,
"learning_rate": 4.595015576323988e-05,
"loss": 0.4432,
"step": 555
},
{
"epoch": 0.523242233123102,
"grad_norm": 0.4651957429166244,
"learning_rate": 4.5863620629975774e-05,
"loss": 0.4443,
"step": 560
},
{
"epoch": 0.527914038775987,
"grad_norm": 0.3721510377381532,
"learning_rate": 4.577708549671167e-05,
"loss": 0.4388,
"step": 565
},
{
"epoch": 0.5325858444288718,
"grad_norm": 0.3533864862523594,
"learning_rate": 4.569055036344756e-05,
"loss": 0.4424,
"step": 570
},
{
"epoch": 0.5372576500817566,
"grad_norm": 0.4051395186816033,
"learning_rate": 4.560401523018345e-05,
"loss": 0.4364,
"step": 575
},
{
"epoch": 0.5419294557346415,
"grad_norm": 0.30845296655300736,
"learning_rate": 4.551748009691935e-05,
"loss": 0.4385,
"step": 580
},
{
"epoch": 0.5466012613875263,
"grad_norm": 0.3581045162580383,
"learning_rate": 4.543094496365525e-05,
"loss": 0.4378,
"step": 585
},
{
"epoch": 0.5512730670404111,
"grad_norm": 0.32580334366304414,
"learning_rate": 4.5344409830391146e-05,
"loss": 0.4481,
"step": 590
},
{
"epoch": 0.555944872693296,
"grad_norm": 0.37673263846977084,
"learning_rate": 4.5257874697127036e-05,
"loss": 0.4361,
"step": 595
},
{
"epoch": 0.5606166783461808,
"grad_norm": 0.31217931672338445,
"learning_rate": 4.517133956386293e-05,
"loss": 0.4372,
"step": 600
},
{
"epoch": 0.5652884839990656,
"grad_norm": 0.36428398552848057,
"learning_rate": 4.5084804430598824e-05,
"loss": 0.4301,
"step": 605
},
{
"epoch": 0.5699602896519504,
"grad_norm": 0.393458134011921,
"learning_rate": 4.499826929733472e-05,
"loss": 0.4482,
"step": 610
},
{
"epoch": 0.5746320953048353,
"grad_norm": 0.3604036012996075,
"learning_rate": 4.491173416407061e-05,
"loss": 0.4313,
"step": 615
},
{
"epoch": 0.5793039009577201,
"grad_norm": 0.39272710261380284,
"learning_rate": 4.482519903080651e-05,
"loss": 0.4443,
"step": 620
},
{
"epoch": 0.583975706610605,
"grad_norm": 0.32017020538872754,
"learning_rate": 4.473866389754241e-05,
"loss": 0.4514,
"step": 625
},
{
"epoch": 0.5886475122634899,
"grad_norm": 0.3925330000488089,
"learning_rate": 4.46521287642783e-05,
"loss": 0.4417,
"step": 630
},
{
"epoch": 0.5933193179163747,
"grad_norm": 0.41970008115691587,
"learning_rate": 4.4565593631014196e-05,
"loss": 0.4419,
"step": 635
},
{
"epoch": 0.5979911235692595,
"grad_norm": 0.36267140296933204,
"learning_rate": 4.4479058497750086e-05,
"loss": 0.4377,
"step": 640
},
{
"epoch": 0.6026629292221444,
"grad_norm": 0.3591178098528886,
"learning_rate": 4.4392523364485984e-05,
"loss": 0.438,
"step": 645
},
{
"epoch": 0.6073347348750292,
"grad_norm": 0.3797895569637972,
"learning_rate": 4.430598823122188e-05,
"loss": 0.4328,
"step": 650
},
{
"epoch": 0.612006540527914,
"grad_norm": 0.3936860116539046,
"learning_rate": 4.421945309795777e-05,
"loss": 0.4526,
"step": 655
},
{
"epoch": 0.6166783461807989,
"grad_norm": 0.328980909424925,
"learning_rate": 4.413291796469366e-05,
"loss": 0.4296,
"step": 660
},
{
"epoch": 0.6213501518336837,
"grad_norm": 0.3266600351726345,
"learning_rate": 4.404638283142956e-05,
"loss": 0.4385,
"step": 665
},
{
"epoch": 0.6260219574865685,
"grad_norm": 0.3657058376931918,
"learning_rate": 4.395984769816546e-05,
"loss": 0.4445,
"step": 670
},
{
"epoch": 0.6306937631394534,
"grad_norm": 0.3356182767817925,
"learning_rate": 4.3873312564901355e-05,
"loss": 0.437,
"step": 675
},
{
"epoch": 0.6353655687923382,
"grad_norm": 0.38655573687506406,
"learning_rate": 4.378677743163725e-05,
"loss": 0.4401,
"step": 680
},
{
"epoch": 0.6400373744452231,
"grad_norm": 0.39008662058137117,
"learning_rate": 4.370024229837314e-05,
"loss": 0.4421,
"step": 685
},
{
"epoch": 0.644709180098108,
"grad_norm": 0.3412974810284902,
"learning_rate": 4.3613707165109034e-05,
"loss": 0.4373,
"step": 690
},
{
"epoch": 0.6493809857509928,
"grad_norm": 0.3401983769863861,
"learning_rate": 4.352717203184493e-05,
"loss": 0.4299,
"step": 695
},
{
"epoch": 0.6540527914038776,
"grad_norm": 0.30501297701259034,
"learning_rate": 4.344063689858083e-05,
"loss": 0.4342,
"step": 700
},
{
"epoch": 0.6587245970567625,
"grad_norm": 0.31484071618032955,
"learning_rate": 4.335410176531672e-05,
"loss": 0.4349,
"step": 705
},
{
"epoch": 0.6633964027096473,
"grad_norm": 0.338094623501772,
"learning_rate": 4.326756663205262e-05,
"loss": 0.4299,
"step": 710
},
{
"epoch": 0.6680682083625321,
"grad_norm": 0.3458704281920345,
"learning_rate": 4.318103149878851e-05,
"loss": 0.4338,
"step": 715
},
{
"epoch": 0.6727400140154169,
"grad_norm": 0.36544157760715257,
"learning_rate": 4.3094496365524405e-05,
"loss": 0.4386,
"step": 720
},
{
"epoch": 0.6774118196683018,
"grad_norm": 0.3913395631524822,
"learning_rate": 4.30079612322603e-05,
"loss": 0.4311,
"step": 725
},
{
"epoch": 0.6820836253211866,
"grad_norm": 0.44864642786524195,
"learning_rate": 4.292142609899619e-05,
"loss": 0.4402,
"step": 730
},
{
"epoch": 0.6867554309740714,
"grad_norm": 0.36192171738267825,
"learning_rate": 4.283489096573209e-05,
"loss": 0.4335,
"step": 735
},
{
"epoch": 0.6914272366269563,
"grad_norm": 0.3298272839211148,
"learning_rate": 4.274835583246799e-05,
"loss": 0.4339,
"step": 740
},
{
"epoch": 0.6960990422798412,
"grad_norm": 0.32921748584135474,
"learning_rate": 4.266182069920388e-05,
"loss": 0.4276,
"step": 745
},
{
"epoch": 0.700770847932726,
"grad_norm": 0.3826068055078561,
"learning_rate": 4.257528556593977e-05,
"loss": 0.4426,
"step": 750
},
{
"epoch": 0.7054426535856109,
"grad_norm": 0.4061362529881364,
"learning_rate": 4.2488750432675667e-05,
"loss": 0.4352,
"step": 755
},
{
"epoch": 0.7101144592384957,
"grad_norm": 0.3267019849696966,
"learning_rate": 4.2402215299411564e-05,
"loss": 0.4288,
"step": 760
},
{
"epoch": 0.7147862648913805,
"grad_norm": 0.37023905138181357,
"learning_rate": 4.231568016614746e-05,
"loss": 0.4321,
"step": 765
},
{
"epoch": 0.7194580705442654,
"grad_norm": 0.3904065707687146,
"learning_rate": 4.222914503288335e-05,
"loss": 0.4372,
"step": 770
},
{
"epoch": 0.7241298761971502,
"grad_norm": 0.35054219903206124,
"learning_rate": 4.214260989961924e-05,
"loss": 0.4262,
"step": 775
},
{
"epoch": 0.728801681850035,
"grad_norm": 0.36278150073448456,
"learning_rate": 4.205607476635514e-05,
"loss": 0.433,
"step": 780
},
{
"epoch": 0.7334734875029199,
"grad_norm": 0.30073849266080294,
"learning_rate": 4.196953963309104e-05,
"loss": 0.4335,
"step": 785
},
{
"epoch": 0.7381452931558047,
"grad_norm": 0.39101708168281585,
"learning_rate": 4.1883004499826935e-05,
"loss": 0.4252,
"step": 790
},
{
"epoch": 0.7428170988086895,
"grad_norm": 0.352387205651067,
"learning_rate": 4.1796469366562826e-05,
"loss": 0.4283,
"step": 795
},
{
"epoch": 0.7474889044615743,
"grad_norm": 0.3883081439670353,
"learning_rate": 4.1709934233298717e-05,
"loss": 0.4279,
"step": 800
},
{
"epoch": 0.7521607101144593,
"grad_norm": 0.317076912997008,
"learning_rate": 4.1623399100034614e-05,
"loss": 0.4453,
"step": 805
},
{
"epoch": 0.7568325157673441,
"grad_norm": 0.3403256109859331,
"learning_rate": 4.153686396677051e-05,
"loss": 0.4392,
"step": 810
},
{
"epoch": 0.761504321420229,
"grad_norm": 0.31666313630482246,
"learning_rate": 4.145032883350641e-05,
"loss": 0.4294,
"step": 815
},
{
"epoch": 0.7661761270731138,
"grad_norm": 0.31206161146063055,
"learning_rate": 4.13637937002423e-05,
"loss": 0.4232,
"step": 820
},
{
"epoch": 0.7708479327259986,
"grad_norm": 0.2796024701694955,
"learning_rate": 4.12772585669782e-05,
"loss": 0.4262,
"step": 825
},
{
"epoch": 0.7755197383788834,
"grad_norm": 0.293380246324644,
"learning_rate": 4.119072343371409e-05,
"loss": 0.4242,
"step": 830
},
{
"epoch": 0.7801915440317683,
"grad_norm": 0.30385440250734963,
"learning_rate": 4.1104188300449985e-05,
"loss": 0.4274,
"step": 835
},
{
"epoch": 0.7848633496846531,
"grad_norm": 0.2826958696397432,
"learning_rate": 4.101765316718588e-05,
"loss": 0.4185,
"step": 840
},
{
"epoch": 0.7895351553375379,
"grad_norm": 0.35518351750249083,
"learning_rate": 4.093111803392177e-05,
"loss": 0.4294,
"step": 845
},
{
"epoch": 0.7942069609904228,
"grad_norm": 0.3372564145059291,
"learning_rate": 4.084458290065767e-05,
"loss": 0.4311,
"step": 850
},
{
"epoch": 0.7988787666433076,
"grad_norm": 0.3766906247548918,
"learning_rate": 4.075804776739356e-05,
"loss": 0.4429,
"step": 855
},
{
"epoch": 0.8035505722961925,
"grad_norm": 0.38315057143223463,
"learning_rate": 4.067151263412946e-05,
"loss": 0.409,
"step": 860
},
{
"epoch": 0.8082223779490774,
"grad_norm": 0.3593790290883937,
"learning_rate": 4.058497750086535e-05,
"loss": 0.4168,
"step": 865
},
{
"epoch": 0.8128941836019622,
"grad_norm": 0.33629163253901395,
"learning_rate": 4.049844236760125e-05,
"loss": 0.4148,
"step": 870
},
{
"epoch": 0.817565989254847,
"grad_norm": 0.37970785823022557,
"learning_rate": 4.0411907234337144e-05,
"loss": 0.4308,
"step": 875
},
{
"epoch": 0.8222377949077319,
"grad_norm": 0.3529519353897875,
"learning_rate": 4.032537210107304e-05,
"loss": 0.4281,
"step": 880
},
{
"epoch": 0.8269096005606167,
"grad_norm": 0.3511500676456135,
"learning_rate": 4.023883696780893e-05,
"loss": 0.41,
"step": 885
},
{
"epoch": 0.8315814062135015,
"grad_norm": 0.2878730551656308,
"learning_rate": 4.015230183454482e-05,
"loss": 0.4243,
"step": 890
},
{
"epoch": 0.8362532118663863,
"grad_norm": 0.30618370050108856,
"learning_rate": 4.006576670128072e-05,
"loss": 0.4212,
"step": 895
},
{
"epoch": 0.8409250175192712,
"grad_norm": 0.33368720068535684,
"learning_rate": 3.997923156801662e-05,
"loss": 0.4225,
"step": 900
},
{
"epoch": 0.845596823172156,
"grad_norm": 0.3242524367445163,
"learning_rate": 3.9892696434752516e-05,
"loss": 0.4206,
"step": 905
},
{
"epoch": 0.8502686288250408,
"grad_norm": 0.2892732599520401,
"learning_rate": 3.9806161301488406e-05,
"loss": 0.4249,
"step": 910
},
{
"epoch": 0.8549404344779257,
"grad_norm": 0.3630657163125305,
"learning_rate": 3.97196261682243e-05,
"loss": 0.4275,
"step": 915
},
{
"epoch": 0.8596122401308106,
"grad_norm": 0.30155530421010024,
"learning_rate": 3.9633091034960194e-05,
"loss": 0.4302,
"step": 920
},
{
"epoch": 0.8642840457836954,
"grad_norm": 0.28276896360183135,
"learning_rate": 3.954655590169609e-05,
"loss": 0.4342,
"step": 925
},
{
"epoch": 0.8689558514365803,
"grad_norm": 0.32456073772439276,
"learning_rate": 3.946002076843199e-05,
"loss": 0.4308,
"step": 930
},
{
"epoch": 0.8736276570894651,
"grad_norm": 0.298935868826097,
"learning_rate": 3.937348563516788e-05,
"loss": 0.416,
"step": 935
},
{
"epoch": 0.8782994627423499,
"grad_norm": 0.3012094323303976,
"learning_rate": 3.928695050190378e-05,
"loss": 0.4139,
"step": 940
},
{
"epoch": 0.8829712683952348,
"grad_norm": 0.29411119167224814,
"learning_rate": 3.920041536863967e-05,
"loss": 0.4188,
"step": 945
},
{
"epoch": 0.8876430740481196,
"grad_norm": 0.30075680097083646,
"learning_rate": 3.9113880235375566e-05,
"loss": 0.4212,
"step": 950
},
{
"epoch": 0.8923148797010044,
"grad_norm": 0.3192580883534557,
"learning_rate": 3.9027345102111456e-05,
"loss": 0.437,
"step": 955
},
{
"epoch": 0.8969866853538893,
"grad_norm": 0.32142737819682643,
"learning_rate": 3.8940809968847354e-05,
"loss": 0.4229,
"step": 960
},
{
"epoch": 0.9016584910067741,
"grad_norm": 0.2954867192111965,
"learning_rate": 3.885427483558325e-05,
"loss": 0.4236,
"step": 965
},
{
"epoch": 0.9063302966596589,
"grad_norm": 0.2665538821089102,
"learning_rate": 3.876773970231914e-05,
"loss": 0.4237,
"step": 970
},
{
"epoch": 0.9110021023125437,
"grad_norm": 0.29473478992397095,
"learning_rate": 3.868120456905504e-05,
"loss": 0.424,
"step": 975
},
{
"epoch": 0.9156739079654287,
"grad_norm": 0.2873899228570041,
"learning_rate": 3.859466943579093e-05,
"loss": 0.4242,
"step": 980
},
{
"epoch": 0.9203457136183135,
"grad_norm": 0.33024714993653614,
"learning_rate": 3.850813430252683e-05,
"loss": 0.4254,
"step": 985
},
{
"epoch": 0.9250175192711984,
"grad_norm": 0.2986837175483655,
"learning_rate": 3.8421599169262725e-05,
"loss": 0.4235,
"step": 990
},
{
"epoch": 0.9296893249240832,
"grad_norm": 0.29001116919390674,
"learning_rate": 3.833506403599862e-05,
"loss": 0.4172,
"step": 995
},
{
"epoch": 0.934361130576968,
"grad_norm": 0.30342924412755984,
"learning_rate": 3.8248528902734506e-05,
"loss": 0.4264,
"step": 1000
},
{
"epoch": 0.9390329362298528,
"grad_norm": 0.29841361641604935,
"learning_rate": 3.8161993769470404e-05,
"loss": 0.4196,
"step": 1005
},
{
"epoch": 0.9437047418827377,
"grad_norm": 0.3550815316731992,
"learning_rate": 3.80754586362063e-05,
"loss": 0.4272,
"step": 1010
},
{
"epoch": 0.9483765475356225,
"grad_norm": 0.33701156045530595,
"learning_rate": 3.79889235029422e-05,
"loss": 0.412,
"step": 1015
},
{
"epoch": 0.9530483531885073,
"grad_norm": 0.35147300124616626,
"learning_rate": 3.7902388369678096e-05,
"loss": 0.4261,
"step": 1020
},
{
"epoch": 0.9577201588413922,
"grad_norm": 0.29529143928010876,
"learning_rate": 3.7815853236413987e-05,
"loss": 0.4138,
"step": 1025
},
{
"epoch": 0.962391964494277,
"grad_norm": 0.3568104417258214,
"learning_rate": 3.772931810314988e-05,
"loss": 0.4308,
"step": 1030
},
{
"epoch": 0.9670637701471618,
"grad_norm": 0.3957278097342003,
"learning_rate": 3.7642782969885775e-05,
"loss": 0.4156,
"step": 1035
},
{
"epoch": 0.9717355758000468,
"grad_norm": 0.29913222381506793,
"learning_rate": 3.755624783662167e-05,
"loss": 0.4236,
"step": 1040
},
{
"epoch": 0.9764073814529316,
"grad_norm": 0.3543369604976565,
"learning_rate": 3.746971270335756e-05,
"loss": 0.4249,
"step": 1045
},
{
"epoch": 0.9810791871058164,
"grad_norm": 0.31732830477378066,
"learning_rate": 3.738317757009346e-05,
"loss": 0.4253,
"step": 1050
},
{
"epoch": 0.9857509927587013,
"grad_norm": 0.27444479513957404,
"learning_rate": 3.729664243682935e-05,
"loss": 0.4232,
"step": 1055
},
{
"epoch": 0.9904227984115861,
"grad_norm": 0.3496931540277006,
"learning_rate": 3.721010730356525e-05,
"loss": 0.4273,
"step": 1060
},
{
"epoch": 0.9950946040644709,
"grad_norm": 0.4065537348528585,
"learning_rate": 3.7123572170301146e-05,
"loss": 0.4361,
"step": 1065
},
{
"epoch": 0.9997664097173558,
"grad_norm": 0.38000713489549176,
"learning_rate": 3.7037037037037037e-05,
"loss": 0.4256,
"step": 1070
},
{
"epoch": 1.003737444522308,
"grad_norm": 0.47113206327807905,
"learning_rate": 3.6950501903772934e-05,
"loss": 0.3781,
"step": 1075
},
{
"epoch": 1.0084092501751927,
"grad_norm": 0.3583857705393172,
"learning_rate": 3.686396677050883e-05,
"loss": 0.3549,
"step": 1080
},
{
"epoch": 1.0130810558280776,
"grad_norm": 8.208246607647503,
"learning_rate": 3.677743163724472e-05,
"loss": 0.3631,
"step": 1085
},
{
"epoch": 1.0177528614809623,
"grad_norm": 0.5464148021876812,
"learning_rate": 3.669089650398061e-05,
"loss": 0.371,
"step": 1090
},
{
"epoch": 1.0224246671338473,
"grad_norm": 0.3431339974687547,
"learning_rate": 3.660436137071651e-05,
"loss": 0.3726,
"step": 1095
},
{
"epoch": 1.027096472786732,
"grad_norm": 0.7801246111914149,
"learning_rate": 3.651782623745241e-05,
"loss": 0.3726,
"step": 1100
},
{
"epoch": 1.031768278439617,
"grad_norm": 0.7682393993244626,
"learning_rate": 3.6431291104188305e-05,
"loss": 0.3753,
"step": 1105
},
{
"epoch": 1.0364400840925017,
"grad_norm": 0.6452333277207267,
"learning_rate": 3.6344755970924196e-05,
"loss": 0.372,
"step": 1110
},
{
"epoch": 1.0411118897453866,
"grad_norm": 0.6406211466626097,
"learning_rate": 3.6258220837660086e-05,
"loss": 0.3605,
"step": 1115
},
{
"epoch": 1.0457836953982715,
"grad_norm": 0.4824710254510864,
"learning_rate": 3.6171685704395984e-05,
"loss": 0.3647,
"step": 1120
},
{
"epoch": 1.0504555010511563,
"grad_norm": 0.41693864235522893,
"learning_rate": 3.608515057113188e-05,
"loss": 0.379,
"step": 1125
},
{
"epoch": 1.0551273067040412,
"grad_norm": 1.581478150681963,
"learning_rate": 3.599861543786778e-05,
"loss": 0.3611,
"step": 1130
},
{
"epoch": 1.059799112356926,
"grad_norm": 0.390189873063051,
"learning_rate": 3.5912080304603676e-05,
"loss": 0.3606,
"step": 1135
},
{
"epoch": 1.0644709180098109,
"grad_norm": 0.29185239276285907,
"learning_rate": 3.582554517133957e-05,
"loss": 0.3614,
"step": 1140
},
{
"epoch": 1.0691427236626956,
"grad_norm": 0.3234375190448062,
"learning_rate": 3.573901003807546e-05,
"loss": 0.3549,
"step": 1145
},
{
"epoch": 1.0738145293155805,
"grad_norm": 0.40185546630002694,
"learning_rate": 3.5652474904811355e-05,
"loss": 0.3657,
"step": 1150
},
{
"epoch": 1.0784863349684652,
"grad_norm": 0.7160969849632728,
"learning_rate": 3.556593977154725e-05,
"loss": 0.3612,
"step": 1155
},
{
"epoch": 1.0831581406213502,
"grad_norm": 0.301962435279091,
"learning_rate": 3.547940463828314e-05,
"loss": 0.3651,
"step": 1160
},
{
"epoch": 1.087829946274235,
"grad_norm": 0.3032042877218907,
"learning_rate": 3.539286950501904e-05,
"loss": 0.3652,
"step": 1165
},
{
"epoch": 1.0925017519271198,
"grad_norm": 0.28212570851757335,
"learning_rate": 3.530633437175493e-05,
"loss": 0.3549,
"step": 1170
},
{
"epoch": 1.0971735575800046,
"grad_norm": 0.2747195601219735,
"learning_rate": 3.521979923849083e-05,
"loss": 0.3523,
"step": 1175
},
{
"epoch": 1.1018453632328895,
"grad_norm": 0.31166214851034807,
"learning_rate": 3.5133264105226726e-05,
"loss": 0.3594,
"step": 1180
},
{
"epoch": 1.1065171688857744,
"grad_norm": 0.2849377003354121,
"learning_rate": 3.504672897196262e-05,
"loss": 0.3616,
"step": 1185
},
{
"epoch": 1.1111889745386592,
"grad_norm": 0.3034186203276653,
"learning_rate": 3.4960193838698514e-05,
"loss": 0.3696,
"step": 1190
},
{
"epoch": 1.115860780191544,
"grad_norm": 0.28010023757179475,
"learning_rate": 3.487365870543441e-05,
"loss": 0.3576,
"step": 1195
},
{
"epoch": 1.1205325858444288,
"grad_norm": 0.28235267927921287,
"learning_rate": 3.47871235721703e-05,
"loss": 0.3589,
"step": 1200
},
{
"epoch": 1.1252043914973138,
"grad_norm": 0.29128959653541914,
"learning_rate": 3.470058843890619e-05,
"loss": 0.3625,
"step": 1205
},
{
"epoch": 1.1298761971501985,
"grad_norm": 0.2863854836440615,
"learning_rate": 3.461405330564209e-05,
"loss": 0.3595,
"step": 1210
},
{
"epoch": 1.1345480028030834,
"grad_norm": 0.2649166931642877,
"learning_rate": 3.452751817237799e-05,
"loss": 0.3588,
"step": 1215
},
{
"epoch": 1.1392198084559682,
"grad_norm": 0.3124948722714658,
"learning_rate": 3.4440983039113886e-05,
"loss": 0.3681,
"step": 1220
},
{
"epoch": 1.143891614108853,
"grad_norm": 0.2896157022013514,
"learning_rate": 3.4354447905849776e-05,
"loss": 0.3613,
"step": 1225
},
{
"epoch": 1.148563419761738,
"grad_norm": 0.2850596388351964,
"learning_rate": 3.426791277258567e-05,
"loss": 0.3529,
"step": 1230
},
{
"epoch": 1.1532352254146228,
"grad_norm": 0.32217666676805523,
"learning_rate": 3.4181377639321564e-05,
"loss": 0.3731,
"step": 1235
},
{
"epoch": 1.1579070310675075,
"grad_norm": 0.28281051296974585,
"learning_rate": 3.409484250605746e-05,
"loss": 0.3444,
"step": 1240
},
{
"epoch": 1.1625788367203924,
"grad_norm": 0.26754822564510194,
"learning_rate": 3.400830737279336e-05,
"loss": 0.3606,
"step": 1245
},
{
"epoch": 1.1672506423732774,
"grad_norm": 0.29739428173086213,
"learning_rate": 3.392177223952925e-05,
"loss": 0.375,
"step": 1250
},
{
"epoch": 1.171922448026162,
"grad_norm": 0.29506048163057214,
"learning_rate": 3.383523710626515e-05,
"loss": 0.3567,
"step": 1255
},
{
"epoch": 1.176594253679047,
"grad_norm": 0.28749319429900416,
"learning_rate": 3.374870197300104e-05,
"loss": 0.3575,
"step": 1260
},
{
"epoch": 1.1812660593319317,
"grad_norm": 0.363817148475934,
"learning_rate": 3.3662166839736935e-05,
"loss": 0.3559,
"step": 1265
},
{
"epoch": 1.1859378649848167,
"grad_norm": 1.7945918554412712,
"learning_rate": 3.357563170647283e-05,
"loss": 0.3629,
"step": 1270
},
{
"epoch": 1.1906096706377014,
"grad_norm": 0.38651927486811116,
"learning_rate": 3.3489096573208724e-05,
"loss": 0.3614,
"step": 1275
},
{
"epoch": 1.1952814762905863,
"grad_norm": 0.2780824349278925,
"learning_rate": 3.340256143994462e-05,
"loss": 0.3664,
"step": 1280
},
{
"epoch": 1.199953281943471,
"grad_norm": 0.2783879841037776,
"learning_rate": 3.331602630668051e-05,
"loss": 0.3575,
"step": 1285
},
{
"epoch": 1.204625087596356,
"grad_norm": 0.27381572421857153,
"learning_rate": 3.322949117341641e-05,
"loss": 0.3561,
"step": 1290
},
{
"epoch": 1.209296893249241,
"grad_norm": 0.6991633413391471,
"learning_rate": 3.31429560401523e-05,
"loss": 0.3642,
"step": 1295
},
{
"epoch": 1.2139686989021257,
"grad_norm": 0.2620850427571531,
"learning_rate": 3.30564209068882e-05,
"loss": 0.3595,
"step": 1300
},
{
"epoch": 1.2186405045550106,
"grad_norm": 0.270994718744177,
"learning_rate": 3.2969885773624095e-05,
"loss": 0.3532,
"step": 1305
},
{
"epoch": 1.2233123102078953,
"grad_norm": 0.2748234539091387,
"learning_rate": 3.288335064035999e-05,
"loss": 0.3545,
"step": 1310
},
{
"epoch": 1.2279841158607803,
"grad_norm": 0.2866745177314814,
"learning_rate": 3.279681550709588e-05,
"loss": 0.3577,
"step": 1315
},
{
"epoch": 1.232655921513665,
"grad_norm": 0.2861293554491327,
"learning_rate": 3.2710280373831774e-05,
"loss": 0.358,
"step": 1320
},
{
"epoch": 1.23732772716655,
"grad_norm": 0.3372700808402862,
"learning_rate": 3.262374524056767e-05,
"loss": 0.3681,
"step": 1325
},
{
"epoch": 1.2419995328194346,
"grad_norm": 0.30762866125096133,
"learning_rate": 3.253721010730357e-05,
"loss": 0.3666,
"step": 1330
},
{
"epoch": 1.2466713384723196,
"grad_norm": 0.28639927233745277,
"learning_rate": 3.2450674974039466e-05,
"loss": 0.3554,
"step": 1335
},
{
"epoch": 1.2513431441252043,
"grad_norm": 0.275340965232692,
"learning_rate": 3.2364139840775357e-05,
"loss": 0.3524,
"step": 1340
},
{
"epoch": 1.2560149497780893,
"grad_norm": 0.267567119630321,
"learning_rate": 3.227760470751125e-05,
"loss": 0.3687,
"step": 1345
},
{
"epoch": 1.260686755430974,
"grad_norm": 0.29332575713838344,
"learning_rate": 3.2191069574247145e-05,
"loss": 0.3571,
"step": 1350
},
{
"epoch": 1.265358561083859,
"grad_norm": 0.2525555599214365,
"learning_rate": 3.210453444098304e-05,
"loss": 0.3568,
"step": 1355
},
{
"epoch": 1.2700303667367439,
"grad_norm": 0.3528635048379817,
"learning_rate": 3.201799930771894e-05,
"loss": 0.3534,
"step": 1360
},
{
"epoch": 1.2747021723896286,
"grad_norm": 0.3239898991058169,
"learning_rate": 3.193146417445483e-05,
"loss": 0.357,
"step": 1365
},
{
"epoch": 1.2793739780425133,
"grad_norm": 0.2869166686179577,
"learning_rate": 3.184492904119072e-05,
"loss": 0.3611,
"step": 1370
},
{
"epoch": 1.2840457836953982,
"grad_norm": 0.24540393338552327,
"learning_rate": 3.175839390792662e-05,
"loss": 0.3491,
"step": 1375
},
{
"epoch": 1.2887175893482832,
"grad_norm": 0.27095178205718407,
"learning_rate": 3.1671858774662516e-05,
"loss": 0.361,
"step": 1380
},
{
"epoch": 1.293389395001168,
"grad_norm": 4.448092834222375,
"learning_rate": 3.1585323641398407e-05,
"loss": 0.3671,
"step": 1385
},
{
"epoch": 1.2980612006540528,
"grad_norm": 0.29030980188514105,
"learning_rate": 3.1498788508134304e-05,
"loss": 0.3657,
"step": 1390
},
{
"epoch": 1.3027330063069376,
"grad_norm": 0.25570751046872936,
"learning_rate": 3.14122533748702e-05,
"loss": 0.3496,
"step": 1395
},
{
"epoch": 1.3074048119598225,
"grad_norm": 0.3041092750342911,
"learning_rate": 3.132571824160609e-05,
"loss": 0.357,
"step": 1400
},
{
"epoch": 1.3120766176127074,
"grad_norm": 0.26998631488256053,
"learning_rate": 3.123918310834199e-05,
"loss": 0.3636,
"step": 1405
},
{
"epoch": 1.3167484232655922,
"grad_norm": 0.31498354527546657,
"learning_rate": 3.115264797507788e-05,
"loss": 0.357,
"step": 1410
},
{
"epoch": 1.3214202289184769,
"grad_norm": 0.2660391767855304,
"learning_rate": 3.106611284181378e-05,
"loss": 0.357,
"step": 1415
},
{
"epoch": 1.3260920345713618,
"grad_norm": 0.27718631765205176,
"learning_rate": 3.0979577708549675e-05,
"loss": 0.3524,
"step": 1420
},
{
"epoch": 1.3307638402242468,
"grad_norm": 0.28859182485644613,
"learning_rate": 3.0893042575285566e-05,
"loss": 0.3473,
"step": 1425
},
{
"epoch": 1.3354356458771315,
"grad_norm": 0.2784672251543859,
"learning_rate": 3.0806507442021456e-05,
"loss": 0.3519,
"step": 1430
},
{
"epoch": 1.3401074515300164,
"grad_norm": 0.28008342661685554,
"learning_rate": 3.0719972308757354e-05,
"loss": 0.3586,
"step": 1435
},
{
"epoch": 1.3447792571829011,
"grad_norm": 0.2930419122109119,
"learning_rate": 3.063343717549325e-05,
"loss": 0.3548,
"step": 1440
},
{
"epoch": 1.349451062835786,
"grad_norm": 0.2680844586193672,
"learning_rate": 3.054690204222915e-05,
"loss": 0.3604,
"step": 1445
},
{
"epoch": 1.3541228684886708,
"grad_norm": 0.2652532733025401,
"learning_rate": 3.0460366908965043e-05,
"loss": 0.3541,
"step": 1450
},
{
"epoch": 1.3587946741415557,
"grad_norm": 0.27794394077314344,
"learning_rate": 3.0373831775700934e-05,
"loss": 0.3545,
"step": 1455
},
{
"epoch": 1.3634664797944405,
"grad_norm": 0.2684193490611374,
"learning_rate": 3.028729664243683e-05,
"loss": 0.3653,
"step": 1460
},
{
"epoch": 1.3681382854473254,
"grad_norm": 0.2636254442479292,
"learning_rate": 3.0200761509172725e-05,
"loss": 0.3588,
"step": 1465
},
{
"epoch": 1.3728100911002104,
"grad_norm": 0.2679525113872143,
"learning_rate": 3.0114226375908622e-05,
"loss": 0.3512,
"step": 1470
},
{
"epoch": 1.377481896753095,
"grad_norm": 0.26819759906115087,
"learning_rate": 3.0027691242644517e-05,
"loss": 0.3505,
"step": 1475
},
{
"epoch": 1.3821537024059798,
"grad_norm": 0.2699938658297075,
"learning_rate": 2.9941156109380407e-05,
"loss": 0.3523,
"step": 1480
},
{
"epoch": 1.3868255080588647,
"grad_norm": 0.33892479933409453,
"learning_rate": 2.9854620976116305e-05,
"loss": 0.3667,
"step": 1485
},
{
"epoch": 1.3914973137117497,
"grad_norm": 0.2920963025389174,
"learning_rate": 2.97680858428522e-05,
"loss": 0.3608,
"step": 1490
},
{
"epoch": 1.3961691193646344,
"grad_norm": 0.2569174080682022,
"learning_rate": 2.9681550709588096e-05,
"loss": 0.3537,
"step": 1495
},
{
"epoch": 1.4008409250175193,
"grad_norm": 0.2958548233616089,
"learning_rate": 2.9595015576323987e-05,
"loss": 0.357,
"step": 1500
},
{
"epoch": 1.405512730670404,
"grad_norm": 0.25733764011116217,
"learning_rate": 2.9508480443059884e-05,
"loss": 0.3569,
"step": 1505
},
{
"epoch": 1.410184536323289,
"grad_norm": 0.25132267401727015,
"learning_rate": 2.942194530979578e-05,
"loss": 0.3568,
"step": 1510
},
{
"epoch": 1.4148563419761737,
"grad_norm": 0.2669832051506726,
"learning_rate": 2.9335410176531676e-05,
"loss": 0.351,
"step": 1515
},
{
"epoch": 1.4195281476290587,
"grad_norm": 0.2613018841035631,
"learning_rate": 2.924887504326757e-05,
"loss": 0.356,
"step": 1520
},
{
"epoch": 1.4241999532819434,
"grad_norm": 0.3090873396882133,
"learning_rate": 2.916233991000346e-05,
"loss": 0.3584,
"step": 1525
},
{
"epoch": 1.4288717589348283,
"grad_norm": 0.3049139538157588,
"learning_rate": 2.9075804776739358e-05,
"loss": 0.365,
"step": 1530
},
{
"epoch": 1.4335435645877133,
"grad_norm": 0.2688570525079533,
"learning_rate": 2.8989269643475252e-05,
"loss": 0.3595,
"step": 1535
},
{
"epoch": 1.438215370240598,
"grad_norm": 0.2624542720616472,
"learning_rate": 2.890273451021115e-05,
"loss": 0.3502,
"step": 1540
},
{
"epoch": 1.4428871758934827,
"grad_norm": 0.27111582469091855,
"learning_rate": 2.881619937694704e-05,
"loss": 0.3603,
"step": 1545
},
{
"epoch": 1.4475589815463676,
"grad_norm": 0.2780354003220708,
"learning_rate": 2.8729664243682934e-05,
"loss": 0.3649,
"step": 1550
},
{
"epoch": 1.4522307871992526,
"grad_norm": 0.260623170766383,
"learning_rate": 2.8643129110418832e-05,
"loss": 0.361,
"step": 1555
},
{
"epoch": 1.4569025928521373,
"grad_norm": 0.28249704920723373,
"learning_rate": 2.855659397715473e-05,
"loss": 0.3509,
"step": 1560
},
{
"epoch": 1.4615743985050222,
"grad_norm": 0.2837000710912426,
"learning_rate": 2.8470058843890623e-05,
"loss": 0.3606,
"step": 1565
},
{
"epoch": 1.466246204157907,
"grad_norm": 0.26910890181120617,
"learning_rate": 2.8383523710626514e-05,
"loss": 0.3559,
"step": 1570
},
{
"epoch": 1.470918009810792,
"grad_norm": 0.24788408369485626,
"learning_rate": 2.829698857736241e-05,
"loss": 0.3545,
"step": 1575
},
{
"epoch": 1.4755898154636768,
"grad_norm": 0.28250052555468164,
"learning_rate": 2.8210453444098305e-05,
"loss": 0.3549,
"step": 1580
},
{
"epoch": 1.4802616211165616,
"grad_norm": 0.2818967756986274,
"learning_rate": 2.8123918310834203e-05,
"loss": 0.3558,
"step": 1585
},
{
"epoch": 1.4849334267694463,
"grad_norm": 0.2654214906895977,
"learning_rate": 2.8037383177570094e-05,
"loss": 0.3527,
"step": 1590
},
{
"epoch": 1.4896052324223312,
"grad_norm": 0.27797835373625635,
"learning_rate": 2.7950848044305988e-05,
"loss": 0.3543,
"step": 1595
},
{
"epoch": 1.4942770380752162,
"grad_norm": 0.2759754122307631,
"learning_rate": 2.7864312911041885e-05,
"loss": 0.3519,
"step": 1600
},
{
"epoch": 1.4989488437281009,
"grad_norm": 0.2676399748360465,
"learning_rate": 2.777777777777778e-05,
"loss": 0.3602,
"step": 1605
},
{
"epoch": 1.5036206493809856,
"grad_norm": 0.29148569741911323,
"learning_rate": 2.7691242644513677e-05,
"loss": 0.3603,
"step": 1610
},
{
"epoch": 1.5082924550338705,
"grad_norm": 0.30555870401526036,
"learning_rate": 2.7604707511249567e-05,
"loss": 0.351,
"step": 1615
},
{
"epoch": 1.5129642606867555,
"grad_norm": 0.24008291867578185,
"learning_rate": 2.751817237798546e-05,
"loss": 0.3537,
"step": 1620
},
{
"epoch": 1.5176360663396404,
"grad_norm": 0.26896039652376746,
"learning_rate": 2.743163724472136e-05,
"loss": 0.3592,
"step": 1625
},
{
"epoch": 1.5223078719925252,
"grad_norm": 0.2656250365980771,
"learning_rate": 2.7345102111457256e-05,
"loss": 0.3443,
"step": 1630
},
{
"epoch": 1.5269796776454099,
"grad_norm": 0.2731809064015199,
"learning_rate": 2.7258566978193147e-05,
"loss": 0.3588,
"step": 1635
},
{
"epoch": 1.5316514832982948,
"grad_norm": 0.25856707440375554,
"learning_rate": 2.717203184492904e-05,
"loss": 0.3516,
"step": 1640
},
{
"epoch": 1.5363232889511798,
"grad_norm": 0.2608276718680239,
"learning_rate": 2.708549671166494e-05,
"loss": 0.3567,
"step": 1645
},
{
"epoch": 1.5409950946040645,
"grad_norm": 0.24006604176521526,
"learning_rate": 2.6998961578400832e-05,
"loss": 0.3525,
"step": 1650
},
{
"epoch": 1.5456669002569492,
"grad_norm": 0.263533118070647,
"learning_rate": 2.691242644513673e-05,
"loss": 0.3617,
"step": 1655
},
{
"epoch": 1.5503387059098341,
"grad_norm": 0.30176717620519383,
"learning_rate": 2.682589131187262e-05,
"loss": 0.3615,
"step": 1660
},
{
"epoch": 1.555010511562719,
"grad_norm": 0.25460999061246914,
"learning_rate": 2.6739356178608515e-05,
"loss": 0.3556,
"step": 1665
},
{
"epoch": 1.5596823172156038,
"grad_norm": 0.2769292746201983,
"learning_rate": 2.6652821045344412e-05,
"loss": 0.361,
"step": 1670
},
{
"epoch": 1.5643541228684885,
"grad_norm": 0.9998644588384835,
"learning_rate": 2.6566285912080306e-05,
"loss": 0.3495,
"step": 1675
},
{
"epoch": 1.5690259285213735,
"grad_norm": 0.26437954267999175,
"learning_rate": 2.6479750778816197e-05,
"loss": 0.3498,
"step": 1680
},
{
"epoch": 1.5736977341742584,
"grad_norm": 0.3080269931674715,
"learning_rate": 2.6393215645552094e-05,
"loss": 0.3523,
"step": 1685
},
{
"epoch": 1.5783695398271433,
"grad_norm": 0.2751400251319551,
"learning_rate": 2.6306680512287992e-05,
"loss": 0.363,
"step": 1690
},
{
"epoch": 1.583041345480028,
"grad_norm": 0.27625088276540716,
"learning_rate": 2.6220145379023886e-05,
"loss": 0.3507,
"step": 1695
},
{
"epoch": 1.5877131511329128,
"grad_norm": 0.24787550008096446,
"learning_rate": 2.6133610245759783e-05,
"loss": 0.3552,
"step": 1700
},
{
"epoch": 1.5923849567857977,
"grad_norm": 0.24885679890801207,
"learning_rate": 2.6047075112495674e-05,
"loss": 0.3485,
"step": 1705
},
{
"epoch": 1.5970567624386827,
"grad_norm": 0.25785490206314426,
"learning_rate": 2.5960539979231568e-05,
"loss": 0.3551,
"step": 1710
},
{
"epoch": 1.6017285680915674,
"grad_norm": 0.2526987125119253,
"learning_rate": 2.5874004845967465e-05,
"loss": 0.36,
"step": 1715
},
{
"epoch": 1.606400373744452,
"grad_norm": 0.23963597328766026,
"learning_rate": 2.578746971270336e-05,
"loss": 0.3615,
"step": 1720
},
{
"epoch": 1.611072179397337,
"grad_norm": 0.2604764012528186,
"learning_rate": 2.570093457943925e-05,
"loss": 0.3484,
"step": 1725
},
{
"epoch": 1.615743985050222,
"grad_norm": 0.27026077600137727,
"learning_rate": 2.5614399446175148e-05,
"loss": 0.3565,
"step": 1730
},
{
"epoch": 1.6204157907031067,
"grad_norm": 0.2544087831975487,
"learning_rate": 2.552786431291104e-05,
"loss": 0.3491,
"step": 1735
},
{
"epoch": 1.6250875963559914,
"grad_norm": 0.25041011980856775,
"learning_rate": 2.544132917964694e-05,
"loss": 0.3507,
"step": 1740
},
{
"epoch": 1.6297594020088764,
"grad_norm": 0.24716417636609062,
"learning_rate": 2.5354794046382833e-05,
"loss": 0.354,
"step": 1745
},
{
"epoch": 1.6344312076617613,
"grad_norm": 0.2602000753124882,
"learning_rate": 2.5268258913118724e-05,
"loss": 0.356,
"step": 1750
},
{
"epoch": 1.6391030133146463,
"grad_norm": 0.25454044050185637,
"learning_rate": 2.518172377985462e-05,
"loss": 0.3531,
"step": 1755
},
{
"epoch": 1.643774818967531,
"grad_norm": 0.2833770824267626,
"learning_rate": 2.509518864659052e-05,
"loss": 0.3497,
"step": 1760
},
{
"epoch": 1.6484466246204157,
"grad_norm": 0.25070331463030265,
"learning_rate": 2.5008653513326413e-05,
"loss": 0.3601,
"step": 1765
},
{
"epoch": 1.6531184302733006,
"grad_norm": 0.2458032214479526,
"learning_rate": 2.4922118380062307e-05,
"loss": 0.3506,
"step": 1770
},
{
"epoch": 1.6577902359261856,
"grad_norm": 0.25586691924590127,
"learning_rate": 2.48355832467982e-05,
"loss": 0.3543,
"step": 1775
},
{
"epoch": 1.6624620415790703,
"grad_norm": 0.2546410430722946,
"learning_rate": 2.4749048113534095e-05,
"loss": 0.3485,
"step": 1780
},
{
"epoch": 1.667133847231955,
"grad_norm": 0.2314147617602589,
"learning_rate": 2.4662512980269992e-05,
"loss": 0.3473,
"step": 1785
},
{
"epoch": 1.67180565288484,
"grad_norm": 0.229752845273799,
"learning_rate": 2.4575977847005887e-05,
"loss": 0.3522,
"step": 1790
},
{
"epoch": 1.676477458537725,
"grad_norm": 0.2566297609627607,
"learning_rate": 2.448944271374178e-05,
"loss": 0.3495,
"step": 1795
},
{
"epoch": 1.6811492641906096,
"grad_norm": 0.24997518582824338,
"learning_rate": 2.4402907580477675e-05,
"loss": 0.3518,
"step": 1800
},
{
"epoch": 1.6858210698434946,
"grad_norm": 0.23492666288431144,
"learning_rate": 2.431637244721357e-05,
"loss": 0.3526,
"step": 1805
},
{
"epoch": 1.6904928754963793,
"grad_norm": 0.261504362105359,
"learning_rate": 2.4229837313949463e-05,
"loss": 0.3578,
"step": 1810
},
{
"epoch": 1.6951646811492642,
"grad_norm": 0.23038766630433052,
"learning_rate": 2.414330218068536e-05,
"loss": 0.3516,
"step": 1815
},
{
"epoch": 1.6998364868021492,
"grad_norm": 0.23631858442788584,
"learning_rate": 2.4056767047421254e-05,
"loss": 0.3569,
"step": 1820
},
{
"epoch": 1.7045082924550339,
"grad_norm": 0.25157564283640055,
"learning_rate": 2.397023191415715e-05,
"loss": 0.3555,
"step": 1825
},
{
"epoch": 1.7091800981079186,
"grad_norm": 0.25879066104554227,
"learning_rate": 2.3883696780893046e-05,
"loss": 0.3496,
"step": 1830
},
{
"epoch": 1.7138519037608035,
"grad_norm": 0.22931371833436698,
"learning_rate": 2.3797161647628936e-05,
"loss": 0.3593,
"step": 1835
},
{
"epoch": 1.7185237094136885,
"grad_norm": 0.2499124013797676,
"learning_rate": 2.3710626514364834e-05,
"loss": 0.3518,
"step": 1840
},
{
"epoch": 1.7231955150665732,
"grad_norm": 0.2473814229439637,
"learning_rate": 2.3624091381100728e-05,
"loss": 0.3541,
"step": 1845
},
{
"epoch": 1.727867320719458,
"grad_norm": 0.24044010726393764,
"learning_rate": 2.3537556247836622e-05,
"loss": 0.3437,
"step": 1850
},
{
"epoch": 1.7325391263723429,
"grad_norm": 0.26246375246193315,
"learning_rate": 2.345102111457252e-05,
"loss": 0.356,
"step": 1855
},
{
"epoch": 1.7372109320252278,
"grad_norm": 0.26557818782226306,
"learning_rate": 2.3364485981308414e-05,
"loss": 0.355,
"step": 1860
},
{
"epoch": 1.7418827376781127,
"grad_norm": 0.256684454105562,
"learning_rate": 2.3277950848044308e-05,
"loss": 0.3497,
"step": 1865
},
{
"epoch": 1.7465545433309975,
"grad_norm": 0.24398080349583323,
"learning_rate": 2.31914157147802e-05,
"loss": 0.353,
"step": 1870
},
{
"epoch": 1.7512263489838822,
"grad_norm": 0.256745449324351,
"learning_rate": 2.3104880581516096e-05,
"loss": 0.3519,
"step": 1875
},
{
"epoch": 1.7558981546367671,
"grad_norm": 0.2556036643262852,
"learning_rate": 2.301834544825199e-05,
"loss": 0.3529,
"step": 1880
},
{
"epoch": 1.760569960289652,
"grad_norm": 0.2614568711313752,
"learning_rate": 2.2931810314987887e-05,
"loss": 0.356,
"step": 1885
},
{
"epoch": 1.7652417659425368,
"grad_norm": 0.28516459215745127,
"learning_rate": 2.284527518172378e-05,
"loss": 0.3585,
"step": 1890
},
{
"epoch": 1.7699135715954215,
"grad_norm": 0.27387023450386516,
"learning_rate": 2.2758740048459675e-05,
"loss": 0.3629,
"step": 1895
},
{
"epoch": 1.7745853772483065,
"grad_norm": 0.2729502926762472,
"learning_rate": 2.2672204915195573e-05,
"loss": 0.3575,
"step": 1900
},
{
"epoch": 1.7792571829011914,
"grad_norm": 0.24916039736502976,
"learning_rate": 2.2585669781931463e-05,
"loss": 0.3522,
"step": 1905
},
{
"epoch": 1.7839289885540761,
"grad_norm": 0.28488813181387246,
"learning_rate": 2.249913464866736e-05,
"loss": 0.3535,
"step": 1910
},
{
"epoch": 1.7886007942069608,
"grad_norm": 0.26130228453505255,
"learning_rate": 2.2412599515403255e-05,
"loss": 0.3466,
"step": 1915
},
{
"epoch": 1.7932725998598458,
"grad_norm": 0.29826118310274563,
"learning_rate": 2.232606438213915e-05,
"loss": 0.3487,
"step": 1920
},
{
"epoch": 1.7979444055127307,
"grad_norm": 0.24057399234165708,
"learning_rate": 2.2239529248875043e-05,
"loss": 0.3463,
"step": 1925
},
{
"epoch": 1.8026162111656157,
"grad_norm": 0.28099984095688735,
"learning_rate": 2.215299411561094e-05,
"loss": 0.3563,
"step": 1930
},
{
"epoch": 1.8072880168185004,
"grad_norm": 0.25003633200863506,
"learning_rate": 2.206645898234683e-05,
"loss": 0.3442,
"step": 1935
},
{
"epoch": 1.811959822471385,
"grad_norm": 0.2397456416268439,
"learning_rate": 2.197992384908273e-05,
"loss": 0.3587,
"step": 1940
},
{
"epoch": 1.81663162812427,
"grad_norm": 0.24746270687518188,
"learning_rate": 2.1893388715818626e-05,
"loss": 0.3474,
"step": 1945
},
{
"epoch": 1.821303433777155,
"grad_norm": 0.2389158619700501,
"learning_rate": 2.1806853582554517e-05,
"loss": 0.3507,
"step": 1950
},
{
"epoch": 1.8259752394300397,
"grad_norm": 0.23840109603392132,
"learning_rate": 2.1720318449290414e-05,
"loss": 0.3559,
"step": 1955
},
{
"epoch": 1.8306470450829244,
"grad_norm": 0.24163055044842047,
"learning_rate": 2.163378331602631e-05,
"loss": 0.3461,
"step": 1960
},
{
"epoch": 1.8353188507358094,
"grad_norm": 0.2411390279052765,
"learning_rate": 2.1547248182762202e-05,
"loss": 0.3507,
"step": 1965
},
{
"epoch": 1.8399906563886943,
"grad_norm": 0.2689246086175121,
"learning_rate": 2.1460713049498096e-05,
"loss": 0.3499,
"step": 1970
},
{
"epoch": 1.844662462041579,
"grad_norm": 0.2575753325596786,
"learning_rate": 2.1374177916233994e-05,
"loss": 0.3573,
"step": 1975
},
{
"epoch": 1.849334267694464,
"grad_norm": 0.22860025701284803,
"learning_rate": 2.1287642782969885e-05,
"loss": 0.3546,
"step": 1980
},
{
"epoch": 1.8540060733473487,
"grad_norm": 0.2473425651945895,
"learning_rate": 2.1201107649705782e-05,
"loss": 0.3443,
"step": 1985
},
{
"epoch": 1.8586778790002336,
"grad_norm": 0.2729706201259369,
"learning_rate": 2.1114572516441676e-05,
"loss": 0.3536,
"step": 1990
},
{
"epoch": 1.8633496846531186,
"grad_norm": 0.2657729323563545,
"learning_rate": 2.102803738317757e-05,
"loss": 0.3615,
"step": 1995
},
{
"epoch": 1.8680214903060033,
"grad_norm": 0.2507639279547885,
"learning_rate": 2.0941502249913468e-05,
"loss": 0.3611,
"step": 2000
},
{
"epoch": 1.872693295958888,
"grad_norm": 0.24604402880476078,
"learning_rate": 2.0854967116649358e-05,
"loss": 0.3492,
"step": 2005
},
{
"epoch": 1.877365101611773,
"grad_norm": 0.2932628949996562,
"learning_rate": 2.0768431983385256e-05,
"loss": 0.3529,
"step": 2010
},
{
"epoch": 1.8820369072646579,
"grad_norm": 0.253627925676872,
"learning_rate": 2.068189685012115e-05,
"loss": 0.3544,
"step": 2015
},
{
"epoch": 1.8867087129175426,
"grad_norm": 0.2354239168303899,
"learning_rate": 2.0595361716857044e-05,
"loss": 0.3452,
"step": 2020
},
{
"epoch": 1.8913805185704273,
"grad_norm": 0.24589941564458,
"learning_rate": 2.050882658359294e-05,
"loss": 0.3508,
"step": 2025
},
{
"epoch": 1.8960523242233123,
"grad_norm": 1.1331048402557922,
"learning_rate": 2.0422291450328835e-05,
"loss": 0.3472,
"step": 2030
},
{
"epoch": 1.9007241298761972,
"grad_norm": 0.2502077991662155,
"learning_rate": 2.033575631706473e-05,
"loss": 0.3574,
"step": 2035
},
{
"epoch": 1.9053959355290822,
"grad_norm": 0.23998015730620764,
"learning_rate": 2.0249221183800623e-05,
"loss": 0.3564,
"step": 2040
},
{
"epoch": 1.9100677411819669,
"grad_norm": 0.24963596950848252,
"learning_rate": 2.016268605053652e-05,
"loss": 0.3476,
"step": 2045
},
{
"epoch": 1.9147395468348516,
"grad_norm": 0.24192663446212184,
"learning_rate": 2.007615091727241e-05,
"loss": 0.3448,
"step": 2050
},
{
"epoch": 1.9194113524877365,
"grad_norm": 0.24444512910562927,
"learning_rate": 1.998961578400831e-05,
"loss": 0.351,
"step": 2055
},
{
"epoch": 1.9240831581406215,
"grad_norm": 0.24898191090908084,
"learning_rate": 1.9903080650744203e-05,
"loss": 0.3498,
"step": 2060
},
{
"epoch": 1.9287549637935062,
"grad_norm": 0.24278405327885177,
"learning_rate": 1.9816545517480097e-05,
"loss": 0.3484,
"step": 2065
},
{
"epoch": 1.933426769446391,
"grad_norm": 0.2462167764412803,
"learning_rate": 1.9730010384215995e-05,
"loss": 0.3454,
"step": 2070
},
{
"epoch": 1.9380985750992759,
"grad_norm": 0.2395040446318053,
"learning_rate": 1.964347525095189e-05,
"loss": 0.3481,
"step": 2075
},
{
"epoch": 1.9427703807521608,
"grad_norm": 0.2709668463672972,
"learning_rate": 1.9556940117687783e-05,
"loss": 0.3556,
"step": 2080
},
{
"epoch": 1.9474421864050455,
"grad_norm": 0.26890519167115307,
"learning_rate": 1.9470404984423677e-05,
"loss": 0.3573,
"step": 2085
},
{
"epoch": 1.9521139920579302,
"grad_norm": 0.2665557191787376,
"learning_rate": 1.938386985115957e-05,
"loss": 0.3564,
"step": 2090
},
{
"epoch": 1.9567857977108152,
"grad_norm": 0.2867599379717776,
"learning_rate": 1.9297334717895465e-05,
"loss": 0.3507,
"step": 2095
},
{
"epoch": 1.9614576033637001,
"grad_norm": 0.24926429000265118,
"learning_rate": 1.9210799584631362e-05,
"loss": 0.3473,
"step": 2100
},
{
"epoch": 1.966129409016585,
"grad_norm": 0.25192232581044366,
"learning_rate": 1.9124264451367253e-05,
"loss": 0.3519,
"step": 2105
},
{
"epoch": 1.9708012146694698,
"grad_norm": 0.23833315952573492,
"learning_rate": 1.903772931810315e-05,
"loss": 0.3528,
"step": 2110
},
{
"epoch": 1.9754730203223545,
"grad_norm": 0.28688121316933135,
"learning_rate": 1.8951194184839048e-05,
"loss": 0.3431,
"step": 2115
},
{
"epoch": 1.9801448259752394,
"grad_norm": 0.22959759632372803,
"learning_rate": 1.886465905157494e-05,
"loss": 0.3463,
"step": 2120
},
{
"epoch": 1.9848166316281244,
"grad_norm": 0.25686912281201,
"learning_rate": 1.8778123918310836e-05,
"loss": 0.3435,
"step": 2125
},
{
"epoch": 1.989488437281009,
"grad_norm": 0.23795225574530124,
"learning_rate": 1.869158878504673e-05,
"loss": 0.3495,
"step": 2130
},
{
"epoch": 1.9941602429338938,
"grad_norm": 0.747482414765369,
"learning_rate": 1.8605053651782624e-05,
"loss": 0.3506,
"step": 2135
},
{
"epoch": 1.9988320485867788,
"grad_norm": 0.28050633953526405,
"learning_rate": 1.8518518518518518e-05,
"loss": 0.3513,
"step": 2140
},
{
"epoch": 2.002803083391731,
"grad_norm": 0.34435077567177486,
"learning_rate": 1.8431983385254416e-05,
"loss": 0.3053,
"step": 2145
},
{
"epoch": 2.007474889044616,
"grad_norm": 0.3232861983470053,
"learning_rate": 1.8345448251990306e-05,
"loss": 0.276,
"step": 2150
},
{
"epoch": 2.0121466946975004,
"grad_norm": 0.2991643901861678,
"learning_rate": 1.8258913118726204e-05,
"loss": 0.2673,
"step": 2155
},
{
"epoch": 2.0168185003503853,
"grad_norm": 0.2741545694404629,
"learning_rate": 1.8172377985462098e-05,
"loss": 0.2739,
"step": 2160
},
{
"epoch": 2.0214903060032703,
"grad_norm": 0.2617834943522712,
"learning_rate": 1.8085842852197992e-05,
"loss": 0.27,
"step": 2165
},
{
"epoch": 2.0261621116561552,
"grad_norm": 0.26830601108259017,
"learning_rate": 1.799930771893389e-05,
"loss": 0.2712,
"step": 2170
},
{
"epoch": 2.0308339173090397,
"grad_norm": 0.2733105802769957,
"learning_rate": 1.7912772585669783e-05,
"loss": 0.2663,
"step": 2175
},
{
"epoch": 2.0355057229619247,
"grad_norm": 0.251566109220412,
"learning_rate": 1.7826237452405678e-05,
"loss": 0.2753,
"step": 2180
},
{
"epoch": 2.0401775286148096,
"grad_norm": 0.27175452190308935,
"learning_rate": 1.773970231914157e-05,
"loss": 0.2709,
"step": 2185
},
{
"epoch": 2.0448493342676946,
"grad_norm": 0.24647565952032985,
"learning_rate": 1.7653167185877466e-05,
"loss": 0.2716,
"step": 2190
},
{
"epoch": 2.0495211399205795,
"grad_norm": 0.27010143574994433,
"learning_rate": 1.7566632052613363e-05,
"loss": 0.2651,
"step": 2195
},
{
"epoch": 2.054192945573464,
"grad_norm": 0.25716921848079094,
"learning_rate": 1.7480096919349257e-05,
"loss": 0.2667,
"step": 2200
},
{
"epoch": 2.058864751226349,
"grad_norm": 0.24574201714503535,
"learning_rate": 1.739356178608515e-05,
"loss": 0.265,
"step": 2205
},
{
"epoch": 2.063536556879234,
"grad_norm": 0.27472795148588064,
"learning_rate": 1.7307026652821045e-05,
"loss": 0.2784,
"step": 2210
},
{
"epoch": 2.068208362532119,
"grad_norm": 0.26861665903194765,
"learning_rate": 1.7220491519556943e-05,
"loss": 0.2725,
"step": 2215
},
{
"epoch": 2.0728801681850033,
"grad_norm": 0.25294960940453676,
"learning_rate": 1.7133956386292833e-05,
"loss": 0.2744,
"step": 2220
},
{
"epoch": 2.0775519738378883,
"grad_norm": 0.2709157246182442,
"learning_rate": 1.704742125302873e-05,
"loss": 0.2673,
"step": 2225
},
{
"epoch": 2.082223779490773,
"grad_norm": 0.23791548921338562,
"learning_rate": 1.6960886119764625e-05,
"loss": 0.2731,
"step": 2230
},
{
"epoch": 2.086895585143658,
"grad_norm": 0.2580146491292611,
"learning_rate": 1.687435098650052e-05,
"loss": 0.263,
"step": 2235
},
{
"epoch": 2.091567390796543,
"grad_norm": 0.25645006753439864,
"learning_rate": 1.6787815853236416e-05,
"loss": 0.2715,
"step": 2240
},
{
"epoch": 2.0962391964494276,
"grad_norm": 0.24389642866586017,
"learning_rate": 1.670128071997231e-05,
"loss": 0.2656,
"step": 2245
},
{
"epoch": 2.1009110021023125,
"grad_norm": 0.25677612105078285,
"learning_rate": 1.6614745586708205e-05,
"loss": 0.2684,
"step": 2250
},
{
"epoch": 2.1055828077551975,
"grad_norm": 0.25580810186269193,
"learning_rate": 1.65282104534441e-05,
"loss": 0.2706,
"step": 2255
},
{
"epoch": 2.1102546134080824,
"grad_norm": 1.220797185763052,
"learning_rate": 1.6441675320179996e-05,
"loss": 0.274,
"step": 2260
},
{
"epoch": 2.114926419060967,
"grad_norm": 0.2559597552026823,
"learning_rate": 1.6355140186915887e-05,
"loss": 0.2712,
"step": 2265
},
{
"epoch": 2.119598224713852,
"grad_norm": 0.24963492503244736,
"learning_rate": 1.6268605053651784e-05,
"loss": 0.2712,
"step": 2270
},
{
"epoch": 2.124270030366737,
"grad_norm": 0.2626845091250949,
"learning_rate": 1.6182069920387678e-05,
"loss": 0.27,
"step": 2275
},
{
"epoch": 2.1289418360196217,
"grad_norm": 0.26563248949871293,
"learning_rate": 1.6095534787123572e-05,
"loss": 0.2661,
"step": 2280
},
{
"epoch": 2.1336136416725062,
"grad_norm": 0.25452400807638254,
"learning_rate": 1.600899965385947e-05,
"loss": 0.2775,
"step": 2285
},
{
"epoch": 2.138285447325391,
"grad_norm": 0.25150023437927815,
"learning_rate": 1.592246452059536e-05,
"loss": 0.268,
"step": 2290
},
{
"epoch": 2.142957252978276,
"grad_norm": 0.26496698727100276,
"learning_rate": 1.5835929387331258e-05,
"loss": 0.27,
"step": 2295
},
{
"epoch": 2.147629058631161,
"grad_norm": 0.26274053883641585,
"learning_rate": 1.5749394254067152e-05,
"loss": 0.2626,
"step": 2300
},
{
"epoch": 2.152300864284046,
"grad_norm": 0.24417182547206637,
"learning_rate": 1.5662859120803046e-05,
"loss": 0.268,
"step": 2305
},
{
"epoch": 2.1569726699369305,
"grad_norm": 0.25099784143140763,
"learning_rate": 1.557632398753894e-05,
"loss": 0.2661,
"step": 2310
},
{
"epoch": 2.1616444755898154,
"grad_norm": 0.2726937615081363,
"learning_rate": 1.5489788854274838e-05,
"loss": 0.2758,
"step": 2315
},
{
"epoch": 2.1663162812427004,
"grad_norm": 0.2534142205531808,
"learning_rate": 1.5403253721010728e-05,
"loss": 0.2718,
"step": 2320
},
{
"epoch": 2.1709880868955853,
"grad_norm": 0.26619731721174505,
"learning_rate": 1.5316718587746626e-05,
"loss": 0.2711,
"step": 2325
},
{
"epoch": 2.17565989254847,
"grad_norm": 0.25487363737119223,
"learning_rate": 1.5230183454482521e-05,
"loss": 0.2678,
"step": 2330
},
{
"epoch": 2.1803316982013548,
"grad_norm": 0.2451055063080652,
"learning_rate": 1.5143648321218415e-05,
"loss": 0.2732,
"step": 2335
},
{
"epoch": 2.1850035038542397,
"grad_norm": 0.25600856881621503,
"learning_rate": 1.5057113187954311e-05,
"loss": 0.2669,
"step": 2340
},
{
"epoch": 2.1896753095071246,
"grad_norm": 0.2549061650595312,
"learning_rate": 1.4970578054690204e-05,
"loss": 0.2648,
"step": 2345
},
{
"epoch": 2.194347115160009,
"grad_norm": 0.28081165004339137,
"learning_rate": 1.48840429214261e-05,
"loss": 0.2744,
"step": 2350
},
{
"epoch": 2.199018920812894,
"grad_norm": 0.2597182832277994,
"learning_rate": 1.4797507788161993e-05,
"loss": 0.2707,
"step": 2355
},
{
"epoch": 2.203690726465779,
"grad_norm": 0.2561877721328425,
"learning_rate": 1.471097265489789e-05,
"loss": 0.2689,
"step": 2360
},
{
"epoch": 2.208362532118664,
"grad_norm": 0.2563076388294137,
"learning_rate": 1.4624437521633785e-05,
"loss": 0.268,
"step": 2365
},
{
"epoch": 2.213034337771549,
"grad_norm": 0.25765046375047895,
"learning_rate": 1.4537902388369679e-05,
"loss": 0.2656,
"step": 2370
},
{
"epoch": 2.2177061434244334,
"grad_norm": 0.28518733350470343,
"learning_rate": 1.4451367255105575e-05,
"loss": 0.2685,
"step": 2375
},
{
"epoch": 2.2223779490773183,
"grad_norm": 0.25368307928496053,
"learning_rate": 1.4364832121841467e-05,
"loss": 0.2725,
"step": 2380
},
{
"epoch": 2.2270497547302033,
"grad_norm": 0.5281956077578304,
"learning_rate": 1.4278296988577365e-05,
"loss": 0.2714,
"step": 2385
},
{
"epoch": 2.231721560383088,
"grad_norm": 0.25641285413166875,
"learning_rate": 1.4191761855313257e-05,
"loss": 0.2667,
"step": 2390
},
{
"epoch": 2.2363933660359727,
"grad_norm": 0.2544566354072801,
"learning_rate": 1.4105226722049153e-05,
"loss": 0.2776,
"step": 2395
},
{
"epoch": 2.2410651716888577,
"grad_norm": 0.2654527447971898,
"learning_rate": 1.4018691588785047e-05,
"loss": 0.2749,
"step": 2400
},
{
"epoch": 2.2457369773417426,
"grad_norm": 0.26498542303356154,
"learning_rate": 1.3932156455520943e-05,
"loss": 0.2768,
"step": 2405
},
{
"epoch": 2.2504087829946275,
"grad_norm": 0.2617880252712014,
"learning_rate": 1.3845621322256838e-05,
"loss": 0.2668,
"step": 2410
},
{
"epoch": 2.2550805886475125,
"grad_norm": 0.27521153124310155,
"learning_rate": 1.375908618899273e-05,
"loss": 0.2716,
"step": 2415
},
{
"epoch": 2.259752394300397,
"grad_norm": 0.2464659107912401,
"learning_rate": 1.3672551055728628e-05,
"loss": 0.2699,
"step": 2420
},
{
"epoch": 2.264424199953282,
"grad_norm": 0.2527245766560059,
"learning_rate": 1.358601592246452e-05,
"loss": 0.2654,
"step": 2425
},
{
"epoch": 2.269096005606167,
"grad_norm": 0.2414395786265717,
"learning_rate": 1.3499480789200416e-05,
"loss": 0.2649,
"step": 2430
},
{
"epoch": 2.273767811259052,
"grad_norm": 0.24272525169309794,
"learning_rate": 1.341294565593631e-05,
"loss": 0.2735,
"step": 2435
},
{
"epoch": 2.2784396169119363,
"grad_norm": 0.2628720042423123,
"learning_rate": 1.3326410522672206e-05,
"loss": 0.273,
"step": 2440
},
{
"epoch": 2.2831114225648212,
"grad_norm": 0.25812541448702475,
"learning_rate": 1.3239875389408098e-05,
"loss": 0.2747,
"step": 2445
},
{
"epoch": 2.287783228217706,
"grad_norm": 0.2560376789403056,
"learning_rate": 1.3153340256143996e-05,
"loss": 0.2698,
"step": 2450
},
{
"epoch": 2.292455033870591,
"grad_norm": 0.26399251912690264,
"learning_rate": 1.3066805122879892e-05,
"loss": 0.2751,
"step": 2455
},
{
"epoch": 2.297126839523476,
"grad_norm": 0.23814257755917403,
"learning_rate": 1.2980269989615784e-05,
"loss": 0.261,
"step": 2460
},
{
"epoch": 2.3017986451763606,
"grad_norm": 0.2504090582744426,
"learning_rate": 1.289373485635168e-05,
"loss": 0.2763,
"step": 2465
},
{
"epoch": 2.3064704508292455,
"grad_norm": 0.2522405112875564,
"learning_rate": 1.2807199723087574e-05,
"loss": 0.2692,
"step": 2470
},
{
"epoch": 2.3111422564821305,
"grad_norm": 0.247482478204774,
"learning_rate": 1.272066458982347e-05,
"loss": 0.271,
"step": 2475
},
{
"epoch": 2.315814062135015,
"grad_norm": 0.25208432697718447,
"learning_rate": 1.2634129456559362e-05,
"loss": 0.2713,
"step": 2480
},
{
"epoch": 2.3204858677879,
"grad_norm": 0.2596669522816459,
"learning_rate": 1.254759432329526e-05,
"loss": 0.273,
"step": 2485
},
{
"epoch": 2.325157673440785,
"grad_norm": 0.2752122403792597,
"learning_rate": 1.2461059190031153e-05,
"loss": 0.2679,
"step": 2490
},
{
"epoch": 2.3298294790936698,
"grad_norm": 0.25305424916450886,
"learning_rate": 1.2374524056767047e-05,
"loss": 0.2699,
"step": 2495
},
{
"epoch": 2.3345012847465547,
"grad_norm": 0.24986399865013822,
"learning_rate": 1.2287988923502943e-05,
"loss": 0.2693,
"step": 2500
},
{
"epoch": 2.339173090399439,
"grad_norm": 0.24892959776998544,
"learning_rate": 1.2201453790238837e-05,
"loss": 0.2709,
"step": 2505
},
{
"epoch": 2.343844896052324,
"grad_norm": 0.2599132285443273,
"learning_rate": 1.2114918656974731e-05,
"loss": 0.2712,
"step": 2510
},
{
"epoch": 2.348516701705209,
"grad_norm": 0.24891162059815367,
"learning_rate": 1.2028383523710627e-05,
"loss": 0.2719,
"step": 2515
},
{
"epoch": 2.353188507358094,
"grad_norm": 0.2839857686291248,
"learning_rate": 1.1941848390446523e-05,
"loss": 0.2664,
"step": 2520
},
{
"epoch": 2.3578603130109785,
"grad_norm": 0.23816482029814176,
"learning_rate": 1.1855313257182417e-05,
"loss": 0.2622,
"step": 2525
},
{
"epoch": 2.3625321186638635,
"grad_norm": 0.23631470622212025,
"learning_rate": 1.1768778123918311e-05,
"loss": 0.2659,
"step": 2530
},
{
"epoch": 2.3672039243167484,
"grad_norm": 0.27010251736647883,
"learning_rate": 1.1682242990654207e-05,
"loss": 0.2716,
"step": 2535
},
{
"epoch": 2.3718757299696334,
"grad_norm": 0.25205702970219457,
"learning_rate": 1.15957078573901e-05,
"loss": 0.2723,
"step": 2540
},
{
"epoch": 2.3765475356225183,
"grad_norm": 0.25640449727385856,
"learning_rate": 1.1509172724125995e-05,
"loss": 0.2728,
"step": 2545
},
{
"epoch": 2.381219341275403,
"grad_norm": 0.24484226039145476,
"learning_rate": 1.142263759086189e-05,
"loss": 0.2726,
"step": 2550
},
{
"epoch": 2.3858911469282877,
"grad_norm": 0.2464355083120681,
"learning_rate": 1.1336102457597786e-05,
"loss": 0.2756,
"step": 2555
},
{
"epoch": 2.3905629525811727,
"grad_norm": 0.24744875149727918,
"learning_rate": 1.124956732433368e-05,
"loss": 0.2728,
"step": 2560
},
{
"epoch": 2.3952347582340576,
"grad_norm": 0.2403777893691903,
"learning_rate": 1.1163032191069575e-05,
"loss": 0.2718,
"step": 2565
},
{
"epoch": 2.399906563886942,
"grad_norm": 0.23574384761543135,
"learning_rate": 1.107649705780547e-05,
"loss": 0.2697,
"step": 2570
},
{
"epoch": 2.404578369539827,
"grad_norm": 0.25155760301757596,
"learning_rate": 1.0989961924541364e-05,
"loss": 0.2679,
"step": 2575
},
{
"epoch": 2.409250175192712,
"grad_norm": 0.24645374017868094,
"learning_rate": 1.0903426791277258e-05,
"loss": 0.2619,
"step": 2580
},
{
"epoch": 2.413921980845597,
"grad_norm": 0.2533648296600673,
"learning_rate": 1.0816891658013154e-05,
"loss": 0.2771,
"step": 2585
},
{
"epoch": 2.418593786498482,
"grad_norm": 0.2659076363497246,
"learning_rate": 1.0730356524749048e-05,
"loss": 0.2726,
"step": 2590
},
{
"epoch": 2.4232655921513664,
"grad_norm": 0.2385243006791889,
"learning_rate": 1.0643821391484942e-05,
"loss": 0.2681,
"step": 2595
},
{
"epoch": 2.4279373978042513,
"grad_norm": 0.26799927060856843,
"learning_rate": 1.0557286258220838e-05,
"loss": 0.2678,
"step": 2600
},
{
"epoch": 2.4326092034571363,
"grad_norm": 0.24972324275198554,
"learning_rate": 1.0470751124956734e-05,
"loss": 0.2737,
"step": 2605
},
{
"epoch": 2.437281009110021,
"grad_norm": 0.25419077865301654,
"learning_rate": 1.0384215991692628e-05,
"loss": 0.2664,
"step": 2610
},
{
"epoch": 2.4419528147629057,
"grad_norm": 0.2482159696188828,
"learning_rate": 1.0297680858428522e-05,
"loss": 0.2669,
"step": 2615
},
{
"epoch": 2.4466246204157907,
"grad_norm": 0.24839635814020702,
"learning_rate": 1.0211145725164418e-05,
"loss": 0.2757,
"step": 2620
},
{
"epoch": 2.4512964260686756,
"grad_norm": 0.25556105512354754,
"learning_rate": 1.0124610591900312e-05,
"loss": 0.271,
"step": 2625
},
{
"epoch": 2.4559682317215605,
"grad_norm": 0.2502049105887111,
"learning_rate": 1.0038075458636206e-05,
"loss": 0.2678,
"step": 2630
},
{
"epoch": 2.460640037374445,
"grad_norm": 0.24128844907979236,
"learning_rate": 9.951540325372102e-06,
"loss": 0.2703,
"step": 2635
},
{
"epoch": 2.46531184302733,
"grad_norm": 0.24603993464427193,
"learning_rate": 9.865005192107997e-06,
"loss": 0.2739,
"step": 2640
},
{
"epoch": 2.469983648680215,
"grad_norm": 0.24852247247203335,
"learning_rate": 9.778470058843891e-06,
"loss": 0.2639,
"step": 2645
},
{
"epoch": 2.4746554543331,
"grad_norm": 0.25309457119110484,
"learning_rate": 9.691934925579785e-06,
"loss": 0.2705,
"step": 2650
},
{
"epoch": 2.4793272599859844,
"grad_norm": 0.26264588665783367,
"learning_rate": 9.605399792315681e-06,
"loss": 0.2649,
"step": 2655
},
{
"epoch": 2.4839990656388693,
"grad_norm": 0.2325571711823743,
"learning_rate": 9.518864659051575e-06,
"loss": 0.2656,
"step": 2660
},
{
"epoch": 2.4886708712917542,
"grad_norm": 0.25896318627451675,
"learning_rate": 9.43232952578747e-06,
"loss": 0.2685,
"step": 2665
},
{
"epoch": 2.493342676944639,
"grad_norm": 0.23708628331763792,
"learning_rate": 9.345794392523365e-06,
"loss": 0.2645,
"step": 2670
},
{
"epoch": 2.498014482597524,
"grad_norm": 0.2424821823112542,
"learning_rate": 9.259259259259259e-06,
"loss": 0.2712,
"step": 2675
},
{
"epoch": 2.5026862882504086,
"grad_norm": 0.25133431772500003,
"learning_rate": 9.172724125995153e-06,
"loss": 0.2708,
"step": 2680
},
{
"epoch": 2.5073580939032936,
"grad_norm": 0.23820727708465494,
"learning_rate": 9.086188992731049e-06,
"loss": 0.2664,
"step": 2685
},
{
"epoch": 2.5120298995561785,
"grad_norm": 0.24708577320678266,
"learning_rate": 8.999653859466945e-06,
"loss": 0.2772,
"step": 2690
},
{
"epoch": 2.5167017052090634,
"grad_norm": 0.24669005270173416,
"learning_rate": 8.913118726202839e-06,
"loss": 0.2722,
"step": 2695
},
{
"epoch": 2.521373510861948,
"grad_norm": 7.282485495543622,
"learning_rate": 8.826583592938733e-06,
"loss": 0.2645,
"step": 2700
},
{
"epoch": 2.526045316514833,
"grad_norm": 0.2495842482489034,
"learning_rate": 8.740048459674629e-06,
"loss": 0.267,
"step": 2705
},
{
"epoch": 2.530717122167718,
"grad_norm": 0.46699969282295695,
"learning_rate": 8.653513326410523e-06,
"loss": 0.2733,
"step": 2710
},
{
"epoch": 2.5353889278206028,
"grad_norm": 0.2413515554248541,
"learning_rate": 8.566978193146417e-06,
"loss": 0.272,
"step": 2715
},
{
"epoch": 2.5400607334734877,
"grad_norm": 0.249606900867172,
"learning_rate": 8.480443059882312e-06,
"loss": 0.2681,
"step": 2720
},
{
"epoch": 2.544732539126372,
"grad_norm": 0.24067296655455211,
"learning_rate": 8.393907926618208e-06,
"loss": 0.2733,
"step": 2725
},
{
"epoch": 2.549404344779257,
"grad_norm": 0.23364465714549507,
"learning_rate": 8.307372793354102e-06,
"loss": 0.2623,
"step": 2730
},
{
"epoch": 2.554076150432142,
"grad_norm": 0.2472714440580257,
"learning_rate": 8.220837660089998e-06,
"loss": 0.2704,
"step": 2735
},
{
"epoch": 2.5587479560850266,
"grad_norm": 0.24111394737923167,
"learning_rate": 8.134302526825892e-06,
"loss": 0.2692,
"step": 2740
},
{
"epoch": 2.5634197617379115,
"grad_norm": 0.23176569456995924,
"learning_rate": 8.047767393561786e-06,
"loss": 0.269,
"step": 2745
},
{
"epoch": 2.5680915673907965,
"grad_norm": 0.2483944559685921,
"learning_rate": 7.96123226029768e-06,
"loss": 0.2697,
"step": 2750
},
{
"epoch": 2.5727633730436814,
"grad_norm": 0.24019971595328232,
"learning_rate": 7.874697127033576e-06,
"loss": 0.2639,
"step": 2755
},
{
"epoch": 2.5774351786965664,
"grad_norm": 0.2508812743970466,
"learning_rate": 7.78816199376947e-06,
"loss": 0.2805,
"step": 2760
},
{
"epoch": 2.5821069843494513,
"grad_norm": 0.24801037482956462,
"learning_rate": 7.701626860505364e-06,
"loss": 0.2675,
"step": 2765
},
{
"epoch": 2.586778790002336,
"grad_norm": 0.24198581158038515,
"learning_rate": 7.615091727241261e-06,
"loss": 0.2634,
"step": 2770
},
{
"epoch": 2.5914505956552207,
"grad_norm": 0.23307903892081475,
"learning_rate": 7.528556593977156e-06,
"loss": 0.2658,
"step": 2775
},
{
"epoch": 2.5961224013081057,
"grad_norm": 0.23343425875516594,
"learning_rate": 7.44202146071305e-06,
"loss": 0.2711,
"step": 2780
},
{
"epoch": 2.60079420696099,
"grad_norm": 0.23623800009068652,
"learning_rate": 7.355486327448945e-06,
"loss": 0.2715,
"step": 2785
},
{
"epoch": 2.605466012613875,
"grad_norm": 0.24373994858483788,
"learning_rate": 7.2689511941848395e-06,
"loss": 0.2685,
"step": 2790
},
{
"epoch": 2.61013781826676,
"grad_norm": 0.244684741216987,
"learning_rate": 7.1824160609207336e-06,
"loss": 0.2677,
"step": 2795
},
{
"epoch": 2.614809623919645,
"grad_norm": 0.24075277220196525,
"learning_rate": 7.0958809276566285e-06,
"loss": 0.2619,
"step": 2800
},
{
"epoch": 2.61948142957253,
"grad_norm": 0.2482116649977835,
"learning_rate": 7.009345794392523e-06,
"loss": 0.2663,
"step": 2805
},
{
"epoch": 2.624153235225415,
"grad_norm": 0.25285810002686593,
"learning_rate": 6.922810661128419e-06,
"loss": 0.2755,
"step": 2810
},
{
"epoch": 2.6288250408782994,
"grad_norm": 0.23411122241137453,
"learning_rate": 6.836275527864314e-06,
"loss": 0.2621,
"step": 2815
},
{
"epoch": 2.6334968465311843,
"grad_norm": 0.23822377371856152,
"learning_rate": 6.749740394600208e-06,
"loss": 0.2722,
"step": 2820
},
{
"epoch": 2.6381686521840693,
"grad_norm": 0.243925573785013,
"learning_rate": 6.663205261336103e-06,
"loss": 0.2681,
"step": 2825
},
{
"epoch": 2.6428404578369538,
"grad_norm": 0.22833307130473307,
"learning_rate": 6.576670128071998e-06,
"loss": 0.2698,
"step": 2830
},
{
"epoch": 2.6475122634898387,
"grad_norm": 0.2389187135245686,
"learning_rate": 6.490134994807892e-06,
"loss": 0.267,
"step": 2835
},
{
"epoch": 2.6521840691427236,
"grad_norm": 0.23466236980366012,
"learning_rate": 6.403599861543787e-06,
"loss": 0.2733,
"step": 2840
},
{
"epoch": 2.6568558747956086,
"grad_norm": 0.238596033352745,
"learning_rate": 6.317064728279681e-06,
"loss": 0.2623,
"step": 2845
},
{
"epoch": 2.6615276804484935,
"grad_norm": 0.2530179429486175,
"learning_rate": 6.230529595015577e-06,
"loss": 0.2662,
"step": 2850
},
{
"epoch": 2.666199486101378,
"grad_norm": 0.24655003236876927,
"learning_rate": 6.143994461751472e-06,
"loss": 0.2671,
"step": 2855
},
{
"epoch": 2.670871291754263,
"grad_norm": 0.2305205078252547,
"learning_rate": 6.057459328487366e-06,
"loss": 0.2638,
"step": 2860
},
{
"epoch": 2.675543097407148,
"grad_norm": 0.22907098832576295,
"learning_rate": 5.9709241952232614e-06,
"loss": 0.2643,
"step": 2865
},
{
"epoch": 2.680214903060033,
"grad_norm": 0.2588287051708774,
"learning_rate": 5.8843890619591555e-06,
"loss": 0.2708,
"step": 2870
},
{
"epoch": 2.6848867087129173,
"grad_norm": 0.23540646631994042,
"learning_rate": 5.79785392869505e-06,
"loss": 0.2699,
"step": 2875
},
{
"epoch": 2.6895585143658023,
"grad_norm": 0.24584007285596526,
"learning_rate": 5.711318795430945e-06,
"loss": 0.2664,
"step": 2880
},
{
"epoch": 2.6942303200186872,
"grad_norm": 0.2336146178847908,
"learning_rate": 5.62478366216684e-06,
"loss": 0.2637,
"step": 2885
},
{
"epoch": 2.698902125671572,
"grad_norm": 0.23331275021394282,
"learning_rate": 5.538248528902735e-06,
"loss": 0.2667,
"step": 2890
},
{
"epoch": 2.703573931324457,
"grad_norm": 0.23572768327654953,
"learning_rate": 5.451713395638629e-06,
"loss": 0.2687,
"step": 2895
},
{
"epoch": 2.7082457369773416,
"grad_norm": 0.23385706385348864,
"learning_rate": 5.365178262374524e-06,
"loss": 0.2721,
"step": 2900
},
{
"epoch": 2.7129175426302266,
"grad_norm": 0.25682014432107253,
"learning_rate": 5.278643129110419e-06,
"loss": 0.2702,
"step": 2905
},
{
"epoch": 2.7175893482831115,
"grad_norm": 0.23306411704916485,
"learning_rate": 5.192107995846314e-06,
"loss": 0.2623,
"step": 2910
},
{
"epoch": 2.722261153935996,
"grad_norm": 0.2420830929102186,
"learning_rate": 5.105572862582209e-06,
"loss": 0.2668,
"step": 2915
},
{
"epoch": 2.726932959588881,
"grad_norm": 0.2273406020691912,
"learning_rate": 5.019037729318103e-06,
"loss": 0.2604,
"step": 2920
},
{
"epoch": 2.731604765241766,
"grad_norm": 0.23352701790479655,
"learning_rate": 4.932502596053999e-06,
"loss": 0.2648,
"step": 2925
},
{
"epoch": 2.736276570894651,
"grad_norm": 0.23076697780407698,
"learning_rate": 4.845967462789893e-06,
"loss": 0.2696,
"step": 2930
},
{
"epoch": 2.7409483765475358,
"grad_norm": 0.23337097892917572,
"learning_rate": 4.759432329525788e-06,
"loss": 0.2579,
"step": 2935
},
{
"epoch": 2.7456201822004207,
"grad_norm": 0.22555654000735353,
"learning_rate": 4.6728971962616825e-06,
"loss": 0.2682,
"step": 2940
},
{
"epoch": 2.750291987853305,
"grad_norm": 0.23210086768204669,
"learning_rate": 4.586362062997577e-06,
"loss": 0.2637,
"step": 2945
},
{
"epoch": 2.75496379350619,
"grad_norm": 0.22558500238817342,
"learning_rate": 4.499826929733472e-06,
"loss": 0.2621,
"step": 2950
},
{
"epoch": 2.759635599159075,
"grad_norm": 0.22888822699838837,
"learning_rate": 4.413291796469366e-06,
"loss": 0.2622,
"step": 2955
},
{
"epoch": 2.7643074048119596,
"grad_norm": 0.23422576571530157,
"learning_rate": 4.326756663205261e-06,
"loss": 0.2648,
"step": 2960
},
{
"epoch": 2.7689792104648445,
"grad_norm": 0.25286861588459847,
"learning_rate": 4.240221529941156e-06,
"loss": 0.2709,
"step": 2965
},
{
"epoch": 2.7736510161177295,
"grad_norm": 0.2387247195237333,
"learning_rate": 4.153686396677051e-06,
"loss": 0.2671,
"step": 2970
},
{
"epoch": 2.7783228217706144,
"grad_norm": 0.24079518073816905,
"learning_rate": 4.067151263412946e-06,
"loss": 0.2708,
"step": 2975
},
{
"epoch": 2.7829946274234993,
"grad_norm": 0.24538112862438574,
"learning_rate": 3.98061613014884e-06,
"loss": 0.2672,
"step": 2980
},
{
"epoch": 2.7876664330763843,
"grad_norm": 0.23651173863220593,
"learning_rate": 3.894080996884735e-06,
"loss": 0.2672,
"step": 2985
},
{
"epoch": 2.792338238729269,
"grad_norm": 0.24226578759988052,
"learning_rate": 3.8075458636206304e-06,
"loss": 0.2627,
"step": 2990
},
{
"epoch": 2.7970100443821537,
"grad_norm": 0.2251099184524913,
"learning_rate": 3.721010730356525e-06,
"loss": 0.2582,
"step": 2995
},
{
"epoch": 2.8016818500350387,
"grad_norm": 0.23128992040804194,
"learning_rate": 3.6344755970924198e-06,
"loss": 0.2695,
"step": 3000
},
{
"epoch": 2.806353655687923,
"grad_norm": 0.24338716009456826,
"learning_rate": 3.5479404638283142e-06,
"loss": 0.2621,
"step": 3005
},
{
"epoch": 2.811025461340808,
"grad_norm": 0.23945832552804494,
"learning_rate": 3.4614053305642096e-06,
"loss": 0.2687,
"step": 3010
},
{
"epoch": 2.815697266993693,
"grad_norm": 0.2362116870382255,
"learning_rate": 3.374870197300104e-06,
"loss": 0.2735,
"step": 3015
},
{
"epoch": 2.820369072646578,
"grad_norm": 0.23352086316951318,
"learning_rate": 3.288335064035999e-06,
"loss": 0.2642,
"step": 3020
},
{
"epoch": 2.825040878299463,
"grad_norm": 0.23003009669088084,
"learning_rate": 3.2017999307718934e-06,
"loss": 0.2667,
"step": 3025
},
{
"epoch": 2.8297126839523474,
"grad_norm": 0.23904997734051717,
"learning_rate": 3.1152647975077884e-06,
"loss": 0.2647,
"step": 3030
},
{
"epoch": 2.8343844896052324,
"grad_norm": 0.22067820623162965,
"learning_rate": 3.028729664243683e-06,
"loss": 0.2596,
"step": 3035
},
{
"epoch": 2.8390562952581173,
"grad_norm": 1.0758039941756197,
"learning_rate": 2.9421945309795778e-06,
"loss": 0.2636,
"step": 3040
},
{
"epoch": 2.8437281009110023,
"grad_norm": 0.25884027526247666,
"learning_rate": 2.8556593977154727e-06,
"loss": 0.265,
"step": 3045
},
{
"epoch": 2.8483999065638868,
"grad_norm": 0.23440425991357747,
"learning_rate": 2.7691242644513676e-06,
"loss": 0.2639,
"step": 3050
},
{
"epoch": 2.8530717122167717,
"grad_norm": 0.22478784169345062,
"learning_rate": 2.682589131187262e-06,
"loss": 0.2744,
"step": 3055
},
{
"epoch": 2.8577435178696566,
"grad_norm": 0.2225135139306482,
"learning_rate": 2.596053997923157e-06,
"loss": 0.2688,
"step": 3060
},
{
"epoch": 2.8624153235225416,
"grad_norm": 0.2299647802881338,
"learning_rate": 2.5095188646590515e-06,
"loss": 0.2605,
"step": 3065
},
{
"epoch": 2.8670871291754265,
"grad_norm": 0.23139270482511712,
"learning_rate": 2.4229837313949464e-06,
"loss": 0.2635,
"step": 3070
},
{
"epoch": 2.871758934828311,
"grad_norm": 0.22647130179241579,
"learning_rate": 2.3364485981308413e-06,
"loss": 0.262,
"step": 3075
},
{
"epoch": 2.876430740481196,
"grad_norm": 0.23287256517338606,
"learning_rate": 2.249913464866736e-06,
"loss": 0.2727,
"step": 3080
},
{
"epoch": 2.881102546134081,
"grad_norm": 0.23572446735147778,
"learning_rate": 2.1633783316026307e-06,
"loss": 0.2636,
"step": 3085
},
{
"epoch": 2.8857743517869654,
"grad_norm": 0.2271413877332073,
"learning_rate": 2.0768431983385256e-06,
"loss": 0.262,
"step": 3090
},
{
"epoch": 2.8904461574398503,
"grad_norm": 0.23936041822322549,
"learning_rate": 1.99030806507442e-06,
"loss": 0.2637,
"step": 3095
},
{
"epoch": 2.8951179630927353,
"grad_norm": 0.23520041301905686,
"learning_rate": 1.9037729318103152e-06,
"loss": 0.2627,
"step": 3100
},
{
"epoch": 2.89978976874562,
"grad_norm": 0.23749601853452365,
"learning_rate": 1.8172377985462099e-06,
"loss": 0.2545,
"step": 3105
},
{
"epoch": 2.904461574398505,
"grad_norm": 0.2266785700002975,
"learning_rate": 1.7307026652821048e-06,
"loss": 0.2659,
"step": 3110
},
{
"epoch": 2.90913338005139,
"grad_norm": 0.24240448231800832,
"learning_rate": 1.6441675320179995e-06,
"loss": 0.2699,
"step": 3115
},
{
"epoch": 2.9138051857042746,
"grad_norm": 0.23019882389207785,
"learning_rate": 1.5576323987538942e-06,
"loss": 0.2649,
"step": 3120
},
{
"epoch": 2.9184769913571595,
"grad_norm": 0.23061531614411465,
"learning_rate": 1.4710972654897889e-06,
"loss": 0.2666,
"step": 3125
},
{
"epoch": 2.9231487970100445,
"grad_norm": 0.23604363667124315,
"learning_rate": 1.3845621322256838e-06,
"loss": 0.2716,
"step": 3130
},
{
"epoch": 2.927820602662929,
"grad_norm": 0.23300423400044568,
"learning_rate": 1.2980269989615785e-06,
"loss": 0.27,
"step": 3135
},
{
"epoch": 2.932492408315814,
"grad_norm": 0.22659021119636075,
"learning_rate": 1.2114918656974732e-06,
"loss": 0.2602,
"step": 3140
},
{
"epoch": 2.937164213968699,
"grad_norm": 0.230996061153359,
"learning_rate": 1.124956732433368e-06,
"loss": 0.2688,
"step": 3145
},
{
"epoch": 2.941836019621584,
"grad_norm": 0.22776039278473825,
"learning_rate": 1.0384215991692628e-06,
"loss": 0.2602,
"step": 3150
},
{
"epoch": 2.9465078252744688,
"grad_norm": 0.2388905450285887,
"learning_rate": 9.518864659051576e-07,
"loss": 0.2625,
"step": 3155
},
{
"epoch": 2.9511796309273537,
"grad_norm": 0.21982259002598517,
"learning_rate": 8.653513326410524e-07,
"loss": 0.2659,
"step": 3160
},
{
"epoch": 2.955851436580238,
"grad_norm": 0.23102989732987197,
"learning_rate": 7.788161993769471e-07,
"loss": 0.2627,
"step": 3165
},
{
"epoch": 2.960523242233123,
"grad_norm": 0.23568084348740576,
"learning_rate": 6.922810661128419e-07,
"loss": 0.267,
"step": 3170
},
{
"epoch": 2.965195047886008,
"grad_norm": 0.2349828995926893,
"learning_rate": 6.057459328487366e-07,
"loss": 0.2592,
"step": 3175
},
{
"epoch": 2.9698668535388926,
"grad_norm": 0.2341632542499381,
"learning_rate": 5.192107995846314e-07,
"loss": 0.2667,
"step": 3180
},
{
"epoch": 2.9745386591917775,
"grad_norm": 0.2368470965989083,
"learning_rate": 4.326756663205262e-07,
"loss": 0.2655,
"step": 3185
},
{
"epoch": 2.9792104648446625,
"grad_norm": 0.21703727243518728,
"learning_rate": 3.4614053305642095e-07,
"loss": 0.2614,
"step": 3190
},
{
"epoch": 2.9838822704975474,
"grad_norm": 0.21961380796313681,
"learning_rate": 2.596053997923157e-07,
"loss": 0.262,
"step": 3195
},
{
"epoch": 2.9885540761504323,
"grad_norm": 0.22635011463484578,
"learning_rate": 1.7307026652821047e-07,
"loss": 0.2673,
"step": 3200
},
{
"epoch": 2.993225881803317,
"grad_norm": 0.22800837794998072,
"learning_rate": 8.653513326410524e-08,
"loss": 0.2663,
"step": 3205
},
{
"epoch": 2.9978976874562018,
"grad_norm": 0.2193312660880487,
"learning_rate": 0.0,
"loss": 0.2612,
"step": 3210
},
{
"epoch": 2.9978976874562018,
"step": 3210,
"total_flos": 4.12909997976558e+18,
"train_loss": 0.36047850327328357,
"train_runtime": 33207.5596,
"train_samples_per_second": 3.094,
"train_steps_per_second": 0.097
}
],
"logging_steps": 5,
"max_steps": 3210,
"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": true
},
"attributes": {}
}
},
"total_flos": 4.12909997976558e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}