Files
minerva-dromedario-control-…/trainer_state.json
ModelHub XC e3da098f23 初始化项目,由ModelHub XC社区提供模型
Model: lukfre/minerva-dromedario-control-ckpt-4000
Source: Original Platform
2026-07-18 02:49:11 +08:00

2835 lines
69 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.8177450679750587,
"eval_steps": 500,
"global_step": 4000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002044362669937647,
"grad_norm": 467.8420479859713,
"learning_rate": 1.5306122448979592e-07,
"loss": 3.1603,
"step": 10
},
{
"epoch": 0.004088725339875294,
"grad_norm": 180.6683220314622,
"learning_rate": 3.231292517006803e-07,
"loss": 2.5124,
"step": 20
},
{
"epoch": 0.0061330880098129405,
"grad_norm": 45.788480106501375,
"learning_rate": 4.931972789115647e-07,
"loss": 1.4801,
"step": 30
},
{
"epoch": 0.008177450679750588,
"grad_norm": 5.980040546120992,
"learning_rate": 6.632653061224491e-07,
"loss": 0.9926,
"step": 40
},
{
"epoch": 0.010221813349688235,
"grad_norm": 1.887542595234843,
"learning_rate": 8.333333333333333e-07,
"loss": 0.904,
"step": 50
},
{
"epoch": 0.012266176019625881,
"grad_norm": 1.7279135258903726,
"learning_rate": 1.0034013605442177e-06,
"loss": 0.8435,
"step": 60
},
{
"epoch": 0.014310538689563529,
"grad_norm": 1.629559597810086,
"learning_rate": 1.1734693877551022e-06,
"loss": 0.8412,
"step": 70
},
{
"epoch": 0.016354901359501177,
"grad_norm": 1.6345729999324983,
"learning_rate": 1.3435374149659866e-06,
"loss": 0.8294,
"step": 80
},
{
"epoch": 0.01839926402943882,
"grad_norm": 1.7338287913512809,
"learning_rate": 1.513605442176871e-06,
"loss": 0.8203,
"step": 90
},
{
"epoch": 0.02044362669937647,
"grad_norm": 1.577028629634745,
"learning_rate": 1.6836734693877552e-06,
"loss": 0.8039,
"step": 100
},
{
"epoch": 0.022487989369314117,
"grad_norm": 1.928124610749223,
"learning_rate": 1.8537414965986398e-06,
"loss": 0.8099,
"step": 110
},
{
"epoch": 0.024532352039251762,
"grad_norm": 1.7053596508886784,
"learning_rate": 2.023809523809524e-06,
"loss": 0.8096,
"step": 120
},
{
"epoch": 0.02657671470918941,
"grad_norm": 1.6307623035013743,
"learning_rate": 2.1938775510204083e-06,
"loss": 0.8116,
"step": 130
},
{
"epoch": 0.028621077379127058,
"grad_norm": 1.7910588621004384,
"learning_rate": 2.363945578231293e-06,
"loss": 0.7583,
"step": 140
},
{
"epoch": 0.030665440049064702,
"grad_norm": 1.8412120520453148,
"learning_rate": 2.534013605442177e-06,
"loss": 0.7842,
"step": 150
},
{
"epoch": 0.032709802719002354,
"grad_norm": 1.8170848143732288,
"learning_rate": 2.7040816326530615e-06,
"loss": 0.8038,
"step": 160
},
{
"epoch": 0.034754165388939995,
"grad_norm": 1.6851570594759335,
"learning_rate": 2.8741496598639456e-06,
"loss": 0.8116,
"step": 170
},
{
"epoch": 0.03679852805887764,
"grad_norm": 1.6286844685502662,
"learning_rate": 3.04421768707483e-06,
"loss": 0.8323,
"step": 180
},
{
"epoch": 0.03884289072881529,
"grad_norm": 2.1067235857811375,
"learning_rate": 3.2142857142857147e-06,
"loss": 0.7927,
"step": 190
},
{
"epoch": 0.04088725339875294,
"grad_norm": 1.9402108723372633,
"learning_rate": 3.384353741496599e-06,
"loss": 0.8083,
"step": 200
},
{
"epoch": 0.04293161606869059,
"grad_norm": 1.6590434990945613,
"learning_rate": 3.5544217687074834e-06,
"loss": 0.789,
"step": 210
},
{
"epoch": 0.044975978738628235,
"grad_norm": 1.8035777317142234,
"learning_rate": 3.724489795918368e-06,
"loss": 0.8001,
"step": 220
},
{
"epoch": 0.04702034140856588,
"grad_norm": 1.7382396023521327,
"learning_rate": 3.894557823129252e-06,
"loss": 0.7926,
"step": 230
},
{
"epoch": 0.049064704078503524,
"grad_norm": 1.591750305106014,
"learning_rate": 4.064625850340137e-06,
"loss": 0.7711,
"step": 240
},
{
"epoch": 0.05110906674844117,
"grad_norm": 1.7233564410470317,
"learning_rate": 4.234693877551021e-06,
"loss": 0.7782,
"step": 250
},
{
"epoch": 0.05315342941837882,
"grad_norm": 1.6527760963055946,
"learning_rate": 4.404761904761905e-06,
"loss": 0.7708,
"step": 260
},
{
"epoch": 0.05519779208831647,
"grad_norm": 1.547497269455989,
"learning_rate": 4.5748299319727895e-06,
"loss": 0.7637,
"step": 270
},
{
"epoch": 0.057242154758254116,
"grad_norm": 1.453833604640337,
"learning_rate": 4.744897959183674e-06,
"loss": 0.7623,
"step": 280
},
{
"epoch": 0.059286517428191764,
"grad_norm": 1.713527764723503,
"learning_rate": 4.914965986394558e-06,
"loss": 0.7793,
"step": 290
},
{
"epoch": 0.061330880098129405,
"grad_norm": 1.6982980718912861,
"learning_rate": 4.99736564805058e-06,
"loss": 0.7825,
"step": 300
},
{
"epoch": 0.06337524276806705,
"grad_norm": 1.5810734240640687,
"learning_rate": 4.992096944151739e-06,
"loss": 0.7622,
"step": 310
},
{
"epoch": 0.06541960543800471,
"grad_norm": 1.649938058656486,
"learning_rate": 4.986828240252898e-06,
"loss": 0.7807,
"step": 320
},
{
"epoch": 0.06746396810794235,
"grad_norm": 1.6416865789482933,
"learning_rate": 4.981559536354058e-06,
"loss": 0.759,
"step": 330
},
{
"epoch": 0.06950833077787999,
"grad_norm": 1.66303043782447,
"learning_rate": 4.9762908324552166e-06,
"loss": 0.795,
"step": 340
},
{
"epoch": 0.07155269344781764,
"grad_norm": 1.5710516423316874,
"learning_rate": 4.9710221285563754e-06,
"loss": 0.8054,
"step": 350
},
{
"epoch": 0.07359705611775529,
"grad_norm": 1.6808382388694845,
"learning_rate": 4.965753424657534e-06,
"loss": 0.7696,
"step": 360
},
{
"epoch": 0.07564141878769294,
"grad_norm": 1.5932392419066819,
"learning_rate": 4.960484720758694e-06,
"loss": 0.7182,
"step": 370
},
{
"epoch": 0.07768578145763058,
"grad_norm": 1.6378838327061345,
"learning_rate": 4.955216016859853e-06,
"loss": 0.7673,
"step": 380
},
{
"epoch": 0.07973014412756824,
"grad_norm": 1.5969769402699843,
"learning_rate": 4.949947312961012e-06,
"loss": 0.7975,
"step": 390
},
{
"epoch": 0.08177450679750588,
"grad_norm": 1.5708101803016665,
"learning_rate": 4.9446786090621715e-06,
"loss": 0.77,
"step": 400
},
{
"epoch": 0.08381886946744352,
"grad_norm": 1.5919117823661149,
"learning_rate": 4.93940990516333e-06,
"loss": 0.7798,
"step": 410
},
{
"epoch": 0.08586323213738117,
"grad_norm": 1.572694411382701,
"learning_rate": 4.934141201264489e-06,
"loss": 0.7696,
"step": 420
},
{
"epoch": 0.08790759480731881,
"grad_norm": 1.5502608970510343,
"learning_rate": 4.928872497365648e-06,
"loss": 0.7799,
"step": 430
},
{
"epoch": 0.08995195747725647,
"grad_norm": 1.558281682200284,
"learning_rate": 4.923603793466808e-06,
"loss": 0.7438,
"step": 440
},
{
"epoch": 0.09199632014719411,
"grad_norm": 1.4603571713730354,
"learning_rate": 4.918335089567967e-06,
"loss": 0.7852,
"step": 450
},
{
"epoch": 0.09404068281713177,
"grad_norm": 1.5684816362903635,
"learning_rate": 4.9130663856691255e-06,
"loss": 0.7695,
"step": 460
},
{
"epoch": 0.0960850454870694,
"grad_norm": 1.5370325796477529,
"learning_rate": 4.907797681770285e-06,
"loss": 0.7669,
"step": 470
},
{
"epoch": 0.09812940815700705,
"grad_norm": 1.5865406422170756,
"learning_rate": 4.902528977871444e-06,
"loss": 0.7405,
"step": 480
},
{
"epoch": 0.1001737708269447,
"grad_norm": 1.5387206861673608,
"learning_rate": 4.897260273972603e-06,
"loss": 0.7902,
"step": 490
},
{
"epoch": 0.10221813349688234,
"grad_norm": 1.5961097840129412,
"learning_rate": 4.891991570073763e-06,
"loss": 0.744,
"step": 500
},
{
"epoch": 0.10426249616682,
"grad_norm": 1.4452548803166185,
"learning_rate": 4.8867228661749215e-06,
"loss": 0.7518,
"step": 510
},
{
"epoch": 0.10630685883675764,
"grad_norm": 1.5836573116777863,
"learning_rate": 4.88145416227608e-06,
"loss": 0.7839,
"step": 520
},
{
"epoch": 0.1083512215066953,
"grad_norm": 1.595244724301496,
"learning_rate": 4.876185458377239e-06,
"loss": 0.7259,
"step": 530
},
{
"epoch": 0.11039558417663294,
"grad_norm": 1.6208161175271276,
"learning_rate": 4.870916754478399e-06,
"loss": 0.7729,
"step": 540
},
{
"epoch": 0.11243994684657058,
"grad_norm": 1.4326014712696011,
"learning_rate": 4.865648050579558e-06,
"loss": 0.7093,
"step": 550
},
{
"epoch": 0.11448430951650823,
"grad_norm": 1.4288773041397789,
"learning_rate": 4.860379346680717e-06,
"loss": 0.7238,
"step": 560
},
{
"epoch": 0.11652867218644587,
"grad_norm": 1.460225435382298,
"learning_rate": 4.855110642781876e-06,
"loss": 0.7091,
"step": 570
},
{
"epoch": 0.11857303485638353,
"grad_norm": 1.5070284064954025,
"learning_rate": 4.849841938883035e-06,
"loss": 0.7411,
"step": 580
},
{
"epoch": 0.12061739752632117,
"grad_norm": 1.453711662286274,
"learning_rate": 4.844573234984194e-06,
"loss": 0.7158,
"step": 590
},
{
"epoch": 0.12266176019625881,
"grad_norm": 1.4438297459566305,
"learning_rate": 4.839304531085353e-06,
"loss": 0.7548,
"step": 600
},
{
"epoch": 0.12470612286619646,
"grad_norm": 1.6469833438164023,
"learning_rate": 4.834035827186513e-06,
"loss": 0.7722,
"step": 610
},
{
"epoch": 0.1267504855361341,
"grad_norm": 1.4937909258739377,
"learning_rate": 4.8287671232876716e-06,
"loss": 0.759,
"step": 620
},
{
"epoch": 0.12879484820607176,
"grad_norm": 1.4340846899747106,
"learning_rate": 4.82349841938883e-06,
"loss": 0.7327,
"step": 630
},
{
"epoch": 0.13083921087600942,
"grad_norm": 1.4801844335840377,
"learning_rate": 4.81822971548999e-06,
"loss": 0.7605,
"step": 640
},
{
"epoch": 0.13288357354594704,
"grad_norm": 1.8418348786613965,
"learning_rate": 4.812961011591149e-06,
"loss": 0.769,
"step": 650
},
{
"epoch": 0.1349279362158847,
"grad_norm": 1.448690324275757,
"learning_rate": 4.807692307692308e-06,
"loss": 0.7204,
"step": 660
},
{
"epoch": 0.13697229888582235,
"grad_norm": 1.4459826207280775,
"learning_rate": 4.8024236037934676e-06,
"loss": 0.7299,
"step": 670
},
{
"epoch": 0.13901666155575998,
"grad_norm": 1.5248487556923038,
"learning_rate": 4.7971548998946264e-06,
"loss": 0.7488,
"step": 680
},
{
"epoch": 0.14106102422569763,
"grad_norm": 1.7571942461065573,
"learning_rate": 4.791886195995785e-06,
"loss": 0.7246,
"step": 690
},
{
"epoch": 0.1431053868956353,
"grad_norm": 1.3781928156896945,
"learning_rate": 4.786617492096944e-06,
"loss": 0.7598,
"step": 700
},
{
"epoch": 0.14514974956557294,
"grad_norm": 1.5130148556601295,
"learning_rate": 4.781348788198104e-06,
"loss": 0.7557,
"step": 710
},
{
"epoch": 0.14719411223551057,
"grad_norm": 1.6624429419495477,
"learning_rate": 4.776080084299263e-06,
"loss": 0.7275,
"step": 720
},
{
"epoch": 0.14923847490544823,
"grad_norm": 1.480556750102346,
"learning_rate": 4.770811380400422e-06,
"loss": 0.6984,
"step": 730
},
{
"epoch": 0.15128283757538588,
"grad_norm": 1.4474264771291125,
"learning_rate": 4.765542676501581e-06,
"loss": 0.7477,
"step": 740
},
{
"epoch": 0.1533272002453235,
"grad_norm": 1.4721333818872286,
"learning_rate": 4.76027397260274e-06,
"loss": 0.7072,
"step": 750
},
{
"epoch": 0.15537156291526116,
"grad_norm": 1.748332544785494,
"learning_rate": 4.755005268703899e-06,
"loss": 0.7578,
"step": 760
},
{
"epoch": 0.15741592558519882,
"grad_norm": 1.3723416969650017,
"learning_rate": 4.749736564805058e-06,
"loss": 0.7674,
"step": 770
},
{
"epoch": 0.15946028825513647,
"grad_norm": 1.4921873776942336,
"learning_rate": 4.744467860906218e-06,
"loss": 0.7435,
"step": 780
},
{
"epoch": 0.1615046509250741,
"grad_norm": 1.4956413293314246,
"learning_rate": 4.7391991570073765e-06,
"loss": 0.7559,
"step": 790
},
{
"epoch": 0.16354901359501176,
"grad_norm": 1.5141633289542618,
"learning_rate": 4.733930453108535e-06,
"loss": 0.7541,
"step": 800
},
{
"epoch": 0.1655933762649494,
"grad_norm": 1.4142887761023697,
"learning_rate": 4.728661749209695e-06,
"loss": 0.743,
"step": 810
},
{
"epoch": 0.16763773893488704,
"grad_norm": 1.4433568453896806,
"learning_rate": 4.723393045310854e-06,
"loss": 0.7592,
"step": 820
},
{
"epoch": 0.1696821016048247,
"grad_norm": 1.5287564437758054,
"learning_rate": 4.718124341412013e-06,
"loss": 0.7515,
"step": 830
},
{
"epoch": 0.17172646427476235,
"grad_norm": 1.402433830466752,
"learning_rate": 4.7128556375131725e-06,
"loss": 0.7306,
"step": 840
},
{
"epoch": 0.1737708269447,
"grad_norm": 1.4070937612693248,
"learning_rate": 4.707586933614331e-06,
"loss": 0.7439,
"step": 850
},
{
"epoch": 0.17581518961463763,
"grad_norm": 1.5232291259335822,
"learning_rate": 4.70231822971549e-06,
"loss": 0.7279,
"step": 860
},
{
"epoch": 0.17785955228457528,
"grad_norm": 1.5200573872226932,
"learning_rate": 4.697049525816649e-06,
"loss": 0.7229,
"step": 870
},
{
"epoch": 0.17990391495451294,
"grad_norm": 1.5694234015163817,
"learning_rate": 4.691780821917809e-06,
"loss": 0.7455,
"step": 880
},
{
"epoch": 0.18194827762445057,
"grad_norm": 1.4413323471095139,
"learning_rate": 4.686512118018968e-06,
"loss": 0.7248,
"step": 890
},
{
"epoch": 0.18399264029438822,
"grad_norm": 1.5353111277232367,
"learning_rate": 4.6812434141201265e-06,
"loss": 0.7292,
"step": 900
},
{
"epoch": 0.18603700296432588,
"grad_norm": 1.4848347905739014,
"learning_rate": 4.675974710221286e-06,
"loss": 0.7232,
"step": 910
},
{
"epoch": 0.18808136563426353,
"grad_norm": 1.4730293552186593,
"learning_rate": 4.670706006322445e-06,
"loss": 0.7233,
"step": 920
},
{
"epoch": 0.19012572830420116,
"grad_norm": 1.5075112675107574,
"learning_rate": 4.665437302423604e-06,
"loss": 0.7223,
"step": 930
},
{
"epoch": 0.1921700909741388,
"grad_norm": 1.4578399239413515,
"learning_rate": 4.660168598524764e-06,
"loss": 0.6925,
"step": 940
},
{
"epoch": 0.19421445364407647,
"grad_norm": 1.5627951583472042,
"learning_rate": 4.6548998946259226e-06,
"loss": 0.743,
"step": 950
},
{
"epoch": 0.1962588163140141,
"grad_norm": 1.3306415927004691,
"learning_rate": 4.649631190727081e-06,
"loss": 0.7259,
"step": 960
},
{
"epoch": 0.19830317898395175,
"grad_norm": 1.456084367183415,
"learning_rate": 4.64436248682824e-06,
"loss": 0.6929,
"step": 970
},
{
"epoch": 0.2003475416538894,
"grad_norm": 1.4020218623539584,
"learning_rate": 4.6390937829294e-06,
"loss": 0.7285,
"step": 980
},
{
"epoch": 0.20239190432382706,
"grad_norm": 1.5551020089604901,
"learning_rate": 4.633825079030559e-06,
"loss": 0.7314,
"step": 990
},
{
"epoch": 0.2044362669937647,
"grad_norm": 1.5298861943993591,
"learning_rate": 4.628556375131718e-06,
"loss": 0.7127,
"step": 1000
},
{
"epoch": 0.20648062966370234,
"grad_norm": 1.4346648984001062,
"learning_rate": 4.6232876712328774e-06,
"loss": 0.7252,
"step": 1010
},
{
"epoch": 0.20852499233364,
"grad_norm": 1.2625190905714159,
"learning_rate": 4.618018967334036e-06,
"loss": 0.731,
"step": 1020
},
{
"epoch": 0.21056935500357762,
"grad_norm": 1.474199489062862,
"learning_rate": 4.612750263435195e-06,
"loss": 0.732,
"step": 1030
},
{
"epoch": 0.21261371767351528,
"grad_norm": 1.5771493582097231,
"learning_rate": 4.607481559536354e-06,
"loss": 0.7452,
"step": 1040
},
{
"epoch": 0.21465808034345293,
"grad_norm": 1.5416319010067572,
"learning_rate": 4.602212855637514e-06,
"loss": 0.7223,
"step": 1050
},
{
"epoch": 0.2167024430133906,
"grad_norm": 1.7824520328095972,
"learning_rate": 4.596944151738673e-06,
"loss": 0.7062,
"step": 1060
},
{
"epoch": 0.21874680568332822,
"grad_norm": 1.4470495746001502,
"learning_rate": 4.5916754478398315e-06,
"loss": 0.7062,
"step": 1070
},
{
"epoch": 0.22079116835326587,
"grad_norm": 1.4908213852908343,
"learning_rate": 4.586406743940991e-06,
"loss": 0.7343,
"step": 1080
},
{
"epoch": 0.22283553102320353,
"grad_norm": 1.3792828846951122,
"learning_rate": 4.58113804004215e-06,
"loss": 0.7244,
"step": 1090
},
{
"epoch": 0.22487989369314115,
"grad_norm": 1.4657553268313241,
"learning_rate": 4.575869336143309e-06,
"loss": 0.721,
"step": 1100
},
{
"epoch": 0.2269242563630788,
"grad_norm": 1.5012938958088435,
"learning_rate": 4.570600632244469e-06,
"loss": 0.6967,
"step": 1110
},
{
"epoch": 0.22896861903301646,
"grad_norm": 1.627910357778962,
"learning_rate": 4.5653319283456275e-06,
"loss": 0.71,
"step": 1120
},
{
"epoch": 0.23101298170295412,
"grad_norm": 1.5941181211234938,
"learning_rate": 4.560063224446786e-06,
"loss": 0.7128,
"step": 1130
},
{
"epoch": 0.23305734437289174,
"grad_norm": 1.4817199349041343,
"learning_rate": 4.554794520547945e-06,
"loss": 0.7344,
"step": 1140
},
{
"epoch": 0.2351017070428294,
"grad_norm": 1.6819119521864316,
"learning_rate": 4.549525816649105e-06,
"loss": 0.6833,
"step": 1150
},
{
"epoch": 0.23714606971276705,
"grad_norm": 1.5461532395938835,
"learning_rate": 4.544257112750264e-06,
"loss": 0.7249,
"step": 1160
},
{
"epoch": 0.23919043238270468,
"grad_norm": 1.4558900912809498,
"learning_rate": 4.538988408851423e-06,
"loss": 0.7114,
"step": 1170
},
{
"epoch": 0.24123479505264234,
"grad_norm": 1.538217648039571,
"learning_rate": 4.533719704952582e-06,
"loss": 0.7209,
"step": 1180
},
{
"epoch": 0.24327915772258,
"grad_norm": 1.4732581917622394,
"learning_rate": 4.528451001053741e-06,
"loss": 0.7092,
"step": 1190
},
{
"epoch": 0.24532352039251762,
"grad_norm": 1.7398779706760932,
"learning_rate": 4.523182297154901e-06,
"loss": 0.7416,
"step": 1200
},
{
"epoch": 0.24736788306245527,
"grad_norm": 1.4178208971948765,
"learning_rate": 4.517913593256059e-06,
"loss": 0.7236,
"step": 1210
},
{
"epoch": 0.24941224573239293,
"grad_norm": 1.4163112090352807,
"learning_rate": 4.512644889357219e-06,
"loss": 0.7427,
"step": 1220
},
{
"epoch": 0.2514566084023306,
"grad_norm": 1.423205541684213,
"learning_rate": 4.5073761854583775e-06,
"loss": 0.7097,
"step": 1230
},
{
"epoch": 0.2535009710722682,
"grad_norm": 1.4167827999767075,
"learning_rate": 4.502107481559536e-06,
"loss": 0.6978,
"step": 1240
},
{
"epoch": 0.2555453337422059,
"grad_norm": 1.4520662732845957,
"learning_rate": 4.496838777660696e-06,
"loss": 0.7332,
"step": 1250
},
{
"epoch": 0.2575896964121435,
"grad_norm": 1.3201725701560465,
"learning_rate": 4.491570073761855e-06,
"loss": 0.7214,
"step": 1260
},
{
"epoch": 0.25963405908208115,
"grad_norm": 1.509730775004563,
"learning_rate": 4.486301369863014e-06,
"loss": 0.6871,
"step": 1270
},
{
"epoch": 0.26167842175201883,
"grad_norm": 1.3421161682592169,
"learning_rate": 4.4810326659641736e-06,
"loss": 0.6908,
"step": 1280
},
{
"epoch": 0.26372278442195646,
"grad_norm": 1.3879780869567981,
"learning_rate": 4.475763962065332e-06,
"loss": 0.6687,
"step": 1290
},
{
"epoch": 0.2657671470918941,
"grad_norm": 1.3666502323196792,
"learning_rate": 4.470495258166491e-06,
"loss": 0.7152,
"step": 1300
},
{
"epoch": 0.26781150976183177,
"grad_norm": 1.4418900655185527,
"learning_rate": 4.46522655426765e-06,
"loss": 0.7217,
"step": 1310
},
{
"epoch": 0.2698558724317694,
"grad_norm": 1.4603147808672516,
"learning_rate": 4.45995785036881e-06,
"loss": 0.7049,
"step": 1320
},
{
"epoch": 0.271900235101707,
"grad_norm": 1.5551836750919699,
"learning_rate": 4.454689146469969e-06,
"loss": 0.7251,
"step": 1330
},
{
"epoch": 0.2739445977716447,
"grad_norm": 1.341528981011124,
"learning_rate": 4.449420442571128e-06,
"loss": 0.7066,
"step": 1340
},
{
"epoch": 0.27598896044158233,
"grad_norm": 1.4168699293732625,
"learning_rate": 4.444151738672287e-06,
"loss": 0.715,
"step": 1350
},
{
"epoch": 0.27803332311151996,
"grad_norm": 1.4303524713935523,
"learning_rate": 4.438883034773446e-06,
"loss": 0.689,
"step": 1360
},
{
"epoch": 0.28007768578145764,
"grad_norm": 1.4416520286839531,
"learning_rate": 4.433614330874606e-06,
"loss": 0.701,
"step": 1370
},
{
"epoch": 0.28212204845139527,
"grad_norm": 1.539243487119809,
"learning_rate": 4.428345626975764e-06,
"loss": 0.7282,
"step": 1380
},
{
"epoch": 0.28416641112133295,
"grad_norm": 1.3955768635762926,
"learning_rate": 4.423076923076924e-06,
"loss": 0.7173,
"step": 1390
},
{
"epoch": 0.2862107737912706,
"grad_norm": 1.3765621047548326,
"learning_rate": 4.4178082191780825e-06,
"loss": 0.7066,
"step": 1400
},
{
"epoch": 0.2882551364612082,
"grad_norm": 1.6208346065050854,
"learning_rate": 4.412539515279241e-06,
"loss": 0.7319,
"step": 1410
},
{
"epoch": 0.2902994991311459,
"grad_norm": 1.4342807513410571,
"learning_rate": 4.407270811380401e-06,
"loss": 0.703,
"step": 1420
},
{
"epoch": 0.2923438618010835,
"grad_norm": 1.4261879456535196,
"learning_rate": 4.40200210748156e-06,
"loss": 0.6747,
"step": 1430
},
{
"epoch": 0.29438822447102114,
"grad_norm": 1.48126107172508,
"learning_rate": 4.396733403582719e-06,
"loss": 0.7043,
"step": 1440
},
{
"epoch": 0.2964325871409588,
"grad_norm": 1.5800033463714298,
"learning_rate": 4.3914646996838785e-06,
"loss": 0.7042,
"step": 1450
},
{
"epoch": 0.29847694981089645,
"grad_norm": 1.478938481655847,
"learning_rate": 4.386195995785037e-06,
"loss": 0.678,
"step": 1460
},
{
"epoch": 0.3005213124808341,
"grad_norm": 1.360981414050118,
"learning_rate": 4.380927291886196e-06,
"loss": 0.6778,
"step": 1470
},
{
"epoch": 0.30256567515077176,
"grad_norm": 1.527838194368243,
"learning_rate": 4.375658587987355e-06,
"loss": 0.7299,
"step": 1480
},
{
"epoch": 0.3046100378207094,
"grad_norm": 1.3674067529238303,
"learning_rate": 4.370389884088515e-06,
"loss": 0.6994,
"step": 1490
},
{
"epoch": 0.306654400490647,
"grad_norm": 1.412787104214405,
"learning_rate": 4.365121180189674e-06,
"loss": 0.7101,
"step": 1500
},
{
"epoch": 0.3086987631605847,
"grad_norm": 1.3271182709132339,
"learning_rate": 4.3598524762908325e-06,
"loss": 0.6688,
"step": 1510
},
{
"epoch": 0.3107431258305223,
"grad_norm": 1.3669875778888532,
"learning_rate": 4.354583772391992e-06,
"loss": 0.7254,
"step": 1520
},
{
"epoch": 0.31278748850046,
"grad_norm": 1.5912158667313845,
"learning_rate": 4.349315068493151e-06,
"loss": 0.7045,
"step": 1530
},
{
"epoch": 0.31483185117039764,
"grad_norm": 1.4695496483876862,
"learning_rate": 4.344046364594311e-06,
"loss": 0.7041,
"step": 1540
},
{
"epoch": 0.31687621384033526,
"grad_norm": 1.5122124435761466,
"learning_rate": 4.338777660695469e-06,
"loss": 0.6899,
"step": 1550
},
{
"epoch": 0.31892057651027295,
"grad_norm": 1.4752177906639534,
"learning_rate": 4.3335089567966285e-06,
"loss": 0.7035,
"step": 1560
},
{
"epoch": 0.3209649391802106,
"grad_norm": 1.364249917048718,
"learning_rate": 4.328240252897787e-06,
"loss": 0.7276,
"step": 1570
},
{
"epoch": 0.3230093018501482,
"grad_norm": 1.6253917888345226,
"learning_rate": 4.322971548998946e-06,
"loss": 0.7155,
"step": 1580
},
{
"epoch": 0.3250536645200859,
"grad_norm": 1.4123132557143987,
"learning_rate": 4.317702845100106e-06,
"loss": 0.6822,
"step": 1590
},
{
"epoch": 0.3270980271900235,
"grad_norm": 1.4901871868584207,
"learning_rate": 4.312434141201265e-06,
"loss": 0.7019,
"step": 1600
},
{
"epoch": 0.32914238985996114,
"grad_norm": 1.502909767728457,
"learning_rate": 4.307165437302424e-06,
"loss": 0.7432,
"step": 1610
},
{
"epoch": 0.3311867525298988,
"grad_norm": 1.369411037385717,
"learning_rate": 4.301896733403583e-06,
"loss": 0.6931,
"step": 1620
},
{
"epoch": 0.33323111519983645,
"grad_norm": 1.3805153850385938,
"learning_rate": 4.296628029504742e-06,
"loss": 0.7029,
"step": 1630
},
{
"epoch": 0.3352754778697741,
"grad_norm": 1.3228042449466453,
"learning_rate": 4.291359325605901e-06,
"loss": 0.7242,
"step": 1640
},
{
"epoch": 0.33731984053971176,
"grad_norm": 1.4524707436582085,
"learning_rate": 4.28609062170706e-06,
"loss": 0.7277,
"step": 1650
},
{
"epoch": 0.3393642032096494,
"grad_norm": 1.4855876026211137,
"learning_rate": 4.28082191780822e-06,
"loss": 0.6947,
"step": 1660
},
{
"epoch": 0.341408565879587,
"grad_norm": 1.4679460471333967,
"learning_rate": 4.275553213909379e-06,
"loss": 0.6996,
"step": 1670
},
{
"epoch": 0.3434529285495247,
"grad_norm": 1.4461495003064762,
"learning_rate": 4.2702845100105374e-06,
"loss": 0.6841,
"step": 1680
},
{
"epoch": 0.3454972912194623,
"grad_norm": 1.3587233430593069,
"learning_rate": 4.265015806111697e-06,
"loss": 0.7006,
"step": 1690
},
{
"epoch": 0.3475416538894,
"grad_norm": 1.3558798681441402,
"learning_rate": 4.259747102212856e-06,
"loss": 0.7049,
"step": 1700
},
{
"epoch": 0.34958601655933763,
"grad_norm": 1.3962949631005122,
"learning_rate": 4.254478398314016e-06,
"loss": 0.6955,
"step": 1710
},
{
"epoch": 0.35163037922927526,
"grad_norm": 1.4831026682811195,
"learning_rate": 4.249209694415174e-06,
"loss": 0.701,
"step": 1720
},
{
"epoch": 0.35367474189921294,
"grad_norm": 1.3930139479335426,
"learning_rate": 4.2439409905163335e-06,
"loss": 0.6792,
"step": 1730
},
{
"epoch": 0.35571910456915057,
"grad_norm": 1.488639752569518,
"learning_rate": 4.238672286617492e-06,
"loss": 0.723,
"step": 1740
},
{
"epoch": 0.3577634672390882,
"grad_norm": 1.4158402533620718,
"learning_rate": 4.233403582718651e-06,
"loss": 0.6913,
"step": 1750
},
{
"epoch": 0.3598078299090259,
"grad_norm": 1.5118863040634951,
"learning_rate": 4.228134878819811e-06,
"loss": 0.72,
"step": 1760
},
{
"epoch": 0.3618521925789635,
"grad_norm": 1.4831963606876897,
"learning_rate": 4.22286617492097e-06,
"loss": 0.7044,
"step": 1770
},
{
"epoch": 0.36389655524890113,
"grad_norm": 1.389259571202407,
"learning_rate": 4.217597471022129e-06,
"loss": 0.7143,
"step": 1780
},
{
"epoch": 0.3659409179188388,
"grad_norm": 1.4283093305319636,
"learning_rate": 4.212328767123288e-06,
"loss": 0.7148,
"step": 1790
},
{
"epoch": 0.36798528058877644,
"grad_norm": 1.3678135321696594,
"learning_rate": 4.207060063224447e-06,
"loss": 0.6828,
"step": 1800
},
{
"epoch": 0.37002964325871407,
"grad_norm": 1.4068417828230946,
"learning_rate": 4.201791359325606e-06,
"loss": 0.7012,
"step": 1810
},
{
"epoch": 0.37207400592865175,
"grad_norm": 1.3085992435202767,
"learning_rate": 4.196522655426765e-06,
"loss": 0.6759,
"step": 1820
},
{
"epoch": 0.3741183685985894,
"grad_norm": 1.3434947282833654,
"learning_rate": 4.191253951527925e-06,
"loss": 0.7138,
"step": 1830
},
{
"epoch": 0.37616273126852706,
"grad_norm": 1.43798526438379,
"learning_rate": 4.1859852476290835e-06,
"loss": 0.7237,
"step": 1840
},
{
"epoch": 0.3782070939384647,
"grad_norm": 1.4630801379090654,
"learning_rate": 4.180716543730242e-06,
"loss": 0.7113,
"step": 1850
},
{
"epoch": 0.3802514566084023,
"grad_norm": 1.5869665043991625,
"learning_rate": 4.175447839831402e-06,
"loss": 0.7148,
"step": 1860
},
{
"epoch": 0.38229581927834,
"grad_norm": 1.348970204620055,
"learning_rate": 4.170179135932561e-06,
"loss": 0.6766,
"step": 1870
},
{
"epoch": 0.3843401819482776,
"grad_norm": 1.4708520817147068,
"learning_rate": 4.164910432033721e-06,
"loss": 0.688,
"step": 1880
},
{
"epoch": 0.38638454461821525,
"grad_norm": 1.3821621865030551,
"learning_rate": 4.1596417281348795e-06,
"loss": 0.6806,
"step": 1890
},
{
"epoch": 0.38842890728815294,
"grad_norm": 1.4668646156142915,
"learning_rate": 4.154373024236038e-06,
"loss": 0.705,
"step": 1900
},
{
"epoch": 0.39047326995809056,
"grad_norm": 1.2566997282788497,
"learning_rate": 4.149104320337197e-06,
"loss": 0.6791,
"step": 1910
},
{
"epoch": 0.3925176326280282,
"grad_norm": 1.369357146631333,
"learning_rate": 4.143835616438356e-06,
"loss": 0.7077,
"step": 1920
},
{
"epoch": 0.3945619952979659,
"grad_norm": 1.3926481215272302,
"learning_rate": 4.138566912539516e-06,
"loss": 0.6885,
"step": 1930
},
{
"epoch": 0.3966063579679035,
"grad_norm": 1.3385415725327532,
"learning_rate": 4.133298208640675e-06,
"loss": 0.7027,
"step": 1940
},
{
"epoch": 0.3986507206378411,
"grad_norm": 1.3676483477484742,
"learning_rate": 4.1280295047418336e-06,
"loss": 0.697,
"step": 1950
},
{
"epoch": 0.4006950833077788,
"grad_norm": 1.2959975154526457,
"learning_rate": 4.122760800842993e-06,
"loss": 0.6469,
"step": 1960
},
{
"epoch": 0.40273944597771644,
"grad_norm": 1.303310642152222,
"learning_rate": 4.117492096944152e-06,
"loss": 0.6935,
"step": 1970
},
{
"epoch": 0.4047838086476541,
"grad_norm": 1.3615406761771238,
"learning_rate": 4.112223393045311e-06,
"loss": 0.6832,
"step": 1980
},
{
"epoch": 0.40682817131759175,
"grad_norm": 1.3903122919723363,
"learning_rate": 4.10695468914647e-06,
"loss": 0.6697,
"step": 1990
},
{
"epoch": 0.4088725339875294,
"grad_norm": 1.4491529250242632,
"learning_rate": 4.10168598524763e-06,
"loss": 0.7317,
"step": 2000
},
{
"epoch": 0.41091689665746706,
"grad_norm": 1.3693171675759095,
"learning_rate": 4.0964172813487884e-06,
"loss": 0.7074,
"step": 2010
},
{
"epoch": 0.4129612593274047,
"grad_norm": 1.4159341111137298,
"learning_rate": 4.091148577449947e-06,
"loss": 0.6944,
"step": 2020
},
{
"epoch": 0.4150056219973423,
"grad_norm": 1.362733710994656,
"learning_rate": 4.085879873551107e-06,
"loss": 0.7032,
"step": 2030
},
{
"epoch": 0.41704998466728,
"grad_norm": 1.3380985509779906,
"learning_rate": 4.080611169652266e-06,
"loss": 0.7008,
"step": 2040
},
{
"epoch": 0.4190943473372176,
"grad_norm": 1.3768658965042804,
"learning_rate": 4.075342465753426e-06,
"loss": 0.6743,
"step": 2050
},
{
"epoch": 0.42113871000715525,
"grad_norm": 1.437811895962572,
"learning_rate": 4.0700737618545845e-06,
"loss": 0.6861,
"step": 2060
},
{
"epoch": 0.42318307267709293,
"grad_norm": 1.6772271215495203,
"learning_rate": 4.064805057955743e-06,
"loss": 0.6981,
"step": 2070
},
{
"epoch": 0.42522743534703056,
"grad_norm": 1.4687379384411468,
"learning_rate": 4.059536354056902e-06,
"loss": 0.6792,
"step": 2080
},
{
"epoch": 0.4272717980169682,
"grad_norm": 1.3937264905730289,
"learning_rate": 4.054267650158061e-06,
"loss": 0.6888,
"step": 2090
},
{
"epoch": 0.42931616068690587,
"grad_norm": 1.3257359348954583,
"learning_rate": 4.048998946259221e-06,
"loss": 0.668,
"step": 2100
},
{
"epoch": 0.4313605233568435,
"grad_norm": 1.2930236247816063,
"learning_rate": 4.04373024236038e-06,
"loss": 0.6763,
"step": 2110
},
{
"epoch": 0.4334048860267812,
"grad_norm": 1.3585941560219317,
"learning_rate": 4.0384615384615385e-06,
"loss": 0.671,
"step": 2120
},
{
"epoch": 0.4354492486967188,
"grad_norm": 1.4149435976595832,
"learning_rate": 4.033192834562698e-06,
"loss": 0.6683,
"step": 2130
},
{
"epoch": 0.43749361136665643,
"grad_norm": 1.400067547135137,
"learning_rate": 4.027924130663857e-06,
"loss": 0.6859,
"step": 2140
},
{
"epoch": 0.4395379740365941,
"grad_norm": 1.6072641958383649,
"learning_rate": 4.022655426765016e-06,
"loss": 0.7131,
"step": 2150
},
{
"epoch": 0.44158233670653174,
"grad_norm": 1.489767162533831,
"learning_rate": 4.017386722866175e-06,
"loss": 0.6821,
"step": 2160
},
{
"epoch": 0.44362669937646937,
"grad_norm": 1.4543637031315133,
"learning_rate": 4.0121180189673345e-06,
"loss": 0.6772,
"step": 2170
},
{
"epoch": 0.44567106204640705,
"grad_norm": 1.3570413342060814,
"learning_rate": 4.006849315068493e-06,
"loss": 0.7344,
"step": 2180
},
{
"epoch": 0.4477154247163447,
"grad_norm": 1.4339215297801629,
"learning_rate": 4.001580611169652e-06,
"loss": 0.6694,
"step": 2190
},
{
"epoch": 0.4497597873862823,
"grad_norm": 1.3510307320401556,
"learning_rate": 3.996311907270812e-06,
"loss": 0.6934,
"step": 2200
},
{
"epoch": 0.45180415005622,
"grad_norm": 1.3785304754241854,
"learning_rate": 3.991043203371971e-06,
"loss": 0.6612,
"step": 2210
},
{
"epoch": 0.4538485127261576,
"grad_norm": 1.3833286516401884,
"learning_rate": 3.9857744994731305e-06,
"loss": 0.6912,
"step": 2220
},
{
"epoch": 0.45589287539609524,
"grad_norm": 1.3097825144941337,
"learning_rate": 3.980505795574289e-06,
"loss": 0.6916,
"step": 2230
},
{
"epoch": 0.4579372380660329,
"grad_norm": 1.4586573527766178,
"learning_rate": 3.975237091675448e-06,
"loss": 0.6793,
"step": 2240
},
{
"epoch": 0.45998160073597055,
"grad_norm": 1.3576362407800902,
"learning_rate": 3.969968387776607e-06,
"loss": 0.7128,
"step": 2250
},
{
"epoch": 0.46202596340590824,
"grad_norm": 1.350129350260781,
"learning_rate": 3.964699683877766e-06,
"loss": 0.6754,
"step": 2260
},
{
"epoch": 0.46407032607584586,
"grad_norm": 1.457788945328202,
"learning_rate": 3.959430979978926e-06,
"loss": 0.6694,
"step": 2270
},
{
"epoch": 0.4661146887457835,
"grad_norm": 1.3669637553089713,
"learning_rate": 3.9541622760800846e-06,
"loss": 0.7033,
"step": 2280
},
{
"epoch": 0.4681590514157212,
"grad_norm": 1.4407380691037208,
"learning_rate": 3.9488935721812434e-06,
"loss": 0.6779,
"step": 2290
},
{
"epoch": 0.4702034140856588,
"grad_norm": 1.4212407161362186,
"learning_rate": 3.943624868282403e-06,
"loss": 0.7051,
"step": 2300
},
{
"epoch": 0.4722477767555964,
"grad_norm": 1.3963384583891147,
"learning_rate": 3.938356164383562e-06,
"loss": 0.6876,
"step": 2310
},
{
"epoch": 0.4742921394255341,
"grad_norm": 1.433285390320458,
"learning_rate": 3.933087460484722e-06,
"loss": 0.6969,
"step": 2320
},
{
"epoch": 0.47633650209547174,
"grad_norm": 1.3772450144882966,
"learning_rate": 3.92781875658588e-06,
"loss": 0.6903,
"step": 2330
},
{
"epoch": 0.47838086476540936,
"grad_norm": 1.3847800864838684,
"learning_rate": 3.9225500526870394e-06,
"loss": 0.7042,
"step": 2340
},
{
"epoch": 0.48042522743534705,
"grad_norm": 1.358584865107043,
"learning_rate": 3.917281348788198e-06,
"loss": 0.6526,
"step": 2350
},
{
"epoch": 0.4824695901052847,
"grad_norm": 1.3522365206947125,
"learning_rate": 3.912012644889357e-06,
"loss": 0.6643,
"step": 2360
},
{
"epoch": 0.4845139527752223,
"grad_norm": 1.4404898667839328,
"learning_rate": 3.906743940990517e-06,
"loss": 0.6516,
"step": 2370
},
{
"epoch": 0.48655831544516,
"grad_norm": 1.3327590794794255,
"learning_rate": 3.901475237091676e-06,
"loss": 0.6615,
"step": 2380
},
{
"epoch": 0.4886026781150976,
"grad_norm": 1.4036985321977502,
"learning_rate": 3.896206533192835e-06,
"loss": 0.7056,
"step": 2390
},
{
"epoch": 0.49064704078503524,
"grad_norm": 1.3978682115700392,
"learning_rate": 3.890937829293994e-06,
"loss": 0.6798,
"step": 2400
},
{
"epoch": 0.4926914034549729,
"grad_norm": 1.3694592600826894,
"learning_rate": 3.885669125395153e-06,
"loss": 0.6955,
"step": 2410
},
{
"epoch": 0.49473576612491055,
"grad_norm": 1.534615499212649,
"learning_rate": 3.880400421496312e-06,
"loss": 0.6732,
"step": 2420
},
{
"epoch": 0.49678012879484823,
"grad_norm": 1.6237656566503973,
"learning_rate": 3.875131717597471e-06,
"loss": 0.6691,
"step": 2430
},
{
"epoch": 0.49882449146478586,
"grad_norm": 1.3584905006089194,
"learning_rate": 3.869863013698631e-06,
"loss": 0.671,
"step": 2440
},
{
"epoch": 0.5008688541347235,
"grad_norm": 1.3870453341105176,
"learning_rate": 3.8645943097997895e-06,
"loss": 0.7085,
"step": 2450
},
{
"epoch": 0.5029132168046612,
"grad_norm": 1.480038331338043,
"learning_rate": 3.859325605900948e-06,
"loss": 0.6935,
"step": 2460
},
{
"epoch": 0.5049575794745987,
"grad_norm": 1.4495212398517232,
"learning_rate": 3.854056902002108e-06,
"loss": 0.7247,
"step": 2470
},
{
"epoch": 0.5070019421445364,
"grad_norm": 1.3208820369781498,
"learning_rate": 3.848788198103267e-06,
"loss": 0.6561,
"step": 2480
},
{
"epoch": 0.5090463048144741,
"grad_norm": 1.4863421751902246,
"learning_rate": 3.843519494204427e-06,
"loss": 0.708,
"step": 2490
},
{
"epoch": 0.5110906674844118,
"grad_norm": 1.622089485174737,
"learning_rate": 3.838250790305585e-06,
"loss": 0.6863,
"step": 2500
},
{
"epoch": 0.5131350301543494,
"grad_norm": 1.4321597349043875,
"learning_rate": 3.832982086406744e-06,
"loss": 0.7172,
"step": 2510
},
{
"epoch": 0.515179392824287,
"grad_norm": 1.315011237475376,
"learning_rate": 3.827713382507903e-06,
"loss": 0.6697,
"step": 2520
},
{
"epoch": 0.5172237554942247,
"grad_norm": 1.3379443624554566,
"learning_rate": 3.822444678609062e-06,
"loss": 0.6816,
"step": 2530
},
{
"epoch": 0.5192681181641623,
"grad_norm": 1.5016005034614845,
"learning_rate": 3.817175974710222e-06,
"loss": 0.7061,
"step": 2540
},
{
"epoch": 0.5213124808341,
"grad_norm": 1.5384303170968157,
"learning_rate": 3.8119072708113807e-06,
"loss": 0.6554,
"step": 2550
},
{
"epoch": 0.5233568435040377,
"grad_norm": 1.3588393773478071,
"learning_rate": 3.80663856691254e-06,
"loss": 0.6603,
"step": 2560
},
{
"epoch": 0.5254012061739752,
"grad_norm": 1.3003339383360144,
"learning_rate": 3.8013698630136993e-06,
"loss": 0.6781,
"step": 2570
},
{
"epoch": 0.5274455688439129,
"grad_norm": 1.3916549366064692,
"learning_rate": 3.796101159114858e-06,
"loss": 0.6678,
"step": 2580
},
{
"epoch": 0.5294899315138506,
"grad_norm": 1.2762940290145008,
"learning_rate": 3.790832455216017e-06,
"loss": 0.6987,
"step": 2590
},
{
"epoch": 0.5315342941837882,
"grad_norm": 1.4409139365893981,
"learning_rate": 3.7855637513171763e-06,
"loss": 0.6619,
"step": 2600
},
{
"epoch": 0.5335786568537259,
"grad_norm": 1.323178615914982,
"learning_rate": 3.780295047418335e-06,
"loss": 0.6589,
"step": 2610
},
{
"epoch": 0.5356230195236635,
"grad_norm": 1.4836716706529218,
"learning_rate": 3.7750263435194944e-06,
"loss": 0.6637,
"step": 2620
},
{
"epoch": 0.5376673821936011,
"grad_norm": 1.3362508348859956,
"learning_rate": 3.7697576396206537e-06,
"loss": 0.6737,
"step": 2630
},
{
"epoch": 0.5397117448635388,
"grad_norm": 1.389558927355723,
"learning_rate": 3.764488935721813e-06,
"loss": 0.7072,
"step": 2640
},
{
"epoch": 0.5417561075334765,
"grad_norm": 1.378839791592441,
"learning_rate": 3.759220231822972e-06,
"loss": 0.7119,
"step": 2650
},
{
"epoch": 0.543800470203414,
"grad_norm": 1.3647504902674432,
"learning_rate": 3.753951527924131e-06,
"loss": 0.646,
"step": 2660
},
{
"epoch": 0.5458448328733517,
"grad_norm": 1.3558201072781753,
"learning_rate": 3.74868282402529e-06,
"loss": 0.7069,
"step": 2670
},
{
"epoch": 0.5478891955432894,
"grad_norm": 1.4635109144339562,
"learning_rate": 3.743414120126449e-06,
"loss": 0.7142,
"step": 2680
},
{
"epoch": 0.549933558213227,
"grad_norm": 1.3670228719733817,
"learning_rate": 3.738145416227608e-06,
"loss": 0.6572,
"step": 2690
},
{
"epoch": 0.5519779208831647,
"grad_norm": 1.319641365924929,
"learning_rate": 3.7328767123287675e-06,
"loss": 0.6899,
"step": 2700
},
{
"epoch": 0.5540222835531023,
"grad_norm": 1.5128587318231028,
"learning_rate": 3.7276080084299263e-06,
"loss": 0.668,
"step": 2710
},
{
"epoch": 0.5560666462230399,
"grad_norm": 1.3037349266967604,
"learning_rate": 3.7223393045310856e-06,
"loss": 0.6563,
"step": 2720
},
{
"epoch": 0.5581110088929776,
"grad_norm": 1.3381540109590586,
"learning_rate": 3.717070600632245e-06,
"loss": 0.6978,
"step": 2730
},
{
"epoch": 0.5601553715629153,
"grad_norm": 1.4420842593534189,
"learning_rate": 3.711801896733404e-06,
"loss": 0.6564,
"step": 2740
},
{
"epoch": 0.5621997342328529,
"grad_norm": 1.4132777112379162,
"learning_rate": 3.706533192834563e-06,
"loss": 0.6752,
"step": 2750
},
{
"epoch": 0.5642440969027905,
"grad_norm": 1.2826384702331286,
"learning_rate": 3.701264488935722e-06,
"loss": 0.7069,
"step": 2760
},
{
"epoch": 0.5662884595727282,
"grad_norm": 1.3462842850419952,
"learning_rate": 3.695995785036881e-06,
"loss": 0.6806,
"step": 2770
},
{
"epoch": 0.5683328222426659,
"grad_norm": 1.400135301741758,
"learning_rate": 3.69072708113804e-06,
"loss": 0.67,
"step": 2780
},
{
"epoch": 0.5703771849126035,
"grad_norm": 1.300384767776387,
"learning_rate": 3.6854583772391994e-06,
"loss": 0.6488,
"step": 2790
},
{
"epoch": 0.5724215475825412,
"grad_norm": 1.4420542400469645,
"learning_rate": 3.6801896733403586e-06,
"loss": 0.7177,
"step": 2800
},
{
"epoch": 0.5744659102524788,
"grad_norm": 1.3555437401937929,
"learning_rate": 3.674920969441518e-06,
"loss": 0.6774,
"step": 2810
},
{
"epoch": 0.5765102729224164,
"grad_norm": 1.301363421701744,
"learning_rate": 3.669652265542677e-06,
"loss": 0.6693,
"step": 2820
},
{
"epoch": 0.5785546355923541,
"grad_norm": 1.4158888402469691,
"learning_rate": 3.664383561643836e-06,
"loss": 0.6719,
"step": 2830
},
{
"epoch": 0.5805989982622918,
"grad_norm": 1.4199777950400523,
"learning_rate": 3.6591148577449954e-06,
"loss": 0.6972,
"step": 2840
},
{
"epoch": 0.5826433609322293,
"grad_norm": 1.4123555413227957,
"learning_rate": 3.653846153846154e-06,
"loss": 0.6916,
"step": 2850
},
{
"epoch": 0.584687723602167,
"grad_norm": 1.5101112083737727,
"learning_rate": 3.648577449947313e-06,
"loss": 0.6549,
"step": 2860
},
{
"epoch": 0.5867320862721047,
"grad_norm": 1.3130628951797034,
"learning_rate": 3.6433087460484724e-06,
"loss": 0.6707,
"step": 2870
},
{
"epoch": 0.5887764489420423,
"grad_norm": 1.3776545485057627,
"learning_rate": 3.6380400421496313e-06,
"loss": 0.6714,
"step": 2880
},
{
"epoch": 0.59082081161198,
"grad_norm": 1.5654287916409135,
"learning_rate": 3.6327713382507905e-06,
"loss": 0.7065,
"step": 2890
},
{
"epoch": 0.5928651742819177,
"grad_norm": 1.2458428326623017,
"learning_rate": 3.62750263435195e-06,
"loss": 0.6847,
"step": 2900
},
{
"epoch": 0.5949095369518552,
"grad_norm": 1.330414522380301,
"learning_rate": 3.622233930453109e-06,
"loss": 0.6524,
"step": 2910
},
{
"epoch": 0.5969538996217929,
"grad_norm": 1.437641125230679,
"learning_rate": 3.616965226554268e-06,
"loss": 0.6529,
"step": 2920
},
{
"epoch": 0.5989982622917306,
"grad_norm": 1.682882623947379,
"learning_rate": 3.611696522655427e-06,
"loss": 0.6876,
"step": 2930
},
{
"epoch": 0.6010426249616682,
"grad_norm": 1.3985036409613092,
"learning_rate": 3.606427818756586e-06,
"loss": 0.6792,
"step": 2940
},
{
"epoch": 0.6030869876316058,
"grad_norm": 1.3240771426220619,
"learning_rate": 3.601159114857745e-06,
"loss": 0.6683,
"step": 2950
},
{
"epoch": 0.6051313503015435,
"grad_norm": 1.273673753027315,
"learning_rate": 3.5958904109589043e-06,
"loss": 0.6641,
"step": 2960
},
{
"epoch": 0.6071757129714811,
"grad_norm": 1.2574527692994326,
"learning_rate": 3.5906217070600636e-06,
"loss": 0.6292,
"step": 2970
},
{
"epoch": 0.6092200756414188,
"grad_norm": 1.4215397222201802,
"learning_rate": 3.585353003161223e-06,
"loss": 0.6721,
"step": 2980
},
{
"epoch": 0.6112644383113565,
"grad_norm": 1.3791823195660944,
"learning_rate": 3.5800842992623817e-06,
"loss": 0.6767,
"step": 2990
},
{
"epoch": 0.613308800981294,
"grad_norm": 1.3861129240221604,
"learning_rate": 3.574815595363541e-06,
"loss": 0.6775,
"step": 3000
},
{
"epoch": 0.6153531636512317,
"grad_norm": 1.3301108610445043,
"learning_rate": 3.5695468914647003e-06,
"loss": 0.7034,
"step": 3010
},
{
"epoch": 0.6173975263211694,
"grad_norm": 1.4554776901052031,
"learning_rate": 3.5642781875658588e-06,
"loss": 0.668,
"step": 3020
},
{
"epoch": 0.619441888991107,
"grad_norm": 1.4029675813428577,
"learning_rate": 3.559009483667018e-06,
"loss": 0.6986,
"step": 3030
},
{
"epoch": 0.6214862516610447,
"grad_norm": 1.3743431932306527,
"learning_rate": 3.5537407797681773e-06,
"loss": 0.6578,
"step": 3040
},
{
"epoch": 0.6235306143309823,
"grad_norm": 1.3328295815320794,
"learning_rate": 3.548472075869336e-06,
"loss": 0.6544,
"step": 3050
},
{
"epoch": 0.62557497700092,
"grad_norm": 1.6433267878189008,
"learning_rate": 3.5432033719704955e-06,
"loss": 0.6821,
"step": 3060
},
{
"epoch": 0.6276193396708576,
"grad_norm": 1.2469707346304197,
"learning_rate": 3.5379346680716548e-06,
"loss": 0.7057,
"step": 3070
},
{
"epoch": 0.6296637023407953,
"grad_norm": 1.3487475825280903,
"learning_rate": 3.532665964172814e-06,
"loss": 0.668,
"step": 3080
},
{
"epoch": 0.631708065010733,
"grad_norm": 1.5117978850168854,
"learning_rate": 3.527397260273973e-06,
"loss": 0.677,
"step": 3090
},
{
"epoch": 0.6337524276806705,
"grad_norm": 1.2808584972233925,
"learning_rate": 3.5221285563751318e-06,
"loss": 0.674,
"step": 3100
},
{
"epoch": 0.6357967903506082,
"grad_norm": 1.3775114234163044,
"learning_rate": 3.516859852476291e-06,
"loss": 0.6743,
"step": 3110
},
{
"epoch": 0.6378411530205459,
"grad_norm": 1.341623034846183,
"learning_rate": 3.51159114857745e-06,
"loss": 0.6658,
"step": 3120
},
{
"epoch": 0.6398855156904835,
"grad_norm": 1.6183504834130353,
"learning_rate": 3.5063224446786092e-06,
"loss": 0.6303,
"step": 3130
},
{
"epoch": 0.6419298783604211,
"grad_norm": 1.4263155365983196,
"learning_rate": 3.5010537407797685e-06,
"loss": 0.6285,
"step": 3140
},
{
"epoch": 0.6439742410303588,
"grad_norm": 1.1676883088867758,
"learning_rate": 3.495785036880928e-06,
"loss": 0.6533,
"step": 3150
},
{
"epoch": 0.6460186037002964,
"grad_norm": 1.2961520817478958,
"learning_rate": 3.4905163329820867e-06,
"loss": 0.6555,
"step": 3160
},
{
"epoch": 0.6480629663702341,
"grad_norm": 1.453862702425531,
"learning_rate": 3.485247629083246e-06,
"loss": 0.6785,
"step": 3170
},
{
"epoch": 0.6501073290401718,
"grad_norm": 1.3458416687967194,
"learning_rate": 3.4799789251844052e-06,
"loss": 0.6638,
"step": 3180
},
{
"epoch": 0.6521516917101093,
"grad_norm": 1.3135808692691582,
"learning_rate": 3.4747102212855637e-06,
"loss": 0.6658,
"step": 3190
},
{
"epoch": 0.654196054380047,
"grad_norm": 1.3202178840709589,
"learning_rate": 3.469441517386723e-06,
"loss": 0.6725,
"step": 3200
},
{
"epoch": 0.6562404170499847,
"grad_norm": 1.416930312668468,
"learning_rate": 3.4641728134878823e-06,
"loss": 0.6794,
"step": 3210
},
{
"epoch": 0.6582847797199223,
"grad_norm": 1.3361755884017856,
"learning_rate": 3.458904109589041e-06,
"loss": 0.6367,
"step": 3220
},
{
"epoch": 0.66032914238986,
"grad_norm": 1.329981905426958,
"learning_rate": 3.4536354056902004e-06,
"loss": 0.712,
"step": 3230
},
{
"epoch": 0.6623735050597976,
"grad_norm": 1.3861042268701549,
"learning_rate": 3.4483667017913597e-06,
"loss": 0.6738,
"step": 3240
},
{
"epoch": 0.6644178677297352,
"grad_norm": 1.2121454007876893,
"learning_rate": 3.443097997892519e-06,
"loss": 0.646,
"step": 3250
},
{
"epoch": 0.6664622303996729,
"grad_norm": 1.380721073718238,
"learning_rate": 3.437829293993678e-06,
"loss": 0.6802,
"step": 3260
},
{
"epoch": 0.6685065930696106,
"grad_norm": 1.4054562491580906,
"learning_rate": 3.432560590094837e-06,
"loss": 0.6412,
"step": 3270
},
{
"epoch": 0.6705509557395481,
"grad_norm": 1.3619708416116894,
"learning_rate": 3.427291886195996e-06,
"loss": 0.6758,
"step": 3280
},
{
"epoch": 0.6725953184094858,
"grad_norm": 3.3907787744932185,
"learning_rate": 3.422023182297155e-06,
"loss": 0.6752,
"step": 3290
},
{
"epoch": 0.6746396810794235,
"grad_norm": 1.411041184126084,
"learning_rate": 3.416754478398314e-06,
"loss": 0.6537,
"step": 3300
},
{
"epoch": 0.6766840437493611,
"grad_norm": 1.2585015748132544,
"learning_rate": 3.4114857744994734e-06,
"loss": 0.6845,
"step": 3310
},
{
"epoch": 0.6787284064192988,
"grad_norm": 1.3374486006457051,
"learning_rate": 3.4062170706006327e-06,
"loss": 0.6628,
"step": 3320
},
{
"epoch": 0.6807727690892365,
"grad_norm": 1.4769780124243537,
"learning_rate": 3.4009483667017916e-06,
"loss": 0.6562,
"step": 3330
},
{
"epoch": 0.682817131759174,
"grad_norm": 1.4456460069972545,
"learning_rate": 3.395679662802951e-06,
"loss": 0.6635,
"step": 3340
},
{
"epoch": 0.6848614944291117,
"grad_norm": 1.351297679131553,
"learning_rate": 3.39041095890411e-06,
"loss": 0.692,
"step": 3350
},
{
"epoch": 0.6869058570990494,
"grad_norm": 1.403980210779944,
"learning_rate": 3.3851422550052686e-06,
"loss": 0.6629,
"step": 3360
},
{
"epoch": 0.6889502197689871,
"grad_norm": 1.3925873005398655,
"learning_rate": 3.379873551106428e-06,
"loss": 0.6901,
"step": 3370
},
{
"epoch": 0.6909945824389246,
"grad_norm": 1.353843103054466,
"learning_rate": 3.374604847207587e-06,
"loss": 0.6639,
"step": 3380
},
{
"epoch": 0.6930389451088623,
"grad_norm": 1.4156299358582807,
"learning_rate": 3.369336143308746e-06,
"loss": 0.6508,
"step": 3390
},
{
"epoch": 0.6950833077788,
"grad_norm": 1.3349726637302035,
"learning_rate": 3.3640674394099053e-06,
"loss": 0.6836,
"step": 3400
},
{
"epoch": 0.6971276704487376,
"grad_norm": 1.4120432248927959,
"learning_rate": 3.3587987355110646e-06,
"loss": 0.6776,
"step": 3410
},
{
"epoch": 0.6991720331186753,
"grad_norm": 1.4540824343741436,
"learning_rate": 3.353530031612224e-06,
"loss": 0.6618,
"step": 3420
},
{
"epoch": 0.701216395788613,
"grad_norm": 1.2666549140293257,
"learning_rate": 3.3482613277133828e-06,
"loss": 0.6851,
"step": 3430
},
{
"epoch": 0.7032607584585505,
"grad_norm": 1.5085785030235015,
"learning_rate": 3.342992623814542e-06,
"loss": 0.6547,
"step": 3440
},
{
"epoch": 0.7053051211284882,
"grad_norm": 1.3427244886639838,
"learning_rate": 3.337723919915701e-06,
"loss": 0.6433,
"step": 3450
},
{
"epoch": 0.7073494837984259,
"grad_norm": 1.4282722132086514,
"learning_rate": 3.33245521601686e-06,
"loss": 0.6344,
"step": 3460
},
{
"epoch": 0.7093938464683635,
"grad_norm": 1.3947619878052135,
"learning_rate": 3.327186512118019e-06,
"loss": 0.6805,
"step": 3470
},
{
"epoch": 0.7114382091383011,
"grad_norm": 1.4996950808481269,
"learning_rate": 3.3219178082191784e-06,
"loss": 0.6736,
"step": 3480
},
{
"epoch": 0.7134825718082388,
"grad_norm": 1.3688070406703314,
"learning_rate": 3.3166491043203377e-06,
"loss": 0.6345,
"step": 3490
},
{
"epoch": 0.7155269344781764,
"grad_norm": 1.3435318933810874,
"learning_rate": 3.3113804004214965e-06,
"loss": 0.6743,
"step": 3500
},
{
"epoch": 0.7175712971481141,
"grad_norm": 1.2668395745633274,
"learning_rate": 3.306111696522656e-06,
"loss": 0.6524,
"step": 3510
},
{
"epoch": 0.7196156598180518,
"grad_norm": 1.4066738429925287,
"learning_rate": 3.300842992623815e-06,
"loss": 0.6551,
"step": 3520
},
{
"epoch": 0.7216600224879893,
"grad_norm": 1.2260317922686725,
"learning_rate": 3.2955742887249744e-06,
"loss": 0.6704,
"step": 3530
},
{
"epoch": 0.723704385157927,
"grad_norm": 1.4097940622600351,
"learning_rate": 3.290305584826133e-06,
"loss": 0.6545,
"step": 3540
},
{
"epoch": 0.7257487478278647,
"grad_norm": 1.3319274601648876,
"learning_rate": 3.285036880927292e-06,
"loss": 0.6666,
"step": 3550
},
{
"epoch": 0.7277931104978023,
"grad_norm": 1.2446441952329466,
"learning_rate": 3.279768177028451e-06,
"loss": 0.598,
"step": 3560
},
{
"epoch": 0.72983747316774,
"grad_norm": 1.6559485345957636,
"learning_rate": 3.2744994731296103e-06,
"loss": 0.6945,
"step": 3570
},
{
"epoch": 0.7318818358376776,
"grad_norm": 1.3288784169530907,
"learning_rate": 3.2692307692307696e-06,
"loss": 0.646,
"step": 3580
},
{
"epoch": 0.7339261985076152,
"grad_norm": 1.3619516078308285,
"learning_rate": 3.263962065331929e-06,
"loss": 0.6591,
"step": 3590
},
{
"epoch": 0.7359705611775529,
"grad_norm": 1.3383806915313736,
"learning_rate": 3.2586933614330877e-06,
"loss": 0.6525,
"step": 3600
},
{
"epoch": 0.7380149238474906,
"grad_norm": 1.3645876639036376,
"learning_rate": 3.253424657534247e-06,
"loss": 0.673,
"step": 3610
},
{
"epoch": 0.7400592865174281,
"grad_norm": 1.2799730669144016,
"learning_rate": 3.248155953635406e-06,
"loss": 0.6723,
"step": 3620
},
{
"epoch": 0.7421036491873658,
"grad_norm": 1.2480867531677344,
"learning_rate": 3.2428872497365647e-06,
"loss": 0.6636,
"step": 3630
},
{
"epoch": 0.7441480118573035,
"grad_norm": 1.4486578291682746,
"learning_rate": 3.237618545837724e-06,
"loss": 0.6739,
"step": 3640
},
{
"epoch": 0.7461923745272412,
"grad_norm": 1.3275379704829926,
"learning_rate": 3.2323498419388833e-06,
"loss": 0.6418,
"step": 3650
},
{
"epoch": 0.7482367371971788,
"grad_norm": 1.3285420000897692,
"learning_rate": 3.2270811380400426e-06,
"loss": 0.6336,
"step": 3660
},
{
"epoch": 0.7502810998671164,
"grad_norm": 1.350744754176338,
"learning_rate": 3.2218124341412015e-06,
"loss": 0.6637,
"step": 3670
},
{
"epoch": 0.7523254625370541,
"grad_norm": 1.2446421023001093,
"learning_rate": 3.2165437302423607e-06,
"loss": 0.6652,
"step": 3680
},
{
"epoch": 0.7543698252069917,
"grad_norm": 1.2933151567866419,
"learning_rate": 3.21127502634352e-06,
"loss": 0.6793,
"step": 3690
},
{
"epoch": 0.7564141878769294,
"grad_norm": 1.3100691591421576,
"learning_rate": 3.2060063224446793e-06,
"loss": 0.7005,
"step": 3700
},
{
"epoch": 0.7584585505468671,
"grad_norm": 1.4297250515517603,
"learning_rate": 3.2007376185458378e-06,
"loss": 0.6552,
"step": 3710
},
{
"epoch": 0.7605029132168046,
"grad_norm": 1.3136412816811998,
"learning_rate": 3.195468914646997e-06,
"loss": 0.6527,
"step": 3720
},
{
"epoch": 0.7625472758867423,
"grad_norm": 1.2866483985703137,
"learning_rate": 3.190200210748156e-06,
"loss": 0.6515,
"step": 3730
},
{
"epoch": 0.76459163855668,
"grad_norm": 1.4493342285199264,
"learning_rate": 3.184931506849315e-06,
"loss": 0.6436,
"step": 3740
},
{
"epoch": 0.7666360012266176,
"grad_norm": 1.3721751615473445,
"learning_rate": 3.1796628029504745e-06,
"loss": 0.6734,
"step": 3750
},
{
"epoch": 0.7686803638965553,
"grad_norm": 1.2811156899303597,
"learning_rate": 3.1743940990516338e-06,
"loss": 0.6782,
"step": 3760
},
{
"epoch": 0.7707247265664929,
"grad_norm": 1.30178970259406,
"learning_rate": 3.1691253951527926e-06,
"loss": 0.6275,
"step": 3770
},
{
"epoch": 0.7727690892364305,
"grad_norm": 1.5095337618928353,
"learning_rate": 3.163856691253952e-06,
"loss": 0.6696,
"step": 3780
},
{
"epoch": 0.7748134519063682,
"grad_norm": 1.429993167244863,
"learning_rate": 3.158587987355111e-06,
"loss": 0.6596,
"step": 3790
},
{
"epoch": 0.7768578145763059,
"grad_norm": 1.7528752523245135,
"learning_rate": 3.1533192834562697e-06,
"loss": 0.6682,
"step": 3800
},
{
"epoch": 0.7789021772462434,
"grad_norm": 1.298913314174674,
"learning_rate": 3.148050579557429e-06,
"loss": 0.6765,
"step": 3810
},
{
"epoch": 0.7809465399161811,
"grad_norm": 1.3162522503018863,
"learning_rate": 3.1427818756585882e-06,
"loss": 0.6149,
"step": 3820
},
{
"epoch": 0.7829909025861188,
"grad_norm": 1.2917682564921729,
"learning_rate": 3.1375131717597475e-06,
"loss": 0.6536,
"step": 3830
},
{
"epoch": 0.7850352652560564,
"grad_norm": 1.4173490442727366,
"learning_rate": 3.1322444678609064e-06,
"loss": 0.6759,
"step": 3840
},
{
"epoch": 0.7870796279259941,
"grad_norm": 1.4166016573580116,
"learning_rate": 3.1269757639620657e-06,
"loss": 0.7004,
"step": 3850
},
{
"epoch": 0.7891239905959317,
"grad_norm": 1.390027202470499,
"learning_rate": 3.121707060063225e-06,
"loss": 0.6564,
"step": 3860
},
{
"epoch": 0.7911683532658693,
"grad_norm": 1.2506495966882687,
"learning_rate": 3.1164383561643843e-06,
"loss": 0.6287,
"step": 3870
},
{
"epoch": 0.793212715935807,
"grad_norm": 1.3427227171750724,
"learning_rate": 3.1111696522655427e-06,
"loss": 0.6336,
"step": 3880
},
{
"epoch": 0.7952570786057447,
"grad_norm": 1.400852528206893,
"learning_rate": 3.105900948366702e-06,
"loss": 0.6654,
"step": 3890
},
{
"epoch": 0.7973014412756823,
"grad_norm": 1.296526337933004,
"learning_rate": 3.100632244467861e-06,
"loss": 0.6943,
"step": 3900
},
{
"epoch": 0.7993458039456199,
"grad_norm": 1.4918055672159256,
"learning_rate": 3.09536354056902e-06,
"loss": 0.6859,
"step": 3910
},
{
"epoch": 0.8013901666155576,
"grad_norm": 1.422515134465073,
"learning_rate": 3.0900948366701794e-06,
"loss": 0.6533,
"step": 3920
},
{
"epoch": 0.8034345292854952,
"grad_norm": 1.3446455729455822,
"learning_rate": 3.0848261327713387e-06,
"loss": 0.6428,
"step": 3930
},
{
"epoch": 0.8054788919554329,
"grad_norm": 1.3190379093056441,
"learning_rate": 3.0795574288724976e-06,
"loss": 0.6665,
"step": 3940
},
{
"epoch": 0.8075232546253706,
"grad_norm": 1.3268381015758894,
"learning_rate": 3.074288724973657e-06,
"loss": 0.6769,
"step": 3950
},
{
"epoch": 0.8095676172953082,
"grad_norm": 1.2552688298132784,
"learning_rate": 3.069020021074816e-06,
"loss": 0.6559,
"step": 3960
},
{
"epoch": 0.8116119799652458,
"grad_norm": 1.3472035532664182,
"learning_rate": 3.0637513171759746e-06,
"loss": 0.6485,
"step": 3970
},
{
"epoch": 0.8136563426351835,
"grad_norm": 1.333547649398474,
"learning_rate": 3.058482613277134e-06,
"loss": 0.6471,
"step": 3980
},
{
"epoch": 0.8157007053051212,
"grad_norm": 1.326997920239911,
"learning_rate": 3.053213909378293e-06,
"loss": 0.6829,
"step": 3990
},
{
"epoch": 0.8177450679750587,
"grad_norm": 1.3590306737733382,
"learning_rate": 3.0479452054794525e-06,
"loss": 0.6415,
"step": 4000
}
],
"logging_steps": 10,
"max_steps": 9784,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 2000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2490542473084928.0,
"train_batch_size": 6,
"trial_name": null,
"trial_params": null
}