llama3.2_feedback_3b / trainer_state.json
terry69's picture
Model save
4486379 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1656,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006038647342995169,
"grad_norm": 12.983106547511849,
"learning_rate": 6.02409638554217e-08,
"loss": 1.5929,
"step": 1
},
{
"epoch": 0.0030193236714975845,
"grad_norm": 13.260146238671595,
"learning_rate": 3.0120481927710845e-07,
"loss": 1.5587,
"step": 5
},
{
"epoch": 0.006038647342995169,
"grad_norm": 11.79639694358591,
"learning_rate": 6.024096385542169e-07,
"loss": 1.5735,
"step": 10
},
{
"epoch": 0.009057971014492754,
"grad_norm": 7.277578840066492,
"learning_rate": 9.036144578313254e-07,
"loss": 1.5311,
"step": 15
},
{
"epoch": 0.012077294685990338,
"grad_norm": 2.769327318468503,
"learning_rate": 1.2048192771084338e-06,
"loss": 1.4438,
"step": 20
},
{
"epoch": 0.015096618357487922,
"grad_norm": 2.665518311227292,
"learning_rate": 1.5060240963855425e-06,
"loss": 1.3911,
"step": 25
},
{
"epoch": 0.018115942028985508,
"grad_norm": 1.8820740815152255,
"learning_rate": 1.8072289156626508e-06,
"loss": 1.3317,
"step": 30
},
{
"epoch": 0.021135265700483092,
"grad_norm": 1.6047068311406143,
"learning_rate": 2.1084337349397595e-06,
"loss": 1.2409,
"step": 35
},
{
"epoch": 0.024154589371980676,
"grad_norm": 1.3473128041783777,
"learning_rate": 2.4096385542168676e-06,
"loss": 1.1865,
"step": 40
},
{
"epoch": 0.02717391304347826,
"grad_norm": 0.8992074365325567,
"learning_rate": 2.710843373493976e-06,
"loss": 1.1272,
"step": 45
},
{
"epoch": 0.030193236714975844,
"grad_norm": 0.828221042015975,
"learning_rate": 3.012048192771085e-06,
"loss": 1.0992,
"step": 50
},
{
"epoch": 0.03321256038647343,
"grad_norm": 0.7722295149097933,
"learning_rate": 3.313253012048193e-06,
"loss": 1.077,
"step": 55
},
{
"epoch": 0.036231884057971016,
"grad_norm": 0.7262180334499571,
"learning_rate": 3.6144578313253016e-06,
"loss": 1.0604,
"step": 60
},
{
"epoch": 0.0392512077294686,
"grad_norm": 0.7658586617349729,
"learning_rate": 3.91566265060241e-06,
"loss": 1.0581,
"step": 65
},
{
"epoch": 0.042270531400966184,
"grad_norm": 0.7379283258123409,
"learning_rate": 4.216867469879519e-06,
"loss": 1.0455,
"step": 70
},
{
"epoch": 0.04528985507246377,
"grad_norm": 0.7009661807319135,
"learning_rate": 4.518072289156627e-06,
"loss": 1.0437,
"step": 75
},
{
"epoch": 0.04830917874396135,
"grad_norm": 0.7478867629078089,
"learning_rate": 4.819277108433735e-06,
"loss": 1.0209,
"step": 80
},
{
"epoch": 0.051328502415458936,
"grad_norm": 0.7441839879764262,
"learning_rate": 5.120481927710844e-06,
"loss": 1.0119,
"step": 85
},
{
"epoch": 0.05434782608695652,
"grad_norm": 0.7698711195458738,
"learning_rate": 5.421686746987952e-06,
"loss": 1.0097,
"step": 90
},
{
"epoch": 0.057367149758454104,
"grad_norm": 0.7510808966928294,
"learning_rate": 5.722891566265061e-06,
"loss": 1.0022,
"step": 95
},
{
"epoch": 0.06038647342995169,
"grad_norm": 0.7895560616147937,
"learning_rate": 6.02409638554217e-06,
"loss": 1.0129,
"step": 100
},
{
"epoch": 0.06340579710144928,
"grad_norm": 0.7650744315430823,
"learning_rate": 6.325301204819277e-06,
"loss": 0.9836,
"step": 105
},
{
"epoch": 0.06642512077294686,
"grad_norm": 0.7656590839025053,
"learning_rate": 6.626506024096386e-06,
"loss": 0.9863,
"step": 110
},
{
"epoch": 0.06944444444444445,
"grad_norm": 0.7358171567450749,
"learning_rate": 6.927710843373494e-06,
"loss": 0.9901,
"step": 115
},
{
"epoch": 0.07246376811594203,
"grad_norm": 0.8007259705758099,
"learning_rate": 7.228915662650603e-06,
"loss": 0.9498,
"step": 120
},
{
"epoch": 0.07548309178743962,
"grad_norm": 0.7867873992695736,
"learning_rate": 7.530120481927712e-06,
"loss": 0.9678,
"step": 125
},
{
"epoch": 0.0785024154589372,
"grad_norm": 0.7713133158468599,
"learning_rate": 7.83132530120482e-06,
"loss": 0.9574,
"step": 130
},
{
"epoch": 0.08152173913043478,
"grad_norm": 0.8073513926091085,
"learning_rate": 8.132530120481928e-06,
"loss": 0.9572,
"step": 135
},
{
"epoch": 0.08454106280193237,
"grad_norm": 1.3178364160478768,
"learning_rate": 8.433734939759038e-06,
"loss": 0.9362,
"step": 140
},
{
"epoch": 0.08756038647342995,
"grad_norm": 0.817319025944233,
"learning_rate": 8.734939759036145e-06,
"loss": 0.9559,
"step": 145
},
{
"epoch": 0.09057971014492754,
"grad_norm": 0.9017519299904656,
"learning_rate": 9.036144578313254e-06,
"loss": 0.9318,
"step": 150
},
{
"epoch": 0.09359903381642512,
"grad_norm": 0.7674041844083781,
"learning_rate": 9.337349397590362e-06,
"loss": 0.9201,
"step": 155
},
{
"epoch": 0.0966183574879227,
"grad_norm": 0.7983502874028157,
"learning_rate": 9.63855421686747e-06,
"loss": 0.9226,
"step": 160
},
{
"epoch": 0.09963768115942029,
"grad_norm": 0.83207644117932,
"learning_rate": 9.93975903614458e-06,
"loss": 0.9184,
"step": 165
},
{
"epoch": 0.10265700483091787,
"grad_norm": 0.8377158508227592,
"learning_rate": 9.999822178354131e-06,
"loss": 0.9236,
"step": 170
},
{
"epoch": 0.10567632850241546,
"grad_norm": 0.9471244349318043,
"learning_rate": 9.999099799595088e-06,
"loss": 0.9025,
"step": 175
},
{
"epoch": 0.10869565217391304,
"grad_norm": 1.0575609941440907,
"learning_rate": 9.997821830092095e-06,
"loss": 0.8809,
"step": 180
},
{
"epoch": 0.11171497584541062,
"grad_norm": 0.8575976269736529,
"learning_rate": 9.995988411876328e-06,
"loss": 0.8897,
"step": 185
},
{
"epoch": 0.11473429951690821,
"grad_norm": 0.9091932951623557,
"learning_rate": 9.993599748710505e-06,
"loss": 0.886,
"step": 190
},
{
"epoch": 0.11775362318840579,
"grad_norm": 0.8046734387056934,
"learning_rate": 9.990656106066257e-06,
"loss": 0.8686,
"step": 195
},
{
"epoch": 0.12077294685990338,
"grad_norm": 0.8525822568043689,
"learning_rate": 9.9871578110946e-06,
"loss": 0.8757,
"step": 200
},
{
"epoch": 0.12379227053140096,
"grad_norm": 0.830408259403725,
"learning_rate": 9.983105252589599e-06,
"loss": 0.8577,
"step": 205
},
{
"epoch": 0.12681159420289856,
"grad_norm": 0.8432806800750737,
"learning_rate": 9.978498880945138e-06,
"loss": 0.8551,
"step": 210
},
{
"epoch": 0.12983091787439613,
"grad_norm": 0.786271761268233,
"learning_rate": 9.97333920810488e-06,
"loss": 0.8621,
"step": 215
},
{
"epoch": 0.13285024154589373,
"grad_norm": 0.796593756792945,
"learning_rate": 9.967626807505359e-06,
"loss": 0.8565,
"step": 220
},
{
"epoch": 0.1358695652173913,
"grad_norm": 0.746917425620166,
"learning_rate": 9.961362314012258e-06,
"loss": 0.8505,
"step": 225
},
{
"epoch": 0.1388888888888889,
"grad_norm": 0.8349781723797338,
"learning_rate": 9.954546423849842e-06,
"loss": 0.8422,
"step": 230
},
{
"epoch": 0.14190821256038647,
"grad_norm": 0.7790591361055157,
"learning_rate": 9.947179894523594e-06,
"loss": 0.8328,
"step": 235
},
{
"epoch": 0.14492753623188406,
"grad_norm": 0.767228136687293,
"learning_rate": 9.93926354473601e-06,
"loss": 0.8295,
"step": 240
},
{
"epoch": 0.14794685990338163,
"grad_norm": 0.7961627962502099,
"learning_rate": 9.930798254295628e-06,
"loss": 0.8519,
"step": 245
},
{
"epoch": 0.15096618357487923,
"grad_norm": 0.7643126730967259,
"learning_rate": 9.921784964019234e-06,
"loss": 0.8277,
"step": 250
},
{
"epoch": 0.1539855072463768,
"grad_norm": 0.7916711244024157,
"learning_rate": 9.91222467562731e-06,
"loss": 0.8413,
"step": 255
},
{
"epoch": 0.1570048309178744,
"grad_norm": 0.7566462648042606,
"learning_rate": 9.902118451632694e-06,
"loss": 0.8396,
"step": 260
},
{
"epoch": 0.16002415458937197,
"grad_norm": 0.7231104384801381,
"learning_rate": 9.891467415222511e-06,
"loss": 0.8236,
"step": 265
},
{
"epoch": 0.16304347826086957,
"grad_norm": 0.7475453602566374,
"learning_rate": 9.880272750133328e-06,
"loss": 0.8202,
"step": 270
},
{
"epoch": 0.16606280193236714,
"grad_norm": 0.7356037530580047,
"learning_rate": 9.868535700519605e-06,
"loss": 0.8254,
"step": 275
},
{
"epoch": 0.16908212560386474,
"grad_norm": 0.7098295832189967,
"learning_rate": 9.856257570815415e-06,
"loss": 0.8223,
"step": 280
},
{
"epoch": 0.1721014492753623,
"grad_norm": 0.7371095836853645,
"learning_rate": 9.843439725589481e-06,
"loss": 0.8264,
"step": 285
},
{
"epoch": 0.1751207729468599,
"grad_norm": 0.7319443665926877,
"learning_rate": 9.83008358939351e-06,
"loss": 0.8091,
"step": 290
},
{
"epoch": 0.17814009661835747,
"grad_norm": 0.7200925017038792,
"learning_rate": 9.81619064660388e-06,
"loss": 0.802,
"step": 295
},
{
"epoch": 0.18115942028985507,
"grad_norm": 0.691320410076784,
"learning_rate": 9.801762441256663e-06,
"loss": 0.8167,
"step": 300
},
{
"epoch": 0.18417874396135267,
"grad_norm": 0.7122816261812357,
"learning_rate": 9.786800576876026e-06,
"loss": 0.8061,
"step": 305
},
{
"epoch": 0.18719806763285024,
"grad_norm": 0.6987904560271868,
"learning_rate": 9.77130671629602e-06,
"loss": 0.8101,
"step": 310
},
{
"epoch": 0.19021739130434784,
"grad_norm": 0.7458807360734803,
"learning_rate": 9.755282581475769e-06,
"loss": 0.8106,
"step": 315
},
{
"epoch": 0.1932367149758454,
"grad_norm": 0.7412678256444616,
"learning_rate": 9.738729953308104e-06,
"loss": 0.8238,
"step": 320
},
{
"epoch": 0.196256038647343,
"grad_norm": 0.7383776506829391,
"learning_rate": 9.72165067142163e-06,
"loss": 0.8107,
"step": 325
},
{
"epoch": 0.19927536231884058,
"grad_norm": 0.6870198763568195,
"learning_rate": 9.70404663397628e-06,
"loss": 0.8133,
"step": 330
},
{
"epoch": 0.20229468599033817,
"grad_norm": 0.7353454455310534,
"learning_rate": 9.68591979745235e-06,
"loss": 0.8014,
"step": 335
},
{
"epoch": 0.20531400966183574,
"grad_norm": 0.6831519191417884,
"learning_rate": 9.667272176433063e-06,
"loss": 0.7968,
"step": 340
},
{
"epoch": 0.20833333333333334,
"grad_norm": 0.6847262031072842,
"learning_rate": 9.648105843380674e-06,
"loss": 0.811,
"step": 345
},
{
"epoch": 0.2113526570048309,
"grad_norm": 0.7475440150871968,
"learning_rate": 9.628422928406133e-06,
"loss": 0.8038,
"step": 350
},
{
"epoch": 0.2143719806763285,
"grad_norm": 0.7261579774011471,
"learning_rate": 9.608225619032361e-06,
"loss": 0.805,
"step": 355
},
{
"epoch": 0.21739130434782608,
"grad_norm": 0.6971431141410696,
"learning_rate": 9.587516159951118e-06,
"loss": 0.7923,
"step": 360
},
{
"epoch": 0.22041062801932368,
"grad_norm": 0.6668789906175726,
"learning_rate": 9.566296852773541e-06,
"loss": 0.797,
"step": 365
},
{
"epoch": 0.22342995169082125,
"grad_norm": 0.6993288716876052,
"learning_rate": 9.544570055774348e-06,
"loss": 0.8031,
"step": 370
},
{
"epoch": 0.22644927536231885,
"grad_norm": 0.6998160923378796,
"learning_rate": 9.522338183629737e-06,
"loss": 0.8061,
"step": 375
},
{
"epoch": 0.22946859903381642,
"grad_norm": 0.676468448981335,
"learning_rate": 9.499603707149035e-06,
"loss": 0.7972,
"step": 380
},
{
"epoch": 0.23248792270531402,
"grad_norm": 0.691768375741505,
"learning_rate": 9.476369153000076e-06,
"loss": 0.7937,
"step": 385
},
{
"epoch": 0.23550724637681159,
"grad_norm": 0.69459059647659,
"learning_rate": 9.45263710342842e-06,
"loss": 0.8009,
"step": 390
},
{
"epoch": 0.23852657004830918,
"grad_norm": 0.7295734841939806,
"learning_rate": 9.428410195970337e-06,
"loss": 0.7843,
"step": 395
},
{
"epoch": 0.24154589371980675,
"grad_norm": 0.707734024855164,
"learning_rate": 9.403691123159707e-06,
"loss": 0.7992,
"step": 400
},
{
"epoch": 0.24456521739130435,
"grad_norm": 0.7333982393973975,
"learning_rate": 9.378482632228745e-06,
"loss": 0.8037,
"step": 405
},
{
"epoch": 0.24758454106280192,
"grad_norm": 0.70227413392279,
"learning_rate": 9.352787524802707e-06,
"loss": 0.7937,
"step": 410
},
{
"epoch": 0.2506038647342995,
"grad_norm": 0.6884233411609633,
"learning_rate": 9.326608656588502e-06,
"loss": 0.7746,
"step": 415
},
{
"epoch": 0.2536231884057971,
"grad_norm": 0.6980055553926061,
"learning_rate": 9.299948937057325e-06,
"loss": 0.7831,
"step": 420
},
{
"epoch": 0.25664251207729466,
"grad_norm": 0.6886842350638906,
"learning_rate": 9.272811329121305e-06,
"loss": 0.7782,
"step": 425
},
{
"epoch": 0.25966183574879226,
"grad_norm": 0.6774682617588125,
"learning_rate": 9.245198848804197e-06,
"loss": 0.7676,
"step": 430
},
{
"epoch": 0.26268115942028986,
"grad_norm": 0.6594148286491368,
"learning_rate": 9.217114564906208e-06,
"loss": 0.78,
"step": 435
},
{
"epoch": 0.26570048309178745,
"grad_norm": 0.719815019827257,
"learning_rate": 9.188561598662921e-06,
"loss": 0.7677,
"step": 440
},
{
"epoch": 0.26871980676328505,
"grad_norm": 0.7034720888358402,
"learning_rate": 9.159543123398416e-06,
"loss": 0.7909,
"step": 445
},
{
"epoch": 0.2717391304347826,
"grad_norm": 0.7146788499687661,
"learning_rate": 9.130062364172582e-06,
"loss": 0.7832,
"step": 450
},
{
"epoch": 0.2747584541062802,
"grad_norm": 0.6785027397476343,
"learning_rate": 9.1001225974227e-06,
"loss": 0.7912,
"step": 455
},
{
"epoch": 0.2777777777777778,
"grad_norm": 0.6719234399237637,
"learning_rate": 9.0697271505993e-06,
"loss": 0.7918,
"step": 460
},
{
"epoch": 0.2807971014492754,
"grad_norm": 0.7023247170996745,
"learning_rate": 9.038879401796358e-06,
"loss": 0.7809,
"step": 465
},
{
"epoch": 0.28381642512077293,
"grad_norm": 0.696493093698672,
"learning_rate": 9.00758277937586e-06,
"loss": 0.7794,
"step": 470
},
{
"epoch": 0.28683574879227053,
"grad_norm": 0.6846368584631723,
"learning_rate": 8.975840761586772e-06,
"loss": 0.7734,
"step": 475
},
{
"epoch": 0.2898550724637681,
"grad_norm": 0.7102623044885478,
"learning_rate": 8.94365687617849e-06,
"loss": 0.7781,
"step": 480
},
{
"epoch": 0.2928743961352657,
"grad_norm": 0.7086900307853392,
"learning_rate": 8.911034700008757e-06,
"loss": 0.7764,
"step": 485
},
{
"epoch": 0.29589371980676327,
"grad_norm": 0.6832022733750703,
"learning_rate": 8.87797785864615e-06,
"loss": 0.7791,
"step": 490
},
{
"epoch": 0.29891304347826086,
"grad_norm": 0.6784895232664829,
"learning_rate": 8.844490025967126e-06,
"loss": 0.7734,
"step": 495
},
{
"epoch": 0.30193236714975846,
"grad_norm": 0.7325432498510648,
"learning_rate": 8.810574923747729e-06,
"loss": 0.7638,
"step": 500
},
{
"epoch": 0.30495169082125606,
"grad_norm": 0.6736300693382467,
"learning_rate": 8.776236321249955e-06,
"loss": 0.7848,
"step": 505
},
{
"epoch": 0.3079710144927536,
"grad_norm": 0.7199354344226971,
"learning_rate": 8.741478034802835e-06,
"loss": 0.7672,
"step": 510
},
{
"epoch": 0.3109903381642512,
"grad_norm": 0.7181172806743141,
"learning_rate": 8.706303927378306e-06,
"loss": 0.7726,
"step": 515
},
{
"epoch": 0.3140096618357488,
"grad_norm": 0.6981681510152888,
"learning_rate": 8.670717908161878e-06,
"loss": 0.7815,
"step": 520
},
{
"epoch": 0.3170289855072464,
"grad_norm": 0.6999219057435419,
"learning_rate": 8.634723932118184e-06,
"loss": 0.7766,
"step": 525
},
{
"epoch": 0.32004830917874394,
"grad_norm": 0.6720516411704228,
"learning_rate": 8.598325999551425e-06,
"loss": 0.7774,
"step": 530
},
{
"epoch": 0.32306763285024154,
"grad_norm": 0.6991878290864751,
"learning_rate": 8.56152815566078e-06,
"loss": 0.7659,
"step": 535
},
{
"epoch": 0.32608695652173914,
"grad_norm": 1.1363114984267308,
"learning_rate": 8.524334490090848e-06,
"loss": 0.7669,
"step": 540
},
{
"epoch": 0.32910628019323673,
"grad_norm": 0.7170057225271362,
"learning_rate": 8.486749136477112e-06,
"loss": 0.7607,
"step": 545
},
{
"epoch": 0.3321256038647343,
"grad_norm": 0.7151461285207649,
"learning_rate": 8.448776271986542e-06,
"loss": 0.7556,
"step": 550
},
{
"epoch": 0.3351449275362319,
"grad_norm": 0.7080653566142882,
"learning_rate": 8.41042011685336e-06,
"loss": 0.7801,
"step": 555
},
{
"epoch": 0.33816425120772947,
"grad_norm": 0.7211038917950292,
"learning_rate": 8.371684933909996e-06,
"loss": 0.7819,
"step": 560
},
{
"epoch": 0.34118357487922707,
"grad_norm": 0.6809572707011691,
"learning_rate": 8.33257502811334e-06,
"loss": 0.7623,
"step": 565
},
{
"epoch": 0.3442028985507246,
"grad_norm": 0.705758884996601,
"learning_rate": 8.293094746066283e-06,
"loss": 0.7806,
"step": 570
},
{
"epoch": 0.3472222222222222,
"grad_norm": 0.7010425860566304,
"learning_rate": 8.253248475534656e-06,
"loss": 0.7579,
"step": 575
},
{
"epoch": 0.3502415458937198,
"grad_norm": 0.6961950426104488,
"learning_rate": 8.213040644959572e-06,
"loss": 0.7563,
"step": 580
},
{
"epoch": 0.3532608695652174,
"grad_norm": 0.7380978491549621,
"learning_rate": 8.172475722965263e-06,
"loss": 0.748,
"step": 585
},
{
"epoch": 0.35628019323671495,
"grad_norm": 0.6880280880233399,
"learning_rate": 8.131558217862444e-06,
"loss": 0.7464,
"step": 590
},
{
"epoch": 0.35929951690821255,
"grad_norm": 0.6923751245984867,
"learning_rate": 8.090292677147268e-06,
"loss": 0.7654,
"step": 595
},
{
"epoch": 0.36231884057971014,
"grad_norm": 0.6875828804508053,
"learning_rate": 8.048683686995921e-06,
"loss": 0.7575,
"step": 600
},
{
"epoch": 0.36533816425120774,
"grad_norm": 0.7271877934985621,
"learning_rate": 8.006735871754932e-06,
"loss": 0.7521,
"step": 605
},
{
"epoch": 0.36835748792270534,
"grad_norm": 0.6919263020022135,
"learning_rate": 7.96445389342722e-06,
"loss": 0.7522,
"step": 610
},
{
"epoch": 0.3713768115942029,
"grad_norm": 0.7079839592582174,
"learning_rate": 7.921842451153982e-06,
"loss": 0.7624,
"step": 615
},
{
"epoch": 0.3743961352657005,
"grad_norm": 0.7185016239152293,
"learning_rate": 7.878906280692424e-06,
"loss": 0.7638,
"step": 620
},
{
"epoch": 0.3774154589371981,
"grad_norm": 0.6692678069942056,
"learning_rate": 7.835650153889449e-06,
"loss": 0.7599,
"step": 625
},
{
"epoch": 0.3804347826086957,
"grad_norm": 0.6901273422065264,
"learning_rate": 7.792078878151318e-06,
"loss": 0.7438,
"step": 630
},
{
"epoch": 0.3834541062801932,
"grad_norm": 0.7025129368063263,
"learning_rate": 7.748197295909359e-06,
"loss": 0.7544,
"step": 635
},
{
"epoch": 0.3864734299516908,
"grad_norm": 0.7452661788017922,
"learning_rate": 7.704010284081801e-06,
"loss": 0.7542,
"step": 640
},
{
"epoch": 0.3894927536231884,
"grad_norm": 0.6838884146151244,
"learning_rate": 7.65952275353175e-06,
"loss": 0.7421,
"step": 645
},
{
"epoch": 0.392512077294686,
"grad_norm": 0.7050616814337126,
"learning_rate": 7.614739648521412e-06,
"loss": 0.7455,
"step": 650
},
{
"epoch": 0.39553140096618356,
"grad_norm": 0.7027226698051754,
"learning_rate": 7.56966594616259e-06,
"loss": 0.7526,
"step": 655
},
{
"epoch": 0.39855072463768115,
"grad_norm": 0.7289432071080068,
"learning_rate": 7.524306655863544e-06,
"loss": 0.7586,
"step": 660
},
{
"epoch": 0.40157004830917875,
"grad_norm": 0.70492218137627,
"learning_rate": 7.478666818772252e-06,
"loss": 0.7517,
"step": 665
},
{
"epoch": 0.40458937198067635,
"grad_norm": 0.6844740192077617,
"learning_rate": 7.432751507216146e-06,
"loss": 0.7515,
"step": 670
},
{
"epoch": 0.4076086956521739,
"grad_norm": 0.7034634050055147,
"learning_rate": 7.386565824138378e-06,
"loss": 0.7389,
"step": 675
},
{
"epoch": 0.4106280193236715,
"grad_norm": 0.7020675334607289,
"learning_rate": 7.3401149025306995e-06,
"loss": 0.7573,
"step": 680
},
{
"epoch": 0.4136473429951691,
"grad_norm": 0.7185467171376326,
"learning_rate": 7.293403904862981e-06,
"loss": 0.7525,
"step": 685
},
{
"epoch": 0.4166666666666667,
"grad_norm": 0.6904148869691911,
"learning_rate": 7.246438022509465e-06,
"loss": 0.7394,
"step": 690
},
{
"epoch": 0.4196859903381642,
"grad_norm": 0.721590691620625,
"learning_rate": 7.199222475171812e-06,
"loss": 0.7539,
"step": 695
},
{
"epoch": 0.4227053140096618,
"grad_norm": 0.6707901440948312,
"learning_rate": 7.151762510298985e-06,
"loss": 0.7364,
"step": 700
},
{
"epoch": 0.4257246376811594,
"grad_norm": 0.6984624075001936,
"learning_rate": 7.104063402504065e-06,
"loss": 0.7475,
"step": 705
},
{
"epoch": 0.428743961352657,
"grad_norm": 0.6893220380004709,
"learning_rate": 7.056130452978039e-06,
"loss": 0.7381,
"step": 710
},
{
"epoch": 0.43176328502415456,
"grad_norm": 0.6680081009254123,
"learning_rate": 7.0079689889006275e-06,
"loss": 0.7277,
"step": 715
},
{
"epoch": 0.43478260869565216,
"grad_norm": 0.7193168024456026,
"learning_rate": 6.959584362848239e-06,
"loss": 0.7366,
"step": 720
},
{
"epoch": 0.43780193236714976,
"grad_norm": 0.6897586291398505,
"learning_rate": 6.910981952199097e-06,
"loss": 0.741,
"step": 725
},
{
"epoch": 0.44082125603864736,
"grad_norm": 0.6942232877339098,
"learning_rate": 6.862167158535599e-06,
"loss": 0.7322,
"step": 730
},
{
"epoch": 0.4438405797101449,
"grad_norm": 0.6892818670366425,
"learning_rate": 6.813145407044003e-06,
"loss": 0.7245,
"step": 735
},
{
"epoch": 0.4468599033816425,
"grad_norm": 0.7158811742929143,
"learning_rate": 6.763922145911474e-06,
"loss": 0.7386,
"step": 740
},
{
"epoch": 0.4498792270531401,
"grad_norm": 0.6973013952551397,
"learning_rate": 6.714502845720595e-06,
"loss": 0.7405,
"step": 745
},
{
"epoch": 0.4528985507246377,
"grad_norm": 0.7452913378431993,
"learning_rate": 6.664892998841361e-06,
"loss": 0.7269,
"step": 750
},
{
"epoch": 0.4559178743961353,
"grad_norm": 0.6976841641342625,
"learning_rate": 6.61509811882078e-06,
"loss": 0.7265,
"step": 755
},
{
"epoch": 0.45893719806763283,
"grad_norm": 0.7477615859131558,
"learning_rate": 6.565123739770102e-06,
"loss": 0.7276,
"step": 760
},
{
"epoch": 0.46195652173913043,
"grad_norm": 0.7014375922011533,
"learning_rate": 6.5149754157497645e-06,
"loss": 0.7325,
"step": 765
},
{
"epoch": 0.46497584541062803,
"grad_norm": 0.7005292253227544,
"learning_rate": 6.464658720152135e-06,
"loss": 0.7225,
"step": 770
},
{
"epoch": 0.46799516908212563,
"grad_norm": 0.6943536700561682,
"learning_rate": 6.41417924508208e-06,
"loss": 0.7467,
"step": 775
},
{
"epoch": 0.47101449275362317,
"grad_norm": 0.7080470812325974,
"learning_rate": 6.363542600735486e-06,
"loss": 0.7247,
"step": 780
},
{
"epoch": 0.47403381642512077,
"grad_norm": 0.7489724802271261,
"learning_rate": 6.312754414775737e-06,
"loss": 0.7364,
"step": 785
},
{
"epoch": 0.47705314009661837,
"grad_norm": 0.7162417629371682,
"learning_rate": 6.261820331708275e-06,
"loss": 0.7443,
"step": 790
},
{
"epoch": 0.48007246376811596,
"grad_norm": 0.7148119129623506,
"learning_rate": 6.210746012253277e-06,
"loss": 0.7413,
"step": 795
},
{
"epoch": 0.4830917874396135,
"grad_norm": 0.735396159989601,
"learning_rate": 6.159537132716532e-06,
"loss": 0.7177,
"step": 800
},
{
"epoch": 0.4861111111111111,
"grad_norm": 0.6991347342976139,
"learning_rate": 6.108199384358595e-06,
"loss": 0.7236,
"step": 805
},
{
"epoch": 0.4891304347826087,
"grad_norm": 0.7041418452567949,
"learning_rate": 6.0567384727622566e-06,
"loss": 0.7248,
"step": 810
},
{
"epoch": 0.4921497584541063,
"grad_norm": 0.7260414438520773,
"learning_rate": 6.005160117198448e-06,
"loss": 0.7361,
"step": 815
},
{
"epoch": 0.49516908212560384,
"grad_norm": 0.7168927241401117,
"learning_rate": 5.953470049990605e-06,
"loss": 0.7244,
"step": 820
},
{
"epoch": 0.49818840579710144,
"grad_norm": 0.7099109469035322,
"learning_rate": 5.90167401587759e-06,
"loss": 0.726,
"step": 825
},
{
"epoch": 0.501207729468599,
"grad_norm": 0.703918614629836,
"learning_rate": 5.84977777137523e-06,
"loss": 0.7192,
"step": 830
},
{
"epoch": 0.5042270531400966,
"grad_norm": 0.7334401309257407,
"learning_rate": 5.797787084136556e-06,
"loss": 0.7316,
"step": 835
},
{
"epoch": 0.5072463768115942,
"grad_norm": 0.7182027362178113,
"learning_rate": 5.745707732310781e-06,
"loss": 0.7281,
"step": 840
},
{
"epoch": 0.5102657004830918,
"grad_norm": 0.7174224552847884,
"learning_rate": 5.693545503901149e-06,
"loss": 0.7246,
"step": 845
},
{
"epoch": 0.5132850241545893,
"grad_norm": 0.7156314475885969,
"learning_rate": 5.641306196121643e-06,
"loss": 0.725,
"step": 850
},
{
"epoch": 0.5163043478260869,
"grad_norm": 0.7364143027978156,
"learning_rate": 5.5889956147527156e-06,
"loss": 0.7213,
"step": 855
},
{
"epoch": 0.5193236714975845,
"grad_norm": 0.7655767182130154,
"learning_rate": 5.536619573496027e-06,
"loss": 0.7243,
"step": 860
},
{
"epoch": 0.5223429951690821,
"grad_norm": 0.7155331967411032,
"learning_rate": 5.484183893328332e-06,
"loss": 0.7238,
"step": 865
},
{
"epoch": 0.5253623188405797,
"grad_norm": 0.7094631414851184,
"learning_rate": 5.431694401854545e-06,
"loss": 0.7022,
"step": 870
},
{
"epoch": 0.5283816425120773,
"grad_norm": 0.701366365553249,
"learning_rate": 5.379156932660067e-06,
"loss": 0.7196,
"step": 875
},
{
"epoch": 0.5314009661835749,
"grad_norm": 0.690941783111536,
"learning_rate": 5.326577324662459e-06,
"loss": 0.7128,
"step": 880
},
{
"epoch": 0.5344202898550725,
"grad_norm": 0.7003221590794428,
"learning_rate": 5.273961421462505e-06,
"loss": 0.726,
"step": 885
},
{
"epoch": 0.5374396135265701,
"grad_norm": 0.6987939906127779,
"learning_rate": 5.221315070694775e-06,
"loss": 0.7083,
"step": 890
},
{
"epoch": 0.5404589371980676,
"grad_norm": 0.7416781371272098,
"learning_rate": 5.168644123377725e-06,
"loss": 0.7056,
"step": 895
},
{
"epoch": 0.5434782608695652,
"grad_norm": 0.7259377894484753,
"learning_rate": 5.1159544332634256e-06,
"loss": 0.7224,
"step": 900
},
{
"epoch": 0.5464975845410628,
"grad_norm": 0.7502013953347851,
"learning_rate": 5.063251856186991e-06,
"loss": 0.7029,
"step": 905
},
{
"epoch": 0.5495169082125604,
"grad_norm": 0.704558454200649,
"learning_rate": 5.010542249415761e-06,
"loss": 0.7151,
"step": 910
},
{
"epoch": 0.552536231884058,
"grad_norm": 0.7097627470098892,
"learning_rate": 4.95783147099835e-06,
"loss": 0.7056,
"step": 915
},
{
"epoch": 0.5555555555555556,
"grad_norm": 0.73696211091724,
"learning_rate": 4.90512537911358e-06,
"loss": 0.7162,
"step": 920
},
{
"epoch": 0.5585748792270532,
"grad_norm": 0.7233738791549937,
"learning_rate": 4.852429831419428e-06,
"loss": 0.7244,
"step": 925
},
{
"epoch": 0.5615942028985508,
"grad_norm": 0.7505624253150545,
"learning_rate": 4.799750684402006e-06,
"loss": 0.7059,
"step": 930
},
{
"epoch": 0.5646135265700483,
"grad_norm": 0.7189320855898085,
"learning_rate": 4.747093792724679e-06,
"loss": 0.7123,
"step": 935
},
{
"epoch": 0.5676328502415459,
"grad_norm": 0.7501646767976898,
"learning_rate": 4.6944650085774095e-06,
"loss": 0.7035,
"step": 940
},
{
"epoch": 0.5706521739130435,
"grad_norm": 0.7162880585290501,
"learning_rate": 4.641870181026322e-06,
"loss": 0.7179,
"step": 945
},
{
"epoch": 0.5736714975845411,
"grad_norm": 1.3960570086002833,
"learning_rate": 4.589315155363683e-06,
"loss": 0.7139,
"step": 950
},
{
"epoch": 0.5766908212560387,
"grad_norm": 0.7062944321865234,
"learning_rate": 4.53680577245824e-06,
"loss": 0.704,
"step": 955
},
{
"epoch": 0.5797101449275363,
"grad_norm": 0.7185669957370462,
"learning_rate": 4.484347868106097e-06,
"loss": 0.7023,
"step": 960
},
{
"epoch": 0.5827294685990339,
"grad_norm": 0.7187434418978588,
"learning_rate": 4.431947272382118e-06,
"loss": 0.712,
"step": 965
},
{
"epoch": 0.5857487922705314,
"grad_norm": 0.7079462407351277,
"learning_rate": 4.379609808992e-06,
"loss": 0.7181,
"step": 970
},
{
"epoch": 0.5887681159420289,
"grad_norm": 0.7482705608146033,
"learning_rate": 4.327341294625019e-06,
"loss": 0.7023,
"step": 975
},
{
"epoch": 0.5917874396135265,
"grad_norm": 0.7490533325460069,
"learning_rate": 4.275147538307594e-06,
"loss": 0.7137,
"step": 980
},
{
"epoch": 0.5948067632850241,
"grad_norm": 0.7242674704325197,
"learning_rate": 4.223034340757666e-06,
"loss": 0.7031,
"step": 985
},
{
"epoch": 0.5978260869565217,
"grad_norm": 0.6939633292606721,
"learning_rate": 4.171007493740023e-06,
"loss": 0.6977,
"step": 990
},
{
"epoch": 0.6008454106280193,
"grad_norm": 0.7133068235360875,
"learning_rate": 4.1190727794226175e-06,
"loss": 0.7162,
"step": 995
},
{
"epoch": 0.6038647342995169,
"grad_norm": 0.754448396241569,
"learning_rate": 4.067235969733937e-06,
"loss": 0.6997,
"step": 1000
},
{
"epoch": 0.6068840579710145,
"grad_norm": 0.7010857082662448,
"learning_rate": 4.015502825721537e-06,
"loss": 0.7066,
"step": 1005
},
{
"epoch": 0.6099033816425121,
"grad_norm": 0.7220837983225042,
"learning_rate": 3.963879096911751e-06,
"loss": 0.7086,
"step": 1010
},
{
"epoch": 0.6129227053140096,
"grad_norm": 0.7251109346616071,
"learning_rate": 3.91237052067072e-06,
"loss": 0.7066,
"step": 1015
},
{
"epoch": 0.6159420289855072,
"grad_norm": 0.72480781238386,
"learning_rate": 3.860982821566729e-06,
"loss": 0.7182,
"step": 1020
},
{
"epoch": 0.6189613526570048,
"grad_norm": 3.2413869749561592,
"learning_rate": 3.8097217107340107e-06,
"loss": 0.6877,
"step": 1025
},
{
"epoch": 0.6219806763285024,
"grad_norm": 0.7252475452320206,
"learning_rate": 3.7585928852380025e-06,
"loss": 0.6922,
"step": 1030
},
{
"epoch": 0.625,
"grad_norm": 0.7225890571453955,
"learning_rate": 3.7076020274421996e-06,
"loss": 0.7111,
"step": 1035
},
{
"epoch": 0.6280193236714976,
"grad_norm": 0.7398879600723324,
"learning_rate": 3.6567548043766157e-06,
"loss": 0.6938,
"step": 1040
},
{
"epoch": 0.6310386473429952,
"grad_norm": 0.7320766149333374,
"learning_rate": 3.6060568671079658e-06,
"loss": 0.6994,
"step": 1045
},
{
"epoch": 0.6340579710144928,
"grad_norm": 0.7313655291651611,
"learning_rate": 3.5555138501116247e-06,
"loss": 0.7059,
"step": 1050
},
{
"epoch": 0.6370772946859904,
"grad_norm": 0.7250580708434604,
"learning_rate": 3.5051313706453995e-06,
"loss": 0.7086,
"step": 1055
},
{
"epoch": 0.6400966183574879,
"grad_norm": 0.7365679502351182,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.702,
"step": 1060
},
{
"epoch": 0.6431159420289855,
"grad_norm": 0.7193521177653698,
"learning_rate": 3.40487040350303e-06,
"loss": 0.7027,
"step": 1065
},
{
"epoch": 0.6461352657004831,
"grad_norm": 0.7308160022070641,
"learning_rate": 3.355003058646105e-06,
"loss": 0.717,
"step": 1070
},
{
"epoch": 0.6491545893719807,
"grad_norm": 0.7141660531113965,
"learning_rate": 3.305318535719343e-06,
"loss": 0.6754,
"step": 1075
},
{
"epoch": 0.6521739130434783,
"grad_norm": 0.7027978905252191,
"learning_rate": 3.2558223565691104e-06,
"loss": 0.6984,
"step": 1080
},
{
"epoch": 0.6551932367149759,
"grad_norm": 0.733945795767204,
"learning_rate": 3.2065200221095905e-06,
"loss": 0.6916,
"step": 1085
},
{
"epoch": 0.6582125603864735,
"grad_norm": 0.7409311097771059,
"learning_rate": 3.1574170117114293e-06,
"loss": 0.7043,
"step": 1090
},
{
"epoch": 0.6612318840579711,
"grad_norm": 0.7161483851129942,
"learning_rate": 3.1085187825927555e-06,
"loss": 0.6941,
"step": 1095
},
{
"epoch": 0.6642512077294686,
"grad_norm": 0.7744670260939269,
"learning_rate": 3.0598307692126904e-06,
"loss": 0.7011,
"step": 1100
},
{
"epoch": 0.6672705314009661,
"grad_norm": 0.7237442202307557,
"learning_rate": 3.0113583826673655e-06,
"loss": 0.7007,
"step": 1105
},
{
"epoch": 0.6702898550724637,
"grad_norm": 0.7450399768587923,
"learning_rate": 2.9631070100885373e-06,
"loss": 0.7107,
"step": 1110
},
{
"epoch": 0.6733091787439613,
"grad_norm": 0.7517737488150603,
"learning_rate": 2.915082014044883e-06,
"loss": 0.7067,
"step": 1115
},
{
"epoch": 0.6763285024154589,
"grad_norm": 0.7254996669994487,
"learning_rate": 2.867288731946004e-06,
"loss": 0.7006,
"step": 1120
},
{
"epoch": 0.6793478260869565,
"grad_norm": 0.7421072709801319,
"learning_rate": 2.8197324754492456e-06,
"loss": 0.6946,
"step": 1125
},
{
"epoch": 0.6823671497584541,
"grad_norm": 0.7531577228122275,
"learning_rate": 2.7724185298693596e-06,
"loss": 0.6893,
"step": 1130
},
{
"epoch": 0.6853864734299517,
"grad_norm": 0.7295101380197203,
"learning_rate": 2.7253521535911144e-06,
"loss": 0.708,
"step": 1135
},
{
"epoch": 0.6884057971014492,
"grad_norm": 0.7706880360337985,
"learning_rate": 2.678538577484871e-06,
"loss": 0.6986,
"step": 1140
},
{
"epoch": 0.6914251207729468,
"grad_norm": 0.7950469957317675,
"learning_rate": 2.6319830043252616e-06,
"loss": 0.6905,
"step": 1145
},
{
"epoch": 0.6944444444444444,
"grad_norm": 0.7435238103186514,
"learning_rate": 2.5856906082129313e-06,
"loss": 0.705,
"step": 1150
},
{
"epoch": 0.697463768115942,
"grad_norm": 0.7599556075082218,
"learning_rate": 2.53966653399952e-06,
"loss": 0.6885,
"step": 1155
},
{
"epoch": 0.7004830917874396,
"grad_norm": 0.7576946845112895,
"learning_rate": 2.4939158967158657e-06,
"loss": 0.7052,
"step": 1160
},
{
"epoch": 0.7035024154589372,
"grad_norm": 0.7524201974305594,
"learning_rate": 2.448443781003527e-06,
"loss": 0.6897,
"step": 1165
},
{
"epoch": 0.7065217391304348,
"grad_norm": 0.7449907814909446,
"learning_rate": 2.403255240549693e-06,
"loss": 0.6965,
"step": 1170
},
{
"epoch": 0.7095410628019324,
"grad_norm": 0.7506825678278848,
"learning_rate": 2.3583552975255108e-06,
"loss": 0.6826,
"step": 1175
},
{
"epoch": 0.7125603864734299,
"grad_norm": 0.7427190867626492,
"learning_rate": 2.313748942027956e-06,
"loss": 0.6969,
"step": 1180
},
{
"epoch": 0.7155797101449275,
"grad_norm": 0.7351464262809558,
"learning_rate": 2.269441131525213e-06,
"loss": 0.6848,
"step": 1185
},
{
"epoch": 0.7185990338164251,
"grad_norm": 0.7679289820138984,
"learning_rate": 2.225436790305733e-06,
"loss": 0.6941,
"step": 1190
},
{
"epoch": 0.7216183574879227,
"grad_norm": 0.7203078445690018,
"learning_rate": 2.181740808930947e-06,
"loss": 0.6949,
"step": 1195
},
{
"epoch": 0.7246376811594203,
"grad_norm": 0.7480753757617126,
"learning_rate": 2.1383580436917452e-06,
"loss": 0.6853,
"step": 1200
},
{
"epoch": 0.7276570048309179,
"grad_norm": 0.7342063399543391,
"learning_rate": 2.0952933160687456e-06,
"loss": 0.688,
"step": 1205
},
{
"epoch": 0.7306763285024155,
"grad_norm": 0.7421683945425792,
"learning_rate": 2.052551412196456e-06,
"loss": 0.6889,
"step": 1210
},
{
"epoch": 0.7336956521739131,
"grad_norm": 0.7453132341130935,
"learning_rate": 2.010137082331354e-06,
"loss": 0.6835,
"step": 1215
},
{
"epoch": 0.7367149758454107,
"grad_norm": 0.7382993949100215,
"learning_rate": 1.96805504032393e-06,
"loss": 0.6846,
"step": 1220
},
{
"epoch": 0.7397342995169082,
"grad_norm": 0.725949395619062,
"learning_rate": 1.9263099630948274e-06,
"loss": 0.6903,
"step": 1225
},
{
"epoch": 0.7427536231884058,
"grad_norm": 0.7405823327274087,
"learning_rate": 1.8849064901150372e-06,
"loss": 0.6855,
"step": 1230
},
{
"epoch": 0.7457729468599034,
"grad_norm": 0.727724200500573,
"learning_rate": 1.8438492228902893e-06,
"loss": 0.6872,
"step": 1235
},
{
"epoch": 0.748792270531401,
"grad_norm": 0.7535334426389806,
"learning_rate": 1.8031427244496357e-06,
"loss": 0.6829,
"step": 1240
},
{
"epoch": 0.7518115942028986,
"grad_norm": 0.7659798574727297,
"learning_rate": 1.7627915188383382e-06,
"loss": 0.6959,
"step": 1245
},
{
"epoch": 0.7548309178743962,
"grad_norm": 0.7326466364617145,
"learning_rate": 1.7228000906150672e-06,
"loss": 0.6877,
"step": 1250
},
{
"epoch": 0.7578502415458938,
"grad_norm": 0.7382801872464264,
"learning_rate": 1.6831728843534962e-06,
"loss": 0.6823,
"step": 1255
},
{
"epoch": 0.7608695652173914,
"grad_norm": 0.752216196056158,
"learning_rate": 1.6439143041483352e-06,
"loss": 0.6934,
"step": 1260
},
{
"epoch": 0.7638888888888888,
"grad_norm": 0.7525714660213496,
"learning_rate": 1.6050287131258862e-06,
"loss": 0.677,
"step": 1265
},
{
"epoch": 0.7669082125603864,
"grad_norm": 0.7264048871030653,
"learning_rate": 1.5665204329591066e-06,
"loss": 0.6724,
"step": 1270
},
{
"epoch": 0.769927536231884,
"grad_norm": 0.8059548153091458,
"learning_rate": 1.528393743387328e-06,
"loss": 0.6827,
"step": 1275
},
{
"epoch": 0.7729468599033816,
"grad_norm": 0.713705103268977,
"learning_rate": 1.4906528817406052e-06,
"loss": 0.6877,
"step": 1280
},
{
"epoch": 0.7759661835748792,
"grad_norm": 0.7174894260477515,
"learning_rate": 1.453302042468786e-06,
"loss": 0.6747,
"step": 1285
},
{
"epoch": 0.7789855072463768,
"grad_norm": 0.7505147750107104,
"learning_rate": 1.4163453766753537e-06,
"loss": 0.6813,
"step": 1290
},
{
"epoch": 0.7820048309178744,
"grad_norm": 0.7257832334807974,
"learning_rate": 1.3797869916560692e-06,
"loss": 0.676,
"step": 1295
},
{
"epoch": 0.785024154589372,
"grad_norm": 0.7311403168001216,
"learning_rate": 1.3436309504425137e-06,
"loss": 0.6834,
"step": 1300
},
{
"epoch": 0.7880434782608695,
"grad_norm": 0.7776182112232197,
"learning_rate": 1.3078812713505079e-06,
"loss": 0.6882,
"step": 1305
},
{
"epoch": 0.7910628019323671,
"grad_norm": 0.7460577014828186,
"learning_rate": 1.2725419275335404e-06,
"loss": 0.6883,
"step": 1310
},
{
"epoch": 0.7940821256038647,
"grad_norm": 0.7681283165774995,
"learning_rate": 1.237616846541192e-06,
"loss": 0.6854,
"step": 1315
},
{
"epoch": 0.7971014492753623,
"grad_norm": 0.7519891668239808,
"learning_rate": 1.2031099098826376e-06,
"loss": 0.6889,
"step": 1320
},
{
"epoch": 0.8001207729468599,
"grad_norm": 0.7499150329840207,
"learning_rate": 1.1690249525952569e-06,
"loss": 0.6886,
"step": 1325
},
{
"epoch": 0.8031400966183575,
"grad_norm": 0.7336555062881699,
"learning_rate": 1.1353657628184217e-06,
"loss": 0.6665,
"step": 1330
},
{
"epoch": 0.8061594202898551,
"grad_norm": 0.747198848048635,
"learning_rate": 1.1021360813724924e-06,
"loss": 0.6887,
"step": 1335
},
{
"epoch": 0.8091787439613527,
"grad_norm": 0.7494833783636606,
"learning_rate": 1.0693396013430552e-06,
"loss": 0.6863,
"step": 1340
},
{
"epoch": 0.8121980676328503,
"grad_norm": 0.762604670610555,
"learning_rate": 1.036979967670494e-06,
"loss": 0.691,
"step": 1345
},
{
"epoch": 0.8152173913043478,
"grad_norm": 0.7514022588829053,
"learning_rate": 1.0050607767448928e-06,
"loss": 0.6721,
"step": 1350
},
{
"epoch": 0.8182367149758454,
"grad_norm": 0.7556434497258411,
"learning_rate": 9.735855760063412e-07,
"loss": 0.6865,
"step": 1355
},
{
"epoch": 0.821256038647343,
"grad_norm": 0.7654410426606472,
"learning_rate": 9.425578635506721e-07,
"loss": 0.6779,
"step": 1360
},
{
"epoch": 0.8242753623188406,
"grad_norm": 0.7454705942741552,
"learning_rate": 9.119810877406998e-07,
"loss": 0.6833,
"step": 1365
},
{
"epoch": 0.8272946859903382,
"grad_norm": 0.7683757765928639,
"learning_rate": 8.818586468229695e-07,
"loss": 0.683,
"step": 1370
},
{
"epoch": 0.8303140096618358,
"grad_norm": 0.7518004833253231,
"learning_rate": 8.521938885500825e-07,
"loss": 0.6817,
"step": 1375
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.7391679840808719,
"learning_rate": 8.229901098086335e-07,
"loss": 0.6834,
"step": 1380
},
{
"epoch": 0.836352657004831,
"grad_norm": 0.7265358575650989,
"learning_rate": 7.942505562528024e-07,
"loss": 0.6828,
"step": 1385
},
{
"epoch": 0.8393719806763285,
"grad_norm": 0.7475584495954996,
"learning_rate": 7.659784219436373e-07,
"loss": 0.6688,
"step": 1390
},
{
"epoch": 0.842391304347826,
"grad_norm": 0.7491195632530819,
"learning_rate": 7.381768489940678e-07,
"loss": 0.6675,
"step": 1395
},
{
"epoch": 0.8454106280193237,
"grad_norm": 0.7526229496466609,
"learning_rate": 7.108489272197089e-07,
"loss": 0.6836,
"step": 1400
},
{
"epoch": 0.8484299516908212,
"grad_norm": 0.7346071246989833,
"learning_rate": 6.839976937954479e-07,
"loss": 0.6844,
"step": 1405
},
{
"epoch": 0.8514492753623188,
"grad_norm": 0.7472181278592462,
"learning_rate": 6.576261329179123e-07,
"loss": 0.6802,
"step": 1410
},
{
"epoch": 0.8544685990338164,
"grad_norm": 0.760752650514086,
"learning_rate": 6.317371754738044e-07,
"loss": 0.6891,
"step": 1415
},
{
"epoch": 0.857487922705314,
"grad_norm": 0.7504755271401121,
"learning_rate": 6.06333698714171e-07,
"loss": 0.6736,
"step": 1420
},
{
"epoch": 0.8605072463768116,
"grad_norm": 0.7362415396628262,
"learning_rate": 5.814185259346267e-07,
"loss": 0.6747,
"step": 1425
},
{
"epoch": 0.8635265700483091,
"grad_norm": 0.7605711326403803,
"learning_rate": 5.56994426161584e-07,
"loss": 0.6862,
"step": 1430
},
{
"epoch": 0.8665458937198067,
"grad_norm": 0.7361924708882572,
"learning_rate": 5.330641138445064e-07,
"loss": 0.6727,
"step": 1435
},
{
"epoch": 0.8695652173913043,
"grad_norm": 0.7535196260958092,
"learning_rate": 5.096302485542265e-07,
"loss": 0.6855,
"step": 1440
},
{
"epoch": 0.8725845410628019,
"grad_norm": 0.779260026378618,
"learning_rate": 4.866954346873715e-07,
"loss": 0.6831,
"step": 1445
},
{
"epoch": 0.8756038647342995,
"grad_norm": 0.7516986282609862,
"learning_rate": 4.642622211769099e-07,
"loss": 0.6799,
"step": 1450
},
{
"epoch": 0.8786231884057971,
"grad_norm": 0.732975469848745,
"learning_rate": 4.4233310120887387e-07,
"loss": 0.6773,
"step": 1455
},
{
"epoch": 0.8816425120772947,
"grad_norm": 0.7423980700987695,
"learning_rate": 4.209105119452628e-07,
"loss": 0.6775,
"step": 1460
},
{
"epoch": 0.8846618357487923,
"grad_norm": 0.7479313416746293,
"learning_rate": 3.999968342531918e-07,
"loss": 0.6635,
"step": 1465
},
{
"epoch": 0.8876811594202898,
"grad_norm": 0.7798737107901984,
"learning_rate": 3.7959439244027727e-07,
"loss": 0.6745,
"step": 1470
},
{
"epoch": 0.8907004830917874,
"grad_norm": 0.7387373917982181,
"learning_rate": 3.5970545399632574e-07,
"loss": 0.6744,
"step": 1475
},
{
"epoch": 0.893719806763285,
"grad_norm": 0.7337344483315793,
"learning_rate": 3.4033222934131914e-07,
"loss": 0.6755,
"step": 1480
},
{
"epoch": 0.8967391304347826,
"grad_norm": 0.7378866951920182,
"learning_rate": 3.214768715797656e-07,
"loss": 0.6642,
"step": 1485
},
{
"epoch": 0.8997584541062802,
"grad_norm": 0.7461379146470514,
"learning_rate": 3.0314147626139543e-07,
"loss": 0.6808,
"step": 1490
},
{
"epoch": 0.9027777777777778,
"grad_norm": 0.757193556184507,
"learning_rate": 2.853280811482734e-07,
"loss": 0.6693,
"step": 1495
},
{
"epoch": 0.9057971014492754,
"grad_norm": 0.7415417205678285,
"learning_rate": 2.6803866598832216e-07,
"loss": 0.6758,
"step": 1500
},
{
"epoch": 0.908816425120773,
"grad_norm": 0.771131359954168,
"learning_rate": 2.5127515229529665e-07,
"loss": 0.682,
"step": 1505
},
{
"epoch": 0.9118357487922706,
"grad_norm": 0.7370379289888651,
"learning_rate": 2.350394031352343e-07,
"loss": 0.6842,
"step": 1510
},
{
"epoch": 0.9148550724637681,
"grad_norm": 0.738116305395423,
"learning_rate": 2.1933322291938897e-07,
"loss": 0.6763,
"step": 1515
},
{
"epoch": 0.9178743961352657,
"grad_norm": 0.7664616804871662,
"learning_rate": 2.041583572037037e-07,
"loss": 0.6884,
"step": 1520
},
{
"epoch": 0.9208937198067633,
"grad_norm": 0.7361049624043503,
"learning_rate": 1.8951649249480287e-07,
"loss": 0.6854,
"step": 1525
},
{
"epoch": 0.9239130434782609,
"grad_norm": 0.73972951148021,
"learning_rate": 1.7540925606256088e-07,
"loss": 0.6956,
"step": 1530
},
{
"epoch": 0.9269323671497585,
"grad_norm": 0.7382244412309273,
"learning_rate": 1.6183821575925186e-07,
"loss": 0.6738,
"step": 1535
},
{
"epoch": 0.9299516908212561,
"grad_norm": 0.7276987617989338,
"learning_rate": 1.4880487984529846e-07,
"loss": 0.6815,
"step": 1540
},
{
"epoch": 0.9329710144927537,
"grad_norm": 0.7527483196420391,
"learning_rate": 1.363106968216482e-07,
"loss": 0.6863,
"step": 1545
},
{
"epoch": 0.9359903381642513,
"grad_norm": 0.7602203209981125,
"learning_rate": 1.243570552687895e-07,
"loss": 0.6869,
"step": 1550
},
{
"epoch": 0.9390096618357487,
"grad_norm": 0.7591679238485205,
"learning_rate": 1.1294528369242663e-07,
"loss": 0.6897,
"step": 1555
},
{
"epoch": 0.9420289855072463,
"grad_norm": 0.7580197201897864,
"learning_rate": 1.020766503758347e-07,
"loss": 0.6862,
"step": 1560
},
{
"epoch": 0.9450483091787439,
"grad_norm": 0.7274440508843062,
"learning_rate": 9.175236323890058e-08,
"loss": 0.6798,
"step": 1565
},
{
"epoch": 0.9480676328502415,
"grad_norm": 0.7517750389382779,
"learning_rate": 8.197356970388148e-08,
"loss": 0.6758,
"step": 1570
},
{
"epoch": 0.9510869565217391,
"grad_norm": 0.7581934148074195,
"learning_rate": 7.274135656787917e-08,
"loss": 0.6871,
"step": 1575
},
{
"epoch": 0.9541062801932367,
"grad_norm": 0.7349286966931822,
"learning_rate": 6.405674988205602e-08,
"loss": 0.6709,
"step": 1580
},
{
"epoch": 0.9571256038647343,
"grad_norm": 0.7492094371596764,
"learning_rate": 5.592071483760397e-08,
"loss": 0.6948,
"step": 1585
},
{
"epoch": 0.9601449275362319,
"grad_norm": 0.7600332357347815,
"learning_rate": 4.833415565847155e-08,
"loss": 0.6798,
"step": 1590
},
{
"epoch": 0.9631642512077294,
"grad_norm": 0.7647034888713604,
"learning_rate": 4.1297915500873034e-08,
"loss": 0.6843,
"step": 1595
},
{
"epoch": 0.966183574879227,
"grad_norm": 0.7434041279358656,
"learning_rate": 3.481277635957903e-08,
"loss": 0.6875,
"step": 1600
},
{
"epoch": 0.9692028985507246,
"grad_norm": 0.7363545932702806,
"learning_rate": 2.88794589810093e-08,
"loss": 0.686,
"step": 1605
},
{
"epoch": 0.9722222222222222,
"grad_norm": 0.7276212514725846,
"learning_rate": 2.3498622783128533e-08,
"loss": 0.6743,
"step": 1610
},
{
"epoch": 0.9752415458937198,
"grad_norm": 0.7350554163815118,
"learning_rate": 1.8670865782161042e-08,
"loss": 0.6838,
"step": 1615
},
{
"epoch": 0.9782608695652174,
"grad_norm": 0.7452729852364519,
"learning_rate": 1.4396724526127282e-08,
"loss": 0.679,
"step": 1620
},
{
"epoch": 0.981280193236715,
"grad_norm": 0.7662828104651594,
"learning_rate": 1.067667403521433e-08,
"loss": 0.6936,
"step": 1625
},
{
"epoch": 0.9842995169082126,
"grad_norm": 0.738399747105122,
"learning_rate": 7.51112774898144e-09,
"loss": 0.6699,
"step": 1630
},
{
"epoch": 0.9873188405797102,
"grad_norm": 0.754232541083393,
"learning_rate": 4.900437480413467e-09,
"loss": 0.6742,
"step": 1635
},
{
"epoch": 0.9903381642512077,
"grad_norm": 0.7350270953242116,
"learning_rate": 2.844893376816593e-09,
"loss": 0.6628,
"step": 1640
},
{
"epoch": 0.9933574879227053,
"grad_norm": 0.7236271614987632,
"learning_rate": 1.3447238875774482e-09,
"loss": 0.6716,
"step": 1645
},
{
"epoch": 0.9963768115942029,
"grad_norm": 0.7622443185044537,
"learning_rate": 4.000957387700899e-10,
"loss": 0.6832,
"step": 1650
},
{
"epoch": 0.9993961352657005,
"grad_norm": 0.7520131293111865,
"learning_rate": 1.1113914626381672e-11,
"loss": 0.6816,
"step": 1655
},
{
"epoch": 1.0,
"eval_runtime": 1.6679,
"eval_samples_per_second": 5.996,
"eval_steps_per_second": 1.799,
"step": 1656
},
{
"epoch": 1.0,
"step": 1656,
"total_flos": 114021718032384.0,
"train_loss": 0.7685830109361289,
"train_runtime": 7351.6585,
"train_samples_per_second": 3.603,
"train_steps_per_second": 0.225
}
],
"logging_steps": 5,
"max_steps": 1656,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 114021718032384.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}