adpretko's picture
final save
70b2ded verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9782037409182254,
"eval_steps": 500,
"global_step": 3200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006183335909723296,
"grad_norm": 0.04143473133444786,
"learning_rate": 5.555555555555555e-07,
"loss": 0.0078,
"step": 10
},
{
"epoch": 0.012366671819446591,
"grad_norm": 0.039171501994132996,
"learning_rate": 1.1728395061728396e-06,
"loss": 0.0073,
"step": 20
},
{
"epoch": 0.018550007729169888,
"grad_norm": 0.045620836317539215,
"learning_rate": 1.7901234567901235e-06,
"loss": 0.0077,
"step": 30
},
{
"epoch": 0.024733343638893183,
"grad_norm": 0.042504891753196716,
"learning_rate": 2.4074074074074075e-06,
"loss": 0.0074,
"step": 40
},
{
"epoch": 0.030916679548616478,
"grad_norm": 0.04532310739159584,
"learning_rate": 3.0246913580246917e-06,
"loss": 0.0078,
"step": 50
},
{
"epoch": 0.037100015458339776,
"grad_norm": 0.046271760016679764,
"learning_rate": 3.641975308641976e-06,
"loss": 0.008,
"step": 60
},
{
"epoch": 0.04328335136806307,
"grad_norm": 0.046998776495456696,
"learning_rate": 4.2592592592592596e-06,
"loss": 0.0082,
"step": 70
},
{
"epoch": 0.049466687277786366,
"grad_norm": 0.06645342707633972,
"learning_rate": 4.876543209876544e-06,
"loss": 0.008,
"step": 80
},
{
"epoch": 0.05565002318750966,
"grad_norm": 0.05185055360198021,
"learning_rate": 5.493827160493828e-06,
"loss": 0.0077,
"step": 90
},
{
"epoch": 0.061833359097232955,
"grad_norm": 0.06297408044338226,
"learning_rate": 6.111111111111112e-06,
"loss": 0.0084,
"step": 100
},
{
"epoch": 0.06801669500695626,
"grad_norm": 0.08234790712594986,
"learning_rate": 6.728395061728395e-06,
"loss": 0.008,
"step": 110
},
{
"epoch": 0.07420003091667955,
"grad_norm": 0.05353984236717224,
"learning_rate": 7.34567901234568e-06,
"loss": 0.0083,
"step": 120
},
{
"epoch": 0.08038336682640285,
"grad_norm": 0.060335684567689896,
"learning_rate": 7.962962962962963e-06,
"loss": 0.0081,
"step": 130
},
{
"epoch": 0.08656670273612614,
"grad_norm": 0.06284262984991074,
"learning_rate": 8.580246913580249e-06,
"loss": 0.0084,
"step": 140
},
{
"epoch": 0.09275003864584944,
"grad_norm": 0.08134513348340988,
"learning_rate": 9.197530864197531e-06,
"loss": 0.0087,
"step": 150
},
{
"epoch": 0.09893337455557273,
"grad_norm": 0.06377002596855164,
"learning_rate": 9.814814814814815e-06,
"loss": 0.0087,
"step": 160
},
{
"epoch": 0.10511671046529603,
"grad_norm": 0.07254844158887863,
"learning_rate": 1.04320987654321e-05,
"loss": 0.0089,
"step": 170
},
{
"epoch": 0.11130004637501932,
"grad_norm": 0.07614797353744507,
"learning_rate": 1.1049382716049384e-05,
"loss": 0.0091,
"step": 180
},
{
"epoch": 0.11748338228474262,
"grad_norm": 0.09067586809396744,
"learning_rate": 1.1666666666666668e-05,
"loss": 0.0084,
"step": 190
},
{
"epoch": 0.12366671819446591,
"grad_norm": 0.06307026743888855,
"learning_rate": 1.228395061728395e-05,
"loss": 0.0095,
"step": 200
},
{
"epoch": 0.1298500541041892,
"grad_norm": 0.05498047545552254,
"learning_rate": 1.2901234567901235e-05,
"loss": 0.0087,
"step": 210
},
{
"epoch": 0.13603339001391251,
"grad_norm": 0.08366186171770096,
"learning_rate": 1.351851851851852e-05,
"loss": 0.009,
"step": 220
},
{
"epoch": 0.1422167259236358,
"grad_norm": 0.08589412271976471,
"learning_rate": 1.4135802469135805e-05,
"loss": 0.009,
"step": 230
},
{
"epoch": 0.1484000618333591,
"grad_norm": 0.05334419757127762,
"learning_rate": 1.4753086419753087e-05,
"loss": 0.0093,
"step": 240
},
{
"epoch": 0.15458339774308238,
"grad_norm": 0.06228677183389664,
"learning_rate": 1.537037037037037e-05,
"loss": 0.0091,
"step": 250
},
{
"epoch": 0.1607667336528057,
"grad_norm": 0.059001028537750244,
"learning_rate": 1.5987654320987655e-05,
"loss": 0.0088,
"step": 260
},
{
"epoch": 0.16695006956252897,
"grad_norm": 0.056203559041023254,
"learning_rate": 1.660493827160494e-05,
"loss": 0.0092,
"step": 270
},
{
"epoch": 0.17313340547225228,
"grad_norm": 0.08194007724523544,
"learning_rate": 1.7222222222222224e-05,
"loss": 0.0095,
"step": 280
},
{
"epoch": 0.17931674138197556,
"grad_norm": 1.0690827369689941,
"learning_rate": 1.7839506172839506e-05,
"loss": 0.0282,
"step": 290
},
{
"epoch": 0.18550007729169887,
"grad_norm": 0.09711379557847977,
"learning_rate": 1.8456790123456792e-05,
"loss": 0.016,
"step": 300
},
{
"epoch": 0.19168341320142215,
"grad_norm": 0.05902743339538574,
"learning_rate": 1.9074074074074075e-05,
"loss": 0.0118,
"step": 310
},
{
"epoch": 0.19786674911114546,
"grad_norm": 0.058570075780153275,
"learning_rate": 1.969135802469136e-05,
"loss": 0.0105,
"step": 320
},
{
"epoch": 0.20405008502086877,
"grad_norm": 0.05820102244615555,
"learning_rate": 1.9999854512405252e-05,
"loss": 0.0098,
"step": 330
},
{
"epoch": 0.21023342093059205,
"grad_norm": 0.05199410021305084,
"learning_rate": 1.9998690637047088e-05,
"loss": 0.0092,
"step": 340
},
{
"epoch": 0.21641675684031536,
"grad_norm": 0.0505528561770916,
"learning_rate": 1.999636302179332e-05,
"loss": 0.0093,
"step": 350
},
{
"epoch": 0.22260009275003864,
"grad_norm": 0.052951931953430176,
"learning_rate": 1.9992871937553292e-05,
"loss": 0.0089,
"step": 360
},
{
"epoch": 0.22878342865976195,
"grad_norm": 0.053663648664951324,
"learning_rate": 1.9988217790651605e-05,
"loss": 0.0086,
"step": 370
},
{
"epoch": 0.23496676456948523,
"grad_norm": 0.052866872400045395,
"learning_rate": 1.9982401122780834e-05,
"loss": 0.0093,
"step": 380
},
{
"epoch": 0.24115010047920854,
"grad_norm": 0.044779375195503235,
"learning_rate": 1.9975422610938463e-05,
"loss": 0.0087,
"step": 390
},
{
"epoch": 0.24733343638893182,
"grad_norm": 0.05273430049419403,
"learning_rate": 1.9967283067348106e-05,
"loss": 0.0087,
"step": 400
},
{
"epoch": 0.2535167722986551,
"grad_norm": 0.05107806622982025,
"learning_rate": 1.9957983439364973e-05,
"loss": 0.0084,
"step": 410
},
{
"epoch": 0.2597001082083784,
"grad_norm": 0.05193234607577324,
"learning_rate": 1.994752480936559e-05,
"loss": 0.0085,
"step": 420
},
{
"epoch": 0.2658834441181017,
"grad_norm": 0.04062167927622795,
"learning_rate": 1.9935908394621844e-05,
"loss": 0.0085,
"step": 430
},
{
"epoch": 0.27206678002782503,
"grad_norm": 0.04734641686081886,
"learning_rate": 1.992313554715929e-05,
"loss": 0.0087,
"step": 440
},
{
"epoch": 0.2782501159375483,
"grad_norm": 0.0560256652534008,
"learning_rate": 1.99092077535998e-05,
"loss": 0.0086,
"step": 450
},
{
"epoch": 0.2844334518472716,
"grad_norm": 0.05063237249851227,
"learning_rate": 1.989412663498853e-05,
"loss": 0.0084,
"step": 460
},
{
"epoch": 0.2906167877569949,
"grad_norm": 0.05622229725122452,
"learning_rate": 1.9877893946605247e-05,
"loss": 0.0085,
"step": 470
},
{
"epoch": 0.2968001236667182,
"grad_norm": 0.06787122786045074,
"learning_rate": 1.9860511577760042e-05,
"loss": 0.0084,
"step": 480
},
{
"epoch": 0.3029834595764415,
"grad_norm": 0.05340688303112984,
"learning_rate": 1.9841981551573424e-05,
"loss": 0.0084,
"step": 490
},
{
"epoch": 0.30916679548616477,
"grad_norm": 0.05143291503190994,
"learning_rate": 1.9822306024740855e-05,
"loss": 0.0085,
"step": 500
},
{
"epoch": 0.3153501313958881,
"grad_norm": 0.04688792675733566,
"learning_rate": 1.980148728728174e-05,
"loss": 0.0085,
"step": 510
},
{
"epoch": 0.3215334673056114,
"grad_norm": 0.04486978054046631,
"learning_rate": 1.9779527762272877e-05,
"loss": 0.0082,
"step": 520
},
{
"epoch": 0.3277168032153347,
"grad_norm": 0.04129200056195259,
"learning_rate": 1.975643000556646e-05,
"loss": 0.0083,
"step": 530
},
{
"epoch": 0.33390013912505795,
"grad_norm": 0.04800448194146156,
"learning_rate": 1.973219670549259e-05,
"loss": 0.0083,
"step": 540
},
{
"epoch": 0.34008347503478126,
"grad_norm": 0.05428948998451233,
"learning_rate": 1.970683068254637e-05,
"loss": 0.0081,
"step": 550
},
{
"epoch": 0.34626681094450457,
"grad_norm": 0.062362976372241974,
"learning_rate": 1.9680334889059668e-05,
"loss": 0.0081,
"step": 560
},
{
"epoch": 0.3524501468542279,
"grad_norm": 0.04794848710298538,
"learning_rate": 1.9652712408857452e-05,
"loss": 0.0081,
"step": 570
},
{
"epoch": 0.35863348276395113,
"grad_norm": 0.04279520735144615,
"learning_rate": 1.9623966456898914e-05,
"loss": 0.0078,
"step": 580
},
{
"epoch": 0.36481681867367444,
"grad_norm": 0.04028687626123428,
"learning_rate": 1.959410037890323e-05,
"loss": 0.0083,
"step": 590
},
{
"epoch": 0.37100015458339775,
"grad_norm": 0.04868888854980469,
"learning_rate": 1.9563117650960208e-05,
"loss": 0.0085,
"step": 600
},
{
"epoch": 0.37718349049312105,
"grad_norm": 0.048066020011901855,
"learning_rate": 1.9531021879125654e-05,
"loss": 0.0082,
"step": 610
},
{
"epoch": 0.3833668264028443,
"grad_norm": 0.051193494349718094,
"learning_rate": 1.9497816799001726e-05,
"loss": 0.0081,
"step": 620
},
{
"epoch": 0.3895501623125676,
"grad_norm": 0.046053316444158554,
"learning_rate": 1.9463506275302094e-05,
"loss": 0.0078,
"step": 630
},
{
"epoch": 0.3957334982222909,
"grad_norm": 0.04008670151233673,
"learning_rate": 1.9428094301402164e-05,
"loss": 0.008,
"step": 640
},
{
"epoch": 0.40191683413201423,
"grad_norm": 0.048819512128829956,
"learning_rate": 1.939158499887428e-05,
"loss": 0.0084,
"step": 650
},
{
"epoch": 0.40810017004173754,
"grad_norm": 0.04148195683956146,
"learning_rate": 1.9353982617008024e-05,
"loss": 0.0077,
"step": 660
},
{
"epoch": 0.4142835059514608,
"grad_norm": 0.04567483067512512,
"learning_rate": 1.9315291532315626e-05,
"loss": 0.0082,
"step": 670
},
{
"epoch": 0.4204668418611841,
"grad_norm": 0.0510326586663723,
"learning_rate": 1.9275516248022613e-05,
"loss": 0.0082,
"step": 680
},
{
"epoch": 0.4266501777709074,
"grad_norm": 0.053588561713695526,
"learning_rate": 1.9234661393543668e-05,
"loss": 0.0078,
"step": 690
},
{
"epoch": 0.4328335136806307,
"grad_norm": 0.0437391996383667,
"learning_rate": 1.91927317239438e-05,
"loss": 0.0082,
"step": 700
},
{
"epoch": 0.439016849590354,
"grad_norm": 0.05658455193042755,
"learning_rate": 1.9149732119384942e-05,
"loss": 0.0083,
"step": 710
},
{
"epoch": 0.4452001855000773,
"grad_norm": 0.039588961750268936,
"learning_rate": 1.9105667584557914e-05,
"loss": 0.0077,
"step": 720
},
{
"epoch": 0.4513835214098006,
"grad_norm": 0.05383581295609474,
"learning_rate": 1.9060543248099958e-05,
"loss": 0.0081,
"step": 730
},
{
"epoch": 0.4575668573195239,
"grad_norm": 0.047476038336753845,
"learning_rate": 1.9014364361997813e-05,
"loss": 0.0082,
"step": 740
},
{
"epoch": 0.46375019322924715,
"grad_norm": 0.057258907705545425,
"learning_rate": 1.896713630097643e-05,
"loss": 0.0081,
"step": 750
},
{
"epoch": 0.46993352913897046,
"grad_norm": 0.046649232506752014,
"learning_rate": 1.8918864561873428e-05,
"loss": 0.0078,
"step": 760
},
{
"epoch": 0.47611686504869377,
"grad_norm": 0.04877932369709015,
"learning_rate": 1.8869554762999307e-05,
"loss": 0.0079,
"step": 770
},
{
"epoch": 0.4823002009584171,
"grad_norm": 0.048988137394189835,
"learning_rate": 1.881921264348355e-05,
"loss": 0.008,
"step": 780
},
{
"epoch": 0.4884835368681404,
"grad_norm": 0.04841630905866623,
"learning_rate": 1.876784406260664e-05,
"loss": 0.0078,
"step": 790
},
{
"epoch": 0.49466687277786364,
"grad_norm": 0.05180073156952858,
"learning_rate": 1.871545499911811e-05,
"loss": 0.0079,
"step": 800
},
{
"epoch": 0.500850208687587,
"grad_norm": 0.04108748957514763,
"learning_rate": 1.866205155054067e-05,
"loss": 0.0079,
"step": 810
},
{
"epoch": 0.5070335445973102,
"grad_norm": 0.03657471388578415,
"learning_rate": 1.860763993246053e-05,
"loss": 0.0077,
"step": 820
},
{
"epoch": 0.5132168805070335,
"grad_norm": 0.04646613076329231,
"learning_rate": 1.8552226477803984e-05,
"loss": 0.0076,
"step": 830
},
{
"epoch": 0.5194002164167568,
"grad_norm": 0.050987377762794495,
"learning_rate": 1.8495817636100293e-05,
"loss": 0.0079,
"step": 840
},
{
"epoch": 0.5255835523264801,
"grad_norm": 0.0483928844332695,
"learning_rate": 1.8438419972731066e-05,
"loss": 0.0076,
"step": 850
},
{
"epoch": 0.5317668882362034,
"grad_norm": 0.05614335834980011,
"learning_rate": 1.8380040168166096e-05,
"loss": 0.0079,
"step": 860
},
{
"epoch": 0.5379502241459267,
"grad_norm": 0.045122936367988586,
"learning_rate": 1.8320685017185832e-05,
"loss": 0.0077,
"step": 870
},
{
"epoch": 0.5441335600556501,
"grad_norm": 0.03678731620311737,
"learning_rate": 1.8260361428090542e-05,
"loss": 0.0077,
"step": 880
},
{
"epoch": 0.5503168959653734,
"grad_norm": 0.04048831760883331,
"learning_rate": 1.8199076421896256e-05,
"loss": 0.0075,
"step": 890
},
{
"epoch": 0.5565002318750966,
"grad_norm": 0.04655023291707039,
"learning_rate": 1.813683713151761e-05,
"loss": 0.0076,
"step": 900
},
{
"epoch": 0.5626835677848199,
"grad_norm": 0.04041978716850281,
"learning_rate": 1.8073650800937627e-05,
"loss": 0.0076,
"step": 910
},
{
"epoch": 0.5688669036945432,
"grad_norm": 0.051544759422540665,
"learning_rate": 1.8009524784364615e-05,
"loss": 0.0076,
"step": 920
},
{
"epoch": 0.5750502396042665,
"grad_norm": 0.04109196737408638,
"learning_rate": 1.7944466545376215e-05,
"loss": 0.0079,
"step": 930
},
{
"epoch": 0.5812335755139898,
"grad_norm": 0.04459492862224579,
"learning_rate": 1.7878483656050713e-05,
"loss": 0.0075,
"step": 940
},
{
"epoch": 0.5874169114237131,
"grad_norm": 0.051292601972818375,
"learning_rate": 1.7811583796085722e-05,
"loss": 0.0077,
"step": 950
},
{
"epoch": 0.5936002473334364,
"grad_norm": 0.04666479676961899,
"learning_rate": 1.7743774751904385e-05,
"loss": 0.0077,
"step": 960
},
{
"epoch": 0.5997835832431597,
"grad_norm": 0.043166983872652054,
"learning_rate": 1.7675064415749078e-05,
"loss": 0.0074,
"step": 970
},
{
"epoch": 0.605966919152883,
"grad_norm": 0.0443737618625164,
"learning_rate": 1.760546078476286e-05,
"loss": 0.0077,
"step": 980
},
{
"epoch": 0.6121502550626062,
"grad_norm": 0.04099385812878609,
"learning_rate": 1.7534971960058684e-05,
"loss": 0.0076,
"step": 990
},
{
"epoch": 0.6183335909723295,
"grad_norm": 0.04541454464197159,
"learning_rate": 1.7463606145776525e-05,
"loss": 0.0076,
"step": 1000
},
{
"epoch": 0.6245169268820528,
"grad_norm": 0.0451682023704052,
"learning_rate": 1.739137164812849e-05,
"loss": 0.0075,
"step": 1010
},
{
"epoch": 0.6307002627917762,
"grad_norm": 0.04592340439558029,
"learning_rate": 1.731827687443208e-05,
"loss": 0.0077,
"step": 1020
},
{
"epoch": 0.6368835987014995,
"grad_norm": 0.04665357992053032,
"learning_rate": 1.7244330332131657e-05,
"loss": 0.0072,
"step": 1030
},
{
"epoch": 0.6430669346112228,
"grad_norm": 0.046974148601293564,
"learning_rate": 1.7169540627808276e-05,
"loss": 0.0074,
"step": 1040
},
{
"epoch": 0.6492502705209461,
"grad_norm": 0.04424068331718445,
"learning_rate": 1.7093916466177972e-05,
"loss": 0.0073,
"step": 1050
},
{
"epoch": 0.6554336064306694,
"grad_norm": 0.04412393271923065,
"learning_rate": 1.701746664907862e-05,
"loss": 0.0072,
"step": 1060
},
{
"epoch": 0.6616169423403926,
"grad_norm": 0.046212710440158844,
"learning_rate": 1.6940200074445503e-05,
"loss": 0.0073,
"step": 1070
},
{
"epoch": 0.6678002782501159,
"grad_norm": 0.05339352414011955,
"learning_rate": 1.6862125735275675e-05,
"loss": 0.0077,
"step": 1080
},
{
"epoch": 0.6739836141598392,
"grad_norm": 0.053103432059288025,
"learning_rate": 1.6783252718581296e-05,
"loss": 0.0078,
"step": 1090
},
{
"epoch": 0.6801669500695625,
"grad_norm": 0.04800717905163765,
"learning_rate": 1.6703590204331967e-05,
"loss": 0.0073,
"step": 1100
},
{
"epoch": 0.6863502859792858,
"grad_norm": 0.04264698177576065,
"learning_rate": 1.6623147464386324e-05,
"loss": 0.0075,
"step": 1110
},
{
"epoch": 0.6925336218890091,
"grad_norm": 0.040609877556562424,
"learning_rate": 1.6541933861412848e-05,
"loss": 0.0073,
"step": 1120
},
{
"epoch": 0.6987169577987324,
"grad_norm": 0.03998512774705887,
"learning_rate": 1.645995884780019e-05,
"loss": 0.0076,
"step": 1130
},
{
"epoch": 0.7049002937084557,
"grad_norm": 0.03671044111251831,
"learning_rate": 1.637723196455699e-05,
"loss": 0.0073,
"step": 1140
},
{
"epoch": 0.7110836296181791,
"grad_norm": 0.041929442435503006,
"learning_rate": 1.629376284020143e-05,
"loss": 0.0075,
"step": 1150
},
{
"epoch": 0.7172669655279023,
"grad_norm": 0.042013250291347504,
"learning_rate": 1.620956118964053e-05,
"loss": 0.0074,
"step": 1160
},
{
"epoch": 0.7234503014376256,
"grad_norm": 0.04549422115087509,
"learning_rate": 1.6124636813039502e-05,
"loss": 0.0073,
"step": 1170
},
{
"epoch": 0.7296336373473489,
"grad_norm": 0.04200827330350876,
"learning_rate": 1.6038999594681074e-05,
"loss": 0.0072,
"step": 1180
},
{
"epoch": 0.7358169732570722,
"grad_norm": 0.03912513703107834,
"learning_rate": 1.5952659501815073e-05,
"loss": 0.0073,
"step": 1190
},
{
"epoch": 0.7420003091667955,
"grad_norm": 0.03626884147524834,
"learning_rate": 1.5865626583498355e-05,
"loss": 0.0073,
"step": 1200
},
{
"epoch": 0.7481836450765188,
"grad_norm": 0.04432886838912964,
"learning_rate": 1.5777910969425184e-05,
"loss": 0.0071,
"step": 1210
},
{
"epoch": 0.7543669809862421,
"grad_norm": 0.03677462041378021,
"learning_rate": 1.5689522868748276e-05,
"loss": 0.0074,
"step": 1220
},
{
"epoch": 0.7605503168959654,
"grad_norm": 0.03566281870007515,
"learning_rate": 1.5600472568890523e-05,
"loss": 0.0071,
"step": 1230
},
{
"epoch": 0.7667336528056886,
"grad_norm": 0.037937846034765244,
"learning_rate": 1.551077043434769e-05,
"loss": 0.0072,
"step": 1240
},
{
"epoch": 0.7729169887154119,
"grad_norm": 0.034936949610710144,
"learning_rate": 1.5420426905482056e-05,
"loss": 0.0071,
"step": 1250
},
{
"epoch": 0.7791003246251352,
"grad_norm": 0.0420398972928524,
"learning_rate": 1.5329452497307316e-05,
"loss": 0.0072,
"step": 1260
},
{
"epoch": 0.7852836605348585,
"grad_norm": 0.04124725982546806,
"learning_rate": 1.5237857798264701e-05,
"loss": 0.0072,
"step": 1270
},
{
"epoch": 0.7914669964445818,
"grad_norm": 0.042076483368873596,
"learning_rate": 1.5145653468990626e-05,
"loss": 0.0071,
"step": 1280
},
{
"epoch": 0.7976503323543052,
"grad_norm": 0.03725574165582657,
"learning_rate": 1.50528502410759e-05,
"loss": 0.0069,
"step": 1290
},
{
"epoch": 0.8038336682640285,
"grad_norm": 0.04050353169441223,
"learning_rate": 1.4959458915816681e-05,
"loss": 0.0072,
"step": 1300
},
{
"epoch": 0.8100170041737518,
"grad_norm": 0.03882165625691414,
"learning_rate": 1.4865490362957317e-05,
"loss": 0.0073,
"step": 1310
},
{
"epoch": 0.8162003400834751,
"grad_norm": 0.03756791725754738,
"learning_rate": 1.477095551942525e-05,
"loss": 0.0074,
"step": 1320
},
{
"epoch": 0.8223836759931983,
"grad_norm": 0.05022745579481125,
"learning_rate": 1.4675865388058037e-05,
"loss": 0.007,
"step": 1330
},
{
"epoch": 0.8285670119029216,
"grad_norm": 0.04937414452433586,
"learning_rate": 1.458023103632277e-05,
"loss": 0.0071,
"step": 1340
},
{
"epoch": 0.8347503478126449,
"grad_norm": 0.051321037113666534,
"learning_rate": 1.4484063595027924e-05,
"loss": 0.0071,
"step": 1350
},
{
"epoch": 0.8409336837223682,
"grad_norm": 0.035162657499313354,
"learning_rate": 1.438737425702786e-05,
"loss": 0.0071,
"step": 1360
},
{
"epoch": 0.8471170196320915,
"grad_norm": 0.03311579301953316,
"learning_rate": 1.4290174275920095e-05,
"loss": 0.0072,
"step": 1370
},
{
"epoch": 0.8533003555418148,
"grad_norm": 0.04103536903858185,
"learning_rate": 1.41924749647355e-05,
"loss": 0.0071,
"step": 1380
},
{
"epoch": 0.8594836914515381,
"grad_norm": 0.051637377589941025,
"learning_rate": 1.4094287694621589e-05,
"loss": 0.0069,
"step": 1390
},
{
"epoch": 0.8656670273612614,
"grad_norm": 0.033984847366809845,
"learning_rate": 1.3995623893519044e-05,
"loss": 0.007,
"step": 1400
},
{
"epoch": 0.8718503632709848,
"grad_norm": 0.03714025020599365,
"learning_rate": 1.3896495044831622e-05,
"loss": 0.0066,
"step": 1410
},
{
"epoch": 0.878033699180708,
"grad_norm": 0.041570328176021576,
"learning_rate": 1.37969126860896e-05,
"loss": 0.0067,
"step": 1420
},
{
"epoch": 0.8842170350904313,
"grad_norm": 0.04341095685958862,
"learning_rate": 1.3696888407606952e-05,
"loss": 0.0069,
"step": 1430
},
{
"epoch": 0.8904003710001546,
"grad_norm": 0.039629433304071426,
"learning_rate": 1.3596433851132342e-05,
"loss": 0.0069,
"step": 1440
},
{
"epoch": 0.8965837069098779,
"grad_norm": 0.04006121680140495,
"learning_rate": 1.3495560708494167e-05,
"loss": 0.007,
"step": 1450
},
{
"epoch": 0.9027670428196012,
"grad_norm": 0.04321839660406113,
"learning_rate": 1.3394280720239733e-05,
"loss": 0.007,
"step": 1460
},
{
"epoch": 0.9089503787293245,
"grad_norm": 0.038396500051021576,
"learning_rate": 1.3292605674268803e-05,
"loss": 0.0069,
"step": 1470
},
{
"epoch": 0.9151337146390478,
"grad_norm": 0.040967196226119995,
"learning_rate": 1.31905474044616e-05,
"loss": 0.007,
"step": 1480
},
{
"epoch": 0.9213170505487711,
"grad_norm": 0.048842981457710266,
"learning_rate": 1.3088117789301473e-05,
"loss": 0.0067,
"step": 1490
},
{
"epoch": 0.9275003864584943,
"grad_norm": 0.03724457696080208,
"learning_rate": 1.298532875049237e-05,
"loss": 0.0069,
"step": 1500
},
{
"epoch": 0.9336837223682176,
"grad_norm": 0.039507653564214706,
"learning_rate": 1.2882192251571282e-05,
"loss": 0.0068,
"step": 1510
},
{
"epoch": 0.9398670582779409,
"grad_norm": 0.035300228744745255,
"learning_rate": 1.2778720296515816e-05,
"loss": 0.0068,
"step": 1520
},
{
"epoch": 0.9460503941876642,
"grad_norm": 0.05400003492832184,
"learning_rate": 1.2674924928347055e-05,
"loss": 0.0067,
"step": 1530
},
{
"epoch": 0.9522337300973875,
"grad_norm": 0.04334673658013344,
"learning_rate": 1.2570818227727889e-05,
"loss": 0.0068,
"step": 1540
},
{
"epoch": 0.9584170660071109,
"grad_norm": 0.03568002209067345,
"learning_rate": 1.2466412311556952e-05,
"loss": 0.0068,
"step": 1550
},
{
"epoch": 0.9646004019168342,
"grad_norm": 0.047103166580200195,
"learning_rate": 1.2361719331558346e-05,
"loss": 0.0068,
"step": 1560
},
{
"epoch": 0.9707837378265575,
"grad_norm": 0.036910735070705414,
"learning_rate": 1.2256751472867304e-05,
"loss": 0.0067,
"step": 1570
},
{
"epoch": 0.9769670737362808,
"grad_norm": 0.04214445874094963,
"learning_rate": 1.2151520952611984e-05,
"loss": 0.0067,
"step": 1580
},
{
"epoch": 0.983150409646004,
"grad_norm": 0.05339273810386658,
"learning_rate": 1.2046040018491523e-05,
"loss": 0.0066,
"step": 1590
},
{
"epoch": 0.9893337455557273,
"grad_norm": 0.04332943633198738,
"learning_rate": 1.1940320947350532e-05,
"loss": 0.007,
"step": 1600
},
{
"epoch": 0.9955170814654506,
"grad_norm": 0.0429212786257267,
"learning_rate": 1.1834376043750201e-05,
"loss": 0.0063,
"step": 1610
},
{
"epoch": 1.0012366671819446,
"grad_norm": 0.04421928524971008,
"learning_rate": 1.1728217638536196e-05,
"loss": 0.0071,
"step": 1620
},
{
"epoch": 1.007420003091668,
"grad_norm": 0.04365362226963043,
"learning_rate": 1.1621858087403466e-05,
"loss": 0.0062,
"step": 1630
},
{
"epoch": 1.0136033390013912,
"grad_norm": 0.03700665757060051,
"learning_rate": 1.1515309769458174e-05,
"loss": 0.0059,
"step": 1640
},
{
"epoch": 1.0197866749111146,
"grad_norm": 0.0455857552587986,
"learning_rate": 1.1408585085776916e-05,
"loss": 0.0059,
"step": 1650
},
{
"epoch": 1.0259700108208378,
"grad_norm": 0.03983158990740776,
"learning_rate": 1.1301696457963352e-05,
"loss": 0.0062,
"step": 1660
},
{
"epoch": 1.0321533467305612,
"grad_norm": 0.03655843436717987,
"learning_rate": 1.1194656326702474e-05,
"loss": 0.0062,
"step": 1670
},
{
"epoch": 1.0383366826402844,
"grad_norm": 0.048476699739694595,
"learning_rate": 1.1087477150312653e-05,
"loss": 0.006,
"step": 1680
},
{
"epoch": 1.0445200185500076,
"grad_norm": 0.04713127389550209,
"learning_rate": 1.098017140329561e-05,
"loss": 0.0059,
"step": 1690
},
{
"epoch": 1.050703354459731,
"grad_norm": 0.03435790538787842,
"learning_rate": 1.0872751574884512e-05,
"loss": 0.006,
"step": 1700
},
{
"epoch": 1.0568866903694543,
"grad_norm": 0.04033854603767395,
"learning_rate": 1.0765230167590385e-05,
"loss": 0.0056,
"step": 1710
},
{
"epoch": 1.0630700262791777,
"grad_norm": 0.03484448045492172,
"learning_rate": 1.0657619695746944e-05,
"loss": 0.0061,
"step": 1720
},
{
"epoch": 1.0692533621889009,
"grad_norm": 0.041448500007390976,
"learning_rate": 1.054993268405405e-05,
"loss": 0.006,
"step": 1730
},
{
"epoch": 1.0754366980986243,
"grad_norm": 0.045591264963150024,
"learning_rate": 1.0442181666119984e-05,
"loss": 0.006,
"step": 1740
},
{
"epoch": 1.0816200340083475,
"grad_norm": 0.036942582577466965,
"learning_rate": 1.033437918300266e-05,
"loss": 0.006,
"step": 1750
},
{
"epoch": 1.087803369918071,
"grad_norm": 0.03570098802447319,
"learning_rate": 1.0226537781749988e-05,
"loss": 0.0058,
"step": 1760
},
{
"epoch": 1.0939867058277941,
"grad_norm": 0.0382462739944458,
"learning_rate": 1.0118670013939526e-05,
"loss": 0.0058,
"step": 1770
},
{
"epoch": 1.1001700417375173,
"grad_norm": 0.03386927768588066,
"learning_rate": 1.0010788434217625e-05,
"loss": 0.0059,
"step": 1780
},
{
"epoch": 1.1063533776472407,
"grad_norm": 0.04100915789604187,
"learning_rate": 9.902905598838177e-06,
"loss": 0.0059,
"step": 1790
},
{
"epoch": 1.112536713556964,
"grad_norm": 0.03817896544933319,
"learning_rate": 9.795034064201235e-06,
"loss": 0.0058,
"step": 1800
},
{
"epoch": 1.1187200494666873,
"grad_norm": 0.04328783601522446,
"learning_rate": 9.68718638539156e-06,
"loss": 0.0059,
"step": 1810
},
{
"epoch": 1.1249033853764105,
"grad_norm": 0.038214631378650665,
"learning_rate": 9.579375114717352e-06,
"loss": 0.0058,
"step": 1820
},
{
"epoch": 1.1310867212861337,
"grad_norm": 0.03657764196395874,
"learning_rate": 9.471612800249295e-06,
"loss": 0.0059,
"step": 1830
},
{
"epoch": 1.1372700571958572,
"grad_norm": 0.045281436294317245,
"learning_rate": 9.363911984360104e-06,
"loss": 0.0059,
"step": 1840
},
{
"epoch": 1.1434533931055806,
"grad_norm": 0.03649020567536354,
"learning_rate": 9.256285202264716e-06,
"loss": 0.0061,
"step": 1850
},
{
"epoch": 1.1496367290153038,
"grad_norm": 0.03974698483943939,
"learning_rate": 9.148744980561332e-06,
"loss": 0.0056,
"step": 1860
},
{
"epoch": 1.155820064925027,
"grad_norm": 0.041133228689432144,
"learning_rate": 9.041303835773458e-06,
"loss": 0.006,
"step": 1870
},
{
"epoch": 1.1620034008347504,
"grad_norm": 0.039168015122413635,
"learning_rate": 8.933974272893113e-06,
"loss": 0.0058,
"step": 1880
},
{
"epoch": 1.1681867367444736,
"grad_norm": 0.06035231426358223,
"learning_rate": 8.826768783925385e-06,
"loss": 0.0058,
"step": 1890
},
{
"epoch": 1.174370072654197,
"grad_norm": 0.04310515522956848,
"learning_rate": 8.719699846434493e-06,
"loss": 0.0057,
"step": 1900
},
{
"epoch": 1.1805534085639202,
"grad_norm": 0.03753894194960594,
"learning_rate": 8.612779922091534e-06,
"loss": 0.0055,
"step": 1910
},
{
"epoch": 1.1867367444736434,
"grad_norm": 0.0496654249727726,
"learning_rate": 8.506021455224083e-06,
"loss": 0.0056,
"step": 1920
},
{
"epoch": 1.1929200803833668,
"grad_norm": 0.03786466270685196,
"learning_rate": 8.399436871367804e-06,
"loss": 0.0059,
"step": 1930
},
{
"epoch": 1.19910341629309,
"grad_norm": 0.034229736775159836,
"learning_rate": 8.293038575820249e-06,
"loss": 0.0058,
"step": 1940
},
{
"epoch": 1.2052867522028134,
"grad_norm": 0.03900885581970215,
"learning_rate": 8.186838952197019e-06,
"loss": 0.0057,
"step": 1950
},
{
"epoch": 1.2114700881125366,
"grad_norm": 0.036625415086746216,
"learning_rate": 8.080850360990439e-06,
"loss": 0.0056,
"step": 1960
},
{
"epoch": 1.21765342402226,
"grad_norm": 0.032340049743652344,
"learning_rate": 7.975085138130938e-06,
"loss": 0.0056,
"step": 1970
},
{
"epoch": 1.2238367599319833,
"grad_norm": 0.03607684373855591,
"learning_rate": 7.869555593551277e-06,
"loss": 0.0059,
"step": 1980
},
{
"epoch": 1.2300200958417067,
"grad_norm": 0.032486408948898315,
"learning_rate": 7.764274009753796e-06,
"loss": 0.0059,
"step": 1990
},
{
"epoch": 1.2362034317514299,
"grad_norm": 0.03523161634802818,
"learning_rate": 7.659252640380876e-06,
"loss": 0.0056,
"step": 2000
},
{
"epoch": 1.242386767661153,
"grad_norm": 0.036453727632761,
"learning_rate": 7.554503708788738e-06,
"loss": 0.0055,
"step": 2010
},
{
"epoch": 1.2485701035708765,
"grad_norm": 0.040288910269737244,
"learning_rate": 7.450039406624792e-06,
"loss": 0.0057,
"step": 2020
},
{
"epoch": 1.2547534394806,
"grad_norm": 0.0430009663105011,
"learning_rate": 7.345871892408631e-06,
"loss": 0.0056,
"step": 2030
},
{
"epoch": 1.2609367753903231,
"grad_norm": 0.03574775531888008,
"learning_rate": 7.242013290116944e-06,
"loss": 0.0057,
"step": 2040
},
{
"epoch": 1.2671201113000463,
"grad_norm": 0.03812326118350029,
"learning_rate": 7.138475687772381e-06,
"loss": 0.0056,
"step": 2050
},
{
"epoch": 1.2733034472097697,
"grad_norm": 0.03299403935670853,
"learning_rate": 7.03527113603667e-06,
"loss": 0.0058,
"step": 2060
},
{
"epoch": 1.279486783119493,
"grad_norm": 0.03786725923418999,
"learning_rate": 6.9324116468080305e-06,
"loss": 0.0056,
"step": 2070
},
{
"epoch": 1.2856701190292164,
"grad_norm": 0.03449017554521561,
"learning_rate": 6.829909191823121e-06,
"loss": 0.0054,
"step": 2080
},
{
"epoch": 1.2918534549389395,
"grad_norm": 0.03949199244379997,
"learning_rate": 6.727775701263666e-06,
"loss": 0.0058,
"step": 2090
},
{
"epoch": 1.2980367908486627,
"grad_norm": 0.03537849709391594,
"learning_rate": 6.626023062367894e-06,
"loss": 0.006,
"step": 2100
},
{
"epoch": 1.3042201267583862,
"grad_norm": 0.03561354801058769,
"learning_rate": 6.524663118047015e-06,
"loss": 0.0058,
"step": 2110
},
{
"epoch": 1.3104034626681094,
"grad_norm": 0.037109751254320145,
"learning_rate": 6.423707665506812e-06,
"loss": 0.0056,
"step": 2120
},
{
"epoch": 1.3165867985778328,
"grad_norm": 0.03227913752198219,
"learning_rate": 6.323168454874582e-06,
"loss": 0.0054,
"step": 2130
},
{
"epoch": 1.322770134487556,
"grad_norm": 0.034376900643110275,
"learning_rate": 6.223057187831548e-06,
"loss": 0.0057,
"step": 2140
},
{
"epoch": 1.3289534703972794,
"grad_norm": 0.042623914778232574,
"learning_rate": 6.123385516250911e-06,
"loss": 0.0054,
"step": 2150
},
{
"epoch": 1.3351368063070026,
"grad_norm": 0.03284977748990059,
"learning_rate": 6.02416504084169e-06,
"loss": 0.0056,
"step": 2160
},
{
"epoch": 1.341320142216726,
"grad_norm": 0.03773827850818634,
"learning_rate": 5.925407309798535e-06,
"loss": 0.0061,
"step": 2170
},
{
"epoch": 1.3475034781264492,
"grad_norm": 0.03684335947036743,
"learning_rate": 5.827123817457631e-06,
"loss": 0.0053,
"step": 2180
},
{
"epoch": 1.3536868140361724,
"grad_norm": 0.03831595927476883,
"learning_rate": 5.729326002958889e-06,
"loss": 0.0057,
"step": 2190
},
{
"epoch": 1.3598701499458958,
"grad_norm": 0.03576178103685379,
"learning_rate": 5.632025248914546e-06,
"loss": 0.0054,
"step": 2200
},
{
"epoch": 1.366053485855619,
"grad_norm": 0.04004263877868652,
"learning_rate": 5.535232880084373e-06,
"loss": 0.0057,
"step": 2210
},
{
"epoch": 1.3722368217653425,
"grad_norm": 0.03518558666110039,
"learning_rate": 5.438960162057548e-06,
"loss": 0.0054,
"step": 2220
},
{
"epoch": 1.3784201576750656,
"grad_norm": 0.03575367107987404,
"learning_rate": 5.343218299941517e-06,
"loss": 0.0055,
"step": 2230
},
{
"epoch": 1.384603493584789,
"grad_norm": 0.03193347156047821,
"learning_rate": 5.248018437057807e-06,
"loss": 0.0055,
"step": 2240
},
{
"epoch": 1.3907868294945123,
"grad_norm": 0.03565206378698349,
"learning_rate": 5.15337165364507e-06,
"loss": 0.0054,
"step": 2250
},
{
"epoch": 1.3969701654042357,
"grad_norm": 0.0318601168692112,
"learning_rate": 5.059288965569471e-06,
"loss": 0.0054,
"step": 2260
},
{
"epoch": 1.4031535013139589,
"grad_norm": 0.03948667645454407,
"learning_rate": 4.965781323042547e-06,
"loss": 0.0054,
"step": 2270
},
{
"epoch": 1.409336837223682,
"grad_norm": 0.03156143054366112,
"learning_rate": 4.872859609346733e-06,
"loss": 0.0053,
"step": 2280
},
{
"epoch": 1.4155201731334055,
"grad_norm": 0.03467686474323273,
"learning_rate": 4.780534639568651e-06,
"loss": 0.0053,
"step": 2290
},
{
"epoch": 1.4217035090431287,
"grad_norm": 0.03605823591351509,
"learning_rate": 4.68881715934037e-06,
"loss": 0.0055,
"step": 2300
},
{
"epoch": 1.4278868449528521,
"grad_norm": 0.03542206436395645,
"learning_rate": 4.597717843588708e-06,
"loss": 0.0054,
"step": 2310
},
{
"epoch": 1.4340701808625753,
"grad_norm": 0.03765241429209709,
"learning_rate": 4.5072472952928015e-06,
"loss": 0.0055,
"step": 2320
},
{
"epoch": 1.4402535167722987,
"grad_norm": 0.03652080148458481,
"learning_rate": 4.417416044250026e-06,
"loss": 0.0057,
"step": 2330
},
{
"epoch": 1.446436852682022,
"grad_norm": 0.03262539580464363,
"learning_rate": 4.328234545850441e-06,
"loss": 0.0052,
"step": 2340
},
{
"epoch": 1.4526201885917454,
"grad_norm": 0.03427337110042572,
"learning_rate": 4.239713179859901e-06,
"loss": 0.0053,
"step": 2350
},
{
"epoch": 1.4588035245014686,
"grad_norm": 0.03885317221283913,
"learning_rate": 4.151862249211953e-06,
"loss": 0.0054,
"step": 2360
},
{
"epoch": 1.4649868604111917,
"grad_norm": 0.03729204460978508,
"learning_rate": 4.064691978808699e-06,
"loss": 0.0053,
"step": 2370
},
{
"epoch": 1.4711701963209152,
"grad_norm": 0.032668232917785645,
"learning_rate": 3.978212514330703e-06,
"loss": 0.0053,
"step": 2380
},
{
"epoch": 1.4773535322306384,
"grad_norm": 0.03400380164384842,
"learning_rate": 3.892433921056184e-06,
"loss": 0.0053,
"step": 2390
},
{
"epoch": 1.4835368681403618,
"grad_norm": 0.03154066205024719,
"learning_rate": 3.8073661826894837e-06,
"loss": 0.0052,
"step": 2400
},
{
"epoch": 1.489720204050085,
"grad_norm": 0.0407693088054657,
"learning_rate": 3.7230192001990963e-06,
"loss": 0.0047,
"step": 2410
},
{
"epoch": 1.4959035399598082,
"grad_norm": 0.03409425541758537,
"learning_rate": 3.6394027906652927e-06,
"loss": 0.0048,
"step": 2420
},
{
"epoch": 1.5020868758695316,
"grad_norm": 0.03952354937791824,
"learning_rate": 3.5565266861375236e-06,
"loss": 0.0052,
"step": 2430
},
{
"epoch": 1.508270211779255,
"grad_norm": 0.035029612481594086,
"learning_rate": 3.4744005325017073e-06,
"loss": 0.0046,
"step": 2440
},
{
"epoch": 1.5144535476889782,
"grad_norm": 0.04004015028476715,
"learning_rate": 3.3930338883575564e-06,
"loss": 0.0048,
"step": 2450
},
{
"epoch": 1.5206368835987014,
"grad_norm": 0.04242485761642456,
"learning_rate": 3.3124362239060627e-06,
"loss": 0.0051,
"step": 2460
},
{
"epoch": 1.5268202195084248,
"grad_norm": 0.035334572196006775,
"learning_rate": 3.2326169198472555e-06,
"loss": 0.005,
"step": 2470
},
{
"epoch": 1.5330035554181483,
"grad_norm": 0.03957268223166466,
"learning_rate": 3.1535852662884138e-06,
"loss": 0.0048,
"step": 2480
},
{
"epoch": 1.5391868913278715,
"grad_norm": 0.03457095846533775,
"learning_rate": 3.0753504616627694e-06,
"loss": 0.0048,
"step": 2490
},
{
"epoch": 1.5453702272375947,
"grad_norm": 0.03342267498373985,
"learning_rate": 2.9979216116589203e-06,
"loss": 0.0048,
"step": 2500
},
{
"epoch": 1.5515535631473178,
"grad_norm": 0.0400426909327507,
"learning_rate": 2.9213077281610336e-06,
"loss": 0.0046,
"step": 2510
},
{
"epoch": 1.5577368990570413,
"grad_norm": 0.034529346972703934,
"learning_rate": 2.8455177281999515e-06,
"loss": 0.0048,
"step": 2520
},
{
"epoch": 1.5639202349667647,
"grad_norm": 0.042129479348659515,
"learning_rate": 2.7705604329153434e-06,
"loss": 0.0048,
"step": 2530
},
{
"epoch": 1.5701035708764879,
"grad_norm": 0.03391699492931366,
"learning_rate": 2.6964445665290274e-06,
"loss": 0.0048,
"step": 2540
},
{
"epoch": 1.576286906786211,
"grad_norm": 0.03239929676055908,
"learning_rate": 2.6231787553295583e-06,
"loss": 0.005,
"step": 2550
},
{
"epoch": 1.5824702426959345,
"grad_norm": 0.03297235071659088,
"learning_rate": 2.550771526668221e-06,
"loss": 0.0047,
"step": 2560
},
{
"epoch": 1.5886535786056577,
"grad_norm": 0.03652636706829071,
"learning_rate": 2.47923130796655e-06,
"loss": 0.0048,
"step": 2570
},
{
"epoch": 1.5948369145153811,
"grad_norm": 0.03381171077489853,
"learning_rate": 2.408566425735446e-06,
"loss": 0.0047,
"step": 2580
},
{
"epoch": 1.6010202504251043,
"grad_norm": 0.03702075779438019,
"learning_rate": 2.338785104606082e-06,
"loss": 0.0046,
"step": 2590
},
{
"epoch": 1.6072035863348275,
"grad_norm": 0.03340402618050575,
"learning_rate": 2.26989546637263e-06,
"loss": 0.0047,
"step": 2600
},
{
"epoch": 1.613386922244551,
"grad_norm": 0.030743056908249855,
"learning_rate": 2.2019055290469816e-06,
"loss": 0.0048,
"step": 2610
},
{
"epoch": 1.6195702581542744,
"grad_norm": 0.03447384014725685,
"learning_rate": 2.1348232059255293e-06,
"loss": 0.0049,
"step": 2620
},
{
"epoch": 1.6257535940639976,
"grad_norm": 0.03731811046600342,
"learning_rate": 2.0686563046681507e-06,
"loss": 0.0047,
"step": 2630
},
{
"epoch": 1.6319369299737208,
"grad_norm": 0.03979882597923279,
"learning_rate": 2.0034125263894777e-06,
"loss": 0.0047,
"step": 2640
},
{
"epoch": 1.6381202658834442,
"grad_norm": 0.030915752053260803,
"learning_rate": 1.9390994647625693e-06,
"loss": 0.0043,
"step": 2650
},
{
"epoch": 1.6443036017931674,
"grad_norm": 0.03616530820727348,
"learning_rate": 1.8757246051350908e-06,
"loss": 0.0047,
"step": 2660
},
{
"epoch": 1.6504869377028908,
"grad_norm": 0.03190992400050163,
"learning_rate": 1.813295323658103e-06,
"loss": 0.0045,
"step": 2670
},
{
"epoch": 1.656670273612614,
"grad_norm": 0.032364994287490845,
"learning_rate": 1.7518188864275498e-06,
"loss": 0.0045,
"step": 2680
},
{
"epoch": 1.6628536095223372,
"grad_norm": 0.03670560568571091,
"learning_rate": 1.6913024486385733e-06,
"loss": 0.0047,
"step": 2690
},
{
"epoch": 1.6690369454320606,
"grad_norm": 0.0347522608935833,
"learning_rate": 1.6317530537527148e-06,
"loss": 0.0047,
"step": 2700
},
{
"epoch": 1.675220281341784,
"grad_norm": 0.03402375429868698,
"learning_rate": 1.5731776326781435e-06,
"loss": 0.0047,
"step": 2710
},
{
"epoch": 1.6814036172515072,
"grad_norm": 0.034099772572517395,
"learning_rate": 1.5155830029629614e-06,
"loss": 0.0045,
"step": 2720
},
{
"epoch": 1.6875869531612304,
"grad_norm": 0.035209812223911285,
"learning_rate": 1.4589758680017263e-06,
"loss": 0.0047,
"step": 2730
},
{
"epoch": 1.6937702890709536,
"grad_norm": 0.03288722783327103,
"learning_rate": 1.403362816255236e-06,
"loss": 0.0044,
"step": 2740
},
{
"epoch": 1.699953624980677,
"grad_norm": 0.031503114849328995,
"learning_rate": 1.3487503204837182e-06,
"loss": 0.005,
"step": 2750
},
{
"epoch": 1.7061369608904005,
"grad_norm": 0.03957793861627579,
"learning_rate": 1.2951447369934544e-06,
"loss": 0.0046,
"step": 2760
},
{
"epoch": 1.7123202968001237,
"grad_norm": 0.03468756750226021,
"learning_rate": 1.2425523048969835e-06,
"loss": 0.0048,
"step": 2770
},
{
"epoch": 1.7185036327098469,
"grad_norm": 0.03237258270382881,
"learning_rate": 1.190979145386939e-06,
"loss": 0.0046,
"step": 2780
},
{
"epoch": 1.7246869686195703,
"grad_norm": 0.032560866326093674,
"learning_rate": 1.1404312610235968e-06,
"loss": 0.0045,
"step": 2790
},
{
"epoch": 1.7308703045292937,
"grad_norm": 0.03548423945903778,
"learning_rate": 1.0909145350362516e-06,
"loss": 0.0046,
"step": 2800
},
{
"epoch": 1.7370536404390169,
"grad_norm": 0.04064382612705231,
"learning_rate": 1.042434730638473e-06,
"loss": 0.0046,
"step": 2810
},
{
"epoch": 1.74323697634874,
"grad_norm": 0.03297872841358185,
"learning_rate": 9.949974903573223e-07,
"loss": 0.0044,
"step": 2820
},
{
"epoch": 1.7494203122584633,
"grad_norm": 0.03142615780234337,
"learning_rate": 9.486083353766262e-07,
"loss": 0.0045,
"step": 2830
},
{
"epoch": 1.7556036481681867,
"grad_norm": 0.032335978001356125,
"learning_rate": 9.032726648943779e-07,
"loss": 0.0047,
"step": 2840
},
{
"epoch": 1.7617869840779101,
"grad_norm": 0.032847121357917786,
"learning_rate": 8.58995755494314e-07,
"loss": 0.0043,
"step": 2850
},
{
"epoch": 1.7679703199876333,
"grad_norm": 0.03723368048667908,
"learning_rate": 8.157827605317892e-07,
"loss": 0.0045,
"step": 2860
},
{
"epoch": 1.7741536558973565,
"grad_norm": 0.03200025483965874,
"learning_rate": 7.736387095339748e-07,
"loss": 0.0046,
"step": 2870
},
{
"epoch": 1.78033699180708,
"grad_norm": 0.03175206854939461,
"learning_rate": 7.325685076144795e-07,
"loss": 0.0044,
"step": 2880
},
{
"epoch": 1.7865203277168034,
"grad_norm": 0.03391426429152489,
"learning_rate": 6.925769349024459e-07,
"loss": 0.0046,
"step": 2890
},
{
"epoch": 1.7927036636265266,
"grad_norm": 0.033382490277290344,
"learning_rate": 6.536686459861918e-07,
"loss": 0.0044,
"step": 2900
},
{
"epoch": 1.7988869995362498,
"grad_norm": 0.03503987565636635,
"learning_rate": 6.158481693714735e-07,
"loss": 0.0045,
"step": 2910
},
{
"epoch": 1.805070335445973,
"grad_norm": 0.03369370475411415,
"learning_rate": 5.791199069544074e-07,
"loss": 0.0043,
"step": 2920
},
{
"epoch": 1.8112536713556964,
"grad_norm": 0.03411993384361267,
"learning_rate": 5.434881335091458e-07,
"loss": 0.0047,
"step": 2930
},
{
"epoch": 1.8174370072654198,
"grad_norm": 0.031462300568819046,
"learning_rate": 5.089569961903318e-07,
"loss": 0.0048,
"step": 2940
},
{
"epoch": 1.823620343175143,
"grad_norm": 0.03133223205804825,
"learning_rate": 4.755305140504185e-07,
"loss": 0.0044,
"step": 2950
},
{
"epoch": 1.8298036790848662,
"grad_norm": 0.03286576643586159,
"learning_rate": 4.432125775718965e-07,
"loss": 0.0044,
"step": 2960
},
{
"epoch": 1.8359870149945896,
"grad_norm": 0.0382905974984169,
"learning_rate": 4.120069482144773e-07,
"loss": 0.0045,
"step": 2970
},
{
"epoch": 1.842170350904313,
"grad_norm": 0.0318513922393322,
"learning_rate": 3.8191725797730897e-07,
"loss": 0.0045,
"step": 2980
},
{
"epoch": 1.8483536868140362,
"grad_norm": 0.03260161355137825,
"learning_rate": 3.529470089762421e-07,
"loss": 0.0045,
"step": 2990
},
{
"epoch": 1.8545370227237594,
"grad_norm": 0.03626921400427818,
"learning_rate": 3.2509957303622496e-07,
"loss": 0.0045,
"step": 3000
},
{
"epoch": 1.8607203586334826,
"grad_norm": 0.033349767327308655,
"learning_rate": 2.983781912988604e-07,
"loss": 0.0044,
"step": 3010
},
{
"epoch": 1.866903694543206,
"grad_norm": 0.03295992314815521,
"learning_rate": 2.727859738451721e-07,
"loss": 0.0044,
"step": 3020
},
{
"epoch": 1.8730870304529295,
"grad_norm": 0.03412463888525963,
"learning_rate": 2.4832589933361907e-07,
"loss": 0.0042,
"step": 3030
},
{
"epoch": 1.8792703663626527,
"grad_norm": 0.03095577284693718,
"learning_rate": 2.2500081465341993e-07,
"loss": 0.0043,
"step": 3040
},
{
"epoch": 1.8854537022723759,
"grad_norm": 0.03507233411073685,
"learning_rate": 2.0281343459319868e-07,
"loss": 0.0044,
"step": 3050
},
{
"epoch": 1.8916370381820993,
"grad_norm": 0.03278861194849014,
"learning_rate": 1.8176634152501816e-07,
"loss": 0.0043,
"step": 3060
},
{
"epoch": 1.8978203740918227,
"grad_norm": 0.03288375213742256,
"learning_rate": 1.618619851038161e-07,
"loss": 0.0045,
"step": 3070
},
{
"epoch": 1.904003710001546,
"grad_norm": 0.03299954906105995,
"learning_rate": 1.4310268198229115e-07,
"loss": 0.0044,
"step": 3080
},
{
"epoch": 1.910187045911269,
"grad_norm": 0.03464515134692192,
"learning_rate": 1.2549061554127494e-07,
"loss": 0.0043,
"step": 3090
},
{
"epoch": 1.9163703818209923,
"grad_norm": 0.03539302200078964,
"learning_rate": 1.0902783563560226e-07,
"loss": 0.0046,
"step": 3100
},
{
"epoch": 1.9225537177307157,
"grad_norm": 0.03203083202242851,
"learning_rate": 9.371625835553622e-08,
"loss": 0.0045,
"step": 3110
},
{
"epoch": 1.9287370536404391,
"grad_norm": 0.03126191347837448,
"learning_rate": 7.955766580375334e-08,
"loss": 0.0042,
"step": 3120
},
{
"epoch": 1.9349203895501623,
"grad_norm": 0.031476061791181564,
"learning_rate": 6.655370588792732e-08,
"loss": 0.0045,
"step": 3130
},
{
"epoch": 1.9411037254598855,
"grad_norm": 0.03566085919737816,
"learning_rate": 5.470589212892785e-08,
"loss": 0.0044,
"step": 3140
},
{
"epoch": 1.947287061369609,
"grad_norm": 0.030942443758249283,
"learning_rate": 4.401560348466505e-08,
"loss": 0.0043,
"step": 3150
},
{
"epoch": 1.9534703972793324,
"grad_norm": 0.03156135976314545,
"learning_rate": 3.448408418959326e-08,
"loss": 0.0043,
"step": 3160
},
{
"epoch": 1.9596537331890556,
"grad_norm": 0.03293183073401451,
"learning_rate": 2.6112443609895845e-08,
"loss": 0.0044,
"step": 3170
},
{
"epoch": 1.9658370690987788,
"grad_norm": 0.03247111290693283,
"learning_rate": 1.890165611436734e-08,
"loss": 0.0045,
"step": 3180
},
{
"epoch": 1.972020405008502,
"grad_norm": 0.03142822906374931,
"learning_rate": 1.285256096100529e-08,
"loss": 0.0044,
"step": 3190
},
{
"epoch": 1.9782037409182254,
"grad_norm": 0.034565262496471405,
"learning_rate": 7.965862199330598e-09,
"loss": 0.0045,
"step": 3200
}
],
"logging_steps": 10,
"max_steps": 3236,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 7.045427946497612e+19,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}