Kartoffel-1B-v0.3 / trainer_state.json
SebastianBodza's picture
Add files using upload-large-folder tool
2d320be verified
raw
history blame
151 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.6510342497163959,
"eval_steps": 1000,
"global_step": 8991,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0,
"eval_loss": 7.394782543182373,
"eval_runtime": 2.5063,
"eval_samples_per_second": 4.788,
"eval_steps_per_second": 1.197,
"step": 0
},
{
"epoch": 0.000724095484057831,
"grad_norm": 6.9375,
"learning_rate": 1.2048192771084338e-06,
"loss": 7.2474,
"step": 10
},
{
"epoch": 0.001448190968115662,
"grad_norm": 4.28125,
"learning_rate": 2.4096385542168676e-06,
"loss": 7.2559,
"step": 20
},
{
"epoch": 0.0021722864521734934,
"grad_norm": 2.890625,
"learning_rate": 3.614457831325301e-06,
"loss": 7.2363,
"step": 30
},
{
"epoch": 0.002896381936231324,
"grad_norm": 1.9921875,
"learning_rate": 4.819277108433735e-06,
"loss": 7.2322,
"step": 40
},
{
"epoch": 0.0036204774202891555,
"grad_norm": 1.8125,
"learning_rate": 6.024096385542169e-06,
"loss": 7.1863,
"step": 50
},
{
"epoch": 0.004344572904346987,
"grad_norm": 1.5625,
"learning_rate": 7.228915662650602e-06,
"loss": 7.1754,
"step": 60
},
{
"epoch": 0.005068668388404818,
"grad_norm": 1.7265625,
"learning_rate": 8.433734939759036e-06,
"loss": 7.1641,
"step": 70
},
{
"epoch": 0.005792763872462648,
"grad_norm": 1.6328125,
"learning_rate": 9.63855421686747e-06,
"loss": 7.1237,
"step": 80
},
{
"epoch": 0.00651685935652048,
"grad_norm": 1.6640625,
"learning_rate": 1.0843373493975904e-05,
"loss": 7.0891,
"step": 90
},
{
"epoch": 0.007240954840578311,
"grad_norm": 1.5625,
"learning_rate": 1.2048192771084338e-05,
"loss": 7.0901,
"step": 100
},
{
"epoch": 0.007965050324636142,
"grad_norm": 1.4453125,
"learning_rate": 1.3253012048192772e-05,
"loss": 7.0337,
"step": 110
},
{
"epoch": 0.008689145808693973,
"grad_norm": 1.953125,
"learning_rate": 1.4457831325301205e-05,
"loss": 7.0387,
"step": 120
},
{
"epoch": 0.009413241292751805,
"grad_norm": 1.4375,
"learning_rate": 1.566265060240964e-05,
"loss": 7.018,
"step": 130
},
{
"epoch": 0.010137336776809636,
"grad_norm": 1.3515625,
"learning_rate": 1.6867469879518073e-05,
"loss": 6.9834,
"step": 140
},
{
"epoch": 0.010861432260867467,
"grad_norm": 1.4609375,
"learning_rate": 1.8072289156626505e-05,
"loss": 6.9829,
"step": 150
},
{
"epoch": 0.011585527744925297,
"grad_norm": 1.3515625,
"learning_rate": 1.927710843373494e-05,
"loss": 6.9724,
"step": 160
},
{
"epoch": 0.012309623228983128,
"grad_norm": 1.390625,
"learning_rate": 2.0481927710843373e-05,
"loss": 6.9671,
"step": 170
},
{
"epoch": 0.01303371871304096,
"grad_norm": 1.4921875,
"learning_rate": 2.168674698795181e-05,
"loss": 6.9373,
"step": 180
},
{
"epoch": 0.01375781419709879,
"grad_norm": 1.71875,
"learning_rate": 2.289156626506024e-05,
"loss": 6.9478,
"step": 190
},
{
"epoch": 0.014481909681156622,
"grad_norm": 1.40625,
"learning_rate": 2.4096385542168677e-05,
"loss": 6.9258,
"step": 200
},
{
"epoch": 0.015206005165214453,
"grad_norm": 1.5625,
"learning_rate": 2.530120481927711e-05,
"loss": 6.9085,
"step": 210
},
{
"epoch": 0.015930100649272284,
"grad_norm": 1.453125,
"learning_rate": 2.6506024096385545e-05,
"loss": 6.9023,
"step": 220
},
{
"epoch": 0.016654196133330114,
"grad_norm": 2.125,
"learning_rate": 2.7710843373493977e-05,
"loss": 6.9096,
"step": 230
},
{
"epoch": 0.017378291617387947,
"grad_norm": 1.453125,
"learning_rate": 2.891566265060241e-05,
"loss": 6.8828,
"step": 240
},
{
"epoch": 0.018102387101445776,
"grad_norm": 1.4609375,
"learning_rate": 3.012048192771085e-05,
"loss": 6.8763,
"step": 250
},
{
"epoch": 0.01882648258550361,
"grad_norm": 1.4140625,
"learning_rate": 3.132530120481928e-05,
"loss": 6.875,
"step": 260
},
{
"epoch": 0.01955057806956144,
"grad_norm": 1.6484375,
"learning_rate": 3.253012048192771e-05,
"loss": 6.8641,
"step": 270
},
{
"epoch": 0.020274673553619272,
"grad_norm": 1.4453125,
"learning_rate": 3.3734939759036146e-05,
"loss": 6.8531,
"step": 280
},
{
"epoch": 0.0209987690376771,
"grad_norm": 1.59375,
"learning_rate": 3.4939759036144585e-05,
"loss": 6.8722,
"step": 290
},
{
"epoch": 0.021722864521734934,
"grad_norm": 1.5078125,
"learning_rate": 3.614457831325301e-05,
"loss": 6.8478,
"step": 300
},
{
"epoch": 0.022446960005792764,
"grad_norm": 1.5859375,
"learning_rate": 3.734939759036144e-05,
"loss": 6.8374,
"step": 310
},
{
"epoch": 0.023171055489850594,
"grad_norm": 1.4609375,
"learning_rate": 3.855421686746988e-05,
"loss": 6.8136,
"step": 320
},
{
"epoch": 0.023895150973908427,
"grad_norm": 1.453125,
"learning_rate": 3.9759036144578314e-05,
"loss": 6.833,
"step": 330
},
{
"epoch": 0.024619246457966256,
"grad_norm": 1.4765625,
"learning_rate": 4.0963855421686746e-05,
"loss": 6.8183,
"step": 340
},
{
"epoch": 0.02534334194202409,
"grad_norm": 1.53125,
"learning_rate": 4.2168674698795186e-05,
"loss": 6.8071,
"step": 350
},
{
"epoch": 0.02606743742608192,
"grad_norm": 1.921875,
"learning_rate": 4.337349397590362e-05,
"loss": 6.8177,
"step": 360
},
{
"epoch": 0.02679153291013975,
"grad_norm": 1.484375,
"learning_rate": 4.457831325301205e-05,
"loss": 6.797,
"step": 370
},
{
"epoch": 0.02751562839419758,
"grad_norm": 1.515625,
"learning_rate": 4.578313253012048e-05,
"loss": 6.7979,
"step": 380
},
{
"epoch": 0.028239723878255414,
"grad_norm": 1.6015625,
"learning_rate": 4.698795180722892e-05,
"loss": 6.7899,
"step": 390
},
{
"epoch": 0.028963819362313244,
"grad_norm": 1.546875,
"learning_rate": 4.8192771084337354e-05,
"loss": 6.7614,
"step": 400
},
{
"epoch": 0.029687914846371073,
"grad_norm": 1.5234375,
"learning_rate": 4.9397590361445786e-05,
"loss": 6.8134,
"step": 410
},
{
"epoch": 0.030412010330428906,
"grad_norm": 1.5078125,
"learning_rate": 4.999998281045896e-05,
"loss": 6.7845,
"step": 420
},
{
"epoch": 0.031136105814486736,
"grad_norm": 1.765625,
"learning_rate": 4.99998452942724e-05,
"loss": 6.7727,
"step": 430
},
{
"epoch": 0.03186020129854457,
"grad_norm": 1.484375,
"learning_rate": 4.9999570262655714e-05,
"loss": 6.7537,
"step": 440
},
{
"epoch": 0.0325842967826024,
"grad_norm": 1.53125,
"learning_rate": 4.9999157717121745e-05,
"loss": 6.7486,
"step": 450
},
{
"epoch": 0.03330839226666023,
"grad_norm": 1.359375,
"learning_rate": 4.9998607659939787e-05,
"loss": 6.7585,
"step": 460
},
{
"epoch": 0.034032487750718064,
"grad_norm": 1.4375,
"learning_rate": 4.999792009413549e-05,
"loss": 6.7467,
"step": 470
},
{
"epoch": 0.034756583234775894,
"grad_norm": 1.421875,
"learning_rate": 4.999709502349091e-05,
"loss": 6.7446,
"step": 480
},
{
"epoch": 0.03548067871883372,
"grad_norm": 1.453125,
"learning_rate": 4.9996132452544484e-05,
"loss": 6.7382,
"step": 490
},
{
"epoch": 0.03620477420289155,
"grad_norm": 1.4296875,
"learning_rate": 4.9995032386590986e-05,
"loss": 6.7563,
"step": 500
},
{
"epoch": 0.03692886968694938,
"grad_norm": 1.390625,
"learning_rate": 4.999379483168148e-05,
"loss": 6.7614,
"step": 510
},
{
"epoch": 0.03765296517100722,
"grad_norm": 1.46875,
"learning_rate": 4.999241979462333e-05,
"loss": 6.7422,
"step": 520
},
{
"epoch": 0.03837706065506505,
"grad_norm": 1.4296875,
"learning_rate": 4.999090728298014e-05,
"loss": 6.7343,
"step": 530
},
{
"epoch": 0.03910115613912288,
"grad_norm": 1.390625,
"learning_rate": 4.9989257305071704e-05,
"loss": 6.7052,
"step": 540
},
{
"epoch": 0.03982525162318071,
"grad_norm": 1.4609375,
"learning_rate": 4.998746986997397e-05,
"loss": 6.7377,
"step": 550
},
{
"epoch": 0.040549347107238544,
"grad_norm": 1.640625,
"learning_rate": 4.9985544987519e-05,
"loss": 6.7441,
"step": 560
},
{
"epoch": 0.04127344259129637,
"grad_norm": 1.296875,
"learning_rate": 4.9983482668294905e-05,
"loss": 6.7001,
"step": 570
},
{
"epoch": 0.0419975380753542,
"grad_norm": 1.4375,
"learning_rate": 4.998128292364578e-05,
"loss": 6.7151,
"step": 580
},
{
"epoch": 0.04272163355941203,
"grad_norm": 1.453125,
"learning_rate": 4.997894576567166e-05,
"loss": 6.7119,
"step": 590
},
{
"epoch": 0.04344572904346987,
"grad_norm": 1.4609375,
"learning_rate": 4.9976471207228426e-05,
"loss": 6.7282,
"step": 600
},
{
"epoch": 0.0441698245275277,
"grad_norm": 1.28125,
"learning_rate": 4.9973859261927755e-05,
"loss": 6.726,
"step": 610
},
{
"epoch": 0.04489392001158553,
"grad_norm": 1.515625,
"learning_rate": 4.997110994413707e-05,
"loss": 6.7266,
"step": 620
},
{
"epoch": 0.04561801549564336,
"grad_norm": 1.34375,
"learning_rate": 4.996822326897939e-05,
"loss": 6.7016,
"step": 630
},
{
"epoch": 0.04634211097970119,
"grad_norm": 1.4453125,
"learning_rate": 4.9965199252333315e-05,
"loss": 6.6972,
"step": 640
},
{
"epoch": 0.047066206463759024,
"grad_norm": 1.3515625,
"learning_rate": 4.996203791083291e-05,
"loss": 6.6964,
"step": 650
},
{
"epoch": 0.04779030194781685,
"grad_norm": 1.390625,
"learning_rate": 4.99587392618676e-05,
"loss": 6.6926,
"step": 660
},
{
"epoch": 0.04851439743187468,
"grad_norm": 1.34375,
"learning_rate": 4.995530332358211e-05,
"loss": 6.7038,
"step": 670
},
{
"epoch": 0.04923849291593251,
"grad_norm": 1.375,
"learning_rate": 4.995173011487635e-05,
"loss": 6.7012,
"step": 680
},
{
"epoch": 0.04996258839999035,
"grad_norm": 1.375,
"learning_rate": 4.994801965540527e-05,
"loss": 6.7019,
"step": 690
},
{
"epoch": 0.05068668388404818,
"grad_norm": 1.4921875,
"learning_rate": 4.9944171965578836e-05,
"loss": 6.6656,
"step": 700
},
{
"epoch": 0.05141077936810601,
"grad_norm": 1.390625,
"learning_rate": 4.9940187066561825e-05,
"loss": 6.6811,
"step": 710
},
{
"epoch": 0.05213487485216384,
"grad_norm": 1.3515625,
"learning_rate": 4.99360649802738e-05,
"loss": 6.6885,
"step": 720
},
{
"epoch": 0.05285897033622167,
"grad_norm": 1.40625,
"learning_rate": 4.9931805729388916e-05,
"loss": 6.69,
"step": 730
},
{
"epoch": 0.0535830658202795,
"grad_norm": 1.3203125,
"learning_rate": 4.992740933733581e-05,
"loss": 6.675,
"step": 740
},
{
"epoch": 0.05430716130433733,
"grad_norm": 1.3671875,
"learning_rate": 4.992287582829751e-05,
"loss": 6.6997,
"step": 750
},
{
"epoch": 0.05503125678839516,
"grad_norm": 1.390625,
"learning_rate": 4.991820522721128e-05,
"loss": 6.6923,
"step": 760
},
{
"epoch": 0.05575535227245299,
"grad_norm": 1.40625,
"learning_rate": 4.991339755976845e-05,
"loss": 6.6936,
"step": 770
},
{
"epoch": 0.05647944775651083,
"grad_norm": 1.515625,
"learning_rate": 4.990845285241434e-05,
"loss": 6.6684,
"step": 780
},
{
"epoch": 0.05720354324056866,
"grad_norm": 1.328125,
"learning_rate": 4.9903371132348046e-05,
"loss": 6.6656,
"step": 790
},
{
"epoch": 0.05792763872462649,
"grad_norm": 1.34375,
"learning_rate": 4.9898152427522334e-05,
"loss": 6.6807,
"step": 800
},
{
"epoch": 0.05865173420868432,
"grad_norm": 1.25,
"learning_rate": 4.9892796766643503e-05,
"loss": 6.6901,
"step": 810
},
{
"epoch": 0.059375829692742146,
"grad_norm": 1.4140625,
"learning_rate": 4.988730417917115e-05,
"loss": 6.6851,
"step": 820
},
{
"epoch": 0.06009992517679998,
"grad_norm": 1.5859375,
"learning_rate": 4.98816746953181e-05,
"loss": 6.6608,
"step": 830
},
{
"epoch": 0.06082402066085781,
"grad_norm": 1.3359375,
"learning_rate": 4.987590834605017e-05,
"loss": 6.6617,
"step": 840
},
{
"epoch": 0.06154811614491564,
"grad_norm": 1.28125,
"learning_rate": 4.9870005163086045e-05,
"loss": 6.6574,
"step": 850
},
{
"epoch": 0.06227221162897347,
"grad_norm": 1.3359375,
"learning_rate": 4.986396517889706e-05,
"loss": 6.6633,
"step": 860
},
{
"epoch": 0.06299630711303131,
"grad_norm": 1.3203125,
"learning_rate": 4.985778842670707e-05,
"loss": 6.6807,
"step": 870
},
{
"epoch": 0.06372040259708914,
"grad_norm": 1.3515625,
"learning_rate": 4.9851474940492256e-05,
"loss": 6.6425,
"step": 880
},
{
"epoch": 0.06444449808114697,
"grad_norm": 1.3984375,
"learning_rate": 4.9845024754980876e-05,
"loss": 6.658,
"step": 890
},
{
"epoch": 0.0651685935652048,
"grad_norm": 1.4453125,
"learning_rate": 4.983843790565316e-05,
"loss": 6.6633,
"step": 900
},
{
"epoch": 0.06589268904926263,
"grad_norm": 1.3671875,
"learning_rate": 4.983171442874106e-05,
"loss": 6.6828,
"step": 910
},
{
"epoch": 0.06661678453332046,
"grad_norm": 1.3515625,
"learning_rate": 4.982485436122809e-05,
"loss": 6.6251,
"step": 920
},
{
"epoch": 0.06734088001737829,
"grad_norm": 1.2578125,
"learning_rate": 4.981785774084909e-05,
"loss": 6.6573,
"step": 930
},
{
"epoch": 0.06806497550143613,
"grad_norm": 1.3828125,
"learning_rate": 4.981072460609002e-05,
"loss": 6.6677,
"step": 940
},
{
"epoch": 0.06878907098549396,
"grad_norm": 1.3125,
"learning_rate": 4.980345499618778e-05,
"loss": 6.6577,
"step": 950
},
{
"epoch": 0.06951316646955179,
"grad_norm": 1.2890625,
"learning_rate": 4.9796048951129936e-05,
"loss": 6.6628,
"step": 960
},
{
"epoch": 0.07023726195360962,
"grad_norm": 1.3984375,
"learning_rate": 4.978850651165458e-05,
"loss": 6.6443,
"step": 970
},
{
"epoch": 0.07096135743766745,
"grad_norm": 1.421875,
"learning_rate": 4.9780827719250035e-05,
"loss": 6.6617,
"step": 980
},
{
"epoch": 0.07168545292172528,
"grad_norm": 1.3515625,
"learning_rate": 4.977301261615465e-05,
"loss": 6.6679,
"step": 990
},
{
"epoch": 0.0724095484057831,
"grad_norm": 1.34375,
"learning_rate": 4.9765061245356594e-05,
"loss": 6.6385,
"step": 1000
},
{
"epoch": 0.0724095484057831,
"eval_loss": 6.789700984954834,
"eval_runtime": 1.6435,
"eval_samples_per_second": 7.302,
"eval_steps_per_second": 1.825,
"step": 1000
},
{
"epoch": 0.07313364388984094,
"grad_norm": 1.296875,
"learning_rate": 4.9756973650593583e-05,
"loss": 6.6351,
"step": 1010
},
{
"epoch": 0.07385773937389876,
"grad_norm": 1.3828125,
"learning_rate": 4.9748749876352655e-05,
"loss": 6.6602,
"step": 1020
},
{
"epoch": 0.07458183485795661,
"grad_norm": 1.3203125,
"learning_rate": 4.974038996786991e-05,
"loss": 6.653,
"step": 1030
},
{
"epoch": 0.07530593034201444,
"grad_norm": 1.4453125,
"learning_rate": 4.9731893971130306e-05,
"loss": 6.6532,
"step": 1040
},
{
"epoch": 0.07603002582607227,
"grad_norm": 1.421875,
"learning_rate": 4.972326193286736e-05,
"loss": 6.6297,
"step": 1050
},
{
"epoch": 0.0767541213101301,
"grad_norm": 1.3828125,
"learning_rate": 4.971449390056289e-05,
"loss": 6.6379,
"step": 1060
},
{
"epoch": 0.07747821679418793,
"grad_norm": 1.2890625,
"learning_rate": 4.970558992244679e-05,
"loss": 6.6478,
"step": 1070
},
{
"epoch": 0.07820231227824576,
"grad_norm": 1.4609375,
"learning_rate": 4.969655004749674e-05,
"loss": 6.6616,
"step": 1080
},
{
"epoch": 0.07892640776230359,
"grad_norm": 1.4296875,
"learning_rate": 4.968737432543794e-05,
"loss": 6.6314,
"step": 1090
},
{
"epoch": 0.07965050324636141,
"grad_norm": 1.2265625,
"learning_rate": 4.967806280674283e-05,
"loss": 6.6465,
"step": 1100
},
{
"epoch": 0.08037459873041924,
"grad_norm": 1.5390625,
"learning_rate": 4.9668615542630836e-05,
"loss": 6.6517,
"step": 1110
},
{
"epoch": 0.08109869421447709,
"grad_norm": 1.2734375,
"learning_rate": 4.965903258506806e-05,
"loss": 6.6278,
"step": 1120
},
{
"epoch": 0.08182278969853492,
"grad_norm": 1.3125,
"learning_rate": 4.964931398676701e-05,
"loss": 6.6246,
"step": 1130
},
{
"epoch": 0.08254688518259275,
"grad_norm": 1.296875,
"learning_rate": 4.9639459801186306e-05,
"loss": 6.616,
"step": 1140
},
{
"epoch": 0.08327098066665058,
"grad_norm": 1.3515625,
"learning_rate": 4.962947008253038e-05,
"loss": 6.6492,
"step": 1150
},
{
"epoch": 0.0839950761507084,
"grad_norm": 1.390625,
"learning_rate": 4.96193448857492e-05,
"loss": 6.6096,
"step": 1160
},
{
"epoch": 0.08471917163476624,
"grad_norm": 1.421875,
"learning_rate": 4.960908426653793e-05,
"loss": 6.6282,
"step": 1170
},
{
"epoch": 0.08544326711882406,
"grad_norm": 1.296875,
"learning_rate": 4.959868828133667e-05,
"loss": 6.6053,
"step": 1180
},
{
"epoch": 0.0861673626028819,
"grad_norm": 1.3359375,
"learning_rate": 4.95881569873301e-05,
"loss": 6.6031,
"step": 1190
},
{
"epoch": 0.08689145808693974,
"grad_norm": 1.3984375,
"learning_rate": 4.95774904424472e-05,
"loss": 6.609,
"step": 1200
},
{
"epoch": 0.08761555357099757,
"grad_norm": 1.34375,
"learning_rate": 4.9566688705360915e-05,
"loss": 6.6291,
"step": 1210
},
{
"epoch": 0.0883396490550554,
"grad_norm": 1.296875,
"learning_rate": 4.955575183548783e-05,
"loss": 6.6252,
"step": 1220
},
{
"epoch": 0.08906374453911323,
"grad_norm": 1.296875,
"learning_rate": 4.9544679892987855e-05,
"loss": 6.6246,
"step": 1230
},
{
"epoch": 0.08978784002317106,
"grad_norm": 1.28125,
"learning_rate": 4.9533472938763884e-05,
"loss": 6.6155,
"step": 1240
},
{
"epoch": 0.09051193550722889,
"grad_norm": 1.3203125,
"learning_rate": 4.9522131034461464e-05,
"loss": 6.6263,
"step": 1250
},
{
"epoch": 0.09123603099128672,
"grad_norm": 1.328125,
"learning_rate": 4.951065424246845e-05,
"loss": 6.6219,
"step": 1260
},
{
"epoch": 0.09196012647534454,
"grad_norm": 1.375,
"learning_rate": 4.9499042625914674e-05,
"loss": 6.6144,
"step": 1270
},
{
"epoch": 0.09268422195940237,
"grad_norm": 1.3984375,
"learning_rate": 4.9487296248671586e-05,
"loss": 6.6153,
"step": 1280
},
{
"epoch": 0.09340831744346022,
"grad_norm": 1.2421875,
"learning_rate": 4.9475415175351926e-05,
"loss": 6.6136,
"step": 1290
},
{
"epoch": 0.09413241292751805,
"grad_norm": 1.34375,
"learning_rate": 4.946339947130933e-05,
"loss": 6.586,
"step": 1300
},
{
"epoch": 0.09485650841157588,
"grad_norm": 1.359375,
"learning_rate": 4.945124920263797e-05,
"loss": 6.6185,
"step": 1310
},
{
"epoch": 0.0955806038956337,
"grad_norm": 1.3359375,
"learning_rate": 4.9438964436172264e-05,
"loss": 6.6248,
"step": 1320
},
{
"epoch": 0.09630469937969154,
"grad_norm": 1.2890625,
"learning_rate": 4.9426545239486424e-05,
"loss": 6.605,
"step": 1330
},
{
"epoch": 0.09702879486374937,
"grad_norm": 1.3125,
"learning_rate": 4.94139916808941e-05,
"loss": 6.6525,
"step": 1340
},
{
"epoch": 0.0977528903478072,
"grad_norm": 1.2421875,
"learning_rate": 4.9401303829448075e-05,
"loss": 6.6178,
"step": 1350
},
{
"epoch": 0.09847698583186502,
"grad_norm": 1.28125,
"learning_rate": 4.938848175493976e-05,
"loss": 6.5978,
"step": 1360
},
{
"epoch": 0.09920108131592285,
"grad_norm": 1.296875,
"learning_rate": 4.937552552789894e-05,
"loss": 6.6059,
"step": 1370
},
{
"epoch": 0.0999251767999807,
"grad_norm": 1.3203125,
"learning_rate": 4.936243521959329e-05,
"loss": 6.6078,
"step": 1380
},
{
"epoch": 0.10064927228403853,
"grad_norm": 1.2890625,
"learning_rate": 4.934921090202803e-05,
"loss": 6.6135,
"step": 1390
},
{
"epoch": 0.10137336776809636,
"grad_norm": 1.375,
"learning_rate": 4.933585264794552e-05,
"loss": 6.6068,
"step": 1400
},
{
"epoch": 0.10209746325215419,
"grad_norm": 1.34375,
"learning_rate": 4.932236053082486e-05,
"loss": 6.6026,
"step": 1410
},
{
"epoch": 0.10282155873621202,
"grad_norm": 1.2578125,
"learning_rate": 4.930873462488148e-05,
"loss": 6.6064,
"step": 1420
},
{
"epoch": 0.10354565422026984,
"grad_norm": 1.375,
"learning_rate": 4.9294975005066734e-05,
"loss": 6.6075,
"step": 1430
},
{
"epoch": 0.10426974970432767,
"grad_norm": 1.2421875,
"learning_rate": 4.9281081747067494e-05,
"loss": 6.5982,
"step": 1440
},
{
"epoch": 0.1049938451883855,
"grad_norm": 1.484375,
"learning_rate": 4.926705492730572e-05,
"loss": 6.6109,
"step": 1450
},
{
"epoch": 0.10571794067244333,
"grad_norm": 1.3828125,
"learning_rate": 4.925289462293807e-05,
"loss": 6.596,
"step": 1460
},
{
"epoch": 0.10644203615650118,
"grad_norm": 1.3125,
"learning_rate": 4.923860091185542e-05,
"loss": 6.606,
"step": 1470
},
{
"epoch": 0.107166131640559,
"grad_norm": 1.2578125,
"learning_rate": 4.922417387268251e-05,
"loss": 6.6,
"step": 1480
},
{
"epoch": 0.10789022712461684,
"grad_norm": 1.328125,
"learning_rate": 4.9209613584777445e-05,
"loss": 6.5953,
"step": 1490
},
{
"epoch": 0.10861432260867467,
"grad_norm": 1.2890625,
"learning_rate": 4.919492012823128e-05,
"loss": 6.5907,
"step": 1500
},
{
"epoch": 0.1093384180927325,
"grad_norm": 1.28125,
"learning_rate": 4.918009358386761e-05,
"loss": 6.6144,
"step": 1510
},
{
"epoch": 0.11006251357679032,
"grad_norm": 1.3359375,
"learning_rate": 4.916513403324207e-05,
"loss": 6.5911,
"step": 1520
},
{
"epoch": 0.11078660906084815,
"grad_norm": 1.2578125,
"learning_rate": 4.9150041558641947e-05,
"loss": 6.6071,
"step": 1530
},
{
"epoch": 0.11151070454490598,
"grad_norm": 1.2421875,
"learning_rate": 4.913481624308566e-05,
"loss": 6.6028,
"step": 1540
},
{
"epoch": 0.11223480002896381,
"grad_norm": 1.3203125,
"learning_rate": 4.911945817032238e-05,
"loss": 6.6121,
"step": 1550
},
{
"epoch": 0.11295889551302166,
"grad_norm": 1.4140625,
"learning_rate": 4.91039674248315e-05,
"loss": 6.6003,
"step": 1560
},
{
"epoch": 0.11368299099707949,
"grad_norm": 1.421875,
"learning_rate": 4.908834409182221e-05,
"loss": 6.632,
"step": 1570
},
{
"epoch": 0.11440708648113732,
"grad_norm": 1.2578125,
"learning_rate": 4.907258825723301e-05,
"loss": 6.5953,
"step": 1580
},
{
"epoch": 0.11513118196519515,
"grad_norm": 1.34375,
"learning_rate": 4.905670000773126e-05,
"loss": 6.6028,
"step": 1590
},
{
"epoch": 0.11585527744925297,
"grad_norm": 1.3046875,
"learning_rate": 4.904067943071268e-05,
"loss": 6.592,
"step": 1600
},
{
"epoch": 0.1165793729333108,
"grad_norm": 1.359375,
"learning_rate": 4.902452661430087e-05,
"loss": 6.6015,
"step": 1610
},
{
"epoch": 0.11730346841736863,
"grad_norm": 1.28125,
"learning_rate": 4.9008241647346836e-05,
"loss": 6.5767,
"step": 1620
},
{
"epoch": 0.11802756390142646,
"grad_norm": 1.2734375,
"learning_rate": 4.899182461942851e-05,
"loss": 6.5894,
"step": 1630
},
{
"epoch": 0.11875165938548429,
"grad_norm": 1.28125,
"learning_rate": 4.897527562085024e-05,
"loss": 6.5856,
"step": 1640
},
{
"epoch": 0.11947575486954214,
"grad_norm": 1.3046875,
"learning_rate": 4.895859474264229e-05,
"loss": 6.5761,
"step": 1650
},
{
"epoch": 0.12019985035359997,
"grad_norm": 1.4140625,
"learning_rate": 4.894178207656035e-05,
"loss": 6.5964,
"step": 1660
},
{
"epoch": 0.1209239458376578,
"grad_norm": 1.2109375,
"learning_rate": 4.892483771508504e-05,
"loss": 6.5657,
"step": 1670
},
{
"epoch": 0.12164804132171562,
"grad_norm": 1.265625,
"learning_rate": 4.890776175142138e-05,
"loss": 6.5946,
"step": 1680
},
{
"epoch": 0.12237213680577345,
"grad_norm": 1.359375,
"learning_rate": 4.889055427949829e-05,
"loss": 6.5916,
"step": 1690
},
{
"epoch": 0.12309623228983128,
"grad_norm": 1.2890625,
"learning_rate": 4.8873215393968074e-05,
"loss": 6.594,
"step": 1700
},
{
"epoch": 0.12382032777388911,
"grad_norm": 1.359375,
"learning_rate": 4.885574519020589e-05,
"loss": 6.566,
"step": 1710
},
{
"epoch": 0.12454442325794694,
"grad_norm": 1.3125,
"learning_rate": 4.883814376430924e-05,
"loss": 6.5752,
"step": 1720
},
{
"epoch": 0.1252685187420048,
"grad_norm": 1.3515625,
"learning_rate": 4.882041121309743e-05,
"loss": 6.6015,
"step": 1730
},
{
"epoch": 0.12599261422606262,
"grad_norm": 1.328125,
"learning_rate": 4.880254763411103e-05,
"loss": 6.5655,
"step": 1740
},
{
"epoch": 0.12671670971012045,
"grad_norm": 1.3125,
"learning_rate": 4.878455312561138e-05,
"loss": 6.5883,
"step": 1750
},
{
"epoch": 0.12744080519417827,
"grad_norm": 1.3515625,
"learning_rate": 4.8766427786579964e-05,
"loss": 6.6061,
"step": 1760
},
{
"epoch": 0.1281649006782361,
"grad_norm": 1.296875,
"learning_rate": 4.874817171671798e-05,
"loss": 6.5878,
"step": 1770
},
{
"epoch": 0.12888899616229393,
"grad_norm": 1.2265625,
"learning_rate": 4.872978501644569e-05,
"loss": 6.5837,
"step": 1780
},
{
"epoch": 0.12961309164635176,
"grad_norm": 1.3203125,
"learning_rate": 4.8711267786901914e-05,
"loss": 6.5884,
"step": 1790
},
{
"epoch": 0.1303371871304096,
"grad_norm": 1.1953125,
"learning_rate": 4.869262012994348e-05,
"loss": 6.5864,
"step": 1800
},
{
"epoch": 0.13106128261446742,
"grad_norm": 1.3125,
"learning_rate": 4.867384214814465e-05,
"loss": 6.5655,
"step": 1810
},
{
"epoch": 0.13178537809852525,
"grad_norm": 1.1953125,
"learning_rate": 4.8654933944796546e-05,
"loss": 6.5747,
"step": 1820
},
{
"epoch": 0.13250947358258308,
"grad_norm": 1.2578125,
"learning_rate": 4.86358956239066e-05,
"loss": 6.5689,
"step": 1830
},
{
"epoch": 0.1332335690666409,
"grad_norm": 1.234375,
"learning_rate": 4.861672729019797e-05,
"loss": 6.6059,
"step": 1840
},
{
"epoch": 0.13395766455069874,
"grad_norm": 1.2890625,
"learning_rate": 4.859742904910899e-05,
"loss": 6.5985,
"step": 1850
},
{
"epoch": 0.13468176003475657,
"grad_norm": 1.25,
"learning_rate": 4.8578001006792544e-05,
"loss": 6.6006,
"step": 1860
},
{
"epoch": 0.13540585551881443,
"grad_norm": 1.28125,
"learning_rate": 4.8558443270115524e-05,
"loss": 6.5779,
"step": 1870
},
{
"epoch": 0.13612995100287226,
"grad_norm": 1.296875,
"learning_rate": 4.853875594665821e-05,
"loss": 6.566,
"step": 1880
},
{
"epoch": 0.1368540464869301,
"grad_norm": 1.328125,
"learning_rate": 4.85189391447137e-05,
"loss": 6.6019,
"step": 1890
},
{
"epoch": 0.13757814197098792,
"grad_norm": 1.265625,
"learning_rate": 4.8498992973287324e-05,
"loss": 6.5758,
"step": 1900
},
{
"epoch": 0.13830223745504575,
"grad_norm": 1.296875,
"learning_rate": 4.8478917542096005e-05,
"loss": 6.5754,
"step": 1910
},
{
"epoch": 0.13902633293910358,
"grad_norm": 1.21875,
"learning_rate": 4.845871296156769e-05,
"loss": 6.593,
"step": 1920
},
{
"epoch": 0.1397504284231614,
"grad_norm": 1.296875,
"learning_rate": 4.843837934284072e-05,
"loss": 6.5754,
"step": 1930
},
{
"epoch": 0.14047452390721923,
"grad_norm": 1.421875,
"learning_rate": 4.841791679776326e-05,
"loss": 6.5659,
"step": 1940
},
{
"epoch": 0.14119861939127706,
"grad_norm": 1.359375,
"learning_rate": 4.839732543889263e-05,
"loss": 6.5821,
"step": 1950
},
{
"epoch": 0.1419227148753349,
"grad_norm": 1.34375,
"learning_rate": 4.83766053794947e-05,
"loss": 6.58,
"step": 1960
},
{
"epoch": 0.14264681035939272,
"grad_norm": 1.2734375,
"learning_rate": 4.83557567335433e-05,
"loss": 6.5662,
"step": 1970
},
{
"epoch": 0.14337090584345055,
"grad_norm": 1.2265625,
"learning_rate": 4.833477961571956e-05,
"loss": 6.5675,
"step": 1980
},
{
"epoch": 0.14409500132750838,
"grad_norm": 1.1875,
"learning_rate": 4.831367414141129e-05,
"loss": 6.5784,
"step": 1990
},
{
"epoch": 0.1448190968115662,
"grad_norm": 1.265625,
"learning_rate": 4.829244042671235e-05,
"loss": 6.5707,
"step": 2000
},
{
"epoch": 0.1448190968115662,
"eval_loss": 6.703906536102295,
"eval_runtime": 1.4135,
"eval_samples_per_second": 8.489,
"eval_steps_per_second": 2.122,
"step": 2000
},
{
"epoch": 0.14554319229562404,
"grad_norm": 1.3203125,
"learning_rate": 4.827107858842198e-05,
"loss": 6.577,
"step": 2010
},
{
"epoch": 0.14626728777968187,
"grad_norm": 1.21875,
"learning_rate": 4.824958874404422e-05,
"loss": 6.5519,
"step": 2020
},
{
"epoch": 0.1469913832637397,
"grad_norm": 1.2578125,
"learning_rate": 4.8227971011787196e-05,
"loss": 6.5635,
"step": 2030
},
{
"epoch": 0.14771547874779753,
"grad_norm": 1.2578125,
"learning_rate": 4.820622551056252e-05,
"loss": 6.566,
"step": 2040
},
{
"epoch": 0.1484395742318554,
"grad_norm": 1.265625,
"learning_rate": 4.818435235998461e-05,
"loss": 6.5532,
"step": 2050
},
{
"epoch": 0.14916366971591322,
"grad_norm": 1.25,
"learning_rate": 4.8162351680370044e-05,
"loss": 6.5784,
"step": 2060
},
{
"epoch": 0.14988776519997105,
"grad_norm": 1.265625,
"learning_rate": 4.8140223592736885e-05,
"loss": 6.5798,
"step": 2070
},
{
"epoch": 0.15061186068402888,
"grad_norm": 1.25,
"learning_rate": 4.811796821880402e-05,
"loss": 6.5747,
"step": 2080
},
{
"epoch": 0.1513359561680867,
"grad_norm": 1.28125,
"learning_rate": 4.809558568099051e-05,
"loss": 6.5855,
"step": 2090
},
{
"epoch": 0.15206005165214453,
"grad_norm": 1.2734375,
"learning_rate": 4.807307610241488e-05,
"loss": 6.5657,
"step": 2100
},
{
"epoch": 0.15278414713620236,
"grad_norm": 1.28125,
"learning_rate": 4.805043960689447e-05,
"loss": 6.5535,
"step": 2110
},
{
"epoch": 0.1535082426202602,
"grad_norm": 1.3203125,
"learning_rate": 4.8027676318944756e-05,
"loss": 6.5728,
"step": 2120
},
{
"epoch": 0.15423233810431802,
"grad_norm": 1.3515625,
"learning_rate": 4.800478636377865e-05,
"loss": 6.5564,
"step": 2130
},
{
"epoch": 0.15495643358837585,
"grad_norm": 1.296875,
"learning_rate": 4.798176986730579e-05,
"loss": 6.5745,
"step": 2140
},
{
"epoch": 0.15568052907243368,
"grad_norm": 1.328125,
"learning_rate": 4.795862695613192e-05,
"loss": 6.5697,
"step": 2150
},
{
"epoch": 0.1564046245564915,
"grad_norm": 1.2578125,
"learning_rate": 4.793535775755811e-05,
"loss": 6.5961,
"step": 2160
},
{
"epoch": 0.15712872004054934,
"grad_norm": 1.3359375,
"learning_rate": 4.791196239958011e-05,
"loss": 6.5581,
"step": 2170
},
{
"epoch": 0.15785281552460717,
"grad_norm": 1.203125,
"learning_rate": 4.7888441010887625e-05,
"loss": 6.554,
"step": 2180
},
{
"epoch": 0.158576911008665,
"grad_norm": 1.328125,
"learning_rate": 4.786479372086361e-05,
"loss": 6.5572,
"step": 2190
},
{
"epoch": 0.15930100649272283,
"grad_norm": 1.2109375,
"learning_rate": 4.784102065958356e-05,
"loss": 6.5765,
"step": 2200
},
{
"epoch": 0.16002510197678066,
"grad_norm": 1.3203125,
"learning_rate": 4.781712195781479e-05,
"loss": 6.5703,
"step": 2210
},
{
"epoch": 0.1607491974608385,
"grad_norm": 1.21875,
"learning_rate": 4.779309774701574e-05,
"loss": 6.5712,
"step": 2220
},
{
"epoch": 0.16147329294489635,
"grad_norm": 1.3203125,
"learning_rate": 4.77689481593352e-05,
"loss": 6.5733,
"step": 2230
},
{
"epoch": 0.16219738842895418,
"grad_norm": 1.2734375,
"learning_rate": 4.7744673327611625e-05,
"loss": 6.558,
"step": 2240
},
{
"epoch": 0.162921483913012,
"grad_norm": 1.296875,
"learning_rate": 4.7720273385372416e-05,
"loss": 6.5623,
"step": 2250
},
{
"epoch": 0.16364557939706983,
"grad_norm": 1.2890625,
"learning_rate": 4.769574846683313e-05,
"loss": 6.5813,
"step": 2260
},
{
"epoch": 0.16436967488112766,
"grad_norm": 1.6328125,
"learning_rate": 4.76710987068968e-05,
"loss": 6.5531,
"step": 2270
},
{
"epoch": 0.1650937703651855,
"grad_norm": 1.3359375,
"learning_rate": 4.764632424115315e-05,
"loss": 6.5368,
"step": 2280
},
{
"epoch": 0.16581786584924332,
"grad_norm": 1.296875,
"learning_rate": 4.762142520587789e-05,
"loss": 6.5631,
"step": 2290
},
{
"epoch": 0.16654196133330115,
"grad_norm": 1.28125,
"learning_rate": 4.7596401738031916e-05,
"loss": 6.5436,
"step": 2300
},
{
"epoch": 0.16726605681735898,
"grad_norm": 1.2734375,
"learning_rate": 4.75712539752606e-05,
"loss": 6.5749,
"step": 2310
},
{
"epoch": 0.1679901523014168,
"grad_norm": 1.234375,
"learning_rate": 4.754598205589303e-05,
"loss": 6.5359,
"step": 2320
},
{
"epoch": 0.16871424778547464,
"grad_norm": 1.328125,
"learning_rate": 4.752058611894119e-05,
"loss": 6.5382,
"step": 2330
},
{
"epoch": 0.16943834326953247,
"grad_norm": 1.2890625,
"learning_rate": 4.7495066304099294e-05,
"loss": 6.5609,
"step": 2340
},
{
"epoch": 0.1701624387535903,
"grad_norm": 1.2734375,
"learning_rate": 4.7469422751742944e-05,
"loss": 6.5635,
"step": 2350
},
{
"epoch": 0.17088653423764813,
"grad_norm": 1.25,
"learning_rate": 4.744365560292837e-05,
"loss": 6.5632,
"step": 2360
},
{
"epoch": 0.17161062972170596,
"grad_norm": 1.2734375,
"learning_rate": 4.741776499939168e-05,
"loss": 6.5634,
"step": 2370
},
{
"epoch": 0.1723347252057638,
"grad_norm": 1.2734375,
"learning_rate": 4.739175108354804e-05,
"loss": 6.5638,
"step": 2380
},
{
"epoch": 0.17305882068982162,
"grad_norm": 1.203125,
"learning_rate": 4.7365613998490944e-05,
"loss": 6.5594,
"step": 2390
},
{
"epoch": 0.17378291617387948,
"grad_norm": 1.234375,
"learning_rate": 4.733935388799137e-05,
"loss": 6.5716,
"step": 2400
},
{
"epoch": 0.1745070116579373,
"grad_norm": 1.21875,
"learning_rate": 4.731297089649703e-05,
"loss": 6.5383,
"step": 2410
},
{
"epoch": 0.17523110714199513,
"grad_norm": 1.2734375,
"learning_rate": 4.728646516913157e-05,
"loss": 6.5484,
"step": 2420
},
{
"epoch": 0.17595520262605296,
"grad_norm": 1.3828125,
"learning_rate": 4.725983685169374e-05,
"loss": 6.5473,
"step": 2430
},
{
"epoch": 0.1766792981101108,
"grad_norm": 1.265625,
"learning_rate": 4.723308609065663e-05,
"loss": 6.5431,
"step": 2440
},
{
"epoch": 0.17740339359416862,
"grad_norm": 1.3125,
"learning_rate": 4.720621303316685e-05,
"loss": 6.5569,
"step": 2450
},
{
"epoch": 0.17812748907822645,
"grad_norm": 1.28125,
"learning_rate": 4.717921782704371e-05,
"loss": 6.5659,
"step": 2460
},
{
"epoch": 0.17885158456228428,
"grad_norm": 1.3515625,
"learning_rate": 4.715210062077844e-05,
"loss": 6.5329,
"step": 2470
},
{
"epoch": 0.1795756800463421,
"grad_norm": 1.2109375,
"learning_rate": 4.712486156353331e-05,
"loss": 6.5778,
"step": 2480
},
{
"epoch": 0.18029977553039994,
"grad_norm": 1.1953125,
"learning_rate": 4.7097500805140894e-05,
"loss": 6.5461,
"step": 2490
},
{
"epoch": 0.18102387101445777,
"grad_norm": 1.2265625,
"learning_rate": 4.7070018496103176e-05,
"loss": 6.5448,
"step": 2500
},
{
"epoch": 0.1817479664985156,
"grad_norm": 1.265625,
"learning_rate": 4.704241478759076e-05,
"loss": 6.5525,
"step": 2510
},
{
"epoch": 0.18247206198257343,
"grad_norm": 1.2890625,
"learning_rate": 4.7014689831442004e-05,
"loss": 6.5503,
"step": 2520
},
{
"epoch": 0.18319615746663126,
"grad_norm": 1.265625,
"learning_rate": 4.698684378016222e-05,
"loss": 6.5854,
"step": 2530
},
{
"epoch": 0.1839202529506891,
"grad_norm": 1.2578125,
"learning_rate": 4.695887678692286e-05,
"loss": 6.5733,
"step": 2540
},
{
"epoch": 0.18464434843474692,
"grad_norm": 1.2734375,
"learning_rate": 4.693078900556056e-05,
"loss": 6.5334,
"step": 2550
},
{
"epoch": 0.18536844391880475,
"grad_norm": 1.2890625,
"learning_rate": 4.690258059057643e-05,
"loss": 6.5561,
"step": 2560
},
{
"epoch": 0.18609253940286258,
"grad_norm": 1.296875,
"learning_rate": 4.687425169713512e-05,
"loss": 6.5502,
"step": 2570
},
{
"epoch": 0.18681663488692044,
"grad_norm": 1.265625,
"learning_rate": 4.684580248106399e-05,
"loss": 6.543,
"step": 2580
},
{
"epoch": 0.18754073037097826,
"grad_norm": 1.265625,
"learning_rate": 4.6817233098852255e-05,
"loss": 6.5674,
"step": 2590
},
{
"epoch": 0.1882648258550361,
"grad_norm": 1.296875,
"learning_rate": 4.6788543707650124e-05,
"loss": 6.5527,
"step": 2600
},
{
"epoch": 0.18898892133909392,
"grad_norm": 1.296875,
"learning_rate": 4.675973446526793e-05,
"loss": 6.5686,
"step": 2610
},
{
"epoch": 0.18971301682315175,
"grad_norm": 1.328125,
"learning_rate": 4.673080553017528e-05,
"loss": 6.5365,
"step": 2620
},
{
"epoch": 0.19043711230720958,
"grad_norm": 1.1796875,
"learning_rate": 4.670175706150015e-05,
"loss": 6.5584,
"step": 2630
},
{
"epoch": 0.1911612077912674,
"grad_norm": 1.3125,
"learning_rate": 4.667258921902803e-05,
"loss": 6.5567,
"step": 2640
},
{
"epoch": 0.19188530327532524,
"grad_norm": 1.21875,
"learning_rate": 4.664330216320105e-05,
"loss": 6.5569,
"step": 2650
},
{
"epoch": 0.19260939875938307,
"grad_norm": 1.2890625,
"learning_rate": 4.661389605511709e-05,
"loss": 6.5434,
"step": 2660
},
{
"epoch": 0.1933334942434409,
"grad_norm": 1.3046875,
"learning_rate": 4.65843710565289e-05,
"loss": 6.5305,
"step": 2670
},
{
"epoch": 0.19405758972749873,
"grad_norm": 1.3671875,
"learning_rate": 4.65547273298432e-05,
"loss": 6.542,
"step": 2680
},
{
"epoch": 0.19478168521155656,
"grad_norm": 1.3125,
"learning_rate": 4.6524965038119786e-05,
"loss": 6.5415,
"step": 2690
},
{
"epoch": 0.1955057806956144,
"grad_norm": 1.28125,
"learning_rate": 4.6495084345070656e-05,
"loss": 6.5114,
"step": 2700
},
{
"epoch": 0.19622987617967222,
"grad_norm": 1.21875,
"learning_rate": 4.646508541505907e-05,
"loss": 6.5511,
"step": 2710
},
{
"epoch": 0.19695397166373005,
"grad_norm": 1.2421875,
"learning_rate": 4.643496841309868e-05,
"loss": 6.5372,
"step": 2720
},
{
"epoch": 0.19767806714778788,
"grad_norm": 1.296875,
"learning_rate": 4.640473350485261e-05,
"loss": 6.5409,
"step": 2730
},
{
"epoch": 0.1984021626318457,
"grad_norm": 1.1953125,
"learning_rate": 4.637438085663257e-05,
"loss": 6.5408,
"step": 2740
},
{
"epoch": 0.19912625811590354,
"grad_norm": 1.3203125,
"learning_rate": 4.634391063539786e-05,
"loss": 6.5488,
"step": 2750
},
{
"epoch": 0.1998503535999614,
"grad_norm": 1.21875,
"learning_rate": 4.631332300875455e-05,
"loss": 6.5651,
"step": 2760
},
{
"epoch": 0.20057444908401922,
"grad_norm": 1.265625,
"learning_rate": 4.6282618144954516e-05,
"loss": 6.536,
"step": 2770
},
{
"epoch": 0.20129854456807705,
"grad_norm": 1.3828125,
"learning_rate": 4.6251796212894486e-05,
"loss": 6.5438,
"step": 2780
},
{
"epoch": 0.20202264005213488,
"grad_norm": 1.203125,
"learning_rate": 4.622085738211518e-05,
"loss": 6.5701,
"step": 2790
},
{
"epoch": 0.2027467355361927,
"grad_norm": 1.4140625,
"learning_rate": 4.618980182280031e-05,
"loss": 6.5484,
"step": 2800
},
{
"epoch": 0.20347083102025054,
"grad_norm": 1.453125,
"learning_rate": 4.615862970577567e-05,
"loss": 6.5297,
"step": 2810
},
{
"epoch": 0.20419492650430837,
"grad_norm": 1.2890625,
"learning_rate": 4.612734120250821e-05,
"loss": 6.5334,
"step": 2820
},
{
"epoch": 0.2049190219883662,
"grad_norm": 1.2734375,
"learning_rate": 4.609593648510507e-05,
"loss": 6.5359,
"step": 2830
},
{
"epoch": 0.20564311747242403,
"grad_norm": 1.296875,
"learning_rate": 4.606441572631266e-05,
"loss": 6.534,
"step": 2840
},
{
"epoch": 0.20636721295648186,
"grad_norm": 1.2890625,
"learning_rate": 4.603277909951566e-05,
"loss": 6.5344,
"step": 2850
},
{
"epoch": 0.2070913084405397,
"grad_norm": 1.28125,
"learning_rate": 4.600102677873614e-05,
"loss": 6.5294,
"step": 2860
},
{
"epoch": 0.20781540392459752,
"grad_norm": 1.390625,
"learning_rate": 4.596915893863253e-05,
"loss": 6.5422,
"step": 2870
},
{
"epoch": 0.20853949940865535,
"grad_norm": 1.28125,
"learning_rate": 4.593717575449872e-05,
"loss": 6.5522,
"step": 2880
},
{
"epoch": 0.20926359489271318,
"grad_norm": 1.28125,
"learning_rate": 4.5905077402263034e-05,
"loss": 6.5464,
"step": 2890
},
{
"epoch": 0.209987690376771,
"grad_norm": 1.3046875,
"learning_rate": 4.587286405848732e-05,
"loss": 6.5167,
"step": 2900
},
{
"epoch": 0.21071178586082884,
"grad_norm": 1.2734375,
"learning_rate": 4.584053590036594e-05,
"loss": 6.5263,
"step": 2910
},
{
"epoch": 0.21143588134488667,
"grad_norm": 1.234375,
"learning_rate": 4.580809310572484e-05,
"loss": 6.547,
"step": 2920
},
{
"epoch": 0.2121599768289445,
"grad_norm": 1.296875,
"learning_rate": 4.57755358530205e-05,
"loss": 6.5113,
"step": 2930
},
{
"epoch": 0.21288407231300235,
"grad_norm": 1.25,
"learning_rate": 4.574286432133901e-05,
"loss": 6.5458,
"step": 2940
},
{
"epoch": 0.21360816779706018,
"grad_norm": 1.234375,
"learning_rate": 4.571007869039509e-05,
"loss": 6.5221,
"step": 2950
},
{
"epoch": 0.214332263281118,
"grad_norm": 1.203125,
"learning_rate": 4.567717914053103e-05,
"loss": 6.5321,
"step": 2960
},
{
"epoch": 0.21505635876517584,
"grad_norm": 1.25,
"learning_rate": 4.564416585271582e-05,
"loss": 6.5424,
"step": 2970
},
{
"epoch": 0.21578045424923367,
"grad_norm": 1.2890625,
"learning_rate": 4.561103900854401e-05,
"loss": 6.5315,
"step": 2980
},
{
"epoch": 0.2165045497332915,
"grad_norm": 1.28125,
"learning_rate": 4.557779879023483e-05,
"loss": 6.5202,
"step": 2990
},
{
"epoch": 0.21722864521734933,
"grad_norm": 1.328125,
"learning_rate": 4.554444538063113e-05,
"loss": 6.5246,
"step": 3000
},
{
"epoch": 0.21722864521734933,
"eval_loss": 6.682950973510742,
"eval_runtime": 1.4758,
"eval_samples_per_second": 8.131,
"eval_steps_per_second": 2.033,
"step": 3000
},
{
"epoch": 0.21795274070140716,
"grad_norm": 1.3125,
"learning_rate": 4.551097896319838e-05,
"loss": 6.5144,
"step": 3010
},
{
"epoch": 0.218676836185465,
"grad_norm": 1.3671875,
"learning_rate": 4.5477399722023674e-05,
"loss": 6.5295,
"step": 3020
},
{
"epoch": 0.21940093166952282,
"grad_norm": 1.3125,
"learning_rate": 4.5443707841814695e-05,
"loss": 6.5278,
"step": 3030
},
{
"epoch": 0.22012502715358065,
"grad_norm": 1.2734375,
"learning_rate": 4.540990350789875e-05,
"loss": 6.5584,
"step": 3040
},
{
"epoch": 0.22084912263763848,
"grad_norm": 1.3046875,
"learning_rate": 4.537598690622167e-05,
"loss": 6.5528,
"step": 3050
},
{
"epoch": 0.2215732181216963,
"grad_norm": 1.203125,
"learning_rate": 4.534195822334686e-05,
"loss": 6.5392,
"step": 3060
},
{
"epoch": 0.22229731360575414,
"grad_norm": 1.3984375,
"learning_rate": 4.530781764645422e-05,
"loss": 6.5403,
"step": 3070
},
{
"epoch": 0.22302140908981197,
"grad_norm": 1.3125,
"learning_rate": 4.5273565363339185e-05,
"loss": 6.5396,
"step": 3080
},
{
"epoch": 0.2237455045738698,
"grad_norm": 1.3125,
"learning_rate": 4.52392015624116e-05,
"loss": 6.5249,
"step": 3090
},
{
"epoch": 0.22446960005792763,
"grad_norm": 1.25,
"learning_rate": 4.520472643269477e-05,
"loss": 6.5199,
"step": 3100
},
{
"epoch": 0.22519369554198548,
"grad_norm": 1.375,
"learning_rate": 4.517014016382432e-05,
"loss": 6.5302,
"step": 3110
},
{
"epoch": 0.2259177910260433,
"grad_norm": 1.28125,
"learning_rate": 4.5135442946047294e-05,
"loss": 6.5274,
"step": 3120
},
{
"epoch": 0.22664188651010114,
"grad_norm": 1.1953125,
"learning_rate": 4.5100634970220967e-05,
"loss": 6.5431,
"step": 3130
},
{
"epoch": 0.22736598199415897,
"grad_norm": 1.2421875,
"learning_rate": 4.5065716427811874e-05,
"loss": 6.5362,
"step": 3140
},
{
"epoch": 0.2280900774782168,
"grad_norm": 1.3046875,
"learning_rate": 4.503068751089474e-05,
"loss": 6.5307,
"step": 3150
},
{
"epoch": 0.22881417296227463,
"grad_norm": 1.296875,
"learning_rate": 4.499554841215143e-05,
"loss": 6.5172,
"step": 3160
},
{
"epoch": 0.22953826844633246,
"grad_norm": 1.203125,
"learning_rate": 4.496029932486986e-05,
"loss": 6.521,
"step": 3170
},
{
"epoch": 0.2302623639303903,
"grad_norm": 1.1796875,
"learning_rate": 4.492494044294297e-05,
"loss": 6.5346,
"step": 3180
},
{
"epoch": 0.23098645941444812,
"grad_norm": 1.2578125,
"learning_rate": 4.4889471960867635e-05,
"loss": 6.5422,
"step": 3190
},
{
"epoch": 0.23171055489850595,
"grad_norm": 1.34375,
"learning_rate": 4.485389407374361e-05,
"loss": 6.5498,
"step": 3200
},
{
"epoch": 0.23243465038256378,
"grad_norm": 1.265625,
"learning_rate": 4.481820697727244e-05,
"loss": 6.523,
"step": 3210
},
{
"epoch": 0.2331587458666216,
"grad_norm": 1.1875,
"learning_rate": 4.47824108677564e-05,
"loss": 6.5477,
"step": 3220
},
{
"epoch": 0.23388284135067944,
"grad_norm": 1.3359375,
"learning_rate": 4.47465059420974e-05,
"loss": 6.5379,
"step": 3230
},
{
"epoch": 0.23460693683473727,
"grad_norm": 1.25,
"learning_rate": 4.471049239779592e-05,
"loss": 6.5389,
"step": 3240
},
{
"epoch": 0.2353310323187951,
"grad_norm": 1.3125,
"learning_rate": 4.4674370432949905e-05,
"loss": 6.5552,
"step": 3250
},
{
"epoch": 0.23605512780285293,
"grad_norm": 1.3046875,
"learning_rate": 4.463814024625368e-05,
"loss": 6.5114,
"step": 3260
},
{
"epoch": 0.23677922328691076,
"grad_norm": 1.28125,
"learning_rate": 4.460180203699688e-05,
"loss": 6.5101,
"step": 3270
},
{
"epoch": 0.23750331877096859,
"grad_norm": 1.1953125,
"learning_rate": 4.4565356005063304e-05,
"loss": 6.5051,
"step": 3280
},
{
"epoch": 0.23822741425502644,
"grad_norm": 1.3046875,
"learning_rate": 4.452880235092987e-05,
"loss": 6.5213,
"step": 3290
},
{
"epoch": 0.23895150973908427,
"grad_norm": 1.25,
"learning_rate": 4.449214127566549e-05,
"loss": 6.5246,
"step": 3300
},
{
"epoch": 0.2396756052231421,
"grad_norm": 1.2578125,
"learning_rate": 4.4455372980929935e-05,
"loss": 6.5309,
"step": 3310
},
{
"epoch": 0.24039970070719993,
"grad_norm": 1.359375,
"learning_rate": 4.4418497668972785e-05,
"loss": 6.5349,
"step": 3320
},
{
"epoch": 0.24112379619125776,
"grad_norm": 1.2421875,
"learning_rate": 4.4381515542632274e-05,
"loss": 6.5232,
"step": 3330
},
{
"epoch": 0.2418478916753156,
"grad_norm": 1.21875,
"learning_rate": 4.434442680533417e-05,
"loss": 6.518,
"step": 3340
},
{
"epoch": 0.24257198715937342,
"grad_norm": 1.2734375,
"learning_rate": 4.430723166109069e-05,
"loss": 6.5465,
"step": 3350
},
{
"epoch": 0.24329608264343125,
"grad_norm": 1.28125,
"learning_rate": 4.426993031449934e-05,
"loss": 6.5353,
"step": 3360
},
{
"epoch": 0.24402017812748908,
"grad_norm": 1.234375,
"learning_rate": 4.423252297074183e-05,
"loss": 6.5499,
"step": 3370
},
{
"epoch": 0.2447442736115469,
"grad_norm": 1.3203125,
"learning_rate": 4.41950098355829e-05,
"loss": 6.5203,
"step": 3380
},
{
"epoch": 0.24546836909560474,
"grad_norm": 1.3125,
"learning_rate": 4.415739111536924e-05,
"loss": 6.5392,
"step": 3390
},
{
"epoch": 0.24619246457966257,
"grad_norm": 1.3515625,
"learning_rate": 4.4119667017028297e-05,
"loss": 6.5197,
"step": 3400
},
{
"epoch": 0.2469165600637204,
"grad_norm": 1.2578125,
"learning_rate": 4.4081837748067186e-05,
"loss": 6.4986,
"step": 3410
},
{
"epoch": 0.24764065554777823,
"grad_norm": 1.21875,
"learning_rate": 4.404390351657153e-05,
"loss": 6.5238,
"step": 3420
},
{
"epoch": 0.24836475103183606,
"grad_norm": 1.25,
"learning_rate": 4.4005864531204285e-05,
"loss": 6.5473,
"step": 3430
},
{
"epoch": 0.24908884651589389,
"grad_norm": 1.1953125,
"learning_rate": 4.396772100120466e-05,
"loss": 6.5189,
"step": 3440
},
{
"epoch": 0.24981294199995172,
"grad_norm": 1.3046875,
"learning_rate": 4.39294731363869e-05,
"loss": 6.5228,
"step": 3450
},
{
"epoch": 0.2505370374840096,
"grad_norm": 1.296875,
"learning_rate": 4.389112114713918e-05,
"loss": 6.5224,
"step": 3460
},
{
"epoch": 0.2512611329680674,
"grad_norm": 1.3359375,
"learning_rate": 4.385266524442241e-05,
"loss": 6.5229,
"step": 3470
},
{
"epoch": 0.25198522845212523,
"grad_norm": 1.2578125,
"learning_rate": 4.3814105639769106e-05,
"loss": 6.54,
"step": 3480
},
{
"epoch": 0.25270932393618306,
"grad_norm": 1.34375,
"learning_rate": 4.37754425452822e-05,
"loss": 6.5226,
"step": 3490
},
{
"epoch": 0.2534334194202409,
"grad_norm": 1.21875,
"learning_rate": 4.373667617363389e-05,
"loss": 6.5276,
"step": 3500
},
{
"epoch": 0.2541575149042987,
"grad_norm": 1.3828125,
"learning_rate": 4.369780673806447e-05,
"loss": 6.534,
"step": 3510
},
{
"epoch": 0.25488161038835655,
"grad_norm": 1.203125,
"learning_rate": 4.365883445238116e-05,
"loss": 6.5317,
"step": 3520
},
{
"epoch": 0.2556057058724144,
"grad_norm": 1.2734375,
"learning_rate": 4.361975953095689e-05,
"loss": 6.5119,
"step": 3530
},
{
"epoch": 0.2563298013564722,
"grad_norm": 1.265625,
"learning_rate": 4.358058218872918e-05,
"loss": 6.5281,
"step": 3540
},
{
"epoch": 0.25705389684053004,
"grad_norm": 1.2734375,
"learning_rate": 4.354130264119894e-05,
"loss": 6.5284,
"step": 3550
},
{
"epoch": 0.25777799232458787,
"grad_norm": 1.3046875,
"learning_rate": 4.350192110442926e-05,
"loss": 6.5324,
"step": 3560
},
{
"epoch": 0.2585020878086457,
"grad_norm": 1.2265625,
"learning_rate": 4.346243779504421e-05,
"loss": 6.523,
"step": 3570
},
{
"epoch": 0.2592261832927035,
"grad_norm": 1.28125,
"learning_rate": 4.342285293022775e-05,
"loss": 6.5267,
"step": 3580
},
{
"epoch": 0.25995027877676136,
"grad_norm": 1.2265625,
"learning_rate": 4.338316672772238e-05,
"loss": 6.5263,
"step": 3590
},
{
"epoch": 0.2606743742608192,
"grad_norm": 1.21875,
"learning_rate": 4.334337940582808e-05,
"loss": 6.5438,
"step": 3600
},
{
"epoch": 0.261398469744877,
"grad_norm": 1.21875,
"learning_rate": 4.330349118340102e-05,
"loss": 6.5315,
"step": 3610
},
{
"epoch": 0.26212256522893485,
"grad_norm": 1.203125,
"learning_rate": 4.326350227985241e-05,
"loss": 6.5183,
"step": 3620
},
{
"epoch": 0.2628466607129927,
"grad_norm": 1.28125,
"learning_rate": 4.3223412915147254e-05,
"loss": 6.5214,
"step": 3630
},
{
"epoch": 0.2635707561970505,
"grad_norm": 1.3515625,
"learning_rate": 4.318322330980317e-05,
"loss": 6.531,
"step": 3640
},
{
"epoch": 0.26429485168110833,
"grad_norm": 1.2578125,
"learning_rate": 4.314293368488915e-05,
"loss": 6.5133,
"step": 3650
},
{
"epoch": 0.26501894716516616,
"grad_norm": 1.25,
"learning_rate": 4.3102544262024394e-05,
"loss": 6.5288,
"step": 3660
},
{
"epoch": 0.265743042649224,
"grad_norm": 1.1953125,
"learning_rate": 4.3062055263377e-05,
"loss": 6.4998,
"step": 3670
},
{
"epoch": 0.2664671381332818,
"grad_norm": 1.3203125,
"learning_rate": 4.302146691166286e-05,
"loss": 6.5302,
"step": 3680
},
{
"epoch": 0.26719123361733965,
"grad_norm": 1.265625,
"learning_rate": 4.298077943014431e-05,
"loss": 6.5053,
"step": 3690
},
{
"epoch": 0.2679153291013975,
"grad_norm": 1.2734375,
"learning_rate": 4.293999304262902e-05,
"loss": 6.5242,
"step": 3700
},
{
"epoch": 0.2686394245854553,
"grad_norm": 1.21875,
"learning_rate": 4.289910797346868e-05,
"loss": 6.5061,
"step": 3710
},
{
"epoch": 0.26936352006951314,
"grad_norm": 1.25,
"learning_rate": 4.285812444755779e-05,
"loss": 6.5257,
"step": 3720
},
{
"epoch": 0.27008761555357097,
"grad_norm": 1.1953125,
"learning_rate": 4.281704269033242e-05,
"loss": 6.5269,
"step": 3730
},
{
"epoch": 0.27081171103762885,
"grad_norm": 1.3046875,
"learning_rate": 4.2775862927769025e-05,
"loss": 6.4974,
"step": 3740
},
{
"epoch": 0.2715358065216867,
"grad_norm": 1.3671875,
"learning_rate": 4.2734585386383086e-05,
"loss": 6.4876,
"step": 3750
},
{
"epoch": 0.2722599020057445,
"grad_norm": 1.2734375,
"learning_rate": 4.269321029322797e-05,
"loss": 6.5099,
"step": 3760
},
{
"epoch": 0.27298399748980234,
"grad_norm": 1.265625,
"learning_rate": 4.265173787589364e-05,
"loss": 6.5222,
"step": 3770
},
{
"epoch": 0.2737080929738602,
"grad_norm": 1.2109375,
"learning_rate": 4.2610168362505395e-05,
"loss": 6.5077,
"step": 3780
},
{
"epoch": 0.274432188457918,
"grad_norm": 1.203125,
"learning_rate": 4.256850198172263e-05,
"loss": 6.5283,
"step": 3790
},
{
"epoch": 0.27515628394197583,
"grad_norm": 1.171875,
"learning_rate": 4.252673896273758e-05,
"loss": 6.5135,
"step": 3800
},
{
"epoch": 0.27588037942603366,
"grad_norm": 1.203125,
"learning_rate": 4.248487953527404e-05,
"loss": 6.5123,
"step": 3810
},
{
"epoch": 0.2766044749100915,
"grad_norm": 1.21875,
"learning_rate": 4.244292392958613e-05,
"loss": 6.5163,
"step": 3820
},
{
"epoch": 0.2773285703941493,
"grad_norm": 1.3203125,
"learning_rate": 4.2400872376457e-05,
"loss": 6.5063,
"step": 3830
},
{
"epoch": 0.27805266587820715,
"grad_norm": 1.2734375,
"learning_rate": 4.2358725107197576e-05,
"loss": 6.5214,
"step": 3840
},
{
"epoch": 0.278776761362265,
"grad_norm": 1.3046875,
"learning_rate": 4.231648235364529e-05,
"loss": 6.5375,
"step": 3850
},
{
"epoch": 0.2795008568463228,
"grad_norm": 1.2578125,
"learning_rate": 4.227414434816279e-05,
"loss": 6.5003,
"step": 3860
},
{
"epoch": 0.28022495233038064,
"grad_norm": 1.234375,
"learning_rate": 4.22317113236367e-05,
"loss": 6.5425,
"step": 3870
},
{
"epoch": 0.28094904781443847,
"grad_norm": 1.1953125,
"learning_rate": 4.218918351347626e-05,
"loss": 6.5073,
"step": 3880
},
{
"epoch": 0.2816731432984963,
"grad_norm": 1.3125,
"learning_rate": 4.214656115161215e-05,
"loss": 6.5127,
"step": 3890
},
{
"epoch": 0.2823972387825541,
"grad_norm": 1.2265625,
"learning_rate": 4.210384447249509e-05,
"loss": 6.5198,
"step": 3900
},
{
"epoch": 0.28312133426661196,
"grad_norm": 1.234375,
"learning_rate": 4.2061033711094655e-05,
"loss": 6.5206,
"step": 3910
},
{
"epoch": 0.2838454297506698,
"grad_norm": 1.3125,
"learning_rate": 4.2018129102897904e-05,
"loss": 6.4952,
"step": 3920
},
{
"epoch": 0.2845695252347276,
"grad_norm": 1.2265625,
"learning_rate": 4.197513088390813e-05,
"loss": 6.5065,
"step": 3930
},
{
"epoch": 0.28529362071878545,
"grad_norm": 1.265625,
"learning_rate": 4.193203929064353e-05,
"loss": 6.4917,
"step": 3940
},
{
"epoch": 0.2860177162028433,
"grad_norm": 1.28125,
"learning_rate": 4.1888854560135934e-05,
"loss": 6.5252,
"step": 3950
},
{
"epoch": 0.2867418116869011,
"grad_norm": 1.2578125,
"learning_rate": 4.1845576929929486e-05,
"loss": 6.4974,
"step": 3960
},
{
"epoch": 0.28746590717095893,
"grad_norm": 1.234375,
"learning_rate": 4.180220663807934e-05,
"loss": 6.5253,
"step": 3970
},
{
"epoch": 0.28819000265501676,
"grad_norm": 1.265625,
"learning_rate": 4.175874392315033e-05,
"loss": 6.4986,
"step": 3980
},
{
"epoch": 0.2889140981390746,
"grad_norm": 1.3203125,
"learning_rate": 4.1715189024215716e-05,
"loss": 6.5147,
"step": 3990
},
{
"epoch": 0.2896381936231324,
"grad_norm": 1.3125,
"learning_rate": 4.1671542180855796e-05,
"loss": 6.5102,
"step": 4000
},
{
"epoch": 0.2896381936231324,
"eval_loss": 6.658437252044678,
"eval_runtime": 1.632,
"eval_samples_per_second": 7.353,
"eval_steps_per_second": 1.838,
"step": 4000
},
{
"epoch": 0.29036228910719025,
"grad_norm": 1.2421875,
"learning_rate": 4.162780363315662e-05,
"loss": 6.5095,
"step": 4010
},
{
"epoch": 0.2910863845912481,
"grad_norm": 1.3125,
"learning_rate": 4.158397362170871e-05,
"loss": 6.4945,
"step": 4020
},
{
"epoch": 0.2918104800753059,
"grad_norm": 1.359375,
"learning_rate": 4.154005238760565e-05,
"loss": 6.4935,
"step": 4030
},
{
"epoch": 0.29253457555936374,
"grad_norm": 1.28125,
"learning_rate": 4.149604017244285e-05,
"loss": 6.5137,
"step": 4040
},
{
"epoch": 0.29325867104342157,
"grad_norm": 1.25,
"learning_rate": 4.145193721831615e-05,
"loss": 6.5045,
"step": 4050
},
{
"epoch": 0.2939827665274794,
"grad_norm": 1.296875,
"learning_rate": 4.1407743767820504e-05,
"loss": 6.5433,
"step": 4060
},
{
"epoch": 0.29470686201153723,
"grad_norm": 1.3671875,
"learning_rate": 4.136346006404869e-05,
"loss": 6.5018,
"step": 4070
},
{
"epoch": 0.29543095749559506,
"grad_norm": 1.234375,
"learning_rate": 4.1319086350589916e-05,
"loss": 6.5069,
"step": 4080
},
{
"epoch": 0.29615505297965294,
"grad_norm": 1.3359375,
"learning_rate": 4.127462287152849e-05,
"loss": 6.5224,
"step": 4090
},
{
"epoch": 0.2968791484637108,
"grad_norm": 1.234375,
"learning_rate": 4.1230069871442525e-05,
"loss": 6.5215,
"step": 4100
},
{
"epoch": 0.2976032439477686,
"grad_norm": 1.265625,
"learning_rate": 4.118542759540251e-05,
"loss": 6.5162,
"step": 4110
},
{
"epoch": 0.29832733943182643,
"grad_norm": 1.2734375,
"learning_rate": 4.114069628897006e-05,
"loss": 6.4849,
"step": 4120
},
{
"epoch": 0.29905143491588426,
"grad_norm": 1.265625,
"learning_rate": 4.109587619819648e-05,
"loss": 6.5217,
"step": 4130
},
{
"epoch": 0.2997755303999421,
"grad_norm": 1.3515625,
"learning_rate": 4.1050967569621465e-05,
"loss": 6.5011,
"step": 4140
},
{
"epoch": 0.3004996258839999,
"grad_norm": 1.1640625,
"learning_rate": 4.1005970650271705e-05,
"loss": 6.4931,
"step": 4150
},
{
"epoch": 0.30122372136805775,
"grad_norm": 1.1796875,
"learning_rate": 4.096088568765958e-05,
"loss": 6.5058,
"step": 4160
},
{
"epoch": 0.3019478168521156,
"grad_norm": 1.28125,
"learning_rate": 4.091571292978173e-05,
"loss": 6.4999,
"step": 4170
},
{
"epoch": 0.3026719123361734,
"grad_norm": 1.21875,
"learning_rate": 4.087045262511775e-05,
"loss": 6.5029,
"step": 4180
},
{
"epoch": 0.30339600782023124,
"grad_norm": 1.3125,
"learning_rate": 4.082510502262879e-05,
"loss": 6.4853,
"step": 4190
},
{
"epoch": 0.30412010330428907,
"grad_norm": 1.21875,
"learning_rate": 4.0779670371756184e-05,
"loss": 6.4962,
"step": 4200
},
{
"epoch": 0.3048441987883469,
"grad_norm": 1.2578125,
"learning_rate": 4.073414892242011e-05,
"loss": 6.5212,
"step": 4210
},
{
"epoch": 0.30556829427240473,
"grad_norm": 1.28125,
"learning_rate": 4.068854092501819e-05,
"loss": 6.5144,
"step": 4220
},
{
"epoch": 0.30629238975646256,
"grad_norm": 1.203125,
"learning_rate": 4.0642846630424114e-05,
"loss": 6.5349,
"step": 4230
},
{
"epoch": 0.3070164852405204,
"grad_norm": 1.28125,
"learning_rate": 4.059706628998625e-05,
"loss": 6.4897,
"step": 4240
},
{
"epoch": 0.3077405807245782,
"grad_norm": 1.2265625,
"learning_rate": 4.055120015552628e-05,
"loss": 6.5038,
"step": 4250
},
{
"epoch": 0.30846467620863605,
"grad_norm": 1.2421875,
"learning_rate": 4.050524847933783e-05,
"loss": 6.4993,
"step": 4260
},
{
"epoch": 0.3091887716926939,
"grad_norm": 1.234375,
"learning_rate": 4.045921151418504e-05,
"loss": 6.5261,
"step": 4270
},
{
"epoch": 0.3099128671767517,
"grad_norm": 1.2890625,
"learning_rate": 4.041308951330119e-05,
"loss": 6.5095,
"step": 4280
},
{
"epoch": 0.31063696266080953,
"grad_norm": 1.21875,
"learning_rate": 4.036688273038734e-05,
"loss": 6.5032,
"step": 4290
},
{
"epoch": 0.31136105814486736,
"grad_norm": 1.3203125,
"learning_rate": 4.032059141961086e-05,
"loss": 6.4758,
"step": 4300
},
{
"epoch": 0.3120851536289252,
"grad_norm": 1.2421875,
"learning_rate": 4.027421583560414e-05,
"loss": 6.4878,
"step": 4310
},
{
"epoch": 0.312809249112983,
"grad_norm": 1.2890625,
"learning_rate": 4.022775623346306e-05,
"loss": 6.5128,
"step": 4320
},
{
"epoch": 0.31353334459704085,
"grad_norm": 1.328125,
"learning_rate": 4.018121286874571e-05,
"loss": 6.5149,
"step": 4330
},
{
"epoch": 0.3142574400810987,
"grad_norm": 1.2890625,
"learning_rate": 4.013458599747091e-05,
"loss": 6.5083,
"step": 4340
},
{
"epoch": 0.3149815355651565,
"grad_norm": 1.234375,
"learning_rate": 4.00878758761168e-05,
"loss": 6.516,
"step": 4350
},
{
"epoch": 0.31570563104921434,
"grad_norm": 1.15625,
"learning_rate": 4.0041082761619476e-05,
"loss": 6.4992,
"step": 4360
},
{
"epoch": 0.31642972653327217,
"grad_norm": 1.1796875,
"learning_rate": 3.9994206911371544e-05,
"loss": 6.492,
"step": 4370
},
{
"epoch": 0.31715382201733,
"grad_norm": 1.2421875,
"learning_rate": 3.994724858322071e-05,
"loss": 6.5087,
"step": 4380
},
{
"epoch": 0.31787791750138783,
"grad_norm": 1.25,
"learning_rate": 3.990020803546835e-05,
"loss": 6.485,
"step": 4390
},
{
"epoch": 0.31860201298544566,
"grad_norm": 1.2734375,
"learning_rate": 3.985308552686812e-05,
"loss": 6.4696,
"step": 4400
},
{
"epoch": 0.3193261084695035,
"grad_norm": 1.234375,
"learning_rate": 3.9805881316624506e-05,
"loss": 6.4946,
"step": 4410
},
{
"epoch": 0.3200502039535613,
"grad_norm": 1.2578125,
"learning_rate": 3.9758595664391396e-05,
"loss": 6.5153,
"step": 4420
},
{
"epoch": 0.32077429943761915,
"grad_norm": 1.2421875,
"learning_rate": 3.9711228830270684e-05,
"loss": 6.5096,
"step": 4430
},
{
"epoch": 0.321498394921677,
"grad_norm": 1.1640625,
"learning_rate": 3.9663781074810796e-05,
"loss": 6.52,
"step": 4440
},
{
"epoch": 0.32222249040573486,
"grad_norm": 1.234375,
"learning_rate": 3.961625265900529e-05,
"loss": 6.4891,
"step": 4450
},
{
"epoch": 0.3229465858897927,
"grad_norm": 1.25,
"learning_rate": 3.956864384429142e-05,
"loss": 6.4985,
"step": 4460
},
{
"epoch": 0.3236706813738505,
"grad_norm": 1.265625,
"learning_rate": 3.952095489254864e-05,
"loss": 6.5256,
"step": 4470
},
{
"epoch": 0.32439477685790835,
"grad_norm": 1.203125,
"learning_rate": 3.947318606609727e-05,
"loss": 6.4904,
"step": 4480
},
{
"epoch": 0.3251188723419662,
"grad_norm": 1.21875,
"learning_rate": 3.942533762769695e-05,
"loss": 6.4722,
"step": 4490
},
{
"epoch": 0.325842967826024,
"grad_norm": 1.2734375,
"learning_rate": 3.937740984054526e-05,
"loss": 6.4933,
"step": 4500
},
{
"epoch": 0.32656706331008184,
"grad_norm": 1.2265625,
"learning_rate": 3.932940296827624e-05,
"loss": 6.5028,
"step": 4510
},
{
"epoch": 0.32729115879413967,
"grad_norm": 1.15625,
"learning_rate": 3.928131727495895e-05,
"loss": 6.5001,
"step": 4520
},
{
"epoch": 0.3280152542781975,
"grad_norm": 1.1953125,
"learning_rate": 3.923315302509602e-05,
"loss": 6.488,
"step": 4530
},
{
"epoch": 0.32873934976225533,
"grad_norm": 1.2109375,
"learning_rate": 3.918491048362219e-05,
"loss": 6.4929,
"step": 4540
},
{
"epoch": 0.32946344524631316,
"grad_norm": 1.2890625,
"learning_rate": 3.9136589915902856e-05,
"loss": 6.5015,
"step": 4550
},
{
"epoch": 0.330187540730371,
"grad_norm": 1.3359375,
"learning_rate": 3.9088191587732606e-05,
"loss": 6.4959,
"step": 4560
},
{
"epoch": 0.3309116362144288,
"grad_norm": 1.234375,
"learning_rate": 3.9039715765333766e-05,
"loss": 6.4823,
"step": 4570
},
{
"epoch": 0.33163573169848665,
"grad_norm": 1.234375,
"learning_rate": 3.8991162715354936e-05,
"loss": 6.4936,
"step": 4580
},
{
"epoch": 0.3323598271825445,
"grad_norm": 1.1640625,
"learning_rate": 3.894253270486951e-05,
"loss": 6.5096,
"step": 4590
},
{
"epoch": 0.3330839226666023,
"grad_norm": 1.2421875,
"learning_rate": 3.88938260013742e-05,
"loss": 6.4944,
"step": 4600
},
{
"epoch": 0.33380801815066014,
"grad_norm": 1.28125,
"learning_rate": 3.884504287278761e-05,
"loss": 6.5183,
"step": 4610
},
{
"epoch": 0.33453211363471796,
"grad_norm": 1.203125,
"learning_rate": 3.879618358744871e-05,
"loss": 6.4898,
"step": 4620
},
{
"epoch": 0.3352562091187758,
"grad_norm": 1.21875,
"learning_rate": 3.874724841411539e-05,
"loss": 6.4888,
"step": 4630
},
{
"epoch": 0.3359803046028336,
"grad_norm": 1.2265625,
"learning_rate": 3.8698237621962974e-05,
"loss": 6.504,
"step": 4640
},
{
"epoch": 0.33670440008689145,
"grad_norm": 1.2421875,
"learning_rate": 3.8649151480582745e-05,
"loss": 6.514,
"step": 4650
},
{
"epoch": 0.3374284955709493,
"grad_norm": 1.28125,
"learning_rate": 3.8599990259980435e-05,
"loss": 6.5113,
"step": 4660
},
{
"epoch": 0.3381525910550071,
"grad_norm": 1.28125,
"learning_rate": 3.855075423057477e-05,
"loss": 6.494,
"step": 4670
},
{
"epoch": 0.33887668653906494,
"grad_norm": 1.2421875,
"learning_rate": 3.850144366319599e-05,
"loss": 6.4887,
"step": 4680
},
{
"epoch": 0.33960078202312277,
"grad_norm": 1.2578125,
"learning_rate": 3.845205882908432e-05,
"loss": 6.4935,
"step": 4690
},
{
"epoch": 0.3403248775071806,
"grad_norm": 1.265625,
"learning_rate": 3.8402599999888514e-05,
"loss": 6.5083,
"step": 4700
},
{
"epoch": 0.34104897299123843,
"grad_norm": 1.21875,
"learning_rate": 3.835306744766434e-05,
"loss": 6.4812,
"step": 4710
},
{
"epoch": 0.34177306847529626,
"grad_norm": 1.234375,
"learning_rate": 3.830346144487309e-05,
"loss": 6.4605,
"step": 4720
},
{
"epoch": 0.3424971639593541,
"grad_norm": 1.296875,
"learning_rate": 3.825378226438009e-05,
"loss": 6.4931,
"step": 4730
},
{
"epoch": 0.3432212594434119,
"grad_norm": 1.1953125,
"learning_rate": 3.820403017945319e-05,
"loss": 6.4958,
"step": 4740
},
{
"epoch": 0.34394535492746975,
"grad_norm": 1.21875,
"learning_rate": 3.815420546376124e-05,
"loss": 6.4939,
"step": 4750
},
{
"epoch": 0.3446694504115276,
"grad_norm": 1.2578125,
"learning_rate": 3.8104308391372646e-05,
"loss": 6.494,
"step": 4760
},
{
"epoch": 0.3453935458955854,
"grad_norm": 1.2421875,
"learning_rate": 3.805433923675379e-05,
"loss": 6.4772,
"step": 4770
},
{
"epoch": 0.34611764137964324,
"grad_norm": 1.265625,
"learning_rate": 3.800429827476757e-05,
"loss": 6.497,
"step": 4780
},
{
"epoch": 0.34684173686370107,
"grad_norm": 1.1953125,
"learning_rate": 3.795418578067185e-05,
"loss": 6.4894,
"step": 4790
},
{
"epoch": 0.34756583234775895,
"grad_norm": 1.203125,
"learning_rate": 3.7904002030118e-05,
"loss": 6.5089,
"step": 4800
},
{
"epoch": 0.3482899278318168,
"grad_norm": 1.2578125,
"learning_rate": 3.7853747299149325e-05,
"loss": 6.5015,
"step": 4810
},
{
"epoch": 0.3490140233158746,
"grad_norm": 1.1953125,
"learning_rate": 3.7803421864199576e-05,
"loss": 6.4898,
"step": 4820
},
{
"epoch": 0.34973811879993244,
"grad_norm": 1.234375,
"learning_rate": 3.775302600209141e-05,
"loss": 6.4994,
"step": 4830
},
{
"epoch": 0.35046221428399027,
"grad_norm": 1.2265625,
"learning_rate": 3.7702559990034906e-05,
"loss": 6.4958,
"step": 4840
},
{
"epoch": 0.3511863097680481,
"grad_norm": 1.2734375,
"learning_rate": 3.765202410562598e-05,
"loss": 6.5129,
"step": 4850
},
{
"epoch": 0.35191040525210593,
"grad_norm": 1.2734375,
"learning_rate": 3.760141862684492e-05,
"loss": 6.5063,
"step": 4860
},
{
"epoch": 0.35263450073616376,
"grad_norm": 1.2421875,
"learning_rate": 3.75507438320548e-05,
"loss": 6.4901,
"step": 4870
},
{
"epoch": 0.3533585962202216,
"grad_norm": 1.203125,
"learning_rate": 3.7500000000000003e-05,
"loss": 6.5077,
"step": 4880
},
{
"epoch": 0.3540826917042794,
"grad_norm": 1.3125,
"learning_rate": 3.7449187409804655e-05,
"loss": 6.4971,
"step": 4890
},
{
"epoch": 0.35480678718833725,
"grad_norm": 1.21875,
"learning_rate": 3.739830634097109e-05,
"loss": 6.4796,
"step": 4900
},
{
"epoch": 0.3555308826723951,
"grad_norm": 1.234375,
"learning_rate": 3.734735707337831e-05,
"loss": 6.4935,
"step": 4910
},
{
"epoch": 0.3562549781564529,
"grad_norm": 1.234375,
"learning_rate": 3.729633988728049e-05,
"loss": 6.4704,
"step": 4920
},
{
"epoch": 0.35697907364051074,
"grad_norm": 1.265625,
"learning_rate": 3.724525506330536e-05,
"loss": 6.51,
"step": 4930
},
{
"epoch": 0.35770316912456857,
"grad_norm": 1.1875,
"learning_rate": 3.719410288245273e-05,
"loss": 6.4905,
"step": 4940
},
{
"epoch": 0.3584272646086264,
"grad_norm": 1.171875,
"learning_rate": 3.7142883626092914e-05,
"loss": 6.4786,
"step": 4950
},
{
"epoch": 0.3591513600926842,
"grad_norm": 1.3046875,
"learning_rate": 3.709159757596516e-05,
"loss": 6.4824,
"step": 4960
},
{
"epoch": 0.35987545557674205,
"grad_norm": 1.25,
"learning_rate": 3.704024501417615e-05,
"loss": 6.5007,
"step": 4970
},
{
"epoch": 0.3605995510607999,
"grad_norm": 1.1875,
"learning_rate": 3.6988826223198425e-05,
"loss": 6.4778,
"step": 4980
},
{
"epoch": 0.3613236465448577,
"grad_norm": 1.265625,
"learning_rate": 3.693734148586882e-05,
"loss": 6.4788,
"step": 4990
},
{
"epoch": 0.36204774202891554,
"grad_norm": 1.1796875,
"learning_rate": 3.6885791085386905e-05,
"loss": 6.5081,
"step": 5000
},
{
"epoch": 0.36204774202891554,
"eval_loss": 6.645383834838867,
"eval_runtime": 1.5549,
"eval_samples_per_second": 7.717,
"eval_steps_per_second": 1.929,
"step": 5000
},
{
"epoch": 0.36277183751297337,
"grad_norm": 1.2265625,
"learning_rate": 3.683417530531348e-05,
"loss": 6.4752,
"step": 5010
},
{
"epoch": 0.3634959329970312,
"grad_norm": 1.3203125,
"learning_rate": 3.678249442956892e-05,
"loss": 6.4981,
"step": 5020
},
{
"epoch": 0.36422002848108903,
"grad_norm": 1.2109375,
"learning_rate": 3.673074874243172e-05,
"loss": 6.4765,
"step": 5030
},
{
"epoch": 0.36494412396514686,
"grad_norm": 1.234375,
"learning_rate": 3.667893852853685e-05,
"loss": 6.4939,
"step": 5040
},
{
"epoch": 0.3656682194492047,
"grad_norm": 1.1796875,
"learning_rate": 3.662706407287424e-05,
"loss": 6.4985,
"step": 5050
},
{
"epoch": 0.3663923149332625,
"grad_norm": 1.234375,
"learning_rate": 3.6575125660787156e-05,
"loss": 6.4764,
"step": 5060
},
{
"epoch": 0.36711641041732035,
"grad_norm": 1.21875,
"learning_rate": 3.6523123577970694e-05,
"loss": 6.4896,
"step": 5070
},
{
"epoch": 0.3678405059013782,
"grad_norm": 1.296875,
"learning_rate": 3.64710581104702e-05,
"loss": 6.4858,
"step": 5080
},
{
"epoch": 0.368564601385436,
"grad_norm": 1.2109375,
"learning_rate": 3.6418929544679635e-05,
"loss": 6.4957,
"step": 5090
},
{
"epoch": 0.36928869686949384,
"grad_norm": 1.1796875,
"learning_rate": 3.6366738167340055e-05,
"loss": 6.5179,
"step": 5100
},
{
"epoch": 0.37001279235355167,
"grad_norm": 1.3046875,
"learning_rate": 3.631448426553803e-05,
"loss": 6.489,
"step": 5110
},
{
"epoch": 0.3707368878376095,
"grad_norm": 1.2734375,
"learning_rate": 3.626216812670405e-05,
"loss": 6.4871,
"step": 5120
},
{
"epoch": 0.3714609833216673,
"grad_norm": 1.2421875,
"learning_rate": 3.620979003861095e-05,
"loss": 6.4833,
"step": 5130
},
{
"epoch": 0.37218507880572516,
"grad_norm": 1.359375,
"learning_rate": 3.615735028937232e-05,
"loss": 6.4712,
"step": 5140
},
{
"epoch": 0.372909174289783,
"grad_norm": 1.203125,
"learning_rate": 3.610484916744094e-05,
"loss": 6.49,
"step": 5150
},
{
"epoch": 0.37363326977384087,
"grad_norm": 1.2265625,
"learning_rate": 3.6052286961607166e-05,
"loss": 6.4859,
"step": 5160
},
{
"epoch": 0.3743573652578987,
"grad_norm": 1.265625,
"learning_rate": 3.599966396099737e-05,
"loss": 6.4697,
"step": 5170
},
{
"epoch": 0.37508146074195653,
"grad_norm": 1.25,
"learning_rate": 3.594698045507231e-05,
"loss": 6.4875,
"step": 5180
},
{
"epoch": 0.37580555622601436,
"grad_norm": 1.2421875,
"learning_rate": 3.5894236733625594e-05,
"loss": 6.4719,
"step": 5190
},
{
"epoch": 0.3765296517100722,
"grad_norm": 1.1875,
"learning_rate": 3.584143308678203e-05,
"loss": 6.4674,
"step": 5200
},
{
"epoch": 0.37725374719413,
"grad_norm": 1.21875,
"learning_rate": 3.578856980499606e-05,
"loss": 6.4724,
"step": 5210
},
{
"epoch": 0.37797784267818785,
"grad_norm": 1.2265625,
"learning_rate": 3.573564717905017e-05,
"loss": 6.4982,
"step": 5220
},
{
"epoch": 0.3787019381622457,
"grad_norm": 1.2578125,
"learning_rate": 3.568266550005327e-05,
"loss": 6.496,
"step": 5230
},
{
"epoch": 0.3794260336463035,
"grad_norm": 1.1796875,
"learning_rate": 3.5629625059439084e-05,
"loss": 6.4764,
"step": 5240
},
{
"epoch": 0.38015012913036134,
"grad_norm": 1.234375,
"learning_rate": 3.557652614896459e-05,
"loss": 6.5149,
"step": 5250
},
{
"epoch": 0.38087422461441917,
"grad_norm": 1.203125,
"learning_rate": 3.552336906070838e-05,
"loss": 6.489,
"step": 5260
},
{
"epoch": 0.381598320098477,
"grad_norm": 1.25,
"learning_rate": 3.547015408706904e-05,
"loss": 6.455,
"step": 5270
},
{
"epoch": 0.3823224155825348,
"grad_norm": 1.25,
"learning_rate": 3.54168815207636e-05,
"loss": 6.481,
"step": 5280
},
{
"epoch": 0.38304651106659265,
"grad_norm": 1.28125,
"learning_rate": 3.536355165482586e-05,
"loss": 6.466,
"step": 5290
},
{
"epoch": 0.3837706065506505,
"grad_norm": 1.3125,
"learning_rate": 3.531016478260481e-05,
"loss": 6.4746,
"step": 5300
},
{
"epoch": 0.3844947020347083,
"grad_norm": 1.2421875,
"learning_rate": 3.525672119776302e-05,
"loss": 6.5012,
"step": 5310
},
{
"epoch": 0.38521879751876614,
"grad_norm": 1.265625,
"learning_rate": 3.5203221194275e-05,
"loss": 6.4712,
"step": 5320
},
{
"epoch": 0.385942893002824,
"grad_norm": 1.25,
"learning_rate": 3.5149665066425636e-05,
"loss": 6.4787,
"step": 5330
},
{
"epoch": 0.3866669884868818,
"grad_norm": 1.234375,
"learning_rate": 3.5096053108808486e-05,
"loss": 6.4847,
"step": 5340
},
{
"epoch": 0.38739108397093963,
"grad_norm": 1.171875,
"learning_rate": 3.504238561632424e-05,
"loss": 6.4859,
"step": 5350
},
{
"epoch": 0.38811517945499746,
"grad_norm": 1.234375,
"learning_rate": 3.498866288417906e-05,
"loss": 6.482,
"step": 5360
},
{
"epoch": 0.3888392749390553,
"grad_norm": 1.1640625,
"learning_rate": 3.493488520788295e-05,
"loss": 6.4888,
"step": 5370
},
{
"epoch": 0.3895633704231131,
"grad_norm": 1.234375,
"learning_rate": 3.4881052883248155e-05,
"loss": 6.49,
"step": 5380
},
{
"epoch": 0.39028746590717095,
"grad_norm": 1.265625,
"learning_rate": 3.482716620638753e-05,
"loss": 6.4738,
"step": 5390
},
{
"epoch": 0.3910115613912288,
"grad_norm": 1.1953125,
"learning_rate": 3.477322547371288e-05,
"loss": 6.4959,
"step": 5400
},
{
"epoch": 0.3917356568752866,
"grad_norm": 1.21875,
"learning_rate": 3.4719230981933366e-05,
"loss": 6.5087,
"step": 5410
},
{
"epoch": 0.39245975235934444,
"grad_norm": 1.1953125,
"learning_rate": 3.466518302805386e-05,
"loss": 6.4963,
"step": 5420
},
{
"epoch": 0.39318384784340227,
"grad_norm": 1.265625,
"learning_rate": 3.46110819093733e-05,
"loss": 6.4831,
"step": 5430
},
{
"epoch": 0.3939079433274601,
"grad_norm": 1.203125,
"learning_rate": 3.455692792348307e-05,
"loss": 6.4731,
"step": 5440
},
{
"epoch": 0.3946320388115179,
"grad_norm": 1.2265625,
"learning_rate": 3.450272136826537e-05,
"loss": 6.4805,
"step": 5450
},
{
"epoch": 0.39535613429557576,
"grad_norm": 1.2578125,
"learning_rate": 3.444846254189156e-05,
"loss": 6.4445,
"step": 5460
},
{
"epoch": 0.3960802297796336,
"grad_norm": 1.296875,
"learning_rate": 3.439415174282049e-05,
"loss": 6.4815,
"step": 5470
},
{
"epoch": 0.3968043252636914,
"grad_norm": 1.21875,
"learning_rate": 3.4339789269796964e-05,
"loss": 6.479,
"step": 5480
},
{
"epoch": 0.39752842074774924,
"grad_norm": 1.2109375,
"learning_rate": 3.4285375421849954e-05,
"loss": 6.4897,
"step": 5490
},
{
"epoch": 0.3982525162318071,
"grad_norm": 1.21875,
"learning_rate": 3.423091049829108e-05,
"loss": 6.4966,
"step": 5500
},
{
"epoch": 0.39897661171586496,
"grad_norm": 1.28125,
"learning_rate": 3.4176394798712886e-05,
"loss": 6.5001,
"step": 5510
},
{
"epoch": 0.3997007071999228,
"grad_norm": 1.21875,
"learning_rate": 3.412182862298723e-05,
"loss": 6.4965,
"step": 5520
},
{
"epoch": 0.4004248026839806,
"grad_norm": 1.1171875,
"learning_rate": 3.406721227126361e-05,
"loss": 6.4859,
"step": 5530
},
{
"epoch": 0.40114889816803845,
"grad_norm": 1.2578125,
"learning_rate": 3.401254604396751e-05,
"loss": 6.4876,
"step": 5540
},
{
"epoch": 0.4018729936520963,
"grad_norm": 1.2265625,
"learning_rate": 3.3957830241798807e-05,
"loss": 6.493,
"step": 5550
},
{
"epoch": 0.4025970891361541,
"grad_norm": 1.171875,
"learning_rate": 3.390306516573004e-05,
"loss": 6.4825,
"step": 5560
},
{
"epoch": 0.40332118462021194,
"grad_norm": 1.171875,
"learning_rate": 3.3848251117004786e-05,
"loss": 6.4948,
"step": 5570
},
{
"epoch": 0.40404528010426977,
"grad_norm": 1.1875,
"learning_rate": 3.379338839713601e-05,
"loss": 6.465,
"step": 5580
},
{
"epoch": 0.4047693755883276,
"grad_norm": 1.234375,
"learning_rate": 3.373847730790442e-05,
"loss": 6.4826,
"step": 5590
},
{
"epoch": 0.4054934710723854,
"grad_norm": 1.2421875,
"learning_rate": 3.368351815135674e-05,
"loss": 6.4876,
"step": 5600
},
{
"epoch": 0.40621756655644325,
"grad_norm": 1.3125,
"learning_rate": 3.362851122980413e-05,
"loss": 6.4724,
"step": 5610
},
{
"epoch": 0.4069416620405011,
"grad_norm": 1.21875,
"learning_rate": 3.35734568458205e-05,
"loss": 6.4725,
"step": 5620
},
{
"epoch": 0.4076657575245589,
"grad_norm": 1.1640625,
"learning_rate": 3.351835530224079e-05,
"loss": 6.4914,
"step": 5630
},
{
"epoch": 0.40838985300861674,
"grad_norm": 1.15625,
"learning_rate": 3.3463206902159395e-05,
"loss": 6.4716,
"step": 5640
},
{
"epoch": 0.4091139484926746,
"grad_norm": 1.2109375,
"learning_rate": 3.3408011948928415e-05,
"loss": 6.4575,
"step": 5650
},
{
"epoch": 0.4098380439767324,
"grad_norm": 1.25,
"learning_rate": 3.335277074615605e-05,
"loss": 6.5197,
"step": 5660
},
{
"epoch": 0.41056213946079023,
"grad_norm": 1.2265625,
"learning_rate": 3.329748359770488e-05,
"loss": 6.4902,
"step": 5670
},
{
"epoch": 0.41128623494484806,
"grad_norm": 1.2890625,
"learning_rate": 3.3242150807690234e-05,
"loss": 6.481,
"step": 5680
},
{
"epoch": 0.4120103304289059,
"grad_norm": 1.203125,
"learning_rate": 3.3186772680478496e-05,
"loss": 6.4396,
"step": 5690
},
{
"epoch": 0.4127344259129637,
"grad_norm": 1.3203125,
"learning_rate": 3.313134952068543e-05,
"loss": 6.4679,
"step": 5700
},
{
"epoch": 0.41345852139702155,
"grad_norm": 1.25,
"learning_rate": 3.307588163317449e-05,
"loss": 6.4912,
"step": 5710
},
{
"epoch": 0.4141826168810794,
"grad_norm": 1.2734375,
"learning_rate": 3.30203693230552e-05,
"loss": 6.4688,
"step": 5720
},
{
"epoch": 0.4149067123651372,
"grad_norm": 1.2578125,
"learning_rate": 3.296481289568141e-05,
"loss": 6.494,
"step": 5730
},
{
"epoch": 0.41563080784919504,
"grad_norm": 1.2265625,
"learning_rate": 3.290921265664966e-05,
"loss": 6.458,
"step": 5740
},
{
"epoch": 0.41635490333325287,
"grad_norm": 1.234375,
"learning_rate": 3.2853568911797465e-05,
"loss": 6.4874,
"step": 5750
},
{
"epoch": 0.4170789988173107,
"grad_norm": 1.234375,
"learning_rate": 3.279788196720167e-05,
"loss": 6.4784,
"step": 5760
},
{
"epoch": 0.4178030943013685,
"grad_norm": 1.2421875,
"learning_rate": 3.274215212917672e-05,
"loss": 6.4754,
"step": 5770
},
{
"epoch": 0.41852718978542636,
"grad_norm": 1.1796875,
"learning_rate": 3.268637970427302e-05,
"loss": 6.4881,
"step": 5780
},
{
"epoch": 0.4192512852694842,
"grad_norm": 1.2421875,
"learning_rate": 3.263056499927524e-05,
"loss": 6.4943,
"step": 5790
},
{
"epoch": 0.419975380753542,
"grad_norm": 1.234375,
"learning_rate": 3.2574708321200606e-05,
"loss": 6.4685,
"step": 5800
},
{
"epoch": 0.42069947623759985,
"grad_norm": 1.2890625,
"learning_rate": 3.251880997729722e-05,
"loss": 6.4982,
"step": 5810
},
{
"epoch": 0.4214235717216577,
"grad_norm": 1.171875,
"learning_rate": 3.246287027504237e-05,
"loss": 6.4781,
"step": 5820
},
{
"epoch": 0.4221476672057155,
"grad_norm": 1.203125,
"learning_rate": 3.2406889522140856e-05,
"loss": 6.4914,
"step": 5830
},
{
"epoch": 0.42287176268977333,
"grad_norm": 1.203125,
"learning_rate": 3.235086802652327e-05,
"loss": 6.4663,
"step": 5840
},
{
"epoch": 0.42359585817383116,
"grad_norm": 1.1953125,
"learning_rate": 3.2294806096344324e-05,
"loss": 6.4821,
"step": 5850
},
{
"epoch": 0.424319953657889,
"grad_norm": 1.1953125,
"learning_rate": 3.223870403998116e-05,
"loss": 6.4862,
"step": 5860
},
{
"epoch": 0.4250440491419469,
"grad_norm": 1.203125,
"learning_rate": 3.2182562166031606e-05,
"loss": 6.4912,
"step": 5870
},
{
"epoch": 0.4257681446260047,
"grad_norm": 1.1640625,
"learning_rate": 3.2126380783312525e-05,
"loss": 6.4966,
"step": 5880
},
{
"epoch": 0.42649224011006254,
"grad_norm": 1.234375,
"learning_rate": 3.2070160200858136e-05,
"loss": 6.4732,
"step": 5890
},
{
"epoch": 0.42721633559412037,
"grad_norm": 1.2265625,
"learning_rate": 3.2013900727918236e-05,
"loss": 6.4787,
"step": 5900
},
{
"epoch": 0.4279404310781782,
"grad_norm": 1.265625,
"learning_rate": 3.195760267395656e-05,
"loss": 6.4929,
"step": 5910
},
{
"epoch": 0.428664526562236,
"grad_norm": 1.21875,
"learning_rate": 3.190126634864908e-05,
"loss": 6.4908,
"step": 5920
},
{
"epoch": 0.42938862204629386,
"grad_norm": 1.1953125,
"learning_rate": 3.184489206188228e-05,
"loss": 6.4957,
"step": 5930
},
{
"epoch": 0.4301127175303517,
"grad_norm": 1.1953125,
"learning_rate": 3.1788480123751426e-05,
"loss": 6.4702,
"step": 5940
},
{
"epoch": 0.4308368130144095,
"grad_norm": 1.3125,
"learning_rate": 3.1732030844558936e-05,
"loss": 6.4674,
"step": 5950
},
{
"epoch": 0.43156090849846734,
"grad_norm": 1.296875,
"learning_rate": 3.167554453481261e-05,
"loss": 6.4738,
"step": 5960
},
{
"epoch": 0.4322850039825252,
"grad_norm": 1.21875,
"learning_rate": 3.1619021505223936e-05,
"loss": 6.4568,
"step": 5970
},
{
"epoch": 0.433009099466583,
"grad_norm": 1.203125,
"learning_rate": 3.156246206670637e-05,
"loss": 6.443,
"step": 5980
},
{
"epoch": 0.43373319495064083,
"grad_norm": 1.1953125,
"learning_rate": 3.150586653037367e-05,
"loss": 6.4858,
"step": 5990
},
{
"epoch": 0.43445729043469866,
"grad_norm": 1.1640625,
"learning_rate": 3.1449235207538155e-05,
"loss": 6.4682,
"step": 6000
},
{
"epoch": 0.43445729043469866,
"eval_loss": 6.624475955963135,
"eval_runtime": 1.4312,
"eval_samples_per_second": 8.385,
"eval_steps_per_second": 2.096,
"step": 6000
},
{
"epoch": 0.4351813859187565,
"grad_norm": 1.1796875,
"learning_rate": 3.139256840970897e-05,
"loss": 6.4852,
"step": 6010
},
{
"epoch": 0.4359054814028143,
"grad_norm": 1.1796875,
"learning_rate": 3.13358664485904e-05,
"loss": 6.4892,
"step": 6020
},
{
"epoch": 0.43662957688687215,
"grad_norm": 1.296875,
"learning_rate": 3.127912963608016e-05,
"loss": 6.4724,
"step": 6030
},
{
"epoch": 0.43735367237093,
"grad_norm": 1.265625,
"learning_rate": 3.1222358284267685e-05,
"loss": 6.4783,
"step": 6040
},
{
"epoch": 0.4380777678549878,
"grad_norm": 1.2265625,
"learning_rate": 3.116555270543236e-05,
"loss": 6.4949,
"step": 6050
},
{
"epoch": 0.43880186333904564,
"grad_norm": 1.2265625,
"learning_rate": 3.110871321204188e-05,
"loss": 6.458,
"step": 6060
},
{
"epoch": 0.43952595882310347,
"grad_norm": 1.2109375,
"learning_rate": 3.105184011675047e-05,
"loss": 6.4576,
"step": 6070
},
{
"epoch": 0.4402500543071613,
"grad_norm": 1.25,
"learning_rate": 3.099493373239719e-05,
"loss": 6.4929,
"step": 6080
},
{
"epoch": 0.4409741497912191,
"grad_norm": 1.2578125,
"learning_rate": 3.093799437200421e-05,
"loss": 6.462,
"step": 6090
},
{
"epoch": 0.44169824527527696,
"grad_norm": 1.1875,
"learning_rate": 3.0881022348775104e-05,
"loss": 6.4752,
"step": 6100
},
{
"epoch": 0.4424223407593348,
"grad_norm": 1.28125,
"learning_rate": 3.082401797609309e-05,
"loss": 6.4974,
"step": 6110
},
{
"epoch": 0.4431464362433926,
"grad_norm": 1.1640625,
"learning_rate": 3.0766981567519346e-05,
"loss": 6.4688,
"step": 6120
},
{
"epoch": 0.44387053172745045,
"grad_norm": 1.21875,
"learning_rate": 3.070991343679127e-05,
"loss": 6.4683,
"step": 6130
},
{
"epoch": 0.4445946272115083,
"grad_norm": 1.234375,
"learning_rate": 3.065281389782075e-05,
"loss": 6.4841,
"step": 6140
},
{
"epoch": 0.4453187226955661,
"grad_norm": 1.21875,
"learning_rate": 3.059568326469242e-05,
"loss": 6.4724,
"step": 6150
},
{
"epoch": 0.44604281817962393,
"grad_norm": 1.28125,
"learning_rate": 3.0538521851661986e-05,
"loss": 6.4475,
"step": 6160
},
{
"epoch": 0.44676691366368176,
"grad_norm": 1.25,
"learning_rate": 3.0481329973154427e-05,
"loss": 6.4788,
"step": 6170
},
{
"epoch": 0.4474910091477396,
"grad_norm": 1.1875,
"learning_rate": 3.0424107943762333e-05,
"loss": 6.4613,
"step": 6180
},
{
"epoch": 0.4482151046317974,
"grad_norm": 1.2265625,
"learning_rate": 3.036685607824412e-05,
"loss": 6.4948,
"step": 6190
},
{
"epoch": 0.44893920011585525,
"grad_norm": 1.2109375,
"learning_rate": 3.030957469152235e-05,
"loss": 6.4943,
"step": 6200
},
{
"epoch": 0.4496632955999131,
"grad_norm": 1.25,
"learning_rate": 3.0252264098681947e-05,
"loss": 6.4713,
"step": 6210
},
{
"epoch": 0.45038739108397097,
"grad_norm": 1.25,
"learning_rate": 3.0194924614968495e-05,
"loss": 6.4844,
"step": 6220
},
{
"epoch": 0.4511114865680288,
"grad_norm": 1.140625,
"learning_rate": 3.013755655578649e-05,
"loss": 6.4724,
"step": 6230
},
{
"epoch": 0.4518355820520866,
"grad_norm": 1.25,
"learning_rate": 3.008016023669763e-05,
"loss": 6.4964,
"step": 6240
},
{
"epoch": 0.45255967753614446,
"grad_norm": 1.203125,
"learning_rate": 3.002273597341903e-05,
"loss": 6.4579,
"step": 6250
},
{
"epoch": 0.4532837730202023,
"grad_norm": 1.1640625,
"learning_rate": 2.996528408182156e-05,
"loss": 6.4774,
"step": 6260
},
{
"epoch": 0.4540078685042601,
"grad_norm": 1.1953125,
"learning_rate": 2.990780487792802e-05,
"loss": 6.485,
"step": 6270
},
{
"epoch": 0.45473196398831794,
"grad_norm": 1.21875,
"learning_rate": 2.9850298677911475e-05,
"loss": 6.4758,
"step": 6280
},
{
"epoch": 0.4554560594723758,
"grad_norm": 1.2578125,
"learning_rate": 2.9792765798093465e-05,
"loss": 6.479,
"step": 6290
},
{
"epoch": 0.4561801549564336,
"grad_norm": 1.2421875,
"learning_rate": 2.97352065549423e-05,
"loss": 6.4656,
"step": 6300
},
{
"epoch": 0.45690425044049143,
"grad_norm": 1.2109375,
"learning_rate": 2.9677621265071313e-05,
"loss": 6.4761,
"step": 6310
},
{
"epoch": 0.45762834592454926,
"grad_norm": 1.2109375,
"learning_rate": 2.9620010245237094e-05,
"loss": 6.4661,
"step": 6320
},
{
"epoch": 0.4583524414086071,
"grad_norm": 1.2890625,
"learning_rate": 2.956237381233777e-05,
"loss": 6.467,
"step": 6330
},
{
"epoch": 0.4590765368926649,
"grad_norm": 1.1953125,
"learning_rate": 2.9504712283411256e-05,
"loss": 6.4593,
"step": 6340
},
{
"epoch": 0.45980063237672275,
"grad_norm": 1.203125,
"learning_rate": 2.9447025975633517e-05,
"loss": 6.4832,
"step": 6350
},
{
"epoch": 0.4605247278607806,
"grad_norm": 1.2109375,
"learning_rate": 2.938931520631681e-05,
"loss": 6.4839,
"step": 6360
},
{
"epoch": 0.4612488233448384,
"grad_norm": 1.2421875,
"learning_rate": 2.933158029290795e-05,
"loss": 6.4612,
"step": 6370
},
{
"epoch": 0.46197291882889624,
"grad_norm": 1.203125,
"learning_rate": 2.927382155298657e-05,
"loss": 6.4639,
"step": 6380
},
{
"epoch": 0.46269701431295407,
"grad_norm": 1.2578125,
"learning_rate": 2.921603930426333e-05,
"loss": 6.4811,
"step": 6390
},
{
"epoch": 0.4634211097970119,
"grad_norm": 1.21875,
"learning_rate": 2.9158233864578254e-05,
"loss": 6.4671,
"step": 6400
},
{
"epoch": 0.46414520528106973,
"grad_norm": 1.2421875,
"learning_rate": 2.9100405551898896e-05,
"loss": 6.4636,
"step": 6410
},
{
"epoch": 0.46486930076512756,
"grad_norm": 1.1953125,
"learning_rate": 2.9042554684318636e-05,
"loss": 6.4711,
"step": 6420
},
{
"epoch": 0.4655933962491854,
"grad_norm": 1.125,
"learning_rate": 2.8984681580054917e-05,
"loss": 6.4705,
"step": 6430
},
{
"epoch": 0.4663174917332432,
"grad_norm": 1.28125,
"learning_rate": 2.892678655744751e-05,
"loss": 6.4933,
"step": 6440
},
{
"epoch": 0.46704158721730105,
"grad_norm": 1.1796875,
"learning_rate": 2.886886993495674e-05,
"loss": 6.4649,
"step": 6450
},
{
"epoch": 0.4677656827013589,
"grad_norm": 1.1640625,
"learning_rate": 2.8810932031161742e-05,
"loss": 6.4719,
"step": 6460
},
{
"epoch": 0.4684897781854167,
"grad_norm": 1.1796875,
"learning_rate": 2.875297316475873e-05,
"loss": 6.4689,
"step": 6470
},
{
"epoch": 0.46921387366947453,
"grad_norm": 1.21875,
"learning_rate": 2.869499365455921e-05,
"loss": 6.4806,
"step": 6480
},
{
"epoch": 0.46993796915353236,
"grad_norm": 1.2109375,
"learning_rate": 2.863699381948824e-05,
"loss": 6.4796,
"step": 6490
},
{
"epoch": 0.4706620646375902,
"grad_norm": 1.171875,
"learning_rate": 2.8578973978582695e-05,
"loss": 6.4612,
"step": 6500
},
{
"epoch": 0.471386160121648,
"grad_norm": 1.2109375,
"learning_rate": 2.8520934450989485e-05,
"loss": 6.4452,
"step": 6510
},
{
"epoch": 0.47211025560570585,
"grad_norm": 1.1640625,
"learning_rate": 2.8462875555963797e-05,
"loss": 6.4682,
"step": 6520
},
{
"epoch": 0.4728343510897637,
"grad_norm": 1.1875,
"learning_rate": 2.840479761286738e-05,
"loss": 6.4683,
"step": 6530
},
{
"epoch": 0.4735584465738215,
"grad_norm": 1.171875,
"learning_rate": 2.8346700941166725e-05,
"loss": 6.4716,
"step": 6540
},
{
"epoch": 0.47428254205787934,
"grad_norm": 1.1640625,
"learning_rate": 2.8288585860431387e-05,
"loss": 6.4644,
"step": 6550
},
{
"epoch": 0.47500663754193717,
"grad_norm": 1.2421875,
"learning_rate": 2.8230452690332132e-05,
"loss": 6.4855,
"step": 6560
},
{
"epoch": 0.475730733025995,
"grad_norm": 1.2109375,
"learning_rate": 2.8172301750639274e-05,
"loss": 6.4506,
"step": 6570
},
{
"epoch": 0.4764548285100529,
"grad_norm": 1.171875,
"learning_rate": 2.8114133361220845e-05,
"loss": 6.4714,
"step": 6580
},
{
"epoch": 0.4771789239941107,
"grad_norm": 1.140625,
"learning_rate": 2.8055947842040862e-05,
"loss": 6.4555,
"step": 6590
},
{
"epoch": 0.47790301947816854,
"grad_norm": 1.21875,
"learning_rate": 2.7997745513157582e-05,
"loss": 6.4689,
"step": 6600
},
{
"epoch": 0.4786271149622264,
"grad_norm": 1.2265625,
"learning_rate": 2.793952669472172e-05,
"loss": 6.4681,
"step": 6610
},
{
"epoch": 0.4793512104462842,
"grad_norm": 1.234375,
"learning_rate": 2.7881291706974676e-05,
"loss": 6.4709,
"step": 6620
},
{
"epoch": 0.48007530593034203,
"grad_norm": 1.2109375,
"learning_rate": 2.782304087024682e-05,
"loss": 6.4703,
"step": 6630
},
{
"epoch": 0.48079940141439986,
"grad_norm": 1.21875,
"learning_rate": 2.7764774504955694e-05,
"loss": 6.4614,
"step": 6640
},
{
"epoch": 0.4815234968984577,
"grad_norm": 1.21875,
"learning_rate": 2.7706492931604238e-05,
"loss": 6.4632,
"step": 6650
},
{
"epoch": 0.4822475923825155,
"grad_norm": 1.2578125,
"learning_rate": 2.7648196470779075e-05,
"loss": 6.4594,
"step": 6660
},
{
"epoch": 0.48297168786657335,
"grad_norm": 1.21875,
"learning_rate": 2.75898854431487e-05,
"loss": 6.4725,
"step": 6670
},
{
"epoch": 0.4836957833506312,
"grad_norm": 1.2109375,
"learning_rate": 2.753156016946174e-05,
"loss": 6.4721,
"step": 6680
},
{
"epoch": 0.484419878834689,
"grad_norm": 1.25,
"learning_rate": 2.7473220970545182e-05,
"loss": 6.4438,
"step": 6690
},
{
"epoch": 0.48514397431874684,
"grad_norm": 1.140625,
"learning_rate": 2.7414868167302614e-05,
"loss": 6.4764,
"step": 6700
},
{
"epoch": 0.48586806980280467,
"grad_norm": 1.234375,
"learning_rate": 2.735650208071246e-05,
"loss": 6.4539,
"step": 6710
},
{
"epoch": 0.4865921652868625,
"grad_norm": 1.1953125,
"learning_rate": 2.72981230318262e-05,
"loss": 6.4654,
"step": 6720
},
{
"epoch": 0.48731626077092033,
"grad_norm": 1.25,
"learning_rate": 2.723973134176662e-05,
"loss": 6.4338,
"step": 6730
},
{
"epoch": 0.48804035625497816,
"grad_norm": 1.140625,
"learning_rate": 2.7181327331726058e-05,
"loss": 6.4534,
"step": 6740
},
{
"epoch": 0.488764451739036,
"grad_norm": 1.265625,
"learning_rate": 2.7122911322964594e-05,
"loss": 6.4736,
"step": 6750
},
{
"epoch": 0.4894885472230938,
"grad_norm": 1.2421875,
"learning_rate": 2.7064483636808313e-05,
"loss": 6.4707,
"step": 6760
},
{
"epoch": 0.49021264270715165,
"grad_norm": 1.1796875,
"learning_rate": 2.7006044594647543e-05,
"loss": 6.4474,
"step": 6770
},
{
"epoch": 0.4909367381912095,
"grad_norm": 1.15625,
"learning_rate": 2.6947594517935083e-05,
"loss": 6.4672,
"step": 6780
},
{
"epoch": 0.4916608336752673,
"grad_norm": 1.171875,
"learning_rate": 2.688913372818442e-05,
"loss": 6.4718,
"step": 6790
},
{
"epoch": 0.49238492915932514,
"grad_norm": 1.21875,
"learning_rate": 2.683066254696795e-05,
"loss": 6.4976,
"step": 6800
},
{
"epoch": 0.49310902464338296,
"grad_norm": 1.2890625,
"learning_rate": 2.6772181295915272e-05,
"loss": 6.4536,
"step": 6810
},
{
"epoch": 0.4938331201274408,
"grad_norm": 1.1796875,
"learning_rate": 2.6713690296711346e-05,
"loss": 6.4684,
"step": 6820
},
{
"epoch": 0.4945572156114986,
"grad_norm": 1.171875,
"learning_rate": 2.665518987109475e-05,
"loss": 6.4794,
"step": 6830
},
{
"epoch": 0.49528131109555645,
"grad_norm": 1.2265625,
"learning_rate": 2.6596680340855935e-05,
"loss": 6.477,
"step": 6840
},
{
"epoch": 0.4960054065796143,
"grad_norm": 1.2109375,
"learning_rate": 2.6538162027835416e-05,
"loss": 6.4537,
"step": 6850
},
{
"epoch": 0.4967295020636721,
"grad_norm": 1.171875,
"learning_rate": 2.6479635253922018e-05,
"loss": 6.4503,
"step": 6860
},
{
"epoch": 0.49745359754772994,
"grad_norm": 1.109375,
"learning_rate": 2.6421100341051125e-05,
"loss": 6.4848,
"step": 6870
},
{
"epoch": 0.49817769303178777,
"grad_norm": 1.21875,
"learning_rate": 2.636255761120287e-05,
"loss": 6.4634,
"step": 6880
},
{
"epoch": 0.4989017885158456,
"grad_norm": 1.2421875,
"learning_rate": 2.6304007386400398e-05,
"loss": 6.4821,
"step": 6890
},
{
"epoch": 0.49962588399990343,
"grad_norm": 1.1953125,
"learning_rate": 2.6245449988708066e-05,
"loss": 6.4534,
"step": 6900
},
{
"epoch": 0.5003499794839613,
"grad_norm": 1.265625,
"learning_rate": 2.618688574022971e-05,
"loss": 6.4562,
"step": 6910
},
{
"epoch": 0.5010740749680191,
"grad_norm": 1.234375,
"learning_rate": 2.6128314963106832e-05,
"loss": 6.4691,
"step": 6920
},
{
"epoch": 0.501798170452077,
"grad_norm": 1.1875,
"learning_rate": 2.6069737979516845e-05,
"loss": 6.4536,
"step": 6930
},
{
"epoch": 0.5025222659361348,
"grad_norm": 1.1796875,
"learning_rate": 2.601115511167131e-05,
"loss": 6.4474,
"step": 6940
},
{
"epoch": 0.5032463614201926,
"grad_norm": 1.2734375,
"learning_rate": 2.5952566681814156e-05,
"loss": 6.4382,
"step": 6950
},
{
"epoch": 0.5039704569042505,
"grad_norm": 1.2265625,
"learning_rate": 2.589397301221989e-05,
"loss": 6.465,
"step": 6960
},
{
"epoch": 0.5046945523883083,
"grad_norm": 1.234375,
"learning_rate": 2.5835374425191866e-05,
"loss": 6.4716,
"step": 6970
},
{
"epoch": 0.5054186478723661,
"grad_norm": 1.1875,
"learning_rate": 2.5776771243060472e-05,
"loss": 6.4443,
"step": 6980
},
{
"epoch": 0.506142743356424,
"grad_norm": 1.265625,
"learning_rate": 2.5718163788181375e-05,
"loss": 6.4442,
"step": 6990
},
{
"epoch": 0.5068668388404818,
"grad_norm": 1.2421875,
"learning_rate": 2.5659552382933732e-05,
"loss": 6.4818,
"step": 7000
},
{
"epoch": 0.5068668388404818,
"eval_loss": 6.613395690917969,
"eval_runtime": 1.5015,
"eval_samples_per_second": 7.992,
"eval_steps_per_second": 1.998,
"step": 7000
},
{
"epoch": 0.5075909343245396,
"grad_norm": 1.21875,
"learning_rate": 2.5600937349718462e-05,
"loss": 6.4476,
"step": 7010
},
{
"epoch": 0.5083150298085974,
"grad_norm": 1.1953125,
"learning_rate": 2.554231901095641e-05,
"loss": 6.459,
"step": 7020
},
{
"epoch": 0.5090391252926553,
"grad_norm": 1.25,
"learning_rate": 2.5483697689086615e-05,
"loss": 6.4414,
"step": 7030
},
{
"epoch": 0.5097632207767131,
"grad_norm": 1.1796875,
"learning_rate": 2.5425073706564528e-05,
"loss": 6.4691,
"step": 7040
},
{
"epoch": 0.5104873162607709,
"grad_norm": 1.2421875,
"learning_rate": 2.5366447385860225e-05,
"loss": 6.4414,
"step": 7050
},
{
"epoch": 0.5112114117448288,
"grad_norm": 1.21875,
"learning_rate": 2.5307819049456655e-05,
"loss": 6.478,
"step": 7060
},
{
"epoch": 0.5119355072288866,
"grad_norm": 1.2109375,
"learning_rate": 2.524918901984784e-05,
"loss": 6.4605,
"step": 7070
},
{
"epoch": 0.5126596027129444,
"grad_norm": 1.2421875,
"learning_rate": 2.519055761953714e-05,
"loss": 6.4395,
"step": 7080
},
{
"epoch": 0.5133836981970022,
"grad_norm": 1.2109375,
"learning_rate": 2.5131925171035438e-05,
"loss": 6.4684,
"step": 7090
},
{
"epoch": 0.5141077936810601,
"grad_norm": 1.2265625,
"learning_rate": 2.5073291996859372e-05,
"loss": 6.4657,
"step": 7100
},
{
"epoch": 0.5148318891651179,
"grad_norm": 1.2109375,
"learning_rate": 2.501465841952959e-05,
"loss": 6.4963,
"step": 7110
},
{
"epoch": 0.5155559846491757,
"grad_norm": 1.1875,
"learning_rate": 2.4956024761568965e-05,
"loss": 6.4635,
"step": 7120
},
{
"epoch": 0.5162800801332336,
"grad_norm": 1.2890625,
"learning_rate": 2.4897391345500777e-05,
"loss": 6.4686,
"step": 7130
},
{
"epoch": 0.5170041756172914,
"grad_norm": 1.1875,
"learning_rate": 2.483875849384702e-05,
"loss": 6.4515,
"step": 7140
},
{
"epoch": 0.5177282711013492,
"grad_norm": 1.1953125,
"learning_rate": 2.4780126529126555e-05,
"loss": 6.4655,
"step": 7150
},
{
"epoch": 0.518452366585407,
"grad_norm": 1.2109375,
"learning_rate": 2.4721495773853366e-05,
"loss": 6.457,
"step": 7160
},
{
"epoch": 0.5191764620694649,
"grad_norm": 1.1796875,
"learning_rate": 2.4662866550534798e-05,
"loss": 6.4679,
"step": 7170
},
{
"epoch": 0.5199005575535227,
"grad_norm": 1.25,
"learning_rate": 2.460423918166975e-05,
"loss": 6.4634,
"step": 7180
},
{
"epoch": 0.5206246530375805,
"grad_norm": 1.15625,
"learning_rate": 2.4545613989746953e-05,
"loss": 6.4335,
"step": 7190
},
{
"epoch": 0.5213487485216384,
"grad_norm": 1.2421875,
"learning_rate": 2.4486991297243123e-05,
"loss": 6.4535,
"step": 7200
},
{
"epoch": 0.5220728440056962,
"grad_norm": 1.171875,
"learning_rate": 2.4428371426621248e-05,
"loss": 6.4829,
"step": 7210
},
{
"epoch": 0.522796939489754,
"grad_norm": 1.2265625,
"learning_rate": 2.4369754700328805e-05,
"loss": 6.4657,
"step": 7220
},
{
"epoch": 0.5235210349738119,
"grad_norm": 1.1875,
"learning_rate": 2.4311141440795953e-05,
"loss": 6.4412,
"step": 7230
},
{
"epoch": 0.5242451304578697,
"grad_norm": 1.1796875,
"learning_rate": 2.4252531970433786e-05,
"loss": 6.4693,
"step": 7240
},
{
"epoch": 0.5249692259419275,
"grad_norm": 1.1640625,
"learning_rate": 2.419392661163257e-05,
"loss": 6.46,
"step": 7250
},
{
"epoch": 0.5256933214259853,
"grad_norm": 1.171875,
"learning_rate": 2.4135325686759942e-05,
"loss": 6.4613,
"step": 7260
},
{
"epoch": 0.5264174169100432,
"grad_norm": 1.1796875,
"learning_rate": 2.4076729518159146e-05,
"loss": 6.463,
"step": 7270
},
{
"epoch": 0.527141512394101,
"grad_norm": 1.265625,
"learning_rate": 2.401813842814728e-05,
"loss": 6.4323,
"step": 7280
},
{
"epoch": 0.5278656078781588,
"grad_norm": 1.203125,
"learning_rate": 2.395955273901349e-05,
"loss": 6.4312,
"step": 7290
},
{
"epoch": 0.5285897033622167,
"grad_norm": 1.2109375,
"learning_rate": 2.390097277301722e-05,
"loss": 6.4483,
"step": 7300
},
{
"epoch": 0.5293137988462745,
"grad_norm": 1.1796875,
"learning_rate": 2.3842398852386434e-05,
"loss": 6.468,
"step": 7310
},
{
"epoch": 0.5300378943303323,
"grad_norm": 1.234375,
"learning_rate": 2.3783831299315845e-05,
"loss": 6.4574,
"step": 7320
},
{
"epoch": 0.5307619898143902,
"grad_norm": 1.21875,
"learning_rate": 2.3725270435965123e-05,
"loss": 6.4518,
"step": 7330
},
{
"epoch": 0.531486085298448,
"grad_norm": 1.2265625,
"learning_rate": 2.3666716584457174e-05,
"loss": 6.4598,
"step": 7340
},
{
"epoch": 0.5322101807825058,
"grad_norm": 1.21875,
"learning_rate": 2.36081700668763e-05,
"loss": 6.4511,
"step": 7350
},
{
"epoch": 0.5329342762665636,
"grad_norm": 1.328125,
"learning_rate": 2.3549631205266482e-05,
"loss": 6.477,
"step": 7360
},
{
"epoch": 0.5336583717506215,
"grad_norm": 1.1484375,
"learning_rate": 2.3491100321629573e-05,
"loss": 6.4593,
"step": 7370
},
{
"epoch": 0.5343824672346793,
"grad_norm": 1.203125,
"learning_rate": 2.3432577737923564e-05,
"loss": 6.4507,
"step": 7380
},
{
"epoch": 0.5351065627187371,
"grad_norm": 1.1640625,
"learning_rate": 2.337406377606077e-05,
"loss": 6.4771,
"step": 7390
},
{
"epoch": 0.535830658202795,
"grad_norm": 1.2265625,
"learning_rate": 2.3315558757906087e-05,
"loss": 6.4647,
"step": 7400
},
{
"epoch": 0.5365547536868528,
"grad_norm": 1.1640625,
"learning_rate": 2.325706300527521e-05,
"loss": 6.4545,
"step": 7410
},
{
"epoch": 0.5372788491709106,
"grad_norm": 1.21875,
"learning_rate": 2.3198576839932893e-05,
"loss": 6.4633,
"step": 7420
},
{
"epoch": 0.5380029446549685,
"grad_norm": 1.203125,
"learning_rate": 2.3140100583591124e-05,
"loss": 6.4553,
"step": 7430
},
{
"epoch": 0.5387270401390263,
"grad_norm": 1.171875,
"learning_rate": 2.3081634557907388e-05,
"loss": 6.4596,
"step": 7440
},
{
"epoch": 0.5394511356230841,
"grad_norm": 1.2109375,
"learning_rate": 2.3023179084482916e-05,
"loss": 6.4385,
"step": 7450
},
{
"epoch": 0.5401752311071419,
"grad_norm": 1.21875,
"learning_rate": 2.296473448486087e-05,
"loss": 6.4597,
"step": 7460
},
{
"epoch": 0.5408993265911999,
"grad_norm": 1.1328125,
"learning_rate": 2.2906301080524614e-05,
"loss": 6.4686,
"step": 7470
},
{
"epoch": 0.5416234220752577,
"grad_norm": 1.1484375,
"learning_rate": 2.2847879192895922e-05,
"loss": 6.4299,
"step": 7480
},
{
"epoch": 0.5423475175593155,
"grad_norm": 1.15625,
"learning_rate": 2.278946914333323e-05,
"loss": 6.4546,
"step": 7490
},
{
"epoch": 0.5430716130433734,
"grad_norm": 1.1484375,
"learning_rate": 2.2731071253129844e-05,
"loss": 6.4454,
"step": 7500
},
{
"epoch": 0.5437957085274312,
"grad_norm": 1.234375,
"learning_rate": 2.2672685843512187e-05,
"loss": 6.471,
"step": 7510
},
{
"epoch": 0.544519804011489,
"grad_norm": 1.234375,
"learning_rate": 2.2614313235638047e-05,
"loss": 6.4412,
"step": 7520
},
{
"epoch": 0.5452438994955469,
"grad_norm": 1.2109375,
"learning_rate": 2.2555953750594777e-05,
"loss": 6.4591,
"step": 7530
},
{
"epoch": 0.5459679949796047,
"grad_norm": 1.1484375,
"learning_rate": 2.2497607709397543e-05,
"loss": 6.4524,
"step": 7540
},
{
"epoch": 0.5466920904636625,
"grad_norm": 1.2265625,
"learning_rate": 2.2439275432987585e-05,
"loss": 6.4796,
"step": 7550
},
{
"epoch": 0.5474161859477203,
"grad_norm": 1.2734375,
"learning_rate": 2.23809572422304e-05,
"loss": 6.4641,
"step": 7560
},
{
"epoch": 0.5481402814317782,
"grad_norm": 1.125,
"learning_rate": 2.2322653457914014e-05,
"loss": 6.4564,
"step": 7570
},
{
"epoch": 0.548864376915836,
"grad_norm": 1.1796875,
"learning_rate": 2.2264364400747226e-05,
"loss": 6.4588,
"step": 7580
},
{
"epoch": 0.5495884723998938,
"grad_norm": 1.296875,
"learning_rate": 2.2206090391357797e-05,
"loss": 6.4472,
"step": 7590
},
{
"epoch": 0.5503125678839517,
"grad_norm": 1.21875,
"learning_rate": 2.2147831750290732e-05,
"loss": 6.4609,
"step": 7600
},
{
"epoch": 0.5510366633680095,
"grad_norm": 1.171875,
"learning_rate": 2.20895887980065e-05,
"loss": 6.4575,
"step": 7610
},
{
"epoch": 0.5517607588520673,
"grad_norm": 1.140625,
"learning_rate": 2.203136185487927e-05,
"loss": 6.4575,
"step": 7620
},
{
"epoch": 0.5524848543361252,
"grad_norm": 1.1875,
"learning_rate": 2.1973151241195143e-05,
"loss": 6.4774,
"step": 7630
},
{
"epoch": 0.553208949820183,
"grad_norm": 1.2421875,
"learning_rate": 2.191495727715041e-05,
"loss": 6.4593,
"step": 7640
},
{
"epoch": 0.5539330453042408,
"grad_norm": 1.1875,
"learning_rate": 2.1856780282849765e-05,
"loss": 6.4552,
"step": 7650
},
{
"epoch": 0.5546571407882986,
"grad_norm": 1.1640625,
"learning_rate": 2.1798620578304575e-05,
"loss": 6.4388,
"step": 7660
},
{
"epoch": 0.5553812362723565,
"grad_norm": 1.171875,
"learning_rate": 2.1740478483431088e-05,
"loss": 6.4495,
"step": 7670
},
{
"epoch": 0.5561053317564143,
"grad_norm": 1.1640625,
"learning_rate": 2.1682354318048676e-05,
"loss": 6.4629,
"step": 7680
},
{
"epoch": 0.5568294272404721,
"grad_norm": 1.2265625,
"learning_rate": 2.1624248401878123e-05,
"loss": 6.4558,
"step": 7690
},
{
"epoch": 0.55755352272453,
"grad_norm": 1.1328125,
"learning_rate": 2.1566161054539798e-05,
"loss": 6.4757,
"step": 7700
},
{
"epoch": 0.5582776182085878,
"grad_norm": 1.2578125,
"learning_rate": 2.1508092595551938e-05,
"loss": 6.4555,
"step": 7710
},
{
"epoch": 0.5590017136926456,
"grad_norm": 1.1953125,
"learning_rate": 2.1450043344328892e-05,
"loss": 6.4687,
"step": 7720
},
{
"epoch": 0.5597258091767034,
"grad_norm": 1.171875,
"learning_rate": 2.1392013620179337e-05,
"loss": 6.4549,
"step": 7730
},
{
"epoch": 0.5604499046607613,
"grad_norm": 1.1875,
"learning_rate": 2.133400374230455e-05,
"loss": 6.4584,
"step": 7740
},
{
"epoch": 0.5611740001448191,
"grad_norm": 1.15625,
"learning_rate": 2.127601402979665e-05,
"loss": 6.4676,
"step": 7750
},
{
"epoch": 0.5618980956288769,
"grad_norm": 1.21875,
"learning_rate": 2.1218044801636805e-05,
"loss": 6.4536,
"step": 7760
},
{
"epoch": 0.5626221911129348,
"grad_norm": 1.1953125,
"learning_rate": 2.116009637669353e-05,
"loss": 6.4334,
"step": 7770
},
{
"epoch": 0.5633462865969926,
"grad_norm": 1.1953125,
"learning_rate": 2.1102169073720894e-05,
"loss": 6.4507,
"step": 7780
},
{
"epoch": 0.5640703820810504,
"grad_norm": 1.203125,
"learning_rate": 2.1044263211356797e-05,
"loss": 6.436,
"step": 7790
},
{
"epoch": 0.5647944775651083,
"grad_norm": 1.1328125,
"learning_rate": 2.098637910812119e-05,
"loss": 6.4561,
"step": 7800
},
{
"epoch": 0.5655185730491661,
"grad_norm": 1.1953125,
"learning_rate": 2.0928517082414335e-05,
"loss": 6.454,
"step": 7810
},
{
"epoch": 0.5662426685332239,
"grad_norm": 1.1796875,
"learning_rate": 2.0870677452515057e-05,
"loss": 6.4499,
"step": 7820
},
{
"epoch": 0.5669667640172817,
"grad_norm": 1.2265625,
"learning_rate": 2.081286053657899e-05,
"loss": 6.4552,
"step": 7830
},
{
"epoch": 0.5676908595013396,
"grad_norm": 1.203125,
"learning_rate": 2.0755066652636817e-05,
"loss": 6.4605,
"step": 7840
},
{
"epoch": 0.5684149549853974,
"grad_norm": 1.203125,
"learning_rate": 2.0697296118592553e-05,
"loss": 6.4492,
"step": 7850
},
{
"epoch": 0.5691390504694552,
"grad_norm": 1.2734375,
"learning_rate": 2.0639549252221745e-05,
"loss": 6.4476,
"step": 7860
},
{
"epoch": 0.5698631459535131,
"grad_norm": 1.1953125,
"learning_rate": 2.058182637116976e-05,
"loss": 6.4568,
"step": 7870
},
{
"epoch": 0.5705872414375709,
"grad_norm": 1.2734375,
"learning_rate": 2.0524127792950056e-05,
"loss": 6.4512,
"step": 7880
},
{
"epoch": 0.5713113369216287,
"grad_norm": 1.1875,
"learning_rate": 2.046645383494237e-05,
"loss": 6.4383,
"step": 7890
},
{
"epoch": 0.5720354324056866,
"grad_norm": 1.2265625,
"learning_rate": 2.0408804814391043e-05,
"loss": 6.4694,
"step": 7900
},
{
"epoch": 0.5727595278897444,
"grad_norm": 1.1640625,
"learning_rate": 2.0351181048403227e-05,
"loss": 6.4535,
"step": 7910
},
{
"epoch": 0.5734836233738022,
"grad_norm": 1.2109375,
"learning_rate": 2.0293582853947164e-05,
"loss": 6.4736,
"step": 7920
},
{
"epoch": 0.57420771885786,
"grad_norm": 1.203125,
"learning_rate": 2.0236010547850432e-05,
"loss": 6.4463,
"step": 7930
},
{
"epoch": 0.5749318143419179,
"grad_norm": 1.2421875,
"learning_rate": 2.0178464446798206e-05,
"loss": 6.462,
"step": 7940
},
{
"epoch": 0.5756559098259757,
"grad_norm": 1.2109375,
"learning_rate": 2.0120944867331522e-05,
"loss": 6.4542,
"step": 7950
},
{
"epoch": 0.5763800053100335,
"grad_norm": 1.171875,
"learning_rate": 2.0063452125845522e-05,
"loss": 6.4458,
"step": 7960
},
{
"epoch": 0.5771041007940914,
"grad_norm": 1.1484375,
"learning_rate": 2.0005986538587735e-05,
"loss": 6.4731,
"step": 7970
},
{
"epoch": 0.5778281962781492,
"grad_norm": 1.1328125,
"learning_rate": 1.9948548421656303e-05,
"loss": 6.4686,
"step": 7980
},
{
"epoch": 0.578552291762207,
"grad_norm": 1.1953125,
"learning_rate": 1.9891138090998285e-05,
"loss": 6.4521,
"step": 7990
},
{
"epoch": 0.5792763872462648,
"grad_norm": 1.1875,
"learning_rate": 1.983375586240789e-05,
"loss": 6.4232,
"step": 8000
},
{
"epoch": 0.5792763872462648,
"eval_loss": 6.610226154327393,
"eval_runtime": 1.4712,
"eval_samples_per_second": 8.157,
"eval_steps_per_second": 2.039,
"step": 8000
},
{
"epoch": 0.5800004827303227,
"grad_norm": 1.25,
"learning_rate": 1.9776402051524724e-05,
"loss": 6.4368,
"step": 8010
},
{
"epoch": 0.5807245782143805,
"grad_norm": 1.1875,
"learning_rate": 1.971907697383213e-05,
"loss": 6.4723,
"step": 8020
},
{
"epoch": 0.5814486736984383,
"grad_norm": 1.1953125,
"learning_rate": 1.9661780944655345e-05,
"loss": 6.4643,
"step": 8030
},
{
"epoch": 0.5821727691824962,
"grad_norm": 1.1796875,
"learning_rate": 1.9604514279159843e-05,
"loss": 6.4756,
"step": 8040
},
{
"epoch": 0.582896864666554,
"grad_norm": 1.140625,
"learning_rate": 1.95472772923496e-05,
"loss": 6.4411,
"step": 8050
},
{
"epoch": 0.5836209601506118,
"grad_norm": 1.25,
"learning_rate": 1.9490070299065293e-05,
"loss": 6.4344,
"step": 8060
},
{
"epoch": 0.5843450556346697,
"grad_norm": 1.1953125,
"learning_rate": 1.943289361398266e-05,
"loss": 6.4451,
"step": 8070
},
{
"epoch": 0.5850691511187275,
"grad_norm": 1.25,
"learning_rate": 1.9375747551610688e-05,
"loss": 6.4609,
"step": 8080
},
{
"epoch": 0.5857932466027853,
"grad_norm": 1.2421875,
"learning_rate": 1.931863242628995e-05,
"loss": 6.4627,
"step": 8090
},
{
"epoch": 0.5865173420868431,
"grad_norm": 1.1484375,
"learning_rate": 1.9261548552190827e-05,
"loss": 6.4676,
"step": 8100
},
{
"epoch": 0.587241437570901,
"grad_norm": 1.1875,
"learning_rate": 1.920449624331179e-05,
"loss": 6.457,
"step": 8110
},
{
"epoch": 0.5879655330549588,
"grad_norm": 1.140625,
"learning_rate": 1.9147475813477718e-05,
"loss": 6.4659,
"step": 8120
},
{
"epoch": 0.5886896285390166,
"grad_norm": 1.21875,
"learning_rate": 1.909048757633809e-05,
"loss": 6.4445,
"step": 8130
},
{
"epoch": 0.5894137240230745,
"grad_norm": 1.265625,
"learning_rate": 1.9033531845365334e-05,
"loss": 6.4637,
"step": 8140
},
{
"epoch": 0.5901378195071323,
"grad_norm": 1.171875,
"learning_rate": 1.897660893385305e-05,
"loss": 6.4351,
"step": 8150
},
{
"epoch": 0.5908619149911901,
"grad_norm": 1.1953125,
"learning_rate": 1.891971915491434e-05,
"loss": 6.4607,
"step": 8160
},
{
"epoch": 0.591586010475248,
"grad_norm": 1.1796875,
"learning_rate": 1.8862862821480025e-05,
"loss": 6.4635,
"step": 8170
},
{
"epoch": 0.5923101059593059,
"grad_norm": 1.2578125,
"learning_rate": 1.8806040246296966e-05,
"loss": 6.4431,
"step": 8180
},
{
"epoch": 0.5930342014433637,
"grad_norm": 1.1796875,
"learning_rate": 1.8749251741926326e-05,
"loss": 6.4586,
"step": 8190
},
{
"epoch": 0.5937582969274215,
"grad_norm": 1.1484375,
"learning_rate": 1.8692497620741884e-05,
"loss": 6.4613,
"step": 8200
},
{
"epoch": 0.5944823924114794,
"grad_norm": 1.2734375,
"learning_rate": 1.8635778194928246e-05,
"loss": 6.4493,
"step": 8210
},
{
"epoch": 0.5952064878955372,
"grad_norm": 1.203125,
"learning_rate": 1.85790937764792e-05,
"loss": 6.4421,
"step": 8220
},
{
"epoch": 0.595930583379595,
"grad_norm": 1.2109375,
"learning_rate": 1.8522444677195965e-05,
"loss": 6.4376,
"step": 8230
},
{
"epoch": 0.5966546788636529,
"grad_norm": 1.171875,
"learning_rate": 1.8465831208685476e-05,
"loss": 6.4449,
"step": 8240
},
{
"epoch": 0.5973787743477107,
"grad_norm": 1.1796875,
"learning_rate": 1.8409253682358678e-05,
"loss": 6.4748,
"step": 8250
},
{
"epoch": 0.5981028698317685,
"grad_norm": 1.2265625,
"learning_rate": 1.835271240942882e-05,
"loss": 6.4677,
"step": 8260
},
{
"epoch": 0.5988269653158264,
"grad_norm": 1.1953125,
"learning_rate": 1.8296207700909717e-05,
"loss": 6.4513,
"step": 8270
},
{
"epoch": 0.5995510607998842,
"grad_norm": 1.1328125,
"learning_rate": 1.823973986761406e-05,
"loss": 6.4442,
"step": 8280
},
{
"epoch": 0.600275156283942,
"grad_norm": 1.09375,
"learning_rate": 1.8183309220151717e-05,
"loss": 6.4554,
"step": 8290
},
{
"epoch": 0.6009992517679998,
"grad_norm": 1.1484375,
"learning_rate": 1.8126916068928e-05,
"loss": 6.4365,
"step": 8300
},
{
"epoch": 0.6017233472520577,
"grad_norm": 1.2109375,
"learning_rate": 1.807056072414195e-05,
"loss": 6.441,
"step": 8310
},
{
"epoch": 0.6024474427361155,
"grad_norm": 1.203125,
"learning_rate": 1.801424349578469e-05,
"loss": 6.4688,
"step": 8320
},
{
"epoch": 0.6031715382201733,
"grad_norm": 1.15625,
"learning_rate": 1.795796469363763e-05,
"loss": 6.4426,
"step": 8330
},
{
"epoch": 0.6038956337042312,
"grad_norm": 1.15625,
"learning_rate": 1.7901724627270833e-05,
"loss": 6.4416,
"step": 8340
},
{
"epoch": 0.604619729188289,
"grad_norm": 1.1953125,
"learning_rate": 1.784552360604131e-05,
"loss": 6.4545,
"step": 8350
},
{
"epoch": 0.6053438246723468,
"grad_norm": 1.171875,
"learning_rate": 1.778936193909126e-05,
"loss": 6.422,
"step": 8360
},
{
"epoch": 0.6060679201564046,
"grad_norm": 1.1796875,
"learning_rate": 1.773323993534644e-05,
"loss": 6.4194,
"step": 8370
},
{
"epoch": 0.6067920156404625,
"grad_norm": 1.1484375,
"learning_rate": 1.7677157903514412e-05,
"loss": 6.4644,
"step": 8380
},
{
"epoch": 0.6075161111245203,
"grad_norm": 1.203125,
"learning_rate": 1.762111615208287e-05,
"loss": 6.4715,
"step": 8390
},
{
"epoch": 0.6082402066085781,
"grad_norm": 1.2421875,
"learning_rate": 1.756511498931795e-05,
"loss": 6.4393,
"step": 8400
},
{
"epoch": 0.608964302092636,
"grad_norm": 1.171875,
"learning_rate": 1.750915472326252e-05,
"loss": 6.4506,
"step": 8410
},
{
"epoch": 0.6096883975766938,
"grad_norm": 1.1875,
"learning_rate": 1.7453235661734478e-05,
"loss": 6.4369,
"step": 8420
},
{
"epoch": 0.6104124930607516,
"grad_norm": 1.171875,
"learning_rate": 1.7397358112325085e-05,
"loss": 6.4276,
"step": 8430
},
{
"epoch": 0.6111365885448095,
"grad_norm": 1.15625,
"learning_rate": 1.7341522382397256e-05,
"loss": 6.4483,
"step": 8440
},
{
"epoch": 0.6118606840288673,
"grad_norm": 1.1328125,
"learning_rate": 1.7285728779083862e-05,
"loss": 6.4641,
"step": 8450
},
{
"epoch": 0.6125847795129251,
"grad_norm": 1.125,
"learning_rate": 1.7229977609286062e-05,
"loss": 6.4545,
"step": 8460
},
{
"epoch": 0.613308874996983,
"grad_norm": 1.1640625,
"learning_rate": 1.7174269179671593e-05,
"loss": 6.454,
"step": 8470
},
{
"epoch": 0.6140329704810408,
"grad_norm": 1.1640625,
"learning_rate": 1.7118603796673102e-05,
"loss": 6.4447,
"step": 8480
},
{
"epoch": 0.6147570659650986,
"grad_norm": 1.21875,
"learning_rate": 1.7062981766486437e-05,
"loss": 6.4641,
"step": 8490
},
{
"epoch": 0.6154811614491564,
"grad_norm": 1.1796875,
"learning_rate": 1.7007403395069005e-05,
"loss": 6.4613,
"step": 8500
},
{
"epoch": 0.6162052569332143,
"grad_norm": 1.2109375,
"learning_rate": 1.695186898813803e-05,
"loss": 6.4616,
"step": 8510
},
{
"epoch": 0.6169293524172721,
"grad_norm": 1.2421875,
"learning_rate": 1.6896378851168914e-05,
"loss": 6.4392,
"step": 8520
},
{
"epoch": 0.6176534479013299,
"grad_norm": 1.109375,
"learning_rate": 1.6840933289393564e-05,
"loss": 6.4388,
"step": 8530
},
{
"epoch": 0.6183775433853878,
"grad_norm": 1.15625,
"learning_rate": 1.6785532607798664e-05,
"loss": 6.4475,
"step": 8540
},
{
"epoch": 0.6191016388694456,
"grad_norm": 1.2421875,
"learning_rate": 1.673017711112405e-05,
"loss": 6.4399,
"step": 8550
},
{
"epoch": 0.6198257343535034,
"grad_norm": 1.1953125,
"learning_rate": 1.6674867103861e-05,
"loss": 6.4491,
"step": 8560
},
{
"epoch": 0.6205498298375612,
"grad_norm": 1.203125,
"learning_rate": 1.6619602890250565e-05,
"loss": 6.4558,
"step": 8570
},
{
"epoch": 0.6212739253216191,
"grad_norm": 1.2421875,
"learning_rate": 1.656438477428193e-05,
"loss": 6.4644,
"step": 8580
},
{
"epoch": 0.6219980208056769,
"grad_norm": 1.1640625,
"learning_rate": 1.6509213059690683e-05,
"loss": 6.4744,
"step": 8590
},
{
"epoch": 0.6227221162897347,
"grad_norm": 1.140625,
"learning_rate": 1.645408804995719e-05,
"loss": 6.4486,
"step": 8600
},
{
"epoch": 0.6234462117737926,
"grad_norm": 1.171875,
"learning_rate": 1.6399010048304897e-05,
"loss": 6.4742,
"step": 8610
},
{
"epoch": 0.6241703072578504,
"grad_norm": 1.1796875,
"learning_rate": 1.6343979357698687e-05,
"loss": 6.4486,
"step": 8620
},
{
"epoch": 0.6248944027419082,
"grad_norm": 1.1640625,
"learning_rate": 1.6288996280843188e-05,
"loss": 6.4436,
"step": 8630
},
{
"epoch": 0.625618498225966,
"grad_norm": 1.2109375,
"learning_rate": 1.6234061120181142e-05,
"loss": 6.4286,
"step": 8640
},
{
"epoch": 0.6263425937100239,
"grad_norm": 1.1796875,
"learning_rate": 1.6179174177891703e-05,
"loss": 6.4383,
"step": 8650
},
{
"epoch": 0.6270666891940817,
"grad_norm": 1.1484375,
"learning_rate": 1.6124335755888797e-05,
"loss": 6.4773,
"step": 8660
},
{
"epoch": 0.6277907846781395,
"grad_norm": 1.203125,
"learning_rate": 1.6069546155819464e-05,
"loss": 6.415,
"step": 8670
},
{
"epoch": 0.6285148801621974,
"grad_norm": 1.1875,
"learning_rate": 1.6014805679062185e-05,
"loss": 6.4114,
"step": 8680
},
{
"epoch": 0.6292389756462552,
"grad_norm": 1.1953125,
"learning_rate": 1.596011462672522e-05,
"loss": 6.4498,
"step": 8690
},
{
"epoch": 0.629963071130313,
"grad_norm": 1.2265625,
"learning_rate": 1.590547329964499e-05,
"loss": 6.4754,
"step": 8700
},
{
"epoch": 0.6306871666143709,
"grad_norm": 1.1953125,
"learning_rate": 1.585088199838437e-05,
"loss": 6.4334,
"step": 8710
},
{
"epoch": 0.6314112620984287,
"grad_norm": 1.21875,
"learning_rate": 1.5796341023231066e-05,
"loss": 6.4284,
"step": 8720
},
{
"epoch": 0.6321353575824865,
"grad_norm": 1.2890625,
"learning_rate": 1.5741850674195975e-05,
"loss": 6.4493,
"step": 8730
},
{
"epoch": 0.6328594530665443,
"grad_norm": 1.15625,
"learning_rate": 1.5687411251011487e-05,
"loss": 6.4531,
"step": 8740
},
{
"epoch": 0.6335835485506022,
"grad_norm": 1.1328125,
"learning_rate": 1.563302305312989e-05,
"loss": 6.4532,
"step": 8750
},
{
"epoch": 0.63430764403466,
"grad_norm": 1.15625,
"learning_rate": 1.5578686379721698e-05,
"loss": 6.4465,
"step": 8760
},
{
"epoch": 0.6350317395187178,
"grad_norm": 1.171875,
"learning_rate": 1.5524401529673994e-05,
"loss": 6.4508,
"step": 8770
},
{
"epoch": 0.6357558350027757,
"grad_norm": 1.1484375,
"learning_rate": 1.5470168801588805e-05,
"loss": 6.4562,
"step": 8780
},
{
"epoch": 0.6364799304868335,
"grad_norm": 1.2109375,
"learning_rate": 1.541598849378145e-05,
"loss": 6.4499,
"step": 8790
},
{
"epoch": 0.6372040259708913,
"grad_norm": 1.1875,
"learning_rate": 1.5361860904278927e-05,
"loss": 6.4544,
"step": 8800
},
{
"epoch": 0.6379281214549491,
"grad_norm": 1.1953125,
"learning_rate": 1.53077863308182e-05,
"loss": 6.4484,
"step": 8810
},
{
"epoch": 0.638652216939007,
"grad_norm": 1.1015625,
"learning_rate": 1.525376507084466e-05,
"loss": 6.4392,
"step": 8820
},
{
"epoch": 0.6393763124230648,
"grad_norm": 1.1875,
"learning_rate": 1.5199797421510415e-05,
"loss": 6.4545,
"step": 8830
},
{
"epoch": 0.6401004079071226,
"grad_norm": 1.203125,
"learning_rate": 1.514588367967268e-05,
"loss": 6.4504,
"step": 8840
},
{
"epoch": 0.6408245033911805,
"grad_norm": 1.1640625,
"learning_rate": 1.5092024141892142e-05,
"loss": 6.4662,
"step": 8850
},
{
"epoch": 0.6415485988752383,
"grad_norm": 1.1484375,
"learning_rate": 1.5038219104431334e-05,
"loss": 6.4296,
"step": 8860
},
{
"epoch": 0.6422726943592961,
"grad_norm": 1.2109375,
"learning_rate": 1.498446886325301e-05,
"loss": 6.4654,
"step": 8870
},
{
"epoch": 0.642996789843354,
"grad_norm": 1.1640625,
"learning_rate": 1.49307737140185e-05,
"loss": 6.4722,
"step": 8880
},
{
"epoch": 0.6437208853274119,
"grad_norm": 1.2265625,
"learning_rate": 1.4877133952086089e-05,
"loss": 6.4471,
"step": 8890
},
{
"epoch": 0.6444449808114697,
"grad_norm": 1.1796875,
"learning_rate": 1.4823549872509396e-05,
"loss": 6.4602,
"step": 8900
},
{
"epoch": 0.6451690762955276,
"grad_norm": 1.171875,
"learning_rate": 1.4770021770035767e-05,
"loss": 6.4455,
"step": 8910
},
{
"epoch": 0.6458931717795854,
"grad_norm": 1.1640625,
"learning_rate": 1.4716549939104618e-05,
"loss": 6.4532,
"step": 8920
},
{
"epoch": 0.6466172672636432,
"grad_norm": 1.1328125,
"learning_rate": 1.4663134673845838e-05,
"loss": 6.4466,
"step": 8930
},
{
"epoch": 0.647341362747701,
"grad_norm": 1.2265625,
"learning_rate": 1.4609776268078184e-05,
"loss": 6.439,
"step": 8940
},
{
"epoch": 0.6480654582317589,
"grad_norm": 1.21875,
"learning_rate": 1.455647501530763e-05,
"loss": 6.4401,
"step": 8950
},
{
"epoch": 0.6487895537158167,
"grad_norm": 1.1171875,
"learning_rate": 1.4503231208725781e-05,
"loss": 6.4521,
"step": 8960
},
{
"epoch": 0.6495136491998745,
"grad_norm": 1.203125,
"learning_rate": 1.4450045141208234e-05,
"loss": 6.452,
"step": 8970
},
{
"epoch": 0.6502377446839324,
"grad_norm": 1.109375,
"learning_rate": 1.4396917105313018e-05,
"loss": 6.481,
"step": 8980
},
{
"epoch": 0.6509618401679902,
"grad_norm": 1.171875,
"learning_rate": 1.4343847393278925e-05,
"loss": 6.474,
"step": 8990
}
],
"logging_steps": 10,
"max_steps": 13810,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 999,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.935261929621055e+19,
"train_batch_size": 30,
"trial_name": null,
"trial_params": null
}