cimol's picture
Training in progress, step 200, checkpoint
3075495 verified
{
"best_metric": 0.0007781803724355996,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 0.12682308180088775,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006341154090044388,
"grad_norm": 13.361982345581055,
"learning_rate": 7e-06,
"loss": 4.1214,
"step": 1
},
{
"epoch": 0.0006341154090044388,
"eval_loss": 4.043550491333008,
"eval_runtime": 193.3572,
"eval_samples_per_second": 13.736,
"eval_steps_per_second": 3.434,
"step": 1
},
{
"epoch": 0.0012682308180088776,
"grad_norm": 15.551319122314453,
"learning_rate": 1.4e-05,
"loss": 4.0177,
"step": 2
},
{
"epoch": 0.0019023462270133164,
"grad_norm": 16.52377700805664,
"learning_rate": 2.1e-05,
"loss": 3.9746,
"step": 3
},
{
"epoch": 0.0025364616360177552,
"grad_norm": 14.99045181274414,
"learning_rate": 2.8e-05,
"loss": 3.518,
"step": 4
},
{
"epoch": 0.0031705770450221942,
"grad_norm": 11.617705345153809,
"learning_rate": 3.5e-05,
"loss": 2.547,
"step": 5
},
{
"epoch": 0.003804692454026633,
"grad_norm": 10.286513328552246,
"learning_rate": 4.2e-05,
"loss": 1.7266,
"step": 6
},
{
"epoch": 0.004438807863031071,
"grad_norm": 13.773072242736816,
"learning_rate": 4.899999999999999e-05,
"loss": 0.9743,
"step": 7
},
{
"epoch": 0.0050729232720355105,
"grad_norm": 17.539566040039062,
"learning_rate": 5.6e-05,
"loss": 0.3935,
"step": 8
},
{
"epoch": 0.0057070386810399495,
"grad_norm": 8.249722480773926,
"learning_rate": 6.3e-05,
"loss": 0.1085,
"step": 9
},
{
"epoch": 0.0063411540900443885,
"grad_norm": 0.2928180396556854,
"learning_rate": 7e-05,
"loss": 0.0033,
"step": 10
},
{
"epoch": 0.006975269499048827,
"grad_norm": 0.022791218012571335,
"learning_rate": 6.999521567473641e-05,
"loss": 0.0003,
"step": 11
},
{
"epoch": 0.007609384908053266,
"grad_norm": 0.014727280475199223,
"learning_rate": 6.998086400693241e-05,
"loss": 0.0001,
"step": 12
},
{
"epoch": 0.008243500317057704,
"grad_norm": 0.01638036221265793,
"learning_rate": 6.995694892019065e-05,
"loss": 0.0001,
"step": 13
},
{
"epoch": 0.008877615726062143,
"grad_norm": 0.012314870022237301,
"learning_rate": 6.99234769526571e-05,
"loss": 0.0001,
"step": 14
},
{
"epoch": 0.009511731135066582,
"grad_norm": 0.007553283590823412,
"learning_rate": 6.988045725523343e-05,
"loss": 0.0,
"step": 15
},
{
"epoch": 0.010145846544071021,
"grad_norm": 0.005499022081494331,
"learning_rate": 6.982790158907539e-05,
"loss": 0.0,
"step": 16
},
{
"epoch": 0.01077996195307546,
"grad_norm": 0.004889129661023617,
"learning_rate": 6.976582432237733e-05,
"loss": 0.0,
"step": 17
},
{
"epoch": 0.011414077362079899,
"grad_norm": 0.0041648754850029945,
"learning_rate": 6.969424242644413e-05,
"loss": 0.0,
"step": 18
},
{
"epoch": 0.012048192771084338,
"grad_norm": 0.004347125068306923,
"learning_rate": 6.961317547105138e-05,
"loss": 0.0,
"step": 19
},
{
"epoch": 0.012682308180088777,
"grad_norm": 0.0046438840217888355,
"learning_rate": 6.952264561909527e-05,
"loss": 0.0,
"step": 20
},
{
"epoch": 0.013316423589093214,
"grad_norm": 0.00324826268479228,
"learning_rate": 6.942267762053337e-05,
"loss": 0.0,
"step": 21
},
{
"epoch": 0.013950538998097653,
"grad_norm": 0.0035093065816909075,
"learning_rate": 6.931329880561832e-05,
"loss": 0.0,
"step": 22
},
{
"epoch": 0.014584654407102092,
"grad_norm": 0.004882355220615864,
"learning_rate": 6.919453907742597e-05,
"loss": 0.0,
"step": 23
},
{
"epoch": 0.015218769816106531,
"grad_norm": 0.012776009738445282,
"learning_rate": 6.90664309036802e-05,
"loss": 0.0,
"step": 24
},
{
"epoch": 0.01585288522511097,
"grad_norm": 0.006269507110118866,
"learning_rate": 6.892900930787656e-05,
"loss": 0.0,
"step": 25
},
{
"epoch": 0.016487000634115408,
"grad_norm": 0.0025567272678017616,
"learning_rate": 6.87823118597072e-05,
"loss": 0.0,
"step": 26
},
{
"epoch": 0.017121116043119847,
"grad_norm": 0.0007885888335295022,
"learning_rate": 6.862637866478969e-05,
"loss": 0.0,
"step": 27
},
{
"epoch": 0.017755231452124286,
"grad_norm": 0.0003666975535452366,
"learning_rate": 6.846125235370252e-05,
"loss": 0.0,
"step": 28
},
{
"epoch": 0.018389346861128725,
"grad_norm": 0.000302265165373683,
"learning_rate": 6.828697807033038e-05,
"loss": 0.0,
"step": 29
},
{
"epoch": 0.019023462270133164,
"grad_norm": 0.00023431959562003613,
"learning_rate": 6.81036034595222e-05,
"loss": 0.0,
"step": 30
},
{
"epoch": 0.019657577679137603,
"grad_norm": 0.0002289388794451952,
"learning_rate": 6.791117865406564e-05,
"loss": 0.0,
"step": 31
},
{
"epoch": 0.020291693088142042,
"grad_norm": 0.00023351862910203636,
"learning_rate": 6.770975626098112e-05,
"loss": 0.0,
"step": 32
},
{
"epoch": 0.02092580849714648,
"grad_norm": 0.0002028094750130549,
"learning_rate": 6.749939134713974e-05,
"loss": 0.0,
"step": 33
},
{
"epoch": 0.02155992390615092,
"grad_norm": 0.00021913103410042822,
"learning_rate": 6.728014142420846e-05,
"loss": 0.0,
"step": 34
},
{
"epoch": 0.02219403931515536,
"grad_norm": 0.00020211738592479378,
"learning_rate": 6.7052066432927e-05,
"loss": 0.0,
"step": 35
},
{
"epoch": 0.022828154724159798,
"grad_norm": 0.00020341432536952198,
"learning_rate": 6.681522872672069e-05,
"loss": 0.0,
"step": 36
},
{
"epoch": 0.023462270133164237,
"grad_norm": 0.00019278828403912485,
"learning_rate": 6.656969305465356e-05,
"loss": 0.0,
"step": 37
},
{
"epoch": 0.024096385542168676,
"grad_norm": 0.00019703310681506991,
"learning_rate": 6.631552654372672e-05,
"loss": 0.0,
"step": 38
},
{
"epoch": 0.024730500951173115,
"grad_norm": 0.00019626211724244058,
"learning_rate": 6.60527986805264e-05,
"loss": 0.0,
"step": 39
},
{
"epoch": 0.025364616360177554,
"grad_norm": 0.00019462616182863712,
"learning_rate": 6.578158129222711e-05,
"loss": 0.0,
"step": 40
},
{
"epoch": 0.02599873176918199,
"grad_norm": 0.00020131043856963515,
"learning_rate": 6.550194852695469e-05,
"loss": 0.0,
"step": 41
},
{
"epoch": 0.02663284717818643,
"grad_norm": 0.00016793268150649965,
"learning_rate": 6.521397683351509e-05,
"loss": 0.0,
"step": 42
},
{
"epoch": 0.027266962587190868,
"grad_norm": 0.00018054968677461147,
"learning_rate": 6.491774494049386e-05,
"loss": 0.0,
"step": 43
},
{
"epoch": 0.027901077996195307,
"grad_norm": 0.00016843696357682347,
"learning_rate": 6.461333383473272e-05,
"loss": 0.0,
"step": 44
},
{
"epoch": 0.028535193405199746,
"grad_norm": 0.0001459317863918841,
"learning_rate": 6.430082673918849e-05,
"loss": 0.0,
"step": 45
},
{
"epoch": 0.029169308814204185,
"grad_norm": 0.00034006230998784304,
"learning_rate": 6.398030909018069e-05,
"loss": 0.0,
"step": 46
},
{
"epoch": 0.029803424223208624,
"grad_norm": 0.00014603414456360042,
"learning_rate": 6.365186851403423e-05,
"loss": 0.0,
"step": 47
},
{
"epoch": 0.030437539632213063,
"grad_norm": 0.0001704819151200354,
"learning_rate": 6.331559480312315e-05,
"loss": 0.0,
"step": 48
},
{
"epoch": 0.031071655041217502,
"grad_norm": 0.00017141974240075797,
"learning_rate": 6.297157989132236e-05,
"loss": 0.0,
"step": 49
},
{
"epoch": 0.03170577045022194,
"grad_norm": 0.00016341384616680443,
"learning_rate": 6.261991782887377e-05,
"loss": 0.0,
"step": 50
},
{
"epoch": 0.03170577045022194,
"eval_loss": 0.002813413506373763,
"eval_runtime": 195.2376,
"eval_samples_per_second": 13.604,
"eval_steps_per_second": 3.401,
"step": 50
},
{
"epoch": 0.032339885859226376,
"grad_norm": 2.238180637359619,
"learning_rate": 6.226070475667393e-05,
"loss": 0.0814,
"step": 51
},
{
"epoch": 0.032974001268230815,
"grad_norm": 0.00020237680291756988,
"learning_rate": 6.189403887999006e-05,
"loss": 0.0,
"step": 52
},
{
"epoch": 0.033608116677235254,
"grad_norm": 0.0005459522362798452,
"learning_rate": 6.152002044161171e-05,
"loss": 0.0,
"step": 53
},
{
"epoch": 0.03424223208623969,
"grad_norm": 0.0011731649283319712,
"learning_rate": 6.113875169444539e-05,
"loss": 0.0,
"step": 54
},
{
"epoch": 0.03487634749524413,
"grad_norm": 0.0058555202558636665,
"learning_rate": 6.0750336873559605e-05,
"loss": 0.0,
"step": 55
},
{
"epoch": 0.03551046290424857,
"grad_norm": 0.012653470039367676,
"learning_rate": 6.035488216768811e-05,
"loss": 0.0001,
"step": 56
},
{
"epoch": 0.03614457831325301,
"grad_norm": 0.02792506478726864,
"learning_rate": 5.9952495690198894e-05,
"loss": 0.0001,
"step": 57
},
{
"epoch": 0.03677869372225745,
"grad_norm": 0.022193649783730507,
"learning_rate": 5.954328744953709e-05,
"loss": 0.0001,
"step": 58
},
{
"epoch": 0.03741280913126189,
"grad_norm": 0.008992429822683334,
"learning_rate": 5.91273693191498e-05,
"loss": 0.0001,
"step": 59
},
{
"epoch": 0.03804692454026633,
"grad_norm": 0.004805019125342369,
"learning_rate": 5.870485500690094e-05,
"loss": 0.0,
"step": 60
},
{
"epoch": 0.03868103994927077,
"grad_norm": 0.0020239141304045916,
"learning_rate": 5.827586002398468e-05,
"loss": 0.0,
"step": 61
},
{
"epoch": 0.039315155358275206,
"grad_norm": 0.003996492829173803,
"learning_rate": 5.784050165334589e-05,
"loss": 0.0,
"step": 62
},
{
"epoch": 0.039949270767279645,
"grad_norm": 0.0012353772763162851,
"learning_rate": 5.739889891761608e-05,
"loss": 0.0,
"step": 63
},
{
"epoch": 0.040583386176284084,
"grad_norm": 0.0008236349676735699,
"learning_rate": 5.6951172546573794e-05,
"loss": 0.0,
"step": 64
},
{
"epoch": 0.04121750158528852,
"grad_norm": 0.0006633040611632168,
"learning_rate": 5.6497444944138376e-05,
"loss": 0.0,
"step": 65
},
{
"epoch": 0.04185161699429296,
"grad_norm": 0.0005764969973824918,
"learning_rate": 5.603784015490587e-05,
"loss": 0.0,
"step": 66
},
{
"epoch": 0.0424857324032974,
"grad_norm": 0.0003586241800803691,
"learning_rate": 5.557248383023655e-05,
"loss": 0.0,
"step": 67
},
{
"epoch": 0.04311984781230184,
"grad_norm": 0.0003223729436285794,
"learning_rate": 5.510150319390302e-05,
"loss": 0.0,
"step": 68
},
{
"epoch": 0.04375396322130628,
"grad_norm": 0.00028920508339069784,
"learning_rate": 5.4625027007308546e-05,
"loss": 0.0,
"step": 69
},
{
"epoch": 0.04438807863031072,
"grad_norm": 0.0002675386203918606,
"learning_rate": 5.414318553428494e-05,
"loss": 0.0,
"step": 70
},
{
"epoch": 0.04502219403931516,
"grad_norm": 0.00023136410163715482,
"learning_rate": 5.3656110505479776e-05,
"loss": 0.0,
"step": 71
},
{
"epoch": 0.045656309448319596,
"grad_norm": 0.0002023459383053705,
"learning_rate": 5.316393508234253e-05,
"loss": 0.0,
"step": 72
},
{
"epoch": 0.046290424857324035,
"grad_norm": 0.00023924530250951648,
"learning_rate": 5.266679382071953e-05,
"loss": 0.0,
"step": 73
},
{
"epoch": 0.046924540266328474,
"grad_norm": 0.00016957716434262693,
"learning_rate": 5.216482263406778e-05,
"loss": 0.0,
"step": 74
},
{
"epoch": 0.04755865567533291,
"grad_norm": 0.00015784092829562724,
"learning_rate": 5.1658158756297576e-05,
"loss": 0.0,
"step": 75
},
{
"epoch": 0.04819277108433735,
"grad_norm": 0.00014106677554082125,
"learning_rate": 5.114694070425407e-05,
"loss": 0.0,
"step": 76
},
{
"epoch": 0.04882688649334179,
"grad_norm": 0.00013483190559782088,
"learning_rate": 5.063130823984823e-05,
"loss": 0.0,
"step": 77
},
{
"epoch": 0.04946100190234623,
"grad_norm": 0.0001336677378276363,
"learning_rate": 5.011140233184724e-05,
"loss": 0.0,
"step": 78
},
{
"epoch": 0.05009511731135067,
"grad_norm": 0.00011928391904802993,
"learning_rate": 4.958736511733516e-05,
"loss": 0.0,
"step": 79
},
{
"epoch": 0.05072923272035511,
"grad_norm": 0.00012521824100986123,
"learning_rate": 4.905933986285393e-05,
"loss": 0.0,
"step": 80
},
{
"epoch": 0.05136334812935954,
"grad_norm": 0.00018319839728064835,
"learning_rate": 4.8527470925235824e-05,
"loss": 0.0,
"step": 81
},
{
"epoch": 0.05199746353836398,
"grad_norm": 0.0001144005436799489,
"learning_rate": 4.799190371213772e-05,
"loss": 0.0,
"step": 82
},
{
"epoch": 0.05263157894736842,
"grad_norm": 0.00012895939289592206,
"learning_rate": 4.745278464228808e-05,
"loss": 0.0,
"step": 83
},
{
"epoch": 0.05326569435637286,
"grad_norm": 0.0001076037879101932,
"learning_rate": 4.69102611054575e-05,
"loss": 0.0,
"step": 84
},
{
"epoch": 0.053899809765377296,
"grad_norm": 0.0001227714092237875,
"learning_rate": 4.6364481422163926e-05,
"loss": 0.0,
"step": 85
},
{
"epoch": 0.054533925174381735,
"grad_norm": 0.00011340944183757529,
"learning_rate": 4.581559480312316e-05,
"loss": 0.0,
"step": 86
},
{
"epoch": 0.055168040583386174,
"grad_norm": 0.00010645420843502507,
"learning_rate": 4.526375130845627e-05,
"loss": 0.0,
"step": 87
},
{
"epoch": 0.05580215599239061,
"grad_norm": 0.00010441164340591058,
"learning_rate": 4.4709101806664554e-05,
"loss": 0.0,
"step": 88
},
{
"epoch": 0.05643627140139505,
"grad_norm": 9.799294639378786e-05,
"learning_rate": 4.4151797933383685e-05,
"loss": 0.0,
"step": 89
},
{
"epoch": 0.05707038681039949,
"grad_norm": 0.00010191464389208704,
"learning_rate": 4.359199204992797e-05,
"loss": 0.0,
"step": 90
},
{
"epoch": 0.05770450221940393,
"grad_norm": 0.00010126228880835697,
"learning_rate": 4.30298372016363e-05,
"loss": 0.0,
"step": 91
},
{
"epoch": 0.05833861762840837,
"grad_norm": 0.00010101842781296,
"learning_rate": 4.246548707603114e-05,
"loss": 0.0,
"step": 92
},
{
"epoch": 0.05897273303741281,
"grad_norm": 0.00011214387632207945,
"learning_rate": 4.1899095960801805e-05,
"loss": 0.0,
"step": 93
},
{
"epoch": 0.05960684844641725,
"grad_norm": 0.0001037925758282654,
"learning_rate": 4.133081870162385e-05,
"loss": 0.0,
"step": 94
},
{
"epoch": 0.060240963855421686,
"grad_norm": 0.00011326379899401218,
"learning_rate": 4.076081065982569e-05,
"loss": 0.0,
"step": 95
},
{
"epoch": 0.060875079264426125,
"grad_norm": 0.00011256959260208532,
"learning_rate": 4.018922766991447e-05,
"loss": 0.0,
"step": 96
},
{
"epoch": 0.061509194673430564,
"grad_norm": 0.0001253296941285953,
"learning_rate": 3.961622599697241e-05,
"loss": 0.0,
"step": 97
},
{
"epoch": 0.062143310082435003,
"grad_norm": 0.0001224202278535813,
"learning_rate": 3.9041962293935516e-05,
"loss": 0.0,
"step": 98
},
{
"epoch": 0.06277742549143944,
"grad_norm": 0.00011974848894169554,
"learning_rate": 3.84665935587662e-05,
"loss": 0.0,
"step": 99
},
{
"epoch": 0.06341154090044387,
"grad_norm": 0.004142427816987038,
"learning_rate": 3.7890277091531636e-05,
"loss": 0.0,
"step": 100
},
{
"epoch": 0.06341154090044387,
"eval_loss": 0.0027051952201873064,
"eval_runtime": 195.558,
"eval_samples_per_second": 13.582,
"eval_steps_per_second": 3.395,
"step": 100
},
{
"epoch": 0.06404565630944832,
"grad_norm": 0.4534345865249634,
"learning_rate": 3.7313170451399475e-05,
"loss": 0.0532,
"step": 101
},
{
"epoch": 0.06467977171845275,
"grad_norm": 7.396758155664429e-05,
"learning_rate": 3.673543141356278e-05,
"loss": 0.0,
"step": 102
},
{
"epoch": 0.0653138871274572,
"grad_norm": 8.378738857572898e-05,
"learning_rate": 3.6157217926105783e-05,
"loss": 0.0,
"step": 103
},
{
"epoch": 0.06594800253646163,
"grad_norm": 8.676236029714346e-05,
"learning_rate": 3.557868806682255e-05,
"loss": 0.0,
"step": 104
},
{
"epoch": 0.06658211794546608,
"grad_norm": 9.078224684344605e-05,
"learning_rate": 3.5e-05,
"loss": 0.0,
"step": 105
},
{
"epoch": 0.06721623335447051,
"grad_norm": 0.00010050006676465273,
"learning_rate": 3.442131193317745e-05,
"loss": 0.0,
"step": 106
},
{
"epoch": 0.06785034876347495,
"grad_norm": 0.00010982064122799784,
"learning_rate": 3.384278207389421e-05,
"loss": 0.0,
"step": 107
},
{
"epoch": 0.06848446417247939,
"grad_norm": 0.00010909491538768634,
"learning_rate": 3.3264568586437216e-05,
"loss": 0.0,
"step": 108
},
{
"epoch": 0.06911857958148383,
"grad_norm": 0.00012787252489943057,
"learning_rate": 3.268682954860052e-05,
"loss": 0.0,
"step": 109
},
{
"epoch": 0.06975269499048826,
"grad_norm": 0.00012508960207924247,
"learning_rate": 3.210972290846837e-05,
"loss": 0.0,
"step": 110
},
{
"epoch": 0.07038681039949271,
"grad_norm": 0.00012248492566868663,
"learning_rate": 3.15334064412338e-05,
"loss": 0.0,
"step": 111
},
{
"epoch": 0.07102092580849714,
"grad_norm": 0.00013955520989838988,
"learning_rate": 3.0958037706064485e-05,
"loss": 0.0,
"step": 112
},
{
"epoch": 0.07165504121750159,
"grad_norm": 0.00014822160301264375,
"learning_rate": 3.038377400302758e-05,
"loss": 0.0,
"step": 113
},
{
"epoch": 0.07228915662650602,
"grad_norm": 0.0001431052660336718,
"learning_rate": 2.9810772330085524e-05,
"loss": 0.0,
"step": 114
},
{
"epoch": 0.07292327203551047,
"grad_norm": 0.00015100237214937806,
"learning_rate": 2.9239189340174306e-05,
"loss": 0.0,
"step": 115
},
{
"epoch": 0.0735573874445149,
"grad_norm": 0.0001577723160153255,
"learning_rate": 2.8669181298376163e-05,
"loss": 0.0,
"step": 116
},
{
"epoch": 0.07419150285351934,
"grad_norm": 0.00015383222489617765,
"learning_rate": 2.8100904039198193e-05,
"loss": 0.0,
"step": 117
},
{
"epoch": 0.07482561826252378,
"grad_norm": 0.00015594527940265834,
"learning_rate": 2.7534512923968863e-05,
"loss": 0.0,
"step": 118
},
{
"epoch": 0.07545973367152822,
"grad_norm": 0.0001666178577579558,
"learning_rate": 2.6970162798363695e-05,
"loss": 0.0,
"step": 119
},
{
"epoch": 0.07609384908053266,
"grad_norm": 0.0001624413562240079,
"learning_rate": 2.640800795007203e-05,
"loss": 0.0,
"step": 120
},
{
"epoch": 0.0767279644895371,
"grad_norm": 0.00015609640104230493,
"learning_rate": 2.5848202066616305e-05,
"loss": 0.0,
"step": 121
},
{
"epoch": 0.07736207989854153,
"grad_norm": 0.0001663492585066706,
"learning_rate": 2.5290898193335446e-05,
"loss": 0.0,
"step": 122
},
{
"epoch": 0.07799619530754598,
"grad_norm": 0.00017570104682818055,
"learning_rate": 2.4736248691543736e-05,
"loss": 0.0,
"step": 123
},
{
"epoch": 0.07863031071655041,
"grad_norm": 0.00016999592480715364,
"learning_rate": 2.4184405196876842e-05,
"loss": 0.0,
"step": 124
},
{
"epoch": 0.07926442612555486,
"grad_norm": 0.00017873845354188234,
"learning_rate": 2.363551857783608e-05,
"loss": 0.0,
"step": 125
},
{
"epoch": 0.07989854153455929,
"grad_norm": 0.00016306425095535815,
"learning_rate": 2.308973889454249e-05,
"loss": 0.0,
"step": 126
},
{
"epoch": 0.08053265694356374,
"grad_norm": 0.0001576338690938428,
"learning_rate": 2.2547215357711918e-05,
"loss": 0.0,
"step": 127
},
{
"epoch": 0.08116677235256817,
"grad_norm": 0.0001510537986177951,
"learning_rate": 2.2008096287862266e-05,
"loss": 0.0,
"step": 128
},
{
"epoch": 0.0818008877615726,
"grad_norm": 0.0001824547944124788,
"learning_rate": 2.1472529074764177e-05,
"loss": 0.0,
"step": 129
},
{
"epoch": 0.08243500317057705,
"grad_norm": 0.00018089111836161464,
"learning_rate": 2.0940660137146074e-05,
"loss": 0.0,
"step": 130
},
{
"epoch": 0.08306911857958148,
"grad_norm": 0.00017745721561368555,
"learning_rate": 2.041263488266484e-05,
"loss": 0.0,
"step": 131
},
{
"epoch": 0.08370323398858592,
"grad_norm": 0.00019485809025354683,
"learning_rate": 1.988859766815275e-05,
"loss": 0.0,
"step": 132
},
{
"epoch": 0.08433734939759036,
"grad_norm": 0.00018570972315501422,
"learning_rate": 1.9368691760151773e-05,
"loss": 0.0,
"step": 133
},
{
"epoch": 0.0849714648065948,
"grad_norm": 0.00018695491598919034,
"learning_rate": 1.885305929574593e-05,
"loss": 0.0,
"step": 134
},
{
"epoch": 0.08560558021559923,
"grad_norm": 0.00021328394359443337,
"learning_rate": 1.8341841243702424e-05,
"loss": 0.0,
"step": 135
},
{
"epoch": 0.08623969562460368,
"grad_norm": 0.00017723164637573063,
"learning_rate": 1.7835177365932225e-05,
"loss": 0.0,
"step": 136
},
{
"epoch": 0.08687381103360811,
"grad_norm": 0.00018894548702519387,
"learning_rate": 1.7333206179280478e-05,
"loss": 0.0,
"step": 137
},
{
"epoch": 0.08750792644261256,
"grad_norm": 0.0001940389338415116,
"learning_rate": 1.6836064917657478e-05,
"loss": 0.0,
"step": 138
},
{
"epoch": 0.08814204185161699,
"grad_norm": 0.00017241237219423056,
"learning_rate": 1.6343889494520224e-05,
"loss": 0.0,
"step": 139
},
{
"epoch": 0.08877615726062144,
"grad_norm": 0.00018744268163572997,
"learning_rate": 1.5856814465715064e-05,
"loss": 0.0,
"step": 140
},
{
"epoch": 0.08941027266962587,
"grad_norm": 0.00019515029271133244,
"learning_rate": 1.5374972992691458e-05,
"loss": 0.0,
"step": 141
},
{
"epoch": 0.09004438807863031,
"grad_norm": 0.00017881295934785157,
"learning_rate": 1.4898496806096974e-05,
"loss": 0.0,
"step": 142
},
{
"epoch": 0.09067850348763475,
"grad_norm": 0.00021232447761576623,
"learning_rate": 1.4427516169763444e-05,
"loss": 0.0,
"step": 143
},
{
"epoch": 0.09131261889663919,
"grad_norm": 0.00019485698430798948,
"learning_rate": 1.396215984509412e-05,
"loss": 0.0,
"step": 144
},
{
"epoch": 0.09194673430564362,
"grad_norm": 0.00022662342234980315,
"learning_rate": 1.3502555055861625e-05,
"loss": 0.0,
"step": 145
},
{
"epoch": 0.09258084971464807,
"grad_norm": 0.0002101845311699435,
"learning_rate": 1.3048827453426203e-05,
"loss": 0.0,
"step": 146
},
{
"epoch": 0.0932149651236525,
"grad_norm": 0.00027142782346345484,
"learning_rate": 1.2601101082383917e-05,
"loss": 0.0,
"step": 147
},
{
"epoch": 0.09384908053265695,
"grad_norm": 0.00023996931849978864,
"learning_rate": 1.2159498346654094e-05,
"loss": 0.0,
"step": 148
},
{
"epoch": 0.09448319594166138,
"grad_norm": 0.0002598701976239681,
"learning_rate": 1.1724139976015306e-05,
"loss": 0.0,
"step": 149
},
{
"epoch": 0.09511731135066583,
"grad_norm": 0.00027565957861952484,
"learning_rate": 1.1295144993099068e-05,
"loss": 0.0,
"step": 150
},
{
"epoch": 0.09511731135066583,
"eval_loss": 0.0010700421407818794,
"eval_runtime": 195.9764,
"eval_samples_per_second": 13.553,
"eval_steps_per_second": 3.388,
"step": 150
},
{
"epoch": 0.09575142675967026,
"grad_norm": 1.567829966545105,
"learning_rate": 1.0872630680850196e-05,
"loss": 0.0652,
"step": 151
},
{
"epoch": 0.0963855421686747,
"grad_norm": 0.00010371788084739819,
"learning_rate": 1.0456712550462898e-05,
"loss": 0.0,
"step": 152
},
{
"epoch": 0.09701965757767914,
"grad_norm": 0.00011418973735999316,
"learning_rate": 1.0047504309801104e-05,
"loss": 0.0,
"step": 153
},
{
"epoch": 0.09765377298668358,
"grad_norm": 0.00012190837878733873,
"learning_rate": 9.645117832311886e-06,
"loss": 0.0,
"step": 154
},
{
"epoch": 0.09828788839568801,
"grad_norm": 0.00011330062261549756,
"learning_rate": 9.249663126440394e-06,
"loss": 0.0,
"step": 155
},
{
"epoch": 0.09892200380469246,
"grad_norm": 0.00012431383947841823,
"learning_rate": 8.861248305554624e-06,
"loss": 0.0,
"step": 156
},
{
"epoch": 0.09955611921369689,
"grad_norm": 0.00012524866906460375,
"learning_rate": 8.47997955838829e-06,
"loss": 0.0,
"step": 157
},
{
"epoch": 0.10019023462270134,
"grad_norm": 0.00012364127906039357,
"learning_rate": 8.10596112000994e-06,
"loss": 0.0,
"step": 158
},
{
"epoch": 0.10082435003170577,
"grad_norm": 0.00011792991426773369,
"learning_rate": 7.739295243326067e-06,
"loss": 0.0,
"step": 159
},
{
"epoch": 0.10145846544071022,
"grad_norm": 0.0001275044633075595,
"learning_rate": 7.380082171126228e-06,
"loss": 0.0,
"step": 160
},
{
"epoch": 0.10209258084971465,
"grad_norm": 0.00011679629824357107,
"learning_rate": 7.028420108677635e-06,
"loss": 0.0,
"step": 161
},
{
"epoch": 0.10272669625871908,
"grad_norm": 0.00013719015987589955,
"learning_rate": 6.684405196876842e-06,
"loss": 0.0,
"step": 162
},
{
"epoch": 0.10336081166772353,
"grad_norm": 0.0001346962817478925,
"learning_rate": 6.3481314859657675e-06,
"loss": 0.0,
"step": 163
},
{
"epoch": 0.10399492707672796,
"grad_norm": 0.00014347612159326673,
"learning_rate": 6.019690909819298e-06,
"loss": 0.0,
"step": 164
},
{
"epoch": 0.1046290424857324,
"grad_norm": 0.0001298694551223889,
"learning_rate": 5.6991732608115e-06,
"loss": 0.0,
"step": 165
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.0001310157822445035,
"learning_rate": 5.386666165267256e-06,
"loss": 0.0,
"step": 166
},
{
"epoch": 0.10589727330374128,
"grad_norm": 0.0001319856382906437,
"learning_rate": 5.08225505950613e-06,
"loss": 0.0,
"step": 167
},
{
"epoch": 0.10653138871274571,
"grad_norm": 0.00013899788609705865,
"learning_rate": 4.786023166484913e-06,
"loss": 0.0,
"step": 168
},
{
"epoch": 0.10716550412175016,
"grad_norm": 0.00013763431343249977,
"learning_rate": 4.498051473045291e-06,
"loss": 0.0,
"step": 169
},
{
"epoch": 0.10779961953075459,
"grad_norm": 0.00014283151540439576,
"learning_rate": 4.218418707772886e-06,
"loss": 0.0,
"step": 170
},
{
"epoch": 0.10843373493975904,
"grad_norm": 0.00014707644004374743,
"learning_rate": 3.947201319473587e-06,
"loss": 0.0,
"step": 171
},
{
"epoch": 0.10906785034876347,
"grad_norm": 0.00016239233082160354,
"learning_rate": 3.684473456273278e-06,
"loss": 0.0,
"step": 172
},
{
"epoch": 0.10970196575776792,
"grad_norm": 0.0001635009830351919,
"learning_rate": 3.4303069453464383e-06,
"loss": 0.0,
"step": 173
},
{
"epoch": 0.11033608116677235,
"grad_norm": 0.00016097123443614691,
"learning_rate": 3.184771273279312e-06,
"loss": 0.0,
"step": 174
},
{
"epoch": 0.1109701965757768,
"grad_norm": 0.00013572271564044058,
"learning_rate": 2.947933567072987e-06,
"loss": 0.0,
"step": 175
},
{
"epoch": 0.11160431198478123,
"grad_norm": 0.00015159006579779088,
"learning_rate": 2.719858575791534e-06,
"loss": 0.0,
"step": 176
},
{
"epoch": 0.11223842739378567,
"grad_norm": 0.0001638934772927314,
"learning_rate": 2.500608652860256e-06,
"loss": 0.0,
"step": 177
},
{
"epoch": 0.1128725428027901,
"grad_norm": 0.00014194424147717655,
"learning_rate": 2.2902437390188737e-06,
"loss": 0.0,
"step": 178
},
{
"epoch": 0.11350665821179455,
"grad_norm": 0.00015119832823984325,
"learning_rate": 2.0888213459343587e-06,
"loss": 0.0,
"step": 179
},
{
"epoch": 0.11414077362079898,
"grad_norm": 0.00015596729645039886,
"learning_rate": 1.8963965404777875e-06,
"loss": 0.0,
"step": 180
},
{
"epoch": 0.11477488902980343,
"grad_norm": 0.00015777695807628334,
"learning_rate": 1.7130219296696263e-06,
"loss": 0.0,
"step": 181
},
{
"epoch": 0.11540900443880786,
"grad_norm": 0.00014426129928324372,
"learning_rate": 1.5387476462974824e-06,
"loss": 0.0,
"step": 182
},
{
"epoch": 0.1160431198478123,
"grad_norm": 0.00016764002793934196,
"learning_rate": 1.3736213352103147e-06,
"loss": 0.0,
"step": 183
},
{
"epoch": 0.11667723525681674,
"grad_norm": 0.00016620867245364934,
"learning_rate": 1.2176881402928002e-06,
"loss": 0.0,
"step": 184
},
{
"epoch": 0.11731135066582118,
"grad_norm": 0.00017529992328491062,
"learning_rate": 1.0709906921234367e-06,
"loss": 0.0,
"step": 185
},
{
"epoch": 0.11794546607482562,
"grad_norm": 0.00016166568093467504,
"learning_rate": 9.33569096319799e-07,
"loss": 0.0,
"step": 186
},
{
"epoch": 0.11857958148383006,
"grad_norm": 0.00017907556321006268,
"learning_rate": 8.054609225740255e-07,
"loss": 0.0,
"step": 187
},
{
"epoch": 0.1192136968928345,
"grad_norm": 0.00016855564899742603,
"learning_rate": 6.867011943816724e-07,
"loss": 0.0,
"step": 188
},
{
"epoch": 0.11984781230183894,
"grad_norm": 0.00017594834207557142,
"learning_rate": 5.77322379466617e-07,
"loss": 0.0,
"step": 189
},
{
"epoch": 0.12048192771084337,
"grad_norm": 0.000184446937055327,
"learning_rate": 4.773543809047186e-07,
"loss": 0.0,
"step": 190
},
{
"epoch": 0.12111604311984782,
"grad_norm": 0.000200022550416179,
"learning_rate": 3.868245289486027e-07,
"loss": 0.0,
"step": 191
},
{
"epoch": 0.12175015852885225,
"grad_norm": 0.00020334366126917303,
"learning_rate": 3.0575757355586817e-07,
"loss": 0.0,
"step": 192
},
{
"epoch": 0.1223842739378567,
"grad_norm": 0.00020971783669665456,
"learning_rate": 2.3417567762266497e-07,
"loss": 0.0,
"step": 193
},
{
"epoch": 0.12301838934686113,
"grad_norm": 0.00019827681535389274,
"learning_rate": 1.7209841092460043e-07,
"loss": 0.0,
"step": 194
},
{
"epoch": 0.12365250475586556,
"grad_norm": 0.0002055370423477143,
"learning_rate": 1.1954274476655534e-07,
"loss": 0.0,
"step": 195
},
{
"epoch": 0.12428662016487001,
"grad_norm": 0.0002557662082836032,
"learning_rate": 7.652304734289127e-08,
"loss": 0.0,
"step": 196
},
{
"epoch": 0.12492073557387444,
"grad_norm": 0.00027073186356574297,
"learning_rate": 4.30510798093342e-08,
"loss": 0.0,
"step": 197
},
{
"epoch": 0.12555485098287889,
"grad_norm": 0.0002866918221116066,
"learning_rate": 1.9135993067588284e-08,
"loss": 0.0,
"step": 198
},
{
"epoch": 0.12618896639188332,
"grad_norm": 0.00026750980759970844,
"learning_rate": 4.784325263584854e-09,
"loss": 0.0,
"step": 199
},
{
"epoch": 0.12682308180088775,
"grad_norm": 0.00027732347371056676,
"learning_rate": 0.0,
"loss": 0.0,
"step": 200
},
{
"epoch": 0.12682308180088775,
"eval_loss": 0.0007781803724355996,
"eval_runtime": 195.8366,
"eval_samples_per_second": 13.562,
"eval_steps_per_second": 3.391,
"step": 200
}
],
"logging_steps": 1,
"max_steps": 200,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 4,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 0
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.861550979252224e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}