nttx's picture
Training in progress, step 100, checkpoint
15be878 verified
{
"best_metric": 10.37281322479248,
"best_model_checkpoint": "miner_id_24/checkpoint-100",
"epoch": 0.02929222657537256,
"eval_steps": 50,
"global_step": 100,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0002929222657537256,
"grad_norm": 0.023211592808365822,
"learning_rate": 5.000000000000001e-07,
"loss": 10.3728,
"step": 1
},
{
"epoch": 0.0002929222657537256,
"eval_loss": 10.373064994812012,
"eval_runtime": 13.8949,
"eval_samples_per_second": 413.822,
"eval_steps_per_second": 103.491,
"step": 1
},
{
"epoch": 0.0005858445315074512,
"grad_norm": 0.0207655169069767,
"learning_rate": 1.0000000000000002e-06,
"loss": 10.3758,
"step": 2
},
{
"epoch": 0.0008787667972611768,
"grad_norm": 0.02462274208664894,
"learning_rate": 1.5e-06,
"loss": 10.3737,
"step": 3
},
{
"epoch": 0.0011716890630149024,
"grad_norm": 0.022382507100701332,
"learning_rate": 2.0000000000000003e-06,
"loss": 10.3739,
"step": 4
},
{
"epoch": 0.001464611328768628,
"grad_norm": 0.02478136122226715,
"learning_rate": 2.5e-06,
"loss": 10.3741,
"step": 5
},
{
"epoch": 0.0017575335945223536,
"grad_norm": 0.02159055881202221,
"learning_rate": 3e-06,
"loss": 10.3734,
"step": 6
},
{
"epoch": 0.002050455860276079,
"grad_norm": 0.02282402291893959,
"learning_rate": 3.5e-06,
"loss": 10.3743,
"step": 7
},
{
"epoch": 0.0023433781260298048,
"grad_norm": 0.024356847628951073,
"learning_rate": 4.000000000000001e-06,
"loss": 10.3753,
"step": 8
},
{
"epoch": 0.0026363003917835304,
"grad_norm": 0.024056624621152878,
"learning_rate": 4.5e-06,
"loss": 10.3759,
"step": 9
},
{
"epoch": 0.002929222657537256,
"grad_norm": 0.02597019262611866,
"learning_rate": 5e-06,
"loss": 10.3731,
"step": 10
},
{
"epoch": 0.0032221449232909816,
"grad_norm": 0.022780271247029305,
"learning_rate": 4.99847706754774e-06,
"loss": 10.3737,
"step": 11
},
{
"epoch": 0.003515067189044707,
"grad_norm": 0.027100862935185432,
"learning_rate": 4.993910125649561e-06,
"loss": 10.3741,
"step": 12
},
{
"epoch": 0.0038079894547984328,
"grad_norm": 0.029738787561655045,
"learning_rate": 4.986304738420684e-06,
"loss": 10.3725,
"step": 13
},
{
"epoch": 0.004100911720552158,
"grad_norm": 0.02589118853211403,
"learning_rate": 4.975670171853926e-06,
"loss": 10.3745,
"step": 14
},
{
"epoch": 0.004393833986305884,
"grad_norm": 0.0212496779859066,
"learning_rate": 4.962019382530521e-06,
"loss": 10.3723,
"step": 15
},
{
"epoch": 0.0046867562520596096,
"grad_norm": 0.028709905222058296,
"learning_rate": 4.9453690018345144e-06,
"loss": 10.3753,
"step": 16
},
{
"epoch": 0.004979678517813335,
"grad_norm": 0.026526963338255882,
"learning_rate": 4.925739315689991e-06,
"loss": 10.376,
"step": 17
},
{
"epoch": 0.005272600783567061,
"grad_norm": 0.02661576308310032,
"learning_rate": 4.903154239845798e-06,
"loss": 10.3722,
"step": 18
},
{
"epoch": 0.005565523049320786,
"grad_norm": 0.026910118758678436,
"learning_rate": 4.8776412907378845e-06,
"loss": 10.3746,
"step": 19
},
{
"epoch": 0.005858445315074512,
"grad_norm": 0.023959338665008545,
"learning_rate": 4.849231551964771e-06,
"loss": 10.3727,
"step": 20
},
{
"epoch": 0.0061513675808282375,
"grad_norm": 0.02778114750981331,
"learning_rate": 4.817959636416969e-06,
"loss": 10.3739,
"step": 21
},
{
"epoch": 0.006444289846581963,
"grad_norm": 0.027889959514141083,
"learning_rate": 4.783863644106502e-06,
"loss": 10.371,
"step": 22
},
{
"epoch": 0.006737212112335689,
"grad_norm": 0.027227042242884636,
"learning_rate": 4.746985115747918e-06,
"loss": 10.3725,
"step": 23
},
{
"epoch": 0.007030134378089414,
"grad_norm": 0.029192369431257248,
"learning_rate": 4.707368982147318e-06,
"loss": 10.3733,
"step": 24
},
{
"epoch": 0.00732305664384314,
"grad_norm": 0.030428782105445862,
"learning_rate": 4.665063509461098e-06,
"loss": 10.3728,
"step": 25
},
{
"epoch": 0.0076159789095968655,
"grad_norm": 0.029366305097937584,
"learning_rate": 4.620120240391065e-06,
"loss": 10.3729,
"step": 26
},
{
"epoch": 0.007908901175350591,
"grad_norm": 0.02809171751141548,
"learning_rate": 4.572593931387604e-06,
"loss": 10.3724,
"step": 27
},
{
"epoch": 0.008201823441104317,
"grad_norm": 0.02870216779410839,
"learning_rate": 4.522542485937369e-06,
"loss": 10.3726,
"step": 28
},
{
"epoch": 0.008494745706858042,
"grad_norm": 0.02702483907341957,
"learning_rate": 4.470026884016805e-06,
"loss": 10.3725,
"step": 29
},
{
"epoch": 0.008787667972611768,
"grad_norm": 0.028651194646954536,
"learning_rate": 4.415111107797445e-06,
"loss": 10.3753,
"step": 30
},
{
"epoch": 0.009080590238365494,
"grad_norm": 0.02481980249285698,
"learning_rate": 4.357862063693486e-06,
"loss": 10.3756,
"step": 31
},
{
"epoch": 0.009373512504119219,
"grad_norm": 0.029130522161722183,
"learning_rate": 4.2983495008466285e-06,
"loss": 10.3714,
"step": 32
},
{
"epoch": 0.009666434769872945,
"grad_norm": 0.03254735469818115,
"learning_rate": 4.236645926147493e-06,
"loss": 10.3722,
"step": 33
},
{
"epoch": 0.00995935703562667,
"grad_norm": 0.030360201373696327,
"learning_rate": 4.172826515897146e-06,
"loss": 10.3727,
"step": 34
},
{
"epoch": 0.010252279301380396,
"grad_norm": 0.029250526800751686,
"learning_rate": 4.106969024216348e-06,
"loss": 10.3737,
"step": 35
},
{
"epoch": 0.010545201567134121,
"grad_norm": 0.028940195217728615,
"learning_rate": 4.039153688314146e-06,
"loss": 10.3741,
"step": 36
},
{
"epoch": 0.010838123832887847,
"grad_norm": 0.03432109206914902,
"learning_rate": 3.969463130731183e-06,
"loss": 10.3732,
"step": 37
},
{
"epoch": 0.011131046098641573,
"grad_norm": 0.03054642491042614,
"learning_rate": 3.897982258676867e-06,
"loss": 10.3737,
"step": 38
},
{
"epoch": 0.011423968364395298,
"grad_norm": 0.03171557933092117,
"learning_rate": 3.824798160583012e-06,
"loss": 10.3728,
"step": 39
},
{
"epoch": 0.011716890630149024,
"grad_norm": 0.03246544674038887,
"learning_rate": 3.7500000000000005e-06,
"loss": 10.3727,
"step": 40
},
{
"epoch": 0.01200981289590275,
"grad_norm": 0.03463643416762352,
"learning_rate": 3.6736789069647273e-06,
"loss": 10.371,
"step": 41
},
{
"epoch": 0.012302735161656475,
"grad_norm": 0.03365543857216835,
"learning_rate": 3.595927866972694e-06,
"loss": 10.3698,
"step": 42
},
{
"epoch": 0.0125956574274102,
"grad_norm": 0.03245304524898529,
"learning_rate": 3.516841607689501e-06,
"loss": 10.3709,
"step": 43
},
{
"epoch": 0.012888579693163926,
"grad_norm": 0.03359043970704079,
"learning_rate": 3.436516483539781e-06,
"loss": 10.3693,
"step": 44
},
{
"epoch": 0.013181501958917652,
"grad_norm": 0.034752264618873596,
"learning_rate": 3.3550503583141726e-06,
"loss": 10.3731,
"step": 45
},
{
"epoch": 0.013474424224671377,
"grad_norm": 0.03718126192688942,
"learning_rate": 3.272542485937369e-06,
"loss": 10.3696,
"step": 46
},
{
"epoch": 0.013767346490425103,
"grad_norm": 0.03321147710084915,
"learning_rate": 3.189093389542498e-06,
"loss": 10.3722,
"step": 47
},
{
"epoch": 0.014060268756178829,
"grad_norm": 0.03554551675915718,
"learning_rate": 3.1048047389991693e-06,
"loss": 10.3675,
"step": 48
},
{
"epoch": 0.014353191021932554,
"grad_norm": 0.04130363464355469,
"learning_rate": 3.019779227044398e-06,
"loss": 10.3713,
"step": 49
},
{
"epoch": 0.01464611328768628,
"grad_norm": 0.04217420145869255,
"learning_rate": 2.9341204441673267e-06,
"loss": 10.3714,
"step": 50
},
{
"epoch": 0.01464611328768628,
"eval_loss": 10.37286376953125,
"eval_runtime": 13.8895,
"eval_samples_per_second": 413.982,
"eval_steps_per_second": 103.531,
"step": 50
},
{
"epoch": 0.014939035553440005,
"grad_norm": 0.021093251183629036,
"learning_rate": 2.847932752400164e-06,
"loss": 10.3754,
"step": 51
},
{
"epoch": 0.015231957819193731,
"grad_norm": 0.019990691915154457,
"learning_rate": 2.761321158169134e-06,
"loss": 10.3743,
"step": 52
},
{
"epoch": 0.015524880084947457,
"grad_norm": 0.025241075083613396,
"learning_rate": 2.6743911843603134e-06,
"loss": 10.3738,
"step": 53
},
{
"epoch": 0.015817802350701182,
"grad_norm": 0.02429122105240822,
"learning_rate": 2.587248741756253e-06,
"loss": 10.3742,
"step": 54
},
{
"epoch": 0.016110724616454908,
"grad_norm": 0.021294623613357544,
"learning_rate": 2.5e-06,
"loss": 10.3751,
"step": 55
},
{
"epoch": 0.016403646882208633,
"grad_norm": 0.022526361048221588,
"learning_rate": 2.4127512582437486e-06,
"loss": 10.3742,
"step": 56
},
{
"epoch": 0.01669656914796236,
"grad_norm": 0.02915373258292675,
"learning_rate": 2.325608815639687e-06,
"loss": 10.3752,
"step": 57
},
{
"epoch": 0.016989491413716085,
"grad_norm": 0.024015026167035103,
"learning_rate": 2.238678841830867e-06,
"loss": 10.3737,
"step": 58
},
{
"epoch": 0.01728241367946981,
"grad_norm": 0.026769675314426422,
"learning_rate": 2.1520672475998374e-06,
"loss": 10.3723,
"step": 59
},
{
"epoch": 0.017575335945223536,
"grad_norm": 0.027045700699090958,
"learning_rate": 2.0658795558326745e-06,
"loss": 10.3757,
"step": 60
},
{
"epoch": 0.01786825821097726,
"grad_norm": 0.02503349259495735,
"learning_rate": 1.9802207729556023e-06,
"loss": 10.3728,
"step": 61
},
{
"epoch": 0.018161180476730987,
"grad_norm": 0.026984136551618576,
"learning_rate": 1.895195261000831e-06,
"loss": 10.374,
"step": 62
},
{
"epoch": 0.018454102742484713,
"grad_norm": 0.02307491935789585,
"learning_rate": 1.8109066104575023e-06,
"loss": 10.3736,
"step": 63
},
{
"epoch": 0.018747025008238438,
"grad_norm": 0.029337333515286446,
"learning_rate": 1.7274575140626318e-06,
"loss": 10.3731,
"step": 64
},
{
"epoch": 0.019039947273992164,
"grad_norm": 0.02527059242129326,
"learning_rate": 1.6449496416858285e-06,
"loss": 10.3746,
"step": 65
},
{
"epoch": 0.01933286953974589,
"grad_norm": 0.02719458006322384,
"learning_rate": 1.56348351646022e-06,
"loss": 10.3726,
"step": 66
},
{
"epoch": 0.019625791805499615,
"grad_norm": 0.027298184111714363,
"learning_rate": 1.4831583923105e-06,
"loss": 10.3724,
"step": 67
},
{
"epoch": 0.01991871407125334,
"grad_norm": 0.024050690233707428,
"learning_rate": 1.4040721330273063e-06,
"loss": 10.3739,
"step": 68
},
{
"epoch": 0.020211636337007066,
"grad_norm": 0.026336120441555977,
"learning_rate": 1.3263210930352737e-06,
"loss": 10.3725,
"step": 69
},
{
"epoch": 0.020504558602760792,
"grad_norm": 0.02476734109222889,
"learning_rate": 1.2500000000000007e-06,
"loss": 10.3735,
"step": 70
},
{
"epoch": 0.020797480868514517,
"grad_norm": 0.030105607584118843,
"learning_rate": 1.1752018394169882e-06,
"loss": 10.3733,
"step": 71
},
{
"epoch": 0.021090403134268243,
"grad_norm": 0.027879927307367325,
"learning_rate": 1.1020177413231334e-06,
"loss": 10.3726,
"step": 72
},
{
"epoch": 0.02138332540002197,
"grad_norm": 0.023449528962373734,
"learning_rate": 1.0305368692688175e-06,
"loss": 10.373,
"step": 73
},
{
"epoch": 0.021676247665775694,
"grad_norm": 0.02750302664935589,
"learning_rate": 9.608463116858544e-07,
"loss": 10.3713,
"step": 74
},
{
"epoch": 0.02196916993152942,
"grad_norm": 0.02971792221069336,
"learning_rate": 8.930309757836517e-07,
"loss": 10.3702,
"step": 75
},
{
"epoch": 0.022262092197283145,
"grad_norm": 0.02820058912038803,
"learning_rate": 8.271734841028553e-07,
"loss": 10.3744,
"step": 76
},
{
"epoch": 0.02255501446303687,
"grad_norm": 0.02563357725739479,
"learning_rate": 7.633540738525066e-07,
"loss": 10.3719,
"step": 77
},
{
"epoch": 0.022847936728790597,
"grad_norm": 0.026806334033608437,
"learning_rate": 7.016504991533727e-07,
"loss": 10.3731,
"step": 78
},
{
"epoch": 0.023140858994544322,
"grad_norm": 0.03033982403576374,
"learning_rate": 6.421379363065142e-07,
"loss": 10.372,
"step": 79
},
{
"epoch": 0.023433781260298048,
"grad_norm": 0.02846497856080532,
"learning_rate": 5.848888922025553e-07,
"loss": 10.371,
"step": 80
},
{
"epoch": 0.023726703526051773,
"grad_norm": 0.026067426428198814,
"learning_rate": 5.299731159831953e-07,
"loss": 10.3729,
"step": 81
},
{
"epoch": 0.0240196257918055,
"grad_norm": 0.029144365340471268,
"learning_rate": 4.774575140626317e-07,
"loss": 10.3702,
"step": 82
},
{
"epoch": 0.024312548057559225,
"grad_norm": 0.028662119060754776,
"learning_rate": 4.27406068612396e-07,
"loss": 10.3753,
"step": 83
},
{
"epoch": 0.02460547032331295,
"grad_norm": 0.033134784549474716,
"learning_rate": 3.798797596089351e-07,
"loss": 10.3728,
"step": 84
},
{
"epoch": 0.024898392589066676,
"grad_norm": 0.028966225683689117,
"learning_rate": 3.3493649053890325e-07,
"loss": 10.372,
"step": 85
},
{
"epoch": 0.0251913148548204,
"grad_norm": 0.031356845051050186,
"learning_rate": 2.9263101785268253e-07,
"loss": 10.3712,
"step": 86
},
{
"epoch": 0.025484237120574127,
"grad_norm": 0.03198644891381264,
"learning_rate": 2.53014884252083e-07,
"loss": 10.3727,
"step": 87
},
{
"epoch": 0.025777159386327853,
"grad_norm": 0.0301873330026865,
"learning_rate": 2.1613635589349756e-07,
"loss": 10.375,
"step": 88
},
{
"epoch": 0.026070081652081578,
"grad_norm": 0.03230826184153557,
"learning_rate": 1.8204036358303173e-07,
"loss": 10.3722,
"step": 89
},
{
"epoch": 0.026363003917835304,
"grad_norm": 0.03326515480875969,
"learning_rate": 1.507684480352292e-07,
"loss": 10.3751,
"step": 90
},
{
"epoch": 0.02665592618358903,
"grad_norm": 0.03078848496079445,
"learning_rate": 1.223587092621162e-07,
"loss": 10.3708,
"step": 91
},
{
"epoch": 0.026948848449342755,
"grad_norm": 0.030252933502197266,
"learning_rate": 9.684576015420277e-08,
"loss": 10.3706,
"step": 92
},
{
"epoch": 0.02724177071509648,
"grad_norm": 0.03237851709127426,
"learning_rate": 7.426068431000883e-08,
"loss": 10.3729,
"step": 93
},
{
"epoch": 0.027534692980850206,
"grad_norm": 0.03713244944810867,
"learning_rate": 5.463099816548578e-08,
"loss": 10.3726,
"step": 94
},
{
"epoch": 0.02782761524660393,
"grad_norm": 0.032373908907175064,
"learning_rate": 3.798061746947995e-08,
"loss": 10.3699,
"step": 95
},
{
"epoch": 0.028120537512357657,
"grad_norm": 0.033183254301548004,
"learning_rate": 2.4329828146074096e-08,
"loss": 10.3712,
"step": 96
},
{
"epoch": 0.028413459778111383,
"grad_norm": 0.0401228629052639,
"learning_rate": 1.3695261579316776e-08,
"loss": 10.3702,
"step": 97
},
{
"epoch": 0.02870638204386511,
"grad_norm": 0.03960711508989334,
"learning_rate": 6.089874350439507e-09,
"loss": 10.3711,
"step": 98
},
{
"epoch": 0.028999304309618834,
"grad_norm": 0.037786684930324554,
"learning_rate": 1.5229324522605949e-09,
"loss": 10.3693,
"step": 99
},
{
"epoch": 0.02929222657537256,
"grad_norm": 0.03953954577445984,
"learning_rate": 0.0,
"loss": 10.3714,
"step": 100
},
{
"epoch": 0.02929222657537256,
"eval_loss": 10.37281322479248,
"eval_runtime": 13.8949,
"eval_samples_per_second": 413.82,
"eval_steps_per_second": 103.491,
"step": 100
}
],
"logging_steps": 1,
"max_steps": 100,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 5,
"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": 11022740619264.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}