36f8fe9f41
Model: alinerodrigues/wav2vec2-large-xlsr-mecita-coraa-portuguese-2-all-03 Source: Original Platform
986 lines
24 KiB
JSON
986 lines
24 KiB
JSON
{
|
|
"best_metric": 0.1349739283323288,
|
|
"best_model_checkpoint": "wav2vec2-large-xlsr-mecita-coraa-portuguese-2-all-03/checkpoint-3698",
|
|
"epoch": 63.0,
|
|
"global_step": 5418,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.01,
|
|
"learning_rate": 3e-05,
|
|
"loss": 28.65,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"eval_cer": 1.0,
|
|
"eval_loss": 3.1940298080444336,
|
|
"eval_runtime": 19.0983,
|
|
"eval_samples_per_second": 17.855,
|
|
"eval_steps_per_second": 2.252,
|
|
"eval_wer": 1.0,
|
|
"step": 86
|
|
},
|
|
{
|
|
"epoch": 1.16,
|
|
"learning_rate": 2.966511627906977e-05,
|
|
"loss": 8.1771,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"eval_cer": 1.0,
|
|
"eval_loss": 2.9853453636169434,
|
|
"eval_runtime": 18.8941,
|
|
"eval_samples_per_second": 18.048,
|
|
"eval_steps_per_second": 2.276,
|
|
"eval_wer": 1.0,
|
|
"step": 172
|
|
},
|
|
{
|
|
"epoch": 2.33,
|
|
"learning_rate": 2.9316279069767443e-05,
|
|
"loss": 3.0358,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"eval_cer": 1.0,
|
|
"eval_loss": 2.895540237426758,
|
|
"eval_runtime": 19.0821,
|
|
"eval_samples_per_second": 17.87,
|
|
"eval_steps_per_second": 2.253,
|
|
"eval_wer": 1.0,
|
|
"step": 258
|
|
},
|
|
{
|
|
"epoch": 3.49,
|
|
"learning_rate": 2.8967441860465116e-05,
|
|
"loss": 2.9359,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"eval_cer": 1.0,
|
|
"eval_loss": 2.9086179733276367,
|
|
"eval_runtime": 19.1659,
|
|
"eval_samples_per_second": 17.792,
|
|
"eval_steps_per_second": 2.244,
|
|
"eval_wer": 1.0,
|
|
"step": 344
|
|
},
|
|
{
|
|
"epoch": 4.65,
|
|
"learning_rate": 2.861860465116279e-05,
|
|
"loss": 2.9026,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 5.0,
|
|
"eval_cer": 0.9998215162197135,
|
|
"eval_loss": 2.694617509841919,
|
|
"eval_runtime": 18.8982,
|
|
"eval_samples_per_second": 18.044,
|
|
"eval_steps_per_second": 2.275,
|
|
"eval_wer": 0.9997623009270263,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 5.81,
|
|
"learning_rate": 2.8269767441860464e-05,
|
|
"loss": 2.4934,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 6.0,
|
|
"eval_cer": 0.1256525813216724,
|
|
"eval_loss": 0.9106346964836121,
|
|
"eval_runtime": 18.9322,
|
|
"eval_samples_per_second": 18.012,
|
|
"eval_steps_per_second": 2.271,
|
|
"eval_wer": 0.49441407178512004,
|
|
"step": 516
|
|
},
|
|
{
|
|
"epoch": 6.98,
|
|
"learning_rate": 2.7920930232558138e-05,
|
|
"loss": 1.1576,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 7.0,
|
|
"eval_cer": 0.06273704877069296,
|
|
"eval_loss": 0.4618076980113983,
|
|
"eval_runtime": 19.1545,
|
|
"eval_samples_per_second": 17.803,
|
|
"eval_steps_per_second": 2.245,
|
|
"eval_wer": 0.22914190634656526,
|
|
"step": 602
|
|
},
|
|
{
|
|
"epoch": 8.0,
|
|
"eval_cer": 0.05381285975636964,
|
|
"eval_loss": 0.3426876664161682,
|
|
"eval_runtime": 19.3029,
|
|
"eval_samples_per_second": 17.666,
|
|
"eval_steps_per_second": 2.228,
|
|
"eval_wer": 0.18706917043023533,
|
|
"step": 688
|
|
},
|
|
{
|
|
"epoch": 8.14,
|
|
"learning_rate": 2.7572093023255812e-05,
|
|
"loss": 0.7498,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 9.0,
|
|
"eval_cer": 0.04801213689705948,
|
|
"eval_loss": 0.283805251121521,
|
|
"eval_runtime": 18.9222,
|
|
"eval_samples_per_second": 18.021,
|
|
"eval_steps_per_second": 2.272,
|
|
"eval_wer": 0.16734014737342523,
|
|
"step": 774
|
|
},
|
|
{
|
|
"epoch": 9.3,
|
|
"learning_rate": 2.722325581395349e-05,
|
|
"loss": 0.5493,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 10.0,
|
|
"eval_cer": 0.04417473562090045,
|
|
"eval_loss": 0.25037863850593567,
|
|
"eval_runtime": 19.1667,
|
|
"eval_samples_per_second": 17.791,
|
|
"eval_steps_per_second": 2.243,
|
|
"eval_wer": 0.14903731875445686,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 10.47,
|
|
"learning_rate": 2.6874418604651163e-05,
|
|
"loss": 0.4735,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 11.0,
|
|
"eval_cer": 0.041185132301102136,
|
|
"eval_loss": 0.23181051015853882,
|
|
"eval_runtime": 18.9289,
|
|
"eval_samples_per_second": 18.015,
|
|
"eval_steps_per_second": 2.272,
|
|
"eval_wer": 0.1388162586165914,
|
|
"step": 946
|
|
},
|
|
{
|
|
"epoch": 11.63,
|
|
"learning_rate": 2.6525581395348837e-05,
|
|
"loss": 0.4183,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 12.0,
|
|
"eval_cer": 0.037392351970014726,
|
|
"eval_loss": 0.2174120843410492,
|
|
"eval_runtime": 18.9003,
|
|
"eval_samples_per_second": 18.042,
|
|
"eval_steps_per_second": 2.275,
|
|
"eval_wer": 0.12146422628951747,
|
|
"step": 1032
|
|
},
|
|
{
|
|
"epoch": 12.79,
|
|
"learning_rate": 2.617674418604651e-05,
|
|
"loss": 0.3924,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 13.0,
|
|
"eval_cer": 0.03565213511222168,
|
|
"eval_loss": 0.20193208754062653,
|
|
"eval_runtime": 19.1675,
|
|
"eval_samples_per_second": 17.791,
|
|
"eval_steps_per_second": 2.243,
|
|
"eval_wer": 0.1126693605894937,
|
|
"step": 1118
|
|
},
|
|
{
|
|
"epoch": 13.95,
|
|
"learning_rate": 2.5827906976744185e-05,
|
|
"loss": 0.3497,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 14.0,
|
|
"eval_cer": 0.03431350676007318,
|
|
"eval_loss": 0.18761980533599854,
|
|
"eval_runtime": 19.269,
|
|
"eval_samples_per_second": 17.697,
|
|
"eval_steps_per_second": 2.232,
|
|
"eval_wer": 0.11124316615165201,
|
|
"step": 1204
|
|
},
|
|
{
|
|
"epoch": 15.0,
|
|
"eval_cer": 0.03507206282629066,
|
|
"eval_loss": 0.18645305931568146,
|
|
"eval_runtime": 18.9659,
|
|
"eval_samples_per_second": 17.98,
|
|
"eval_steps_per_second": 2.267,
|
|
"eval_wer": 0.1110054670786784,
|
|
"step": 1290
|
|
},
|
|
{
|
|
"epoch": 15.12,
|
|
"learning_rate": 2.5479069767441862e-05,
|
|
"loss": 0.3335,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 16.0,
|
|
"eval_cer": 0.0327517736825666,
|
|
"eval_loss": 0.17586949467658997,
|
|
"eval_runtime": 19.1261,
|
|
"eval_samples_per_second": 17.829,
|
|
"eval_steps_per_second": 2.248,
|
|
"eval_wer": 0.1064891846921797,
|
|
"step": 1376
|
|
},
|
|
{
|
|
"epoch": 16.28,
|
|
"learning_rate": 2.5130232558139536e-05,
|
|
"loss": 0.3259,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 17.0,
|
|
"eval_cer": 0.031011556824773547,
|
|
"eval_loss": 0.17175641655921936,
|
|
"eval_runtime": 18.8952,
|
|
"eval_samples_per_second": 18.047,
|
|
"eval_steps_per_second": 2.276,
|
|
"eval_wer": 0.0941288328975517,
|
|
"step": 1462
|
|
},
|
|
{
|
|
"epoch": 17.44,
|
|
"learning_rate": 2.478139534883721e-05,
|
|
"loss": 0.2822,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 18.0,
|
|
"eval_cer": 0.03074383115434385,
|
|
"eval_loss": 0.1663752645254135,
|
|
"eval_runtime": 19.0306,
|
|
"eval_samples_per_second": 17.919,
|
|
"eval_steps_per_second": 2.26,
|
|
"eval_wer": 0.09650582362728785,
|
|
"step": 1548
|
|
},
|
|
{
|
|
"epoch": 18.6,
|
|
"learning_rate": 2.4432558139534884e-05,
|
|
"loss": 0.2689,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 19.0,
|
|
"eval_cer": 0.030163758868412834,
|
|
"eval_loss": 0.16375425457954407,
|
|
"eval_runtime": 19.1028,
|
|
"eval_samples_per_second": 17.851,
|
|
"eval_steps_per_second": 2.251,
|
|
"eval_wer": 0.09507962918944617,
|
|
"step": 1634
|
|
},
|
|
{
|
|
"epoch": 19.77,
|
|
"learning_rate": 2.4083720930232558e-05,
|
|
"loss": 0.262,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"epoch": 20.0,
|
|
"eval_cer": 0.030699210209272233,
|
|
"eval_loss": 0.1615932136774063,
|
|
"eval_runtime": 18.925,
|
|
"eval_samples_per_second": 18.019,
|
|
"eval_steps_per_second": 2.272,
|
|
"eval_wer": 0.09365343475160447,
|
|
"step": 1720
|
|
},
|
|
{
|
|
"epoch": 20.93,
|
|
"learning_rate": 2.373488372093023e-05,
|
|
"loss": 0.2623,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"epoch": 21.0,
|
|
"eval_cer": 0.029896033197983132,
|
|
"eval_loss": 0.1573890894651413,
|
|
"eval_runtime": 18.8701,
|
|
"eval_samples_per_second": 18.071,
|
|
"eval_steps_per_second": 2.279,
|
|
"eval_wer": 0.09270263845971001,
|
|
"step": 1806
|
|
},
|
|
{
|
|
"epoch": 22.0,
|
|
"eval_cer": 0.030565347374057382,
|
|
"eval_loss": 0.15589450299739838,
|
|
"eval_runtime": 19.0532,
|
|
"eval_samples_per_second": 17.897,
|
|
"eval_steps_per_second": 2.257,
|
|
"eval_wer": 0.09436653197052532,
|
|
"step": 1892
|
|
},
|
|
{
|
|
"epoch": 22.09,
|
|
"learning_rate": 2.338604651162791e-05,
|
|
"loss": 0.2344,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"epoch": 23.0,
|
|
"eval_cer": 0.029315960912052116,
|
|
"eval_loss": 0.15381626784801483,
|
|
"eval_runtime": 19.0478,
|
|
"eval_samples_per_second": 17.902,
|
|
"eval_steps_per_second": 2.257,
|
|
"eval_wer": 0.09198954124078916,
|
|
"step": 1978
|
|
},
|
|
{
|
|
"epoch": 23.26,
|
|
"learning_rate": 2.3037209302325583e-05,
|
|
"loss": 0.2315,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 24.0,
|
|
"eval_cer": 0.0298067913078399,
|
|
"eval_loss": 0.1532616764307022,
|
|
"eval_runtime": 19.3337,
|
|
"eval_samples_per_second": 17.638,
|
|
"eval_steps_per_second": 2.224,
|
|
"eval_wer": 0.09270263845971001,
|
|
"step": 2064
|
|
},
|
|
{
|
|
"epoch": 24.42,
|
|
"learning_rate": 2.2688372093023257e-05,
|
|
"loss": 0.2411,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"epoch": 25.0,
|
|
"eval_cer": 0.029360581857123733,
|
|
"eval_loss": 0.15203775465488434,
|
|
"eval_runtime": 18.8831,
|
|
"eval_samples_per_second": 18.058,
|
|
"eval_steps_per_second": 2.277,
|
|
"eval_wer": 0.09080104587592108,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"epoch": 25.58,
|
|
"learning_rate": 2.233953488372093e-05,
|
|
"loss": 0.2102,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"epoch": 26.0,
|
|
"eval_cer": 0.02971754941769667,
|
|
"eval_loss": 0.14963804185390472,
|
|
"eval_runtime": 18.9765,
|
|
"eval_samples_per_second": 17.97,
|
|
"eval_steps_per_second": 2.266,
|
|
"eval_wer": 0.09222724031376278,
|
|
"step": 2236
|
|
},
|
|
{
|
|
"epoch": 26.74,
|
|
"learning_rate": 2.1990697674418605e-05,
|
|
"loss": 0.2115,
|
|
"step": 2300
|
|
},
|
|
{
|
|
"epoch": 27.0,
|
|
"eval_cer": 0.029762170362768285,
|
|
"eval_loss": 0.1473499834537506,
|
|
"eval_runtime": 18.9863,
|
|
"eval_samples_per_second": 17.96,
|
|
"eval_steps_per_second": 2.265,
|
|
"eval_wer": 0.09793201806512955,
|
|
"step": 2322
|
|
},
|
|
{
|
|
"epoch": 27.91,
|
|
"learning_rate": 2.1641860465116282e-05,
|
|
"loss": 0.2139,
|
|
"step": 2400
|
|
},
|
|
{
|
|
"epoch": 28.0,
|
|
"eval_cer": 0.029985275088126366,
|
|
"eval_loss": 0.14958611130714417,
|
|
"eval_runtime": 18.9624,
|
|
"eval_samples_per_second": 17.983,
|
|
"eval_steps_per_second": 2.268,
|
|
"eval_wer": 0.095792726408367,
|
|
"step": 2408
|
|
},
|
|
{
|
|
"epoch": 29.0,
|
|
"eval_cer": 0.02886975146133595,
|
|
"eval_loss": 0.14455682039260864,
|
|
"eval_runtime": 19.1059,
|
|
"eval_samples_per_second": 17.848,
|
|
"eval_steps_per_second": 2.251,
|
|
"eval_wer": 0.09032564772997385,
|
|
"step": 2494
|
|
},
|
|
{
|
|
"epoch": 29.07,
|
|
"learning_rate": 2.1293023255813956e-05,
|
|
"loss": 0.2218,
|
|
"step": 2500
|
|
},
|
|
{
|
|
"epoch": 30.0,
|
|
"eval_cer": 0.029137477131765652,
|
|
"eval_loss": 0.1481563001871109,
|
|
"eval_runtime": 19.0476,
|
|
"eval_samples_per_second": 17.903,
|
|
"eval_steps_per_second": 2.258,
|
|
"eval_wer": 0.09127644402186831,
|
|
"step": 2580
|
|
},
|
|
{
|
|
"epoch": 30.23,
|
|
"learning_rate": 2.094418604651163e-05,
|
|
"loss": 0.1982,
|
|
"step": 2600
|
|
},
|
|
{
|
|
"epoch": 31.0,
|
|
"eval_cer": 0.02837892106554817,
|
|
"eval_loss": 0.14385315775871277,
|
|
"eval_runtime": 18.9549,
|
|
"eval_samples_per_second": 17.99,
|
|
"eval_steps_per_second": 2.269,
|
|
"eval_wer": 0.08913715236510578,
|
|
"step": 2666
|
|
},
|
|
{
|
|
"epoch": 31.4,
|
|
"learning_rate": 2.0595348837209304e-05,
|
|
"loss": 0.196,
|
|
"step": 2700
|
|
},
|
|
{
|
|
"epoch": 32.0,
|
|
"eval_cer": 0.02860202579090625,
|
|
"eval_loss": 0.14276552200317383,
|
|
"eval_runtime": 19.1086,
|
|
"eval_samples_per_second": 17.845,
|
|
"eval_steps_per_second": 2.25,
|
|
"eval_wer": 0.09341573567863086,
|
|
"step": 2752
|
|
},
|
|
{
|
|
"epoch": 32.56,
|
|
"learning_rate": 2.0246511627906977e-05,
|
|
"loss": 0.1846,
|
|
"step": 2800
|
|
},
|
|
{
|
|
"epoch": 33.0,
|
|
"eval_cer": 0.028111195395118468,
|
|
"eval_loss": 0.14039289951324463,
|
|
"eval_runtime": 19.1208,
|
|
"eval_samples_per_second": 17.834,
|
|
"eval_steps_per_second": 2.249,
|
|
"eval_wer": 0.09056334680294748,
|
|
"step": 2838
|
|
},
|
|
{
|
|
"epoch": 33.72,
|
|
"learning_rate": 1.9897674418604655e-05,
|
|
"loss": 0.1873,
|
|
"step": 2900
|
|
},
|
|
{
|
|
"epoch": 34.0,
|
|
"eval_cer": 0.02717415554861452,
|
|
"eval_loss": 0.1395137906074524,
|
|
"eval_runtime": 19.1432,
|
|
"eval_samples_per_second": 17.813,
|
|
"eval_steps_per_second": 2.246,
|
|
"eval_wer": 0.08557166627050154,
|
|
"step": 2924
|
|
},
|
|
{
|
|
"epoch": 34.88,
|
|
"learning_rate": 1.954883720930233e-05,
|
|
"loss": 0.1854,
|
|
"step": 3000
|
|
},
|
|
{
|
|
"epoch": 35.0,
|
|
"eval_cer": 0.02677256704296997,
|
|
"eval_loss": 0.1353740692138672,
|
|
"eval_runtime": 19.2143,
|
|
"eval_samples_per_second": 17.747,
|
|
"eval_steps_per_second": 2.238,
|
|
"eval_wer": 0.08462086997860709,
|
|
"step": 3010
|
|
},
|
|
{
|
|
"epoch": 36.0,
|
|
"eval_cer": 0.027129534603542903,
|
|
"eval_loss": 0.1359284222126007,
|
|
"eval_runtime": 19.183,
|
|
"eval_samples_per_second": 17.776,
|
|
"eval_steps_per_second": 2.242,
|
|
"eval_wer": 0.08580936534347516,
|
|
"step": 3096
|
|
},
|
|
{
|
|
"epoch": 36.05,
|
|
"learning_rate": 1.9200000000000003e-05,
|
|
"loss": 0.1851,
|
|
"step": 3100
|
|
},
|
|
{
|
|
"epoch": 37.0,
|
|
"eval_cer": 0.027620364999330686,
|
|
"eval_loss": 0.13563700020313263,
|
|
"eval_runtime": 19.0319,
|
|
"eval_samples_per_second": 17.917,
|
|
"eval_steps_per_second": 2.259,
|
|
"eval_wer": 0.09032564772997385,
|
|
"step": 3182
|
|
},
|
|
{
|
|
"epoch": 37.21,
|
|
"learning_rate": 1.8851162790697673e-05,
|
|
"loss": 0.1733,
|
|
"step": 3200
|
|
},
|
|
{
|
|
"epoch": 38.0,
|
|
"eval_cer": 0.02704029271339967,
|
|
"eval_loss": 0.13802382349967957,
|
|
"eval_runtime": 19.2183,
|
|
"eval_samples_per_second": 17.744,
|
|
"eval_steps_per_second": 2.237,
|
|
"eval_wer": 0.08509626812455431,
|
|
"step": 3268
|
|
},
|
|
{
|
|
"epoch": 38.37,
|
|
"learning_rate": 1.8502325581395347e-05,
|
|
"loss": 0.1781,
|
|
"step": 3300
|
|
},
|
|
{
|
|
"epoch": 39.0,
|
|
"eval_cer": 0.02583552719646602,
|
|
"eval_loss": 0.13587680459022522,
|
|
"eval_runtime": 19.0209,
|
|
"eval_samples_per_second": 17.928,
|
|
"eval_steps_per_second": 2.261,
|
|
"eval_wer": 0.08129308295697647,
|
|
"step": 3354
|
|
},
|
|
{
|
|
"epoch": 39.53,
|
|
"learning_rate": 1.815348837209302e-05,
|
|
"loss": 0.1821,
|
|
"step": 3400
|
|
},
|
|
{
|
|
"epoch": 40.0,
|
|
"eval_cer": 0.027441881219044218,
|
|
"eval_loss": 0.13828882575035095,
|
|
"eval_runtime": 19.1283,
|
|
"eval_samples_per_second": 17.827,
|
|
"eval_steps_per_second": 2.248,
|
|
"eval_wer": 0.08628476348942239,
|
|
"step": 3440
|
|
},
|
|
{
|
|
"epoch": 40.7,
|
|
"learning_rate": 1.7804651162790698e-05,
|
|
"loss": 0.1724,
|
|
"step": 3500
|
|
},
|
|
{
|
|
"epoch": 41.0,
|
|
"eval_cer": 0.02650484137254027,
|
|
"eval_loss": 0.13766087591648102,
|
|
"eval_runtime": 19.2872,
|
|
"eval_samples_per_second": 17.68,
|
|
"eval_steps_per_second": 2.229,
|
|
"eval_wer": 0.08224387924887093,
|
|
"step": 3526
|
|
},
|
|
{
|
|
"epoch": 41.86,
|
|
"learning_rate": 1.7455813953488372e-05,
|
|
"loss": 0.1591,
|
|
"step": 3600
|
|
},
|
|
{
|
|
"epoch": 42.0,
|
|
"eval_cer": 0.026594083262683504,
|
|
"eval_loss": 0.13794943690299988,
|
|
"eval_runtime": 19.0157,
|
|
"eval_samples_per_second": 17.933,
|
|
"eval_steps_per_second": 2.261,
|
|
"eval_wer": 0.08438317090563346,
|
|
"step": 3612
|
|
},
|
|
{
|
|
"epoch": 43.0,
|
|
"eval_cer": 0.026103252866895722,
|
|
"eval_loss": 0.1349739283323288,
|
|
"eval_runtime": 18.9996,
|
|
"eval_samples_per_second": 17.948,
|
|
"eval_steps_per_second": 2.263,
|
|
"eval_wer": 0.08414547183265986,
|
|
"step": 3698
|
|
},
|
|
{
|
|
"epoch": 43.02,
|
|
"learning_rate": 1.7106976744186046e-05,
|
|
"loss": 0.168,
|
|
"step": 3700
|
|
},
|
|
{
|
|
"epoch": 44.0,
|
|
"eval_cer": 0.02650484137254027,
|
|
"eval_loss": 0.1362057477235794,
|
|
"eval_runtime": 19.2987,
|
|
"eval_samples_per_second": 17.67,
|
|
"eval_steps_per_second": 2.228,
|
|
"eval_wer": 0.08462086997860709,
|
|
"step": 3784
|
|
},
|
|
{
|
|
"epoch": 44.19,
|
|
"learning_rate": 1.675813953488372e-05,
|
|
"loss": 0.162,
|
|
"step": 3800
|
|
},
|
|
{
|
|
"epoch": 45.0,
|
|
"eval_cer": 0.02663870420775512,
|
|
"eval_loss": 0.13529056310653687,
|
|
"eval_runtime": 19.8929,
|
|
"eval_samples_per_second": 17.142,
|
|
"eval_steps_per_second": 2.162,
|
|
"eval_wer": 0.08200618017589731,
|
|
"step": 3870
|
|
},
|
|
{
|
|
"epoch": 45.35,
|
|
"learning_rate": 1.6409302325581394e-05,
|
|
"loss": 0.145,
|
|
"step": 3900
|
|
},
|
|
{
|
|
"epoch": 46.0,
|
|
"eval_cer": 0.026995671768328053,
|
|
"eval_loss": 0.14138084650039673,
|
|
"eval_runtime": 19.2935,
|
|
"eval_samples_per_second": 17.674,
|
|
"eval_steps_per_second": 2.229,
|
|
"eval_wer": 0.08485856905158069,
|
|
"step": 3956
|
|
},
|
|
{
|
|
"epoch": 46.51,
|
|
"learning_rate": 1.606046511627907e-05,
|
|
"loss": 0.1639,
|
|
"step": 4000
|
|
},
|
|
{
|
|
"epoch": 47.0,
|
|
"eval_cer": 0.026594083262683504,
|
|
"eval_loss": 0.13999956846237183,
|
|
"eval_runtime": 19.0568,
|
|
"eval_samples_per_second": 17.894,
|
|
"eval_steps_per_second": 2.256,
|
|
"eval_wer": 0.08367007368671261,
|
|
"step": 4042
|
|
},
|
|
{
|
|
"epoch": 47.67,
|
|
"learning_rate": 1.5711627906976745e-05,
|
|
"loss": 0.1572,
|
|
"step": 4100
|
|
},
|
|
{
|
|
"epoch": 48.0,
|
|
"eval_cer": 0.026058631921824105,
|
|
"eval_loss": 0.1387248933315277,
|
|
"eval_runtime": 19.0666,
|
|
"eval_samples_per_second": 17.885,
|
|
"eval_steps_per_second": 2.255,
|
|
"eval_wer": 0.08153078202995008,
|
|
"step": 4128
|
|
},
|
|
{
|
|
"epoch": 48.84,
|
|
"learning_rate": 1.536279069767442e-05,
|
|
"loss": 0.14,
|
|
"step": 4200
|
|
},
|
|
{
|
|
"epoch": 49.0,
|
|
"eval_cer": 0.02583552719646602,
|
|
"eval_loss": 0.1383592039346695,
|
|
"eval_runtime": 19.025,
|
|
"eval_samples_per_second": 17.924,
|
|
"eval_steps_per_second": 2.26,
|
|
"eval_wer": 0.08224387924887093,
|
|
"step": 4214
|
|
},
|
|
{
|
|
"epoch": 50.0,
|
|
"learning_rate": 1.5013953488372093e-05,
|
|
"loss": 0.1473,
|
|
"step": 4300
|
|
},
|
|
{
|
|
"epoch": 50.0,
|
|
"eval_cer": 0.026549462317611888,
|
|
"eval_loss": 0.13988161087036133,
|
|
"eval_runtime": 19.1678,
|
|
"eval_samples_per_second": 17.79,
|
|
"eval_steps_per_second": 2.243,
|
|
"eval_wer": 0.08271927739481816,
|
|
"step": 4300
|
|
},
|
|
{
|
|
"epoch": 51.0,
|
|
"eval_cer": 0.02650484137254027,
|
|
"eval_loss": 0.14077438414096832,
|
|
"eval_runtime": 19.0777,
|
|
"eval_samples_per_second": 17.874,
|
|
"eval_steps_per_second": 2.254,
|
|
"eval_wer": 0.08224387924887093,
|
|
"step": 4386
|
|
},
|
|
{
|
|
"epoch": 51.16,
|
|
"learning_rate": 1.4665116279069768e-05,
|
|
"loss": 0.1722,
|
|
"step": 4400
|
|
},
|
|
{
|
|
"epoch": 52.0,
|
|
"eval_cer": 0.026683325152826738,
|
|
"eval_loss": 0.1410902589559555,
|
|
"eval_runtime": 19.0905,
|
|
"eval_samples_per_second": 17.862,
|
|
"eval_steps_per_second": 2.252,
|
|
"eval_wer": 0.08509626812455431,
|
|
"step": 4472
|
|
},
|
|
{
|
|
"epoch": 52.33,
|
|
"learning_rate": 1.4316279069767442e-05,
|
|
"loss": 0.1518,
|
|
"step": 4500
|
|
},
|
|
{
|
|
"epoch": 53.0,
|
|
"eval_cer": 0.02730801838382937,
|
|
"eval_loss": 0.14102402329444885,
|
|
"eval_runtime": 19.0143,
|
|
"eval_samples_per_second": 17.934,
|
|
"eval_steps_per_second": 2.261,
|
|
"eval_wer": 0.08604706441644877,
|
|
"step": 4558
|
|
},
|
|
{
|
|
"epoch": 53.49,
|
|
"learning_rate": 1.3967441860465116e-05,
|
|
"loss": 0.1506,
|
|
"step": 4600
|
|
},
|
|
{
|
|
"epoch": 54.0,
|
|
"eval_cer": 0.02704029271339967,
|
|
"eval_loss": 0.14034318923950195,
|
|
"eval_runtime": 19.0734,
|
|
"eval_samples_per_second": 17.878,
|
|
"eval_steps_per_second": 2.254,
|
|
"eval_wer": 0.08580936534347516,
|
|
"step": 4644
|
|
},
|
|
{
|
|
"epoch": 54.65,
|
|
"learning_rate": 1.3618604651162792e-05,
|
|
"loss": 0.1442,
|
|
"step": 4700
|
|
},
|
|
{
|
|
"epoch": 55.0,
|
|
"eval_cer": 0.02623711570211057,
|
|
"eval_loss": 0.1380011886358261,
|
|
"eval_runtime": 19.092,
|
|
"eval_samples_per_second": 17.861,
|
|
"eval_steps_per_second": 2.252,
|
|
"eval_wer": 0.08319467554076539,
|
|
"step": 4730
|
|
},
|
|
{
|
|
"epoch": 55.81,
|
|
"learning_rate": 1.3269767441860466e-05,
|
|
"loss": 0.1427,
|
|
"step": 4800
|
|
},
|
|
{
|
|
"epoch": 56.0,
|
|
"eval_cer": 0.026549462317611888,
|
|
"eval_loss": 0.13952679932117462,
|
|
"eval_runtime": 19.0512,
|
|
"eval_samples_per_second": 17.899,
|
|
"eval_steps_per_second": 2.257,
|
|
"eval_wer": 0.08153078202995008,
|
|
"step": 4816
|
|
},
|
|
{
|
|
"epoch": 56.98,
|
|
"learning_rate": 1.292093023255814e-05,
|
|
"loss": 0.132,
|
|
"step": 4900
|
|
},
|
|
{
|
|
"epoch": 57.0,
|
|
"eval_cer": 0.025924769086609255,
|
|
"eval_loss": 0.13835428655147552,
|
|
"eval_runtime": 19.1303,
|
|
"eval_samples_per_second": 17.825,
|
|
"eval_steps_per_second": 2.248,
|
|
"eval_wer": 0.08129308295697647,
|
|
"step": 4902
|
|
},
|
|
{
|
|
"epoch": 58.0,
|
|
"eval_cer": 0.026058631921824105,
|
|
"eval_loss": 0.1352924406528473,
|
|
"eval_runtime": 19.3096,
|
|
"eval_samples_per_second": 17.66,
|
|
"eval_steps_per_second": 2.227,
|
|
"eval_wer": 0.0817684811029237,
|
|
"step": 4988
|
|
},
|
|
{
|
|
"epoch": 58.14,
|
|
"learning_rate": 1.2572093023255815e-05,
|
|
"loss": 0.1374,
|
|
"step": 5000
|
|
},
|
|
{
|
|
"epoch": 59.0,
|
|
"eval_cer": 0.02650484137254027,
|
|
"eval_loss": 0.140494242310524,
|
|
"eval_runtime": 19.0811,
|
|
"eval_samples_per_second": 17.871,
|
|
"eval_steps_per_second": 2.254,
|
|
"eval_wer": 0.08343237461373901,
|
|
"step": 5074
|
|
},
|
|
{
|
|
"epoch": 59.3,
|
|
"learning_rate": 1.2223255813953489e-05,
|
|
"loss": 0.1318,
|
|
"step": 5100
|
|
},
|
|
{
|
|
"epoch": 60.0,
|
|
"eval_cer": 0.02583552719646602,
|
|
"eval_loss": 0.1383277028799057,
|
|
"eval_runtime": 19.2077,
|
|
"eval_samples_per_second": 17.753,
|
|
"eval_steps_per_second": 2.239,
|
|
"eval_wer": 0.08105538388400285,
|
|
"step": 5160
|
|
},
|
|
{
|
|
"epoch": 60.47,
|
|
"learning_rate": 1.1874418604651165e-05,
|
|
"loss": 0.1339,
|
|
"step": 5200
|
|
},
|
|
{
|
|
"epoch": 61.0,
|
|
"eval_cer": 0.026058631921824105,
|
|
"eval_loss": 0.1385628879070282,
|
|
"eval_runtime": 19.3076,
|
|
"eval_samples_per_second": 17.661,
|
|
"eval_steps_per_second": 2.227,
|
|
"eval_wer": 0.08129308295697647,
|
|
"step": 5246
|
|
},
|
|
{
|
|
"epoch": 61.63,
|
|
"learning_rate": 1.1525581395348838e-05,
|
|
"loss": 0.1346,
|
|
"step": 5300
|
|
},
|
|
{
|
|
"epoch": 62.0,
|
|
"eval_cer": 0.025255454910535005,
|
|
"eval_loss": 0.13994893431663513,
|
|
"eval_runtime": 19.1977,
|
|
"eval_samples_per_second": 17.763,
|
|
"eval_steps_per_second": 2.24,
|
|
"eval_wer": 0.0801045875921084,
|
|
"step": 5332
|
|
},
|
|
{
|
|
"epoch": 62.79,
|
|
"learning_rate": 1.117674418604651e-05,
|
|
"loss": 0.1261,
|
|
"step": 5400
|
|
},
|
|
{
|
|
"epoch": 63.0,
|
|
"eval_cer": 0.026326357592253803,
|
|
"eval_loss": 0.14302104711532593,
|
|
"eval_runtime": 19.1245,
|
|
"eval_samples_per_second": 17.831,
|
|
"eval_steps_per_second": 2.248,
|
|
"eval_wer": 0.08319467554076539,
|
|
"step": 5418
|
|
},
|
|
{
|
|
"epoch": 63.0,
|
|
"step": 5418,
|
|
"total_flos": 4.528287283105284e+19,
|
|
"train_loss": 0.5869265782432654,
|
|
"train_runtime": 15592.601,
|
|
"train_samples_per_second": 17.592,
|
|
"train_steps_per_second": 0.552
|
|
}
|
|
],
|
|
"max_steps": 8600,
|
|
"num_train_epochs": 100,
|
|
"total_flos": 4.528287283105284e+19,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|