678d6709bb
Model: alinerodrigues/wav2vec2-large-xlsr-mecita-coraa-portuguese-2-all-04 Source: Original Platform
1002 lines
25 KiB
JSON
1002 lines
25 KiB
JSON
{
|
|
"best_metric": 0.1832132190465927,
|
|
"best_model_checkpoint": "wav2vec2-large-xlsr-mecita-coraa-portuguese-2-all-04/checkpoint-3784",
|
|
"epoch": 64.0,
|
|
"global_step": 5504,
|
|
"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": 28.4874,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"eval_cer": 1.0,
|
|
"eval_loss": 3.3048207759857178,
|
|
"eval_runtime": 19.5313,
|
|
"eval_samples_per_second": 17.459,
|
|
"eval_steps_per_second": 2.202,
|
|
"eval_wer": 1.0,
|
|
"step": 86
|
|
},
|
|
{
|
|
"epoch": 1.16,
|
|
"learning_rate": 2.9661627906976745e-05,
|
|
"loss": 7.6481,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"eval_cer": 1.0,
|
|
"eval_loss": 2.9769535064697266,
|
|
"eval_runtime": 19.125,
|
|
"eval_samples_per_second": 17.83,
|
|
"eval_steps_per_second": 2.248,
|
|
"eval_wer": 1.0,
|
|
"step": 172
|
|
},
|
|
{
|
|
"epoch": 2.33,
|
|
"learning_rate": 2.931279069767442e-05,
|
|
"loss": 3.0297,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"eval_cer": 1.0,
|
|
"eval_loss": 2.9361252784729004,
|
|
"eval_runtime": 18.9766,
|
|
"eval_samples_per_second": 17.969,
|
|
"eval_steps_per_second": 2.266,
|
|
"eval_wer": 1.0,
|
|
"step": 258
|
|
},
|
|
{
|
|
"epoch": 3.49,
|
|
"learning_rate": 2.8963953488372093e-05,
|
|
"loss": 2.9192,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"eval_cer": 1.0,
|
|
"eval_loss": 2.9206249713897705,
|
|
"eval_runtime": 19.307,
|
|
"eval_samples_per_second": 17.662,
|
|
"eval_steps_per_second": 2.227,
|
|
"eval_wer": 1.0,
|
|
"step": 344
|
|
},
|
|
{
|
|
"epoch": 4.65,
|
|
"learning_rate": 2.861511627906977e-05,
|
|
"loss": 2.8826,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 5.0,
|
|
"eval_cer": 0.9992641986663601,
|
|
"eval_loss": 2.6372246742248535,
|
|
"eval_runtime": 19.1582,
|
|
"eval_samples_per_second": 17.799,
|
|
"eval_steps_per_second": 2.244,
|
|
"eval_wer": 0.9990157480314961,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 5.81,
|
|
"learning_rate": 2.8266279069767444e-05,
|
|
"loss": 2.3512,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 6.0,
|
|
"eval_cer": 0.11133593929638998,
|
|
"eval_loss": 0.8686644434928894,
|
|
"eval_runtime": 19.3918,
|
|
"eval_samples_per_second": 17.585,
|
|
"eval_steps_per_second": 2.217,
|
|
"eval_wer": 0.4357775590551181,
|
|
"step": 516
|
|
},
|
|
{
|
|
"epoch": 6.98,
|
|
"learning_rate": 2.7917441860465118e-05,
|
|
"loss": 1.0373,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 7.0,
|
|
"eval_cer": 0.06645205794435502,
|
|
"eval_loss": 0.5094822645187378,
|
|
"eval_runtime": 19.0781,
|
|
"eval_samples_per_second": 17.874,
|
|
"eval_steps_per_second": 2.254,
|
|
"eval_wer": 0.2418799212598425,
|
|
"step": 602
|
|
},
|
|
{
|
|
"epoch": 8.0,
|
|
"eval_cer": 0.05840423085766843,
|
|
"eval_loss": 0.40014487504959106,
|
|
"eval_runtime": 19.2136,
|
|
"eval_samples_per_second": 17.748,
|
|
"eval_steps_per_second": 2.238,
|
|
"eval_wer": 0.21013779527559054,
|
|
"step": 688
|
|
},
|
|
{
|
|
"epoch": 8.14,
|
|
"learning_rate": 2.7568604651162792e-05,
|
|
"loss": 0.6758,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 9.0,
|
|
"eval_cer": 0.05021844102092435,
|
|
"eval_loss": 0.3354460895061493,
|
|
"eval_runtime": 19.074,
|
|
"eval_samples_per_second": 17.878,
|
|
"eval_steps_per_second": 2.254,
|
|
"eval_wer": 0.1798720472440945,
|
|
"step": 774
|
|
},
|
|
{
|
|
"epoch": 9.3,
|
|
"learning_rate": 2.7219767441860466e-05,
|
|
"loss": 0.5667,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 10.0,
|
|
"eval_cer": 0.04681535985283973,
|
|
"eval_loss": 0.2961549162864685,
|
|
"eval_runtime": 19.1133,
|
|
"eval_samples_per_second": 17.841,
|
|
"eval_steps_per_second": 2.25,
|
|
"eval_wer": 0.16166338582677164,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 10.47,
|
|
"learning_rate": 2.6870930232558143e-05,
|
|
"loss": 0.4908,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 11.0,
|
|
"eval_cer": 0.043044378017935155,
|
|
"eval_loss": 0.2820357084274292,
|
|
"eval_runtime": 19.2073,
|
|
"eval_samples_per_second": 17.754,
|
|
"eval_steps_per_second": 2.239,
|
|
"eval_wer": 0.14714566929133857,
|
|
"step": 946
|
|
},
|
|
{
|
|
"epoch": 11.63,
|
|
"learning_rate": 2.6522093023255817e-05,
|
|
"loss": 0.4084,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 12.0,
|
|
"eval_cer": 0.039871234766613015,
|
|
"eval_loss": 0.25810757279396057,
|
|
"eval_runtime": 19.5082,
|
|
"eval_samples_per_second": 17.48,
|
|
"eval_steps_per_second": 2.204,
|
|
"eval_wer": 0.13508858267716536,
|
|
"step": 1032
|
|
},
|
|
{
|
|
"epoch": 12.79,
|
|
"learning_rate": 2.617325581395349e-05,
|
|
"loss": 0.3728,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 13.0,
|
|
"eval_cer": 0.037387905265578295,
|
|
"eval_loss": 0.2444266527891159,
|
|
"eval_runtime": 19.0415,
|
|
"eval_samples_per_second": 17.908,
|
|
"eval_steps_per_second": 2.258,
|
|
"eval_wer": 0.12303149606299213,
|
|
"step": 1118
|
|
},
|
|
{
|
|
"epoch": 13.95,
|
|
"learning_rate": 2.5824418604651165e-05,
|
|
"loss": 0.3577,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 14.0,
|
|
"eval_cer": 0.035502414348126,
|
|
"eval_loss": 0.24084249138832092,
|
|
"eval_runtime": 19.2097,
|
|
"eval_samples_per_second": 17.751,
|
|
"eval_steps_per_second": 2.238,
|
|
"eval_wer": 0.11441929133858268,
|
|
"step": 1204
|
|
},
|
|
{
|
|
"epoch": 15.0,
|
|
"eval_cer": 0.03398482409749368,
|
|
"eval_loss": 0.23529472947120667,
|
|
"eval_runtime": 19.9753,
|
|
"eval_samples_per_second": 17.071,
|
|
"eval_steps_per_second": 2.153,
|
|
"eval_wer": 0.10999015748031496,
|
|
"step": 1290
|
|
},
|
|
{
|
|
"epoch": 15.12,
|
|
"learning_rate": 2.547558139534884e-05,
|
|
"loss": 0.3237,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 16.0,
|
|
"eval_cer": 0.03324902276385376,
|
|
"eval_loss": 0.22266538441181183,
|
|
"eval_runtime": 19.26,
|
|
"eval_samples_per_second": 17.705,
|
|
"eval_steps_per_second": 2.233,
|
|
"eval_wer": 0.10359251968503937,
|
|
"step": 1376
|
|
},
|
|
{
|
|
"epoch": 16.28,
|
|
"learning_rate": 2.512674418604651e-05,
|
|
"loss": 0.3043,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 17.0,
|
|
"eval_cer": 0.032191308346746376,
|
|
"eval_loss": 0.22414861619472504,
|
|
"eval_runtime": 19.2388,
|
|
"eval_samples_per_second": 17.725,
|
|
"eval_steps_per_second": 2.235,
|
|
"eval_wer": 0.10039370078740158,
|
|
"step": 1462
|
|
},
|
|
{
|
|
"epoch": 17.44,
|
|
"learning_rate": 2.4777906976744186e-05,
|
|
"loss": 0.2876,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 18.0,
|
|
"eval_cer": 0.0329271096803863,
|
|
"eval_loss": 0.22596892714500427,
|
|
"eval_runtime": 19.2615,
|
|
"eval_samples_per_second": 17.704,
|
|
"eval_steps_per_second": 2.232,
|
|
"eval_wer": 0.10482283464566929,
|
|
"step": 1548
|
|
},
|
|
{
|
|
"epoch": 18.6,
|
|
"learning_rate": 2.442906976744186e-05,
|
|
"loss": 0.2842,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 19.0,
|
|
"eval_cer": 0.032513221430213846,
|
|
"eval_loss": 0.20861704647541046,
|
|
"eval_runtime": 19.253,
|
|
"eval_samples_per_second": 17.711,
|
|
"eval_steps_per_second": 2.233,
|
|
"eval_wer": 0.10703740157480315,
|
|
"step": 1634
|
|
},
|
|
{
|
|
"epoch": 19.77,
|
|
"learning_rate": 2.4080232558139534e-05,
|
|
"loss": 0.269,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"epoch": 20.0,
|
|
"eval_cer": 0.03237525868015636,
|
|
"eval_loss": 0.21557514369487762,
|
|
"eval_runtime": 19.2787,
|
|
"eval_samples_per_second": 17.688,
|
|
"eval_steps_per_second": 2.23,
|
|
"eval_wer": 0.10063976377952756,
|
|
"step": 1720
|
|
},
|
|
{
|
|
"epoch": 20.93,
|
|
"learning_rate": 2.3731395348837208e-05,
|
|
"loss": 0.2396,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"epoch": 21.0,
|
|
"eval_cer": 0.03067371809611405,
|
|
"eval_loss": 0.20651517808437347,
|
|
"eval_runtime": 19.1026,
|
|
"eval_samples_per_second": 17.851,
|
|
"eval_steps_per_second": 2.251,
|
|
"eval_wer": 0.0952263779527559,
|
|
"step": 1806
|
|
},
|
|
{
|
|
"epoch": 22.0,
|
|
"eval_cer": 0.03117958151299149,
|
|
"eval_loss": 0.21441449224948883,
|
|
"eval_runtime": 19.5818,
|
|
"eval_samples_per_second": 17.414,
|
|
"eval_steps_per_second": 2.196,
|
|
"eval_wer": 0.09596456692913385,
|
|
"step": 1892
|
|
},
|
|
{
|
|
"epoch": 22.09,
|
|
"learning_rate": 2.3382558139534882e-05,
|
|
"loss": 0.2411,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"epoch": 23.0,
|
|
"eval_cer": 0.030719705679466545,
|
|
"eval_loss": 0.20180490612983704,
|
|
"eval_runtime": 19.3167,
|
|
"eval_samples_per_second": 17.653,
|
|
"eval_steps_per_second": 2.226,
|
|
"eval_wer": 0.09325787401574803,
|
|
"step": 1978
|
|
},
|
|
{
|
|
"epoch": 23.26,
|
|
"learning_rate": 2.303372093023256e-05,
|
|
"loss": 0.2357,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 24.0,
|
|
"eval_cer": 0.030535755346056563,
|
|
"eval_loss": 0.20546293258666992,
|
|
"eval_runtime": 19.3401,
|
|
"eval_samples_per_second": 17.632,
|
|
"eval_steps_per_second": 2.223,
|
|
"eval_wer": 0.09399606299212598,
|
|
"step": 2064
|
|
},
|
|
{
|
|
"epoch": 24.42,
|
|
"learning_rate": 2.2684883720930233e-05,
|
|
"loss": 0.2268,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"epoch": 25.0,
|
|
"eval_cer": 0.03067371809611405,
|
|
"eval_loss": 0.20989486575126648,
|
|
"eval_runtime": 19.2814,
|
|
"eval_samples_per_second": 17.685,
|
|
"eval_steps_per_second": 2.23,
|
|
"eval_wer": 0.09276574803149606,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"epoch": 25.58,
|
|
"learning_rate": 2.2336046511627907e-05,
|
|
"loss": 0.2053,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"epoch": 26.0,
|
|
"eval_cer": 0.030167854679236605,
|
|
"eval_loss": 0.204420268535614,
|
|
"eval_runtime": 19.0634,
|
|
"eval_samples_per_second": 17.888,
|
|
"eval_steps_per_second": 2.256,
|
|
"eval_wer": 0.09325787401574803,
|
|
"step": 2236
|
|
},
|
|
{
|
|
"epoch": 26.74,
|
|
"learning_rate": 2.198720930232558e-05,
|
|
"loss": 0.2188,
|
|
"step": 2300
|
|
},
|
|
{
|
|
"epoch": 27.0,
|
|
"eval_cer": 0.03012186709588411,
|
|
"eval_loss": 0.20365406572818756,
|
|
"eval_runtime": 19.1283,
|
|
"eval_samples_per_second": 17.827,
|
|
"eval_steps_per_second": 2.248,
|
|
"eval_wer": 0.09325787401574803,
|
|
"step": 2322
|
|
},
|
|
{
|
|
"epoch": 27.91,
|
|
"learning_rate": 2.1638372093023255e-05,
|
|
"loss": 0.2152,
|
|
"step": 2400
|
|
},
|
|
{
|
|
"epoch": 28.0,
|
|
"eval_cer": 0.029799954012416648,
|
|
"eval_loss": 0.2068387269973755,
|
|
"eval_runtime": 19.1207,
|
|
"eval_samples_per_second": 17.834,
|
|
"eval_steps_per_second": 2.249,
|
|
"eval_wer": 0.09325787401574803,
|
|
"step": 2408
|
|
},
|
|
{
|
|
"epoch": 29.0,
|
|
"eval_cer": 0.028282363761784318,
|
|
"eval_loss": 0.20444060862064362,
|
|
"eval_runtime": 19.0848,
|
|
"eval_samples_per_second": 17.868,
|
|
"eval_steps_per_second": 2.253,
|
|
"eval_wer": 0.08981299212598425,
|
|
"step": 2494
|
|
},
|
|
{
|
|
"epoch": 29.07,
|
|
"learning_rate": 2.128953488372093e-05,
|
|
"loss": 0.1998,
|
|
"step": 2500
|
|
},
|
|
{
|
|
"epoch": 30.0,
|
|
"eval_cer": 0.029386065762244194,
|
|
"eval_loss": 0.20385394990444183,
|
|
"eval_runtime": 19.6397,
|
|
"eval_samples_per_second": 17.363,
|
|
"eval_steps_per_second": 2.189,
|
|
"eval_wer": 0.08882874015748031,
|
|
"step": 2580
|
|
},
|
|
{
|
|
"epoch": 30.23,
|
|
"learning_rate": 2.0940697674418606e-05,
|
|
"loss": 0.1934,
|
|
"step": 2600
|
|
},
|
|
{
|
|
"epoch": 31.0,
|
|
"eval_cer": 0.029570016095654173,
|
|
"eval_loss": 0.1964925080537796,
|
|
"eval_runtime": 19.4249,
|
|
"eval_samples_per_second": 17.555,
|
|
"eval_steps_per_second": 2.214,
|
|
"eval_wer": 0.09153543307086615,
|
|
"step": 2666
|
|
},
|
|
{
|
|
"epoch": 31.4,
|
|
"learning_rate": 2.059186046511628e-05,
|
|
"loss": 0.19,
|
|
"step": 2700
|
|
},
|
|
{
|
|
"epoch": 32.0,
|
|
"eval_cer": 0.02906415267877673,
|
|
"eval_loss": 0.19938160479068756,
|
|
"eval_runtime": 19.2102,
|
|
"eval_samples_per_second": 17.751,
|
|
"eval_steps_per_second": 2.238,
|
|
"eval_wer": 0.0907972440944882,
|
|
"step": 2752
|
|
},
|
|
{
|
|
"epoch": 32.56,
|
|
"learning_rate": 2.0243023255813954e-05,
|
|
"loss": 0.2108,
|
|
"step": 2800
|
|
},
|
|
{
|
|
"epoch": 33.0,
|
|
"eval_cer": 0.029202115428834215,
|
|
"eval_loss": 0.20258696377277374,
|
|
"eval_runtime": 19.1989,
|
|
"eval_samples_per_second": 17.761,
|
|
"eval_steps_per_second": 2.24,
|
|
"eval_wer": 0.08882874015748031,
|
|
"step": 2838
|
|
},
|
|
{
|
|
"epoch": 33.72,
|
|
"learning_rate": 1.9894186046511628e-05,
|
|
"loss": 0.1922,
|
|
"step": 2900
|
|
},
|
|
{
|
|
"epoch": 34.0,
|
|
"eval_cer": 0.02924810301218671,
|
|
"eval_loss": 0.1956785023212433,
|
|
"eval_runtime": 19.1262,
|
|
"eval_samples_per_second": 17.829,
|
|
"eval_steps_per_second": 2.248,
|
|
"eval_wer": 0.09055118110236221,
|
|
"step": 2924
|
|
},
|
|
{
|
|
"epoch": 34.88,
|
|
"learning_rate": 1.9545348837209302e-05,
|
|
"loss": 0.1827,
|
|
"step": 3000
|
|
},
|
|
{
|
|
"epoch": 35.0,
|
|
"eval_cer": 0.028742239595309266,
|
|
"eval_loss": 0.19348372519016266,
|
|
"eval_runtime": 19.2699,
|
|
"eval_samples_per_second": 17.696,
|
|
"eval_steps_per_second": 2.231,
|
|
"eval_wer": 0.08932086614173228,
|
|
"step": 3010
|
|
},
|
|
{
|
|
"epoch": 36.0,
|
|
"eval_cer": 0.029524028512301678,
|
|
"eval_loss": 0.19330757856369019,
|
|
"eval_runtime": 19.3036,
|
|
"eval_samples_per_second": 17.665,
|
|
"eval_steps_per_second": 2.228,
|
|
"eval_wer": 0.09055118110236221,
|
|
"step": 3096
|
|
},
|
|
{
|
|
"epoch": 36.05,
|
|
"learning_rate": 1.919651162790698e-05,
|
|
"loss": 0.1848,
|
|
"step": 3100
|
|
},
|
|
{
|
|
"epoch": 37.0,
|
|
"eval_cer": 0.02869625201195677,
|
|
"eval_loss": 0.19993430376052856,
|
|
"eval_runtime": 19.2145,
|
|
"eval_samples_per_second": 17.747,
|
|
"eval_steps_per_second": 2.238,
|
|
"eval_wer": 0.08833661417322834,
|
|
"step": 3182
|
|
},
|
|
{
|
|
"epoch": 37.21,
|
|
"learning_rate": 1.8847674418604653e-05,
|
|
"loss": 0.1807,
|
|
"step": 3200
|
|
},
|
|
{
|
|
"epoch": 38.0,
|
|
"eval_cer": 0.02869625201195677,
|
|
"eval_loss": 0.1949012279510498,
|
|
"eval_runtime": 19.2564,
|
|
"eval_samples_per_second": 17.708,
|
|
"eval_steps_per_second": 2.233,
|
|
"eval_wer": 0.08809055118110236,
|
|
"step": 3268
|
|
},
|
|
{
|
|
"epoch": 38.37,
|
|
"learning_rate": 1.8498837209302327e-05,
|
|
"loss": 0.1695,
|
|
"step": 3300
|
|
},
|
|
{
|
|
"epoch": 39.0,
|
|
"eval_cer": 0.029478040928949185,
|
|
"eval_loss": 0.1967935413122177,
|
|
"eval_runtime": 19.3052,
|
|
"eval_samples_per_second": 17.664,
|
|
"eval_steps_per_second": 2.227,
|
|
"eval_wer": 0.09055118110236221,
|
|
"step": 3354
|
|
},
|
|
{
|
|
"epoch": 39.53,
|
|
"learning_rate": 1.815e-05,
|
|
"loss": 0.1764,
|
|
"step": 3400
|
|
},
|
|
{
|
|
"epoch": 40.0,
|
|
"eval_cer": 0.02837433892848931,
|
|
"eval_loss": 0.18935783207416534,
|
|
"eval_runtime": 19.4348,
|
|
"eval_samples_per_second": 17.546,
|
|
"eval_steps_per_second": 2.213,
|
|
"eval_wer": 0.08636811023622047,
|
|
"step": 3440
|
|
},
|
|
{
|
|
"epoch": 40.7,
|
|
"learning_rate": 1.7801162790697675e-05,
|
|
"loss": 0.1812,
|
|
"step": 3500
|
|
},
|
|
{
|
|
"epoch": 41.0,
|
|
"eval_cer": 0.028558289261899288,
|
|
"eval_loss": 0.19317786395549774,
|
|
"eval_runtime": 19.3277,
|
|
"eval_samples_per_second": 17.643,
|
|
"eval_steps_per_second": 2.225,
|
|
"eval_wer": 0.08710629921259842,
|
|
"step": 3526
|
|
},
|
|
{
|
|
"epoch": 41.86,
|
|
"learning_rate": 1.7452325581395352e-05,
|
|
"loss": 0.1566,
|
|
"step": 3600
|
|
},
|
|
{
|
|
"epoch": 42.0,
|
|
"eval_cer": 0.02809841342837434,
|
|
"eval_loss": 0.18891684710979462,
|
|
"eval_runtime": 19.5468,
|
|
"eval_samples_per_second": 17.445,
|
|
"eval_steps_per_second": 2.2,
|
|
"eval_wer": 0.08686023622047244,
|
|
"step": 3612
|
|
},
|
|
{
|
|
"epoch": 43.0,
|
|
"eval_cer": 0.02819038859507933,
|
|
"eval_loss": 0.19649028778076172,
|
|
"eval_runtime": 19.4284,
|
|
"eval_samples_per_second": 17.552,
|
|
"eval_steps_per_second": 2.213,
|
|
"eval_wer": 0.08636811023622047,
|
|
"step": 3698
|
|
},
|
|
{
|
|
"epoch": 43.02,
|
|
"learning_rate": 1.7103488372093026e-05,
|
|
"loss": 0.161,
|
|
"step": 3700
|
|
},
|
|
{
|
|
"epoch": 44.0,
|
|
"eval_cer": 0.028558289261899288,
|
|
"eval_loss": 0.1832132190465927,
|
|
"eval_runtime": 19.3446,
|
|
"eval_samples_per_second": 17.628,
|
|
"eval_steps_per_second": 2.223,
|
|
"eval_wer": 0.08956692913385826,
|
|
"step": 3784
|
|
},
|
|
{
|
|
"epoch": 44.19,
|
|
"learning_rate": 1.67546511627907e-05,
|
|
"loss": 0.167,
|
|
"step": 3800
|
|
},
|
|
{
|
|
"epoch": 45.0,
|
|
"eval_cer": 0.028236376178431822,
|
|
"eval_loss": 0.19680964946746826,
|
|
"eval_runtime": 19.1457,
|
|
"eval_samples_per_second": 17.811,
|
|
"eval_steps_per_second": 2.246,
|
|
"eval_wer": 0.08858267716535433,
|
|
"step": 3870
|
|
},
|
|
{
|
|
"epoch": 45.35,
|
|
"learning_rate": 1.6405813953488374e-05,
|
|
"loss": 0.1556,
|
|
"step": 3900
|
|
},
|
|
{
|
|
"epoch": 46.0,
|
|
"eval_cer": 0.028328351345136813,
|
|
"eval_loss": 0.18351908028125763,
|
|
"eval_runtime": 19.2512,
|
|
"eval_samples_per_second": 17.713,
|
|
"eval_steps_per_second": 2.234,
|
|
"eval_wer": 0.08809055118110236,
|
|
"step": 3956
|
|
},
|
|
{
|
|
"epoch": 46.51,
|
|
"learning_rate": 1.6056976744186047e-05,
|
|
"loss": 0.157,
|
|
"step": 4000
|
|
},
|
|
{
|
|
"epoch": 47.0,
|
|
"eval_cer": 0.027868475511611864,
|
|
"eval_loss": 0.19117018580436707,
|
|
"eval_runtime": 19.4147,
|
|
"eval_samples_per_second": 17.564,
|
|
"eval_steps_per_second": 2.215,
|
|
"eval_wer": 0.08636811023622047,
|
|
"step": 4042
|
|
},
|
|
{
|
|
"epoch": 47.67,
|
|
"learning_rate": 1.570813953488372e-05,
|
|
"loss": 0.1588,
|
|
"step": 4100
|
|
},
|
|
{
|
|
"epoch": 48.0,
|
|
"eval_cer": 0.02791446309496436,
|
|
"eval_loss": 0.19263863563537598,
|
|
"eval_runtime": 19.509,
|
|
"eval_samples_per_second": 17.479,
|
|
"eval_steps_per_second": 2.204,
|
|
"eval_wer": 0.08833661417322834,
|
|
"step": 4128
|
|
},
|
|
{
|
|
"epoch": 48.84,
|
|
"learning_rate": 1.5359302325581395e-05,
|
|
"loss": 0.1543,
|
|
"step": 4200
|
|
},
|
|
{
|
|
"epoch": 49.0,
|
|
"eval_cer": 0.02819038859507933,
|
|
"eval_loss": 0.1961362212896347,
|
|
"eval_runtime": 19.2346,
|
|
"eval_samples_per_second": 17.728,
|
|
"eval_steps_per_second": 2.236,
|
|
"eval_wer": 0.08686023622047244,
|
|
"step": 4214
|
|
},
|
|
{
|
|
"epoch": 50.0,
|
|
"learning_rate": 1.5010465116279071e-05,
|
|
"loss": 0.1528,
|
|
"step": 4300
|
|
},
|
|
{
|
|
"epoch": 50.0,
|
|
"eval_cer": 0.028558289261899288,
|
|
"eval_loss": 0.1958739012479782,
|
|
"eval_runtime": 19.2236,
|
|
"eval_samples_per_second": 17.739,
|
|
"eval_steps_per_second": 2.237,
|
|
"eval_wer": 0.08858267716535433,
|
|
"step": 4300
|
|
},
|
|
{
|
|
"epoch": 51.0,
|
|
"eval_cer": 0.02773051276155438,
|
|
"eval_loss": 0.19377769529819489,
|
|
"eval_runtime": 19.7288,
|
|
"eval_samples_per_second": 17.284,
|
|
"eval_steps_per_second": 2.18,
|
|
"eval_wer": 0.08686023622047244,
|
|
"step": 4386
|
|
},
|
|
{
|
|
"epoch": 51.16,
|
|
"learning_rate": 1.4661627906976745e-05,
|
|
"loss": 0.1477,
|
|
"step": 4400
|
|
},
|
|
{
|
|
"epoch": 52.0,
|
|
"eval_cer": 0.02717866176132444,
|
|
"eval_loss": 0.19839388132095337,
|
|
"eval_runtime": 19.2855,
|
|
"eval_samples_per_second": 17.682,
|
|
"eval_steps_per_second": 2.23,
|
|
"eval_wer": 0.08390748031496063,
|
|
"step": 4472
|
|
},
|
|
{
|
|
"epoch": 52.33,
|
|
"learning_rate": 1.4312790697674419e-05,
|
|
"loss": 0.1533,
|
|
"step": 4500
|
|
},
|
|
{
|
|
"epoch": 53.0,
|
|
"eval_cer": 0.027362612094734423,
|
|
"eval_loss": 0.19690458476543427,
|
|
"eval_runtime": 19.5093,
|
|
"eval_samples_per_second": 17.479,
|
|
"eval_steps_per_second": 2.204,
|
|
"eval_wer": 0.08636811023622047,
|
|
"step": 4558
|
|
},
|
|
{
|
|
"epoch": 53.49,
|
|
"learning_rate": 1.3963953488372094e-05,
|
|
"loss": 0.1351,
|
|
"step": 4600
|
|
},
|
|
{
|
|
"epoch": 54.0,
|
|
"eval_cer": 0.027500574844791906,
|
|
"eval_loss": 0.19829140603542328,
|
|
"eval_runtime": 19.311,
|
|
"eval_samples_per_second": 17.658,
|
|
"eval_steps_per_second": 2.227,
|
|
"eval_wer": 0.08562992125984252,
|
|
"step": 4644
|
|
},
|
|
{
|
|
"epoch": 54.65,
|
|
"learning_rate": 1.3615116279069768e-05,
|
|
"loss": 0.1468,
|
|
"step": 4700
|
|
},
|
|
{
|
|
"epoch": 55.0,
|
|
"eval_cer": 0.02717866176132444,
|
|
"eval_loss": 0.1969507783651352,
|
|
"eval_runtime": 19.3127,
|
|
"eval_samples_per_second": 17.657,
|
|
"eval_steps_per_second": 2.227,
|
|
"eval_wer": 0.08341535433070867,
|
|
"step": 4730
|
|
},
|
|
{
|
|
"epoch": 55.81,
|
|
"learning_rate": 1.3266279069767444e-05,
|
|
"loss": 0.1406,
|
|
"step": 4800
|
|
},
|
|
{
|
|
"epoch": 56.0,
|
|
"eval_cer": 0.026994711427914462,
|
|
"eval_loss": 0.19397075474262238,
|
|
"eval_runtime": 19.5659,
|
|
"eval_samples_per_second": 17.428,
|
|
"eval_steps_per_second": 2.198,
|
|
"eval_wer": 0.08169291338582677,
|
|
"step": 4816
|
|
},
|
|
{
|
|
"epoch": 56.98,
|
|
"learning_rate": 1.2917441860465116e-05,
|
|
"loss": 0.1412,
|
|
"step": 4900
|
|
},
|
|
{
|
|
"epoch": 57.0,
|
|
"eval_cer": 0.027684525178201885,
|
|
"eval_loss": 0.18938837945461273,
|
|
"eval_runtime": 19.2752,
|
|
"eval_samples_per_second": 17.691,
|
|
"eval_steps_per_second": 2.231,
|
|
"eval_wer": 0.0858759842519685,
|
|
"step": 4902
|
|
},
|
|
{
|
|
"epoch": 58.0,
|
|
"eval_cer": 0.027546562428144402,
|
|
"eval_loss": 0.1928883194923401,
|
|
"eval_runtime": 19.254,
|
|
"eval_samples_per_second": 17.711,
|
|
"eval_steps_per_second": 2.233,
|
|
"eval_wer": 0.08562992125984252,
|
|
"step": 4988
|
|
},
|
|
{
|
|
"epoch": 58.14,
|
|
"learning_rate": 1.256860465116279e-05,
|
|
"loss": 0.1435,
|
|
"step": 5000
|
|
},
|
|
{
|
|
"epoch": 59.0,
|
|
"eval_cer": 0.02782248792825937,
|
|
"eval_loss": 0.1976550668478012,
|
|
"eval_runtime": 19.3356,
|
|
"eval_samples_per_second": 17.636,
|
|
"eval_steps_per_second": 2.224,
|
|
"eval_wer": 0.0858759842519685,
|
|
"step": 5074
|
|
},
|
|
{
|
|
"epoch": 59.3,
|
|
"learning_rate": 1.2219767441860465e-05,
|
|
"loss": 0.1389,
|
|
"step": 5100
|
|
},
|
|
{
|
|
"epoch": 60.0,
|
|
"eval_cer": 0.027684525178201885,
|
|
"eval_loss": 0.19864533841609955,
|
|
"eval_runtime": 19.4727,
|
|
"eval_samples_per_second": 17.512,
|
|
"eval_steps_per_second": 2.208,
|
|
"eval_wer": 0.08489173228346457,
|
|
"step": 5160
|
|
},
|
|
{
|
|
"epoch": 60.47,
|
|
"learning_rate": 1.187093023255814e-05,
|
|
"loss": 0.1366,
|
|
"step": 5200
|
|
},
|
|
{
|
|
"epoch": 61.0,
|
|
"eval_cer": 0.027224649344676936,
|
|
"eval_loss": 0.1977253407239914,
|
|
"eval_runtime": 19.4637,
|
|
"eval_samples_per_second": 17.52,
|
|
"eval_steps_per_second": 2.209,
|
|
"eval_wer": 0.08562992125984252,
|
|
"step": 5246
|
|
},
|
|
{
|
|
"epoch": 61.63,
|
|
"learning_rate": 1.1522093023255813e-05,
|
|
"loss": 0.1367,
|
|
"step": 5300
|
|
},
|
|
{
|
|
"epoch": 62.0,
|
|
"eval_cer": 0.02745458726143941,
|
|
"eval_loss": 0.19484524428844452,
|
|
"eval_runtime": 19.3058,
|
|
"eval_samples_per_second": 17.663,
|
|
"eval_steps_per_second": 2.227,
|
|
"eval_wer": 0.08513779527559055,
|
|
"step": 5332
|
|
},
|
|
{
|
|
"epoch": 62.79,
|
|
"learning_rate": 1.1173255813953489e-05,
|
|
"loss": 0.1268,
|
|
"step": 5400
|
|
},
|
|
{
|
|
"epoch": 63.0,
|
|
"eval_cer": 0.027500574844791906,
|
|
"eval_loss": 0.1947014331817627,
|
|
"eval_runtime": 19.4659,
|
|
"eval_samples_per_second": 17.518,
|
|
"eval_steps_per_second": 2.209,
|
|
"eval_wer": 0.08562992125984252,
|
|
"step": 5418
|
|
},
|
|
{
|
|
"epoch": 63.95,
|
|
"learning_rate": 1.0824418604651163e-05,
|
|
"loss": 0.1441,
|
|
"step": 5500
|
|
},
|
|
{
|
|
"epoch": 64.0,
|
|
"eval_cer": 0.02773051276155438,
|
|
"eval_loss": 0.19555877149105072,
|
|
"eval_runtime": 19.4278,
|
|
"eval_samples_per_second": 17.552,
|
|
"eval_steps_per_second": 2.213,
|
|
"eval_wer": 0.08562992125984252,
|
|
"step": 5504
|
|
},
|
|
{
|
|
"epoch": 64.0,
|
|
"step": 5504,
|
|
"total_flos": 4.598243749292761e+19,
|
|
"train_loss": 0.5636214655932299,
|
|
"train_runtime": 15910.8756,
|
|
"train_samples_per_second": 17.24,
|
|
"train_steps_per_second": 0.541
|
|
}
|
|
],
|
|
"max_steps": 8600,
|
|
"num_train_epochs": 100,
|
|
"total_flos": 4.598243749292761e+19,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|