|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 20.752797558494404, |
|
"global_step": 20400, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.41, |
|
"learning_rate": 0.00023999999999999998, |
|
"loss": 4.7996, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.41, |
|
"eval_loss": 3.018367290496826, |
|
"eval_runtime": 230.2288, |
|
"eval_samples_per_second": 8.127, |
|
"eval_wer": 1.0, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"learning_rate": 0.0002968954812004139, |
|
"loss": 2.0295, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"eval_loss": 0.71845942735672, |
|
"eval_runtime": 234.0394, |
|
"eval_samples_per_second": 7.994, |
|
"eval_wer": 0.5917076167076167, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 1.22, |
|
"learning_rate": 0.0002927561228009658, |
|
"loss": 0.4632, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 1.22, |
|
"eval_loss": 0.4624195396900177, |
|
"eval_runtime": 419.7523, |
|
"eval_samples_per_second": 4.457, |
|
"eval_wer": 0.4606879606879607, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 1.63, |
|
"learning_rate": 0.0002886167644015177, |
|
"loss": 0.3595, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 1.63, |
|
"eval_loss": 0.4238700568675995, |
|
"eval_runtime": 235.4681, |
|
"eval_samples_per_second": 7.946, |
|
"eval_wer": 0.4348894348894349, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 2.03, |
|
"learning_rate": 0.0002844774060020697, |
|
"loss": 0.3181, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 2.03, |
|
"eval_loss": 0.4156078100204468, |
|
"eval_runtime": 234.1423, |
|
"eval_samples_per_second": 7.991, |
|
"eval_wer": 0.42106879606879605, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 2.44, |
|
"learning_rate": 0.0002803380476026216, |
|
"loss": 0.2606, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 2.44, |
|
"eval_loss": 0.39613404870033264, |
|
"eval_runtime": 233.8224, |
|
"eval_samples_per_second": 8.002, |
|
"eval_wer": 0.4057125307125307, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 2.85, |
|
"learning_rate": 0.0002761986892031735, |
|
"loss": 0.2475, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 2.85, |
|
"eval_loss": 0.3680589497089386, |
|
"eval_runtime": 235.255, |
|
"eval_samples_per_second": 7.953, |
|
"eval_wer": 0.39256756756756755, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 3.26, |
|
"learning_rate": 0.0002720593308037254, |
|
"loss": 0.2208, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 3.26, |
|
"eval_loss": 0.40793511271476746, |
|
"eval_runtime": 233.2565, |
|
"eval_samples_per_second": 8.021, |
|
"eval_wer": 0.3984029484029484, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 3.66, |
|
"learning_rate": 0.0002679199724042773, |
|
"loss": 0.2168, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 3.66, |
|
"eval_loss": 0.40217384696006775, |
|
"eval_runtime": 233.5419, |
|
"eval_samples_per_second": 8.011, |
|
"eval_wer": 0.40958230958230957, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 4.07, |
|
"learning_rate": 0.0002637806140048292, |
|
"loss": 0.2059, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 4.07, |
|
"eval_loss": 0.42218777537345886, |
|
"eval_runtime": 233.047, |
|
"eval_samples_per_second": 8.028, |
|
"eval_wer": 0.40128992628992627, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 4.48, |
|
"learning_rate": 0.0002596412556053811, |
|
"loss": 0.1808, |
|
"step": 4400 |
|
}, |
|
{ |
|
"epoch": 4.48, |
|
"eval_loss": 0.43475455045700073, |
|
"eval_runtime": 235.2511, |
|
"eval_samples_per_second": 7.953, |
|
"eval_wer": 0.39213759213759214, |
|
"step": 4400 |
|
}, |
|
{ |
|
"epoch": 4.88, |
|
"learning_rate": 0.0002555018972059331, |
|
"loss": 0.1874, |
|
"step": 4800 |
|
}, |
|
{ |
|
"epoch": 4.88, |
|
"eval_loss": 0.40788909792900085, |
|
"eval_runtime": 237.3453, |
|
"eval_samples_per_second": 7.883, |
|
"eval_wer": 0.3885749385749386, |
|
"step": 4800 |
|
}, |
|
{ |
|
"epoch": 5.29, |
|
"learning_rate": 0.000251362538806485, |
|
"loss": 0.1671, |
|
"step": 5200 |
|
}, |
|
{ |
|
"epoch": 5.29, |
|
"eval_loss": 0.42425239086151123, |
|
"eval_runtime": 235.3303, |
|
"eval_samples_per_second": 7.951, |
|
"eval_wer": 0.3864864864864865, |
|
"step": 5200 |
|
}, |
|
{ |
|
"epoch": 5.7, |
|
"learning_rate": 0.0002472231804070369, |
|
"loss": 0.168, |
|
"step": 5600 |
|
}, |
|
{ |
|
"epoch": 5.7, |
|
"eval_loss": 0.39308613538742065, |
|
"eval_runtime": 234.7832, |
|
"eval_samples_per_second": 7.969, |
|
"eval_wer": 0.38114250614250617, |
|
"step": 5600 |
|
}, |
|
{ |
|
"epoch": 6.1, |
|
"learning_rate": 0.0002430838220075888, |
|
"loss": 0.1603, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 6.1, |
|
"eval_loss": 0.4091956615447998, |
|
"eval_runtime": 234.4154, |
|
"eval_samples_per_second": 7.982, |
|
"eval_wer": 0.38175675675675674, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 6.51, |
|
"learning_rate": 0.00023894446360814072, |
|
"loss": 0.1458, |
|
"step": 6400 |
|
}, |
|
{ |
|
"epoch": 6.51, |
|
"eval_loss": 0.4052203893661499, |
|
"eval_runtime": 236.4087, |
|
"eval_samples_per_second": 7.914, |
|
"eval_wer": 0.3844594594594595, |
|
"step": 6400 |
|
}, |
|
{ |
|
"epoch": 6.92, |
|
"learning_rate": 0.00023480510520869263, |
|
"loss": 0.1466, |
|
"step": 6800 |
|
}, |
|
{ |
|
"epoch": 6.92, |
|
"eval_loss": 0.4150303900241852, |
|
"eval_runtime": 234.9457, |
|
"eval_samples_per_second": 7.964, |
|
"eval_wer": 0.3832923832923833, |
|
"step": 6800 |
|
}, |
|
{ |
|
"epoch": 7.32, |
|
"learning_rate": 0.00023066574680924453, |
|
"loss": 0.1347, |
|
"step": 7200 |
|
}, |
|
{ |
|
"epoch": 7.32, |
|
"eval_loss": 0.40190571546554565, |
|
"eval_runtime": 232.7808, |
|
"eval_samples_per_second": 8.038, |
|
"eval_wer": 0.3687346437346437, |
|
"step": 7200 |
|
}, |
|
{ |
|
"epoch": 7.73, |
|
"learning_rate": 0.00022652638840979646, |
|
"loss": 0.1358, |
|
"step": 7600 |
|
}, |
|
{ |
|
"epoch": 7.73, |
|
"eval_loss": 0.40535062551498413, |
|
"eval_runtime": 234.3447, |
|
"eval_samples_per_second": 7.984, |
|
"eval_wer": 0.3745085995085995, |
|
"step": 7600 |
|
}, |
|
{ |
|
"epoch": 8.14, |
|
"learning_rate": 0.0002223870300103484, |
|
"loss": 0.1334, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 8.14, |
|
"eval_loss": 0.3982257843017578, |
|
"eval_runtime": 234.9167, |
|
"eval_samples_per_second": 7.965, |
|
"eval_wer": 0.371990171990172, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 8.55, |
|
"learning_rate": 0.0002182476716109003, |
|
"loss": 0.1228, |
|
"step": 8400 |
|
}, |
|
{ |
|
"epoch": 8.55, |
|
"eval_loss": 0.38730743527412415, |
|
"eval_runtime": 236.5628, |
|
"eval_samples_per_second": 7.909, |
|
"eval_wer": 0.3726044226044226, |
|
"step": 8400 |
|
}, |
|
{ |
|
"epoch": 8.95, |
|
"learning_rate": 0.0002141083132114522, |
|
"loss": 0.1226, |
|
"step": 8800 |
|
}, |
|
{ |
|
"epoch": 8.95, |
|
"eval_loss": 0.3983035087585449, |
|
"eval_runtime": 236.8485, |
|
"eval_samples_per_second": 7.9, |
|
"eval_wer": 0.36437346437346435, |
|
"step": 8800 |
|
}, |
|
{ |
|
"epoch": 9.36, |
|
"learning_rate": 0.0002099689548120041, |
|
"loss": 0.1091, |
|
"step": 9200 |
|
}, |
|
{ |
|
"epoch": 9.36, |
|
"eval_loss": 0.4045446813106537, |
|
"eval_runtime": 234.4792, |
|
"eval_samples_per_second": 7.979, |
|
"eval_wer": 0.36406633906633906, |
|
"step": 9200 |
|
}, |
|
{ |
|
"epoch": 9.77, |
|
"learning_rate": 0.00020582959641255602, |
|
"loss": 0.1124, |
|
"step": 9600 |
|
}, |
|
{ |
|
"epoch": 9.77, |
|
"eval_loss": 0.41637349128723145, |
|
"eval_runtime": 234.3486, |
|
"eval_samples_per_second": 7.984, |
|
"eval_wer": 0.36812039312039313, |
|
"step": 9600 |
|
}, |
|
{ |
|
"epoch": 10.17, |
|
"learning_rate": 0.00020169023801310795, |
|
"loss": 0.1117, |
|
"step": 10000 |
|
}, |
|
{ |
|
"epoch": 10.17, |
|
"eval_loss": 0.42693623900413513, |
|
"eval_runtime": 235.3, |
|
"eval_samples_per_second": 7.952, |
|
"eval_wer": 0.36732186732186733, |
|
"step": 10000 |
|
}, |
|
{ |
|
"epoch": 10.58, |
|
"learning_rate": 0.00019755087961365985, |
|
"loss": 0.1015, |
|
"step": 10400 |
|
}, |
|
{ |
|
"epoch": 10.58, |
|
"eval_loss": 0.45069122314453125, |
|
"eval_runtime": 236.7508, |
|
"eval_samples_per_second": 7.903, |
|
"eval_wer": 0.36357493857493856, |
|
"step": 10400 |
|
}, |
|
{ |
|
"epoch": 10.99, |
|
"learning_rate": 0.0001934115212142118, |
|
"loss": 0.1062, |
|
"step": 10800 |
|
}, |
|
{ |
|
"epoch": 10.99, |
|
"eval_loss": 0.4379144608974457, |
|
"eval_runtime": 237.8786, |
|
"eval_samples_per_second": 7.865, |
|
"eval_wer": 0.3648034398034398, |
|
"step": 10800 |
|
}, |
|
{ |
|
"epoch": 11.39, |
|
"learning_rate": 0.00018927216281476372, |
|
"loss": 0.0945, |
|
"step": 11200 |
|
}, |
|
{ |
|
"epoch": 11.39, |
|
"eval_loss": 0.4285587966442108, |
|
"eval_runtime": 234.2891, |
|
"eval_samples_per_second": 7.986, |
|
"eval_wer": 0.36437346437346435, |
|
"step": 11200 |
|
}, |
|
{ |
|
"epoch": 11.8, |
|
"learning_rate": 0.00018513280441531562, |
|
"loss": 0.0996, |
|
"step": 11600 |
|
}, |
|
{ |
|
"epoch": 11.8, |
|
"eval_loss": 0.4398898780345917, |
|
"eval_runtime": 235.5528, |
|
"eval_samples_per_second": 7.943, |
|
"eval_wer": 0.367014742014742, |
|
"step": 11600 |
|
}, |
|
{ |
|
"epoch": 12.21, |
|
"learning_rate": 0.00018099344601586753, |
|
"loss": 0.0911, |
|
"step": 12000 |
|
}, |
|
{ |
|
"epoch": 12.21, |
|
"eval_loss": 0.4221023619174957, |
|
"eval_runtime": 234.4385, |
|
"eval_samples_per_second": 7.981, |
|
"eval_wer": 0.3568181818181818, |
|
"step": 12000 |
|
}, |
|
{ |
|
"epoch": 12.61, |
|
"learning_rate": 0.00017685408761641943, |
|
"loss": 0.09, |
|
"step": 12400 |
|
}, |
|
{ |
|
"epoch": 12.61, |
|
"eval_loss": 0.4387373924255371, |
|
"eval_runtime": 238.8017, |
|
"eval_samples_per_second": 7.835, |
|
"eval_wer": 0.3641891891891892, |
|
"step": 12400 |
|
}, |
|
{ |
|
"epoch": 13.02, |
|
"learning_rate": 0.00017271472921697134, |
|
"loss": 0.0894, |
|
"step": 12800 |
|
}, |
|
{ |
|
"epoch": 13.02, |
|
"eval_loss": 0.4392118752002716, |
|
"eval_runtime": 235.2986, |
|
"eval_samples_per_second": 7.952, |
|
"eval_wer": 0.3574938574938575, |
|
"step": 12800 |
|
}, |
|
{ |
|
"epoch": 13.43, |
|
"learning_rate": 0.00016857537081752327, |
|
"loss": 0.0816, |
|
"step": 13200 |
|
}, |
|
{ |
|
"epoch": 13.43, |
|
"eval_loss": 0.44565021991729736, |
|
"eval_runtime": 235.1666, |
|
"eval_samples_per_second": 7.956, |
|
"eval_wer": 0.3616707616707617, |
|
"step": 13200 |
|
}, |
|
{ |
|
"epoch": 13.84, |
|
"learning_rate": 0.00016443601241807518, |
|
"loss": 0.0813, |
|
"step": 13600 |
|
}, |
|
{ |
|
"epoch": 13.84, |
|
"eval_loss": 0.4428552985191345, |
|
"eval_runtime": 236.4759, |
|
"eval_samples_per_second": 7.912, |
|
"eval_wer": 0.3593980343980344, |
|
"step": 13600 |
|
}, |
|
{ |
|
"epoch": 14.24, |
|
"learning_rate": 0.0001602966540186271, |
|
"loss": 0.078, |
|
"step": 14000 |
|
}, |
|
{ |
|
"epoch": 14.24, |
|
"eval_loss": 0.46921756863594055, |
|
"eval_runtime": 236.081, |
|
"eval_samples_per_second": 7.925, |
|
"eval_wer": 0.35909090909090907, |
|
"step": 14000 |
|
}, |
|
{ |
|
"epoch": 14.65, |
|
"learning_rate": 0.00015615729561917901, |
|
"loss": 0.0777, |
|
"step": 14400 |
|
}, |
|
{ |
|
"epoch": 14.65, |
|
"eval_loss": 0.45899245142936707, |
|
"eval_runtime": 236.6098, |
|
"eval_samples_per_second": 7.908, |
|
"eval_wer": 0.35448402948402946, |
|
"step": 14400 |
|
}, |
|
{ |
|
"epoch": 15.06, |
|
"learning_rate": 0.00015201793721973095, |
|
"loss": 0.076, |
|
"step": 14800 |
|
}, |
|
{ |
|
"epoch": 15.06, |
|
"eval_loss": 0.48056069016456604, |
|
"eval_runtime": 234.5593, |
|
"eval_samples_per_second": 7.977, |
|
"eval_wer": 0.3517199017199017, |
|
"step": 14800 |
|
}, |
|
{ |
|
"epoch": 15.46, |
|
"learning_rate": 0.00014787857882028285, |
|
"loss": 0.0717, |
|
"step": 15200 |
|
}, |
|
{ |
|
"epoch": 15.46, |
|
"eval_loss": 0.49670103192329407, |
|
"eval_runtime": 236.0843, |
|
"eval_samples_per_second": 7.925, |
|
"eval_wer": 0.3600737100737101, |
|
"step": 15200 |
|
}, |
|
{ |
|
"epoch": 15.87, |
|
"learning_rate": 0.00014373922042083476, |
|
"loss": 0.0708, |
|
"step": 15600 |
|
}, |
|
{ |
|
"epoch": 15.87, |
|
"eval_loss": 0.46058785915374756, |
|
"eval_runtime": 242.6156, |
|
"eval_samples_per_second": 7.712, |
|
"eval_wer": 0.35982800982800983, |
|
"step": 15600 |
|
}, |
|
{ |
|
"epoch": 16.28, |
|
"learning_rate": 0.00013959986202138666, |
|
"loss": 0.0673, |
|
"step": 16000 |
|
}, |
|
{ |
|
"epoch": 16.28, |
|
"eval_loss": 0.45084264874458313, |
|
"eval_runtime": 234.0195, |
|
"eval_samples_per_second": 7.995, |
|
"eval_wer": 0.3546683046683047, |
|
"step": 16000 |
|
}, |
|
{ |
|
"epoch": 16.68, |
|
"learning_rate": 0.0001354605036219386, |
|
"loss": 0.0664, |
|
"step": 16400 |
|
}, |
|
{ |
|
"epoch": 16.68, |
|
"eval_loss": 0.4838450849056244, |
|
"eval_runtime": 237.8079, |
|
"eval_samples_per_second": 7.868, |
|
"eval_wer": 0.35614250614250614, |
|
"step": 16400 |
|
}, |
|
{ |
|
"epoch": 17.09, |
|
"learning_rate": 0.0001313211452224905, |
|
"loss": 0.0639, |
|
"step": 16800 |
|
}, |
|
{ |
|
"epoch": 17.09, |
|
"eval_loss": 0.4703587591648102, |
|
"eval_runtime": 235.9394, |
|
"eval_samples_per_second": 7.93, |
|
"eval_wer": 0.35657248157248156, |
|
"step": 16800 |
|
}, |
|
{ |
|
"epoch": 17.5, |
|
"learning_rate": 0.0001271817868230424, |
|
"loss": 0.0597, |
|
"step": 17200 |
|
}, |
|
{ |
|
"epoch": 17.5, |
|
"eval_loss": 0.47286155819892883, |
|
"eval_runtime": 237.4364, |
|
"eval_samples_per_second": 7.88, |
|
"eval_wer": 0.35657248157248156, |
|
"step": 17200 |
|
}, |
|
{ |
|
"epoch": 17.9, |
|
"learning_rate": 0.00012304242842359434, |
|
"loss": 0.068, |
|
"step": 17600 |
|
}, |
|
{ |
|
"epoch": 17.9, |
|
"eval_loss": 0.42778506875038147, |
|
"eval_runtime": 235.8032, |
|
"eval_samples_per_second": 7.935, |
|
"eval_wer": 0.3484029484029484, |
|
"step": 17600 |
|
}, |
|
{ |
|
"epoch": 18.31, |
|
"learning_rate": 0.00011890307002414625, |
|
"loss": 0.0582, |
|
"step": 18000 |
|
}, |
|
{ |
|
"epoch": 18.31, |
|
"eval_loss": 0.47386595606803894, |
|
"eval_runtime": 238.3257, |
|
"eval_samples_per_second": 7.851, |
|
"eval_wer": 0.34864864864864864, |
|
"step": 18000 |
|
}, |
|
{ |
|
"epoch": 18.72, |
|
"learning_rate": 0.00011476371162469816, |
|
"loss": 0.0593, |
|
"step": 18400 |
|
}, |
|
{ |
|
"epoch": 18.72, |
|
"eval_loss": 0.46009212732315063, |
|
"eval_runtime": 238.7244, |
|
"eval_samples_per_second": 7.837, |
|
"eval_wer": 0.3468673218673219, |
|
"step": 18400 |
|
}, |
|
{ |
|
"epoch": 19.13, |
|
"learning_rate": 0.00011062435322525008, |
|
"loss": 0.0554, |
|
"step": 18800 |
|
}, |
|
{ |
|
"epoch": 19.13, |
|
"eval_loss": 0.506970226764679, |
|
"eval_runtime": 236.9238, |
|
"eval_samples_per_second": 7.897, |
|
"eval_wer": 0.34594594594594597, |
|
"step": 18800 |
|
}, |
|
{ |
|
"epoch": 19.53, |
|
"learning_rate": 0.00010648499482580198, |
|
"loss": 0.0567, |
|
"step": 19200 |
|
}, |
|
{ |
|
"epoch": 19.53, |
|
"eval_loss": 0.47986453771591187, |
|
"eval_runtime": 237.1448, |
|
"eval_samples_per_second": 7.89, |
|
"eval_wer": 0.3435503685503685, |
|
"step": 19200 |
|
}, |
|
{ |
|
"epoch": 19.94, |
|
"learning_rate": 0.00010234563642635392, |
|
"loss": 0.0554, |
|
"step": 19600 |
|
}, |
|
{ |
|
"epoch": 19.94, |
|
"eval_loss": 0.47748732566833496, |
|
"eval_runtime": 236.6967, |
|
"eval_samples_per_second": 7.905, |
|
"eval_wer": 0.34877149877149877, |
|
"step": 19600 |
|
}, |
|
{ |
|
"epoch": 20.35, |
|
"learning_rate": 9.820627802690582e-05, |
|
"loss": 0.0532, |
|
"step": 20000 |
|
}, |
|
{ |
|
"epoch": 20.35, |
|
"eval_loss": 0.5038613677024841, |
|
"eval_runtime": 237.1224, |
|
"eval_samples_per_second": 7.89, |
|
"eval_wer": 0.34342751842751845, |
|
"step": 20000 |
|
}, |
|
{ |
|
"epoch": 20.75, |
|
"learning_rate": 9.406691962745773e-05, |
|
"loss": 0.0503, |
|
"step": 20400 |
|
}, |
|
{ |
|
"epoch": 20.75, |
|
"eval_loss": 0.48931312561035156, |
|
"eval_runtime": 236.3616, |
|
"eval_samples_per_second": 7.916, |
|
"eval_wer": 0.3406633906633907, |
|
"step": 20400 |
|
} |
|
], |
|
"max_steps": 29490, |
|
"num_train_epochs": 30, |
|
"total_flos": 1.1319628335303795e+20, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|