|
{ |
|
"best_metric": 31.96095635007677, |
|
"best_model_checkpoint": "./whisper-medium-bn/checkpoint-64000", |
|
"epoch": 4.046023517511696, |
|
"global_step": 64000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.13, |
|
"learning_rate": 9.936816319635089e-06, |
|
"loss": 0.2519, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"eval_loss": 0.10757352411746979, |
|
"eval_runtime": 1082.7593, |
|
"eval_samples_per_second": 5.962, |
|
"eval_steps_per_second": 0.373, |
|
"eval_wer": 100.07530891277328, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"learning_rate": 9.852346158719434e-06, |
|
"loss": 0.1112, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"eval_loss": 0.08318958431482315, |
|
"eval_runtime": 1158.0641, |
|
"eval_samples_per_second": 5.574, |
|
"eval_steps_per_second": 0.349, |
|
"eval_wer": 99.66732470571031, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"learning_rate": 9.767875997803777e-06, |
|
"loss": 0.0909, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"eval_loss": 0.0710902065038681, |
|
"eval_runtime": 1331.5659, |
|
"eval_samples_per_second": 4.848, |
|
"eval_steps_per_second": 0.303, |
|
"eval_wer": 98.73729619068509, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"learning_rate": 9.683490307049036e-06, |
|
"loss": 0.0824, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"eval_loss": 0.06057652831077576, |
|
"eval_runtime": 2277.9347, |
|
"eval_samples_per_second": 2.834, |
|
"eval_steps_per_second": 0.177, |
|
"eval_wer": 71.25392995539957, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"learning_rate": 9.599062381213838e-06, |
|
"loss": 0.0766, |
|
"step": 10000 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"eval_loss": 0.054338764399290085, |
|
"eval_runtime": 2259.5514, |
|
"eval_samples_per_second": 2.857, |
|
"eval_steps_per_second": 0.179, |
|
"eval_wer": 72.27827740001462, |
|
"step": 10000 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"learning_rate": 9.51459222029818e-06, |
|
"loss": 0.0697, |
|
"step": 12000 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"eval_loss": 0.05125707387924194, |
|
"eval_runtime": 2175.3979, |
|
"eval_samples_per_second": 2.967, |
|
"eval_steps_per_second": 0.186, |
|
"eval_wer": 81.96680558602033, |
|
"step": 12000 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"learning_rate": 9.43016429446298e-06, |
|
"loss": 0.0628, |
|
"step": 14000 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"eval_loss": 0.04882914572954178, |
|
"eval_runtime": 2201.7571, |
|
"eval_samples_per_second": 2.932, |
|
"eval_steps_per_second": 0.183, |
|
"eval_wer": 78.58814067412445, |
|
"step": 14000 |
|
}, |
|
{ |
|
"epoch": 1.01, |
|
"learning_rate": 9.345736368627784e-06, |
|
"loss": 0.0584, |
|
"step": 16000 |
|
}, |
|
{ |
|
"epoch": 1.01, |
|
"eval_loss": 0.045878272503614426, |
|
"eval_runtime": 2300.4041, |
|
"eval_samples_per_second": 2.806, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 48.36294509029758, |
|
"step": 16000 |
|
}, |
|
{ |
|
"epoch": 1.14, |
|
"learning_rate": 9.261266207712126e-06, |
|
"loss": 0.048, |
|
"step": 18000 |
|
}, |
|
{ |
|
"epoch": 1.14, |
|
"eval_loss": 0.0436786450445652, |
|
"eval_runtime": 2280.8033, |
|
"eval_samples_per_second": 2.83, |
|
"eval_steps_per_second": 0.177, |
|
"eval_wer": 55.72201506178256, |
|
"step": 18000 |
|
}, |
|
{ |
|
"epoch": 1.26, |
|
"learning_rate": 9.17679604679647e-06, |
|
"loss": 0.0479, |
|
"step": 20000 |
|
}, |
|
{ |
|
"epoch": 1.26, |
|
"eval_loss": 0.04214460402727127, |
|
"eval_runtime": 2296.865, |
|
"eval_samples_per_second": 2.81, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 49.24910433574614, |
|
"step": 20000 |
|
}, |
|
{ |
|
"epoch": 1.39, |
|
"learning_rate": 9.09236812096127e-06, |
|
"loss": 0.0449, |
|
"step": 22000 |
|
}, |
|
{ |
|
"epoch": 1.39, |
|
"eval_loss": 0.041120000183582306, |
|
"eval_runtime": 2298.4349, |
|
"eval_samples_per_second": 2.808, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 49.43774219492579, |
|
"step": 22000 |
|
}, |
|
{ |
|
"epoch": 1.52, |
|
"learning_rate": 9.007940195126072e-06, |
|
"loss": 0.0442, |
|
"step": 24000 |
|
}, |
|
{ |
|
"epoch": 1.52, |
|
"eval_loss": 0.03955331817269325, |
|
"eval_runtime": 2308.5361, |
|
"eval_samples_per_second": 2.796, |
|
"eval_steps_per_second": 0.175, |
|
"eval_wer": 42.16494845360825, |
|
"step": 24000 |
|
}, |
|
{ |
|
"epoch": 1.64, |
|
"learning_rate": 8.923512269290874e-06, |
|
"loss": 0.0434, |
|
"step": 26000 |
|
}, |
|
{ |
|
"epoch": 1.64, |
|
"eval_loss": 0.037342339754104614, |
|
"eval_runtime": 2302.7779, |
|
"eval_samples_per_second": 2.803, |
|
"eval_steps_per_second": 0.175, |
|
"eval_wer": 45.68765080061417, |
|
"step": 26000 |
|
}, |
|
{ |
|
"epoch": 1.77, |
|
"learning_rate": 8.839084343455674e-06, |
|
"loss": 0.0421, |
|
"step": 28000 |
|
}, |
|
{ |
|
"epoch": 1.77, |
|
"eval_loss": 0.0369366817176342, |
|
"eval_runtime": 2289.7529, |
|
"eval_samples_per_second": 2.819, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 56.33399137237698, |
|
"step": 28000 |
|
}, |
|
{ |
|
"epoch": 1.9, |
|
"learning_rate": 8.754656417620476e-06, |
|
"loss": 0.0417, |
|
"step": 30000 |
|
}, |
|
{ |
|
"epoch": 1.9, |
|
"eval_loss": 0.03590318188071251, |
|
"eval_runtime": 2299.447, |
|
"eval_samples_per_second": 2.807, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 40.42114498793595, |
|
"step": 30000 |
|
}, |
|
{ |
|
"epoch": 2.02, |
|
"learning_rate": 8.670228491785278e-06, |
|
"loss": 0.0396, |
|
"step": 32000 |
|
}, |
|
{ |
|
"epoch": 2.02, |
|
"eval_loss": 0.03535410389304161, |
|
"eval_runtime": 2295.7712, |
|
"eval_samples_per_second": 2.812, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 39.69218395847042, |
|
"step": 32000 |
|
}, |
|
{ |
|
"epoch": 2.15, |
|
"learning_rate": 8.585758330869621e-06, |
|
"loss": 0.0316, |
|
"step": 34000 |
|
}, |
|
{ |
|
"epoch": 2.15, |
|
"eval_loss": 0.035752419382333755, |
|
"eval_runtime": 2290.0476, |
|
"eval_samples_per_second": 2.819, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 48.914966732470575, |
|
"step": 34000 |
|
}, |
|
{ |
|
"epoch": 2.28, |
|
"learning_rate": 8.501330405034423e-06, |
|
"loss": 0.0304, |
|
"step": 36000 |
|
}, |
|
{ |
|
"epoch": 2.28, |
|
"eval_loss": 0.03435269743204117, |
|
"eval_runtime": 2306.3737, |
|
"eval_samples_per_second": 2.799, |
|
"eval_steps_per_second": 0.175, |
|
"eval_wer": 37.936682020911014, |
|
"step": 36000 |
|
}, |
|
{ |
|
"epoch": 2.4, |
|
"learning_rate": 8.416944714279682e-06, |
|
"loss": 0.0308, |
|
"step": 38000 |
|
}, |
|
{ |
|
"epoch": 2.4, |
|
"eval_loss": 0.034886814653873444, |
|
"eval_runtime": 2296.8318, |
|
"eval_samples_per_second": 2.81, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 38.98735102727206, |
|
"step": 38000 |
|
}, |
|
{ |
|
"epoch": 2.53, |
|
"learning_rate": 8.332474553364025e-06, |
|
"loss": 0.031, |
|
"step": 40000 |
|
}, |
|
{ |
|
"epoch": 2.53, |
|
"eval_loss": 0.03347828984260559, |
|
"eval_runtime": 2306.1956, |
|
"eval_samples_per_second": 2.799, |
|
"eval_steps_per_second": 0.175, |
|
"eval_wer": 34.57117788988813, |
|
"step": 40000 |
|
}, |
|
{ |
|
"epoch": 2.66, |
|
"learning_rate": 8.24800439244837e-06, |
|
"loss": 0.0299, |
|
"step": 42000 |
|
}, |
|
{ |
|
"epoch": 2.66, |
|
"eval_loss": 0.03452066704630852, |
|
"eval_runtime": 2295.8518, |
|
"eval_samples_per_second": 2.812, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 37.69832565621115, |
|
"step": 42000 |
|
}, |
|
{ |
|
"epoch": 2.78, |
|
"learning_rate": 8.16357646661317e-06, |
|
"loss": 0.0304, |
|
"step": 44000 |
|
}, |
|
{ |
|
"epoch": 2.78, |
|
"eval_loss": 0.03322349861264229, |
|
"eval_runtime": 2305.8554, |
|
"eval_samples_per_second": 2.799, |
|
"eval_steps_per_second": 0.175, |
|
"eval_wer": 36.211157417562326, |
|
"step": 44000 |
|
}, |
|
{ |
|
"epoch": 2.91, |
|
"learning_rate": 8.079148540777972e-06, |
|
"loss": 0.028, |
|
"step": 46000 |
|
}, |
|
{ |
|
"epoch": 2.91, |
|
"eval_loss": 0.03227932006120682, |
|
"eval_runtime": 2303.1451, |
|
"eval_samples_per_second": 2.803, |
|
"eval_steps_per_second": 0.175, |
|
"eval_wer": 35.43905827301309, |
|
"step": 46000 |
|
}, |
|
{ |
|
"epoch": 3.03, |
|
"learning_rate": 7.994678379862315e-06, |
|
"loss": 0.0281, |
|
"step": 48000 |
|
}, |
|
{ |
|
"epoch": 3.03, |
|
"eval_loss": 0.03418450057506561, |
|
"eval_runtime": 2296.1947, |
|
"eval_samples_per_second": 2.811, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 35.552387219419465, |
|
"step": 48000 |
|
}, |
|
{ |
|
"epoch": 3.16, |
|
"learning_rate": 7.910250454027115e-06, |
|
"loss": 0.0208, |
|
"step": 50000 |
|
}, |
|
{ |
|
"epoch": 3.16, |
|
"eval_loss": 0.0337488129734993, |
|
"eval_runtime": 2291.6062, |
|
"eval_samples_per_second": 2.817, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 37.779483804927985, |
|
"step": 50000 |
|
}, |
|
{ |
|
"epoch": 3.29, |
|
"learning_rate": 7.82578029311146e-06, |
|
"loss": 0.021, |
|
"step": 52000 |
|
}, |
|
{ |
|
"epoch": 3.29, |
|
"eval_loss": 0.033615775406360626, |
|
"eval_runtime": 2304.2255, |
|
"eval_samples_per_second": 2.801, |
|
"eval_steps_per_second": 0.175, |
|
"eval_wer": 33.51612195656942, |
|
"step": 52000 |
|
}, |
|
{ |
|
"epoch": 3.41, |
|
"learning_rate": 7.74135236727626e-06, |
|
"loss": 0.021, |
|
"step": 54000 |
|
}, |
|
{ |
|
"epoch": 3.41, |
|
"eval_loss": 0.03333365172147751, |
|
"eval_runtime": 2311.8192, |
|
"eval_samples_per_second": 2.792, |
|
"eval_steps_per_second": 0.175, |
|
"eval_wer": 32.82591211522995, |
|
"step": 54000 |
|
}, |
|
{ |
|
"epoch": 3.54, |
|
"learning_rate": 7.656882206360604e-06, |
|
"loss": 0.0202, |
|
"step": 56000 |
|
}, |
|
{ |
|
"epoch": 3.54, |
|
"eval_loss": 0.032315392047166824, |
|
"eval_runtime": 2307.3986, |
|
"eval_samples_per_second": 2.798, |
|
"eval_steps_per_second": 0.175, |
|
"eval_wer": 33.00870073846604, |
|
"step": 56000 |
|
}, |
|
{ |
|
"epoch": 3.67, |
|
"learning_rate": 7.5724965156058624e-06, |
|
"loss": 0.0214, |
|
"step": 58000 |
|
}, |
|
{ |
|
"epoch": 3.67, |
|
"eval_loss": 0.0318850576877594, |
|
"eval_runtime": 2301.2104, |
|
"eval_samples_per_second": 2.805, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 32.453023323828326, |
|
"step": 58000 |
|
}, |
|
{ |
|
"epoch": 3.79, |
|
"learning_rate": 7.488026354690206e-06, |
|
"loss": 0.0206, |
|
"step": 60000 |
|
}, |
|
{ |
|
"epoch": 3.79, |
|
"eval_loss": 0.031195413321256638, |
|
"eval_runtime": 2300.0768, |
|
"eval_samples_per_second": 2.806, |
|
"eval_steps_per_second": 0.176, |
|
"eval_wer": 32.26511661914163, |
|
"step": 60000 |
|
}, |
|
{ |
|
"epoch": 3.92, |
|
"learning_rate": 7.403598428855008e-06, |
|
"loss": 0.0213, |
|
"step": 62000 |
|
}, |
|
{ |
|
"epoch": 3.92, |
|
"eval_loss": 0.03142615407705307, |
|
"eval_runtime": 2307.3794, |
|
"eval_samples_per_second": 2.798, |
|
"eval_steps_per_second": 0.175, |
|
"eval_wer": 32.072822987497254, |
|
"step": 62000 |
|
}, |
|
{ |
|
"epoch": 4.05, |
|
"learning_rate": 7.319170503019809e-06, |
|
"loss": 0.0169, |
|
"step": 64000 |
|
}, |
|
{ |
|
"epoch": 4.05, |
|
"eval_loss": 0.03290534391999245, |
|
"eval_runtime": 2313.6442, |
|
"eval_samples_per_second": 2.79, |
|
"eval_steps_per_second": 0.175, |
|
"eval_wer": 31.96095635007677, |
|
"step": 64000 |
|
} |
|
], |
|
"max_steps": 237270, |
|
"num_train_epochs": 15, |
|
"total_flos": 2.6127077390548992e+20, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|