| { | |
| "best_metric": 0.30534103512763977, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-bemgen-male-model-test/checkpoint-2300", | |
| "epoch": 2.688728024819028, | |
| "eval_steps": 100, | |
| "global_step": 2600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.10341261633919338, | |
| "grad_norm": 3.1888983249664307, | |
| "learning_rate": 0.00028799999999999995, | |
| "loss": 6.9809, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10341261633919338, | |
| "eval_loss": 1.3138622045516968, | |
| "eval_runtime": 31.0692, | |
| "eval_samples_per_second": 15.803, | |
| "eval_steps_per_second": 3.959, | |
| "eval_wer": 0.9957016978293574, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.20682523267838676, | |
| "grad_norm": 2.4408092498779297, | |
| "learning_rate": 0.0002990038049117952, | |
| "loss": 0.745, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.20682523267838676, | |
| "eval_loss": 0.4297034740447998, | |
| "eval_runtime": 31.0581, | |
| "eval_samples_per_second": 15.809, | |
| "eval_steps_per_second": 3.96, | |
| "eval_wer": 0.5882226520524393, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.31023784901758017, | |
| "grad_norm": 1.8350099325180054, | |
| "learning_rate": 0.0002979661016949152, | |
| "loss": 0.5423, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.31023784901758017, | |
| "eval_loss": 0.3886265754699707, | |
| "eval_runtime": 30.9037, | |
| "eval_samples_per_second": 15.888, | |
| "eval_steps_per_second": 3.98, | |
| "eval_wer": 0.5643670750053729, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4136504653567735, | |
| "grad_norm": 4.089741230010986, | |
| "learning_rate": 0.00029692839847803525, | |
| "loss": 0.539, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4136504653567735, | |
| "eval_loss": 0.36830800771713257, | |
| "eval_runtime": 30.9142, | |
| "eval_samples_per_second": 15.883, | |
| "eval_steps_per_second": 3.979, | |
| "eval_wer": 0.5448098001289491, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5170630816959669, | |
| "grad_norm": 2.093287706375122, | |
| "learning_rate": 0.0002958906952611553, | |
| "loss": 0.5277, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5170630816959669, | |
| "eval_loss": 0.3528771698474884, | |
| "eval_runtime": 31.284, | |
| "eval_samples_per_second": 15.695, | |
| "eval_steps_per_second": 3.932, | |
| "eval_wer": 0.508274231678487, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6204756980351603, | |
| "grad_norm": 1.6442524194717407, | |
| "learning_rate": 0.0002948529920442753, | |
| "loss": 0.4708, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6204756980351603, | |
| "eval_loss": 0.3492579162120819, | |
| "eval_runtime": 31.108, | |
| "eval_samples_per_second": 15.784, | |
| "eval_steps_per_second": 3.954, | |
| "eval_wer": 0.4996776273372018, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7238883143743536, | |
| "grad_norm": 2.5362725257873535, | |
| "learning_rate": 0.00029381528882739533, | |
| "loss": 0.4889, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7238883143743536, | |
| "eval_loss": 0.3467372953891754, | |
| "eval_runtime": 31.1038, | |
| "eval_samples_per_second": 15.786, | |
| "eval_steps_per_second": 3.954, | |
| "eval_wer": 0.5095637223296797, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.827300930713547, | |
| "grad_norm": 4.026537895202637, | |
| "learning_rate": 0.00029277758561051535, | |
| "loss": 0.4793, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.827300930713547, | |
| "eval_loss": 0.34066715836524963, | |
| "eval_runtime": 30.9432, | |
| "eval_samples_per_second": 15.868, | |
| "eval_steps_per_second": 3.975, | |
| "eval_wer": 0.48183967332903505, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.9307135470527405, | |
| "grad_norm": 4.126669406890869, | |
| "learning_rate": 0.0002917398823936354, | |
| "loss": 0.469, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9307135470527405, | |
| "eval_loss": 0.3455369174480438, | |
| "eval_runtime": 31.0926, | |
| "eval_samples_per_second": 15.792, | |
| "eval_steps_per_second": 3.956, | |
| "eval_wer": 0.49580915538362347, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.0341261633919339, | |
| "grad_norm": 1.1861900091171265, | |
| "learning_rate": 0.0002907021791767554, | |
| "loss": 0.4407, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0341261633919339, | |
| "eval_loss": 0.3328763544559479, | |
| "eval_runtime": 31.2336, | |
| "eval_samples_per_second": 15.72, | |
| "eval_steps_per_second": 3.938, | |
| "eval_wer": 0.473457984096282, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.1375387797311272, | |
| "grad_norm": 1.2025549411773682, | |
| "learning_rate": 0.00028967485299204426, | |
| "loss": 0.4524, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.1375387797311272, | |
| "eval_loss": 0.32889580726623535, | |
| "eval_runtime": 31.3882, | |
| "eval_samples_per_second": 15.643, | |
| "eval_steps_per_second": 3.919, | |
| "eval_wer": 0.4878572963679347, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.2409513960703205, | |
| "grad_norm": 1.4938141107559204, | |
| "learning_rate": 0.0002886371497751643, | |
| "loss": 0.4416, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.2409513960703205, | |
| "eval_loss": 0.327963262796402, | |
| "eval_runtime": 30.9988, | |
| "eval_samples_per_second": 15.839, | |
| "eval_steps_per_second": 3.968, | |
| "eval_wer": 0.4910810229959166, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.344364012409514, | |
| "grad_norm": 1.22685706615448, | |
| "learning_rate": 0.0002875994465582843, | |
| "loss": 0.4599, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.344364012409514, | |
| "eval_loss": 0.3284839391708374, | |
| "eval_runtime": 31.2403, | |
| "eval_samples_per_second": 15.717, | |
| "eval_steps_per_second": 3.937, | |
| "eval_wer": 0.4764667956157318, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.4477766287487073, | |
| "grad_norm": 1.3746376037597656, | |
| "learning_rate": 0.00028656174334140434, | |
| "loss": 0.4739, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.4477766287487073, | |
| "eval_loss": 0.3221452236175537, | |
| "eval_runtime": 31.2054, | |
| "eval_samples_per_second": 15.734, | |
| "eval_steps_per_second": 3.942, | |
| "eval_wer": 0.4693745970341715, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.5511892450879006, | |
| "grad_norm": 1.6073061227798462, | |
| "learning_rate": 0.00028552404012452436, | |
| "loss": 0.4466, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.5511892450879006, | |
| "eval_loss": 0.3196486234664917, | |
| "eval_runtime": 31.0094, | |
| "eval_samples_per_second": 15.834, | |
| "eval_steps_per_second": 3.967, | |
| "eval_wer": 0.4588437567160971, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.654601861427094, | |
| "grad_norm": 1.1827540397644043, | |
| "learning_rate": 0.0002844967139398132, | |
| "loss": 0.4483, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.654601861427094, | |
| "eval_loss": 0.3144252896308899, | |
| "eval_runtime": 31.0445, | |
| "eval_samples_per_second": 15.816, | |
| "eval_steps_per_second": 3.962, | |
| "eval_wer": 0.4526112185686654, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.7580144777662876, | |
| "grad_norm": 1.335554599761963, | |
| "learning_rate": 0.0002834590107229332, | |
| "loss": 0.4543, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.7580144777662876, | |
| "eval_loss": 0.31697967648506165, | |
| "eval_runtime": 31.4988, | |
| "eval_samples_per_second": 15.588, | |
| "eval_steps_per_second": 3.905, | |
| "eval_wer": 0.4528261336771975, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.861427094105481, | |
| "grad_norm": 2.555410385131836, | |
| "learning_rate": 0.00028242130750605327, | |
| "loss": 0.4537, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.861427094105481, | |
| "eval_loss": 0.3140595555305481, | |
| "eval_runtime": 31.3179, | |
| "eval_samples_per_second": 15.678, | |
| "eval_steps_per_second": 3.927, | |
| "eval_wer": 0.4521813883516011, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.9648397104446742, | |
| "grad_norm": 1.150947093963623, | |
| "learning_rate": 0.0002813836042891733, | |
| "loss": 0.4293, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.9648397104446742, | |
| "eval_loss": 0.31063735485076904, | |
| "eval_runtime": 31.1712, | |
| "eval_samples_per_second": 15.752, | |
| "eval_steps_per_second": 3.946, | |
| "eval_wer": 0.44530410487857297, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.0682523267838677, | |
| "grad_norm": 0.674537181854248, | |
| "learning_rate": 0.00028034590107229326, | |
| "loss": 0.4457, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.0682523267838677, | |
| "eval_loss": 0.3134273290634155, | |
| "eval_runtime": 31.1408, | |
| "eval_samples_per_second": 15.767, | |
| "eval_steps_per_second": 3.95, | |
| "eval_wer": 0.4650762948635289, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.1716649431230612, | |
| "grad_norm": 0.9923437237739563, | |
| "learning_rate": 0.00027930819785541334, | |
| "loss": 0.4214, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.1716649431230612, | |
| "eval_loss": 0.3119480609893799, | |
| "eval_runtime": 31.4395, | |
| "eval_samples_per_second": 15.617, | |
| "eval_steps_per_second": 3.912, | |
| "eval_wer": 0.4543305394369224, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.2750775594622543, | |
| "grad_norm": 0.9173412919044495, | |
| "learning_rate": 0.00027827049463853337, | |
| "loss": 0.4103, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.2750775594622543, | |
| "eval_loss": 0.3089054822921753, | |
| "eval_runtime": 31.248, | |
| "eval_samples_per_second": 15.713, | |
| "eval_steps_per_second": 3.936, | |
| "eval_wer": 0.43907156673114117, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.378490175801448, | |
| "grad_norm": 3.1410484313964844, | |
| "learning_rate": 0.0002772327914216534, | |
| "loss": 0.407, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.378490175801448, | |
| "eval_loss": 0.30534103512763977, | |
| "eval_runtime": 31.5782, | |
| "eval_samples_per_second": 15.549, | |
| "eval_steps_per_second": 3.895, | |
| "eval_wer": 0.43305394369224154, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.481902792140641, | |
| "grad_norm": 1.2900283336639404, | |
| "learning_rate": 0.0002761950882047734, | |
| "loss": 0.4314, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.481902792140641, | |
| "eval_loss": 0.30588600039482117, | |
| "eval_runtime": 31.1067, | |
| "eval_samples_per_second": 15.784, | |
| "eval_steps_per_second": 3.954, | |
| "eval_wer": 0.433698689017838, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.5853154084798344, | |
| "grad_norm": 1.6208014488220215, | |
| "learning_rate": 0.00027515738498789345, | |
| "loss": 0.4144, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.5853154084798344, | |
| "eval_loss": 0.30535265803337097, | |
| "eval_runtime": 31.4136, | |
| "eval_samples_per_second": 15.63, | |
| "eval_steps_per_second": 3.915, | |
| "eval_wer": 0.4382119062970127, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.688728024819028, | |
| "grad_norm": 1.7971055507659912, | |
| "learning_rate": 0.0002741196817710135, | |
| "loss": 0.4099, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.688728024819028, | |
| "eval_loss": 0.3059903085231781, | |
| "eval_runtime": 31.3346, | |
| "eval_samples_per_second": 15.67, | |
| "eval_steps_per_second": 3.925, | |
| "eval_wer": 0.4446593595529766, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.688728024819028, | |
| "step": 2600, | |
| "total_flos": 6.22653785519667e+18, | |
| "train_loss": 0.718606931246244, | |
| "train_runtime": 2586.1395, | |
| "train_samples_per_second": 44.847, | |
| "train_steps_per_second": 11.217 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 29010, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.22653785519667e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |