| { | |
| "best_metric": 0.473453164100647, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-bigcgen-male-15hrs-model/checkpoint-1700", | |
| "epoch": 2.169421487603306, | |
| "eval_steps": 100, | |
| "global_step": 2100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.10330578512396695, | |
| "grad_norm": 5.168842792510986, | |
| "learning_rate": 0.000285, | |
| "loss": 14.7965, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10330578512396695, | |
| "eval_loss": 3.522738456726074, | |
| "eval_runtime": 30.9773, | |
| "eval_samples_per_second": 14.236, | |
| "eval_steps_per_second": 3.583, | |
| "eval_wer": 1.0007213272421256, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2066115702479339, | |
| "grad_norm": 4.0135955810546875, | |
| "learning_rate": 0.000299015203870076, | |
| "loss": 6.2688, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2066115702479339, | |
| "eval_loss": 2.770048141479492, | |
| "eval_runtime": 30.9601, | |
| "eval_samples_per_second": 14.244, | |
| "eval_steps_per_second": 3.585, | |
| "eval_wer": 1.032459725895648, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.30991735537190085, | |
| "grad_norm": 2.5909135341644287, | |
| "learning_rate": 0.00029797857636489284, | |
| "loss": 3.6179, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.30991735537190085, | |
| "eval_loss": 0.753160834312439, | |
| "eval_runtime": 30.8437, | |
| "eval_samples_per_second": 14.298, | |
| "eval_steps_per_second": 3.599, | |
| "eval_wer": 0.6068766530415965, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4132231404958678, | |
| "grad_norm": 4.135544776916504, | |
| "learning_rate": 0.0002969419488597097, | |
| "loss": 1.7779, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4132231404958678, | |
| "eval_loss": 0.650780200958252, | |
| "eval_runtime": 30.8789, | |
| "eval_samples_per_second": 14.282, | |
| "eval_steps_per_second": 3.595, | |
| "eval_wer": 0.5821110843952874, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5165289256198347, | |
| "grad_norm": 2.8424060344696045, | |
| "learning_rate": 0.0002959053213545266, | |
| "loss": 1.5595, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5165289256198347, | |
| "eval_loss": 0.6249143481254578, | |
| "eval_runtime": 31.2101, | |
| "eval_samples_per_second": 14.13, | |
| "eval_steps_per_second": 3.557, | |
| "eval_wer": 0.5578264005770618, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6198347107438017, | |
| "grad_norm": 3.2064597606658936, | |
| "learning_rate": 0.00029486869384934343, | |
| "loss": 1.5884, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6198347107438017, | |
| "eval_loss": 0.6142242550849915, | |
| "eval_runtime": 30.7532, | |
| "eval_samples_per_second": 14.34, | |
| "eval_steps_per_second": 3.609, | |
| "eval_wer": 0.5282519836499159, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7231404958677686, | |
| "grad_norm": 3.0032007694244385, | |
| "learning_rate": 0.0002938320663441603, | |
| "loss": 1.5532, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7231404958677686, | |
| "eval_loss": 0.5929429531097412, | |
| "eval_runtime": 31.0147, | |
| "eval_samples_per_second": 14.219, | |
| "eval_steps_per_second": 3.579, | |
| "eval_wer": 0.5171916326039914, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.8264462809917356, | |
| "grad_norm": 3.4867029190063477, | |
| "learning_rate": 0.00029279543883897714, | |
| "loss": 1.4021, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8264462809917356, | |
| "eval_loss": 0.5995635390281677, | |
| "eval_runtime": 31.1021, | |
| "eval_samples_per_second": 14.179, | |
| "eval_steps_per_second": 3.569, | |
| "eval_wer": 0.5181534022601587, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.9297520661157025, | |
| "grad_norm": 1.893242359161377, | |
| "learning_rate": 0.000291758811333794, | |
| "loss": 1.507, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9297520661157025, | |
| "eval_loss": 0.5824074149131775, | |
| "eval_runtime": 31.2256, | |
| "eval_samples_per_second": 14.123, | |
| "eval_steps_per_second": 3.555, | |
| "eval_wer": 0.5121423419091128, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.0330578512396693, | |
| "grad_norm": 5.389484882354736, | |
| "learning_rate": 0.0002907221838286109, | |
| "loss": 1.5374, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0330578512396693, | |
| "eval_loss": 0.5614578723907471, | |
| "eval_runtime": 31.1303, | |
| "eval_samples_per_second": 14.166, | |
| "eval_steps_per_second": 3.566, | |
| "eval_wer": 0.5061312815580669, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.1363636363636362, | |
| "grad_norm": 3.9391493797302246, | |
| "learning_rate": 0.0002896855563234278, | |
| "loss": 1.4139, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.1363636363636362, | |
| "eval_loss": 0.5456417798995972, | |
| "eval_runtime": 30.8696, | |
| "eval_samples_per_second": 14.286, | |
| "eval_steps_per_second": 3.596, | |
| "eval_wer": 0.5066121663861505, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.2396694214876034, | |
| "grad_norm": 2.858807325363159, | |
| "learning_rate": 0.0002886489288182446, | |
| "loss": 1.4472, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.2396694214876034, | |
| "eval_loss": 0.5177425146102905, | |
| "eval_runtime": 31.4742, | |
| "eval_samples_per_second": 14.011, | |
| "eval_steps_per_second": 3.527, | |
| "eval_wer": 0.48737677326280354, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.3429752066115703, | |
| "grad_norm": 2.0307440757751465, | |
| "learning_rate": 0.0002876123013130615, | |
| "loss": 1.2958, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.3429752066115703, | |
| "eval_loss": 0.5022083520889282, | |
| "eval_runtime": 31.2404, | |
| "eval_samples_per_second": 14.116, | |
| "eval_steps_per_second": 3.553, | |
| "eval_wer": 0.48713633084876173, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.4462809917355373, | |
| "grad_norm": 1.2417099475860596, | |
| "learning_rate": 0.0002865756738078783, | |
| "loss": 1.3292, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.4462809917355373, | |
| "eval_loss": 0.49844667315483093, | |
| "eval_runtime": 31.0915, | |
| "eval_samples_per_second": 14.184, | |
| "eval_steps_per_second": 3.57, | |
| "eval_wer": 0.48713633084876173, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.549586776859504, | |
| "grad_norm": 1.5981268882751465, | |
| "learning_rate": 0.0002855390463026952, | |
| "loss": 1.2062, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.549586776859504, | |
| "eval_loss": 0.48859983682632446, | |
| "eval_runtime": 31.0946, | |
| "eval_samples_per_second": 14.183, | |
| "eval_steps_per_second": 3.57, | |
| "eval_wer": 0.4799230584275066, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.6528925619834711, | |
| "grad_norm": 10.061776161193848, | |
| "learning_rate": 0.0002845024187975121, | |
| "loss": 1.1623, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.6528925619834711, | |
| "eval_loss": 0.4811255931854248, | |
| "eval_runtime": 31.525, | |
| "eval_samples_per_second": 13.989, | |
| "eval_steps_per_second": 3.521, | |
| "eval_wer": 0.48232748256792496, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.756198347107438, | |
| "grad_norm": 3.169590473175049, | |
| "learning_rate": 0.0002834657912923289, | |
| "loss": 1.2759, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.756198347107438, | |
| "eval_loss": 0.473453164100647, | |
| "eval_runtime": 31.3702, | |
| "eval_samples_per_second": 14.058, | |
| "eval_steps_per_second": 3.538, | |
| "eval_wer": 0.46766049531137294, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.859504132231405, | |
| "grad_norm": 3.227973222732544, | |
| "learning_rate": 0.0002824291637871458, | |
| "loss": 1.1852, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.859504132231405, | |
| "eval_loss": 0.4986236095428467, | |
| "eval_runtime": 31.0563, | |
| "eval_samples_per_second": 14.2, | |
| "eval_steps_per_second": 3.574, | |
| "eval_wer": 0.4669391680692474, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.962809917355372, | |
| "grad_norm": 3.3289122581481934, | |
| "learning_rate": 0.0002813925362819626, | |
| "loss": 1.0712, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.962809917355372, | |
| "eval_loss": 0.5045417547225952, | |
| "eval_runtime": 31.1996, | |
| "eval_samples_per_second": 14.135, | |
| "eval_steps_per_second": 3.558, | |
| "eval_wer": 0.4844914642943015, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.0661157024793386, | |
| "grad_norm": 2.1973116397857666, | |
| "learning_rate": 0.0002803559087767795, | |
| "loss": 1.2023, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.0661157024793386, | |
| "eval_loss": 0.475466787815094, | |
| "eval_runtime": 31.3451, | |
| "eval_samples_per_second": 14.069, | |
| "eval_steps_per_second": 3.541, | |
| "eval_wer": 0.4782399615292138, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.169421487603306, | |
| "grad_norm": 1.9049010276794434, | |
| "learning_rate": 0.0002793192812715964, | |
| "loss": 1.2275, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.169421487603306, | |
| "eval_loss": 0.47560110688209534, | |
| "eval_runtime": 31.4981, | |
| "eval_samples_per_second": 14.001, | |
| "eval_steps_per_second": 3.524, | |
| "eval_wer": 0.47054580427987497, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.169421487603306, | |
| "step": 2100, | |
| "total_flos": 1.0989151242100367e+19, | |
| "train_loss": 2.3535848345075334, | |
| "train_runtime": 3049.1563, | |
| "train_samples_per_second": 76.192, | |
| "train_steps_per_second": 9.524 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 29040, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 4, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 3 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.0989151242100367e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |