| { | |
| "best_metric": 0.1408591866493225, | |
| "best_model_checkpoint": "./mms-1b-bem-male-sv/checkpoint-4400", | |
| "epoch": 5.0, | |
| "eval_steps": 200, | |
| "global_step": 4580, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2183406113537118, | |
| "eval_loss": 0.1927431970834732, | |
| "eval_runtime": 127.6058, | |
| "eval_samples_per_second": 7.97, | |
| "eval_steps_per_second": 1.003, | |
| "eval_wer": 0.4257075471698113, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4366812227074236, | |
| "eval_loss": 0.1712573915719986, | |
| "eval_runtime": 126.4823, | |
| "eval_samples_per_second": 8.041, | |
| "eval_steps_per_second": 1.012, | |
| "eval_wer": 0.3884958071278826, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5458515283842795, | |
| "grad_norm": 1.9703088998794556, | |
| "learning_rate": 0.000992, | |
| "loss": 2.0358, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6550218340611353, | |
| "eval_loss": 0.17603513598442078, | |
| "eval_runtime": 127.0067, | |
| "eval_samples_per_second": 8.007, | |
| "eval_steps_per_second": 1.008, | |
| "eval_wer": 0.39072327044025157, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8733624454148472, | |
| "eval_loss": 0.18193688988685608, | |
| "eval_runtime": 126.7253, | |
| "eval_samples_per_second": 8.025, | |
| "eval_steps_per_second": 1.01, | |
| "eval_wer": 0.4143081761006289, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.091703056768559, | |
| "grad_norm": 4.094836235046387, | |
| "learning_rate": 0.0008784313725490196, | |
| "loss": 0.519, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.091703056768559, | |
| "eval_loss": 0.16109386086463928, | |
| "eval_runtime": 127.0612, | |
| "eval_samples_per_second": 8.004, | |
| "eval_steps_per_second": 1.007, | |
| "eval_wer": 0.38692348008385746, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.3100436681222707, | |
| "eval_loss": 0.15501286089420319, | |
| "eval_runtime": 126.7082, | |
| "eval_samples_per_second": 8.026, | |
| "eval_steps_per_second": 1.01, | |
| "eval_wer": 0.3735587002096436, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.5283842794759825, | |
| "eval_loss": 0.15383152663707733, | |
| "eval_runtime": 127.2176, | |
| "eval_samples_per_second": 7.994, | |
| "eval_steps_per_second": 1.006, | |
| "eval_wer": 0.3770964360587002, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.6375545851528384, | |
| "grad_norm": 1.7256165742874146, | |
| "learning_rate": 0.0007558823529411764, | |
| "loss": 0.4764, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.7467248908296944, | |
| "eval_loss": 0.1743510365486145, | |
| "eval_runtime": 127.1222, | |
| "eval_samples_per_second": 8.0, | |
| "eval_steps_per_second": 1.007, | |
| "eval_wer": 0.417583857442348, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.965065502183406, | |
| "eval_loss": 0.15977127850055695, | |
| "eval_runtime": 127.3519, | |
| "eval_samples_per_second": 7.986, | |
| "eval_steps_per_second": 1.005, | |
| "eval_wer": 0.38836477987421386, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.183406113537118, | |
| "grad_norm": 2.758544683456421, | |
| "learning_rate": 0.0006333333333333333, | |
| "loss": 0.4501, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.183406113537118, | |
| "eval_loss": 0.15066786110401154, | |
| "eval_runtime": 127.104, | |
| "eval_samples_per_second": 8.001, | |
| "eval_steps_per_second": 1.007, | |
| "eval_wer": 0.3577044025157233, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.4017467248908297, | |
| "eval_loss": 0.15350954234600067, | |
| "eval_runtime": 127.7067, | |
| "eval_samples_per_second": 7.964, | |
| "eval_steps_per_second": 1.002, | |
| "eval_wer": 0.37631027253668764, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.6200873362445414, | |
| "eval_loss": 0.15018954873085022, | |
| "eval_runtime": 127.3689, | |
| "eval_samples_per_second": 7.985, | |
| "eval_steps_per_second": 1.005, | |
| "eval_wer": 0.36491090146750527, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.7292576419213974, | |
| "grad_norm": 0.7292295694351196, | |
| "learning_rate": 0.0005107843137254902, | |
| "loss": 0.4422, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.8384279475982535, | |
| "eval_loss": 0.14573481678962708, | |
| "eval_runtime": 127.0532, | |
| "eval_samples_per_second": 8.005, | |
| "eval_steps_per_second": 1.007, | |
| "eval_wer": 0.35023584905660377, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.056768558951965, | |
| "eval_loss": 0.1484854817390442, | |
| "eval_runtime": 128.0576, | |
| "eval_samples_per_second": 7.942, | |
| "eval_steps_per_second": 1.0, | |
| "eval_wer": 0.3579664570230608, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.2751091703056767, | |
| "grad_norm": 0.6407122015953064, | |
| "learning_rate": 0.0003884803921568628, | |
| "loss": 0.4217, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.2751091703056767, | |
| "eval_loss": 0.148036390542984, | |
| "eval_runtime": 128.302, | |
| "eval_samples_per_second": 7.927, | |
| "eval_steps_per_second": 0.998, | |
| "eval_wer": 0.3546907756813417, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.493449781659389, | |
| "eval_loss": 0.14975149929523468, | |
| "eval_runtime": 127.0597, | |
| "eval_samples_per_second": 8.004, | |
| "eval_steps_per_second": 1.007, | |
| "eval_wer": 0.3666142557651992, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 3.7117903930131004, | |
| "eval_loss": 0.14578010141849518, | |
| "eval_runtime": 127.8692, | |
| "eval_samples_per_second": 7.953, | |
| "eval_steps_per_second": 1.001, | |
| "eval_wer": 0.3494496855345912, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 3.8209606986899565, | |
| "grad_norm": 1.9009268283843994, | |
| "learning_rate": 0.0002659313725490196, | |
| "loss": 0.4144, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.930131004366812, | |
| "eval_loss": 0.1427353024482727, | |
| "eval_runtime": 127.8119, | |
| "eval_samples_per_second": 7.957, | |
| "eval_steps_per_second": 1.001, | |
| "eval_wer": 0.35744234800838576, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 4.148471615720524, | |
| "eval_loss": 0.14451348781585693, | |
| "eval_runtime": 127.6381, | |
| "eval_samples_per_second": 7.968, | |
| "eval_steps_per_second": 1.003, | |
| "eval_wer": 0.3594077568134172, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 4.366812227074236, | |
| "grad_norm": 1.5231894254684448, | |
| "learning_rate": 0.00014338235294117645, | |
| "loss": 0.3926, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.366812227074236, | |
| "eval_loss": 0.14618775248527527, | |
| "eval_runtime": 127.9874, | |
| "eval_samples_per_second": 7.946, | |
| "eval_steps_per_second": 1.0, | |
| "eval_wer": 0.3666142557651992, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.585152838427947, | |
| "eval_loss": 0.14320309460163116, | |
| "eval_runtime": 128.1452, | |
| "eval_samples_per_second": 7.936, | |
| "eval_steps_per_second": 0.999, | |
| "eval_wer": 0.3527253668763103, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 4.8034934497816595, | |
| "eval_loss": 0.1408591866493225, | |
| "eval_runtime": 127.7676, | |
| "eval_samples_per_second": 7.96, | |
| "eval_steps_per_second": 1.002, | |
| "eval_wer": 0.3498427672955975, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 4.9126637554585155, | |
| "grad_norm": 6.477123737335205, | |
| "learning_rate": 2.0833333333333333e-05, | |
| "loss": 0.3928, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4580, | |
| "total_flos": 2.179329538942206e+19, | |
| "train_loss": 0.6123242174173547, | |
| "train_runtime": 13424.2744, | |
| "train_samples_per_second": 2.729, | |
| "train_steps_per_second": 0.341 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4580, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.179329538942206e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |