| { | |
| "best_metric": 0.21319285035133362, | |
| "best_model_checkpoint": "./mms-1b-bem-female-sv/checkpoint-2400", | |
| "epoch": 5.0, | |
| "eval_steps": 200, | |
| "global_step": 2505, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.3992015968063872, | |
| "eval_loss": 0.3566432595252991, | |
| "eval_runtime": 69.3187, | |
| "eval_samples_per_second": 6.852, | |
| "eval_steps_per_second": 0.866, | |
| "eval_wer": 0.5019098548510313, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7984031936127745, | |
| "eval_loss": 0.26201513409614563, | |
| "eval_runtime": 68.6891, | |
| "eval_samples_per_second": 6.915, | |
| "eval_steps_per_second": 0.874, | |
| "eval_wer": 0.4028520499108734, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.998003992015968, | |
| "grad_norm": 1.1227352619171143, | |
| "learning_rate": 0.000994, | |
| "loss": 1.7214, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.1976047904191618, | |
| "eval_loss": 0.2546118199825287, | |
| "eval_runtime": 69.4229, | |
| "eval_samples_per_second": 6.842, | |
| "eval_steps_per_second": 0.864, | |
| "eval_wer": 0.40998217468805703, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.596806387225549, | |
| "eval_loss": 0.2358996421098709, | |
| "eval_runtime": 68.6925, | |
| "eval_samples_per_second": 6.915, | |
| "eval_steps_per_second": 0.873, | |
| "eval_wer": 0.3964858670741024, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.996007984031936, | |
| "grad_norm": 2.2609593868255615, | |
| "learning_rate": 0.0007531172069825436, | |
| "loss": 0.2801, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.996007984031936, | |
| "eval_loss": 0.23218725621700287, | |
| "eval_runtime": 68.6733, | |
| "eval_samples_per_second": 6.917, | |
| "eval_steps_per_second": 0.874, | |
| "eval_wer": 0.38095238095238093, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.3952095808383236, | |
| "eval_loss": 0.23050391674041748, | |
| "eval_runtime": 68.6828, | |
| "eval_samples_per_second": 6.916, | |
| "eval_steps_per_second": 0.874, | |
| "eval_wer": 0.3745861981156099, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.7944111776447107, | |
| "eval_loss": 0.22583143413066864, | |
| "eval_runtime": 68.7708, | |
| "eval_samples_per_second": 6.907, | |
| "eval_steps_per_second": 0.872, | |
| "eval_wer": 0.33358798064680417, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.9940119760479043, | |
| "grad_norm": 0.9439337849617004, | |
| "learning_rate": 0.0005037406483790524, | |
| "loss": 0.2528, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.193612774451098, | |
| "eval_loss": 0.22624839842319489, | |
| "eval_runtime": 68.8386, | |
| "eval_samples_per_second": 6.9, | |
| "eval_steps_per_second": 0.872, | |
| "eval_wer": 0.43086325439266615, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.592814371257485, | |
| "eval_loss": 0.2164432853460312, | |
| "eval_runtime": 68.775, | |
| "eval_samples_per_second": 6.907, | |
| "eval_steps_per_second": 0.872, | |
| "eval_wer": 0.35141329258976317, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.992015968063872, | |
| "grad_norm": 1.3000344038009644, | |
| "learning_rate": 0.0002543640897755611, | |
| "loss": 0.2351, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.992015968063872, | |
| "eval_loss": 0.22146381437778473, | |
| "eval_runtime": 68.601, | |
| "eval_samples_per_second": 6.924, | |
| "eval_steps_per_second": 0.875, | |
| "eval_wer": 0.38935574229691877, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.391217564870259, | |
| "eval_loss": 0.21647760272026062, | |
| "eval_runtime": 68.6439, | |
| "eval_samples_per_second": 6.92, | |
| "eval_steps_per_second": 0.874, | |
| "eval_wer": 0.3623631270690094, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.790419161676647, | |
| "eval_loss": 0.21319285035133362, | |
| "eval_runtime": 68.6858, | |
| "eval_samples_per_second": 6.916, | |
| "eval_steps_per_second": 0.874, | |
| "eval_wer": 0.3557422969187675, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.99001996007984, | |
| "grad_norm": 14.332290649414062, | |
| "learning_rate": 4.987531172069825e-06, | |
| "loss": 0.2222, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 2505, | |
| "total_flos": 1.1971485322227634e+19, | |
| "train_loss": 0.5417278716188229, | |
| "train_runtime": 6598.5921, | |
| "train_samples_per_second": 3.036, | |
| "train_steps_per_second": 0.38 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2505, | |
| "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": 1.1971485322227634e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |