| { | |
| "best_metric": 0.24462078511714935, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/xls-r-1b-bemgen-combined-model/checkpoint-1500", | |
| "epoch": 3.3871543264942017, | |
| "eval_steps": 100, | |
| "global_step": 1900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1784121320249777, | |
| "eval_loss": 3.441302537918091, | |
| "eval_runtime": 39.9204, | |
| "eval_samples_per_second": 13.552, | |
| "eval_steps_per_second": 3.407, | |
| "eval_wer": 1.0002904865649964, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3568242640499554, | |
| "eval_loss": 2.914942979812622, | |
| "eval_runtime": 39.4152, | |
| "eval_samples_per_second": 13.726, | |
| "eval_steps_per_second": 3.45, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5352363960749331, | |
| "eval_loss": 0.7767665386199951, | |
| "eval_runtime": 39.9772, | |
| "eval_samples_per_second": 13.533, | |
| "eval_steps_per_second": 3.402, | |
| "eval_wer": 0.9234567901234568, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7136485280999108, | |
| "eval_loss": 0.605692446231842, | |
| "eval_runtime": 39.422, | |
| "eval_samples_per_second": 13.723, | |
| "eval_steps_per_second": 3.45, | |
| "eval_wer": 0.904720406681191, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8920606601248885, | |
| "grad_norm": 3.2083628177642822, | |
| "learning_rate": 2.976e-05, | |
| "loss": 5.3372, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8920606601248885, | |
| "eval_loss": 0.43167629837989807, | |
| "eval_runtime": 39.6193, | |
| "eval_samples_per_second": 13.655, | |
| "eval_steps_per_second": 3.433, | |
| "eval_wer": 0.6720406681190995, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0695807314897412, | |
| "eval_loss": 0.3997327983379364, | |
| "eval_runtime": 39.7537, | |
| "eval_samples_per_second": 13.609, | |
| "eval_steps_per_second": 3.421, | |
| "eval_wer": 0.6704429920116195, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.247992863514719, | |
| "eval_loss": 0.3610936999320984, | |
| "eval_runtime": 39.4379, | |
| "eval_samples_per_second": 13.718, | |
| "eval_steps_per_second": 3.448, | |
| "eval_wer": 0.6405228758169934, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.4264049955396967, | |
| "eval_loss": 0.34406763315200806, | |
| "eval_runtime": 39.4181, | |
| "eval_samples_per_second": 13.725, | |
| "eval_steps_per_second": 3.45, | |
| "eval_wer": 0.5603485838779957, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.6048171275646745, | |
| "eval_loss": 0.2945486903190613, | |
| "eval_runtime": 39.8239, | |
| "eval_samples_per_second": 13.585, | |
| "eval_steps_per_second": 3.415, | |
| "eval_wer": 0.49135802469135803, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.783229259589652, | |
| "grad_norm": 3.3745229244232178, | |
| "learning_rate": 2.908711656441718e-05, | |
| "loss": 0.6459, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.783229259589652, | |
| "eval_loss": 0.3041314482688904, | |
| "eval_runtime": 39.339, | |
| "eval_samples_per_second": 13.752, | |
| "eval_steps_per_second": 3.457, | |
| "eval_wer": 0.4923747276688453, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9616413916146298, | |
| "eval_loss": 0.2805267870426178, | |
| "eval_runtime": 39.2725, | |
| "eval_samples_per_second": 13.776, | |
| "eval_steps_per_second": 3.463, | |
| "eval_wer": 0.46811909949164854, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.1391614629794824, | |
| "eval_loss": 0.2774401605129242, | |
| "eval_runtime": 39.4398, | |
| "eval_samples_per_second": 13.717, | |
| "eval_steps_per_second": 3.448, | |
| "eval_wer": 0.5108206245461148, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.3175735950044603, | |
| "eval_loss": 0.26826903223991394, | |
| "eval_runtime": 39.6616, | |
| "eval_samples_per_second": 13.64, | |
| "eval_steps_per_second": 3.429, | |
| "eval_wer": 0.42541757443718226, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.495985727029438, | |
| "eval_loss": 0.2643994987010956, | |
| "eval_runtime": 39.3737, | |
| "eval_samples_per_second": 13.74, | |
| "eval_steps_per_second": 3.454, | |
| "eval_wer": 0.4381989832970225, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.674397859054416, | |
| "grad_norm": 2.350830554962158, | |
| "learning_rate": 2.816687116564417e-05, | |
| "loss": 0.4599, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.674397859054416, | |
| "eval_loss": 0.24462078511714935, | |
| "eval_runtime": 39.5477, | |
| "eval_samples_per_second": 13.68, | |
| "eval_steps_per_second": 3.439, | |
| "eval_wer": 0.4142338416848221, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.8528099910793934, | |
| "eval_loss": 0.24734348058700562, | |
| "eval_runtime": 39.4462, | |
| "eval_samples_per_second": 13.715, | |
| "eval_steps_per_second": 3.448, | |
| "eval_wer": 0.4117647058823529, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.0303300624442464, | |
| "eval_loss": 0.24915921688079834, | |
| "eval_runtime": 39.9106, | |
| "eval_samples_per_second": 13.555, | |
| "eval_steps_per_second": 3.408, | |
| "eval_wer": 0.396078431372549, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.208742194469224, | |
| "eval_loss": 0.24668627977371216, | |
| "eval_runtime": 39.4647, | |
| "eval_samples_per_second": 13.708, | |
| "eval_steps_per_second": 3.446, | |
| "eval_wer": 0.40697167755991287, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.3871543264942017, | |
| "eval_loss": 0.25092557072639465, | |
| "eval_runtime": 39.3684, | |
| "eval_samples_per_second": 13.742, | |
| "eval_steps_per_second": 3.455, | |
| "eval_wer": 0.39230210602759624, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.3871543264942017, | |
| "step": 1900, | |
| "total_flos": 1.1803978917941375e+19, | |
| "train_loss": 1.775066086618524, | |
| "train_runtime": 3182.4103, | |
| "train_samples_per_second": 42.251, | |
| "train_steps_per_second": 5.279 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 16800, | |
| "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": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.1803978917941375e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |