| { | |
| "best_metric": 0.6841106414794922, | |
| "best_model_checkpoint": "/scratch/skscla001/results/xls-r-1b-bem-natbed-native-model/checkpoint-1000", | |
| "epoch": 7.303370786516854, | |
| "eval_steps": 100, | |
| "global_step": 1300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5617977528089888, | |
| "grad_norm": 4.858776092529297, | |
| "learning_rate": 0.00029099999999999997, | |
| "loss": 4.5137, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5617977528089888, | |
| "eval_loss": 2.554856777191162, | |
| "eval_runtime": 50.0091, | |
| "eval_samples_per_second": 12.998, | |
| "eval_steps_per_second": 1.64, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.1235955056179776, | |
| "grad_norm": 3.9340174198150635, | |
| "learning_rate": 0.00029444656488549615, | |
| "loss": 1.3916, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.1235955056179776, | |
| "eval_loss": 1.088287115097046, | |
| "eval_runtime": 49.3058, | |
| "eval_samples_per_second": 13.183, | |
| "eval_steps_per_second": 1.663, | |
| "eval_wer": 0.9840240430243594, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6853932584269664, | |
| "grad_norm": 1.637488603591919, | |
| "learning_rate": 0.00028872137404580147, | |
| "loss": 0.9962, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.6853932584269664, | |
| "eval_loss": 0.8152701258659363, | |
| "eval_runtime": 50.2981, | |
| "eval_samples_per_second": 12.923, | |
| "eval_steps_per_second": 1.63, | |
| "eval_wer": 0.8190446061372983, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.247191011235955, | |
| "grad_norm": 1.2791478633880615, | |
| "learning_rate": 0.00028299618320610685, | |
| "loss": 0.8625, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.247191011235955, | |
| "eval_loss": 0.8690391182899475, | |
| "eval_runtime": 49.7576, | |
| "eval_samples_per_second": 13.063, | |
| "eval_steps_per_second": 1.648, | |
| "eval_wer": 0.8418222081619741, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.808988764044944, | |
| "grad_norm": 1.049700379371643, | |
| "learning_rate": 0.00027727099236641217, | |
| "loss": 0.8168, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.808988764044944, | |
| "eval_loss": 0.7395117282867432, | |
| "eval_runtime": 49.4925, | |
| "eval_samples_per_second": 13.133, | |
| "eval_steps_per_second": 1.657, | |
| "eval_wer": 0.7390066434672572, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.370786516853933, | |
| "grad_norm": 2.349531650543213, | |
| "learning_rate": 0.00027154580152671755, | |
| "loss": 0.7197, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.370786516853933, | |
| "eval_loss": 0.7596462965011597, | |
| "eval_runtime": 49.3475, | |
| "eval_samples_per_second": 13.172, | |
| "eval_steps_per_second": 1.662, | |
| "eval_wer": 0.7366339765896868, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.932584269662921, | |
| "grad_norm": 0.8924151062965393, | |
| "learning_rate": 0.00026582061068702287, | |
| "loss": 0.6848, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.932584269662921, | |
| "eval_loss": 0.7033310532569885, | |
| "eval_runtime": 50.4297, | |
| "eval_samples_per_second": 12.889, | |
| "eval_steps_per_second": 1.626, | |
| "eval_wer": 0.7228725086997786, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.49438202247191, | |
| "grad_norm": 1.047180414199829, | |
| "learning_rate": 0.0002600954198473282, | |
| "loss": 0.6134, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.49438202247191, | |
| "eval_loss": 0.830004870891571, | |
| "eval_runtime": 50.1876, | |
| "eval_samples_per_second": 12.951, | |
| "eval_steps_per_second": 1.634, | |
| "eval_wer": 0.7662132236633976, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.056179775280899, | |
| "grad_norm": 0.8339403867721558, | |
| "learning_rate": 0.00025437022900763357, | |
| "loss": 0.6303, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.056179775280899, | |
| "eval_loss": 0.73649001121521, | |
| "eval_runtime": 49.3956, | |
| "eval_samples_per_second": 13.159, | |
| "eval_steps_per_second": 1.66, | |
| "eval_wer": 0.7896235368554255, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.617977528089888, | |
| "grad_norm": 0.746986448764801, | |
| "learning_rate": 0.0002486450381679389, | |
| "loss": 0.5467, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.617977528089888, | |
| "eval_loss": 0.6841106414794922, | |
| "eval_runtime": 49.7948, | |
| "eval_samples_per_second": 13.054, | |
| "eval_steps_per_second": 1.647, | |
| "eval_wer": 0.7486554887693768, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.179775280898877, | |
| "grad_norm": 0.8229517936706543, | |
| "learning_rate": 0.00024291984732824427, | |
| "loss": 0.5194, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 6.179775280898877, | |
| "eval_loss": 0.7867633700370789, | |
| "eval_runtime": 49.8846, | |
| "eval_samples_per_second": 13.03, | |
| "eval_steps_per_second": 1.644, | |
| "eval_wer": 0.694875039544448, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 6.741573033707866, | |
| "grad_norm": 0.969409704208374, | |
| "learning_rate": 0.0002371946564885496, | |
| "loss": 0.4617, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 6.741573033707866, | |
| "eval_loss": 0.7563472986221313, | |
| "eval_runtime": 49.4271, | |
| "eval_samples_per_second": 13.151, | |
| "eval_steps_per_second": 1.659, | |
| "eval_wer": 0.7277760202467574, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 7.303370786516854, | |
| "grad_norm": 0.5057498812675476, | |
| "learning_rate": 0.00023146946564885494, | |
| "loss": 0.4525, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 7.303370786516854, | |
| "eval_loss": 0.7276196479797363, | |
| "eval_runtime": 49.3168, | |
| "eval_samples_per_second": 13.18, | |
| "eval_steps_per_second": 1.663, | |
| "eval_wer": 0.6730465042708004, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 7.303370786516854, | |
| "step": 1300, | |
| "total_flos": 1.4772414251163556e+19, | |
| "train_loss": 1.0161105111929087, | |
| "train_runtime": 3259.7695, | |
| "train_samples_per_second": 26.183, | |
| "train_steps_per_second": 1.638 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 5340, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 2 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4772414251163556e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |