| { | |
| "best_metric": 0.6817988753318787, | |
| "best_model_checkpoint": "/scratch/skscla001/results/xls-r-1b-bem-natbed-non-native-model/checkpoint-1000", | |
| "epoch": 5.8690744920993225, | |
| "eval_steps": 100, | |
| "global_step": 1300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.45146726862302483, | |
| "grad_norm": 2.7790069580078125, | |
| "learning_rate": 0.00029099999999999997, | |
| "loss": 3.8975, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.45146726862302483, | |
| "eval_loss": 1.1342732906341553, | |
| "eval_runtime": 44.5102, | |
| "eval_samples_per_second": 15.906, | |
| "eval_steps_per_second": 2.0, | |
| "eval_wer": 0.9315557065217391, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9029345372460497, | |
| "grad_norm": 1.7560482025146484, | |
| "learning_rate": 0.00029554364471669214, | |
| "loss": 1.0247, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9029345372460497, | |
| "eval_loss": 0.9810584187507629, | |
| "eval_runtime": 43.9801, | |
| "eval_samples_per_second": 16.098, | |
| "eval_steps_per_second": 2.024, | |
| "eval_wer": 0.8826426630434783, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3544018058690745, | |
| "grad_norm": 1.5388858318328857, | |
| "learning_rate": 0.00029094946401225113, | |
| "loss": 0.9631, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.3544018058690745, | |
| "eval_loss": 0.810827374458313, | |
| "eval_runtime": 43.7967, | |
| "eval_samples_per_second": 16.166, | |
| "eval_steps_per_second": 2.032, | |
| "eval_wer": 0.7523777173913043, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.8058690744920993, | |
| "grad_norm": 2.79054856300354, | |
| "learning_rate": 0.00028635528330781006, | |
| "loss": 0.8711, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.8058690744920993, | |
| "eval_loss": 0.7748640775680542, | |
| "eval_runtime": 43.9031, | |
| "eval_samples_per_second": 16.126, | |
| "eval_steps_per_second": 2.027, | |
| "eval_wer": 0.7637567934782609, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.2573363431151243, | |
| "grad_norm": 2.239819288253784, | |
| "learning_rate": 0.00028176110260336905, | |
| "loss": 0.7935, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.2573363431151243, | |
| "eval_loss": 0.8008168339729309, | |
| "eval_runtime": 44.1964, | |
| "eval_samples_per_second": 16.019, | |
| "eval_steps_per_second": 2.014, | |
| "eval_wer": 0.8231997282608695, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.708803611738149, | |
| "grad_norm": 1.130868673324585, | |
| "learning_rate": 0.000277166921898928, | |
| "loss": 0.7869, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.708803611738149, | |
| "eval_loss": 0.7333548069000244, | |
| "eval_runtime": 44.3134, | |
| "eval_samples_per_second": 15.977, | |
| "eval_steps_per_second": 2.008, | |
| "eval_wer": 0.7262228260869565, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.160270880361174, | |
| "grad_norm": 1.1080917119979858, | |
| "learning_rate": 0.000272572741194487, | |
| "loss": 0.7265, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.160270880361174, | |
| "eval_loss": 0.7238579988479614, | |
| "eval_runtime": 44.0849, | |
| "eval_samples_per_second": 16.06, | |
| "eval_steps_per_second": 2.019, | |
| "eval_wer": 0.6949728260869565, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.6117381489841986, | |
| "grad_norm": 1.219346523284912, | |
| "learning_rate": 0.0002679785604900459, | |
| "loss": 0.7011, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.6117381489841986, | |
| "eval_loss": 0.70769864320755, | |
| "eval_runtime": 44.0752, | |
| "eval_samples_per_second": 16.063, | |
| "eval_steps_per_second": 2.019, | |
| "eval_wer": 0.6807065217391305, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.063205417607223, | |
| "grad_norm": 0.6558771729469299, | |
| "learning_rate": 0.00026338437978560485, | |
| "loss": 0.7343, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.063205417607223, | |
| "eval_loss": 0.6984588503837585, | |
| "eval_runtime": 44.1425, | |
| "eval_samples_per_second": 16.039, | |
| "eval_steps_per_second": 2.016, | |
| "eval_wer": 0.69140625, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.514672686230249, | |
| "grad_norm": 1.1139057874679565, | |
| "learning_rate": 0.00025879019908116383, | |
| "loss": 0.6624, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.514672686230249, | |
| "eval_loss": 0.6817988753318787, | |
| "eval_runtime": 44.1617, | |
| "eval_samples_per_second": 16.032, | |
| "eval_steps_per_second": 2.015, | |
| "eval_wer": 0.7141644021739131, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.966139954853273, | |
| "grad_norm": 0.7626570463180542, | |
| "learning_rate": 0.0002541960183767228, | |
| "loss": 0.6642, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.966139954853273, | |
| "eval_loss": 0.7096632122993469, | |
| "eval_runtime": 44.5476, | |
| "eval_samples_per_second": 15.893, | |
| "eval_steps_per_second": 1.998, | |
| "eval_wer": 0.6968410326086957, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 5.417607223476298, | |
| "grad_norm": 0.8168011903762817, | |
| "learning_rate": 0.00024960183767228176, | |
| "loss": 0.6064, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.417607223476298, | |
| "eval_loss": 0.7194671630859375, | |
| "eval_runtime": 44.5485, | |
| "eval_samples_per_second": 15.893, | |
| "eval_steps_per_second": 1.998, | |
| "eval_wer": 0.6856317934782609, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.8690744920993225, | |
| "grad_norm": 1.018847942352295, | |
| "learning_rate": 0.0002450076569678407, | |
| "loss": 0.5867, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.8690744920993225, | |
| "eval_loss": 0.6899004578590393, | |
| "eval_runtime": 43.9975, | |
| "eval_samples_per_second": 16.092, | |
| "eval_steps_per_second": 2.023, | |
| "eval_wer": 0.6880095108695652, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.8690744920993225, | |
| "step": 1300, | |
| "total_flos": 1.2230713692520196e+19, | |
| "train_loss": 1.001419219970703, | |
| "train_runtime": 2976.6499, | |
| "train_samples_per_second": 35.698, | |
| "train_steps_per_second": 2.227 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 6630, | |
| "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": 1.2230713692520196e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |