| { | |
| "best_metric": 0.5530011653900146, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/xls-r-1b-bigcgen-combined-30hrs-model/checkpoint-2700", | |
| "epoch": 2.180028129395218, | |
| "eval_steps": 100, | |
| "global_step": 3100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07032348804500703, | |
| "eval_loss": 3.5623645782470703, | |
| "eval_runtime": 51.151, | |
| "eval_samples_per_second": 13.333, | |
| "eval_steps_per_second": 3.343, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14064697609001406, | |
| "eval_loss": 2.7976131439208984, | |
| "eval_runtime": 50.6538, | |
| "eval_samples_per_second": 13.464, | |
| "eval_steps_per_second": 3.376, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2109704641350211, | |
| "eval_loss": 1.7476502656936646, | |
| "eval_runtime": 50.7435, | |
| "eval_samples_per_second": 13.44, | |
| "eval_steps_per_second": 3.37, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2812939521800281, | |
| "eval_loss": 0.955557644367218, | |
| "eval_runtime": 50.779, | |
| "eval_samples_per_second": 13.431, | |
| "eval_steps_per_second": 3.368, | |
| "eval_wer": 0.9632130063378341, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.35161744022503516, | |
| "grad_norm": 5.9262189865112305, | |
| "learning_rate": 3e-05, | |
| "loss": 5.9145, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.35161744022503516, | |
| "eval_loss": 0.8886582851409912, | |
| "eval_runtime": 51.3778, | |
| "eval_samples_per_second": 13.274, | |
| "eval_steps_per_second": 3.328, | |
| "eval_wer": 0.8955635161201433, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4219409282700422, | |
| "eval_loss": 0.8702635765075684, | |
| "eval_runtime": 50.7264, | |
| "eval_samples_per_second": 13.445, | |
| "eval_steps_per_second": 3.371, | |
| "eval_wer": 0.9243593276384678, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.49226441631504925, | |
| "eval_loss": 0.6984609961509705, | |
| "eval_runtime": 50.919, | |
| "eval_samples_per_second": 13.394, | |
| "eval_steps_per_second": 3.358, | |
| "eval_wer": 0.7485533204739597, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5625879043600562, | |
| "eval_loss": 0.8052944540977478, | |
| "eval_runtime": 51.2331, | |
| "eval_samples_per_second": 13.312, | |
| "eval_steps_per_second": 3.338, | |
| "eval_wer": 0.8899145770184624, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6329113924050633, | |
| "eval_loss": 0.6571323871612549, | |
| "eval_runtime": 51.5133, | |
| "eval_samples_per_second": 13.239, | |
| "eval_steps_per_second": 3.32, | |
| "eval_wer": 0.6398456875172224, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7032348804500703, | |
| "grad_norm": 16.431087493896484, | |
| "learning_rate": 2.9644212523719166e-05, | |
| "loss": 1.3395, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7032348804500703, | |
| "eval_loss": 0.6412575840950012, | |
| "eval_runtime": 51.0276, | |
| "eval_samples_per_second": 13.365, | |
| "eval_steps_per_second": 3.351, | |
| "eval_wer": 0.6394323505097823, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7735583684950773, | |
| "eval_loss": 0.600770115852356, | |
| "eval_runtime": 51.3069, | |
| "eval_samples_per_second": 13.293, | |
| "eval_steps_per_second": 3.333, | |
| "eval_wer": 0.5698539542573712, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.8438818565400844, | |
| "eval_loss": 0.6437661051750183, | |
| "eval_runtime": 51.3281, | |
| "eval_samples_per_second": 13.287, | |
| "eval_steps_per_second": 3.332, | |
| "eval_wer": 0.6576191788371453, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9142053445850914, | |
| "eval_loss": 0.6002921462059021, | |
| "eval_runtime": 51.7544, | |
| "eval_samples_per_second": 13.178, | |
| "eval_steps_per_second": 3.304, | |
| "eval_wer": 0.6022320198401764, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.9845288326300985, | |
| "eval_loss": 0.7025433778762817, | |
| "eval_runtime": 51.575, | |
| "eval_samples_per_second": 13.223, | |
| "eval_steps_per_second": 3.316, | |
| "eval_wer": 0.7080462937448333, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.0548523206751055, | |
| "grad_norm": 2.604250431060791, | |
| "learning_rate": 2.928842504743833e-05, | |
| "loss": 1.1915, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.0548523206751055, | |
| "eval_loss": 0.8042856454849243, | |
| "eval_runtime": 51.7495, | |
| "eval_samples_per_second": 13.179, | |
| "eval_steps_per_second": 3.304, | |
| "eval_wer": 0.8568476164232571, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.1251758087201125, | |
| "eval_loss": 0.6106212139129639, | |
| "eval_runtime": 51.3078, | |
| "eval_samples_per_second": 13.292, | |
| "eval_steps_per_second": 3.333, | |
| "eval_wer": 0.6685037200330669, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.1954992967651195, | |
| "eval_loss": 0.6002780795097351, | |
| "eval_runtime": 51.6079, | |
| "eval_samples_per_second": 13.215, | |
| "eval_steps_per_second": 3.313, | |
| "eval_wer": 0.6194543951501791, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.2658227848101267, | |
| "eval_loss": 0.6331446766853333, | |
| "eval_runtime": 51.2866, | |
| "eval_samples_per_second": 13.298, | |
| "eval_steps_per_second": 3.334, | |
| "eval_wer": 0.6742904381372279, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.3361462728551337, | |
| "eval_loss": 0.7776533961296082, | |
| "eval_runtime": 51.4336, | |
| "eval_samples_per_second": 13.26, | |
| "eval_steps_per_second": 3.325, | |
| "eval_wer": 0.8179939377238908, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.4064697609001406, | |
| "grad_norm": 1.66465163230896, | |
| "learning_rate": 2.8932637571157497e-05, | |
| "loss": 1.0769, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.4064697609001406, | |
| "eval_loss": 0.5906365513801575, | |
| "eval_runtime": 51.362, | |
| "eval_samples_per_second": 13.278, | |
| "eval_steps_per_second": 3.329, | |
| "eval_wer": 0.6095343069716175, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.4767932489451476, | |
| "eval_loss": 0.6689415574073792, | |
| "eval_runtime": 51.5463, | |
| "eval_samples_per_second": 13.231, | |
| "eval_steps_per_second": 3.317, | |
| "eval_wer": 0.7241664370349958, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.5471167369901546, | |
| "eval_loss": 0.6466124653816223, | |
| "eval_runtime": 51.5035, | |
| "eval_samples_per_second": 13.242, | |
| "eval_steps_per_second": 3.32, | |
| "eval_wer": 0.6489391016809039, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.6174402250351618, | |
| "eval_loss": 0.5825116038322449, | |
| "eval_runtime": 51.1787, | |
| "eval_samples_per_second": 13.326, | |
| "eval_steps_per_second": 3.341, | |
| "eval_wer": 0.6113254340038579, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.6877637130801688, | |
| "eval_loss": 0.7114273905754089, | |
| "eval_runtime": 50.9372, | |
| "eval_samples_per_second": 13.389, | |
| "eval_steps_per_second": 3.357, | |
| "eval_wer": 0.7517222375310003, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.7580872011251758, | |
| "grad_norm": 2.698178768157959, | |
| "learning_rate": 2.857685009487666e-05, | |
| "loss": 1.0964, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.7580872011251758, | |
| "eval_loss": 0.8008682131767273, | |
| "eval_runtime": 51.3979, | |
| "eval_samples_per_second": 13.269, | |
| "eval_steps_per_second": 3.327, | |
| "eval_wer": 0.7965004133370074, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.8284106891701828, | |
| "eval_loss": 0.6260038614273071, | |
| "eval_runtime": 51.2623, | |
| "eval_samples_per_second": 13.304, | |
| "eval_steps_per_second": 3.336, | |
| "eval_wer": 0.6364012124552219, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.8987341772151898, | |
| "eval_loss": 0.5530011653900146, | |
| "eval_runtime": 51.1589, | |
| "eval_samples_per_second": 13.331, | |
| "eval_steps_per_second": 3.343, | |
| "eval_wer": 0.5519426839349683, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.969057665260197, | |
| "eval_loss": 0.6974045038223267, | |
| "eval_runtime": 51.0833, | |
| "eval_samples_per_second": 13.351, | |
| "eval_steps_per_second": 3.347, | |
| "eval_wer": 0.7149352438688343, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.0393811533052038, | |
| "eval_loss": 0.5646800994873047, | |
| "eval_runtime": 51.3226, | |
| "eval_samples_per_second": 13.289, | |
| "eval_steps_per_second": 3.332, | |
| "eval_wer": 0.5954808487186553, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.109704641350211, | |
| "grad_norm": 1.9064384698867798, | |
| "learning_rate": 2.8221062618595825e-05, | |
| "loss": 1.0268, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.109704641350211, | |
| "eval_loss": 0.5556230545043945, | |
| "eval_runtime": 51.0134, | |
| "eval_samples_per_second": 13.369, | |
| "eval_steps_per_second": 3.352, | |
| "eval_wer": 0.5487737668779278, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.180028129395218, | |
| "eval_loss": 0.6443986892700195, | |
| "eval_runtime": 50.9797, | |
| "eval_samples_per_second": 13.378, | |
| "eval_steps_per_second": 3.354, | |
| "eval_wer": 0.650179112703224, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.180028129395218, | |
| "step": 3100, | |
| "total_flos": 1.8402257527802782e+19, | |
| "train_loss": 1.9087347387498426, | |
| "train_runtime": 5842.2823, | |
| "train_samples_per_second": 58.416, | |
| "train_steps_per_second": 7.302 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 42660, | |
| "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": 1.8402257527802782e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |