| { | |
| "best_metric": 0.1726989597082138, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-nyagen-combined-model/checkpoint-2000", | |
| "epoch": 3.1377899045020463, | |
| "eval_steps": 100, | |
| "global_step": 2300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1364256480218281, | |
| "grad_norm": 2.10461163520813, | |
| "learning_rate": 0.00029099999999999997, | |
| "loss": 6.9978, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1364256480218281, | |
| "eval_loss": 0.6384420394897461, | |
| "eval_runtime": 30.8578, | |
| "eval_samples_per_second": 11.148, | |
| "eval_steps_per_second": 2.787, | |
| "eval_wer": 0.5015432098765432, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2728512960436562, | |
| "grad_norm": 0.9924084544181824, | |
| "learning_rate": 0.0002986843307446322, | |
| "loss": 0.482, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2728512960436562, | |
| "eval_loss": 0.27768996357917786, | |
| "eval_runtime": 30.8437, | |
| "eval_samples_per_second": 11.153, | |
| "eval_steps_per_second": 2.788, | |
| "eval_wer": 0.3712522045855379, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4092769440654843, | |
| "grad_norm": 1.641776442527771, | |
| "learning_rate": 0.0002973275468250342, | |
| "loss": 0.3907, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4092769440654843, | |
| "eval_loss": 0.24835941195487976, | |
| "eval_runtime": 30.9832, | |
| "eval_samples_per_second": 11.103, | |
| "eval_steps_per_second": 2.776, | |
| "eval_wer": 0.3481040564373898, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5457025920873124, | |
| "grad_norm": 1.1754199266433716, | |
| "learning_rate": 0.00029597076290543625, | |
| "loss": 0.3782, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5457025920873124, | |
| "eval_loss": 0.22902728617191315, | |
| "eval_runtime": 30.7597, | |
| "eval_samples_per_second": 11.183, | |
| "eval_steps_per_second": 2.796, | |
| "eval_wer": 0.32319223985890655, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6821282401091405, | |
| "grad_norm": 1.491380214691162, | |
| "learning_rate": 0.00029460027409776153, | |
| "loss": 0.3316, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6821282401091405, | |
| "eval_loss": 0.222237691283226, | |
| "eval_runtime": 31.0639, | |
| "eval_samples_per_second": 11.074, | |
| "eval_steps_per_second": 2.768, | |
| "eval_wer": 0.3148148148148148, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8185538881309686, | |
| "grad_norm": 0.943580150604248, | |
| "learning_rate": 0.00029322978529008675, | |
| "loss": 0.3158, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8185538881309686, | |
| "eval_loss": 0.2126626819372177, | |
| "eval_runtime": 31.0228, | |
| "eval_samples_per_second": 11.089, | |
| "eval_steps_per_second": 2.772, | |
| "eval_wer": 0.30423280423280424, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9549795361527967, | |
| "grad_norm": 1.18917715549469, | |
| "learning_rate": 0.000291859296482412, | |
| "loss": 0.3199, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9549795361527967, | |
| "eval_loss": 0.2105857878923416, | |
| "eval_runtime": 30.6684, | |
| "eval_samples_per_second": 11.217, | |
| "eval_steps_per_second": 2.804, | |
| "eval_wer": 0.2932098765432099, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0914051841746248, | |
| "grad_norm": 1.4716788530349731, | |
| "learning_rate": 0.0002904888076747373, | |
| "loss": 0.3223, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.0914051841746248, | |
| "eval_loss": 0.20132741332054138, | |
| "eval_runtime": 30.7244, | |
| "eval_samples_per_second": 11.196, | |
| "eval_steps_per_second": 2.799, | |
| "eval_wer": 0.2826278659611993, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2278308321964528, | |
| "grad_norm": 0.600020706653595, | |
| "learning_rate": 0.0002891183188670626, | |
| "loss": 0.3075, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.2278308321964528, | |
| "eval_loss": 0.19748319685459137, | |
| "eval_runtime": 31.1335, | |
| "eval_samples_per_second": 11.049, | |
| "eval_steps_per_second": 2.762, | |
| "eval_wer": 0.2709435626102293, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.364256480218281, | |
| "grad_norm": 0.36852729320526123, | |
| "learning_rate": 0.00028774783005938785, | |
| "loss": 0.3015, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.364256480218281, | |
| "eval_loss": 0.1942104995250702, | |
| "eval_runtime": 31.1423, | |
| "eval_samples_per_second": 11.046, | |
| "eval_steps_per_second": 2.762, | |
| "eval_wer": 0.2762345679012346, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.500682128240109, | |
| "grad_norm": 0.969207227230072, | |
| "learning_rate": 0.00028639104613978984, | |
| "loss": 0.3049, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.500682128240109, | |
| "eval_loss": 0.18946479260921478, | |
| "eval_runtime": 30.9295, | |
| "eval_samples_per_second": 11.122, | |
| "eval_steps_per_second": 2.781, | |
| "eval_wer": 0.27292768959435626, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6371077762619373, | |
| "grad_norm": 0.7448552250862122, | |
| "learning_rate": 0.00028502055733211506, | |
| "loss": 0.3029, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.6371077762619373, | |
| "eval_loss": 0.1888139843940735, | |
| "eval_runtime": 31.0815, | |
| "eval_samples_per_second": 11.068, | |
| "eval_steps_per_second": 2.767, | |
| "eval_wer": 0.2718253968253968, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.7735334242837655, | |
| "grad_norm": 0.7603126168251038, | |
| "learning_rate": 0.0002836500685244404, | |
| "loss": 0.2626, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.7735334242837655, | |
| "eval_loss": 0.1865960955619812, | |
| "eval_runtime": 31.1543, | |
| "eval_samples_per_second": 11.042, | |
| "eval_steps_per_second": 2.76, | |
| "eval_wer": 0.2682980599647266, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.9099590723055935, | |
| "grad_norm": 0.49969810247421265, | |
| "learning_rate": 0.0002822795797167656, | |
| "loss": 0.2803, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.9099590723055935, | |
| "eval_loss": 0.18303100764751434, | |
| "eval_runtime": 31.2418, | |
| "eval_samples_per_second": 11.011, | |
| "eval_steps_per_second": 2.753, | |
| "eval_wer": 0.26146384479717816, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.0463847203274215, | |
| "grad_norm": 1.5848807096481323, | |
| "learning_rate": 0.0002809090909090909, | |
| "loss": 0.2725, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.0463847203274215, | |
| "eval_loss": 0.1813870072364807, | |
| "eval_runtime": 31.0194, | |
| "eval_samples_per_second": 11.09, | |
| "eval_steps_per_second": 2.772, | |
| "eval_wer": 0.26256613756613756, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.1828103683492497, | |
| "grad_norm": 0.9148170948028564, | |
| "learning_rate": 0.00027953860210141616, | |
| "loss": 0.2732, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.1828103683492497, | |
| "eval_loss": 0.1783067137002945, | |
| "eval_runtime": 31.1451, | |
| "eval_samples_per_second": 11.045, | |
| "eval_steps_per_second": 2.761, | |
| "eval_wer": 0.2641093474426808, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.319236016371078, | |
| "grad_norm": 0.7404142618179321, | |
| "learning_rate": 0.00027816811329374144, | |
| "loss": 0.249, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.319236016371078, | |
| "eval_loss": 0.1828220933675766, | |
| "eval_runtime": 31.1781, | |
| "eval_samples_per_second": 11.033, | |
| "eval_steps_per_second": 2.758, | |
| "eval_wer": 0.25595238095238093, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.4556616643929057, | |
| "grad_norm": 0.795720100402832, | |
| "learning_rate": 0.00027679762448606666, | |
| "loss": 0.2423, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.4556616643929057, | |
| "eval_loss": 0.17622150480747223, | |
| "eval_runtime": 31.3032, | |
| "eval_samples_per_second": 10.989, | |
| "eval_steps_per_second": 2.747, | |
| "eval_wer": 0.24801587301587302, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.592087312414734, | |
| "grad_norm": 2.1790249347686768, | |
| "learning_rate": 0.00027542713567839193, | |
| "loss": 0.2668, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.592087312414734, | |
| "eval_loss": 0.1731557846069336, | |
| "eval_runtime": 31.2391, | |
| "eval_samples_per_second": 11.012, | |
| "eval_steps_per_second": 2.753, | |
| "eval_wer": 0.24581128747795414, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.728512960436562, | |
| "grad_norm": 1.353415608406067, | |
| "learning_rate": 0.0002740566468707172, | |
| "loss": 0.2653, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.728512960436562, | |
| "eval_loss": 0.1726989597082138, | |
| "eval_runtime": 30.843, | |
| "eval_samples_per_second": 11.153, | |
| "eval_steps_per_second": 2.788, | |
| "eval_wer": 0.24603174603174602, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.8649386084583903, | |
| "grad_norm": 0.9100846648216248, | |
| "learning_rate": 0.00027268615806304243, | |
| "loss": 0.2614, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.8649386084583903, | |
| "eval_loss": 0.1748967319726944, | |
| "eval_runtime": 31.2508, | |
| "eval_samples_per_second": 11.008, | |
| "eval_steps_per_second": 2.752, | |
| "eval_wer": 0.2533068783068783, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.001364256480218, | |
| "grad_norm": 0.705278217792511, | |
| "learning_rate": 0.0002713156692553677, | |
| "loss": 0.2474, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.001364256480218, | |
| "eval_loss": 0.1732555478811264, | |
| "eval_runtime": 31.2961, | |
| "eval_samples_per_second": 10.992, | |
| "eval_steps_per_second": 2.748, | |
| "eval_wer": 0.24382716049382716, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.1377899045020463, | |
| "grad_norm": 1.5550167560577393, | |
| "learning_rate": 0.000269945180447693, | |
| "loss": 0.2317, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.1377899045020463, | |
| "eval_loss": 0.17674298584461212, | |
| "eval_runtime": 31.0854, | |
| "eval_samples_per_second": 11.066, | |
| "eval_steps_per_second": 2.767, | |
| "eval_wer": 0.2447089947089947, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.1377899045020463, | |
| "step": 2300, | |
| "total_flos": 8.649929823066914e+18, | |
| "train_loss": 0.5916279270337975, | |
| "train_runtime": 2613.7833, | |
| "train_samples_per_second": 33.618, | |
| "train_steps_per_second": 8.413 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 21990, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.649929823066914e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |