{ "best_metric": 0.5928366184234619, "best_model_checkpoint": "/scratch/skscla001/results/mms-zeroshot-300m-natbed-combined-model/checkpoint-2400", "epoch": 3.7546933667083855, "eval_steps": 200, "global_step": 3000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.2503128911138924, "eval_loss": 2.66123104095459, "eval_runtime": 85.5481, "eval_samples_per_second": 15.874, "eval_steps_per_second": 1.987, "eval_wer": 1.0, "step": 200 }, { "epoch": 0.5006257822277848, "eval_loss": 0.7787109017372131, "eval_runtime": 85.3612, "eval_samples_per_second": 15.909, "eval_steps_per_second": 1.992, "eval_wer": 0.6600606010973712, "step": 400 }, { "epoch": 0.6257822277847309, "grad_norm": 0.983116626739502, "learning_rate": 0.0002950104733975701, "loss": 3.3034, "step": 500 }, { "epoch": 0.7509386733416771, "eval_loss": 0.7195701599121094, "eval_runtime": 85.1972, "eval_samples_per_second": 15.939, "eval_steps_per_second": 1.995, "eval_wer": 0.6193595938088609, "step": 600 }, { "epoch": 1.0012515644555695, "eval_loss": 0.6961039900779724, "eval_runtime": 85.8666, "eval_samples_per_second": 15.815, "eval_steps_per_second": 1.98, "eval_wer": 0.5965932356072394, "step": 800 }, { "epoch": 1.2515644555694618, "grad_norm": 3.753706693649292, "learning_rate": 0.0002887264348554671, "loss": 0.8261, "step": 1000 }, { "epoch": 1.2515644555694618, "eval_loss": 0.6694886684417725, "eval_runtime": 86.9509, "eval_samples_per_second": 15.618, "eval_steps_per_second": 1.955, "eval_wer": 0.5762017852755712, "step": 1000 }, { "epoch": 1.5018773466833542, "eval_loss": 0.6313649415969849, "eval_runtime": 84.9355, "eval_samples_per_second": 15.989, "eval_steps_per_second": 2.002, "eval_wer": 0.5728441569077062, "step": 1200 }, { "epoch": 1.7521902377972465, "eval_loss": 0.6477869749069214, "eval_runtime": 84.1241, "eval_samples_per_second": 16.143, "eval_steps_per_second": 2.021, "eval_wer": 0.5575300958152486, "step": 1400 }, { "epoch": 1.8773466833541927, "grad_norm": 2.5700743198394775, "learning_rate": 0.000282442396313364, "loss": 0.7513, "step": 1500 }, { "epoch": 2.002503128911139, "eval_loss": 0.6373749375343323, "eval_runtime": 85.3519, "eval_samples_per_second": 15.911, "eval_steps_per_second": 1.992, "eval_wer": 0.5554008680697732, "step": 1600 }, { "epoch": 2.252816020025031, "eval_loss": 0.6032695770263672, "eval_runtime": 84.1367, "eval_samples_per_second": 16.14, "eval_steps_per_second": 2.021, "eval_wer": 0.5483580378347391, "step": 1800 }, { "epoch": 2.5031289111389237, "grad_norm": 0.43018773198127747, "learning_rate": 0.00027615835777126095, "loss": 0.7173, "step": 2000 }, { "epoch": 2.5031289111389237, "eval_loss": 0.6269567608833313, "eval_runtime": 84.4441, "eval_samples_per_second": 16.082, "eval_steps_per_second": 2.013, "eval_wer": 0.541888461223487, "step": 2000 }, { "epoch": 2.7534418022528158, "eval_loss": 0.6057294011116028, "eval_runtime": 84.7682, "eval_samples_per_second": 16.02, "eval_steps_per_second": 2.005, "eval_wer": 0.5432806485955286, "step": 2200 }, { "epoch": 3.0037546933667083, "eval_loss": 0.5928366184234619, "eval_runtime": 84.7776, "eval_samples_per_second": 16.018, "eval_steps_per_second": 2.005, "eval_wer": 0.527802800753419, "step": 2400 }, { "epoch": 3.1289111389236544, "grad_norm": 1.3301851749420166, "learning_rate": 0.00026987431922915794, "loss": 0.7092, "step": 2500 }, { "epoch": 3.254067584480601, "eval_loss": 0.5980406403541565, "eval_runtime": 84.3511, "eval_samples_per_second": 16.099, "eval_steps_per_second": 2.015, "eval_wer": 0.5321431496191958, "step": 2600 }, { "epoch": 3.504380475594493, "eval_loss": 0.5975831747055054, "eval_runtime": 85.6923, "eval_samples_per_second": 15.847, "eval_steps_per_second": 1.984, "eval_wer": 0.5260830398820735, "step": 2800 }, { "epoch": 3.7546933667083855, "grad_norm": 0.4005778431892395, "learning_rate": 0.00026359028068705483, "loss": 0.696, "step": 3000 }, { "epoch": 3.7546933667083855, "eval_loss": 0.6369301676750183, "eval_runtime": 85.3682, "eval_samples_per_second": 15.908, "eval_steps_per_second": 1.991, "eval_wer": 0.5247727458848579, "step": 3000 }, { "epoch": 3.7546933667083855, "step": 3000, "total_flos": 5.095058470999591e+18, "train_loss": 1.167226084391276, "train_runtime": 3484.8413, "train_samples_per_second": 54.984, "train_steps_per_second": 6.878 } ], "logging_steps": 500, "max_steps": 23970, "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": 3 } }, "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 5.095058470999591e+18, "train_batch_size": 8, "trial_name": null, "trial_params": null }