| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9619084263178146, | |
| "eval_steps": 500, | |
| "global_step": 15000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03206361421059382, | |
| "grad_norm": 9.123078346252441, | |
| "learning_rate": 0.0002465, | |
| "loss": 4.6156, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.03206361421059382, | |
| "eval_loss": 1.5867419242858887, | |
| "eval_runtime": 188.7881, | |
| "eval_samples_per_second": 37.195, | |
| "eval_steps_per_second": 0.583, | |
| "eval_wer": 0.9176576887814082, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06412722842118763, | |
| "grad_norm": 6.654547691345215, | |
| "learning_rate": 0.00029181249999999997, | |
| "loss": 1.0315, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.06412722842118763, | |
| "eval_loss": 1.1748294830322266, | |
| "eval_runtime": 190.3404, | |
| "eval_samples_per_second": 36.892, | |
| "eval_steps_per_second": 0.578, | |
| "eval_wer": 0.7888358867377988, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.09619084263178146, | |
| "grad_norm": 6.171149253845215, | |
| "learning_rate": 0.0002813958333333333, | |
| "loss": 0.834, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.09619084263178146, | |
| "eval_loss": 1.0392996072769165, | |
| "eval_runtime": 189.2832, | |
| "eval_samples_per_second": 37.098, | |
| "eval_steps_per_second": 0.581, | |
| "eval_wer": 0.7219867391275462, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.12825445684237527, | |
| "grad_norm": 6.896900177001953, | |
| "learning_rate": 0.00027097916666666666, | |
| "loss": 0.7184, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.12825445684237527, | |
| "eval_loss": 0.9616143703460693, | |
| "eval_runtime": 190.9944, | |
| "eval_samples_per_second": 36.765, | |
| "eval_steps_per_second": 0.576, | |
| "eval_wer": 0.663747857399115, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.16031807105296908, | |
| "grad_norm": 9.408955574035645, | |
| "learning_rate": 0.0002605625, | |
| "loss": 0.6655, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.16031807105296908, | |
| "eval_loss": 0.9033711552619934, | |
| "eval_runtime": 190.9851, | |
| "eval_samples_per_second": 36.767, | |
| "eval_steps_per_second": 0.576, | |
| "eval_wer": 0.6331335787081944, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.19238168526356292, | |
| "grad_norm": 6.4334211349487305, | |
| "learning_rate": 0.0002501458333333333, | |
| "loss": 0.6193, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.19238168526356292, | |
| "eval_loss": 0.8614802956581116, | |
| "eval_runtime": 191.2463, | |
| "eval_samples_per_second": 36.717, | |
| "eval_steps_per_second": 0.575, | |
| "eval_wer": 0.6238988028009939, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.22444529947415673, | |
| "grad_norm": 3.711681365966797, | |
| "learning_rate": 0.00023972916666666665, | |
| "loss": 0.5952, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.22444529947415673, | |
| "eval_loss": 0.8161324858665466, | |
| "eval_runtime": 191.2031, | |
| "eval_samples_per_second": 36.725, | |
| "eval_steps_per_second": 0.575, | |
| "eval_wer": 0.5866275129884798, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.25650891368475054, | |
| "grad_norm": 7.527787208557129, | |
| "learning_rate": 0.00022933333333333332, | |
| "loss": 0.5622, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.25650891368475054, | |
| "eval_loss": 0.811023473739624, | |
| "eval_runtime": 190.6985, | |
| "eval_samples_per_second": 36.823, | |
| "eval_steps_per_second": 0.577, | |
| "eval_wer": 0.5850728816487065, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.2885725278953444, | |
| "grad_norm": 11.801218032836914, | |
| "learning_rate": 0.0002189583333333333, | |
| "loss": 0.5341, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.2885725278953444, | |
| "eval_loss": 0.757978618144989, | |
| "eval_runtime": 192.268, | |
| "eval_samples_per_second": 36.522, | |
| "eval_steps_per_second": 0.572, | |
| "eval_wer": 0.5546579146680132, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.32063614210593816, | |
| "grad_norm": 9.381750106811523, | |
| "learning_rate": 0.00020854166666666664, | |
| "loss": 0.522, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.32063614210593816, | |
| "eval_loss": 0.7397128343582153, | |
| "eval_runtime": 191.0373, | |
| "eval_samples_per_second": 36.757, | |
| "eval_steps_per_second": 0.576, | |
| "eval_wer": 0.5411711556092959, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.352699756316532, | |
| "grad_norm": 6.341240882873535, | |
| "learning_rate": 0.00019812499999999998, | |
| "loss": 0.5123, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.352699756316532, | |
| "eval_loss": 0.7228623628616333, | |
| "eval_runtime": 191.6536, | |
| "eval_samples_per_second": 36.639, | |
| "eval_steps_per_second": 0.574, | |
| "eval_wer": 0.531737067991868, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "grad_norm": 6.53903341293335, | |
| "learning_rate": 0.00018772916666666666, | |
| "loss": 0.4884, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "eval_loss": 0.72346431016922, | |
| "eval_runtime": 191.4082, | |
| "eval_samples_per_second": 36.686, | |
| "eval_steps_per_second": 0.575, | |
| "eval_wer": 0.5164830784358017, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.4168269847377196, | |
| "grad_norm": 10.402660369873047, | |
| "learning_rate": 0.00017731249999999998, | |
| "loss": 0.4658, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.4168269847377196, | |
| "eval_loss": 0.681357204914093, | |
| "eval_runtime": 191.0697, | |
| "eval_samples_per_second": 36.751, | |
| "eval_steps_per_second": 0.576, | |
| "eval_wer": 0.5116995973903453, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.44889059894831346, | |
| "grad_norm": 11.663326263427734, | |
| "learning_rate": 0.00016691666666666667, | |
| "loss": 0.4471, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.44889059894831346, | |
| "eval_loss": 0.662290632724762, | |
| "eval_runtime": 191.4867, | |
| "eval_samples_per_second": 36.671, | |
| "eval_steps_per_second": 0.574, | |
| "eval_wer": 0.4890577871085186, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.4809542131589073, | |
| "grad_norm": 7.363061428070068, | |
| "learning_rate": 0.00015649999999999998, | |
| "loss": 0.4338, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.4809542131589073, | |
| "eval_loss": 0.6449915170669556, | |
| "eval_runtime": 190.9868, | |
| "eval_samples_per_second": 36.767, | |
| "eval_steps_per_second": 0.576, | |
| "eval_wer": 0.4913830903945043, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.5130178273695011, | |
| "grad_norm": 14.478469848632812, | |
| "learning_rate": 0.00014610416666666667, | |
| "loss": 0.4267, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.5130178273695011, | |
| "eval_loss": 0.6256160736083984, | |
| "eval_runtime": 190.8261, | |
| "eval_samples_per_second": 36.798, | |
| "eval_steps_per_second": 0.576, | |
| "eval_wer": 0.4685419684024502, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.5450814415800949, | |
| "grad_norm": 10.456161499023438, | |
| "learning_rate": 0.00013568749999999998, | |
| "loss": 0.4283, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.5450814415800949, | |
| "eval_loss": 0.6342806816101074, | |
| "eval_runtime": 190.609, | |
| "eval_samples_per_second": 36.84, | |
| "eval_steps_per_second": 0.577, | |
| "eval_wer": 0.4710665833986633, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.5771450557906888, | |
| "grad_norm": 9.847672462463379, | |
| "learning_rate": 0.00012527083333333333, | |
| "loss": 0.4131, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.5771450557906888, | |
| "eval_loss": 0.5988845229148865, | |
| "eval_runtime": 189.2404, | |
| "eval_samples_per_second": 37.106, | |
| "eval_steps_per_second": 0.581, | |
| "eval_wer": 0.4486506597217608, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.6092086700012825, | |
| "grad_norm": 7.610143661499023, | |
| "learning_rate": 0.00011485416666666666, | |
| "loss": 0.4317, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.6092086700012825, | |
| "eval_loss": 0.7167520523071289, | |
| "eval_runtime": 189.8256, | |
| "eval_samples_per_second": 36.992, | |
| "eval_steps_per_second": 0.579, | |
| "eval_wer": 0.4919677380778379, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.6412722842118763, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.00010691666666666665, | |
| "loss": 0.5904, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.6412722842118763, | |
| "eval_loss": NaN, | |
| "eval_runtime": 190.1563, | |
| "eval_samples_per_second": 36.928, | |
| "eval_steps_per_second": 0.578, | |
| "eval_wer": 0.7309956284298224, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.6733358984224702, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.000106875, | |
| "loss": 0.0513, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.6733358984224702, | |
| "eval_loss": NaN, | |
| "eval_runtime": 185.5416, | |
| "eval_samples_per_second": 37.846, | |
| "eval_steps_per_second": 0.593, | |
| "eval_wer": 1.0, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.705399512633064, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.000106875, | |
| "loss": 0.0, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.705399512633064, | |
| "eval_loss": NaN, | |
| "eval_runtime": 185.2695, | |
| "eval_samples_per_second": 37.902, | |
| "eval_steps_per_second": 0.594, | |
| "eval_wer": 1.0, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.7374631268436578, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.000106875, | |
| "loss": 0.0, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.7374631268436578, | |
| "eval_loss": NaN, | |
| "eval_runtime": 185.1794, | |
| "eval_samples_per_second": 37.92, | |
| "eval_steps_per_second": 0.594, | |
| "eval_wer": 1.0, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.000106875, | |
| "loss": 0.0, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "eval_loss": NaN, | |
| "eval_runtime": 184.8489, | |
| "eval_samples_per_second": 37.988, | |
| "eval_steps_per_second": 0.595, | |
| "eval_wer": 1.0, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.8015903552648455, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.000106875, | |
| "loss": 0.0, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.8015903552648455, | |
| "eval_loss": NaN, | |
| "eval_runtime": 184.8249, | |
| "eval_samples_per_second": 37.993, | |
| "eval_steps_per_second": 0.595, | |
| "eval_wer": 1.0, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.8336539694754392, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.000106875, | |
| "loss": 0.0, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.8336539694754392, | |
| "eval_loss": NaN, | |
| "eval_runtime": 185.2964, | |
| "eval_samples_per_second": 37.896, | |
| "eval_steps_per_second": 0.594, | |
| "eval_wer": 1.0, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.8657175836860331, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.000106875, | |
| "loss": 0.0, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.8657175836860331, | |
| "eval_loss": NaN, | |
| "eval_runtime": 184.7613, | |
| "eval_samples_per_second": 38.006, | |
| "eval_steps_per_second": 0.595, | |
| "eval_wer": 1.0, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.8977811978966269, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.000106875, | |
| "loss": 0.0, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.8977811978966269, | |
| "eval_loss": NaN, | |
| "eval_runtime": 184.7837, | |
| "eval_samples_per_second": 38.001, | |
| "eval_steps_per_second": 0.595, | |
| "eval_wer": 1.0, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.9298448121072207, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.000106875, | |
| "loss": 0.0, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.9298448121072207, | |
| "eval_loss": NaN, | |
| "eval_runtime": 184.6054, | |
| "eval_samples_per_second": 38.038, | |
| "eval_steps_per_second": 0.596, | |
| "eval_wer": 1.0, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.9619084263178146, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.000106875, | |
| "loss": 0.0, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.9619084263178146, | |
| "eval_loss": NaN, | |
| "eval_runtime": 184.8182, | |
| "eval_samples_per_second": 37.994, | |
| "eval_steps_per_second": 0.595, | |
| "eval_wer": 1.0, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.9619084263178146, | |
| "step": 15000, | |
| "total_flos": 1.7109669148845115e+19, | |
| "train_loss": 0.5128920831044514, | |
| "train_runtime": 11433.8652, | |
| "train_samples_per_second": 10.495, | |
| "train_steps_per_second": 1.312 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 15000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.7109669148845115e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |