| { | |
| "best_metric": Infinity, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.8811136192626035, | |
| "eval_steps": 100, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07524454477050414, | |
| "grad_norm": 8.513701438903809, | |
| "learning_rate": 0.000285, | |
| "loss": 13.9528, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07524454477050414, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.8079, | |
| "eval_samples_per_second": 15.418, | |
| "eval_steps_per_second": 3.863, | |
| "eval_wer": 1.043132803632236, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1504890895410083, | |
| "grad_norm": 6.828241348266602, | |
| "learning_rate": 0.00028812499999999997, | |
| "loss": 6.1846, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1504890895410083, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.7299, | |
| "eval_samples_per_second": 15.457, | |
| "eval_steps_per_second": 3.872, | |
| "eval_wer": 1.0029511918274687, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.22573363431151242, | |
| "grad_norm": 7.705227851867676, | |
| "learning_rate": 0.00027562499999999994, | |
| "loss": 5.4651, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.22573363431151242, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.7705, | |
| "eval_samples_per_second": 15.437, | |
| "eval_steps_per_second": 3.867, | |
| "eval_wer": 1.0385925085130534, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3009781790820166, | |
| "grad_norm": 6.132575988769531, | |
| "learning_rate": 0.00026312499999999996, | |
| "loss": 4.4356, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3009781790820166, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.8056, | |
| "eval_samples_per_second": 15.419, | |
| "eval_steps_per_second": 3.863, | |
| "eval_wer": 0.8830874006810443, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3762227238525207, | |
| "grad_norm": 3.9721577167510986, | |
| "learning_rate": 0.000250625, | |
| "loss": 2.2016, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.3762227238525207, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.8903, | |
| "eval_samples_per_second": 15.377, | |
| "eval_steps_per_second": 3.852, | |
| "eval_wer": 0.6217934165720772, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.45146726862302483, | |
| "grad_norm": 4.83721923828125, | |
| "learning_rate": 0.00023812499999999997, | |
| "loss": 1.8013, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.45146726862302483, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.8946, | |
| "eval_samples_per_second": 15.375, | |
| "eval_steps_per_second": 3.852, | |
| "eval_wer": 0.5745743473325766, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.526711813393529, | |
| "grad_norm": 5.4762725830078125, | |
| "learning_rate": 0.00022562499999999997, | |
| "loss": 1.7499, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.526711813393529, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.6173, | |
| "eval_samples_per_second": 15.514, | |
| "eval_steps_per_second": 3.887, | |
| "eval_wer": 0.5793416572077185, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6019563581640331, | |
| "grad_norm": 5.545216083526611, | |
| "learning_rate": 0.000213125, | |
| "loss": 1.6979, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6019563581640331, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.6746, | |
| "eval_samples_per_second": 15.485, | |
| "eval_steps_per_second": 3.879, | |
| "eval_wer": 0.5500567536889898, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6772009029345373, | |
| "grad_norm": 6.604818344116211, | |
| "learning_rate": 0.00020062499999999996, | |
| "loss": 1.5567, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6772009029345373, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 31.0923, | |
| "eval_samples_per_second": 15.277, | |
| "eval_steps_per_second": 3.827, | |
| "eval_wer": 0.5439273552780931, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7524454477050414, | |
| "grad_norm": 5.870357036590576, | |
| "learning_rate": 0.00018824999999999997, | |
| "loss": 1.6301, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7524454477050414, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.7668, | |
| "eval_samples_per_second": 15.439, | |
| "eval_steps_per_second": 3.868, | |
| "eval_wer": 0.535527809307605, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8276899924755455, | |
| "grad_norm": 20.379985809326172, | |
| "learning_rate": 0.00017575, | |
| "loss": 1.6362, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.8276899924755455, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.7224, | |
| "eval_samples_per_second": 15.461, | |
| "eval_steps_per_second": 3.873, | |
| "eval_wer": 0.5366628830874007, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.9029345372460497, | |
| "grad_norm": 7.096904277801514, | |
| "learning_rate": 0.00016324999999999998, | |
| "loss": 1.5247, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9029345372460497, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.6145, | |
| "eval_samples_per_second": 15.516, | |
| "eval_steps_per_second": 3.887, | |
| "eval_wer": 0.5325766174801362, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9781790820165538, | |
| "grad_norm": 6.06346321105957, | |
| "learning_rate": 0.00015074999999999998, | |
| "loss": 1.4012, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.9781790820165538, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.9705, | |
| "eval_samples_per_second": 15.337, | |
| "eval_steps_per_second": 3.842, | |
| "eval_wer": 0.5346197502837684, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.053423626787058, | |
| "grad_norm": 5.044532775878906, | |
| "learning_rate": 0.00013824999999999997, | |
| "loss": 1.6397, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.053423626787058, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.897, | |
| "eval_samples_per_second": 15.374, | |
| "eval_steps_per_second": 3.852, | |
| "eval_wer": 0.5300794551645857, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.1286681715575622, | |
| "grad_norm": 4.185507297515869, | |
| "learning_rate": 0.00012575, | |
| "loss": 1.5258, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.1286681715575622, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.6085, | |
| "eval_samples_per_second": 15.519, | |
| "eval_steps_per_second": 3.888, | |
| "eval_wer": 0.5284903518728717, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.2039127163280663, | |
| "grad_norm": 2.0165092945098877, | |
| "learning_rate": 0.00011324999999999999, | |
| "loss": 1.4144, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.2039127163280663, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.8112, | |
| "eval_samples_per_second": 15.416, | |
| "eval_steps_per_second": 3.862, | |
| "eval_wer": 0.5244040862656073, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.2791572610985704, | |
| "grad_norm": 6.2198486328125, | |
| "learning_rate": 0.00010074999999999998, | |
| "loss": 1.4363, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.2791572610985704, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 31.0265, | |
| "eval_samples_per_second": 15.31, | |
| "eval_steps_per_second": 3.835, | |
| "eval_wer": 0.5144154370034052, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.3544018058690745, | |
| "grad_norm": 6.625217437744141, | |
| "learning_rate": 8.837499999999998e-05, | |
| "loss": 1.3733, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.3544018058690745, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 31.0253, | |
| "eval_samples_per_second": 15.31, | |
| "eval_steps_per_second": 3.836, | |
| "eval_wer": 0.5357548240635641, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.4296463506395787, | |
| "grad_norm": 3.523829460144043, | |
| "learning_rate": 7.5875e-05, | |
| "loss": 1.4592, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.4296463506395787, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.9124, | |
| "eval_samples_per_second": 15.366, | |
| "eval_steps_per_second": 3.85, | |
| "eval_wer": 0.5598183881952327, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.5048908954100828, | |
| "grad_norm": 8.395014762878418, | |
| "learning_rate": 6.3375e-05, | |
| "loss": 1.3499, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.5048908954100828, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.7712, | |
| "eval_samples_per_second": 15.437, | |
| "eval_steps_per_second": 3.867, | |
| "eval_wer": 0.5191827468785472, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.580135440180587, | |
| "grad_norm": 2.0930612087249756, | |
| "learning_rate": 5.0874999999999997e-05, | |
| "loss": 1.4039, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.580135440180587, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 31.265, | |
| "eval_samples_per_second": 15.193, | |
| "eval_steps_per_second": 3.806, | |
| "eval_wer": 0.5228149829738933, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.655379984951091, | |
| "grad_norm": 1.5763949155807495, | |
| "learning_rate": 3.8375e-05, | |
| "loss": 1.4057, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.655379984951091, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.8864, | |
| "eval_samples_per_second": 15.379, | |
| "eval_steps_per_second": 3.853, | |
| "eval_wer": 0.52894438138479, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.7306245297215952, | |
| "grad_norm": 2.9080910682678223, | |
| "learning_rate": 2.5874999999999995e-05, | |
| "loss": 1.4961, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.7306245297215952, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 31.0242, | |
| "eval_samples_per_second": 15.311, | |
| "eval_steps_per_second": 3.836, | |
| "eval_wer": 0.5323496027241771, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.8058690744920993, | |
| "grad_norm": 3.506730079650879, | |
| "learning_rate": 1.3375e-05, | |
| "loss": 1.3975, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.8058690744920993, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 30.8502, | |
| "eval_samples_per_second": 15.397, | |
| "eval_steps_per_second": 3.857, | |
| "eval_wer": 0.5119182746878547, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.8811136192626035, | |
| "grad_norm": 12.351927757263184, | |
| "learning_rate": 8.75e-07, | |
| "loss": 1.4725, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.8811136192626035, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 31.0235, | |
| "eval_samples_per_second": 15.311, | |
| "eval_steps_per_second": 3.836, | |
| "eval_wer": 0.5259931895573212, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.8811136192626035, | |
| "step": 2500, | |
| "total_flos": 1.2622443564523827e+19, | |
| "train_loss": 2.5044897521972658, | |
| "train_runtime": 3534.0194, | |
| "train_samples_per_second": 5.659, | |
| "train_steps_per_second": 0.707 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 2500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.2622443564523827e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |