| { | |
| "best_metric": 103.5799522673031, | |
| "best_model_checkpoint": "ASFEMA_SEV/checkpoint-1600", | |
| "epoch": 25.964630225080384, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.6000000000000004e-07, | |
| "loss": 4.0894, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 9.600000000000001e-07, | |
| "loss": 3.7, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 1.46e-06, | |
| "loss": 3.0759, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 1.9600000000000003e-06, | |
| "loss": 2.5729, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 2.46e-06, | |
| "loss": 2.1885, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 2.96e-06, | |
| "loss": 1.9431, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 3.46e-06, | |
| "loss": 1.5753, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 3.96e-06, | |
| "loss": 1.3319, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "learning_rate": 4.4600000000000005e-06, | |
| "loss": 1.2076, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "learning_rate": 4.960000000000001e-06, | |
| "loss": 0.9211, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 5.460000000000001e-06, | |
| "loss": 0.7332, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "learning_rate": 5.9600000000000005e-06, | |
| "loss": 0.6601, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "learning_rate": 6.460000000000001e-06, | |
| "loss": 0.4824, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "learning_rate": 6.96e-06, | |
| "loss": 0.3455, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "learning_rate": 7.4600000000000006e-06, | |
| "loss": 0.3741, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 5.19, | |
| "learning_rate": 7.960000000000002e-06, | |
| "loss": 0.2448, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 5.19, | |
| "eval_loss": 0.9975764751434326, | |
| "eval_runtime": 24.147, | |
| "eval_samples_per_second": 5.756, | |
| "eval_steps_per_second": 0.745, | |
| "eval_wer": 132.57756563245823, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 5.51, | |
| "learning_rate": 8.46e-06, | |
| "loss": 0.1713, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "learning_rate": 8.96e-06, | |
| "loss": 0.1906, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 6.17, | |
| "learning_rate": 9.460000000000001e-06, | |
| "loss": 0.1479, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 6.49, | |
| "learning_rate": 9.960000000000001e-06, | |
| "loss": 0.103, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 6.81, | |
| "learning_rate": 9.975789473684211e-06, | |
| "loss": 0.1271, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 7.14, | |
| "learning_rate": 9.949473684210526e-06, | |
| "loss": 0.1033, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 7.46, | |
| "learning_rate": 9.923157894736844e-06, | |
| "loss": 0.0799, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 7.78, | |
| "learning_rate": 9.89684210526316e-06, | |
| "loss": 0.0814, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 8.12, | |
| "learning_rate": 9.870526315789474e-06, | |
| "loss": 0.069, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 8.44, | |
| "learning_rate": 9.84421052631579e-06, | |
| "loss": 0.0546, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 8.76, | |
| "learning_rate": 9.817894736842106e-06, | |
| "loss": 0.0627, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "learning_rate": 9.791578947368422e-06, | |
| "loss": 0.0459, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 9.41, | |
| "learning_rate": 9.765263157894737e-06, | |
| "loss": 0.034, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 9.73, | |
| "learning_rate": 9.738947368421054e-06, | |
| "loss": 0.036, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 10.06, | |
| "learning_rate": 9.712631578947369e-06, | |
| "loss": 0.0367, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 10.39, | |
| "learning_rate": 9.686315789473684e-06, | |
| "loss": 0.0248, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 10.39, | |
| "eval_loss": 0.8046127557754517, | |
| "eval_runtime": 19.9036, | |
| "eval_samples_per_second": 6.984, | |
| "eval_steps_per_second": 0.904, | |
| "eval_wer": 108.59188544152745, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 10.71, | |
| "learning_rate": 9.66e-06, | |
| "loss": 0.0324, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 11.04, | |
| "learning_rate": 9.633684210526316e-06, | |
| "loss": 0.0374, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 11.36, | |
| "learning_rate": 9.607368421052632e-06, | |
| "loss": 0.0253, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 11.68, | |
| "learning_rate": 9.581052631578947e-06, | |
| "loss": 0.0262, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "learning_rate": 9.554736842105264e-06, | |
| "loss": 0.0374, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 12.33, | |
| "learning_rate": 9.52842105263158e-06, | |
| "loss": 0.0179, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 12.66, | |
| "learning_rate": 9.502105263157896e-06, | |
| "loss": 0.0236, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 12.98, | |
| "learning_rate": 9.475789473684212e-06, | |
| "loss": 0.0187, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 13.31, | |
| "learning_rate": 9.449473684210527e-06, | |
| "loss": 0.0186, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 13.63, | |
| "learning_rate": 9.423157894736842e-06, | |
| "loss": 0.0217, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 13.95, | |
| "learning_rate": 9.396842105263159e-06, | |
| "loss": 0.0206, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 14.28, | |
| "learning_rate": 9.370526315789474e-06, | |
| "loss": 0.0103, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 14.6, | |
| "learning_rate": 9.34421052631579e-06, | |
| "loss": 0.0168, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 14.93, | |
| "learning_rate": 9.317894736842105e-06, | |
| "loss": 0.0214, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 15.26, | |
| "learning_rate": 9.291578947368422e-06, | |
| "loss": 0.0115, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 15.58, | |
| "learning_rate": 9.265263157894737e-06, | |
| "loss": 0.0138, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 15.58, | |
| "eval_loss": 0.8237143158912659, | |
| "eval_runtime": 22.295, | |
| "eval_samples_per_second": 6.235, | |
| "eval_steps_per_second": 0.807, | |
| "eval_wer": 117.89976133651552, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 15.9, | |
| "learning_rate": 9.238947368421052e-06, | |
| "loss": 0.0216, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 16.23, | |
| "learning_rate": 9.21263157894737e-06, | |
| "loss": 0.014, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 16.55, | |
| "learning_rate": 9.186315789473685e-06, | |
| "loss": 0.0117, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 16.87, | |
| "learning_rate": 9.16e-06, | |
| "loss": 0.0104, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 17.21, | |
| "learning_rate": 9.133684210526317e-06, | |
| "loss": 0.0086, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 17.53, | |
| "learning_rate": 9.107368421052632e-06, | |
| "loss": 0.0092, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 17.85, | |
| "learning_rate": 9.081052631578949e-06, | |
| "loss": 0.0098, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 18.18, | |
| "learning_rate": 9.054736842105264e-06, | |
| "loss": 0.0125, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 18.5, | |
| "learning_rate": 9.02842105263158e-06, | |
| "loss": 0.0055, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 18.82, | |
| "learning_rate": 9.002105263157895e-06, | |
| "loss": 0.011, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 19.15, | |
| "learning_rate": 8.97578947368421e-06, | |
| "loss": 0.0088, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 19.48, | |
| "learning_rate": 8.949473684210527e-06, | |
| "loss": 0.0065, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 19.8, | |
| "learning_rate": 8.923157894736842e-06, | |
| "loss": 0.0098, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 20.13, | |
| "learning_rate": 8.896842105263159e-06, | |
| "loss": 0.0038, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 20.45, | |
| "learning_rate": 8.870526315789474e-06, | |
| "loss": 0.0053, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 20.77, | |
| "learning_rate": 8.84421052631579e-06, | |
| "loss": 0.0075, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 20.77, | |
| "eval_loss": 0.8213781118392944, | |
| "eval_runtime": 18.9322, | |
| "eval_samples_per_second": 7.342, | |
| "eval_steps_per_second": 0.951, | |
| "eval_wer": 103.5799522673031, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 21.1, | |
| "learning_rate": 8.817894736842107e-06, | |
| "loss": 0.0111, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 21.42, | |
| "learning_rate": 8.791578947368422e-06, | |
| "loss": 0.0083, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 21.75, | |
| "learning_rate": 8.765263157894739e-06, | |
| "loss": 0.0056, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 22.08, | |
| "learning_rate": 8.738947368421053e-06, | |
| "loss": 0.0055, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 22.4, | |
| "learning_rate": 8.712631578947368e-06, | |
| "loss": 0.0031, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 22.72, | |
| "learning_rate": 8.686315789473685e-06, | |
| "loss": 0.0075, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 23.05, | |
| "learning_rate": 8.66e-06, | |
| "loss": 0.0078, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 23.37, | |
| "learning_rate": 8.633684210526317e-06, | |
| "loss": 0.0063, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 23.69, | |
| "learning_rate": 8.607368421052632e-06, | |
| "loss": 0.0012, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 24.03, | |
| "learning_rate": 8.581052631578948e-06, | |
| "loss": 0.0045, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 24.35, | |
| "learning_rate": 8.554736842105263e-06, | |
| "loss": 0.004, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 24.67, | |
| "learning_rate": 8.528421052631578e-06, | |
| "loss": 0.006, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 24.99, | |
| "learning_rate": 8.502105263157897e-06, | |
| "loss": 0.0034, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 25.32, | |
| "learning_rate": 8.475789473684212e-06, | |
| "loss": 0.0024, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 25.64, | |
| "learning_rate": 8.449473684210527e-06, | |
| "loss": 0.0032, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 25.96, | |
| "learning_rate": 8.423157894736843e-06, | |
| "loss": 0.0021, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 25.96, | |
| "eval_loss": 0.7943580746650696, | |
| "eval_runtime": 20.6652, | |
| "eval_samples_per_second": 6.726, | |
| "eval_steps_per_second": 0.871, | |
| "eval_wer": 106.92124105011933, | |
| "step": 2000 | |
| } | |
| ], | |
| "max_steps": 10000, | |
| "num_train_epochs": 130, | |
| "total_flos": 3.2965538512896e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |