| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "global_step": 5900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 5.88e-05, | |
| "loss": 7.6789, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "learning_rate": 0.0001188, | |
| "loss": 3.2952, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 5.08, | |
| "learning_rate": 0.00017879999999999998, | |
| "loss": 3.0495, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 6.78, | |
| "learning_rate": 0.0002388, | |
| "loss": 2.7948, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 8.47, | |
| "learning_rate": 0.0002988, | |
| "loss": 1.8597, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.47, | |
| "eval_loss": 0.7730758786201477, | |
| "eval_runtime": 49.5825, | |
| "eval_samples_per_second": 16.135, | |
| "eval_steps_per_second": 16.135, | |
| "eval_wer": 0.7211251598238386, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 10.17, | |
| "learning_rate": 0.00029455555555555555, | |
| "loss": 1.4933, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 11.86, | |
| "learning_rate": 0.000289, | |
| "loss": 1.365, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 13.56, | |
| "learning_rate": 0.0002834444444444444, | |
| "loss": 1.3187, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 15.25, | |
| "learning_rate": 0.0002778888888888889, | |
| "loss": 1.2744, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 16.95, | |
| "learning_rate": 0.0002723333333333333, | |
| "loss": 1.2508, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 16.95, | |
| "eval_loss": 0.5367600321769714, | |
| "eval_runtime": 48.7576, | |
| "eval_samples_per_second": 16.408, | |
| "eval_steps_per_second": 16.408, | |
| "eval_wer": 0.5989487143060094, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 18.64, | |
| "learning_rate": 0.0002667777777777778, | |
| "loss": 1.1856, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 20.34, | |
| "learning_rate": 0.0002612222222222222, | |
| "loss": 1.1821, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 22.03, | |
| "learning_rate": 0.00025566666666666663, | |
| "loss": 1.1284, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 23.73, | |
| "learning_rate": 0.0002501111111111111, | |
| "loss": 1.1204, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 25.42, | |
| "learning_rate": 0.00024455555555555553, | |
| "loss": 1.1066, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 25.42, | |
| "eval_loss": 0.5033634901046753, | |
| "eval_runtime": 48.524, | |
| "eval_samples_per_second": 16.487, | |
| "eval_steps_per_second": 16.487, | |
| "eval_wer": 0.553345645688308, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 27.12, | |
| "learning_rate": 0.00023905555555555553, | |
| "loss": 1.0826, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 28.81, | |
| "learning_rate": 0.00023349999999999998, | |
| "loss": 1.0375, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 30.51, | |
| "learning_rate": 0.00022794444444444443, | |
| "loss": 1.0304, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 32.2, | |
| "learning_rate": 0.00022238888888888889, | |
| "loss": 1.0207, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 33.9, | |
| "learning_rate": 0.0002168333333333333, | |
| "loss": 1.0064, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 33.9, | |
| "eval_loss": 0.46856507658958435, | |
| "eval_runtime": 48.9718, | |
| "eval_samples_per_second": 16.336, | |
| "eval_steps_per_second": 16.336, | |
| "eval_wer": 0.5114362835630061, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 35.59, | |
| "learning_rate": 0.00021127777777777776, | |
| "loss": 0.9979, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 37.29, | |
| "learning_rate": 0.0002057222222222222, | |
| "loss": 0.9902, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 38.98, | |
| "learning_rate": 0.00020016666666666666, | |
| "loss": 0.9511, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 40.68, | |
| "learning_rate": 0.00019461111111111109, | |
| "loss": 0.9584, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 42.37, | |
| "learning_rate": 0.00018905555555555554, | |
| "loss": 0.9324, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 42.37, | |
| "eval_loss": 0.49271178245544434, | |
| "eval_runtime": 48.7262, | |
| "eval_samples_per_second": 16.418, | |
| "eval_steps_per_second": 16.418, | |
| "eval_wer": 0.5056115925557607, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 44.07, | |
| "learning_rate": 0.0001835, | |
| "loss": 0.9287, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 45.76, | |
| "learning_rate": 0.00017794444444444444, | |
| "loss": 0.9264, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 47.46, | |
| "learning_rate": 0.00017238888888888886, | |
| "loss": 0.9031, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 49.15, | |
| "learning_rate": 0.00016683333333333331, | |
| "loss": 0.8996, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 50.85, | |
| "learning_rate": 0.00016127777777777776, | |
| "loss": 0.876, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 50.85, | |
| "eval_loss": 0.4733906686306, | |
| "eval_runtime": 48.5334, | |
| "eval_samples_per_second": 16.484, | |
| "eval_steps_per_second": 16.484, | |
| "eval_wer": 0.4794715158403182, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 52.54, | |
| "learning_rate": 0.0001557222222222222, | |
| "loss": 0.8432, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 54.24, | |
| "learning_rate": 0.00015016666666666664, | |
| "loss": 0.8592, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 55.93, | |
| "learning_rate": 0.0001446111111111111, | |
| "loss": 0.8322, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 57.63, | |
| "learning_rate": 0.00013905555555555554, | |
| "loss": 0.8286, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 59.32, | |
| "learning_rate": 0.0001335, | |
| "loss": 0.8082, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 59.32, | |
| "eval_loss": 0.474797785282135, | |
| "eval_runtime": 48.7764, | |
| "eval_samples_per_second": 16.401, | |
| "eval_steps_per_second": 16.401, | |
| "eval_wer": 0.4798977127432874, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 61.02, | |
| "learning_rate": 0.00012794444444444442, | |
| "loss": 0.8101, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 62.71, | |
| "learning_rate": 0.0001223888888888889, | |
| "loss": 0.7805, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 64.41, | |
| "learning_rate": 0.00011683333333333332, | |
| "loss": 0.7741, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 66.1, | |
| "learning_rate": 0.00011127777777777777, | |
| "loss": 0.773, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 67.8, | |
| "learning_rate": 0.0001057222222222222, | |
| "loss": 0.7604, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 67.8, | |
| "eval_loss": 0.49494636058807373, | |
| "eval_runtime": 48.9958, | |
| "eval_samples_per_second": 16.328, | |
| "eval_steps_per_second": 16.328, | |
| "eval_wer": 0.46910072453473506, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 69.49, | |
| "learning_rate": 0.00010016666666666666, | |
| "loss": 0.7527, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 71.19, | |
| "learning_rate": 9.46111111111111e-05, | |
| "loss": 0.7354, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 72.88, | |
| "learning_rate": 8.905555555555555e-05, | |
| "loss": 0.7334, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 74.58, | |
| "learning_rate": 8.349999999999998e-05, | |
| "loss": 0.7286, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 76.27, | |
| "learning_rate": 7.794444444444445e-05, | |
| "loss": 0.7241, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 76.27, | |
| "eval_loss": 0.5090368390083313, | |
| "eval_runtime": 49.0851, | |
| "eval_samples_per_second": 16.298, | |
| "eval_steps_per_second": 16.298, | |
| "eval_wer": 0.4627077709901975, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 77.97, | |
| "learning_rate": 7.238888888888889e-05, | |
| "loss": 0.7065, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 79.66, | |
| "learning_rate": 6.683333333333332e-05, | |
| "loss": 0.6978, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 81.36, | |
| "learning_rate": 6.127777777777777e-05, | |
| "loss": 0.688, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 83.05, | |
| "learning_rate": 5.572222222222222e-05, | |
| "loss": 0.6725, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 84.75, | |
| "learning_rate": 5.016666666666666e-05, | |
| "loss": 0.6739, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 84.75, | |
| "eval_loss": 0.49671733379364014, | |
| "eval_runtime": 50.1856, | |
| "eval_samples_per_second": 15.941, | |
| "eval_steps_per_second": 15.941, | |
| "eval_wer": 0.4452336979684614, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 86.44, | |
| "learning_rate": 4.4611111111111106e-05, | |
| "loss": 0.6675, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 88.14, | |
| "learning_rate": 3.905555555555555e-05, | |
| "loss": 0.6548, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 89.83, | |
| "learning_rate": 3.3499999999999994e-05, | |
| "loss": 0.6465, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 91.53, | |
| "learning_rate": 2.8e-05, | |
| "loss": 0.6471, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 93.22, | |
| "learning_rate": 2.2444444444444444e-05, | |
| "loss": 0.6447, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 93.22, | |
| "eval_loss": 0.5071204900741577, | |
| "eval_runtime": 50.5031, | |
| "eval_samples_per_second": 15.841, | |
| "eval_steps_per_second": 15.841, | |
| "eval_wer": 0.4436709759909078, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 94.92, | |
| "learning_rate": 1.6888888888888888e-05, | |
| "loss": 0.6362, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 96.61, | |
| "learning_rate": 1.1333333333333332e-05, | |
| "loss": 0.6434, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 98.31, | |
| "learning_rate": 5.777777777777778e-06, | |
| "loss": 0.6269, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 2.222222222222222e-07, | |
| "loss": 0.6369, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 5900, | |
| "total_flos": 3.4027105250131104e+19, | |
| "train_loss": 1.1293389116707495, | |
| "train_runtime": 13204.1924, | |
| "train_samples_per_second": 14.064, | |
| "train_steps_per_second": 0.447 | |
| } | |
| ], | |
| "max_steps": 5900, | |
| "num_train_epochs": 100, | |
| "total_flos": 3.4027105250131104e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |