| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 50.0, | |
| "global_step": 7500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.3579999999999999e-05, | |
| "loss": 16.4709, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 2.758e-05, | |
| "loss": 5.6159, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 4.157999999999999e-05, | |
| "loss": 3.8197, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 5.558e-05, | |
| "loss": 3.1231, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 6.958e-05, | |
| "loss": 2.9205, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "eval_loss": 2.865917205810547, | |
| "eval_runtime": 58.3236, | |
| "eval_samples_per_second": 35.646, | |
| "eval_steps_per_second": 1.114, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 6.903e-05, | |
| "loss": 2.8123, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "learning_rate": 6.803e-05, | |
| "loss": 2.4784, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "learning_rate": 6.702999999999999e-05, | |
| "loss": 2.0098, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 6.602999999999999e-05, | |
| "loss": 1.7892, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "learning_rate": 6.502999999999999e-05, | |
| "loss": 1.6403, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "eval_loss": 0.9439539909362793, | |
| "eval_runtime": 58.6422, | |
| "eval_samples_per_second": 35.452, | |
| "eval_steps_per_second": 1.108, | |
| "eval_wer": 0.7593160027676189, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.33, | |
| "learning_rate": 6.403e-05, | |
| "loss": 1.5762, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 6.303e-05, | |
| "loss": 1.5018, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 8.67, | |
| "learning_rate": 6.202999999999999e-05, | |
| "loss": 1.4272, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 9.33, | |
| "learning_rate": 6.103e-05, | |
| "loss": 1.3945, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 6.002999999999999e-05, | |
| "loss": 1.3483, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.7580158710479736, | |
| "eval_runtime": 57.9482, | |
| "eval_samples_per_second": 35.877, | |
| "eval_steps_per_second": 1.122, | |
| "eval_wer": 0.6215281209844815, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 10.67, | |
| "learning_rate": 5.9029999999999996e-05, | |
| "loss": 1.3139, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 11.33, | |
| "learning_rate": 5.802999999999999e-05, | |
| "loss": 1.3011, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 5.7029999999999994e-05, | |
| "loss": 1.2637, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 12.67, | |
| "learning_rate": 5.602999999999999e-05, | |
| "loss": 1.2293, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 13.33, | |
| "learning_rate": 5.502999999999999e-05, | |
| "loss": 1.2255, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 13.33, | |
| "eval_loss": 0.6850859522819519, | |
| "eval_runtime": 58.048, | |
| "eval_samples_per_second": 35.815, | |
| "eval_steps_per_second": 1.12, | |
| "eval_wer": 0.5722051991697144, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "learning_rate": 5.4029999999999996e-05, | |
| "loss": 1.1828, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 14.67, | |
| "learning_rate": 5.302999999999999e-05, | |
| "loss": 1.1484, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 15.33, | |
| "learning_rate": 5.203e-05, | |
| "loss": 1.1475, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 5.102999999999999e-05, | |
| "loss": 1.1336, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 5.003e-05, | |
| "loss": 1.1139, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "eval_loss": 0.6409106850624084, | |
| "eval_runtime": 58.0436, | |
| "eval_samples_per_second": 35.818, | |
| "eval_steps_per_second": 1.12, | |
| "eval_wer": 0.5219926855787289, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 17.33, | |
| "learning_rate": 4.9029999999999996e-05, | |
| "loss": 1.1248, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "learning_rate": 4.803e-05, | |
| "loss": 1.0867, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 18.67, | |
| "learning_rate": 4.7029999999999995e-05, | |
| "loss": 1.0591, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 19.33, | |
| "learning_rate": 4.603e-05, | |
| "loss": 1.0539, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 4.5029999999999994e-05, | |
| "loss": 1.0688, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 0.6245269775390625, | |
| "eval_runtime": 58.2955, | |
| "eval_samples_per_second": 35.663, | |
| "eval_steps_per_second": 1.115, | |
| "eval_wer": 0.5054858159533459, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 20.67, | |
| "learning_rate": 4.403e-05, | |
| "loss": 1.0503, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 21.33, | |
| "learning_rate": 4.303e-05, | |
| "loss": 1.0137, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "learning_rate": 4.2029999999999996e-05, | |
| "loss": 1.0103, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 22.67, | |
| "learning_rate": 4.103e-05, | |
| "loss": 0.9984, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 23.33, | |
| "learning_rate": 4.0029999999999995e-05, | |
| "loss": 0.99, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 23.33, | |
| "eval_loss": 0.6142364144325256, | |
| "eval_runtime": 57.332, | |
| "eval_samples_per_second": 36.262, | |
| "eval_steps_per_second": 1.134, | |
| "eval_wer": 0.4873974498369082, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "learning_rate": 3.903e-05, | |
| "loss": 0.9867, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 24.67, | |
| "learning_rate": 3.802999999999999e-05, | |
| "loss": 0.9613, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 25.33, | |
| "learning_rate": 3.7029999999999996e-05, | |
| "loss": 0.9642, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "learning_rate": 3.602999999999999e-05, | |
| "loss": 0.9616, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 26.67, | |
| "learning_rate": 3.5029999999999995e-05, | |
| "loss": 0.9345, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 26.67, | |
| "eval_loss": 0.5945756435394287, | |
| "eval_runtime": 58.8391, | |
| "eval_samples_per_second": 35.334, | |
| "eval_steps_per_second": 1.105, | |
| "eval_wer": 0.48294949095581696, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 27.33, | |
| "learning_rate": 3.403e-05, | |
| "loss": 0.9469, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "learning_rate": 3.303e-05, | |
| "loss": 0.9252, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 28.67, | |
| "learning_rate": 3.2029999999999997e-05, | |
| "loss": 0.9111, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 29.33, | |
| "learning_rate": 3.103e-05, | |
| "loss": 0.9042, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "learning_rate": 3.0029999999999995e-05, | |
| "loss": 0.9058, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_loss": 0.6228715777397156, | |
| "eval_runtime": 57.5967, | |
| "eval_samples_per_second": 36.096, | |
| "eval_steps_per_second": 1.129, | |
| "eval_wer": 0.47039636255807055, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 30.67, | |
| "learning_rate": 2.9029999999999998e-05, | |
| "loss": 0.8957, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 31.33, | |
| "learning_rate": 2.8029999999999998e-05, | |
| "loss": 0.897, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "learning_rate": 2.704e-05, | |
| "loss": 0.8733, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 32.67, | |
| "learning_rate": 2.6039999999999998e-05, | |
| "loss": 0.8852, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 2.5039999999999997e-05, | |
| "loss": 0.8683, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "eval_loss": 0.6152545809745789, | |
| "eval_runtime": 57.5209, | |
| "eval_samples_per_second": 36.143, | |
| "eval_steps_per_second": 1.13, | |
| "eval_wer": 0.46664030839181575, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "learning_rate": 2.4039999999999997e-05, | |
| "loss": 0.8671, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 34.67, | |
| "learning_rate": 2.3039999999999996e-05, | |
| "loss": 0.869, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 35.33, | |
| "learning_rate": 2.2039999999999996e-05, | |
| "loss": 0.8387, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "learning_rate": 2.1039999999999995e-05, | |
| "loss": 0.8543, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 36.67, | |
| "learning_rate": 2.004e-05, | |
| "loss": 0.8367, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 36.67, | |
| "eval_loss": 0.5952465534210205, | |
| "eval_runtime": 57.7371, | |
| "eval_samples_per_second": 36.008, | |
| "eval_steps_per_second": 1.126, | |
| "eval_wer": 0.4541860235247603, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 37.33, | |
| "learning_rate": 1.904e-05, | |
| "loss": 0.8508, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "learning_rate": 1.804e-05, | |
| "loss": 0.8356, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 38.67, | |
| "learning_rate": 1.704e-05, | |
| "loss": 0.8184, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 39.33, | |
| "learning_rate": 1.604e-05, | |
| "loss": 0.8333, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "learning_rate": 1.5039999999999998e-05, | |
| "loss": 0.8162, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_loss": 0.6029675602912903, | |
| "eval_runtime": 57.519, | |
| "eval_samples_per_second": 36.145, | |
| "eval_steps_per_second": 1.13, | |
| "eval_wer": 0.4540871799940694, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 40.67, | |
| "learning_rate": 1.4039999999999998e-05, | |
| "loss": 0.8091, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 41.33, | |
| "learning_rate": 1.3039999999999999e-05, | |
| "loss": 0.8081, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "learning_rate": 1.2039999999999998e-05, | |
| "loss": 0.8024, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 42.67, | |
| "learning_rate": 1.104e-05, | |
| "loss": 0.7959, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 43.33, | |
| "learning_rate": 1.0039999999999999e-05, | |
| "loss": 0.8042, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 43.33, | |
| "eval_loss": 0.59715735912323, | |
| "eval_runtime": 56.5464, | |
| "eval_samples_per_second": 36.766, | |
| "eval_steps_per_second": 1.149, | |
| "eval_wer": 0.44845309874468714, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "learning_rate": 9.04e-06, | |
| "loss": 0.797, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 44.67, | |
| "learning_rate": 8.04e-06, | |
| "loss": 0.777, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 45.33, | |
| "learning_rate": 7.0399999999999995e-06, | |
| "loss": 0.785, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "learning_rate": 6.04e-06, | |
| "loss": 0.7787, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 46.67, | |
| "learning_rate": 5.039999999999999e-06, | |
| "loss": 0.7836, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 46.67, | |
| "eval_loss": 0.6069726943969727, | |
| "eval_runtime": 57.3149, | |
| "eval_samples_per_second": 36.273, | |
| "eval_steps_per_second": 1.134, | |
| "eval_wer": 0.44973806464366906, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 47.33, | |
| "learning_rate": 4.0399999999999994e-06, | |
| "loss": 0.7594, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "learning_rate": 3.0399999999999997e-06, | |
| "loss": 0.7756, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 48.67, | |
| "learning_rate": 2.04e-06, | |
| "loss": 0.7748, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 49.33, | |
| "learning_rate": 1.04e-06, | |
| "loss": 0.7911, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 4e-08, | |
| "loss": 0.7556, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_loss": 0.6102160811424255, | |
| "eval_runtime": 57.2514, | |
| "eval_samples_per_second": 36.314, | |
| "eval_steps_per_second": 1.135, | |
| "eval_wer": 0.44548779282395967, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 7500, | |
| "total_flos": 2.217517579686915e+19, | |
| "train_loss": 1.4197682708740234, | |
| "train_runtime": 9558.3676, | |
| "train_samples_per_second": 25.057, | |
| "train_steps_per_second": 0.785 | |
| } | |
| ], | |
| "max_steps": 7500, | |
| "num_train_epochs": 50, | |
| "total_flos": 2.217517579686915e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |