| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 60.0, | |
| "global_step": 4920, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 7.35e-06, | |
| "loss": 12.2749, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 1.485e-05, | |
| "loss": 4.8071, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "learning_rate": 2.2349999999999998e-05, | |
| "loss": 3.5841, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "learning_rate": 2.985e-05, | |
| "loss": 3.1829, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "eval_loss": 3.1227903366088867, | |
| "eval_runtime": 33.8472, | |
| "eval_samples_per_second": 35.247, | |
| "eval_steps_per_second": 2.216, | |
| "eval_wer": 1.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "learning_rate": 3.735e-05, | |
| "loss": 3.0323, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 7.32, | |
| "learning_rate": 4.484999999999999e-05, | |
| "loss": 2.9603, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 8.54, | |
| "learning_rate": 5.234999999999999e-05, | |
| "loss": 2.8982, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 9.76, | |
| "learning_rate": 5.985e-05, | |
| "loss": 2.8675, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 9.76, | |
| "eval_loss": 2.8616490364074707, | |
| "eval_runtime": 33.8201, | |
| "eval_samples_per_second": 35.275, | |
| "eval_steps_per_second": 2.218, | |
| "eval_wer": 0.9992680427463037, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 10.98, | |
| "learning_rate": 6.735e-05, | |
| "loss": 2.7893, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "learning_rate": 7.484999999999999e-05, | |
| "loss": 2.2812, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 13.41, | |
| "learning_rate": 7.312499999999999e-05, | |
| "loss": 1.7369, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 14.63, | |
| "learning_rate": 7.121173469387755e-05, | |
| "loss": 1.583, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 14.63, | |
| "eval_loss": 0.6392253637313843, | |
| "eval_runtime": 33.9553, | |
| "eval_samples_per_second": 35.134, | |
| "eval_steps_per_second": 2.209, | |
| "eval_wer": 0.6239203630507978, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 15.85, | |
| "learning_rate": 6.92984693877551e-05, | |
| "loss": 1.4572, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 17.07, | |
| "learning_rate": 6.738520408163265e-05, | |
| "loss": 1.3366, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 18.29, | |
| "learning_rate": 6.547193877551019e-05, | |
| "loss": 1.2668, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 19.51, | |
| "learning_rate": 6.355867346938775e-05, | |
| "loss": 1.1959, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 19.51, | |
| "eval_loss": 0.36018967628479004, | |
| "eval_runtime": 33.3624, | |
| "eval_samples_per_second": 35.759, | |
| "eval_steps_per_second": 2.248, | |
| "eval_wer": 0.3651002781437564, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 20.73, | |
| "learning_rate": 6.164540816326531e-05, | |
| "loss": 1.1694, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 21.95, | |
| "learning_rate": 5.9732142857142846e-05, | |
| "loss": 1.1001, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 23.17, | |
| "learning_rate": 5.7818877551020404e-05, | |
| "loss": 1.0737, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 24.39, | |
| "learning_rate": 5.5905612244897955e-05, | |
| "loss": 1.0276, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 24.39, | |
| "eval_loss": 0.3020930588245392, | |
| "eval_runtime": 33.6823, | |
| "eval_samples_per_second": 35.419, | |
| "eval_steps_per_second": 2.227, | |
| "eval_wer": 0.29805299370516763, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 25.61, | |
| "learning_rate": 5.3992346938775506e-05, | |
| "loss": 1.0293, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 26.83, | |
| "learning_rate": 5.207908163265306e-05, | |
| "loss": 0.9966, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 28.05, | |
| "learning_rate": 5.016581632653061e-05, | |
| "loss": 0.9721, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 29.27, | |
| "learning_rate": 4.8271683673469385e-05, | |
| "loss": 0.9671, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 29.27, | |
| "eval_loss": 0.28717485070228577, | |
| "eval_runtime": 33.2455, | |
| "eval_samples_per_second": 35.885, | |
| "eval_steps_per_second": 2.256, | |
| "eval_wer": 0.27389840433318696, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 30.49, | |
| "learning_rate": 4.635841836734693e-05, | |
| "loss": 0.9536, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 31.71, | |
| "learning_rate": 4.444515306122448e-05, | |
| "loss": 0.934, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 32.93, | |
| "learning_rate": 4.253188775510204e-05, | |
| "loss": 0.9136, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 34.15, | |
| "learning_rate": 4.061862244897959e-05, | |
| "loss": 0.873, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 34.15, | |
| "eval_loss": 0.259307324886322, | |
| "eval_runtime": 33.4272, | |
| "eval_samples_per_second": 35.689, | |
| "eval_steps_per_second": 2.244, | |
| "eval_wer": 0.24593763724198506, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 35.37, | |
| "learning_rate": 3.870535714285714e-05, | |
| "loss": 0.8649, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 36.59, | |
| "learning_rate": 3.679209183673469e-05, | |
| "loss": 0.8464, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 37.8, | |
| "learning_rate": 3.487882653061224e-05, | |
| "loss": 0.8545, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 39.02, | |
| "learning_rate": 3.2965561224489794e-05, | |
| "loss": 0.8513, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 39.02, | |
| "eval_loss": 0.2617344856262207, | |
| "eval_runtime": 33.3367, | |
| "eval_samples_per_second": 35.786, | |
| "eval_steps_per_second": 2.25, | |
| "eval_wer": 0.24725516029863856, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 40.24, | |
| "learning_rate": 3.1052295918367345e-05, | |
| "loss": 0.8424, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 41.46, | |
| "learning_rate": 2.9139030612244896e-05, | |
| "loss": 0.8244, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 42.68, | |
| "learning_rate": 2.7225765306122448e-05, | |
| "loss": 0.8157, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 43.9, | |
| "learning_rate": 2.53125e-05, | |
| "loss": 0.8132, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 43.9, | |
| "eval_loss": 0.25475597381591797, | |
| "eval_runtime": 33.5269, | |
| "eval_samples_per_second": 35.583, | |
| "eval_steps_per_second": 2.237, | |
| "eval_wer": 0.2425706338749817, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 45.12, | |
| "learning_rate": 2.339923469387755e-05, | |
| "loss": 0.8008, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 46.34, | |
| "learning_rate": 2.14859693877551e-05, | |
| "loss": 0.7863, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 47.56, | |
| "learning_rate": 1.9572704081632652e-05, | |
| "loss": 0.7965, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 48.78, | |
| "learning_rate": 1.7659438775510203e-05, | |
| "loss": 0.7935, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 48.78, | |
| "eval_loss": 0.2637428939342499, | |
| "eval_runtime": 33.4727, | |
| "eval_samples_per_second": 35.641, | |
| "eval_steps_per_second": 2.241, | |
| "eval_wer": 0.23525106133801785, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 1.5746173469387754e-05, | |
| "loss": 0.7857, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 51.22, | |
| "learning_rate": 1.3832908163265304e-05, | |
| "loss": 0.7834, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 52.44, | |
| "learning_rate": 1.1919642857142855e-05, | |
| "loss": 0.7732, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 53.66, | |
| "learning_rate": 1.0006377551020406e-05, | |
| "loss": 0.7565, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 53.66, | |
| "eval_loss": 0.2629481256008148, | |
| "eval_runtime": 33.9009, | |
| "eval_samples_per_second": 35.191, | |
| "eval_steps_per_second": 2.212, | |
| "eval_wer": 0.23217684087249305, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 54.88, | |
| "learning_rate": 8.093112244897959e-06, | |
| "loss": 0.7525, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 56.1, | |
| "learning_rate": 6.198979591836734e-06, | |
| "loss": 0.7716, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 57.32, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.7521, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 58.54, | |
| "learning_rate": 2.372448979591836e-06, | |
| "loss": 0.7359, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 58.54, | |
| "eval_loss": 0.2578625977039337, | |
| "eval_runtime": 33.5888, | |
| "eval_samples_per_second": 35.518, | |
| "eval_steps_per_second": 2.233, | |
| "eval_wer": 0.22529644268774704, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 59.76, | |
| "learning_rate": 4.591836734693877e-07, | |
| "loss": 0.7476, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "step": 4920, | |
| "total_flos": 1.5656877251307375e+19, | |
| "train_loss": 1.5886853283982936, | |
| "train_runtime": 6665.2028, | |
| "train_samples_per_second": 23.459, | |
| "train_steps_per_second": 0.738 | |
| } | |
| ], | |
| "max_steps": 4920, | |
| "num_train_epochs": 60, | |
| "total_flos": 1.5656877251307375e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |