| { | |
| "best_metric": 22.969736555628128, | |
| "best_model_checkpoint": "./results/checkpoint-1000", | |
| "epoch": 62.5, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 4.4e-07, | |
| "loss": 1.7533, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "learning_rate": 9.400000000000001e-07, | |
| "loss": 1.6609, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "learning_rate": 1.44e-06, | |
| "loss": 1.3484, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 1.94e-06, | |
| "loss": 1.1098, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "learning_rate": 2.4400000000000004e-06, | |
| "loss": 0.8698, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "learning_rate": 2.9400000000000002e-06, | |
| "loss": 0.662, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 5.47, | |
| "learning_rate": 3.44e-06, | |
| "loss": 0.5234, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "learning_rate": 3.94e-06, | |
| "loss": 0.4031, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "learning_rate": 4.440000000000001e-06, | |
| "loss": 0.3321, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 7.81, | |
| "learning_rate": 4.94e-06, | |
| "loss": 0.2633, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 8.59, | |
| "learning_rate": 5.4400000000000004e-06, | |
| "loss": 0.2306, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 9.38, | |
| "learning_rate": 5.94e-06, | |
| "loss": 0.1706, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 10.16, | |
| "learning_rate": 6.440000000000001e-06, | |
| "loss": 0.1469, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 10.94, | |
| "learning_rate": 6.9400000000000005e-06, | |
| "loss": 0.1107, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 11.72, | |
| "learning_rate": 7.440000000000001e-06, | |
| "loss": 0.0825, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 7.94e-06, | |
| "loss": 0.0597, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 13.28, | |
| "learning_rate": 8.44e-06, | |
| "loss": 0.0456, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 14.06, | |
| "learning_rate": 8.94e-06, | |
| "loss": 0.0321, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 14.84, | |
| "learning_rate": 9.440000000000001e-06, | |
| "loss": 0.0279, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 15.62, | |
| "learning_rate": 9.940000000000001e-06, | |
| "loss": 0.0176, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 16.41, | |
| "learning_rate": 9.937142857142858e-06, | |
| "loss": 0.0186, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 17.19, | |
| "learning_rate": 9.865714285714285e-06, | |
| "loss": 0.0134, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 17.97, | |
| "learning_rate": 9.794285714285714e-06, | |
| "loss": 0.0189, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 18.75, | |
| "learning_rate": 9.722857142857143e-06, | |
| "loss": 0.0195, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 19.53, | |
| "learning_rate": 9.651428571428572e-06, | |
| "loss": 0.0122, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 20.31, | |
| "learning_rate": 9.58e-06, | |
| "loss": 0.0125, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 21.09, | |
| "learning_rate": 9.508571428571429e-06, | |
| "loss": 0.0075, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 21.88, | |
| "learning_rate": 9.437142857142858e-06, | |
| "loss": 0.0074, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 22.66, | |
| "learning_rate": 9.365714285714287e-06, | |
| "loss": 0.0078, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 23.44, | |
| "learning_rate": 9.294285714285714e-06, | |
| "loss": 0.0047, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 24.22, | |
| "learning_rate": 9.222857142857143e-06, | |
| "loss": 0.0067, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 9.151428571428572e-06, | |
| "loss": 0.0046, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 25.78, | |
| "learning_rate": 9.080000000000001e-06, | |
| "loss": 0.0048, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 26.56, | |
| "learning_rate": 9.00857142857143e-06, | |
| "loss": 0.0045, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 27.34, | |
| "learning_rate": 8.937142857142857e-06, | |
| "loss": 0.0054, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 28.12, | |
| "learning_rate": 8.865714285714287e-06, | |
| "loss": 0.0035, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 28.91, | |
| "learning_rate": 8.794285714285716e-06, | |
| "loss": 0.0043, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 29.69, | |
| "learning_rate": 8.722857142857145e-06, | |
| "loss": 0.0041, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 30.47, | |
| "learning_rate": 8.651428571428572e-06, | |
| "loss": 0.0051, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 31.25, | |
| "learning_rate": 8.580000000000001e-06, | |
| "loss": 0.0044, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 31.25, | |
| "eval_loss": 0.5659037828445435, | |
| "eval_runtime": 45.0123, | |
| "eval_samples_per_second": 2.799, | |
| "eval_steps_per_second": 0.355, | |
| "eval_wer": 22.969736555628128, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 32.03, | |
| "learning_rate": 8.50857142857143e-06, | |
| "loss": 0.0075, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 32.81, | |
| "learning_rate": 8.437142857142859e-06, | |
| "loss": 0.0035, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 33.59, | |
| "learning_rate": 8.365714285714286e-06, | |
| "loss": 0.0041, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 34.38, | |
| "learning_rate": 8.294285714285715e-06, | |
| "loss": 0.0034, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 35.16, | |
| "learning_rate": 8.222857142857144e-06, | |
| "loss": 0.003, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 35.94, | |
| "learning_rate": 8.151428571428572e-06, | |
| "loss": 0.0026, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 36.72, | |
| "learning_rate": 8.08e-06, | |
| "loss": 0.003, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 37.5, | |
| "learning_rate": 8.00857142857143e-06, | |
| "loss": 0.002, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 38.28, | |
| "learning_rate": 7.937142857142857e-06, | |
| "loss": 0.0029, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 39.06, | |
| "learning_rate": 7.865714285714286e-06, | |
| "loss": 0.0022, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 39.84, | |
| "learning_rate": 7.794285714285715e-06, | |
| "loss": 0.0019, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 40.62, | |
| "learning_rate": 7.722857142857142e-06, | |
| "loss": 0.0016, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 41.41, | |
| "learning_rate": 7.651428571428571e-06, | |
| "loss": 0.0023, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 42.19, | |
| "learning_rate": 7.58e-06, | |
| "loss": 0.0014, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 42.97, | |
| "learning_rate": 7.508571428571429e-06, | |
| "loss": 0.0009, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 43.75, | |
| "learning_rate": 7.4371428571428575e-06, | |
| "loss": 0.0012, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 44.53, | |
| "learning_rate": 7.365714285714286e-06, | |
| "loss": 0.0011, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 45.31, | |
| "learning_rate": 7.294285714285715e-06, | |
| "loss": 0.0027, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 46.09, | |
| "learning_rate": 7.222857142857144e-06, | |
| "loss": 0.0009, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 46.88, | |
| "learning_rate": 7.151428571428573e-06, | |
| "loss": 0.0011, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 47.66, | |
| "learning_rate": 7.08e-06, | |
| "loss": 0.0009, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 48.44, | |
| "learning_rate": 7.008571428571429e-06, | |
| "loss": 0.0014, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 49.22, | |
| "learning_rate": 6.937142857142858e-06, | |
| "loss": 0.0017, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 6.865714285714287e-06, | |
| "loss": 0.0019, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 50.78, | |
| "learning_rate": 6.794285714285714e-06, | |
| "loss": 0.0009, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 51.56, | |
| "learning_rate": 6.722857142857143e-06, | |
| "loss": 0.0007, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 52.34, | |
| "learning_rate": 6.651428571428572e-06, | |
| "loss": 0.001, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 53.12, | |
| "learning_rate": 6.5800000000000005e-06, | |
| "loss": 0.0004, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 53.91, | |
| "learning_rate": 6.5085714285714295e-06, | |
| "loss": 0.0009, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 54.69, | |
| "learning_rate": 6.437142857142858e-06, | |
| "loss": 0.0004, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 55.47, | |
| "learning_rate": 6.365714285714286e-06, | |
| "loss": 0.0005, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 56.25, | |
| "learning_rate": 6.294285714285715e-06, | |
| "loss": 0.0005, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 57.03, | |
| "learning_rate": 6.222857142857144e-06, | |
| "loss": 0.0006, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 57.81, | |
| "learning_rate": 6.151428571428571e-06, | |
| "loss": 0.0011, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 58.59, | |
| "learning_rate": 6.08e-06, | |
| "loss": 0.001, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 59.38, | |
| "learning_rate": 6.008571428571429e-06, | |
| "loss": 0.0006, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 60.16, | |
| "learning_rate": 5.937142857142858e-06, | |
| "loss": 0.0007, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 60.94, | |
| "learning_rate": 5.865714285714286e-06, | |
| "loss": 0.0005, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 61.72, | |
| "learning_rate": 5.794285714285715e-06, | |
| "loss": 0.0005, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 62.5, | |
| "learning_rate": 5.722857142857144e-06, | |
| "loss": 0.0005, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 62.5, | |
| "eval_loss": 0.6124230027198792, | |
| "eval_runtime": 40.8325, | |
| "eval_samples_per_second": 3.086, | |
| "eval_steps_per_second": 0.392, | |
| "eval_wer": 23.66644894404529, | |
| "step": 2000 | |
| } | |
| ], | |
| "max_steps": 4000, | |
| "num_train_epochs": 125, | |
| "total_flos": 2.03932799926272e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |