| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 100, | |
| "global_step": 900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 0.00019777777777777778, | |
| "loss": 1.9886, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 0.00019555555555555556, | |
| "loss": 1.6024, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 0.00019333333333333333, | |
| "loss": 1.3753, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 0.00019111111111111114, | |
| "loss": 1.3624, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 0.00018888888888888888, | |
| "loss": 1.3255, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 0.0001866666666666667, | |
| "loss": 0.9955, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 0.00018444444444444446, | |
| "loss": 0.9689, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 0.00018222222222222224, | |
| "loss": 0.9057, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.00018, | |
| "loss": 0.8525, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 0.00017777777777777779, | |
| "loss": 0.6539, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 0.8062787652015686, | |
| "eval_runtime": 12.7041, | |
| "eval_samples_per_second": 85.012, | |
| "eval_steps_per_second": 10.626, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 0.00017555555555555556, | |
| "loss": 0.6373, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 0.00017333333333333334, | |
| "loss": 0.5875, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 0.0001711111111111111, | |
| "loss": 0.5479, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "learning_rate": 0.00016888888888888889, | |
| "loss": 0.5463, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 0.5625, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 0.00016444444444444444, | |
| "loss": 0.5033, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 0.00016222222222222224, | |
| "loss": 0.5121, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 0.00016, | |
| "loss": 0.5484, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 0.0001577777777777778, | |
| "loss": 0.3358, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 0.00015555555555555556, | |
| "loss": 0.2553, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "eval_accuracy": 0.8351851851851851, | |
| "eval_loss": 0.5555349588394165, | |
| "eval_runtime": 12.687, | |
| "eval_samples_per_second": 85.127, | |
| "eval_steps_per_second": 10.641, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 0.00015333333333333334, | |
| "loss": 0.3577, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 0.0001511111111111111, | |
| "loss": 0.3141, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 0.0001488888888888889, | |
| "loss": 0.3053, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 0.00014666666666666666, | |
| "loss": 0.2487, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 0.00014444444444444444, | |
| "loss": 0.3129, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 0.00014222222222222224, | |
| "loss": 0.2975, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 0.00014, | |
| "loss": 0.2661, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "learning_rate": 0.0001377777777777778, | |
| "loss": 0.1503, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 0.00013555555555555556, | |
| "loss": 0.1649, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 0.1909, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "eval_accuracy": 0.8453703703703703, | |
| "eval_loss": 0.5217463970184326, | |
| "eval_runtime": 12.8584, | |
| "eval_samples_per_second": 83.992, | |
| "eval_steps_per_second": 10.499, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "learning_rate": 0.00013111111111111111, | |
| "loss": 0.1387, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "learning_rate": 0.00012888888888888892, | |
| "loss": 0.1357, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "learning_rate": 0.00012666666666666666, | |
| "loss": 0.1582, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "learning_rate": 0.00012444444444444444, | |
| "loss": 0.1676, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "learning_rate": 0.00012222222222222224, | |
| "loss": 0.156, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 0.00012, | |
| "loss": 0.1711, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "learning_rate": 0.00011777777777777779, | |
| "loss": 0.1132, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "learning_rate": 0.00011555555555555555, | |
| "loss": 0.1259, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "learning_rate": 0.00011333333333333334, | |
| "loss": 0.1086, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "learning_rate": 0.00011111111111111112, | |
| "loss": 0.0999, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "eval_accuracy": 0.8722222222222222, | |
| "eval_loss": 0.5075178146362305, | |
| "eval_runtime": 15.6251, | |
| "eval_samples_per_second": 69.119, | |
| "eval_steps_per_second": 8.64, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "learning_rate": 0.00010888888888888889, | |
| "loss": 0.0736, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "learning_rate": 0.00010666666666666667, | |
| "loss": 0.0755, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "learning_rate": 0.00010444444444444445, | |
| "loss": 0.0929, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "learning_rate": 0.00010222222222222222, | |
| "loss": 0.0873, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1017, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "learning_rate": 9.777777777777778e-05, | |
| "loss": 0.0312, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "learning_rate": 9.555555555555557e-05, | |
| "loss": 0.0707, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "learning_rate": 9.333333333333334e-05, | |
| "loss": 0.0412, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "learning_rate": 9.111111111111112e-05, | |
| "loss": 0.0315, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 5.56, | |
| "learning_rate": 8.888888888888889e-05, | |
| "loss": 0.0666, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.56, | |
| "eval_accuracy": 0.8768518518518519, | |
| "eval_loss": 0.4633159637451172, | |
| "eval_runtime": 12.8293, | |
| "eval_samples_per_second": 84.183, | |
| "eval_steps_per_second": 10.523, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.67, | |
| "learning_rate": 8.666666666666667e-05, | |
| "loss": 0.0705, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 5.78, | |
| "learning_rate": 8.444444444444444e-05, | |
| "loss": 0.0272, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "learning_rate": 8.222222222222222e-05, | |
| "loss": 0.0291, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 8e-05, | |
| "loss": 0.0503, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 6.11, | |
| "learning_rate": 7.777777777777778e-05, | |
| "loss": 0.031, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 6.22, | |
| "learning_rate": 7.555555555555556e-05, | |
| "loss": 0.0474, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 6.33, | |
| "learning_rate": 7.333333333333333e-05, | |
| "loss": 0.0203, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 6.44, | |
| "learning_rate": 7.111111111111112e-05, | |
| "loss": 0.0235, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "learning_rate": 6.88888888888889e-05, | |
| "loss": 0.0203, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 0.0392, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "eval_accuracy": 0.8740740740740741, | |
| "eval_loss": 0.4613787829875946, | |
| "eval_runtime": 13.0592, | |
| "eval_samples_per_second": 82.7, | |
| "eval_steps_per_second": 10.338, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.78, | |
| "learning_rate": 6.444444444444446e-05, | |
| "loss": 0.0242, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 6.89, | |
| "learning_rate": 6.222222222222222e-05, | |
| "loss": 0.0277, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 6e-05, | |
| "loss": 0.0151, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 7.11, | |
| "learning_rate": 5.7777777777777776e-05, | |
| "loss": 0.0189, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 7.22, | |
| "learning_rate": 5.555555555555556e-05, | |
| "loss": 0.0181, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 7.33, | |
| "learning_rate": 5.333333333333333e-05, | |
| "loss": 0.0133, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 7.44, | |
| "learning_rate": 5.111111111111111e-05, | |
| "loss": 0.0236, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 7.56, | |
| "learning_rate": 4.888888888888889e-05, | |
| "loss": 0.015, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 7.67, | |
| "learning_rate": 4.666666666666667e-05, | |
| "loss": 0.0137, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 7.78, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.0111, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 7.78, | |
| "eval_accuracy": 0.9101851851851852, | |
| "eval_loss": 0.35738176107406616, | |
| "eval_runtime": 12.567, | |
| "eval_samples_per_second": 85.939, | |
| "eval_steps_per_second": 10.742, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 7.89, | |
| "learning_rate": 4.222222222222222e-05, | |
| "loss": 0.0182, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0135, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 8.11, | |
| "learning_rate": 3.777777777777778e-05, | |
| "loss": 0.0121, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "learning_rate": 3.555555555555556e-05, | |
| "loss": 0.013, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.0111, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 8.44, | |
| "learning_rate": 3.111111111111111e-05, | |
| "loss": 0.0104, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 8.56, | |
| "learning_rate": 2.8888888888888888e-05, | |
| "loss": 0.0122, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 8.67, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.0155, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 8.78, | |
| "learning_rate": 2.4444444444444445e-05, | |
| "loss": 0.0089, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 8.89, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.0122, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 8.89, | |
| "eval_accuracy": 0.9166666666666666, | |
| "eval_loss": 0.3158920705318451, | |
| "eval_runtime": 13.6141, | |
| "eval_samples_per_second": 79.33, | |
| "eval_steps_per_second": 9.916, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0082, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 9.11, | |
| "learning_rate": 1.777777777777778e-05, | |
| "loss": 0.0118, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 9.22, | |
| "learning_rate": 1.5555555555555555e-05, | |
| "loss": 0.009, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 9.33, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.0079, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 9.44, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.008, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 9.56, | |
| "learning_rate": 8.88888888888889e-06, | |
| "loss": 0.0154, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 9.67, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.0096, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 9.78, | |
| "learning_rate": 4.444444444444445e-06, | |
| "loss": 0.0078, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 9.89, | |
| "learning_rate": 2.2222222222222225e-06, | |
| "loss": 0.0081, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.0112, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9175925925925926, | |
| "eval_loss": 0.3201213479042053, | |
| "eval_runtime": 12.9343, | |
| "eval_samples_per_second": 83.499, | |
| "eval_steps_per_second": 10.437, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 900, | |
| "total_flos": 3.347863979094835e+18, | |
| "train_loss": 0.24984459381964472, | |
| "train_runtime": 897.5432, | |
| "train_samples_per_second": 48.131, | |
| "train_steps_per_second": 1.003 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 900, | |
| "num_train_epochs": 10, | |
| "save_steps": 1000, | |
| "total_flos": 3.347863979094835e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |