| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "global_step": 109, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0, | |
| "loss": 14.822, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0, | |
| "loss": 15.2906, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1e-06, | |
| "loss": 15.9083, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2e-06, | |
| "loss": 15.062, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 3e-06, | |
| "loss": 14.1132, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 4e-06, | |
| "loss": 15.2949, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5e-06, | |
| "loss": 13.8691, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 6e-06, | |
| "loss": 13.307, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 6e-06, | |
| "loss": 12.5476, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 7e-06, | |
| "loss": 13.2011, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 8e-06, | |
| "loss": 12.3008, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 9e-06, | |
| "loss": 11.5046, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 1e-05, | |
| "loss": 10.6207, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 1.1e-05, | |
| "loss": 10.1038, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 1.2e-05, | |
| "loss": 8.8277, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 1.3e-05, | |
| "loss": 7.8204, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 1.4e-05, | |
| "loss": 6.3586, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 1.5e-05, | |
| "loss": 5.6274, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 1.6e-05, | |
| "loss": 5.6395, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 1.7000000000000003e-05, | |
| "loss": 5.0595, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 1.8e-05, | |
| "loss": 5.2143, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 1.9e-05, | |
| "loss": 5.5995, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 2e-05, | |
| "loss": 4.9733, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 2.1000000000000002e-05, | |
| "loss": 4.9453, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 2.2e-05, | |
| "loss": 4.8826, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 2.3e-05, | |
| "loss": 4.4724, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 2.4e-05, | |
| "loss": 3.7938, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 2.5e-05, | |
| "loss": 3.2653, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 2.6e-05, | |
| "loss": 3.3843, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 2.7e-05, | |
| "loss": 3.5122, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 2.8e-05, | |
| "loss": 3.4207, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 2.9000000000000004e-05, | |
| "loss": 3.4373, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 3e-05, | |
| "loss": 3.2939, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 3.1e-05, | |
| "loss": 3.2784, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 3.2e-05, | |
| "loss": 3.2101, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 3.3e-05, | |
| "loss": 3.1498, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 3.2037, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 3.5000000000000004e-05, | |
| "loss": 3.1997, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 3.6e-05, | |
| "loss": 3.1802, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.7e-05, | |
| "loss": 3.1123, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.8e-05, | |
| "loss": 3.1104, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.9e-05, | |
| "loss": 3.1473, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 4e-05, | |
| "loss": 3.1615, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 4.1e-05, | |
| "loss": 3.1428, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 4.2000000000000004e-05, | |
| "loss": 3.0819, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 4.2999999999999995e-05, | |
| "loss": 3.1113, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 4.4e-05, | |
| "loss": 3.0946, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 4.4999999999999996e-05, | |
| "loss": 3.0418, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 4.6e-05, | |
| "loss": 3.0446, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 4.7000000000000004e-05, | |
| "loss": 3.0657, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 4.8e-05, | |
| "loss": 3.079, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 4.9000000000000005e-05, | |
| "loss": 3.0337, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 5e-05, | |
| "loss": 2.9961, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 5.1e-05, | |
| "loss": 3.0675, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 5.2e-05, | |
| "loss": 3.2045, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 5.3e-05, | |
| "loss": 3.1336, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5.4e-05, | |
| "loss": 2.8979, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 5.5e-05, | |
| "loss": 3.0432, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 5.6e-05, | |
| "loss": 3.1173, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 5.7e-05, | |
| "loss": 2.8984, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 5.800000000000001e-05, | |
| "loss": 3.0313, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 5.9e-05, | |
| "loss": 3.0274, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 6e-05, | |
| "loss": 2.8286, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 6.1e-05, | |
| "loss": 2.7637, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 6.2e-05, | |
| "loss": 2.7111, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 6.3e-05, | |
| "loss": 2.6568, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 6.4e-05, | |
| "loss": 2.5761, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 2.594, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 6.6e-05, | |
| "loss": 2.5337, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 6.7e-05, | |
| "loss": 2.328, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 2.2416, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 6.900000000000001e-05, | |
| "loss": 2.1625, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 7.000000000000001e-05, | |
| "loss": 2.0447, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 7.099999999999999e-05, | |
| "loss": 1.8691, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.8868, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 7.3e-05, | |
| "loss": 1.6314, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 7.4e-05, | |
| "loss": 1.6226, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.7013, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 7.6e-05, | |
| "loss": 1.5557, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 7.7e-05, | |
| "loss": 1.4843, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 7.8e-05, | |
| "loss": 1.563, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 7.9e-05, | |
| "loss": 1.3356, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8e-05, | |
| "loss": 1.3477, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 8.1e-05, | |
| "loss": 1.2732, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 8.2e-05, | |
| "loss": 1.1838, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 8.300000000000001e-05, | |
| "loss": 1.1471, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 8.400000000000001e-05, | |
| "loss": 1.0627, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.9847, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 8.599999999999999e-05, | |
| "loss": 0.9781, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 8.7e-05, | |
| "loss": 0.96, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 8.8e-05, | |
| "loss": 0.9061, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 8.9e-05, | |
| "loss": 0.9697, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 8.999999999999999e-05, | |
| "loss": 1.0049, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 9.1e-05, | |
| "loss": 1.1034, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 9.2e-05, | |
| "loss": 0.9069, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 9.3e-05, | |
| "loss": 1.026, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 9.400000000000001e-05, | |
| "loss": 1.0734, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.9626, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.8731, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.7e-05, | |
| "loss": 1.0878, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 9.800000000000001e-05, | |
| "loss": 0.9423, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 0.9755, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 0.0001, | |
| "loss": 0.9086, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 0.000101, | |
| "loss": 0.8551, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 0.000102, | |
| "loss": 1.0771, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 0.000103, | |
| "loss": 0.9983, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 0.000104, | |
| "loss": 0.9869, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 0.000105, | |
| "loss": 1.1931, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.000106, | |
| "loss": 1.1017, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 109, | |
| "total_flos": 1.3366257233080689e+18, | |
| "train_loss": 4.048810622560868, | |
| "train_runtime": 250.3759, | |
| "train_samples_per_second": 13.891, | |
| "train_steps_per_second": 0.435 | |
| } | |
| ], | |
| "max_steps": 109, | |
| "num_train_epochs": 1, | |
| "total_flos": 1.3366257233080689e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |