| { | |
| "best_metric": 2.237640380859375, | |
| "best_model_checkpoint": "/exports/eddie/scratch/s1970716/models/summarization/longt5_xl_summ_screen_bp_only_30/checkpoint-14", | |
| "epoch": 10.991304347826087, | |
| "eval_steps": 500, | |
| "global_step": 158, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3076, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3217, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3519, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3449, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3197, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3342, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 0.0005, | |
| "loss": 0.324, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "eval_gen_len": 246.7396449704142, | |
| "eval_loss": 2.237640380859375, | |
| "eval_rouge1": 40.4388, | |
| "eval_rouge2": 16.4662, | |
| "eval_rougeL": 28.0771, | |
| "eval_rougeLsum": 38.3405, | |
| "eval_runtime": 1709.2671, | |
| "eval_samples_per_second": 0.198, | |
| "eval_steps_per_second": 0.025, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2825, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2835, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2732, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2928, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2912, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2914, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2707, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "eval_gen_len": 307.3786982248521, | |
| "eval_loss": 2.320437431335449, | |
| "eval_rouge1": 40.2873, | |
| "eval_rouge2": 16.7641, | |
| "eval_rougeL": 27.3895, | |
| "eval_rougeLsum": 38.2689, | |
| "eval_runtime": 1775.692, | |
| "eval_samples_per_second": 0.19, | |
| "eval_steps_per_second": 0.024, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2342, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2326, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2334, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2212, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2161, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2236, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2217, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "eval_gen_len": 501.93195266272187, | |
| "eval_loss": 2.5281190872192383, | |
| "eval_rouge1": 31.9916, | |
| "eval_rouge2": 13.8136, | |
| "eval_rougeL": 22.1895, | |
| "eval_rougeLsum": 30.623, | |
| "eval_runtime": 1780.5368, | |
| "eval_samples_per_second": 0.19, | |
| "eval_steps_per_second": 0.024, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2055, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1943, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1875, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1909, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1881, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1541, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1776, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "eval_gen_len": 489.6183431952663, | |
| "eval_loss": 2.7530014514923096, | |
| "eval_rouge1": 31.7535, | |
| "eval_rouge2": 13.8852, | |
| "eval_rougeL": 22.8653, | |
| "eval_rougeLsum": 30.3796, | |
| "eval_runtime": 1780.5033, | |
| "eval_samples_per_second": 0.19, | |
| "eval_steps_per_second": 0.024, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1876, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1589, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1529, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1464, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1689, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1492, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1424, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "eval_gen_len": 502.11242603550295, | |
| "eval_loss": 2.657783031463623, | |
| "eval_rouge1": 32.117, | |
| "eval_rouge2": 14.2141, | |
| "eval_rougeL": 22.3733, | |
| "eval_rougeLsum": 30.8328, | |
| "eval_runtime": 1780.0926, | |
| "eval_samples_per_second": 0.19, | |
| "eval_steps_per_second": 0.024, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1494, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 5.15, | |
| "learning_rate": 0.0005, | |
| "loss": 0.118, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1499, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1369, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1084, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "learning_rate": 0.0005, | |
| "loss": 0.117, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "learning_rate": 0.0005, | |
| "loss": 0.144, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1449, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "eval_gen_len": 357.31360946745565, | |
| "eval_loss": 2.5508346557617188, | |
| "eval_rouge1": 35.3448, | |
| "eval_rouge2": 13.8478, | |
| "eval_rougeL": 24.9044, | |
| "eval_rougeLsum": 33.6108, | |
| "eval_runtime": 1768.966, | |
| "eval_samples_per_second": 0.191, | |
| "eval_steps_per_second": 0.024, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1101, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 6.26, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0985, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1101, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 6.54, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1013, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 6.68, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1057, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1102, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1191, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "eval_gen_len": 408.86686390532543, | |
| "eval_loss": 3.162177324295044, | |
| "eval_rouge1": 37.2189, | |
| "eval_rouge2": 16.0076, | |
| "eval_rougeL": 25.7011, | |
| "eval_rougeLsum": 35.294, | |
| "eval_runtime": 1778.8704, | |
| "eval_samples_per_second": 0.19, | |
| "eval_steps_per_second": 0.024, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 7.23, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0949, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 7.37, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0974, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 7.51, | |
| "learning_rate": 0.0005, | |
| "loss": 0.096, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 7.65, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1023, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 7.79, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0829, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0879, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_gen_len": 318.2278106508876, | |
| "eval_loss": 2.8510310649871826, | |
| "eval_rouge1": 39.8825, | |
| "eval_rouge2": 16.8073, | |
| "eval_rougeL": 27.2428, | |
| "eval_rougeLsum": 37.9568, | |
| "eval_runtime": 1776.4341, | |
| "eval_samples_per_second": 0.19, | |
| "eval_steps_per_second": 0.024, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 8.07, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0957, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 8.21, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0991, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 8.35, | |
| "learning_rate": 0.0005, | |
| "loss": 0.095, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0999, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 8.63, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0787, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 8.77, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0716, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0899, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "eval_gen_len": 500.405325443787, | |
| "eval_loss": 2.9137675762176514, | |
| "eval_rouge1": 31.7139, | |
| "eval_rouge2": 13.7066, | |
| "eval_rougeL": 21.8844, | |
| "eval_rougeLsum": 30.5075, | |
| "eval_runtime": 1780.3113, | |
| "eval_samples_per_second": 0.19, | |
| "eval_steps_per_second": 0.024, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 9.04, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0945, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 9.18, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0597, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 9.32, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0586, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 9.46, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0757, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0665, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 9.74, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0594, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 9.88, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0656, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 9.95, | |
| "eval_gen_len": 488.1686390532544, | |
| "eval_loss": 3.1616334915161133, | |
| "eval_rouge1": 33.055, | |
| "eval_rouge2": 14.5841, | |
| "eval_rougeL": 22.5883, | |
| "eval_rougeLsum": 31.7565, | |
| "eval_runtime": 1782.2406, | |
| "eval_samples_per_second": 0.19, | |
| "eval_steps_per_second": 0.024, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0613, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 10.16, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0621, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 10.3, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0739, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 10.43, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0841, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 10.57, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0562, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 10.71, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0611, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 10.85, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0607, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 10.99, | |
| "learning_rate": 0.0005, | |
| "loss": 0.0542, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 10.99, | |
| "eval_gen_len": 198.80769230769232, | |
| "eval_loss": 3.3630056381225586, | |
| "eval_rouge1": 43.7514, | |
| "eval_rouge2": 18.9011, | |
| "eval_rougeL": 29.9017, | |
| "eval_rougeLsum": 41.6887, | |
| "eval_runtime": 1460.3937, | |
| "eval_samples_per_second": 0.231, | |
| "eval_steps_per_second": 0.029, | |
| "step": 158 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 210, | |
| "num_train_epochs": 15, | |
| "save_steps": 500, | |
| "total_flos": 2.764359930466935e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |