| { | |
| "best_metric": 2.681736707687378, | |
| "best_model_checkpoint": "longt5_xl_summ_screen_memsum_bp_30/checkpoint-100", | |
| "epoch": 9.73913043478261, | |
| "eval_steps": 500, | |
| "global_step": 140, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 0.001, | |
| "loss": 0.0739, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 0.001, | |
| "loss": 0.0732, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 0.001, | |
| "loss": 0.0657, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 0.001, | |
| "loss": 0.0699, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 0.001, | |
| "loss": 0.0667, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 0.001, | |
| "loss": 0.0705, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 0.001, | |
| "loss": 0.0707, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "eval_gen_len": 453.5295857988166, | |
| "eval_loss": 2.709702730178833, | |
| "eval_rouge1": 41.4751, | |
| "eval_rouge2": 15.5831, | |
| "eval_rougeL": 25.1976, | |
| "eval_rougeLsum": 39.9229, | |
| "eval_runtime": 757.2273, | |
| "eval_samples_per_second": 0.446, | |
| "eval_steps_per_second": 0.057, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 0.001, | |
| "loss": 0.0649, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 0.001, | |
| "loss": 0.0535, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 0.001, | |
| "loss": 0.0715, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 0.001, | |
| "loss": 0.0731, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 0.001, | |
| "loss": 0.0931, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 0.001, | |
| "loss": 0.0583, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 0.001, | |
| "loss": 0.0608, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "eval_gen_len": 387.41715976331363, | |
| "eval_loss": 2.7271180152893066, | |
| "eval_rouge1": 45.691, | |
| "eval_rouge2": 17.905, | |
| "eval_rougeL": 27.9519, | |
| "eval_rougeLsum": 43.8787, | |
| "eval_runtime": 765.753, | |
| "eval_samples_per_second": 0.441, | |
| "eval_steps_per_second": 0.056, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 0.001, | |
| "loss": 0.0521, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 0.001, | |
| "loss": 0.0396, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 0.001, | |
| "loss": 0.0505, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 0.001, | |
| "loss": 0.0496, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 0.001, | |
| "loss": 0.0591, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 0.001, | |
| "loss": 0.061, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 0.001, | |
| "loss": 0.0851, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "eval_gen_len": 261.56804733727813, | |
| "eval_loss": 3.0001442432403564, | |
| "eval_rouge1": 47.1647, | |
| "eval_rouge2": 17.8993, | |
| "eval_rougeL": 28.7561, | |
| "eval_rougeLsum": 45.661, | |
| "eval_runtime": 691.7097, | |
| "eval_samples_per_second": 0.489, | |
| "eval_steps_per_second": 0.062, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 0.001, | |
| "loss": 0.0519, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 0.001, | |
| "loss": 0.0481, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "learning_rate": 0.001, | |
| "loss": 0.0535, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 0.001, | |
| "loss": 0.0404, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "learning_rate": 0.001, | |
| "loss": 0.0468, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "learning_rate": 0.001, | |
| "loss": 0.0535, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "learning_rate": 0.001, | |
| "loss": 0.0697, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "eval_gen_len": 365.3047337278106, | |
| "eval_loss": 2.9297378063201904, | |
| "eval_rouge1": 46.6892, | |
| "eval_rouge2": 17.8922, | |
| "eval_rougeL": 28.0724, | |
| "eval_rougeLsum": 44.8821, | |
| "eval_runtime": 770.7359, | |
| "eval_samples_per_second": 0.439, | |
| "eval_steps_per_second": 0.056, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "learning_rate": 0.001, | |
| "loss": 0.0371, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 0.001, | |
| "loss": 0.0306, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "learning_rate": 0.001, | |
| "loss": 0.0285, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "learning_rate": 0.001, | |
| "loss": 0.033, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "learning_rate": 0.001, | |
| "loss": 0.0411, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "learning_rate": 0.001, | |
| "loss": 0.0326, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "learning_rate": 0.001, | |
| "loss": 0.0296, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "eval_gen_len": 440.6390532544379, | |
| "eval_loss": 2.901674270629883, | |
| "eval_rouge1": 44.2702, | |
| "eval_rouge2": 17.7874, | |
| "eval_rougeL": 26.7598, | |
| "eval_rougeLsum": 42.6857, | |
| "eval_runtime": 764.3617, | |
| "eval_samples_per_second": 0.442, | |
| "eval_steps_per_second": 0.056, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "learning_rate": 0.001, | |
| "loss": 0.0384, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 5.15, | |
| "learning_rate": 0.001, | |
| "loss": 0.0429, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "learning_rate": 0.001, | |
| "loss": 0.0459, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "learning_rate": 0.001, | |
| "loss": 0.0336, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "learning_rate": 0.001, | |
| "loss": 0.0406, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "learning_rate": 0.001, | |
| "loss": 0.0361, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "learning_rate": 0.001, | |
| "loss": 0.0394, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "learning_rate": 0.001, | |
| "loss": 0.0312, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "eval_gen_len": 306.6715976331361, | |
| "eval_loss": 3.04889178276062, | |
| "eval_rouge1": 47.7884, | |
| "eval_rouge2": 18.1788, | |
| "eval_rougeL": 28.6688, | |
| "eval_rougeLsum": 46.0744, | |
| "eval_runtime": 746.8476, | |
| "eval_samples_per_second": 0.453, | |
| "eval_steps_per_second": 0.058, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "learning_rate": 0.001, | |
| "loss": 0.0276, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 6.26, | |
| "learning_rate": 0.001, | |
| "loss": 0.0338, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 0.001, | |
| "loss": 0.0487, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 6.54, | |
| "learning_rate": 0.001, | |
| "loss": 0.0379, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 6.68, | |
| "learning_rate": 0.001, | |
| "loss": 0.0309, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "learning_rate": 0.001, | |
| "loss": 0.0366, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "learning_rate": 0.001, | |
| "loss": 0.0383, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "eval_gen_len": 308.9082840236686, | |
| "eval_loss": 2.681736707687378, | |
| "eval_rouge1": 47.1842, | |
| "eval_rouge2": 18.22, | |
| "eval_rougeL": 28.4626, | |
| "eval_rougeLsum": 45.5778, | |
| "eval_runtime": 764.2847, | |
| "eval_samples_per_second": 0.442, | |
| "eval_steps_per_second": 0.056, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "learning_rate": 0.001, | |
| "loss": 0.0474, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 7.23, | |
| "learning_rate": 0.001, | |
| "loss": 0.0376, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 7.37, | |
| "learning_rate": 0.001, | |
| "loss": 0.0535, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 7.51, | |
| "learning_rate": 0.001, | |
| "loss": 0.036, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 7.65, | |
| "learning_rate": 0.001, | |
| "loss": 0.0339, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 7.79, | |
| "learning_rate": 0.001, | |
| "loss": 0.0557, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "learning_rate": 0.001, | |
| "loss": 0.0367, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_gen_len": 227.8550295857988, | |
| "eval_loss": 3.0245203971862793, | |
| "eval_rouge1": 45.5573, | |
| "eval_rouge2": 17.2161, | |
| "eval_rougeL": 28.0573, | |
| "eval_rougeLsum": 43.7772, | |
| "eval_runtime": 668.0023, | |
| "eval_samples_per_second": 0.506, | |
| "eval_steps_per_second": 0.064, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 8.07, | |
| "learning_rate": 0.001, | |
| "loss": 0.033, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 8.21, | |
| "learning_rate": 0.001, | |
| "loss": 0.0379, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 8.35, | |
| "learning_rate": 0.001, | |
| "loss": 0.0274, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "learning_rate": 0.001, | |
| "loss": 0.026, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 8.63, | |
| "learning_rate": 0.001, | |
| "loss": 0.0269, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 8.77, | |
| "learning_rate": 0.001, | |
| "loss": 0.0314, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "learning_rate": 0.001, | |
| "loss": 0.04, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "eval_gen_len": 429.8757396449704, | |
| "eval_loss": 3.2872769832611084, | |
| "eval_rouge1": 44.0164, | |
| "eval_rouge2": 17.1682, | |
| "eval_rougeL": 26.4769, | |
| "eval_rougeLsum": 42.3752, | |
| "eval_runtime": 779.6058, | |
| "eval_samples_per_second": 0.434, | |
| "eval_steps_per_second": 0.055, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 9.04, | |
| "learning_rate": 0.001, | |
| "loss": 0.0288, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 9.18, | |
| "learning_rate": 0.001, | |
| "loss": 0.0252, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 9.32, | |
| "learning_rate": 0.001, | |
| "loss": 0.0262, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 9.46, | |
| "learning_rate": 0.001, | |
| "loss": 0.0297, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 0.001, | |
| "loss": 0.0286, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 9.74, | |
| "learning_rate": 0.001, | |
| "loss": 0.028, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 9.74, | |
| "eval_gen_len": 337.4822485207101, | |
| "eval_loss": 2.9815244674682617, | |
| "eval_rouge1": 46.6542, | |
| "eval_rouge2": 17.8515, | |
| "eval_rougeL": 28.146, | |
| "eval_rougeLsum": 45.0274, | |
| "eval_runtime": 765.5763, | |
| "eval_samples_per_second": 0.441, | |
| "eval_steps_per_second": 0.056, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 9.74, | |
| "step": 140, | |
| "total_flos": 2.7436549259892326e+17, | |
| "train_loss": 0.045875947869249756, | |
| "train_runtime": 15096.5898, | |
| "train_samples_per_second": 2.433, | |
| "train_steps_per_second": 0.009 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 140, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 2.7436549259892326e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |