| { | |
| "best_metric": 3.1917154788970947, | |
| "best_model_checkpoint": "/exports/eddie/scratch/s1970716/models/summarization/longt5_xl_summ_screen_20/checkpoint-28", | |
| "epoch": 6.956521739130435, | |
| "eval_steps": 500, | |
| "global_step": 100, | |
| "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": 0.411, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 0.001, | |
| "loss": 0.4642, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 0.001, | |
| "loss": 0.4965, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 0.001, | |
| "loss": 0.7455, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 0.001, | |
| "loss": 0.4501, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 0.001, | |
| "loss": 0.3804, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 0.001, | |
| "loss": 0.4063, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "eval_gen_len": 71.90828402366864, | |
| "eval_loss": 3.7384819984436035, | |
| "eval_rouge1": 27.9171, | |
| "eval_rouge2": 6.7215, | |
| "eval_rougeL": 17.9315, | |
| "eval_rougeLsum": 24.363, | |
| "eval_runtime": 823.0541, | |
| "eval_samples_per_second": 0.411, | |
| "eval_steps_per_second": 0.205, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 0.001, | |
| "loss": 0.3201, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 0.001, | |
| "loss": 0.3253, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 0.001, | |
| "loss": 0.3215, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 0.001, | |
| "loss": 0.3175, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 0.001, | |
| "loss": 0.3331, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 0.001, | |
| "loss": 0.2811, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 0.001, | |
| "loss": 0.3125, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "eval_gen_len": 96.20414201183432, | |
| "eval_loss": 3.1917154788970947, | |
| "eval_rouge1": 28.1708, | |
| "eval_rouge2": 6.6895, | |
| "eval_rougeL": 18.1637, | |
| "eval_rougeLsum": 24.3987, | |
| "eval_runtime": 1069.4844, | |
| "eval_samples_per_second": 0.316, | |
| "eval_steps_per_second": 0.158, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 0.001, | |
| "loss": 0.2621, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 0.001, | |
| "loss": 0.2194, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 0.001, | |
| "loss": 0.2386, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 0.001, | |
| "loss": 0.2264, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 0.001, | |
| "loss": 0.2002, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 0.001, | |
| "loss": 0.2477, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 0.001, | |
| "loss": 0.2177, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "eval_gen_len": 198.0473372781065, | |
| "eval_loss": 3.9997544288635254, | |
| "eval_rouge1": 29.3167, | |
| "eval_rouge2": 5.9, | |
| "eval_rougeL": 17.3608, | |
| "eval_rougeLsum": 25.6945, | |
| "eval_runtime": 1900.1301, | |
| "eval_samples_per_second": 0.178, | |
| "eval_steps_per_second": 0.089, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 0.001, | |
| "loss": 0.2069, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 0.001, | |
| "loss": 0.164, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "learning_rate": 0.001, | |
| "loss": 0.1679, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 0.001, | |
| "loss": 0.1736, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "learning_rate": 0.001, | |
| "loss": 0.1688, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "learning_rate": 0.001, | |
| "loss": 0.1749, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "learning_rate": 0.001, | |
| "loss": 0.1753, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "eval_gen_len": 158.6508875739645, | |
| "eval_loss": 4.228714466094971, | |
| "eval_rouge1": 29.0605, | |
| "eval_rouge2": 6.2534, | |
| "eval_rougeL": 17.5744, | |
| "eval_rougeLsum": 25.6415, | |
| "eval_runtime": 1492.9623, | |
| "eval_samples_per_second": 0.226, | |
| "eval_steps_per_second": 0.113, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "learning_rate": 0.001, | |
| "loss": 0.1656, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 0.001, | |
| "loss": 0.1144, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "learning_rate": 0.001, | |
| "loss": 0.161, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "learning_rate": 0.001, | |
| "loss": 0.2169, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "learning_rate": 0.001, | |
| "loss": 0.1943, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "learning_rate": 0.001, | |
| "loss": 0.1777, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "learning_rate": 0.001, | |
| "loss": 0.2747, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "eval_gen_len": 118.44378698224853, | |
| "eval_loss": 4.102721214294434, | |
| "eval_rouge1": 31.2245, | |
| "eval_rouge2": 6.5663, | |
| "eval_rougeL": 18.1588, | |
| "eval_rougeLsum": 26.8996, | |
| "eval_runtime": 1188.6007, | |
| "eval_samples_per_second": 0.284, | |
| "eval_steps_per_second": 0.142, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "learning_rate": 0.001, | |
| "loss": 0.1399, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 5.15, | |
| "learning_rate": 0.001, | |
| "loss": 0.0986, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "learning_rate": 0.001, | |
| "loss": 0.1051, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "learning_rate": 0.001, | |
| "loss": 0.1288, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "learning_rate": 0.001, | |
| "loss": 0.1097, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "learning_rate": 0.001, | |
| "loss": 0.1163, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "learning_rate": 0.001, | |
| "loss": 0.1205, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "learning_rate": 0.001, | |
| "loss": 0.1045, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "eval_gen_len": 92.98816568047337, | |
| "eval_loss": 5.058135986328125, | |
| "eval_rouge1": 30.6056, | |
| "eval_rouge2": 6.8892, | |
| "eval_rougeL": 18.4933, | |
| "eval_rougeLsum": 26.4027, | |
| "eval_runtime": 984.3965, | |
| "eval_samples_per_second": 0.343, | |
| "eval_steps_per_second": 0.172, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "learning_rate": 0.001, | |
| "loss": 0.0767, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 6.26, | |
| "learning_rate": 0.001, | |
| "loss": 0.0678, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 0.001, | |
| "loss": 0.0759, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 6.54, | |
| "learning_rate": 0.001, | |
| "loss": 0.0714, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 6.68, | |
| "learning_rate": 0.001, | |
| "loss": 0.0822, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "learning_rate": 0.001, | |
| "loss": 0.0843, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "learning_rate": 0.001, | |
| "loss": 0.0875, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "eval_gen_len": 160.89644970414201, | |
| "eval_loss": 4.59414529800415, | |
| "eval_rouge1": 32.5234, | |
| "eval_rouge2": 7.3736, | |
| "eval_rougeL": 18.8958, | |
| "eval_rougeLsum": 28.4738, | |
| "eval_runtime": 1504.7392, | |
| "eval_samples_per_second": 0.225, | |
| "eval_steps_per_second": 0.112, | |
| "step": 100 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 140, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 1.7591398064550052e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |