| { | |
| "best_metric": 41.3073, | |
| "best_model_checkpoint": "models/mt0-xl_english_adafactor/checkpoint-27400", | |
| "epoch": 20.0, | |
| "global_step": 27400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4.75e-05, | |
| "loss": 2.1171, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 11.879845515662995, | |
| "eval_loss": 1.8174540996551514, | |
| "eval_rouge1": 27.0261, | |
| "eval_rouge2": 8.6429, | |
| "eval_rougeL": 25.2826, | |
| "eval_rougeLsum": 25.2952, | |
| "eval_runtime": 356.3645, | |
| "eval_samples_per_second": 39.235, | |
| "eval_steps_per_second": 1.226, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.8186, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 11.791946788728366, | |
| "eval_loss": 1.7111847400665283, | |
| "eval_rouge1": 29.1583, | |
| "eval_rouge2": 9.9747, | |
| "eval_rougeL": 27.3432, | |
| "eval_rougeLsum": 27.3647, | |
| "eval_runtime": 305.4337, | |
| "eval_samples_per_second": 45.778, | |
| "eval_steps_per_second": 1.431, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 4.25e-05, | |
| "loss": 1.643, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 12.412530396223715, | |
| "eval_loss": 1.6442108154296875, | |
| "eval_rouge1": 30.9045, | |
| "eval_rouge2": 11.2256, | |
| "eval_rougeL": 28.7826, | |
| "eval_rougeLsum": 28.788, | |
| "eval_runtime": 308.7201, | |
| "eval_samples_per_second": 45.29, | |
| "eval_steps_per_second": 1.416, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 4e-05, | |
| "loss": 1.499, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 12.31290230296095, | |
| "eval_loss": 1.5977873802185059, | |
| "eval_rouge1": 32.1126, | |
| "eval_rouge2": 12.6674, | |
| "eval_rougeL": 29.97, | |
| "eval_rougeLsum": 29.9843, | |
| "eval_runtime": 307.939, | |
| "eval_samples_per_second": 45.405, | |
| "eval_steps_per_second": 1.419, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 1.3772, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 12.688742669146045, | |
| "eval_loss": 1.571955680847168, | |
| "eval_rouge1": 33.6113, | |
| "eval_rouge2": 13.8451, | |
| "eval_rougeL": 31.3468, | |
| "eval_rougeLsum": 31.3599, | |
| "eval_runtime": 320.2331, | |
| "eval_samples_per_second": 43.662, | |
| "eval_steps_per_second": 1.365, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.2742, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_gen_len": 12.200257473895007, | |
| "eval_loss": 1.556434988975525, | |
| "eval_rouge1": 34.4899, | |
| "eval_rouge2": 15.1005, | |
| "eval_rougeL": 32.3177, | |
| "eval_rougeLsum": 32.3291, | |
| "eval_runtime": 308.9469, | |
| "eval_samples_per_second": 45.257, | |
| "eval_steps_per_second": 1.414, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 1.1785, | |
| "step": 9590 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_gen_len": 12.4487197825776, | |
| "eval_loss": 1.5466481447219849, | |
| "eval_rouge1": 35.4729, | |
| "eval_rouge2": 16.2035, | |
| "eval_rougeL": 33.2166, | |
| "eval_rougeLsum": 33.2295, | |
| "eval_runtime": 305.713, | |
| "eval_samples_per_second": 45.736, | |
| "eval_steps_per_second": 1.429, | |
| "step": 9590 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0941, | |
| "step": 10960 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_gen_len": 12.754255471320269, | |
| "eval_loss": 1.5571105480194092, | |
| "eval_rouge1": 36.4885, | |
| "eval_rouge2": 17.5396, | |
| "eval_rougeL": 34.2494, | |
| "eval_rougeLsum": 34.2759, | |
| "eval_runtime": 308.476, | |
| "eval_samples_per_second": 45.326, | |
| "eval_steps_per_second": 1.417, | |
| "step": 10960 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 1.0202, | |
| "step": 12330 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_gen_len": 12.860320411958233, | |
| "eval_loss": 1.5540790557861328, | |
| "eval_rouge1": 37.4019, | |
| "eval_rouge2": 18.5568, | |
| "eval_rougeL": 35.1341, | |
| "eval_rougeLsum": 35.1473, | |
| "eval_runtime": 308.6475, | |
| "eval_samples_per_second": 45.301, | |
| "eval_steps_per_second": 1.416, | |
| "step": 12330 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.9552, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_gen_len": 12.698684022314405, | |
| "eval_loss": 1.5641505718231201, | |
| "eval_rouge1": 38.127, | |
| "eval_rouge2": 19.4057, | |
| "eval_rougeL": 35.9008, | |
| "eval_rougeLsum": 35.9163, | |
| "eval_runtime": 307.6813, | |
| "eval_samples_per_second": 45.443, | |
| "eval_steps_per_second": 1.42, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.8963, | |
| "step": 15070 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_gen_len": 12.705192390215991, | |
| "eval_loss": 1.5771721601486206, | |
| "eval_rouge1": 38.5073, | |
| "eval_rouge2": 20.0584, | |
| "eval_rougeL": 36.3304, | |
| "eval_rougeLsum": 36.3399, | |
| "eval_runtime": 480.883, | |
| "eval_samples_per_second": 29.076, | |
| "eval_steps_per_second": 0.909, | |
| "step": 15070 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 2e-05, | |
| "loss": 0.8443, | |
| "step": 16440 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_gen_len": 13.039479330567874, | |
| "eval_loss": 1.595460057258606, | |
| "eval_rouge1": 39.2323, | |
| "eval_rouge2": 20.9237, | |
| "eval_rougeL": 36.9863, | |
| "eval_rougeLsum": 37.0049, | |
| "eval_runtime": 314.6611, | |
| "eval_samples_per_second": 44.435, | |
| "eval_steps_per_second": 1.389, | |
| "step": 16440 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.7982, | |
| "step": 17810 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_gen_len": 13.140037190673723, | |
| "eval_loss": 1.608866810798645, | |
| "eval_rouge1": 39.7947, | |
| "eval_rouge2": 21.6422, | |
| "eval_rougeL": 37.5619, | |
| "eval_rougeLsum": 37.5815, | |
| "eval_runtime": 465.3407, | |
| "eval_samples_per_second": 30.047, | |
| "eval_steps_per_second": 0.939, | |
| "step": 17810 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.7586, | |
| "step": 19180 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_gen_len": 12.85889000143041, | |
| "eval_loss": 1.6293412446975708, | |
| "eval_rouge1": 40.2922, | |
| "eval_rouge2": 22.2301, | |
| "eval_rougeL": 38.0755, | |
| "eval_rougeLsum": 38.0757, | |
| "eval_runtime": 312.4581, | |
| "eval_samples_per_second": 44.748, | |
| "eval_steps_per_second": 1.399, | |
| "step": 19180 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.7234, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_gen_len": 13.110213131168646, | |
| "eval_loss": 1.6492763757705688, | |
| "eval_rouge1": 40.6358, | |
| "eval_rouge2": 22.5355, | |
| "eval_rougeL": 38.3523, | |
| "eval_rougeLsum": 38.3659, | |
| "eval_runtime": 307.5457, | |
| "eval_samples_per_second": 45.463, | |
| "eval_steps_per_second": 1.421, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6946, | |
| "step": 21920 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_gen_len": 13.103490201687885, | |
| "eval_loss": 1.6700669527053833, | |
| "eval_rouge1": 40.7708, | |
| "eval_rouge2": 22.906, | |
| "eval_rougeL": 38.5037, | |
| "eval_rougeLsum": 38.5174, | |
| "eval_runtime": 306.4464, | |
| "eval_samples_per_second": 45.626, | |
| "eval_steps_per_second": 1.426, | |
| "step": 21920 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.6688, | |
| "step": 23290 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_gen_len": 13.295093691889573, | |
| "eval_loss": 1.690183162689209, | |
| "eval_rouge1": 41.0847, | |
| "eval_rouge2": 23.1663, | |
| "eval_rougeL": 38.8126, | |
| "eval_rougeLsum": 38.8149, | |
| "eval_runtime": 309.6148, | |
| "eval_samples_per_second": 45.159, | |
| "eval_steps_per_second": 1.411, | |
| "step": 23290 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6484, | |
| "step": 24660 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_gen_len": 13.270705192390215, | |
| "eval_loss": 1.7005170583724976, | |
| "eval_rouge1": 41.2075, | |
| "eval_rouge2": 23.3967, | |
| "eval_rougeL": 38.9529, | |
| "eval_rougeLsum": 38.9545, | |
| "eval_runtime": 310.6615, | |
| "eval_samples_per_second": 45.007, | |
| "eval_steps_per_second": 1.407, | |
| "step": 24660 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.6342, | |
| "step": 26030 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_gen_len": 13.217279359176084, | |
| "eval_loss": 1.7115505933761597, | |
| "eval_rouge1": 41.2454, | |
| "eval_rouge2": 23.5187, | |
| "eval_rougeL": 39.0203, | |
| "eval_rougeLsum": 39.0396, | |
| "eval_runtime": 310.561, | |
| "eval_samples_per_second": 45.022, | |
| "eval_steps_per_second": 1.407, | |
| "step": 26030 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.6234, | |
| "step": 27400 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_gen_len": 13.255757402374481, | |
| "eval_loss": 1.7210286855697632, | |
| "eval_rouge1": 41.3073, | |
| "eval_rouge2": 23.5691, | |
| "eval_rougeL": 39.0662, | |
| "eval_rougeLsum": 39.074, | |
| "eval_runtime": 482.7425, | |
| "eval_samples_per_second": 28.964, | |
| "eval_steps_per_second": 0.905, | |
| "step": 27400 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 27400, | |
| "total_flos": 4.389256654374306e+18, | |
| "train_loss": 1.0633717825116902, | |
| "train_runtime": 55751.6068, | |
| "train_samples_per_second": 62.898, | |
| "train_steps_per_second": 0.491 | |
| } | |
| ], | |
| "max_steps": 27400, | |
| "num_train_epochs": 20, | |
| "total_flos": 4.389256654374306e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |