| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9998730481147645, | |
| "global_step": 3938, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 7.5e-05, | |
| "loss": 12.4009, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.00015, | |
| "loss": 1.7002, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.000225, | |
| "loss": 1.374, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 0.0003, | |
| "loss": 1.2745, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 0.0002915206331260599, | |
| "loss": 1.2333, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_bleu": 0.16985441338971674, | |
| "eval_loss": 1.1092561483383179, | |
| "eval_rouge1": 0.49744095525932, | |
| "eval_rouge2": 0.4023858560784097, | |
| "eval_rougeL": 0.4605319249625821, | |
| "eval_rougeLsum": 0.4606110961561354, | |
| "eval_runtime": 448.9761, | |
| "eval_samples_per_second": 10.453, | |
| "eval_steps_per_second": 1.307, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 0.0002830412662521198, | |
| "loss": 1.2179, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 0.0002745618993781798, | |
| "loss": 1.194, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 0.00026608253250423966, | |
| "loss": 1.1763, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 0.0002576031656302996, | |
| "loss": 1.1782, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 0.00024912379875635947, | |
| "loss": 1.1517, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_bleu": 0.167245118666087, | |
| "eval_loss": 1.0618804693222046, | |
| "eval_rouge1": 0.4974694475768039, | |
| "eval_rouge2": 0.40136842689484065, | |
| "eval_rougeL": 0.46015392715352893, | |
| "eval_rougeLsum": 0.4601809444739556, | |
| "eval_runtime": 441.1296, | |
| "eval_samples_per_second": 10.639, | |
| "eval_steps_per_second": 1.331, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 0.0002406444318824194, | |
| "loss": 1.136, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 0.00023216506500847936, | |
| "loss": 1.1522, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 0.00022368569813453927, | |
| "loss": 1.1346, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 0.00021520633126059918, | |
| "loss": 1.1307, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 0.0002067269643866591, | |
| "loss": 1.1281, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_bleu": 0.16267141788045325, | |
| "eval_loss": 1.0443658828735352, | |
| "eval_rouge1": 0.4920999084850846, | |
| "eval_rouge2": 0.3953067823623436, | |
| "eval_rougeL": 0.45460007159317317, | |
| "eval_rougeLsum": 0.4546306350106014, | |
| "eval_runtime": 454.009, | |
| "eval_samples_per_second": 10.337, | |
| "eval_steps_per_second": 1.293, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 0.00019824759751271905, | |
| "loss": 1.1131, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 0.00018976823063877895, | |
| "loss": 1.1105, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 0.00018128886376483889, | |
| "loss": 1.1267, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 0.0001728094968908988, | |
| "loss": 1.0872, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 0.0001643301300169587, | |
| "loss": 1.1054, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_bleu": 0.17300330032744055, | |
| "eval_loss": 1.0234187841415405, | |
| "eval_rouge1": 0.503801949569296, | |
| "eval_rouge2": 0.40895933412262137, | |
| "eval_rougeL": 0.466809748089249, | |
| "eval_rougeLsum": 0.4667488466486195, | |
| "eval_runtime": 443.5894, | |
| "eval_samples_per_second": 10.58, | |
| "eval_steps_per_second": 1.323, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 0.00015585076314301866, | |
| "loss": 1.1017, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 0.00014737139626907857, | |
| "loss": 1.1064, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 0.00013889202939513847, | |
| "loss": 1.1047, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 0.0001304126625211984, | |
| "loss": 1.0758, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 0.00012193329564725833, | |
| "loss": 1.0848, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_bleu": 0.16666637884945387, | |
| "eval_loss": 1.0130037069320679, | |
| "eval_rouge1": 0.49905053867339855, | |
| "eval_rouge2": 0.40300133873229244, | |
| "eval_rougeL": 0.46173781984585727, | |
| "eval_rougeLsum": 0.4617687754853303, | |
| "eval_runtime": 442.0228, | |
| "eval_samples_per_second": 10.617, | |
| "eval_steps_per_second": 1.328, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 0.00011345392877331824, | |
| "loss": 1.0746, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 0.00010497456189937817, | |
| "loss": 1.0777, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 9.649519502543809e-05, | |
| "loss": 1.0644, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 8.801582815149802e-05, | |
| "loss": 1.0662, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 7.953646127755793e-05, | |
| "loss": 1.0737, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_bleu": 0.17037069563841947, | |
| "eval_loss": 1.002275824546814, | |
| "eval_rouge1": 0.5031038227519693, | |
| "eval_rouge2": 0.40777366696302797, | |
| "eval_rougeL": 0.4660246892337898, | |
| "eval_rougeLsum": 0.4660288890427905, | |
| "eval_runtime": 442.0838, | |
| "eval_samples_per_second": 10.616, | |
| "eval_steps_per_second": 1.328, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 7.105709440361785e-05, | |
| "loss": 1.0639, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 6.257772752967779e-05, | |
| "loss": 1.058, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 5.40983606557377e-05, | |
| "loss": 1.0598, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 4.5618993781797626e-05, | |
| "loss": 1.066, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 3.713962690785755e-05, | |
| "loss": 1.0532, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "eval_bleu": 0.16910393271412671, | |
| "eval_loss": 0.995514452457428, | |
| "eval_rouge1": 0.5013148071463731, | |
| "eval_rouge2": 0.405633242849629, | |
| "eval_rougeL": 0.4639778817773035, | |
| "eval_rougeLsum": 0.464049641741316, | |
| "eval_runtime": 441.9704, | |
| "eval_samples_per_second": 10.618, | |
| "eval_steps_per_second": 1.328, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 2.8660260033917467e-05, | |
| "loss": 1.0503, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 2.0180893159977384e-05, | |
| "loss": 1.0415, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 1.1701526286037308e-05, | |
| "loss": 1.0498, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 3.22215941209723e-06, | |
| "loss": 1.0493, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 3938, | |
| "total_flos": 5.10811297256448e+17, | |
| "train_loss": 1.4182514357288336, | |
| "train_runtime": 30065.8698, | |
| "train_samples_per_second": 8.384, | |
| "train_steps_per_second": 0.131 | |
| } | |
| ], | |
| "max_steps": 3938, | |
| "num_train_epochs": 1, | |
| "total_flos": 5.10811297256448e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |