| { | |
| "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": 3.8742, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.00015, | |
| "loss": 1.2473, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.000225, | |
| "loss": 1.1998, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 0.0003, | |
| "loss": 1.1545, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 0.0002915206331260599, | |
| "loss": 1.1364, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_bleu": 0.17166972696083574, | |
| "eval_loss": 1.1582657098770142, | |
| "eval_rouge1": 0.5000074515653312, | |
| "eval_rouge2": 0.40559916777579735, | |
| "eval_rougeL": 0.46317379201759545, | |
| "eval_rougeLsum": 0.4632046447608197, | |
| "eval_runtime": 211.1462, | |
| "eval_samples_per_second": 22.226, | |
| "eval_steps_per_second": 2.78, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 0.0002830412662521198, | |
| "loss": 1.1312, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 0.0002745618993781798, | |
| "loss": 1.1164, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 0.00026608253250423966, | |
| "loss": 1.1037, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 0.0002576031656302996, | |
| "loss": 1.1084, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 0.00024912379875635947, | |
| "loss": 1.0874, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_bleu": 0.174359464623795, | |
| "eval_loss": 1.1108546257019043, | |
| "eval_rouge1": 0.5050901535930163, | |
| "eval_rouge2": 0.41029314654694715, | |
| "eval_rougeL": 0.4678610312798079, | |
| "eval_rougeLsum": 0.4679688863086813, | |
| "eval_runtime": 208.6731, | |
| "eval_samples_per_second": 22.49, | |
| "eval_steps_per_second": 2.813, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 0.0002406444318824194, | |
| "loss": 1.0758, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 0.00023216506500847936, | |
| "loss": 1.0899, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 0.00022368569813453927, | |
| "loss": 1.0728, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 0.00021520633126059918, | |
| "loss": 1.0751, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 0.0002067269643866591, | |
| "loss": 1.072, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_bleu": 0.16964536883217013, | |
| "eval_loss": 1.0908721685409546, | |
| "eval_rouge1": 0.49944919713548785, | |
| "eval_rouge2": 0.4041247246204984, | |
| "eval_rougeL": 0.4622247411282464, | |
| "eval_rougeLsum": 0.4622259005781699, | |
| "eval_runtime": 214.1194, | |
| "eval_samples_per_second": 21.918, | |
| "eval_steps_per_second": 2.741, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 0.00019824759751271905, | |
| "loss": 1.0601, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 0.00018976823063877895, | |
| "loss": 1.067, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 0.00018128886376483889, | |
| "loss": 1.0769, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 0.0001728094968908988, | |
| "loss": 1.0376, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 0.0001643301300169587, | |
| "loss": 1.0589, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_bleu": 0.17592370006590605, | |
| "eval_loss": 1.0714174509048462, | |
| "eval_rouge1": 0.5071117469721906, | |
| "eval_rouge2": 0.412824258369919, | |
| "eval_rougeL": 0.47019269708823, | |
| "eval_rougeLsum": 0.4702462880937972, | |
| "eval_runtime": 208.6946, | |
| "eval_samples_per_second": 22.487, | |
| "eval_steps_per_second": 2.813, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 0.00015585076314301866, | |
| "loss": 1.0589, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 0.00014737139626907857, | |
| "loss": 1.0593, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 0.00013889202939513847, | |
| "loss": 1.0599, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 0.0001304126625211984, | |
| "loss": 1.0335, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 0.00012193329564725833, | |
| "loss": 1.0441, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_bleu": 0.16827637557044198, | |
| "eval_loss": 1.0615065097808838, | |
| "eval_rouge1": 0.500508116585923, | |
| "eval_rouge2": 0.40524827339523467, | |
| "eval_rougeL": 0.46325896066513017, | |
| "eval_rougeLsum": 0.4633425344363471, | |
| "eval_runtime": 211.7269, | |
| "eval_samples_per_second": 22.165, | |
| "eval_steps_per_second": 2.772, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 0.00011345392877331824, | |
| "loss": 1.0273, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 0.00010497456189937817, | |
| "loss": 1.0371, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 9.649519502543809e-05, | |
| "loss": 1.0219, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 8.801582815149802e-05, | |
| "loss": 1.0292, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 7.953646127755793e-05, | |
| "loss": 1.0292, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_bleu": 0.17334929168529137, | |
| "eval_loss": 1.0501199960708618, | |
| "eval_rouge1": 0.5057510392904037, | |
| "eval_rouge2": 0.41127958350841676, | |
| "eval_rougeL": 0.46859672321186563, | |
| "eval_rougeLsum": 0.4686909635105035, | |
| "eval_runtime": 210.649, | |
| "eval_samples_per_second": 22.279, | |
| "eval_steps_per_second": 2.787, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 7.105709440361785e-05, | |
| "loss": 1.0166, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 6.257772752967779e-05, | |
| "loss": 1.015, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 5.40983606557377e-05, | |
| "loss": 1.0142, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 4.5618993781797626e-05, | |
| "loss": 1.0256, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 3.713962690785755e-05, | |
| "loss": 1.0094, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "eval_bleu": 0.1714081170400144, | |
| "eval_loss": 1.0423640012741089, | |
| "eval_rouge1": 0.5035879586288757, | |
| "eval_rouge2": 0.40851924259079697, | |
| "eval_rougeL": 0.46637902927069697, | |
| "eval_rougeLsum": 0.4663610962140706, | |
| "eval_runtime": 210.1257, | |
| "eval_samples_per_second": 22.334, | |
| "eval_steps_per_second": 2.794, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 2.8660260033917467e-05, | |
| "loss": 1.0099, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 2.0180893159977384e-05, | |
| "loss": 0.9959, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 1.1701526286037308e-05, | |
| "loss": 1.0107, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 3.22215941209723e-06, | |
| "loss": 1.0094, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 3938, | |
| "total_flos": 1.5343986185183232e+17, | |
| "train_loss": 1.1361048782216892, | |
| "train_runtime": 10695.8658, | |
| "train_samples_per_second": 23.566, | |
| "train_steps_per_second": 0.368 | |
| } | |
| ], | |
| "max_steps": 3938, | |
| "num_train_epochs": 1, | |
| "total_flos": 1.5343986185183232e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |