| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.0, | |
| "eval_steps": 500, | |
| "global_step": 14016, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.821632420091324e-05, | |
| "loss": 0.3322, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 4.6432648401826485e-05, | |
| "loss": 0.2664, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 4.464897260273973e-05, | |
| "loss": 0.2412, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 4.286529680365297e-05, | |
| "loss": 0.2219, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 21.2068, | |
| "eval_gen_len": 19.2422, | |
| "eval_loss": 0.18159246444702148, | |
| "eval_runtime": 154.1038, | |
| "eval_samples_per_second": 53.886, | |
| "eval_steps_per_second": 1.687, | |
| "step": 2336 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 4.108162100456621e-05, | |
| "loss": 0.1887, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 3.929794520547945e-05, | |
| "loss": 0.1575, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 3.7514269406392696e-05, | |
| "loss": 0.1534, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 3.573059360730594e-05, | |
| "loss": 0.1553, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 3.394691780821918e-05, | |
| "loss": 0.1479, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 21.6783, | |
| "eval_gen_len": 19.2471, | |
| "eval_loss": 0.16088075935840607, | |
| "eval_runtime": 154.8749, | |
| "eval_samples_per_second": 53.617, | |
| "eval_steps_per_second": 1.679, | |
| "step": 4672 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 3.2163242009132423e-05, | |
| "loss": 0.1249, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 3.037956621004566e-05, | |
| "loss": 0.1117, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 2.8595890410958903e-05, | |
| "loss": 0.1119, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 2.681221461187215e-05, | |
| "loss": 0.1137, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 2.502853881278539e-05, | |
| "loss": 0.1092, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 21.8526, | |
| "eval_gen_len": 19.2463, | |
| "eval_loss": 0.15343397855758667, | |
| "eval_runtime": 154.8601, | |
| "eval_samples_per_second": 53.623, | |
| "eval_steps_per_second": 1.679, | |
| "step": 7008 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 2.324486301369863e-05, | |
| "loss": 0.0842, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "learning_rate": 2.1461187214611872e-05, | |
| "loss": 0.085, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "learning_rate": 1.9677511415525117e-05, | |
| "loss": 0.0853, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "learning_rate": 1.7893835616438355e-05, | |
| "loss": 0.0856, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 22.0841, | |
| "eval_gen_len": 19.2482, | |
| "eval_loss": 0.15252342820167542, | |
| "eval_runtime": 155.0415, | |
| "eval_samples_per_second": 53.56, | |
| "eval_steps_per_second": 1.677, | |
| "step": 9344 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 1.61101598173516e-05, | |
| "loss": 0.0792, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.28, | |
| "learning_rate": 1.4326484018264841e-05, | |
| "loss": 0.0664, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "learning_rate": 1.2542808219178081e-05, | |
| "loss": 0.0673, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "learning_rate": 1.0759132420091326e-05, | |
| "loss": 0.0663, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "learning_rate": 8.975456621004565e-06, | |
| "loss": 0.0667, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_bleu": 22.1943, | |
| "eval_gen_len": 19.2467, | |
| "eval_loss": 0.15876752138137817, | |
| "eval_runtime": 155.4094, | |
| "eval_samples_per_second": 53.433, | |
| "eval_steps_per_second": 1.673, | |
| "step": 11680 | |
| }, | |
| { | |
| "epoch": 5.14, | |
| "learning_rate": 7.191780821917809e-06, | |
| "loss": 0.0579, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.35, | |
| "learning_rate": 5.40810502283105e-06, | |
| "loss": 0.0545, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "learning_rate": 3.6244292237442927e-06, | |
| "loss": 0.0543, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.78, | |
| "learning_rate": 1.8407534246575344e-06, | |
| "loss": 0.0534, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.99, | |
| "learning_rate": 5.7077625570776255e-08, | |
| "loss": 0.0549, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_bleu": 22.2237, | |
| "eval_gen_len": 19.2467, | |
| "eval_loss": 0.16120968759059906, | |
| "eval_runtime": 154.8857, | |
| "eval_samples_per_second": 53.614, | |
| "eval_steps_per_second": 1.679, | |
| "step": 14016 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 14016, | |
| "total_flos": 3.006237250179072e+16, | |
| "train_loss": 0.12124856073222204, | |
| "train_runtime": 4033.4181, | |
| "train_samples_per_second": 111.169, | |
| "train_steps_per_second": 3.475 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 14016, | |
| "num_train_epochs": 6, | |
| "save_steps": 500, | |
| "total_flos": 3.006237250179072e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |