| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.911894273127754, | |
| "eval_steps": 500, | |
| "global_step": 9000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 4.7246696035242295e-05, | |
| "loss": 2.9573, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 17.2961, | |
| "eval_gen_len": 53.6889, | |
| "eval_loss": 2.6540744304656982, | |
| "eval_runtime": 49.9025, | |
| "eval_samples_per_second": 4.509, | |
| "eval_steps_per_second": 0.581, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 4.449339207048459e-05, | |
| "loss": 2.927, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 4.1740088105726874e-05, | |
| "loss": 2.7809, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 19.3527, | |
| "eval_gen_len": 42.6667, | |
| "eval_loss": 2.5852272510528564, | |
| "eval_runtime": 37.69, | |
| "eval_samples_per_second": 5.97, | |
| "eval_steps_per_second": 0.769, | |
| "step": 1816 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 3.898678414096916e-05, | |
| "loss": 2.7165, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "learning_rate": 3.623348017621145e-05, | |
| "loss": 2.6527, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 18.53, | |
| "eval_gen_len": 44.8044, | |
| "eval_loss": 2.521505355834961, | |
| "eval_runtime": 44.997, | |
| "eval_samples_per_second": 5.0, | |
| "eval_steps_per_second": 0.644, | |
| "step": 2724 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "learning_rate": 3.3480176211453745e-05, | |
| "loss": 2.5517, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "learning_rate": 3.072687224669604e-05, | |
| "loss": 2.5069, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 21.0151, | |
| "eval_gen_len": 39.6889, | |
| "eval_loss": 2.480680227279663, | |
| "eval_runtime": 35.165, | |
| "eval_samples_per_second": 6.398, | |
| "eval_steps_per_second": 0.825, | |
| "step": 3632 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "learning_rate": 2.7973568281938327e-05, | |
| "loss": 2.4257, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "learning_rate": 2.522026431718062e-05, | |
| "loss": 2.3976, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_bleu": 21.4466, | |
| "eval_gen_len": 39.8667, | |
| "eval_loss": 2.4499943256378174, | |
| "eval_runtime": 31.2343, | |
| "eval_samples_per_second": 7.204, | |
| "eval_steps_per_second": 0.928, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 5.51, | |
| "learning_rate": 2.246696035242291e-05, | |
| "loss": 2.3175, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_bleu": 20.5303, | |
| "eval_gen_len": 42.0044, | |
| "eval_loss": 2.4212148189544678, | |
| "eval_runtime": 35.5561, | |
| "eval_samples_per_second": 6.328, | |
| "eval_steps_per_second": 0.816, | |
| "step": 5448 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "learning_rate": 1.97136563876652e-05, | |
| "loss": 2.3049, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 6.61, | |
| "learning_rate": 1.696035242290749e-05, | |
| "loss": 2.2268, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_bleu": 20.9986, | |
| "eval_gen_len": 42.8578, | |
| "eval_loss": 2.4010252952575684, | |
| "eval_runtime": 35.0954, | |
| "eval_samples_per_second": 6.411, | |
| "eval_steps_per_second": 0.826, | |
| "step": 6356 | |
| }, | |
| { | |
| "epoch": 7.16, | |
| "learning_rate": 1.420704845814978e-05, | |
| "loss": 2.2282, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 7.71, | |
| "learning_rate": 1.1453744493392071e-05, | |
| "loss": 2.1703, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_bleu": 22.8201, | |
| "eval_gen_len": 36.8267, | |
| "eval_loss": 2.3867766857147217, | |
| "eval_runtime": 23.8243, | |
| "eval_samples_per_second": 9.444, | |
| "eval_steps_per_second": 1.217, | |
| "step": 7264 | |
| }, | |
| { | |
| "epoch": 8.26, | |
| "learning_rate": 8.700440528634362e-06, | |
| "loss": 2.1589, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 8.81, | |
| "learning_rate": 5.947136563876652e-06, | |
| "loss": 2.1223, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_bleu": 22.9591, | |
| "eval_gen_len": 37.3422, | |
| "eval_loss": 2.3790900707244873, | |
| "eval_runtime": 28.3931, | |
| "eval_samples_per_second": 7.924, | |
| "eval_steps_per_second": 1.021, | |
| "step": 8172 | |
| }, | |
| { | |
| "epoch": 9.36, | |
| "learning_rate": 3.193832599118943e-06, | |
| "loss": 2.0995, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 9.91, | |
| "learning_rate": 4.4052863436123357e-07, | |
| "loss": 2.1029, | |
| "step": 9000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 9080, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 2122263810146304.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |