| { | |
| "best_metric": 14.6014, | |
| "best_model_checkpoint": "./ko-zh/checkpoint-7500", | |
| "epoch": 7.2337575351640995, | |
| "eval_steps": 1500, | |
| "global_step": 13500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.0954, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 5e-05, | |
| "loss": 1.5161, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 4.966051059206953e-05, | |
| "loss": 1.3824, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_bleu": 11.4445, | |
| "eval_gen_len": 15.774, | |
| "eval_loss": 1.301085352897644, | |
| "eval_runtime": 894.2268, | |
| "eval_samples_per_second": 16.693, | |
| "eval_steps_per_second": 1.043, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 4.9321021184139056e-05, | |
| "loss": 1.245, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 4.898153177620858e-05, | |
| "loss": 1.0757, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 4.8642042368278115e-05, | |
| "loss": 1.0646, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "eval_bleu": 13.1811, | |
| "eval_gen_len": 15.6757, | |
| "eval_loss": 1.1915582418441772, | |
| "eval_runtime": 882.4536, | |
| "eval_samples_per_second": 16.915, | |
| "eval_steps_per_second": 1.057, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 4.830255296034764e-05, | |
| "loss": 1.051, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 4.796306355241716e-05, | |
| "loss": 0.9135, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "learning_rate": 4.7623574144486695e-05, | |
| "loss": 0.8071, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "eval_bleu": 14.1901, | |
| "eval_gen_len": 15.2832, | |
| "eval_loss": 1.186390995979309, | |
| "eval_runtime": 873.6287, | |
| "eval_samples_per_second": 17.086, | |
| "eval_steps_per_second": 1.068, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "learning_rate": 4.728408473655622e-05, | |
| "loss": 0.8236, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "learning_rate": 4.694459532862575e-05, | |
| "loss": 0.8252, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 4.6605105920695275e-05, | |
| "loss": 0.6496, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "eval_bleu": 14.3496, | |
| "eval_gen_len": 15.5238, | |
| "eval_loss": 1.1978733539581299, | |
| "eval_runtime": 857.9115, | |
| "eval_samples_per_second": 17.399, | |
| "eval_steps_per_second": 1.088, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 4.62656165127648e-05, | |
| "loss": 0.6263, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 4.5926127104834335e-05, | |
| "loss": 0.6425, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "learning_rate": 4.558663769690386e-05, | |
| "loss": 0.6365, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "eval_bleu": 14.6014, | |
| "eval_gen_len": 15.5634, | |
| "eval_loss": 1.251100778579712, | |
| "eval_runtime": 857.2309, | |
| "eval_samples_per_second": 17.413, | |
| "eval_steps_per_second": 1.088, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "learning_rate": 4.524714828897338e-05, | |
| "loss": 0.4573, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "learning_rate": 4.4907658881042915e-05, | |
| "loss": 0.4783, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "learning_rate": 4.456816947311244e-05, | |
| "loss": 0.4942, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "eval_bleu": 14.3411, | |
| "eval_gen_len": 15.4888, | |
| "eval_loss": 1.252127766609192, | |
| "eval_runtime": 854.4386, | |
| "eval_samples_per_second": 17.47, | |
| "eval_steps_per_second": 1.092, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "learning_rate": 4.422868006518197e-05, | |
| "loss": 0.4455, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "learning_rate": 4.3889190657251494e-05, | |
| "loss": 0.3442, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 5.63, | |
| "learning_rate": 4.354970124932102e-05, | |
| "loss": 0.3632, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 5.63, | |
| "eval_bleu": 14.204, | |
| "eval_gen_len": 15.4075, | |
| "eval_loss": 1.3326140642166138, | |
| "eval_runtime": 847.6182, | |
| "eval_samples_per_second": 17.611, | |
| "eval_steps_per_second": 1.101, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "learning_rate": 4.3210211841390554e-05, | |
| "loss": 0.3762, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 6.16, | |
| "learning_rate": 4.2870722433460074e-05, | |
| "loss": 0.2971, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "learning_rate": 4.253123302552961e-05, | |
| "loss": 0.2601, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "eval_bleu": 14.1714, | |
| "eval_gen_len": 15.4783, | |
| "eval_loss": 1.4028061628341675, | |
| "eval_runtime": 854.4224, | |
| "eval_samples_per_second": 17.47, | |
| "eval_steps_per_second": 1.092, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "learning_rate": 4.2191743617599134e-05, | |
| "loss": 0.2697, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 6.97, | |
| "learning_rate": 4.185225420966866e-05, | |
| "loss": 0.2834, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 7.23, | |
| "learning_rate": 4.151276480173819e-05, | |
| "loss": 0.1919, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 7.23, | |
| "eval_bleu": 13.9406, | |
| "eval_gen_len": 15.4543, | |
| "eval_loss": 1.4764152765274048, | |
| "eval_runtime": 851.0189, | |
| "eval_samples_per_second": 17.54, | |
| "eval_steps_per_second": 1.096, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 7.23, | |
| "step": 13500, | |
| "total_flos": 1.872400119496704e+18, | |
| "train_loss": 0.7265107291892723, | |
| "train_runtime": 49668.6266, | |
| "train_samples_per_second": 96.178, | |
| "train_steps_per_second": 1.503 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 74640, | |
| "num_train_epochs": 40, | |
| "save_steps": 1500, | |
| "total_flos": 1.872400119496704e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |