| { | |
| "best_metric": 0.26939964294433594, | |
| "best_model_checkpoint": "./checkpoint/checkpoint-4000", | |
| "epoch": 4.790692369111441, | |
| "eval_steps": 500, | |
| "global_step": 10500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 4e-08, | |
| "loss": 12.0623, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 1.9960000000000002e-05, | |
| "loss": 7.4437, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "eval_bleu": 33.6574, | |
| "eval_gen_len": 31.2317, | |
| "eval_loss": 3.4493963718414307, | |
| "eval_runtime": 576.3536, | |
| "eval_samples_per_second": 1.73, | |
| "eval_steps_per_second": 1.73, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 1.904543280726925e-05, | |
| "loss": 1.3107, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_bleu": 33.3513, | |
| "eval_gen_len": 31.7091, | |
| "eval_loss": 0.2903362810611725, | |
| "eval_runtime": 476.1637, | |
| "eval_samples_per_second": 2.094, | |
| "eval_steps_per_second": 2.094, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.8088952654232427e-05, | |
| "loss": 0.1582, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "eval_bleu": 33.7009, | |
| "eval_gen_len": 31.7202, | |
| "eval_loss": 0.2747056484222412, | |
| "eval_runtime": 477.4288, | |
| "eval_samples_per_second": 2.088, | |
| "eval_steps_per_second": 2.088, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.71324725011956e-05, | |
| "loss": 0.1477, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "eval_bleu": 34.2274, | |
| "eval_gen_len": 31.5557, | |
| "eval_loss": 0.2713315188884735, | |
| "eval_runtime": 473.8787, | |
| "eval_samples_per_second": 2.104, | |
| "eval_steps_per_second": 2.104, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 1.617599234815878e-05, | |
| "loss": 0.1413, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "eval_bleu": 34.0664, | |
| "eval_gen_len": 31.674, | |
| "eval_loss": 0.2716849446296692, | |
| "eval_runtime": 477.2223, | |
| "eval_samples_per_second": 2.089, | |
| "eval_steps_per_second": 2.089, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 1.5219512195121952e-05, | |
| "loss": 0.1355, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "eval_bleu": 34.1168, | |
| "eval_gen_len": 31.8506, | |
| "eval_loss": 0.2718922197818756, | |
| "eval_runtime": 479.68, | |
| "eval_samples_per_second": 2.078, | |
| "eval_steps_per_second": 2.078, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 1.4263032042085128e-05, | |
| "loss": 0.136, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_bleu": 34.2638, | |
| "eval_gen_len": 31.7523, | |
| "eval_loss": 0.270623117685318, | |
| "eval_runtime": 479.0203, | |
| "eval_samples_per_second": 2.081, | |
| "eval_steps_per_second": 2.081, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 1.3306551889048302e-05, | |
| "loss": 0.1316, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "eval_bleu": 34.1582, | |
| "eval_gen_len": 31.6931, | |
| "eval_loss": 0.26939964294433594, | |
| "eval_runtime": 477.4424, | |
| "eval_samples_per_second": 2.088, | |
| "eval_steps_per_second": 2.088, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 1.235007173601148e-05, | |
| "loss": 0.1312, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "eval_bleu": 34.4277, | |
| "eval_gen_len": 31.662, | |
| "eval_loss": 0.2704804539680481, | |
| "eval_runtime": 479.1942, | |
| "eval_samples_per_second": 2.081, | |
| "eval_steps_per_second": 2.081, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 1.1393591582974655e-05, | |
| "loss": 0.1258, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "eval_bleu": 34.3594, | |
| "eval_gen_len": 31.651, | |
| "eval_loss": 0.27046987414360046, | |
| "eval_runtime": 478.936, | |
| "eval_samples_per_second": 2.082, | |
| "eval_steps_per_second": 2.082, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "learning_rate": 1.043711142993783e-05, | |
| "loss": 0.1271, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "eval_bleu": 34.3412, | |
| "eval_gen_len": 31.8094, | |
| "eval_loss": 0.27054643630981445, | |
| "eval_runtime": 481.7151, | |
| "eval_samples_per_second": 2.07, | |
| "eval_steps_per_second": 2.07, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 9.480631276901005e-06, | |
| "loss": 0.1249, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "eval_bleu": 34.2387, | |
| "eval_gen_len": 31.7212, | |
| "eval_loss": 0.2704330086708069, | |
| "eval_runtime": 479.9945, | |
| "eval_samples_per_second": 2.077, | |
| "eval_steps_per_second": 2.077, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "learning_rate": 8.52415112386418e-06, | |
| "loss": 0.1245, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "eval_bleu": 34.3033, | |
| "eval_gen_len": 31.8616, | |
| "eval_loss": 0.27082785964012146, | |
| "eval_runtime": 482.2741, | |
| "eval_samples_per_second": 2.067, | |
| "eval_steps_per_second": 2.067, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "learning_rate": 7.5676709708273554e-06, | |
| "loss": 0.1195, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "eval_bleu": 34.2748, | |
| "eval_gen_len": 31.9017, | |
| "eval_loss": 0.27176010608673096, | |
| "eval_runtime": 484.0841, | |
| "eval_samples_per_second": 2.06, | |
| "eval_steps_per_second": 2.06, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "learning_rate": 6.611190817790531e-06, | |
| "loss": 0.1198, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "eval_bleu": 34.2897, | |
| "eval_gen_len": 31.7312, | |
| "eval_loss": 0.27175214886665344, | |
| "eval_runtime": 479.5665, | |
| "eval_samples_per_second": 2.079, | |
| "eval_steps_per_second": 2.079, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "learning_rate": 5.654710664753707e-06, | |
| "loss": 0.1209, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "eval_bleu": 34.4446, | |
| "eval_gen_len": 31.7272, | |
| "eval_loss": 0.2709992527961731, | |
| "eval_runtime": 478.4748, | |
| "eval_samples_per_second": 2.084, | |
| "eval_steps_per_second": 2.084, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "learning_rate": 4.6982305117168825e-06, | |
| "loss": 0.1201, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "eval_bleu": 34.3571, | |
| "eval_gen_len": 31.7432, | |
| "eval_loss": 0.2712614834308624, | |
| "eval_runtime": 478.6295, | |
| "eval_samples_per_second": 2.083, | |
| "eval_steps_per_second": 2.083, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "learning_rate": 3.7417503586800574e-06, | |
| "loss": 0.1201, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "eval_bleu": 34.4398, | |
| "eval_gen_len": 31.7613, | |
| "eval_loss": 0.27223262190818787, | |
| "eval_runtime": 478.6225, | |
| "eval_samples_per_second": 2.083, | |
| "eval_steps_per_second": 2.083, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "learning_rate": 2.785270205643233e-06, | |
| "loss": 0.1178, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "eval_bleu": 34.4074, | |
| "eval_gen_len": 31.7753, | |
| "eval_loss": 0.27177131175994873, | |
| "eval_runtime": 479.0762, | |
| "eval_samples_per_second": 2.081, | |
| "eval_steps_per_second": 2.081, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "learning_rate": 1.8287900526064088e-06, | |
| "loss": 0.1181, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "eval_bleu": 34.4628, | |
| "eval_gen_len": 31.8034, | |
| "eval_loss": 0.2723881006240845, | |
| "eval_runtime": 479.5721, | |
| "eval_samples_per_second": 2.079, | |
| "eval_steps_per_second": 2.079, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "learning_rate": 8.72309899569584e-07, | |
| "loss": 0.1169, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "eval_bleu": 34.563, | |
| "eval_gen_len": 31.7442, | |
| "eval_loss": 0.2720402777194977, | |
| "eval_runtime": 478.0313, | |
| "eval_samples_per_second": 2.086, | |
| "eval_steps_per_second": 2.086, | |
| "step": 10500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10955, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 1.820292017113006e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |