| { | |
| "best_metric": 1.4880632162094116, | |
| "best_model_checkpoint": "./zhko_mbartLarge_19p_run1/checkpoint-5000", | |
| "epoch": 8.973438621679827, | |
| "eval_steps": 5000, | |
| "global_step": 25000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 2.4329, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 1.9318, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 5e-05, | |
| "loss": 1.8063, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 4.9379498634897e-05, | |
| "loss": 1.7429, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 4.875899726979399e-05, | |
| "loss": 1.6586, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 4.8138495904690995e-05, | |
| "loss": 1.5945, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 4.751799453958799e-05, | |
| "loss": 1.4888, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 4.689749317448498e-05, | |
| "loss": 1.3603, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 4.627699180938198e-05, | |
| "loss": 1.2398, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 4.5656490444278984e-05, | |
| "loss": 1.1911, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "eval_bleu": 9.7977, | |
| "eval_gen_len": 14.6128, | |
| "eval_loss": 1.4880632162094116, | |
| "eval_runtime": 656.7417, | |
| "eval_samples_per_second": 16.966, | |
| "eval_steps_per_second": 1.061, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 4.5035989079175975e-05, | |
| "loss": 1.1667, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "learning_rate": 4.441548771407297e-05, | |
| "loss": 1.129, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 4.379498634896997e-05, | |
| "loss": 1.048, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "learning_rate": 4.317448498386697e-05, | |
| "loss": 0.9477, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "learning_rate": 4.2553983618763964e-05, | |
| "loss": 0.8805, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 4.193348225366096e-05, | |
| "loss": 0.8654, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "learning_rate": 4.131298088855796e-05, | |
| "loss": 0.8357, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "learning_rate": 4.069247952345495e-05, | |
| "loss": 0.7993, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 4.0071978158351954e-05, | |
| "loss": 0.7272, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "learning_rate": 3.945147679324895e-05, | |
| "loss": 0.6536, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "eval_bleu": 13.3897, | |
| "eval_gen_len": 14.9179, | |
| "eval_loss": 1.6061058044433594, | |
| "eval_runtime": 661.3374, | |
| "eval_samples_per_second": 16.848, | |
| "eval_steps_per_second": 1.054, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "learning_rate": 3.883097542814594e-05, | |
| "loss": 0.6277, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "learning_rate": 3.821047406304294e-05, | |
| "loss": 0.6063, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "learning_rate": 3.758997269793994e-05, | |
| "loss": 0.589, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "learning_rate": 3.6969471332836934e-05, | |
| "loss": 0.5473, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "learning_rate": 3.634896996773393e-05, | |
| "loss": 0.4878, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "learning_rate": 3.572846860263093e-05, | |
| "loss": 0.4455, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "learning_rate": 3.510796723752792e-05, | |
| "loss": 0.4362, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "learning_rate": 3.4487465872424924e-05, | |
| "loss": 0.4156, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "learning_rate": 3.386696450732192e-05, | |
| "loss": 0.4034, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 5.38, | |
| "learning_rate": 3.324646314221891e-05, | |
| "loss": 0.3665, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 5.38, | |
| "eval_bleu": 14.0018, | |
| "eval_gen_len": 15.2051, | |
| "eval_loss": 1.7928513288497925, | |
| "eval_runtime": 662.4842, | |
| "eval_samples_per_second": 16.819, | |
| "eval_steps_per_second": 1.052, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 5.56, | |
| "learning_rate": 3.262596177711591e-05, | |
| "loss": 0.3223, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 5.74, | |
| "learning_rate": 3.200546041201291e-05, | |
| "loss": 0.3039, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 5.92, | |
| "learning_rate": 3.1384959046909904e-05, | |
| "loss": 0.2964, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "learning_rate": 3.07644576818069e-05, | |
| "loss": 0.2834, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 6.28, | |
| "learning_rate": 3.01439563167039e-05, | |
| "loss": 0.2662, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 6.46, | |
| "learning_rate": 2.9523454951600893e-05, | |
| "loss": 0.2376, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 6.64, | |
| "learning_rate": 2.8902953586497894e-05, | |
| "loss": 0.2152, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "learning_rate": 2.828245222139489e-05, | |
| "loss": 0.2076, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 2.7661950856291885e-05, | |
| "loss": 0.1995, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 7.18, | |
| "learning_rate": 2.704144949118888e-05, | |
| "loss": 0.194, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 7.18, | |
| "eval_bleu": 14.7102, | |
| "eval_gen_len": 14.7308, | |
| "eval_loss": 1.9398521184921265, | |
| "eval_runtime": 638.6619, | |
| "eval_samples_per_second": 17.446, | |
| "eval_steps_per_second": 1.091, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 7.36, | |
| "learning_rate": 2.6420948126085876e-05, | |
| "loss": 0.1763, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 7.54, | |
| "learning_rate": 2.5800446760982877e-05, | |
| "loss": 0.1566, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 7.72, | |
| "learning_rate": 2.517994539587987e-05, | |
| "loss": 0.1471, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "learning_rate": 2.455944403077687e-05, | |
| "loss": 0.1462, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 8.08, | |
| "learning_rate": 2.3938942665673866e-05, | |
| "loss": 0.1364, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 8.26, | |
| "learning_rate": 2.3318441300570863e-05, | |
| "loss": 0.1297, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 8.44, | |
| "learning_rate": 2.269793993546786e-05, | |
| "loss": 0.1174, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 8.61, | |
| "learning_rate": 2.2077438570364855e-05, | |
| "loss": 0.1075, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "learning_rate": 2.1456937205261852e-05, | |
| "loss": 0.1041, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "learning_rate": 2.083643584015885e-05, | |
| "loss": 0.1004, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "eval_bleu": 14.9684, | |
| "eval_gen_len": 14.8811, | |
| "eval_loss": 2.0678608417510986, | |
| "eval_runtime": 646.6568, | |
| "eval_samples_per_second": 17.23, | |
| "eval_steps_per_second": 1.078, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "step": 25000, | |
| "total_flos": 1.7337038143488e+18, | |
| "train_loss": 0.6854593217468262, | |
| "train_runtime": 45689.3015, | |
| "train_samples_per_second": 29.267, | |
| "train_steps_per_second": 0.915 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 41790, | |
| "num_train_epochs": 15, | |
| "save_steps": 5000, | |
| "total_flos": 1.7337038143488e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |