| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 0.0005, | |
| "loss": 2.3712, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8732, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.593873417721519, | |
| "eval_loss": 2.011091947555542, | |
| "eval_runtime": 5.3852, | |
| "eval_samples_per_second": 92.847, | |
| "eval_steps_per_second": 11.699, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 0.0005, | |
| "loss": 1.7329, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 0.0005, | |
| "loss": 1.6233, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 0.0005, | |
| "loss": 1.6142, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6051139240506329, | |
| "eval_loss": 1.844284176826477, | |
| "eval_runtime": 5.3165, | |
| "eval_samples_per_second": 94.047, | |
| "eval_steps_per_second": 11.85, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1625, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 0.0005, | |
| "loss": 1.206, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6007341772151898, | |
| "eval_loss": 1.981817603111267, | |
| "eval_runtime": 5.3466, | |
| "eval_samples_per_second": 93.517, | |
| "eval_steps_per_second": 11.783, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0064, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8141, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8693, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.594126582278481, | |
| "eval_loss": 2.2100281715393066, | |
| "eval_runtime": 4.739, | |
| "eval_samples_per_second": 105.508, | |
| "eval_steps_per_second": 13.294, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5477, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 0.0005, | |
| "loss": 0.6023, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5910126582278481, | |
| "eval_loss": 2.375615119934082, | |
| "eval_runtime": 4.6068, | |
| "eval_samples_per_second": 108.536, | |
| "eval_steps_per_second": 13.676, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "learning_rate": 0.0005, | |
| "loss": 0.514, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4479, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4717, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5895696202531645, | |
| "eval_loss": 2.542104482650757, | |
| "eval_runtime": 4.662, | |
| "eval_samples_per_second": 107.251, | |
| "eval_steps_per_second": 13.514, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3723, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3938, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.5891139240506329, | |
| "eval_loss": 2.658656597137451, | |
| "eval_runtime": 4.9363, | |
| "eval_samples_per_second": 101.29, | |
| "eval_steps_per_second": 12.763, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3727, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3538, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3697, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5872911392405064, | |
| "eval_loss": 2.7531518936157227, | |
| "eval_runtime": 4.7177, | |
| "eval_samples_per_second": 105.985, | |
| "eval_steps_per_second": 13.354, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3431, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3617, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.5870126582278481, | |
| "eval_loss": 2.7664403915405273, | |
| "eval_runtime": 4.7999, | |
| "eval_samples_per_second": 104.169, | |
| "eval_steps_per_second": 13.125, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3424, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3418, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3607, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.5866835443037974, | |
| "eval_loss": 2.8513779640197754, | |
| "eval_runtime": 4.6699, | |
| "eval_samples_per_second": 107.069, | |
| "eval_steps_per_second": 13.491, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "learning_rate": 0.0005, | |
| "loss": 0.339, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3414, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.5860759493670886, | |
| "eval_loss": 2.8931641578674316, | |
| "eval_runtime": 4.8648, | |
| "eval_samples_per_second": 102.78, | |
| "eval_steps_per_second": 12.95, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3302, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "learning_rate": 0.0005, | |
| "loss": 0.329, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3439, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.5854683544303797, | |
| "eval_loss": 2.9544754028320312, | |
| "eval_runtime": 5.1699, | |
| "eval_samples_per_second": 96.713, | |
| "eval_steps_per_second": 12.186, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3182, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "learning_rate": 0.0005, | |
| "loss": 0.335, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.5842784810126582, | |
| "eval_loss": 2.899094343185425, | |
| "eval_runtime": 4.8433, | |
| "eval_samples_per_second": 103.236, | |
| "eval_steps_per_second": 13.008, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "learning_rate": 0.0005, | |
| "loss": 0.327, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3223, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3391, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.5839746835443038, | |
| "eval_loss": 2.879316806793213, | |
| "eval_runtime": 4.9294, | |
| "eval_samples_per_second": 101.433, | |
| "eval_steps_per_second": 12.781, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3128, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 14.8, | |
| "learning_rate": 0.0005, | |
| "loss": 0.328, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.5851392405063292, | |
| "eval_loss": 2.8953680992126465, | |
| "eval_runtime": 4.6969, | |
| "eval_samples_per_second": 106.452, | |
| "eval_steps_per_second": 13.413, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 15.2, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3233, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3216, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3351, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.5838481012658228, | |
| "eval_loss": 2.913999080657959, | |
| "eval_runtime": 4.7522, | |
| "eval_samples_per_second": 105.214, | |
| "eval_steps_per_second": 13.257, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.4, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3087, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 16.8, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3252, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.5825063291139241, | |
| "eval_loss": 2.929701089859009, | |
| "eval_runtime": 4.8518, | |
| "eval_samples_per_second": 103.055, | |
| "eval_steps_per_second": 12.985, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 17.2, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3152, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3161, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "learning_rate": 0.0005, | |
| "loss": 0.332, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.5833924050632912, | |
| "eval_loss": 2.9811997413635254, | |
| "eval_runtime": 5.341, | |
| "eval_samples_per_second": 93.615, | |
| "eval_steps_per_second": 11.796, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.4, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3089, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 18.8, | |
| "learning_rate": 0.0005, | |
| "loss": 0.324, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.5807848101265822, | |
| "eval_loss": 2.982343912124634, | |
| "eval_runtime": 4.9018, | |
| "eval_samples_per_second": 102.002, | |
| "eval_steps_per_second": 12.852, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3169, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 19.6, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3116, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3329, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.5813164556962025, | |
| "eval_loss": 2.9419918060302734, | |
| "eval_runtime": 4.5961, | |
| "eval_samples_per_second": 108.787, | |
| "eval_steps_per_second": 13.707, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 5000, | |
| "total_flos": 3.1967425075347456e+17, | |
| "train_loss": 0.5721206008911133, | |
| "train_runtime": 3939.3187, | |
| "train_samples_per_second": 40.616, | |
| "train_steps_per_second": 1.269 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 5000, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 3.1967425075347456e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |