| { | |
| "best_metric": 0.8897758152173914, | |
| "best_model_checkpoint": "out/checkpoint-6940", | |
| "epoch": 10.0, | |
| "global_step": 6940, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 1.4409221902017292e-06, | |
| "loss": 3.8183, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 2.8818443804034585e-06, | |
| "loss": 3.7614, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 4.322766570605187e-06, | |
| "loss": 3.2003, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 5.749279538904899e-06, | |
| "loss": 2.3107, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 7.190201729106629e-06, | |
| "loss": 1.7861, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 8.631123919308358e-06, | |
| "loss": 1.4103, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7471127717391305, | |
| "eval_loss": 1.1298972368240356, | |
| "eval_runtime": 237.0543, | |
| "eval_samples_per_second": 24.838, | |
| "eval_steps_per_second": 0.776, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 9.993595901376882e-06, | |
| "loss": 1.1941, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 9.833493435798911e-06, | |
| "loss": 0.9417, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 9.673390970220942e-06, | |
| "loss": 0.8036, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 9.513288504642972e-06, | |
| "loss": 0.7488, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 9.353186039065003e-06, | |
| "loss": 0.6777, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 9.193083573487034e-06, | |
| "loss": 0.6306, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 9.032981107909063e-06, | |
| "loss": 0.5324, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8495244565217391, | |
| "eval_loss": 0.6167322397232056, | |
| "eval_runtime": 234.135, | |
| "eval_samples_per_second": 25.148, | |
| "eval_steps_per_second": 0.786, | |
| "step": 1388 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 8.874479666986872e-06, | |
| "loss": 0.5331, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 8.714377201408901e-06, | |
| "loss": 0.3651, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 8.554274735830932e-06, | |
| "loss": 0.3298, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 8.394172270252963e-06, | |
| "loss": 0.3177, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 8.234069804674993e-06, | |
| "loss": 0.3093, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 8.073967339097024e-06, | |
| "loss": 0.317, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 7.913864873519053e-06, | |
| "loss": 0.3164, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8678668478260869, | |
| "eval_loss": 0.5136280655860901, | |
| "eval_runtime": 234.6734, | |
| "eval_samples_per_second": 25.09, | |
| "eval_steps_per_second": 0.784, | |
| "step": 2082 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 7.753762407941084e-06, | |
| "loss": 0.2798, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "learning_rate": 7.593659942363113e-06, | |
| "loss": 0.1962, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "learning_rate": 7.4335574767851426e-06, | |
| "loss": 0.1898, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "learning_rate": 7.2734550112071735e-06, | |
| "loss": 0.1615, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 7.113352545629203e-06, | |
| "loss": 0.1534, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 6.953250080051234e-06, | |
| "loss": 0.162, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "learning_rate": 6.793147614473263e-06, | |
| "loss": 0.1535, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8758491847826086, | |
| "eval_loss": 0.5263639092445374, | |
| "eval_runtime": 233.6811, | |
| "eval_samples_per_second": 25.197, | |
| "eval_steps_per_second": 0.787, | |
| "step": 2776 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "learning_rate": 6.633045148895294e-06, | |
| "loss": 0.1458, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "learning_rate": 6.472942683317323e-06, | |
| "loss": 0.1132, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "learning_rate": 6.3128402177393535e-06, | |
| "loss": 0.1162, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "learning_rate": 6.152737752161384e-06, | |
| "loss": 0.0797, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "learning_rate": 5.992635286583414e-06, | |
| "loss": 0.0832, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "learning_rate": 5.832532821005445e-06, | |
| "loss": 0.0759, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "learning_rate": 5.672430355427474e-06, | |
| "loss": 0.1018, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8766983695652174, | |
| "eval_loss": 0.5890083909034729, | |
| "eval_runtime": 236.5762, | |
| "eval_samples_per_second": 24.888, | |
| "eval_steps_per_second": 0.778, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "learning_rate": 5.512327889849503e-06, | |
| "loss": 0.0816, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.19, | |
| "learning_rate": 5.352225424271534e-06, | |
| "loss": 0.0557, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "learning_rate": 5.1921229586935636e-06, | |
| "loss": 0.0544, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 5.48, | |
| "learning_rate": 5.0320204931155946e-06, | |
| "loss": 0.045, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 5.62, | |
| "learning_rate": 4.871918027537625e-06, | |
| "loss": 0.0541, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "learning_rate": 4.713416586615434e-06, | |
| "loss": 0.0656, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.91, | |
| "learning_rate": 4.553314121037464e-06, | |
| "loss": 0.0555, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8772078804347826, | |
| "eval_loss": 0.612079381942749, | |
| "eval_runtime": 232.3119, | |
| "eval_samples_per_second": 25.345, | |
| "eval_steps_per_second": 0.792, | |
| "step": 4164 | |
| }, | |
| { | |
| "epoch": 6.05, | |
| "learning_rate": 4.393211655459494e-06, | |
| "loss": 0.0439, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "learning_rate": 4.2331091898815245e-06, | |
| "loss": 0.0234, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 6.34, | |
| "learning_rate": 4.073006724303555e-06, | |
| "loss": 0.035, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 6.48, | |
| "learning_rate": 3.912904258725585e-06, | |
| "loss": 0.0376, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 6.63, | |
| "learning_rate": 3.7528017931476145e-06, | |
| "loss": 0.0348, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 6.77, | |
| "learning_rate": 3.5926993275696446e-06, | |
| "loss": 0.0169, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 6.92, | |
| "learning_rate": 3.432596861991675e-06, | |
| "loss": 0.031, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8777173913043478, | |
| "eval_loss": 0.6623581051826477, | |
| "eval_runtime": 235.7029, | |
| "eval_samples_per_second": 24.981, | |
| "eval_steps_per_second": 0.781, | |
| "step": 4858 | |
| }, | |
| { | |
| "epoch": 7.06, | |
| "learning_rate": 3.2724943964137053e-06, | |
| "loss": 0.0305, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 3.1123919308357354e-06, | |
| "loss": 0.0205, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 7.35, | |
| "learning_rate": 2.952289465257765e-06, | |
| "loss": 0.0239, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 7.49, | |
| "learning_rate": 2.7921869996797953e-06, | |
| "loss": 0.0224, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 7.64, | |
| "learning_rate": 2.6320845341018254e-06, | |
| "loss": 0.0187, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 7.78, | |
| "learning_rate": 2.4719820685238555e-06, | |
| "loss": 0.0125, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "learning_rate": 2.3118796029458857e-06, | |
| "loss": 0.0264, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8868885869565217, | |
| "eval_loss": 0.6172327995300293, | |
| "eval_runtime": 233.4791, | |
| "eval_samples_per_second": 25.219, | |
| "eval_steps_per_second": 0.788, | |
| "step": 5552 | |
| }, | |
| { | |
| "epoch": 8.07, | |
| "learning_rate": 2.1517771373679154e-06, | |
| "loss": 0.013, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 8.21, | |
| "learning_rate": 1.9916746717899455e-06, | |
| "loss": 0.0145, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 8.36, | |
| "learning_rate": 1.8315722062119759e-06, | |
| "loss": 0.0146, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "learning_rate": 1.6714697406340058e-06, | |
| "loss": 0.0127, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 8.65, | |
| "learning_rate": 1.511367275056036e-06, | |
| "loss": 0.0112, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "learning_rate": 1.351264809478066e-06, | |
| "loss": 0.0115, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 8.93, | |
| "learning_rate": 1.1911623439000962e-06, | |
| "loss": 0.0079, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8884171195652174, | |
| "eval_loss": 0.6224890947341919, | |
| "eval_runtime": 234.5344, | |
| "eval_samples_per_second": 25.105, | |
| "eval_steps_per_second": 0.785, | |
| "step": 6246 | |
| }, | |
| { | |
| "epoch": 9.08, | |
| "learning_rate": 1.0310598783221263e-06, | |
| "loss": 0.0073, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 9.22, | |
| "learning_rate": 8.709574127441562e-07, | |
| "loss": 0.0084, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 9.37, | |
| "learning_rate": 7.108549471661865e-07, | |
| "loss": 0.0034, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 9.51, | |
| "learning_rate": 5.507524815882165e-07, | |
| "loss": 0.0058, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 9.65, | |
| "learning_rate": 3.906500160102466e-07, | |
| "loss": 0.0048, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "learning_rate": 2.3054755043227666e-07, | |
| "loss": 0.0063, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 9.94, | |
| "learning_rate": 7.044508485430676e-08, | |
| "loss": 0.0058, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8897758152173914, | |
| "eval_loss": 0.6338901519775391, | |
| "eval_runtime": 234.8904, | |
| "eval_samples_per_second": 25.067, | |
| "eval_steps_per_second": 0.783, | |
| "step": 6940 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 6940, | |
| "total_flos": 2.7871930756137603e+19, | |
| "train_loss": 0.3924562724076362, | |
| "train_runtime": 18647.2594, | |
| "train_samples_per_second": 11.902, | |
| "train_steps_per_second": 0.372 | |
| } | |
| ], | |
| "max_steps": 6940, | |
| "num_train_epochs": 10, | |
| "total_flos": 2.7871930756137603e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |