{ "best_metric": null, "best_model_checkpoint": null, "best_supernet_model_checkpoint": null, "epoch": 4.0, "global_step": 45484, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "compression_loss": 0.0, "epoch": 0.04, "learning_rate": 1.9994108544833432e-05, "loss": 0.9256, "step": 500 }, { "compression_loss": 0.0, "epoch": 0.09, "learning_rate": 1.9976346258894502e-05, "loss": 0.6433, "step": 1000 }, { "compression_loss": 0.0, "epoch": 0.13, "learning_rate": 1.9946662925610062e-05, "loss": 0.5994, "step": 1500 }, { "compression_loss": 0.0, "epoch": 0.18, "learning_rate": 1.9905117625959957e-05, "loss": 0.5411, "step": 2000 }, { "compression_loss": 0.0, "epoch": 0.22, "learning_rate": 1.9851759905098248e-05, "loss": 0.5124, "step": 2500 }, { "compression_loss": 0.0, "epoch": 0.26, "learning_rate": 1.97866533951692e-05, "loss": 0.4807, "step": 3000 }, { "compression_loss": 0.0, "epoch": 0.31, "learning_rate": 1.9709875739422284e-05, "loss": 0.4652, "step": 3500 }, { "compression_loss": 0.0, "epoch": 0.35, "learning_rate": 1.9621518499618157e-05, "loss": 0.4457, "step": 4000 }, { "compression_loss": 0.0, "epoch": 0.4, "learning_rate": 1.9521687046836e-05, "loss": 0.4466, "step": 4500 }, { "compression_loss": 0.0, "epoch": 0.44, "learning_rate": 1.9410500435812454e-05, "loss": 0.4176, "step": 5000 }, { "compression_loss": 0.0, "epoch": 0.48, "learning_rate": 1.928834718660774e-05, "loss": 0.4164, "step": 5500 }, { "compression_loss": 0.0, "epoch": 0.53, "learning_rate": 1.9154883430559243e-05, "loss": 0.4014, "step": 6000 }, { "compression_loss": 0.0, "epoch": 0.57, "learning_rate": 1.901080149718364e-05, "loss": 0.3843, "step": 6500 }, { "compression_loss": 0.0, "epoch": 0.62, "learning_rate": 1.8855695787268013e-05, "loss": 0.3758, "step": 7000 }, { "compression_loss": 0.0, "epoch": 0.66, "learning_rate": 1.869002915190846e-05, "loss": 0.369, "step": 7500 }, { "compression_loss": 0.0, "epoch": 0.7, "learning_rate": 1.8513999158075064e-05, "loss": 0.3938, "step": 8000 }, { "compression_loss": 0.0, "epoch": 0.75, "learning_rate": 1.8327815731637612e-05, "loss": 0.3676, "step": 8500 }, { "compression_loss": 0.0, "epoch": 0.79, "learning_rate": 1.8131700907016972e-05, "loss": 0.363, "step": 9000 }, { "compression_loss": 0.0, "epoch": 0.84, "learning_rate": 1.79258885623963e-05, "loss": 0.3566, "step": 9500 }, { "compression_loss": 0.0, "epoch": 0.88, "learning_rate": 1.771062414080787e-05, "loss": 0.3426, "step": 10000 }, { "compression_loss": 0.0, "epoch": 0.92, "learning_rate": 1.7486164357428192e-05, "loss": 0.3246, "step": 10500 }, { "compression_loss": 0.0, "epoch": 0.97, "learning_rate": 1.725277689343042e-05, "loss": 0.3347, "step": 11000 }, { "Minimum SubNet": "OrderedDict([(, {0: 320, 1: 256, 2: 320, 3: 192, 4: 256, 5: 256, 6: 192, 7: 256, 8: 64, 9: 192, 10: 192, 11: 512, 12: 1675, 13: 1666, 14: 1787, 15: 1791, 16: 1772, 17: 1751, 18: 1709, 19: 1590, 20: 1320, 21: 762, 22: 348, 23: 115})])", "epoch": 1.0, "eval_accuracy": 0.8986148899332179, "eval_combined_score": 0.8815576185145908, "eval_f1": 0.8645003470959637, "eval_loss": 0.27813825011253357, "eval_runtime": 27.532, "eval_samples_per_second": 1468.473, "eval_steps_per_second": 11.478, "step": 11371 }, { "SuperNet": "OrderedDict([(, {0: 768, 1: 768, 2: 768, 3: 768, 4: 768, 5: 768, 6: 768, 7: 768, 8: 768, 9: 768, 10: 768, 11: 768, 12: 3072, 13: 3072, 14: 3072, 15: 3072, 16: 3072, 17: 3072, 18: 3072, 19: 3072, 20: 3072, 21: 3072, 22: 3072, 23: 3072})])", "epoch": 1.0, "eval_accuracy": 0.9021766015335148, "eval_combined_score": 0.8865253049131546, "eval_f1": 0.8708740082927945, "eval_loss": 0.2632658779621124, "eval_runtime": 33.2751, "eval_samples_per_second": 1215.023, "eval_steps_per_second": 9.497, "step": 11371 }, { "compression_loss": 0.0, "epoch": 1.01, "learning_rate": 1.7011232592278327e-05, "loss": 0.2935, "step": 11500 }, { "compression_loss": 0.0, "epoch": 1.06, "learning_rate": 1.6760851492396305e-05, "loss": 0.2134, "step": 12000 }, { "compression_loss": 0.0, "epoch": 1.1, "learning_rate": 1.650293242000503e-05, "loss": 0.215, "step": 12500 }, { "compression_loss": 0.0, "epoch": 1.14, "learning_rate": 1.623728921389968e-05, "loss": 0.214, "step": 13000 }, { "compression_loss": 0.0, "epoch": 1.19, "learning_rate": 1.5963682982587413e-05, "loss": 0.2014, "step": 13500 }, { "compression_loss": 0.0, "epoch": 1.23, "learning_rate": 1.5682964717069908e-05, "loss": 0.2065, "step": 14000 }, { "compression_loss": 0.0, "epoch": 1.28, "learning_rate": 1.5395469189988864e-05, "loss": 0.2073, "step": 14500 }, { "compression_loss": 0.0, "epoch": 1.32, "learning_rate": 1.5101539256259345e-05, "loss": 0.2234, "step": 15000 }, { "compression_loss": 0.0, "epoch": 1.36, "learning_rate": 1.4801525444195615e-05, "loss": 0.203, "step": 15500 }, { "compression_loss": 0.0, "epoch": 1.41, "learning_rate": 1.4495785537485976e-05, "loss": 0.1976, "step": 16000 }, { "compression_loss": 0.0, "epoch": 1.45, "learning_rate": 1.418468414851523e-05, "loss": 0.2019, "step": 16500 }, { "compression_loss": 0.0, "epoch": 1.5, "learning_rate": 1.3868592283543443e-05, "loss": 0.2025, "step": 17000 }, { "compression_loss": 0.0, "epoch": 1.54, "learning_rate": 1.3548532661979667e-05, "loss": 0.1949, "step": 17500 }, { "compression_loss": 0.0, "epoch": 1.58, "learning_rate": 1.3223604296950975e-05, "loss": 0.1989, "step": 18000 }, { "compression_loss": 0.0, "epoch": 1.63, "learning_rate": 1.2894831598799443e-05, "loss": 0.2037, "step": 18500 }, { "compression_loss": 0.0, "epoch": 1.67, "learning_rate": 1.2563274280384495e-05, "loss": 0.1902, "step": 19000 }, { "compression_loss": 0.0, "epoch": 1.71, "learning_rate": 1.222799898813265e-05, "loss": 0.1871, "step": 19500 }, { "compression_loss": 0.0, "epoch": 1.76, "learning_rate": 1.1890066679225309e-05, "loss": 0.1857, "step": 20000 }, { "compression_loss": 0.0, "epoch": 1.8, "learning_rate": 1.1549880357334364e-05, "loss": 0.1832, "step": 20500 }, { "compression_loss": 0.0, "epoch": 1.85, "learning_rate": 1.1207845714171553e-05, "loss": 0.1872, "step": 21000 }, { "compression_loss": 0.0, "epoch": 1.89, "learning_rate": 1.0864370645677793e-05, "loss": 0.1788, "step": 21500 }, { "compression_loss": 0.0, "epoch": 1.93, "learning_rate": 1.0519864765583883e-05, "loss": 0.1808, "step": 22000 }, { "compression_loss": 0.0, "epoch": 1.98, "learning_rate": 1.0174738916922611e-05, "loss": 0.1664, "step": 22500 }, { "Minimum SubNet": "OrderedDict([(, {0: 320, 1: 256, 2: 320, 3: 192, 4: 256, 5: 256, 6: 192, 7: 256, 8: 64, 9: 192, 10: 192, 11: 512, 12: 1675, 13: 1666, 14: 1787, 15: 1791, 16: 1772, 17: 1751, 18: 1709, 19: 1590, 20: 1320, 21: 762, 22: 348, 23: 115})])", "epoch": 2.0, "eval_accuracy": 0.9048478852337373, "eval_combined_score": 0.8892236797616033, "eval_f1": 0.8735994742894694, "eval_loss": 0.27244308590888977, "eval_runtime": 32.6391, "eval_samples_per_second": 1238.699, "eval_steps_per_second": 9.682, "step": 22742 }, { "SuperNet": "OrderedDict([(, {0: 768, 1: 768, 2: 768, 3: 768, 4: 768, 5: 768, 6: 768, 7: 768, 8: 768, 9: 768, 10: 768, 11: 768, 12: 3072, 13: 3072, 14: 3072, 15: 3072, 16: 3072, 17: 3072, 18: 3072, 19: 3072, 20: 3072, 21: 3072, 22: 3072, 23: 3072})])", "epoch": 2.0, "eval_accuracy": 0.9106109324758842, "eval_combined_score": 0.8957706744614529, "eval_f1": 0.8809304164470216, "eval_loss": 0.26653361320495605, "eval_runtime": 45.2051, "eval_samples_per_second": 894.367, "eval_steps_per_second": 6.99, "step": 22742 }, { "compression_loss": 0.0, "epoch": 2.02, "learning_rate": 9.830785888286694e-06, "loss": 0.1477, "step": 23000 }, { "compression_loss": 0.0, "epoch": 2.07, "learning_rate": 9.485653464158049e-06, "loss": 0.112, "step": 23500 }, { "compression_loss": 0.0, "epoch": 2.11, "learning_rate": 9.141134427791111e-06, "loss": 0.107, "step": 24000 }, { "compression_loss": 0.0, "epoch": 2.15, "learning_rate": 8.797639637908284e-06, "loss": 0.1161, "step": 24500 }, { "compression_loss": 0.0, "epoch": 2.2, "learning_rate": 8.45626115108603e-06, "loss": 0.1099, "step": 25000 }, { "compression_loss": 0.0, "epoch": 2.24, "learning_rate": 8.116037966577595e-06, "loss": 0.11, "step": 25500 }, { "compression_loss": 0.0, "epoch": 2.29, "learning_rate": 7.778061514912307e-06, "loss": 0.111, "step": 26000 }, { "compression_loss": 0.0, "epoch": 2.33, "learning_rate": 7.4427348524048315e-06, "loss": 0.1096, "step": 26500 }, { "compression_loss": 0.0, "epoch": 2.37, "learning_rate": 7.111119121678447e-06, "loss": 0.1057, "step": 27000 }, { "compression_loss": 0.0, "epoch": 2.42, "learning_rate": 6.782280804725536e-06, "loss": 0.1046, "step": 27500 }, { "compression_loss": 0.0, "epoch": 2.46, "learning_rate": 6.457279802573981e-06, "loss": 0.097, "step": 28000 }, { "compression_loss": 0.0, "epoch": 2.51, "learning_rate": 6.136503697570255e-06, "loss": 0.1122, "step": 28500 }, { "compression_loss": 0.0, "epoch": 2.55, "learning_rate": 5.82033503362853e-06, "loss": 0.1063, "step": 29000 }, { "compression_loss": 0.0, "epoch": 2.59, "learning_rate": 5.509150860025087e-06, "loss": 0.1052, "step": 29500 }, { "compression_loss": 0.0, "epoch": 2.64, "learning_rate": 5.2039283504942894e-06, "loss": 0.1023, "step": 30000 }, { "compression_loss": 0.0, "epoch": 2.68, "learning_rate": 4.9038082855161555e-06, "loss": 0.0949, "step": 30500 }, { "compression_loss": 0.0, "epoch": 2.73, "learning_rate": 4.609765721621778e-06, "loss": 0.0949, "step": 31000 }, { "compression_loss": 0.0, "epoch": 2.77, "learning_rate": 4.322151321443352e-06, "loss": 0.1048, "step": 31500 }, { "compression_loss": 0.0, "epoch": 2.81, "learning_rate": 4.041862786524023e-06, "loss": 0.0938, "step": 32000 }, { "compression_loss": 0.0, "epoch": 2.86, "learning_rate": 3.768111087965475e-06, "loss": 0.0971, "step": 32500 }, { "compression_loss": 0.0, "epoch": 2.9, "learning_rate": 3.501791274575046e-06, "loss": 0.0905, "step": 33000 }, { "compression_loss": 0.0, "epoch": 2.95, "learning_rate": 3.243220948013981e-06, "loss": 0.0934, "step": 33500 }, { "compression_loss": 0.0, "epoch": 2.99, "learning_rate": 2.9927084682357343e-06, "loss": 0.092, "step": 34000 }, { "Minimum SubNet": "OrderedDict([(, {0: 320, 1: 256, 2: 320, 3: 192, 4: 256, 5: 256, 6: 192, 7: 256, 8: 64, 9: 192, 10: 192, 11: 512, 12: 1675, 13: 1666, 14: 1787, 15: 1791, 16: 1772, 17: 1751, 18: 1709, 19: 1590, 20: 1320, 21: 762, 22: 348, 23: 115})])", "epoch": 3.0, "eval_accuracy": 0.9094484293841207, "eval_combined_score": 0.8940211803693942, "eval_f1": 0.8785939313546676, "eval_loss": 0.28715240955352783, "eval_runtime": 25.8149, "eval_samples_per_second": 1566.149, "eval_steps_per_second": 12.241, "step": 34113 }, { "SuperNet": "OrderedDict([(, {0: 768, 1: 768, 2: 768, 3: 768, 4: 768, 5: 768, 6: 768, 7: 768, 8: 768, 9: 768, 10: 768, 11: 768, 12: 3072, 13: 3072, 14: 3072, 15: 3072, 16: 3072, 17: 3072, 18: 3072, 19: 3072, 20: 3072, 21: 3072, 22: 3072, 23: 3072})])", "epoch": 3.0, "eval_accuracy": 0.9141231758595103, "eval_combined_score": 0.8993641670310579, "eval_f1": 0.8846051582026057, "eval_loss": 0.2707538604736328, "eval_runtime": 32.4731, "eval_samples_per_second": 1245.032, "eval_steps_per_second": 9.731, "step": 34113 }, { "compression_loss": 0.0, "epoch": 3.03, "learning_rate": 2.751504176813342e-06, "loss": 0.0737, "step": 34500 }, { "compression_loss": 0.0, "epoch": 3.08, "learning_rate": 2.5179585331698086e-06, "loss": 0.0666, "step": 35000 }, { "compression_loss": 0.0, "epoch": 3.12, "learning_rate": 2.2933356533607244e-06, "loss": 0.0736, "step": 35500 }, { "compression_loss": 0.0, "epoch": 3.17, "learning_rate": 2.0779034130610965e-06, "loss": 0.0682, "step": 36000 }, { "compression_loss": 0.0, "epoch": 3.21, "learning_rate": 1.8723211031117794e-06, "loss": 0.07, "step": 36500 }, { "compression_loss": 0.0, "epoch": 3.25, "learning_rate": 1.6760099970815446e-06, "loss": 0.0701, "step": 37000 }, { "compression_loss": 0.0, "epoch": 3.3, "learning_rate": 1.4896257268334447e-06, "loss": 0.0665, "step": 37500 }, { "compression_loss": 0.0, "epoch": 3.34, "learning_rate": 1.3133905663061163e-06, "loss": 0.068, "step": 38000 }, { "compression_loss": 0.0, "epoch": 3.39, "learning_rate": 1.1475146860424102e-06, "loss": 0.0696, "step": 38500 }, { "compression_loss": 0.0, "epoch": 3.43, "learning_rate": 9.924958789004234e-07, "loss": 0.0687, "step": 39000 }, { "compression_loss": 0.0, "epoch": 3.47, "learning_rate": 8.478977577170089e-07, "loss": 0.0682, "step": 39500 }, { "compression_loss": 0.0, "epoch": 3.52, "learning_rate": 7.144704103835564e-07, "loss": 0.0723, "step": 40000 }, { "compression_loss": 0.0, "epoch": 3.56, "learning_rate": 5.918382309806136e-07, "loss": 0.0707, "step": 40500 }, { "compression_loss": 0.0, "epoch": 3.61, "learning_rate": 4.80425824410179e-07, "loss": 0.074, "step": 41000 }, { "compression_loss": 0.0, "epoch": 3.65, "learning_rate": 3.8036605636091126e-07, "loss": 0.0727, "step": 41500 }, { "compression_loss": 0.0, "epoch": 3.69, "learning_rate": 2.91778253848517e-07, "loss": 0.0687, "step": 42000 }, { "compression_loss": 0.0, "epoch": 3.74, "learning_rate": 2.147680629114379e-07, "loss": 0.0679, "step": 42500 }, { "compression_loss": 0.0, "epoch": 3.78, "learning_rate": 1.4942732262188898e-07, "loss": 0.0659, "step": 43000 }, { "compression_loss": 0.0, "epoch": 3.83, "learning_rate": 9.583395556248875e-08, "loss": 0.07, "step": 43500 }, { "compression_loss": 0.0, "epoch": 3.87, "learning_rate": 5.405187489911701e-08, "loss": 0.0685, "step": 44000 }, { "compression_loss": 0.0, "epoch": 3.91, "learning_rate": 2.4178886625614294e-08, "loss": 0.0739, "step": 44500 }, { "compression_loss": 0.0, "epoch": 3.96, "learning_rate": 6.130896503045991e-09, "loss": 0.0693, "step": 45000 }, { "Minimum SubNet": "OrderedDict([(, {0: 320, 1: 256, 2: 320, 3: 192, 4: 256, 5: 256, 6: 192, 7: 256, 8: 64, 9: 192, 10: 192, 11: 512, 12: 1675, 13: 1666, 14: 1787, 15: 1791, 16: 1772, 17: 1751, 18: 1709, 19: 1590, 20: 1320, 21: 762, 22: 348, 23: 115})])", "epoch": 4.0, "eval_accuracy": 0.9088300766757358, "eval_combined_score": 0.8929735128011296, "eval_f1": 0.8771169489265235, "eval_loss": 0.29656311869621277, "eval_runtime": 26.0405, "eval_samples_per_second": 1552.579, "eval_steps_per_second": 12.135, "step": 45484 }, { "SuperNet": "OrderedDict([(, {0: 768, 1: 768, 2: 768, 3: 768, 4: 768, 5: 768, 6: 768, 7: 768, 8: 768, 9: 768, 10: 768, 11: 768, 12: 3072, 13: 3072, 14: 3072, 15: 3072, 16: 3072, 17: 3072, 18: 3072, 19: 3072, 20: 3072, 21: 3072, 22: 3072, 23: 3072})])", "epoch": 4.0, "eval_accuracy": 0.9143952510511996, "eval_combined_score": 0.8995123693503939, "eval_f1": 0.8846294876495883, "eval_loss": 0.2779307961463928, "eval_runtime": 32.5644, "eval_samples_per_second": 1241.54, "eval_steps_per_second": 9.704, "step": 45484 }, { "epoch": 4.0, "step": 45484, "total_flos": 9.573190504851456e+16, "train_loss": 0.20403898293818326, "train_runtime": 74603.6038, "train_samples_per_second": 19.508, "train_steps_per_second": 0.61 } ], "max_steps": 45484, "min_subnet_acc": null, "min_subnet_best_acc": null, "num_train_epochs": 4, "supernet_acc": null, "supernet_best_acc": null, "total_flos": 9.573190504851456e+16, "trial_name": null, "trial_params": null }