{ "best_metric": 0.030118444934487343, "best_model_checkpoint": "./vit-base-crack-classification/checkpoint-609", "epoch": 3.0, "eval_steps": 100, "global_step": 609, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.05, "learning_rate": 0.00019671592775041052, "loss": 1.4989, "step": 10 }, { "epoch": 0.1, "learning_rate": 0.00019343185550082105, "loss": 1.059, "step": 20 }, { "epoch": 0.15, "learning_rate": 0.00019014778325123153, "loss": 0.8111, "step": 30 }, { "epoch": 0.2, "learning_rate": 0.00018686371100164203, "loss": 0.6727, "step": 40 }, { "epoch": 0.25, "learning_rate": 0.00018357963875205257, "loss": 0.7606, "step": 50 }, { "epoch": 0.3, "learning_rate": 0.00018029556650246307, "loss": 0.688, "step": 60 }, { "epoch": 0.34, "learning_rate": 0.00017701149425287358, "loss": 0.6076, "step": 70 }, { "epoch": 0.39, "learning_rate": 0.00017372742200328408, "loss": 0.5333, "step": 80 }, { "epoch": 0.44, "learning_rate": 0.0001704433497536946, "loss": 0.4982, "step": 90 }, { "epoch": 0.49, "learning_rate": 0.0001671592775041051, "loss": 0.4319, "step": 100 }, { "epoch": 0.54, "learning_rate": 0.0001638752052545156, "loss": 0.5499, "step": 110 }, { "epoch": 0.59, "learning_rate": 0.00016059113300492613, "loss": 0.3999, "step": 120 }, { "epoch": 0.64, "learning_rate": 0.00015730706075533664, "loss": 0.443, "step": 130 }, { "epoch": 0.69, "learning_rate": 0.00015402298850574712, "loss": 0.417, "step": 140 }, { "epoch": 0.74, "learning_rate": 0.00015073891625615765, "loss": 0.3383, "step": 150 }, { "epoch": 0.79, "learning_rate": 0.00014745484400656815, "loss": 0.4005, "step": 160 }, { "epoch": 0.84, "learning_rate": 0.00014417077175697866, "loss": 0.3608, "step": 170 }, { "epoch": 0.89, "learning_rate": 0.00014088669950738917, "loss": 0.3782, "step": 180 }, { "epoch": 0.94, "learning_rate": 0.0001376026272577997, "loss": 0.3411, "step": 190 }, { "epoch": 0.99, "learning_rate": 0.00013431855500821018, "loss": 0.2138, "step": 200 }, { "epoch": 1.0, "eval_accuracy": 0.9027777777777778, "eval_loss": 0.2958821952342987, "eval_runtime": 15.6222, "eval_samples_per_second": 46.088, "eval_steps_per_second": 5.761, "step": 203 }, { "epoch": 1.03, "learning_rate": 0.00013103448275862068, "loss": 0.1832, "step": 210 }, { "epoch": 1.08, "learning_rate": 0.00012775041050903122, "loss": 0.1517, "step": 220 }, { "epoch": 1.13, "learning_rate": 0.00012446633825944172, "loss": 0.1099, "step": 230 }, { "epoch": 1.18, "learning_rate": 0.00012118226600985223, "loss": 0.1762, "step": 240 }, { "epoch": 1.23, "learning_rate": 0.00011789819376026272, "loss": 0.1496, "step": 250 }, { "epoch": 1.28, "learning_rate": 0.00011461412151067324, "loss": 0.1287, "step": 260 }, { "epoch": 1.33, "learning_rate": 0.00011133004926108374, "loss": 0.1389, "step": 270 }, { "epoch": 1.38, "learning_rate": 0.00010804597701149426, "loss": 0.1583, "step": 280 }, { "epoch": 1.43, "learning_rate": 0.00010476190476190477, "loss": 0.1214, "step": 290 }, { "epoch": 1.48, "learning_rate": 0.00010147783251231529, "loss": 0.1566, "step": 300 }, { "epoch": 1.53, "learning_rate": 9.819376026272578e-05, "loss": 0.1256, "step": 310 }, { "epoch": 1.58, "learning_rate": 9.490968801313629e-05, "loss": 0.1123, "step": 320 }, { "epoch": 1.63, "learning_rate": 9.16256157635468e-05, "loss": 0.0853, "step": 330 }, { "epoch": 1.67, "learning_rate": 8.834154351395731e-05, "loss": 0.067, "step": 340 }, { "epoch": 1.72, "learning_rate": 8.505747126436782e-05, "loss": 0.1035, "step": 350 }, { "epoch": 1.77, "learning_rate": 8.177339901477834e-05, "loss": 0.0681, "step": 360 }, { "epoch": 1.82, "learning_rate": 7.848932676518884e-05, "loss": 0.0382, "step": 370 }, { "epoch": 1.87, "learning_rate": 7.520525451559935e-05, "loss": 0.0537, "step": 380 }, { "epoch": 1.92, "learning_rate": 7.192118226600985e-05, "loss": 0.0283, "step": 390 }, { "epoch": 1.97, "learning_rate": 6.863711001642037e-05, "loss": 0.0291, "step": 400 }, { "epoch": 2.0, "eval_accuracy": 0.9833333333333333, "eval_loss": 0.06349290162324905, "eval_runtime": 15.4242, "eval_samples_per_second": 46.68, "eval_steps_per_second": 5.835, "step": 406 }, { "epoch": 2.02, "learning_rate": 6.535303776683088e-05, "loss": 0.0275, "step": 410 }, { "epoch": 2.07, "learning_rate": 6.206896551724138e-05, "loss": 0.0318, "step": 420 }, { "epoch": 2.12, "learning_rate": 5.878489326765189e-05, "loss": 0.0312, "step": 430 }, { "epoch": 2.17, "learning_rate": 5.55008210180624e-05, "loss": 0.0284, "step": 440 }, { "epoch": 2.22, "learning_rate": 5.2216748768472914e-05, "loss": 0.0217, "step": 450 }, { "epoch": 2.27, "learning_rate": 4.893267651888342e-05, "loss": 0.014, "step": 460 }, { "epoch": 2.32, "learning_rate": 4.5648604269293925e-05, "loss": 0.02, "step": 470 }, { "epoch": 2.36, "learning_rate": 4.236453201970443e-05, "loss": 0.0148, "step": 480 }, { "epoch": 2.41, "learning_rate": 3.908045977011495e-05, "loss": 0.0136, "step": 490 }, { "epoch": 2.46, "learning_rate": 3.5796387520525456e-05, "loss": 0.0129, "step": 500 }, { "epoch": 2.51, "learning_rate": 3.251231527093596e-05, "loss": 0.0375, "step": 510 }, { "epoch": 2.56, "learning_rate": 2.9228243021346467e-05, "loss": 0.0124, "step": 520 }, { "epoch": 2.61, "learning_rate": 2.5944170771756983e-05, "loss": 0.0129, "step": 530 }, { "epoch": 2.66, "learning_rate": 2.266009852216749e-05, "loss": 0.0189, "step": 540 }, { "epoch": 2.71, "learning_rate": 1.9376026272577998e-05, "loss": 0.0142, "step": 550 }, { "epoch": 2.76, "learning_rate": 1.6091954022988507e-05, "loss": 0.0332, "step": 560 }, { "epoch": 2.81, "learning_rate": 1.2807881773399016e-05, "loss": 0.0116, "step": 570 }, { "epoch": 2.86, "learning_rate": 9.523809523809523e-06, "loss": 0.0311, "step": 580 }, { "epoch": 2.91, "learning_rate": 6.239737274220033e-06, "loss": 0.012, "step": 590 }, { "epoch": 2.96, "learning_rate": 2.955665024630542e-06, "loss": 0.0128, "step": 600 }, { "epoch": 3.0, "eval_accuracy": 0.9944444444444445, "eval_loss": 0.030118444934487343, "eval_runtime": 15.4638, "eval_samples_per_second": 46.56, "eval_steps_per_second": 5.82, "step": 609 }, { "epoch": 3.0, "step": 609, "total_flos": 1.5064982852419584e+18, "train_loss": 0.23008910012362627, "train_runtime": 477.4194, "train_samples_per_second": 40.719, "train_steps_per_second": 1.276 } ], "logging_steps": 10, "max_steps": 609, "num_train_epochs": 3, "save_steps": 500, "total_flos": 1.5064982852419584e+18, "trial_name": null, "trial_params": null }