<--Spectrogram--> Type : Length : 11138 Length : 319287046 Shape one element : (752, 400) ----------------------------- <--EEG--> Type : Length : 20183 Length : 127492639 Shape one element : (128, 256, 4) Creadting 5 folds of the training data ========== fold: 0 training ========== Epoch 1 - avg_train_loss: 0.4490 avg_val_loss: 0.4243 time: 125s Epoch 1 - Save Best valid loss: 0.4243 Model Epoch 2 - avg_train_loss: 0.3431 avg_val_loss: 0.3847 time: 120s Epoch 2 - Save Best valid loss: 0.3847 Model Epoch 3 - avg_train_loss: 0.3024 avg_val_loss: 0.3708 time: 120s Epoch 3 - Save Best valid loss: 0.3708 Model Epoch 4 - avg_train_loss: 0.2752 avg_val_loss: 0.3599 time: 120s Epoch 4 - Save Best valid loss: 0.3599 Model Epoch 5 - avg_train_loss: 0.2533 avg_val_loss: 0.3708 time: 120s ========== fold: 0 result ========== Score with best loss weights stage1: 0.3598788937469328 ========== fold: 1 training ========== Epoch 1 - avg_train_loss: 0.4582 avg_val_loss: 0.3886 time: 121s Epoch 1 - Save Best valid loss: 0.3886 Model Epoch 2 - avg_train_loss: 0.3445 avg_val_loss: 0.3738 time: 120s Epoch 2 - Save Best valid loss: 0.3738 Model Epoch 3 - avg_train_loss: 0.3068 avg_val_loss: 0.3503 time: 120s Epoch 3 - Save Best valid loss: 0.3503 Model Epoch 4 - avg_train_loss: 0.2764 avg_val_loss: 0.3595 time: 120s Epoch 5 - avg_train_loss: 0.2570 avg_val_loss: 0.3453 time: 120s Epoch 5 - Save Best valid loss: 0.3453 Model ========== fold: 1 result ========== Score with best loss weights stage1: 0.3452556534977812 ========== fold: 2 training ========== Epoch 1 - avg_train_loss: 0.4569 avg_val_loss: 0.3837 time: 121s Epoch 1 - Save Best valid loss: 0.3837 Model Epoch 2 - avg_train_loss: 0.3405 avg_val_loss: 0.3605 time: 120s Epoch 2 - Save Best valid loss: 0.3605 Model Epoch 3 - avg_train_loss: 0.2994 avg_val_loss: 0.3407 time: 121s Epoch 3 - Save Best valid loss: 0.3407 Model Epoch 4 - avg_train_loss: 0.2729 avg_val_loss: 0.3402 time: 120s Epoch 4 - Save Best valid loss: 0.3402 Model Epoch 5 - avg_train_loss: 0.2489 avg_val_loss: 0.3374 time: 121s Epoch 5 - Save Best valid loss: 0.3374 Model ========== fold: 2 result ========== Score with best loss weights stage1: 0.3373754414364581 ========== fold: 3 training ========== Epoch 1 - avg_train_loss: 0.4586 avg_val_loss: 0.3846 time: 121s Epoch 1 - Save Best valid loss: 0.3846 Model Epoch 2 - avg_train_loss: 0.3424 avg_val_loss: 0.3545 time: 120s Epoch 2 - Save Best valid loss: 0.3545 Model Epoch 3 - avg_train_loss: 0.3046 avg_val_loss: 0.3353 time: 120s Epoch 3 - Save Best valid loss: 0.3353 Model Epoch 4 - avg_train_loss: 0.2746 avg_val_loss: 0.3511 time: 120s Epoch 5 - avg_train_loss: 0.2525 avg_val_loss: 0.3423 time: 120s ========== fold: 3 result ========== Score with best loss weights stage1: 0.3353293751642418 ========== fold: 4 training ========== Epoch 1 - avg_train_loss: 0.4668 avg_val_loss: 0.3785 time: 121s Epoch 1 - Save Best valid loss: 0.3785 Model Epoch 2 - avg_train_loss: 0.3545 avg_val_loss: 0.3445 time: 119s Epoch 2 - Save Best valid loss: 0.3445 Model Epoch 3 - avg_train_loss: 0.3113 avg_val_loss: 0.3239 time: 120s Epoch 3 - Save Best valid loss: 0.3239 Model Epoch 4 - avg_train_loss: 0.2863 avg_val_loss: 0.3295 time: 120s Epoch 5 - avg_train_loss: 0.2586 avg_val_loss: 0.3275 time: 120s ========== fold: 4 result ========== Score with best loss weights stage1: 0.3239006738005828 ========== CV ========== Score with best loss weights stage1: 0.34034800752919936 ========== fold: 0 training ========== Epoch 1 - avg_train_loss: 0.2589 avg_val_loss: 0.4813 time: 48s Epoch 1 - Save Best valid loss: 0.4813 Model Epoch 2 - avg_train_loss: 0.2118 avg_val_loss: 0.5373 time: 48s Epoch 3 - avg_train_loss: 0.1913 avg_val_loss: 0.4766 time: 47s Epoch 3 - Save Best valid loss: 0.4766 Model Epoch 4 - avg_train_loss: 0.1719 avg_val_loss: 0.4972 time: 47s Epoch 5 - avg_train_loss: 0.1586 avg_val_loss: 0.5009 time: 47s ========== fold: 0 result ========== Score with best loss weights stage2: 0.47662054733501263 ========== fold: 1 training ========== Epoch 1 - avg_train_loss: 0.2335 avg_val_loss: 0.4232 time: 50s Epoch 1 - Save Best valid loss: 0.4232 Model Epoch 2 - avg_train_loss: 0.1913 avg_val_loss: 0.4603 time: 48s Epoch 3 - avg_train_loss: 0.1755 avg_val_loss: 0.4866 time: 49s Epoch 4 - avg_train_loss: 0.1574 avg_val_loss: 0.4770 time: 49s Epoch 5 - avg_train_loss: 0.1445 avg_val_loss: 0.4739 time: 48s ========== fold: 1 result ========== Score with best loss weights stage2: 0.4232155225685631 ========== fold: 2 training ========== Epoch 1 - avg_train_loss: 0.2305 avg_val_loss: 0.4416 time: 50s Epoch 1 - Save Best valid loss: 0.4416 Model Epoch 2 - avg_train_loss: 0.1891 avg_val_loss: 0.4643 time: 49s Epoch 3 - avg_train_loss: 0.1665 avg_val_loss: 0.4959 time: 49s Epoch 4 - avg_train_loss: 0.1552 avg_val_loss: 0.4505 time: 49s Epoch 5 - avg_train_loss: 0.1433 avg_val_loss: 0.4583 time: 49s ========== fold: 2 result ========== Score with best loss weights stage2: 0.4416244997448457 ========== fold: 3 training ========== Epoch 1 - avg_train_loss: 0.2756 avg_val_loss: 0.4538 time: 50s Epoch 1 - Save Best valid loss: 0.4538 Model Epoch 2 - avg_train_loss: 0.2203 avg_val_loss: 0.4531 time: 49s Epoch 2 - Save Best valid loss: 0.4531 Model Epoch 3 - avg_train_loss: 0.1964 avg_val_loss: 0.4876 time: 50s Epoch 4 - avg_train_loss: 0.1778 avg_val_loss: 0.4753 time: 49s Epoch 5 - avg_train_loss: 0.1682 avg_val_loss: 0.4917 time: 49s ========== fold: 3 result ========== Score with best loss weights stage2: 0.453083156121616 ========== fold: 4 training ========== Epoch 1 - avg_train_loss: 0.2741 avg_val_loss: 0.4381 time: 50s Epoch 1 - Save Best valid loss: 0.4381 Model Epoch 2 - avg_train_loss: 0.2243 avg_val_loss: 0.4424 time: 49s Epoch 3 - avg_train_loss: 0.2020 avg_val_loss: 0.4471 time: 49s Epoch 4 - avg_train_loss: 0.1878 avg_val_loss: 0.4826 time: 50s Epoch 5 - avg_train_loss: 0.1698 avg_val_loss: 0.4790 time: 49s ========== fold: 4 result ========== Score with best loss weights stage2: 0.4381120401343222 ========== CV ========== Score with best loss weights stage2: 0.44653115318087194