Open deeplearining-sheen opened 4 years ago
Hey @deeplearining-sheen,
mhm, interesting bug. Looks like there is still an active process on a temporary preprocessed sample file, which throws an exception when cleaning up the temporary files after finishing the first cross-validation fold.
MIScnn shouldn't has an open process on these files at this point.
What IO interface are you using?
Do you have a third party software running on the data? Like synching with Gdrive or Dropbox?
Cheers, Dominik
Hi,
I run as following part, and then I got error message: PermissionError: [WinError 32] The process cannot access the file because it is being used by another process: 'batches\32891163.case_00010.pickle'
Pls. let me know how to solve it.
Library import
from miscnn.evaluation.cross_validation import cross_validation
Run cross-validation function
cross_validation(validation_samples, model, k_fold=3, epochs=5, iterations=2, evaluation_path="evaluation1", draw_figures=True, save_models=True, run_detailed_evaluation=True, callbacks=[cb_lr])
Epoch 1/5 2/2 [==============================] - ETA: 0s - loss: 2.4823 - dice_soft: 0.1726 - dice_crossentropy: 1.7110 Epoch 00001: val_loss improved from inf to 2.52166, saving model to evaluation1\fold_0\model.hdf5 2/2 [==============================] - 17s 9s/step - loss: 2.4823 - dice_soft: 0.1726 - dice_crossentropy: 1.7110 - val_loss: 2.5217 - val_dice_soft: 0.1605 - val_dice_crossentropy: 1.7466 - lr: 1.0000e-05 Epoch 2/5 2/2 [==============================] - ETA: 0s - loss: 2.4626 - dice_soft: 0.1791 - dice_crossentropy: 1.7857 Epoch 00002: val_loss improved from 2.52166 to 2.51943, saving model to evaluation1\fold_0\model.hdf5 2/2 [==============================] - 16s 8s/step - loss: 2.4626 - dice_soft: 0.1791 - dice_crossentropy: 1.7857 - val_loss: 2.5194 - val_dice_soft: 0.1609 - val_dice_crossentropy: 1.7434 - lr: 1.0000e-05 Epoch 3/5 2/2 [==============================] - ETA: 0s - loss: 2.4510 - dice_soft: 0.1830 - dice_crossentropy: 1.9342 Epoch 00003: val_loss did not improve from 2.51943 2/2 [==============================] - 13s 7s/step - loss: 2.4510 - dice_soft: 0.1830 - dice_crossentropy: 1.9342 - val_loss: 2.5364 - val_dice_soft: 0.1554 - val_dice_crossentropy: 1.6986 - lr: 1.0000e-05 Epoch 4/5 2/2 [==============================] - ETA: 0s - loss: 2.4698 - dice_soft: 0.1767 - dice_crossentropy: 2.3842 Epoch 00004: val_loss improved from 2.51943 to 2.51251, saving model to evaluation1\fold_0\model.hdf5 2/2 [==============================] - 17s 9s/step - loss: 2.4698 - dice_soft: 0.1767 - dice_crossentropy: 2.3842 - val_loss: 2.5125 - val_dice_soft: 0.1609 - val_dice_crossentropy: 1.7242 - lr: 1.0000e-05 Epoch 5/5 2/2 [==============================] - ETA: 0s - loss: 2.4784 - dice_soft: 0.1739 - dice_crossentropy: 1.8819 Epoch 00005: val_loss did not improve from 2.51251 2/2 [==============================] - 20s 10s/step - loss: 2.4784 - dice_soft: 0.1739 - dice_crossentropy: 1.8819 - val_loss: 2.5216 - val_dice_soft: 0.1581 - val_dice_crossentropy: 1.6996 - lr: 1.0000e-05
PermissionError Traceback (most recent call last)