Open chnlyi opened 2 years ago
Sorry, according to the code, these are the variables
WORKFLOW_ROOT="$ROOT_DIR/kaggle_workflow"
VALIDATION="$OUTPUTS_DIR/validation"
MASS_TRAIN_UNET="$OUTPUTS_DIR/train_unet"
TRAIN_MASKRCNN="$WORKFLOW_ROOT/inputs/train_maskrcnn"
and here is the .sh script
CONT_MAT_FILE="$MATLAB_SCRIPTS/generateValidation/validationFileNames.mat"
echo "GENERATE VALIDATION DATA FROM MASS_TRAIN_MASKRCNN:"
matlab -nodisplay -nodesktop -nosplash -nojvm -r "addpath('${MATLAB_SCRIPTS}/generateValidation');moveToValidation('${MASS_TRAIN_MASKRCNN}/','${VALIDATION}/','${CONT_MAT_FILE}');exit;"
if [ $? -ne 0 ]
then
echo ERROR: "Error during generate validation maskrcnn"
exit 1
fi
echo "GENERATE VALIDATION DATA FROM MASS_TRAIN_MASKRCNN DONE"
echo "GENERATE VALIDATION DATA FROM MASS_TRAIN_UNET:"
matlab -nodisplay -nodesktop -nosplash -nojvm -r "addpath('${MATLAB_SCRIPTS}/generateValidation');moveToValidation('${MASS_TRAIN_UNET}/','${VALIDATION}/','${CONT_MAT_FILE}');exit;"
if [ $? -ne 0 ]
then
echo ERROR: "Error during generate validation unet"
exit 1
fi
echo "GENERATE VALIDATION DATA FROM MASS_TRAIN_UNET DONE"
If I put my validation in the $VALIDATION, what do these 2 steps do?
Should I copy my train into $TRAIN_UNET and $TRAIN_MASKRCNN? What about my validation data? (I experimented a few times, but it seems that I have to runGenerateValidationCustom.sh using my validation data and copy both train and validation data into $TRAIN_UNET and $TRAIN_MASKRCNN.)
Yes, copy the train images there. Validation goes in the
$VALIDATION
folder which will be inkaggle_workflow/outputs/validation
by default. Indeed you need to runrunGenerateValidationCustom.sh /path/to/validationfolder
, the$IMAGES_DIR
variable in this script is only used to list the validation images, sorry for the confusing variable names.Originally posted by @spreka in https://github.com/spreka/biomagdsb/issues/14#issuecomment-1150762680