Currently, it's quite a headache to specify in each configuration file the working directory and filenames of each stage of an inference pipeline. This is something we can easily automate via either some global wdir variables or through logging in the saved products.
Currently, it's quite a headache to specify in each configuration file the working directory and filenames of each stage of an inference pipeline. This is something we can easily automate via either some global
wdirvariables or through logging in the saved products.