[Tag NCS_Beta_v1.3] In "training/data_loader/data_loader.py", `getPatchData` raises an `AssertionError` if its second argument `comp` contains 'cu_average' or 'partition_cu_average'.

Problem

If, in the script provided in the README.txt at Section "Data loading", comps is modified to include 'partition_cu_average_Y', and this script is run, the method getPatchData in "training/data_loader/data_loader.py" raises an AssertionError. Yet, 'partition_cu_average_Y' is supposed to refer to a valid type of data to be loaded. The same goes if comps is modified to include 'cu_average'.

As another example, let us say that the following command is run:

$ python training/example/create_unified_dataset.py --input_json dataset.json --components partition_cu_average_Y,pred_Y --nb_patches 12 --output_file file_final.pb

Then, the same kind of exception is raised by the method getPatchData in "training/data_loader/data_loader.py".

Explanation

At this line https://vcgit.hhi.fraunhofer.de/jvet-ahg-nnvc/VVCSoftware_VTM/-/blob/NCS_Beta_v1.3/training/data_loader/data_loader.py#L86, 'suffix_cu_average' should be replaced by 'suffix_partition_cu_average'. At this line https://vcgit.hhi.fraunhofer.de/jvet-ahg-nnvc/VVCSoftware_VTM/-/blob/NCS_Beta_v1.3/training/data_loader/data_loader.py#L156, 'cu_average' should be replaced by 'partition_cu_average'.