Preprocessing for BioMassters

[sandykerr]

I am trying to run the BioMassters example using the BioMassters dataset from HuggingFace. I have downloaded the data and extracted it as best I can into the file structure in the documentation for that example. When I look at the code, though, it seems that the UNET regressor script expects files that are different from the BioMassters data; namely, it expects (100, 100, 128) data, while the BioMassters data is (256, 256, 4). Do I need to get the TESSERA embeddings as inputs instead of the BioMassters data?