search

hassonlab / 247-pickling

Contains code to create pickles from raw/processed data
1 stars 9 forks source link

tfsemb_concat does not need --layer_idx #76

Closed hvgazula closed 1 year ago

hvgazula commented 2 years ago

make changes to code accordingly

zkokaja commented 1 year ago

Remove argument from parser @hvgazula

hvgazula commented 1 year ago

I recommend removing this altogether from the tfsemb_parser.py as we are saving embeddings from all layers now. https://github.com/hassonlab/247-pickling/blob/main/Makefile#L111. What do you think @VeritasJoker?

VeritasJoker commented 1 year ago

I don't know about this. Do we still want the flexibility to be able to generate one single layer (usually the last one)? I don't know if everyone is looking into all layers. It could take up more storage space for those who aren't.

hvgazula commented 1 year ago

Computation time versus storage space. I rate the former higher.