For single anndata inference model can now be run out-of-the-box without needing to specially download any additional files or adjust the paths
Also changed the structure of the evaluation script which essentially just does arg parsing now. Moved all functions to a new evaluate.py file
To-do: a lot of the model argument like sample size, chromosome token index etc. should ideally be inferred directly from the model without needing to be specific manually
For single anndata inference model can now be run out-of-the-box without needing to specially download any additional files or adjust the paths
Also changed the structure of the evaluation script which essentially just does arg parsing now. Moved all functions to a new evaluate.py file
To-do: a lot of the model argument like sample size, chromosome token index etc. should ideally be inferred directly from the model without needing to be specific manually