Can Ergen

I think the main difference is that there is transform_batch in generating the posterior expression, while there is no transform_covariate. In line of this, px_r doesn't support gene times other...

@martinkim0 It's the version of destVI we are using for Curio data (other sparsity regularization). Can you have a pass?

To me it sounds most flexible to be able to pass an obs column with train_indices, valid_indices and test_indices and leave it up to the user/developer how to define those....

There is multi-GPU support. However, your calculation looks vastly off and I don't think your dataset is large enough to benefit largely from multi-GPU. E.g. on the human CELLXGENE census...

We cover the enhancement to use custom dataloader in the recent version of scVI-tools. However, it is not clear yet which minimal checks (integer, gene names) we still want to...

@ori-kron-wis We can merge this one after fixing the n_samples issue.

Hi, I really enjoyed reading your preprint. It's interesting to see your good results with CycleConsistencyLoss as it was overintegrating in some examples in our hand (slightly different implementation and...

@PierreBoyeau used cycle consistency loss in our hands, do you see substantial differences to our trials? In which cases do you recommend GMM over VAMP? @martinkim0 Can you review the...

Hi, to do some expectation management. The code will be part of a stable relese with version 1.2. We are currently finishing version 1.1. Timeline for version 1.2 is on...

Hi @moinfar and @Hrovatin we're sorry for the long time it takes us to integrate the changes. Unfortunately, we were not able to understand all parts of the code yet...