Martin Kim

Results 75 comments of Martin Kim

Thanks for bringing up this issue! I think it makes sense based on how we've named our `ArchesMixin` API to implement `prepare_query_mudata`, which I anticipate will have much different code...

closing in favor of #2673

Hi, thanks for bringing up this up and sorry you're running into this issue. One possible source for these NaN errors are the exponential activations in the model - we've...

Thanks! I'll try to get this running and get back to you

Hi, thanks for this suggestion. I think adding these changes would be a great addition to the codebase. However, for now I am opposed to adding them to the scVI...

@Hrovatin For scArches, compatibility mostly depends on how you register covariates with `AnnDataManager` as well as the names of the base neural net components used in the model. Feel free...

Sorry for the late response. Totally fine with me, please go ahead!

@Hrovatin Sorry for the delay in reviewing - I've taken another pass at it now. I think we'll need to add a couple more changes and trim down code where...

Hi, sorry that you're running into this issue - we're working on making this available in #2318 as part of our next scvi-tools release (v1.1). However, I'm unable to reproduce...

Hi @zhenxingjian, what model are you using for this? I'll note that we have only tested multi-GPU training on scVI.