Rebecca Boiarsky

I want to second this issue!! I just spent many hours digging into the source code to figure out why `filter_rank_genes_groups` was filtering out genes that reported really high fold...

For all those asking whether this has been updated, I'm not a contributor to this repo but from reading the source code, it looks like this has indeed been updated...

Thanks for responding! I don't believe this is related to issue #62 . patients.txt contains text like ``` json/1.0.json json/100.0.json json/101.0.json json/102.0.json json/103.0.json json/104.0.json json/105.0.json json/106.0.json json/107.0.json json/108.0.json ``` whereas...

Hi, I'm surprised to here you are running out of memory with batch size 1 on your GPU, as I was able to fit batch size 8 on an 80GB...

Did you preprocess your data to 16906 genes, as the scBERT authors did?

I'm not exactly sure what's going wrong here, but just as a possible tip to point you in the right direction, the memory is breaking when calling "distributed_concat," which brings...

Same error! ~~At first I thought that I included a cell or gene with 0 expression, but I did not.~~ [EDIT: turns out cells with 0 expression indeed were the...

Actually, I solved it in my case -- the issue was that I was subsetting my anndata object to highly variable genes, and once I did that, there were some...

@scarpio02 thanks so much, your suggestion pointed me in the right direction! Here is what ultimately worked for me: My machine that has NVIDIA A100 80GB GPUs, NVIDIA Driver Version:...

actually, I am getting this warning message even in my environment that has flash-attn 1.0.4, so now I'm not sure why I'm getting the warning, and my updated error message...