Yi Zhang
Yi Zhang
Since pytorch 2.5.1 only supports cuda12.4 in official docs, and we can not change pytorch version easily, we need to update doc to guide user to reinstall pytorch if they...
LGTM cc @zhyncs
@xuzhenqi Still not，if you have any progress, please let me know, thank you very much!
I see https://github.com/NVIDIA/cutlass/pull/2095 has merged, thanks a lot! @LucasWilkinson
> Thanks for the contributions. I left a few comments. > > We also did some refactoring recently (#1541, #1538). Could you rebase? Sorry for the late reply, I am...
> Thanks for the contributions. I left a few comments. > > We also did some refactoring recently (#1541, #1538). Could you rebase? OK, I have rebased code into lastet...
> Can this run correctly now without the modification/update of vllm? If so, we can remove "WIP" in the PR title and merge this soon! I think not, there are...
> It seems this PR is merged to [yizhang2077:support-qwen2-vl](https://github.com/yizhang2077/sglang/tree/support-qwen2-vl) by accident? Should we open a new one? It seems this PR is merge into qwen2vl branch，and when this PR #1711...
Hi @MagiaSN , #3657 seems to have address your issue, could you try it again? I close this issue, and if you still have the problem, you can reopen it...
Do we need raise error for bf16 when enable deepep?