Jee Jee Li

Sorry for the late feedback caused by the holiday. I'll ask maintainers for their opinions

After discussing with the maintainer, we're not considering this PR for now. it can remain as a draft PR to help user understand the details of your work. Thank you...

> Thanks @jeejeelee . Can you also update the "supported models" page for LoRA please. > > https://github.com/vllm-project/vllm/blob/9f1710f1ace3535920c0bb6d4cc329c36289080e/docs/source/models/supported_models.md?plain=1#L339 This model supports LoRA, it's just that the MOE layers don't support...

@aurickq Thanks for your contribution and patience

Currently, all models have their `stacked_params_mapping` in `load_weights`, not just the Gemma model. So I personally think these modifications may not be necessary

> That said, it looks like the same piece of information but repeated twice. Would it make sense to combine them into one? I think we probably can't combine them....