[Feature][torch.compile] Add pass to rearrange AllGather for FP8 models in sequence parallel for better Async TP fusion#26257

jasonlizhengjian wants to merge 12 commits intovllm-project:mainvllm-project/vllm:mainfrom

jasonlizhengjian:fix/fused-scaled-matmul-signature2jasonlizhengjian/vllm:fix/fused-scaled-matmul-signature2Copy head branch name to clipboard