[MODEL] Fix handling of multiple channels for gpt-oss with speculative decoding #26291

astralord wants to merge 8 commits intovllm-project:mainvllm-project/vllm:mainfrom

astralord:fix-gpt-oss-with-speculative-decoding-handle-multiple-channelsastralord/vllm:fix-gpt-oss-with-speculative-decoding-handle-multiple-channelsCopy head branch name to clipboard