feat: enable Qwen35-MoE MTP && sp prefill cuda graph#969
Open
amd-yilizhao wants to merge 2 commits intomainfrom
Open
feat: enable Qwen35-MoE MTP && sp prefill cuda graph#969amd-yilizhao wants to merge 2 commits intomainfrom
amd-yilizhao wants to merge 2 commits intomainfrom
Conversation
Collaborator
AI Code Review - PR #969Status: BLOCKING Summary: P0/0 · P1/2 · P2/0 · P3/0 Blocking IssuesP1
Checklist Violations (12 fail / 81 total)General Principles Checklist
RTP-LLM Checklist
Python Static-First Checklist
Strengths
|
Collaborator
AI Code Review - PR #969Status: BLOCKING Summary: P0/0 · P1/2 · P2/1 · P3/0 Blocking IssuesP1
Non-blocking SuggestionsP2
Checklist Violations (10 fail / 81 total)General Principles Checklist
RTP-LLM Checklist
Python Static-First Checklist
Strengths
|
03b2376 to
32ed093
Compare
Collaborator
AI Code Review - PR #969Status: BLOCKING Summary: P0/0 · P1/3 · P2/0 · P3/0 Blocking IssuesP1
Checklist Violations (9 fail / 97 total)General Principles Checklist
RTP-LLM Checklist
Python Static-First Checklist
Strengths
|
32ed093 to
8f136d8
Compare
Collaborator
AI Code Review - PR #969Status: LGTM Summary: P0/0 · P1/0 · P2/0 · P3/0 lgtm ready to ci Checklist Violations (1 fail / 56 total)Python Static-First Checklist
Strengths
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Enable ROCm paged prefill CUDA graph capture/replay for MTP (speculative decoding) scenarios on Qwen3.5-MoE.
Key changes:
isEmbeddingStylePrefillCudaGraph/isMtpDraftPrefillCudaGraph); skip invalid batch-1 post-check for MTP draft; addfinish_capture_session()to defer trt_allreduce IPC handle exchange and avoid ProcessGroupNCCL watchdog racesupdatePrefillRuntime()pybind method to pass runtimemax_seq_len/max_prefix_lenfrom Python, eliminating.item()host-device sync during graph capture; conditionally skip k/v output buffer allocation in paged modeAiterPrefillAttnOpPagedandAiterPrefillImplPaged— pre-allocate metadata buffers, replace.to(device)with pre-moved tensors, replace.item()with pre-computed values, disable triton short-seq path during graph capturencclAllGatherfailure during graph capture); deferconsume_capture()tofinish_hipgraph_capture_session(); enable trt_allreduce priority path during graph capture; add lazy comm init from ProcessGrouphas_pending_capture()and_capture_handles_pendingflag for deferred IPC handle finalization