Add vLLM DeepSeek-R1 128K/8K recipes for GB300 NVL72#165
Draft
qiching wants to merge 1 commit into
Draft
Conversation
Draft recipes for DeepSeek-R1 / V3 (671B MoE) long-context disaggregated
serving on GB300 NVL72 with vLLM. Covers:
- Canonical DEP8 disagg config (4 prefill / 2 decode nodes)
- Wide-EP scaling: DEP16 / DEP32 with flashinfer_nvlink_one_sided A2A
- MTP variant (speculative decoding)
- Rate-matching ablations (prefill-only / decode-only / full-disagg)
- Prefill experiments (DP4+EP4 vs PP4, chunked prefill 4K/12K/32K)
- LongBench v2 accuracy recipe
Related upstream work:
- vLLM PR #39841 (FP8 cast-order fix for chunked prefill)
- FlashInfer PR #3129 (FP8 E4M3/E5M2 in concat_mla_k)
- vLLM issues #41687 (DeepEP), #41685 (long-context PP),
#41682 (PP scheduler), #40674 (NixlConnector + PP)
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #165 +/- ##
=======================================
Coverage ? 65.08%
=======================================
Files ? 67
Lines ? 8217
Branches ? 0
=======================================
Hits ? 5348
Misses ? 2869
Partials ? 0 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Draft recipes for DeepSeek-R1 long-context disaggregated serving on GB300 NVL72 with vLLM. Covers:
Related upstream work: