BBuf/vllm-mixtral-quark-int4fp8-moe-optimization
skills/model-optimization/vllm/vllm-mixtral-quark-int4fp8-moe-optimization/SKILL.md
PR-backed optimization manual for Mixtral Quark / INT4-FP8 MoE in vLLM. Use when an engineer needs to audit, debug, extend, or document Mixtral MoE, expert parallelism, FP8 / ModelOpt quantization, and EPLB in vLLM, which together form the nearest equivalent to Quark INT4-FP8 Mixtral serving.
$ install --global
skillsauthnpx skillsauth add BBuf/AI-Infra-Auto-Driven-SKILLS vllm-mixtral-quark-int4fp8-moe-optimizationInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: May 2, 2026, 2:43 AM263.4s1 file scanned
SKILL.md
- name:
- vllm-mixtral-quark-int4fp8-moe-optimization
- description:
- PR-backed optimization manual for Mixtral Quark / INT4-FP8 MoE in vLLM. Use when an engineer needs to audit, debug, extend, or document Mixtral MoE, expert parallelism, FP8 / ModelOpt quantization, and EPLB in vLLM, which together form the nearest equivalent to Quark INT4-FP8 Mixtral serving.
vLLM Mixtral Quark / INT4-FP8 MoE Optimization
Overview
This skill covers Mixtral MoE, expert parallelism, FP8 / ModelOpt quantization, and EPLB in vLLM, which together form the nearest equivalent to Quark INT4-FP8 Mixtral serving.
Evidence snapshot:
- vLLM mainline checked around
0f7be0f2f76814f80f9091220a5fbbb53912ad00 - Support status: partially supported or only adjacent architectures landed on current mainline
- Canonical PR notes:
references/pr-history.md - History mirrors:
model-pr-optimization-history/vllm/mixtral-quark-int4fp8-moe/README.zh.mdandREADME.en.md
Non-Negotiable Evidence Rule
Use skills/model-optimization/model-pr-diff-dossier/SKILL.md as the production bar.
Every PR cited for this family must be based on diff reading, not only PR titles.
Runtime Surfaces
vllm/vllm/model_executor/models/mixtral.pyvllm/vllm/model_executor/layers/fused_moe/layer.py
Current Main Summary
- vLLM has rich Mixtral MoE support, but not every Quark-branded checkpoint path is called out by name.
- The closest production evidence is the Mixtral fused-MoE, FP8, ModelOpt, and EPLB work already merged.
Key Landed PRs
- #2011
Mixtral 8x7B support: Initial Mixtral model-family support. - #2090
Optimize Mixtral with expert parallelism: Added early expert-parallel scaling instead of pure TP execution. - #2542
Fused MOE for Mixtral: Brought fused-MoE kernels into the Mixtral serving path. - #4527
Support MoE FP8 checkpoints for Mixtral: Added the first serious FP8 checkpoint path for Mixtral MoE. - #15961
Support ModelOpt quantization of Mixtral model: Extended the family to NVIDIA ModelOpt quantization flows. - #22842
Support EPLB for Mixtral Model: Added expert-parallel load balancing to the Mixtral family.
Open Optimization Items
- Re-run PR search before claiming new support beyond the checked mainline commit.
Validation Lanes
- FP16 / BF16 Mixtral baseline.
- FP8 or ModelOpt quantized checkpoint launch.
- Expert-parallel / EPLB topology checks on multi-GPU setups.
Related Skills
BBuf/sglang-humanize-review
development
VerifiedTrustedCommunity
Perform SGLang code review in the style of human maintainers by consulting the full non-agent PR review episode corpus from project start through the latest refresh (June 2026), including inline review threads, top-level PR comments, review submissions, original multilingual text, and multi-round discussions. Use when reviewing SGLang PRs, diffs, patches, or local changes for correctness, tests, performance, GPU/runtime risks, API compatibility, and maintainability.
531SKILL.mdUpdated May 21, 2026
BBuf/model-pr-history-knowledge
documentation
VerifiedTrustedCommunity
Use when an SGLang, vLLM, or TensorRT-LLM serving/model optimization task needs prior model-family PR evidence. Query and read the PR-driven history docs under model-pr-optimization-history before choosing source paths, fast paths, kernel/fusion ideas, regression risks, or validation lanes.
531SKILL.mdUpdated May 17, 2026
BBuf/vllm-sota-humanize-loop
development
VerifiedTrustedCommunity
Run an autonomous Humanize-governed vLLM SOTA performance loop for one LLM model: first perform the fixed fair vLLM/SGLang/TensorRT-LLM deployment search and benchmark, then start one RLCR loop that repeatedly decides the gap, profiles the current bottleneck, runs layer/kernel pipeline analysis, patches vLLM code, optionally uses ncu-report-skill for kernel evidence, and revalidates until vLLM matches or beats the best observed framework under the same workload and SLA.
423SKILL.mdUpdated May 27, 2026
BBuf/llm-pipeline-analysis
devops
VerifiedTrustedCommunity
Inspect LLM torch profiler traces at forward-pass, layer, and kernel level. Use when you need layer timings, anchor-kernel boundaries, representative kernel flows, or Perfetto time ranges.
423SKILL.mdUpdated May 21, 2026