dit-npu-fusion-analysis
Skill7.8k starsChanged 3 days agoApache-2.0View on GitHub
Route Ascend NPU diffusion model and operator optimization to MindIE-SD skills. Use for NPU DiT fusion, profiling, kernel/pattern work and related parallelism, VAE or host bottlenecks. Supply Omni context and load the applicable external skill. CUDA/ROCm tuning, pure AR optimization and runner-version upgrades use their own workflows.
Adopted by
Steady: 0 new projects took it in the last 4 weeks, none the 4 weeks before.thin data
Who's talking about vllm-project/vllm-omni
- verl-project/verl“[RFC] Multi-modal Generation RL 2026Q2 Roadmap”GitHub · Mar 26, 2026 · 16 reactions
- vllm-project/vllm“[Model][VLM] Add Qwen2.5-Omni model support (end-to-end full support)”GitHub · Apr 9, 2025 · 11 reactions
- verl-project/verl“[RFC] Support Qwen-Image Flow-GRPO Training based on vLLM-Omni”GitHub · Dec 22, 2025 · 10 reactions
- verl-project/verl“[RFC] Support Diffusion Generative Models and Async Reward during Rollout”GitHub · Dec 17, 2025 · 8 reactions
- vllm-project/vllm“[Migration] Migrate GGUF quantization support to plugin”GitHub · Apr 12, 2026 · 7 reactions
The biggest 5 of 51 mentions linking to the repository that our crawl found. Each line is their own headline.
See all 51
Talked about on GitHub and Hacker News (51 times).
On GitHub · 50
- Mar 26, 2026[RFC] Multi-modal Generation RL 2026Q2 Roadmap — verl-project/verl16 reactions · 21 comments
- Apr 9, 2025[Model][VLM] Add Qwen2.5-Omni model support (end-to-end full support) — vllm-project/vllm11 reactions · 11 comments
- Dec 22, 2025[RFC] Support Qwen-Image Flow-GRPO Training based on vLLM-Omni — verl-project/verl10 reactions · 8 comments
- Dec 17, 2025[RFC] Support Diffusion Generative Models and Async Reward during Rollout — verl-project/verl8 reactions · 3 comments
- Apr 12, 2026[Migration] Migrate GGUF quantization support to plugin — vllm-project/vllm7 reactions · 16 comments
45 more on GitHub
- Aug 14, 2026[RFC] Tracking v1 upgrade for verl-omni v0.3.0: remaining recipe coverage — verl-project/verl-omni5 reactions · 19 comments
- Apr 29, 2026在vLLM-Omni中支持SenseNova-U1 — OpenSenseNova/SenseNova-U14 reactions · 6 comments
- Dec 7, 2025RL post training for omni-modality models — ByteDance-Seed/VeOmni4 reactions · 2 comments
- Oct 16, 2026Add MOSS-Image Base and SFT recipe — vllm-project/recipes2 reactions · 2 comments
- Aug 3, 2026[Draft] [Bugfix] Clone safetensors tensors before H2D copy on Ascend NPU to fix DMA hang — vllm-project/vllm2 reactions · 3 comments
- Jun 17, 2026[ROCm][Perf] Fix Qwen3-vLLM audio encoder TP when heads are not divisible by TP size — vllm-project/vllm2 reactions · 14 comments
- Jun 16, 2026[RFC]: Qwen3-Omni RL Support on FSDP with vLLM-Omni rollout — verl-project/verl-omni2 reactions · 1 comment
- May 11, 2026[Metrics] Scope unregister_vllm_metrics() to strictly "vllm:" metrics — vllm-project/vllm2 reactions · 10 comments
- Apr 14, 2026[RFC]: Support Mooncake Based ECConnector for EPD — vllm-project/vllm2 reactions · 3 comments
- Apr 7, 2026Collaboration: Integrate VoxCPM into vllm-omni — a710128/nanovllm-voxcpm2 reactions · 4 comments
- Feb 25, 2026vllm-omni is supporting the model inference — bytedance/mammothmoda2 reactions · 3 comments
- Jan 10, 2026[Models] Add `SharedFusedMoE` support to Qwen3MoE — vllm-project/vllm2 reactions · 2 comments
- Jan 9, 2026[Feature]: RL Integration: Support for the veRL + veOmni + vLLM-Omni stack. — JiusiServe/vllm-omni2 reactions · 3 comments
- Dec 20, 2025[RFC]: Optional Hidden States Transfer in `KVConnectorBase_V1` to Eliminate Decoder Prefix-Prefill in P/D Disaggregation — vllm-project/vllm2 reactions · 2 comments
- Nov 12, 2026Add opt-in single RTX 4090 hardware profile — vllm-project/recipes1 reaction · 1 comment
- Nov 10, 2026Sync recipes from vllm-project/vllm-omni@d8751cd14a85 — vllm-project/recipes1 reaction · 2 comments
- Oct 29, 2026Sync recipes from vllm-project/vllm@9a0f9b45df86, vllm-project/vllm-omni@46766eb27f3e — vllm-project/recipes1 reaction · 2 comments
- Oct 5, 2026feat: upgrade vllm from 0.30.0 to 0.31.0 for CUDA and ROCm backends — gpustack/runner1 reaction · 2 comments
- Oct 2, 2026[Model] Add Nemotron 3.5 ASR transcription support — vllm-project/vllm1 reaction · 9 comments
- Sep 23, 2026feat: upgrade vllm from 0.29.0 to 0.30.0 for CUDA and ROCm backends — gpustack/runner1 reaction · 2 comments
- Sep 21, 2026feat(voice): add VoxCPM as a local TTS provider — moltis-org/moltis1 reaction · 1 comment
- Sep 3, 2026Add FastH3 VSA recipe — vllm-project/recipes1 reaction · 1 comment
- Jul 23, 2026[Bug]: HunyuanImage-3.0-Instruct (AR+DiT) startup hangs when waitting engine core process start on Ascend A3 — vllm-project/vllm-ascend1 reaction · 4 comments
- Jul 21, 2026[vLLM backend][Feature Request] Support vLLM-Omni for native audio/video input — triton-inference-server/server1 reaction · 3 comments
- Jul 9, 2026[RFC] Refactor experimental step execution support by reusing vLLM-Omni interfaces — verl-project/verl-omni1 reaction · 7 comments
- Jun 16, 2026Add user guide for v2v and transfer for vllm-omni — NVIDIA/cosmos1 reaction · 5 comments
- Jun 15, 2026[Quantization] Extend ModelOpt mixed precision and NVFP4 runtime formats — vllm-project/vllm1 reaction · 6 comments
- Jun 12, 2026[Draft]Support Megatron LoRA adapter export for rollout — verl-project/verl1 reaction · 1 comment
- Jun 12, 2026[Draft]Support Qwen3-Omni Megatron LoRA rollout — verl-project/verl-omni1 reaction · 1 comment
- Jun 11, 2026310p 部署 Qwen3-TTS — vllm-project/vllm-ascend1 reaction · 11 comments
- Jun 9, 2026docs(diffusion): add Qwen-Image v0.20→v0.22 H200 retro — hsliuustc0106/vllm-omni-cookbook1 reaction · 6 comments
- Jun 6, 2026[Kernel] FlashInfer FP8 scaled_mm: restore N-D output shape — vllm-project/vllm1 reaction · 3 comments
- May 22, 2026[feat]: Cosmos3 port (Tier-A scaffold; SSIM + Tier-B deferred pending weight publish) — hao-ai-lab/FastVideo1 reaction · 8 comments
- May 20, 2026[RFC] Q3 Road Map (welcome contributions) — verl-project/verl-omni1 reaction · 30 comments
- May 15, 2026Add extra_labelnames to PrometheusStatLogger for extensibility — vllm-project/vllm1 reaction · 3 comments
- May 14, 2026Mxfp4 support — yiliu30/vllm-omni-fork1 reaction
- Apr 20, 2026Custom pipeline class fix — knlnguyen1901/vllm-omni1 reaction
- Mar 18, 2026Add SoulX-Duplug turn-serving support — nemoramo/vllm-omni1 reaction · 1 comment
- Jan 12, 2026vLLM/vLLM-Omni integration — NAVER-Cloud-HyperCLOVA-X/OmniServe1 reaction
- Jan 9, 2026[Feature][Community]: Model Support: Add support for Bagel and HunyuanImage3.0 models in disaggregated setups. — JiusiServe/vllm-omni1 reaction
- Oct 6, 2026[Feature] Support Irodori-TTS Japanese-focused TTS in SGLang-Omni — sgl-project/sglang-omni0 reactions
- Aug 5, 2026[Feature]: Diffusion kv block manager — JiusiServe/vllm-omni0 reactions
- Aug 2, 2026[Bug]: Shared prompt_token_ids causes stale M-RoPE positions — vllm-project/vllm0 reactions · 1 comment
- Apr 1, 2026[RFC]: HunyuanImage DIT Fa3-fp8 quantization — JiusiServe/vllm-omni0 reactions · 1 comment
- Feb 27, 2026Support realtime API for audio streaming — EricLBuehler/mistral.rs0 reactions · 1 comment
On Hacker News · 1
- Dec 2, 2025VLLM-Omni: A framework for efficient model inference with Omni-modality models2 points · 1 comment
In this repository29 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- 3d-animation-short-generatorSkill
- brand-promo-video-generatorSkill
- co-op-game-intro-generatorSkill
- h3-prompt-writingSkill
- handdrawn-live-video-generatorSkill
- minimalist-product-ad-generatorSkill
- music-video-subtitle-generatorSkill
- paper-collage-explainer-generatorSkill
- papercraft-stop-motion-explainerSkill
- vllm-omni-h3-character-replacementSkill
- add-diffusion-modelSkill
- add-omni-modelSkill
- add-recipeSkill
- add-tts-modelSkill
- diffusion-accuracy-testSkill
- diffusion-perf-optSkill
- diffusion-perf-testSkill
- find-simplificationsSkill
- omni-perf-ceilingSkill
- omni-perf-optSkill
- precheck-prSkill
- production-add-diffusion-modelSkill
- quantizationSkill
- review-prSkill
- vllm-omni-npu-model-runner-upgradeSkill
- vllm-omniSkill
- vllm-omni-testSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 5 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 15
Use the task's MindIE-SD checkout, or an existing sibling MindIE-SD checkout.
- Very specific: line 18
Read its AGENTS.md and .agents/README.md, then the selected SKILL.md and required references.
- Distinctive: line 23
If GitCode is unavailable, use the official GitHub dev skill index and GitHub dev tree.
- Distinctive: line 26
Follow the actual index if names changed.
- Distinctive: line 27
If required guidance is unavailable, name the missing resource and continue useful Omni source/configuration inspection.
- Distinctive: line 61
Keep Omni, MindIE-SD and installed provider sources distinct.
- Distinctive: line 75
Do not require a parallel Omni evidence schema.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Ascend Diffusion Optimization Router
MindIE-SD owns optimization methods, fusion analysis and NPU acceptance guidance. This entry keeps its existing name for discovery. Do not maintain a second operator catalogue or validation protocol in Omni. Routing means reading the selected instructions in this task, not messaging a task or starting an agent.
Locate and load
- Use the task's MindIE-SD checkout, or an existing sibling
MindIE-SDcheckout. Record its revision and relevant dirty changes; do not switch or reset its branch to route a task. - Read its
AGENTS.mdand.agents/README.md, then the selectedSKILL.mdand required references. Resolve paths against that checkout. Naming a skill does not mean its instructions have been loaded. - Without a checkout, use the official dev skill index and dev tree. If GitCode is unavailable, use the official GitHub dev skill index and GitHub dev tree. Retrieve only the needed instructions and record their source and revision or access date. Follow the actual index if names changed; do not rely on unpublished additions.
- If required guidance is unavailable, name the missing resource and continue useful Omni source/configuration inspection. Do not claim the SD workflow ran or recreate it here. Reading guidance requires neither installation nor an NPU.
Select the applicable entry
Confirm these paths in the selected SD index; they are relative to its root. A focused operator task need not repeat model onboarding.
| Task | MindIE-SD skill under .agents/skills/ |
| Model optimization with an unknown bottleneck | model-auto-optimization/SKILL.md |
| Known bottleneck needing domain selection | performance-optimization/SKILL.md |
| Profile collection / diagnosis | profiling-collect/SKILL.md / profiling-analyze/SKILL.md |
| Fusion scope, benefit or reusable implementation feasibility | fusion-scope-analyze/SKILL.md |
| Kernel implementation / tuning | operator-dev/SKILL.md |
| Pattern / graph implementation | pattern-dev/SKILL.md / aclgraph-dev/SKILL.md |
| Framework enablement / missing adapters | framework-integration/SKILL.md |
| DiT compute / communication | dit-perf-opt/SKILL.md / dit-parallel-opt/SKILL.md |
| VAE / host optimization | vae-opt/SKILL.md / host-opt/SKILL.md |
| MindIE-SD repository development | dev-workflow/SKILL.md, then its implementation route |
For numerical/performance acceptance, read accuracy-gate/SKILL.md and
perf-gate/SKILL.md from that skill root. Those skills own the criteria;
historical evidence applies only within its recorded scope.
Supply context and preserve ownership
Supply the goal and authorized analysis/implementation scope, Omni revision and dirty diff, model/checkpoint and active pipeline/symbols, hardware/provider versions, launch/request commands, workload/dtype/parallel/cache/graph settings, profiles/baselines, precision constraints and output directory. Keep missing fields unknown; the selected skill determines how to proceed.
Keep Omni, MindIE-SD and installed provider sources distinct. Model files resolve against Omni; vendor implementation resolves against its actual provider. Changes to Omni stay there; MindIE-SD changes follow its development workflow. External instructions do not expand user authorization or repository scope.
For Omni changes, read vllm-omni-test for repository tests/CI and diffusion-accuracy-test when adding an Omni accuracy case. Production qualification stays with production-add-diffusion-model. These checks complement the SD acceptance criteria.
Return the selected skill paths/revision, routing reason, changes, actual checks and artifacts using the SD workflow's existing output format. Distinguish task completion from optimization acceptance; identify unread/unexecuted dependencies. Do not require a parallel Omni evidence schema.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

