From 467c815db6df73f0133e3a66f8119a75a776a8ae Mon Sep 17 00:00:00 2001
From: drizzlezyk <drizzlezyk@163.com>
Date: Tue, 17 Mar 2026 23:28:04 +0800
Subject: [PATCH] [CI] expand issue labeler rules for feature/model triage
 (#7356)

- Replace minimal label rules with a comprehensive keyword-based issue
labeler taxonomy.
- Add grouped labels for core features and advanced capabilities to
improve issue routing.
- Expand model-related matching for LLM, multimodal generation,
multimodal understanding, audio, and omni scenarios.
- Add/normalize regex patterns for common model families (DeepSeek,
Kimi, GLM, Qwen, 310p, etc.) to increase auto-label coverage and
consistency.
### What this PR does / why we need it?
- Expands `.github/issue-labeler.yml` from a minimal set of rules to a
richer keyword-based labeling configuration.
- Adds grouped label dimensions for:
- Core features (e.g., PD disaggregation, KV cache pool, ACLGraph, async
scheduler, CPU binding, quantization)
- Advanced features (e.g., long sequence, DPC/PCP, MTP/speculative
decode)
- Model categories (LLM, multimodal generation, multimodal
understanding, audio, omni, etc.)
  - Specific model families (e.g., DeepSeek, Kimi, GLM, Qwen, 310p)
- Improves automatic issue triage accuracy and reduces manual label
maintenance effort.
- Makes issue categorization more consistent for maintainers and
contributors.

Why needed:
- Existing labeler rules were too limited and could not adequately cover
current feature/model issue distribution.
- Broader and more structured matching helps faster routing,
prioritization, and ownership assignment.

Fixes #N/A

### Does this PR introduce _any_ user-facing change?
- No runtime/API user-facing changes.
- This PR only updates GitHub issue automation rules.

### How was this patch tested?
- Performed static validation and review of `.github/issue-labeler.yml`
structure and regex entries.
- Verified that rule groups and label keys are correctly formatted for
GitHub issue labeler consumption.
- Confirmed that legacy minimal rules were replaced by expanded taxonomy
without syntax-breaking YAML changes.
- No unit/e2e tests were added because this is repository automation
configuration (GitHub labeling rules) rather than application runtime
logic.
- vLLM version: v0.17.0
- vLLM main:
https://github.com/vllm-project/vllm/commit/4034c3d32e30d01639459edd3ab486f56993876d

---------

Signed-off-by: drizzlezyk <drizzlezyk@163.com>
---
 .github/issue-labeler.yml | 84 ++++++++++++++++++++++++++++++++++-----
 1 file changed, 73 insertions(+), 11 deletions(-)

diff --git a/.github/issue-labeler.yml b/.github/issue-labeler.yml
index f6b1679d..bbb2e183 100644
--- a/.github/issue-labeler.yml
+++ b/.github/issue-labeler.yml
@@ -1,12 +1,74 @@
-310p:
-    - '/(310p)/i'
-glm5:
-    - '/(glm5)/i'
-qwen3.5:
-    - '/(qwen3.5)/i'
+core-features:
+  - '/((pd|(prefill[- ]?decode))\s+disaggregation|kv cache pool|aclgraph|async scheduler|cpu binding|quantization)/i'
+pd-disaggregation:
+  - '/((pd|(prefill[- ]?decode))\s+disaggregation)/i'
+kv-cache-pool:
+  - '/(kv cache pool)/i'
+aclgraph:
+  - '/(aclgraph)/i'
+async-scheduler:
+  - '/(async scheduler)/i'
+cpu-binding:
+  - '/(cpu binding)/i'
+quantization:
+  - '/(quantization)/i'
+advanced_features:
+  - '/(long sequence|dpc|pcp|mtp|speculative decode)/i'
+long-seq:
+  - '/(long sequence|dpc|pcp)/i'
+mtp/speculative-decode:
+  - '/(mtp|speculative decode)/i'
+eplb:
+  - '/(eplb)/i'
+llm-model:
+  - '/(deepseek[- ]*(r1|v3(\.2)?)\S*|(kimi k2|kimik2|kimi-k2)(?!\.5)|glm5|qwen3-(?:235b|480b)\S*|Qwen3-(?:32B|8B|30B)\S*|qwen3 next|glm\s*4\.(?![^v\s]*v)\S*)/i'
 deepseek:
-    - '/(deepseek)/i'
-kimi k2:
-    - '/(kimi k2|kimik2)(?!\.5)/i'
-kimi k2.5:
-    - '/(kimi k2.5|kimik2.5)/i'
\ No newline at end of file
+  - '/(deepseek[- ]*(r1|v3(\.2)?)\S*)/i'
+kimi-k2:
+  - '/((kimi k2|kimik2|kimi-k2)(?!\.5))/i'
+kimi-k2.5:
+  - '/((kimi k2\.5|kimik2\.5|kimi-k2\.5))/i'
+glm5:
+  - '/(glm5)/i'
+qwen3-moe:
+  - '/(Qwen3-(?:235B|480B)\S*)/i'
+qwen3-dense:
+  - '/(Qwen3-(?:32B|8B|30B)\S*)/i'
+qwen3-next:
+  - '/(qwen3-next)/i'
+glm-4:
+  - '/(glm\s*4\.(?![^v\s]*v)\S*)/i'
+multi-modality-generate:
+  - '/(seedance\S*|seedream\S*|wan\S*|hunyuan\S*|fLux\S*|kimi k2\.5|kimi-k2\.5|kimik2\.5|minimax\S*|qwen-image\S*)/i'
+seedance:
+  - '/(seedance\S*)/i'
+seedream:
+  - '/(seedream\S*)/i'
+wan:
+  - '/(wan\S*)/i'
+hunyuan:
+  - '/(hunyuan\S*)/i'
+fLux:
+  - '/(fLux\S*)/i'
+qwen-image:
+  - '/(qwen-image\S*)/i'
+minimax:
+  - '/(minimax\S*)/i'
+multimodal_understanding:
+  - '/(glm-?4\.\S*v\b|qwen3\.5\S*|deepseek-ocr\S*)/i'
+glm-4v:
+  - '/(glm-?4\.\S*v\b)/i'
+qwen-3.5:
+  - '/(qwen3\.5\S*)/i'
+deepseek-ocr:
+  - '/(deepseek-ocr\S*)/i'
+audio-model:
+  - '/(qwen3-tts\S*)/i'
+omni-model:
+  - '/(qwen3-Omni\S*)/i'
+multimodal-unified-autoregress:
+  - '/(hunyuan\S*|emu\S*)/i'
+paddle:
+  - '/(paddle\S*)/i'
+310p:
+  - '/(310p\S*)/i'