Skip to content

Actions: vllm-project/vllm

Actions

Label issues based on keywords

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
issue_autolabel.yml
2,500+ workflow runs
2,500+ workflow runs

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
[Bug]: Runtime-varying tl.constexpr params force Triton kernel recompiles on the serving path (unified attention MAX_MM_RANGES et al.) Label issues based on keywords #14243: Issue #48650 opened by sbeurnier
7s
7s
[Bug]: deepseek_v4 parser: reply without </think> routes the whole answer to reasoning_content (content empty), trailing EOS not stripped Label issues based on keywords #14242: Issue #48645 edited by fank
8s
8s
[Bug]: deepseek_v4 parser: reply without </think> routes the whole answer to reasoning_content (content empty), trailing EOS not stripped Label issues based on keywords #14241: Issue #48645 opened by fank
10s
10s
[Feature]: Support fast weight updates from disk Label issues based on keywords #14240: Issue #48644 opened by SumanthRH
6s
6s
[RFC]: Read-only NIXL GDS connector for filesystem KV-cache loads Label issues based on keywords #14239: Issue #48504 edited by maheshrbapatu
5s
5s
[RFC]: Read-only NIXL GDS connector for filesystem KV-cache loads Label issues based on keywords #14238: Issue #48504 edited by maheshrbapatu
5s
5s
[Feature][KV-offloading]: Relax Offloading Block Size asserts Label issues based on keywords #14237: Issue #48635 opened by varun-sundar-rabindranath
2m 2s
2m 2s
NixlPushMode (WRITE) Roadmap - Reliability Issue Inventory Label issues based on keywords #14236: Issue #48633 opened by NickLucche
7m 26s
7m 26s
[RFC]: Context-length-aware speculative token scheduling — extending num_speculative_tokens_per_batch_size with a context-length axis Label issues based on keywords #14235: Issue #48627 opened by seongyun1104
6s
6s
[XPU] Dockerfile.xpu: LD_LIBRARY_PATH misses /opt/venv/lib -> torch sees 0 XPU devices in the built image Label issues based on keywords #14234: Issue #48625 opened by marcorigodanzo
28m 54s
28m 54s
[Bug]: Ministral3 RoPE scaling ignores YaRN mscale and nested Llama 4 fields Label issues based on keywords #14233: Issue #48621 edited by oliverholworthy
7m 23s
7m 23s
[Bug]: Ministral3 RoPE scaling ignores YaRN mscale and nested Llama 4 fields Label issues based on keywords #14232: Issue #48621 opened by oliverholworthy
19m 30s
19m 30s
V1 EngineCore: SIGTERM teardown runs two full-heap gen-2 GC passes (~4-7 s) to collect a few hundred objects Label issues based on keywords #14231: Issue #48620 opened by arjmandi
32m 10s
32m 10s
[RFC]: Derender Endpoints to Provide Detokenization for Disaggregated Serving Label issues based on keywords #14230: Issue #42729 edited by hickeyma
25m 57s
25m 57s
[Feature]: [Batch-Invariant-Kernels] GDN_ATTN does not support batch-invariant mode for Qwen3.5/Qwen3.6 GDN models Label issues based on keywords #14229: Issue #48613 edited by cm2435
45s
45s
[Feature]: [Batch-Invariant-Kernels] GDN_ATTN does not support batch-invariant mode for Qwen3.5/Qwen3.6 GDN models Label issues based on keywords #14228: Issue #48613 opened by cm2435
5m 30s
5m 30s
[Bug]: #47327 dense-MHA split breaks FlashMLA sparse: OOB write in top-k index conversion, corrupted fp8_ds_mla context gather Label issues based on keywords #14227: Issue #48611 opened by drakosha
11m 53s
11m 53s
[Bug]: DeepSeekv4 flash on RTXpro6000 vllm benchmark input 8192 output 1024 encounter a err Label issues based on keywords #14226: Issue #48602 edited by iceCreeam
5m 38s
5m 38s
[Bug]: DeepSeekv4 flash on RTXpro6000 vllm benchmark input 8192 output 1024 encounter a err Label issues based on keywords #14225: Issue #48602 edited by iceCreeam
28s
28s
[Bug]: MiniMax M3 Triton path misinterprets token-major top-k buffer Label issues based on keywords #14224: Issue #48603 opened by RedHeartSecretMan
4m 19s
4m 19s
[Bug]: DeepSeekv4 flash on RTXpro6000 vllm benchmark input 8192 output 1024 encounter a err Label issues based on keywords #14223: Issue #48602 opened by iceCreeam
12s
12s
[Bug]: v0.25.0 torchcodec not compatible with cu129 wheel Label issues based on keywords #14222: Issue #48592 edited by VikramRangarajan
16m 53s
16m 53s
[Bug]: v0.25.0 torchcodec not compatible with cu129 wheel Label issues based on keywords #14221: Issue #48592 opened by VikramRangarajan
7m 29s
7m 29s
还不支持[Usage]: vllm 0.25版本在tp=8并行的时候走dflash 和dspark 命中 attention后端的时候出现错误报错定位 CUDA error (/workspace/.deps/vllm-flash-attn-src/hopper/flash_api.cpp:697): invalid configuration argument 8个worker(TP0-TP7) Label issues based on keywords #14220: Issue #48591 opened by cuihangbin
24m 16s
24m 16s
[Bug]: LoRA lora_expand Triton kernel outputs NaN on Hopper (sm_90) with block_n=128, producing garbled LoRA output Label issues based on keywords #14219: Issue #48590 edited by n0whereRuoxi
12m 32s
12m 32s