Skip to content

Issues: vllm-project/vllm

[Roadmap] vLLM Roadmap Q2 2025
#15735 opened Mar 29, 2025 by simon-mo
Open 9
[Roadmap] vLLM Release/CI/Performance Benchmark Q2 2025
#16284 opened Apr 8, 2025 by khluu
Open 3
[Usage] Qwen3 Usage Guide
#17327 opened Apr 28, 2025 by simon-mo
Open 43
Beta
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

1,778 Open 6,546 Closed
1,778 Open 6,546 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

[Installation]: compilation of flash-attn e4m3 kernels fails due to layout incompatibility in copy_traits.hpp installation Installation problems
#17597 opened May 2, 2025 by m-rds
1 task done
[Bug]: fp8 w8a8 quantized Qwen2.5-VL hits AssertionError bug Something isn't working
#17595 opened May 2, 2025 by cebtenzzre
1 task done
[Bug]: torch._inductor.exc.InductorError: TypeError: cannot pickle 'torch._C.DispatchKeySet' object bug Something isn't working torch.compile
#17593 opened May 2, 2025 by zou3519
1 task done
1
[Bug]: vLLM pre-commit hook doesn't work with git worktree bug Something isn't working
#17592 opened May 2, 2025 by zou3519
[Bug]: Cannot load Gemma3 27b QAT GGUF on RTX 5090 bug Something isn't working
#17587 opened May 2, 2025 by FremyCompany
1 task done
2
[Feature]: benchmarks for vllm, it should support OpenAI Chat Completions API feature request New feature or request
#17586 opened May 2, 2025 by xiayu98020214
1 task done
1
[Bug]: Mistral tool parser & streaming: corrupt tool_calls completions bug Something isn't working
#17585 opened May 2, 2025 by hibukipanim
1
[Bug]: Qwen3 FP8 on 0.8.5: type fp8e4nv not supported in this architecture. bug Something isn't working
#17581 opened May 2, 2025 by AlexBefest
1 task done
1
[Bug]: V1 Engine generates corrupt responses on large batch inference with long sequences and fails in seed control bug Something isn't working
#17580 opened May 2, 2025 by Cakeyan
1 task done
1
[Feature]: support for fp8 marlin with MoE feature request New feature or request
#17579 opened May 2, 2025 by ehartford
1 task done
3
[Usage]: Support Qwen3 inference in vLLM==0.8.5 with CUDA 11.8 (currently only vLLM==0.6.1.post1 works) usage How to use vllm
#17578 opened May 2, 2025 by YihongT
[Bug]: ValueError: The output_size of gate's and up's weight = 192 is not divisible by weight quantization block_n = 128. bug Something isn't working
#17569 opened May 2, 2025 by zzlgreat
1 task done
6
[Performance]: 单次请求速度30t/s ，并发请求只有1.5t/s performance Performance-related issues
#17568 opened May 2, 2025 by nvliajia
15
[Bug]: Function calling does not work with Mistral Small bug Something isn't working
#17557 opened May 1, 2025 by menardorama
1 task done
1
[Bug]: top_k: 0 in generation_config.json can't disable top-k sampling bug Something isn't working
#17553 opened May 1, 2025 by toslunar
1 task done
[Feature]: Support HF-style chat template for multi-modal data in offline chat feature request New feature or request good first issue Good for newcomers
#17551 opened May 1, 2025 by DarkLight1337
1 task done
@sfeng33
2
[Usage]: understanding the vllm's gpu_memory_utilization and cuda graph memory requirement usage How to use vllm
#17549 opened May 1, 2025 by mces89
1 task done
[Bug]: failed to run LMCache example for v0 bug Something isn't working
#17545 opened May 1, 2025 by gaowayne
1 task done
4
[Bug]: cached_get_processor is not actually cached bug Something isn't working
#17543 opened May 1, 2025 by Zazzle516
1 task done
3
[Performance]: Performance comparison for v1 engine and v0 engine performance Performance-related issues
#17540 opened May 1, 2025 by hustxiayang
1 task done
1
[Usage]: CUDA Error with Qwen3-32B Model When Processing larger tokens it leads to model went to non responsive condition / stability concerns usage How to use vllm
#17534 opened May 1, 2025 by nskpro-cmd
1 task done
1
[Bug]: AttributeError: 'MultiprocExecutor' object has no attribute 'workers' when VLLM_USE_V1=1 on rocm platform serve deepseek-r1 671B bug Something isn't working rocm Related to AMD ROCm
#17533 opened May 1, 2025 by GuoxiangZu
1 task done
@hongxiayang
7
[Bug]: Bad requests are not captured as traces bug Something isn't working
#17528 opened May 1, 2025 by frzifus
1 task done
[Bug]: Training with vllm not supports Qwen3 bug Something isn't working
#17527 opened May 1, 2025 by fly-dragon211
1 task done
[Bug]: '_OpNamespace' '_C' object has no attribute 'rms_norm' on docker environment bug Something isn't working
#17526 opened May 1, 2025 by erkintelnyx
1 task done
1
ProTip! Find all open issues with in progress development work with linked:pr.