Search results

[Usage]: How to use Env in yaml

usageHow to use vllmHow to use vllm

Status: Open.

#36465 In vllm-project/vllm;

· ciaoyizhen opened on Mar 9, 2026

[Bug]: Qwen3.5-27B fails to start with CPU KV cache offloading (<code>--kv_offloading_backend native</code>) while Qwen3-32B works fine

bugSomething isn't workingSomething isn't working

Status: Open.

#36463 In vllm-project/vllm;

· bigbear07 opened on Mar 9, 2026

[RFC]: vLLM IR Out-of-Tree (OOT) Kernel Registration

RFC

Status: Open.

#36459 In vllm-project/vllm;

· wxsIcey opened on Mar 9, 2026

[Bug]: Local GGUF path fails with "architecture qwen35 is not supported yet" even when --hf-config-path is provided

bugSomething isn't workingSomething isn't working

Status: Open.

#36456 In vllm-project/vllm;

· shba007 opened on Mar 9, 2026

[Bug]: Unable to run Qwen3.5 on RTX5090

bugSomething isn't workingSomething isn't working

Status: Open.

#36455 In vllm-project/vllm;

· YiJunSachs opened on Mar 9, 2026

[Performance]: vLLM v0.15.0 throughput regression compared to ROCm vLLM v0.14.0

performancePerformance-related issuesPerformance-related issues

rocmRelated to AMD ROCmRelated to AMD ROCm

Status: Open.

#36454 In vllm-project/vllm;

· Spurthi-Bhat-ScalersAI opened on Mar 9, 2026

[Bug]: Qwen3.5 AWQ models crash during inference on RTX 5090 (Blackwell) with Triton OOM in solve_tril despite successful model load

bugSomething isn't workingSomething isn't working

Status: Open.

#36450 In vllm-project/vllm;

· NEWbie0709 opened on Mar 9, 2026

[Bug]: qwen3.5-27b ValueError: Tokenizer class TokenizersBackendFast does not exist or is not currently imported.

bugSomething isn't workingSomething isn't working

Status: Open.

#36443 In vllm-project/vllm;

· xiaotianns opened on Mar 9, 2026

[Bug]: prefix cache bug happens when use w4a16 for GLM5

bugSomething isn't workingSomething isn't working

Status: Open.

#36441 In vllm-project/vllm;

· terenceucla opened on Mar 9, 2026

[Feature] Add energy consumption metrics to benchmark suite

Status: Open.

#36440 In vllm-project/vllm;

· hongping-zh opened on Mar 9, 2026

[Bug]: Responses API streaming emits tool call XML as <code>response.output_text.delta</code> instead of <code>response.function_call_arguments.delta</code> for non-harmony models

bugSomething isn't workingSomething isn't working

Status: Open.

#36435 In vllm-project/vllm;

· herve-ves opened on Mar 9, 2026

[Bug]: matryoshka need gpu-memory???

bugSomething isn't workingSomething isn't working

good first issueGood for newcomersGood for newcomers

Status: Open.

#36433 In vllm-project/vllm;

· ciaoyizhen opened on Mar 9, 2026