Search results

[RFC]: Support Hunyuan Image Model in vLLM-Omni (and Enable Batch Scheduling)

Status: Open.

#653 In vllm-project/vllm-omni;

· Semmer2 opened on Jan 5, 2026

[Bug]: When will it support token2wav streaming output，log：we only use the token2wav.process_chunk for the whole sequence. The stream mode will be implemented in the future.

bugSomething isn't workingSomething isn't working

Status: Open.

#652 In vllm-project/vllm-omni;

· YuiGao opened on Jan 5, 2026

[Feature]: Profiler Trace Size Optimization

Status: Open.

#651 In vllm-project/vllm-omni;

· gcanlin opened on Jan 5, 2026

[RFC]: Support torch profiler support for diffusion

Status: Open.

#650 In vllm-project/vllm-omni;

· lishunyang12 opened on Jan 5, 2026

[Bug]: Qwen3-Omni-30B-A3B-Instruct，send single request， the generated audio contains noise when inputting text + 4 images.

bugSomething isn't workingSomething isn't working

Status: Open.

#649 In vllm-project/vllm-omni;

· yenuo26 opened on Jan 5, 2026

[RFC]: PipeFusion Implementation in vLLM-Omni

Status: Open.

#647 In vllm-project/vllm-omni;

· mxuax opened on Jan 5, 2026

[Bug]: Qwen3-Omni-30B-A3B-Instruct，send single request， "e2e_total_time_ms" of [Summary] and client-side statistical latency show a significant discrepancy.

bugSomething isn't workingSomething isn't working

Status: Open.

#646 In vllm-project/vllm-omni;

· yenuo26 opened on Jan 5, 2026

[Bug]: Unnecessary Error logs in offline inference of Qwen3-omni

bugSomething isn't workingSomething isn't working

Status: Open.

#643 In vllm-project/vllm-omni;

· Gaohan123 opened on Jan 5, 2026

[Bug]: Qwen 2.5 Omni examples facing OOM problem for both online/offline in H800

bugSomething isn't workingSomething isn't working

Status: Open.

#642 In vllm-project/vllm-omni;

· congw729 opened on Jan 5, 2026

[Bug]: Qwen3-Omni-30B-A3B-Instruct，Input 4 audio files + text, and the instance reported a RuntimeError after sending the request: "split_with_sizes expects split_sizes to sum exactly to 1000".

bugSomething isn't workingSomething isn't working

Status: Open.

#640 In vllm-project/vllm-omni;

· yenuo26 opened on Jan 5, 2026

[Bug]: AttributeError: dual_attention_layers in Stable Diffusion 3

bugSomething isn't workingSomething isn't working

Status: Open.

#636 In vllm-project/vllm-omni;

· natureofnature opened on Jan 5, 2026

[RFC]: Supports different numbers of stages

Status: Open.

#632 In vllm-project/vllm-omni;

· amy-why-3459 opened on Jan 5, 2026