-
Notifications
You must be signed in to change notification settings - Fork 1.4k
All issues
Issue creation is restricted in this repository
- #5069 · amy-why-3459 opened
on Jul 13, 2026 24 - #5700 · david6666666 opened
on Aug 3, 2026 5 - #4901 · hsliuustc0106 opened
on Jul 5, 2026 8
Issues
is:issue state:open
is:issue state:open
Search results
- Status: Open.#5731 In vllm-project/vllm-omni;
[Bug]: Per-stage engine knobs have a top-level CLI flag that stamps every stage
bugSomething isn't workingSomething isn't workingStatus: Open.#5729 In vllm-project/vllm-omni;[Bug]: CLI stage-overrides don't validate and reject invalid knobs
bugSomething isn't workingSomething isn't workingStatus: Open.#5728 In vllm-project/vllm-omni;[RFC]: OpenAI-compliant /v1/audio/transcriptions with word-level alignment
frontendcode related to entrypointcode related to entrypointomnicode related to omni modelscode related to omni modelsStatus: Open.#5722 In vllm-project/vllm-omni;- Status: Open.#5707 In vllm-project/vllm-omni;
[Roadmap] MiniMax-H3 follow-up: accuracy/performance CI, quantization, and optimizations
CI/CDcodes related to changes to CI/CDcodes related to changes to CI/CDdiffusioncodes related to diffusion modelscodes related to diffusion modelsHardware Pluginsupport different hardware beyond cudasupport different hardware beyond cudahelp wantedExtra attention is neededExtra attention is neededKernel optimizationCodes related to optimize kernel execution to improve hardware utilizationCodes related to optimize kernel execution to improve hardware utilizationNPUPR related to Ascend NPUPR related to Ascend NPUquantizationCode related to quantizationCode related to quantizationROCmPR related to AMD hardwarePR related to AMD hardwarexpuPR related to XPUPR related to XPUStatus: Open.#5700 In vllm-project/vllm-omni;[Feature]: ROCm H3 support
diffusioncodes related to diffusion modelscodes related to diffusion modelsROCmPR related to AMD hardwarePR related to AMD hardwareStatus: Open.#5697 In vllm-project/vllm-omni;[Bug]: Nightly CI, Wan-AI/Wan2.2-I2V-A14B-Diffusers, performance metrics regressed by more than 10% compared to the baseline in some scenarios
bugSomething isn't workingSomething isn't workingci-failureCI failure issues, expected to be solved asapCI failure issues, expected to be solved asaphigh priorityhigh priority issue, needs to be done asaphigh priority issue, needs to be done asapStatus: Open.#5694 In vllm-project/vllm-omni;[Bug]: audio_ttfp goodput SLO is rejected by the CLI and not applied by Omni benchmark metrics
bugSomething isn't workingSomething isn't workinglow prioritylow priority issuelow priority issueStatus: Open.#5668 In vllm-project/vllm-omni;[Bug]: k2-fsa/OmniVoice single word prompt produces white noise intermittently
bugSomething isn't workingSomething isn't workinglow prioritylow priority issuelow priority issueStatus: Open.#5659 In vllm-project/vllm-omni;[Bug]: AutoRound W4A16 E2E load fails for Qwen2.5/Qwen3-Omni: qweight vs RowParallelLinear.weight
bugSomething isn't workingSomething isn't workinglow prioritylow priority issuelow priority issuequantizationCode related to quantizationCode related to quantizationStatus: Open.#5652 In vllm-project/vllm-omni;[Feature]: Support request-level batching for Wan2.2 pipelines
diffusioncodes related to diffusion modelscodes related to diffusion modelsenhancementNew feature or requestNew feature or requesthelp wantedExtra attention is neededExtra attention is neededStatus: Open.#5649 In vllm-project/vllm-omni;