[Bug]: vLLM Omni import-time Request rebind breaks plain Qwen3 inference

[timzsu]

Your current environment

The output of python collect_env.py

Collecting environment information...
uv is set
INFO 06-19 09:56:27 [patch.py:252] NVFP4 W4A4 weight_scale NaN-clamp: installed.
==============================
        System Info
==============================
OS                           : Ubuntu 22.04.5 LTS (x86_64)
GCC version                  : (Ubuntu 11.4.0-1ubuntu1~22.04.3) 11.4.0
Clang version                : Could not collect
CMake version                : Could not collect
Libc version                 : glibc-2.35

==============================
       PyTorch Info
==============================
PyTorch version              : 2.11.0+cu130
Is debug build               : False
CUDA used to build PyTorch   : 13.0
ROCM used to build PyTorch   : N/A

==============================
      Python Environment
==============================
Python version               : 3.12.12 (main, Dec  9 2025, 19:02:36) [Clang 21.1.4 ] (64-bit runtime)
Python platform              : Linux-6.8.0-106-generic-x86_64-with-glibc2.35

==============================
       CUDA / GPU Info
==============================
Is CUDA available            : True
CUDA runtime version         : 11.5.119
CUDA_MODULE_LOADING set to   : 
GPU models and configuration : 
GPU 0: NVIDIA RTX 6000 Ada Generation
GPU 1: NVIDIA RTX 6000 Ada Generation

Nvidia driver version        : 580.126.09
cuDNN version                : Could not collect
HIP runtime version          : N/A
MIOpen runtime version       : N/A
Is XNNPACK available         : True

==============================
          CPU Info
==============================
Architecture:                            x86_64
CPU op-mode(s):                          32-bit, 64-bit
Address sizes:                           48 bits physical, 48 bits virtual
Byte Order:                              Little Endian
CPU(s):                                  24
On-line CPU(s) list:                     0-23
Vendor ID:                               AuthenticAMD
Model name:                              AMD Ryzen 9 9900X 12-Core Processor
CPU family:                              26
Model:                                   68
Thread(s) per core:                      2
Core(s) per socket:                      12
Socket(s):                               1
Stepping:                                0
CPU max MHz:                             5658.0000
CPU min MHz:                             600.0000
BogoMIPS:                                8782.92
Flags:                                   fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good amd_lbr_v2 nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba perfmon_v2 ibrs ibpb stibp ibrs_enhanced vmmcall fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local user_shstk avx_vnni avx512_bf16 clzero irperf xsaveerptr rdpru wbnoinvd cppc amd_ibpb_ret arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif x2avic v_spec_ctrl vnmi avx512vbmi umip pku ospke avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq rdpid bus_lock_detect movdiri movdir64b overflow_recov succor smca fsrm avx512_vp2intersect flush_l1d srso_user_kernel_no
Virtualization:                          AMD-V
L1d cache:                               576 KiB (12 instances)
L1i cache:                               384 KiB (12 instances)
L2 cache:                                12 MiB (12 instances)
L3 cache:                                64 MiB (2 instances)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-23
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; Enhanced / Automatic IBRS; IBPB conditional; STIBP always-on; PBRSB-eIBRS Not affected; BHI Not affected
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

==============================
Versions of relevant libraries
==============================
[pip3] ema-pytorch==0.7.9
[pip3] flashinfer-python==0.6.11.post2
[pip3] numpy==2.3.5
[pip3] nvidia-cublas==13.1.0.3
[pip3] nvidia-cublas-cu12==12.8.4.1
[pip3] nvidia-cuda-cccl==13.3.3.3.1
[pip3] nvidia-cuda-crt==13.3.33
[pip3] nvidia-cuda-cupti==13.0.85
[pip3] nvidia-cuda-cupti-cu12==12.8.90
[pip3] nvidia-cuda-nvcc==13.3.33
[pip3] nvidia-cuda-nvrtc==13.0.88
[pip3] nvidia-cuda-nvrtc-cu12==12.8.93
[pip3] nvidia-cuda-runtime==13.0.96
[pip3] nvidia-cuda-runtime-cu12==12.8.90
[pip3] nvidia-cudnn-cu12==9.10.2.21
[pip3] nvidia-cudnn-cu13==9.19.0.56
[pip3] nvidia-cudnn-frontend==1.18.0
[pip3] nvidia-cufft==12.0.0.61
[pip3] nvidia-cufft-cu12==11.3.3.83
[pip3] nvidia-cufile==1.15.1.6
[pip3] nvidia-cufile-cu12==1.13.1.3
[pip3] nvidia-curand==10.4.0.35
[pip3] nvidia-curand-cu12==10.3.9.90
[pip3] nvidia-cusolver==12.0.4.66
[pip3] nvidia-cusolver-cu12==11.7.3.90
[pip3] nvidia-cusparse==12.6.3.3
[pip3] nvidia-cusparse-cu12==12.5.8.93
[pip3] nvidia-cusparselt-cu12==0.7.1
[pip3] nvidia-cusparselt-cu13==0.8.0
[pip3] nvidia-cutlass-dsl==4.5.2
[pip3] nvidia-cutlass-dsl-libs-base==4.5.2
[pip3] nvidia-cutlass-dsl-libs-cu13==4.5.2
[pip3] nvidia-ml-py==13.595.45
[pip3] nvidia-nccl-cu13==2.28.9
[pip3] nvidia-nvjitlink==13.0.88
[pip3] nvidia-nvjitlink-cu12==12.8.93
[pip3] nvidia-nvshmem-cu12==3.4.5
[pip3] nvidia-nvshmem-cu13==3.4.5
[pip3] nvidia-nvtx==13.0.85
[pip3] nvidia-nvtx-cu12==12.8.90
[pip3] nvidia-nvvm==13.3.33
[pip3] onnxruntime==1.26.0
[pip3] pyzmq==27.1.0
[pip3] tokenspeed-triton==3.7.10.post20260505
[pip3] torch==2.11.0
[pip3] torch-c-dlpack-ext==0.1.5
[pip3] torch-einops-utils==0.1.0
[pip3] torchaudio==2.11.0+cu130
[pip3] torchdata==0.11.0
[pip3] torchsde==0.2.6
[pip3] torchvision==0.26.0+cu130
[pip3] transformers==5.8.1
[pip3] triton==3.6.0
[pip3] x-transformers==2.19.6
[conda] Could not collect

==============================
         vLLM Info
==============================
ROCM Version                 : Could not collect
vLLM Version                 : 0.22.0
vLLM-Omni Version            : 0.22.0rc2.dev191+gbcf716149 (git sha: bcf716149)
vLLM Build Flags:
  CUDA Archs: Not Set; ROCm: Disabled
GPU Topology:
        GPU0    GPU1    CPU Affinity    NUMA Affinity   GPU NUMA ID
GPU0     X      PHB     0-23    0               N/A
GPU1    PHB      X      0-23    0               N/A

Legend:

  X    = Self
  SYS  = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI)
  NODE = Connection traversing PCIe as well as the interconnect between PCIe Host Bridges within a NUMA node
  PHB  = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU)
  PXB  = Connection traversing multiple PCIe bridges (without traversing the PCIe Host Bridge)
  PIX  = Connection traversing at most a single PCIe bridge
  NV#  = Connection traversing a bonded set of # NVLinks

==============================
     Environment Variables
==============================
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
TORCHINDUCTOR_CACHE_DIR=/tmp/torchinductor_zhengyuan
LD_LIBRARY_PATH=/home/zhengyuan/vllm/vllm-omni/.venv/lib/python3.12/site-packages/cv2/../../lib64:

Your code version

The commit id or version of vllm

0.23.0

The commit id or version of vllm-omni

63ba11b8b1f840b8b5fd7ee8232d7d9cd0a424ff

🐛 Describe the bug

TL;DR. Merely having vllm-omni installed breaks plain vLLM inference for any model that selects the V2 model runner (e.g. Qwen3ForCausalLM) — with no omni model in use. vLLM auto-loads vllm-omni's general plugin during engine init; importing the package runs an import-time monkeypatch that rebinds vllm.v1.request.Request to OmniRequest, which is not positional-substitutable for the base class. GPU kernel warmup then constructs a Request positionally and crashes.

Root cause.

1. vllm-omni registers a general plugin vllm_omni_register_models = vllm_omni.engine.arg_utils:register_omni_models_to_vllm. vLLM's load_general_plugins() (run in the EngineCore process when VLLM_PLUGINS is unset) loads it.
2. That imports the vllm_omni package → vllm_omni/__init__.py runs from . import patch.
3. vllm_omni/patch.py loops over loaded vllm.* modules and does module.Request = OmniRequest — a global, import-time rebind of the canonical vllm.v1.request.Request.
4. OmniRequest.__init__ reorders the constructor — omni params (prompt_embeds, external_req_id, additional_information) come first, whereas base Request requires request_id, prompt_token_ids, sampling_params, pooling_params first → not Liskov-substitutable.
5. vllm/v1/worker/gpu/warmup.py:85 (warmup_kernels, only on the V2 model runner) constructs Request(...) positionally → args land in the wrong omni params → TypeError. Qwen3ForCausalLM is in DEFAULT_V2_MODEL_RUNNER_ARCHITECTURES, so it hits this by default; V1-runner archs (e.g. Qwen2.5) don't run warmup_kernels so don't surface it — the global rebind is latent for them too.

Minimal reproduction.

from vllm import LLM
LLM("Qwen/Qwen3-8B", trust_remote_code=True, max_model_len=2048)

Observed error:

(EngineCore pid=***)   File "***/vllm-omni/.venv/lib/python3.12/site-packages/vllm/v1/worker/gpu/warmup.py", line 85, in warmup_kernels
(EngineCore pid=***)     Request(req_ids[i], prompt_token_ids, sampling_params, pooling_params),
(EngineCore pid=***)     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore pid=***)   File "***/vllm-omni/vllm_omni/request.py", line 41, in __init__
(EngineCore pid=***)     super().__init__(prompt_embeds=prompt_embeds_tensor, *args, **kwargs)
(EngineCore pid=***) TypeError: Request.__init__() missing 3 required positional arguments: 'prompt_token_ids', 'sampling_params', and 'pooling_params'

Impact. Any environment with vllm-omni installed crashes plain inference for V2-runner architectures, with no omni model requested. The trigger (a global import-time rebind from an auto-loaded plugin) is opaque and hard to discover.

Suggested fixes. Make OmniRequest substitutable — keep the base positional order, make omni params keyword-only after *args: def __init__(self, *args, prompt_embeds=None, external_req_id=None, additional_information=None, **kwargs).

Workaround. Set VLLM_PLUGINS to exclude the omni plugin (an allowlist of the other general plugins, or "") when serving plain models — confirmed to prevent the rebind and let init succeed.

Before submitting a new issue...

• Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.