[Bug]: Validation Error: block_size (2096) > max_num_batched_tokens (2048) when enabling prefix caching for Qwen3.5 Mamba architecture #36697
Description
Your current environment
The output of python collect_env.py
==============================
System Info
==============================
OS : Ubuntu 22.04.4 LTS (x86_64)
GCC version : (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
Clang version : Could not collect
CMake version : Could not collect
Libc version : glibc-2.35
==============================
PyTorch Info
==============================
PyTorch version : 2.10.0+cu128
Is debug build : False
CUDA used to build PyTorch : 12.8
ROCM used to build PyTorch : N/A
==============================
Python Environment
==============================
Python version : 3.11.14 (main, Oct 21 2025, 18:31:21) [GCC 11.2.0] (64-bit runtime)
Python platform : Linux-5.15.0-94-generic-x86_64-with-glibc2.35
==============================
CUDA / GPU Info
==============================
Is CUDA available : True
CUDA runtime version : 12.4.99
CUDA_MODULE_LOADING set to :
GPU models and configuration :
GPU 0: NVIDIA GeForce RTX 4090
GPU 1: NVIDIA GeForce RTX 4090
GPU 2: NVIDIA GeForce RTX 4090
GPU 3: NVIDIA GeForce RTX 4090
GPU 4: NVIDIA GeForce RTX 4090
GPU 5: NVIDIA GeForce RTX 4090
GPU 6: NVIDIA GeForce RTX 4090
GPU 7: NVIDIA GeForce RTX 4090
Nvidia driver version : 570.158.01
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: 52 bits physical, 57 bits virtual
Byte Order: Little Endian
CPU(s): 192
On-line CPU(s) list: 0-191
Vendor ID: GenuineIntel
Model name: Intel(R) Xeon(R) Platinum 8468
CPU family: 6
Model: 143
Thread(s) per core: 2
Core(s) per socket: 48
Socket(s): 2
Stepping: 8
CPU max MHz: 3800.0000
CPU min MHz: 800.0000
BogoMIPS: 4200.00
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 cat_l2 cdp_l3 invpcid_single intel_ppin cdp_l2 ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local split_lock_detect avx_vnni avx512_bf16 wbnoinvd dtherm ida arat pln pts hwp hwp_act_window hwp_epp hwp_pkg_req avx512vbmi umip pku ospke waitpkg avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq la57 rdpid bus_lock_detect cldemote movdiri movdir64b enqcmd fsrm md_clear serialize tsxldtrk pconfig arch_lbr amx_bf16 avx512_fp16 amx_tile amx_int8 flush_l1d arch_capabilities
Virtualization: VT-x
L1d cache: 4.5 MiB (96 instances)
L1i cache: 3 MiB (96 instances)
L2 cache: 192 MiB (96 instances)
L3 cache: 210 MiB (2 instances)
NUMA node(s): 4
NUMA node0 CPU(s): 0-23,96-119
NUMA node1 CPU(s): 24-47,120-143
NUMA node2 CPU(s): 48-71,144-167
NUMA node3 CPU(s): 72-95,168-191
Vulnerability Gather data sampling: 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 Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; Enhanced IBRS, IBPB conditional, RSB filling, PBRSB-eIBRS SW sequence
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
==============================
Versions of relevant libraries
==============================
[pip3] flashinfer-python==0.6.4
[pip3] numpy==2.2.6
[pip3] nvidia-cublas-cu12==12.8.4.1
[pip3] nvidia-cuda-cupti-cu12==12.8.90
[pip3] nvidia-cuda-nvrtc-cu12==12.8.93
[pip3] nvidia-cuda-runtime-cu12==12.8.90
[pip3] nvidia-cudnn-cu12==9.10.2.21
[pip3] nvidia-cudnn-frontend==1.19.0
[pip3] nvidia-cufft-cu12==11.3.3.83
[pip3] nvidia-cufile-cu12==1.13.1.3
[pip3] nvidia-curand-cu12==10.3.9.90
[pip3] nvidia-cusolver-cu12==11.7.3.90
[pip3] nvidia-cusparse-cu12==12.5.8.93
[pip3] nvidia-cusparselt-cu12==0.7.1
[pip3] nvidia-cutlass-dsl==4.4.1
[pip3] nvidia-cutlass-dsl-libs-base==4.4.1
[pip3] nvidia-ml-py==13.590.48
[pip3] nvidia-nccl-cu12==2.27.5
[pip3] nvidia-nvjitlink-cu12==12.8.93
[pip3] nvidia-nvshmem-cu12==3.4.5
[pip3] nvidia-nvtx-cu12==12.8.90
[pip3] pyzmq==27.1.0
[pip3] torch==2.10.0
[pip3] torch_c_dlpack_ext==0.1.5
[pip3] torchaudio==2.10.0
[pip3] torchvision==0.25.0
[pip3] transformers==4.57.6
[pip3] triton==3.6.0
[conda] flashinfer-python 0.6.4 pypi_0 pypi
[conda] numpy 2.2.6 pypi_0 pypi
[conda] nvidia-cublas-cu12 12.8.4.1 pypi_0 pypi
[conda] nvidia-cuda-cupti-cu12 12.8.90 pypi_0 pypi
[conda] nvidia-cuda-nvrtc-cu12 12.8.93 pypi_0 pypi
[conda] nvidia-cuda-runtime-cu12 12.8.90 pypi_0 pypi
[conda] nvidia-cudnn-cu12 9.10.2.21 pypi_0 pypi
[conda] nvidia-cudnn-frontend 1.19.0 pypi_0 pypi
[conda] nvidia-cufft-cu12 11.3.3.83 pypi_0 pypi
[conda] nvidia-cufile-cu12 1.13.1.3 pypi_0 pypi
[conda] nvidia-curand-cu12 10.3.9.90 pypi_0 pypi
[conda] nvidia-cusolver-cu12 11.7.3.90 pypi_0 pypi
[conda] nvidia-cusparse-cu12 12.5.8.93 pypi_0 pypi
[conda] nvidia-cusparselt-cu12 0.7.1 pypi_0 pypi
[conda] nvidia-cutlass-dsl 4.4.1 pypi_0 pypi
[conda] nvidia-cutlass-dsl-libs-base 4.4.1 pypi_0 pypi
[conda] nvidia-ml-py 13.590.48 pypi_0 pypi
[conda] nvidia-nccl-cu12 2.27.5 pypi_0 pypi
[conda] nvidia-nvjitlink-cu12 12.8.93 pypi_0 pypi
[conda] nvidia-nvshmem-cu12 3.4.5 pypi_0 pypi
[conda] nvidia-nvtx-cu12 12.8.90 pypi_0 pypi
[conda] pyzmq 27.1.0 pypi_0 pypi
[conda] torch 2.10.0 pypi_0 pypi
[conda] torch-c-dlpack-ext 0.1.5 pypi_0 pypi
[conda] torchaudio 2.10.0 pypi_0 pypi
[conda] torchvision 0.25.0 pypi_0 pypi
[conda] transformers 4.57.6 pypi_0 pypi
[conda] triton 3.6.0 pypi_0 pypi
==============================
vLLM Info
==============================
ROCM Version : Could not collect
vLLM Version : 0.17.0
vLLM Build Flags:
CUDA Archs: Not Set; ROCm: Disabled
GPU Topology:
GPU0 GPU1 GPU2 GPU3 GPU4 GPU5 GPU6 GPU7 NIC0 CPU Affinity NUMA Affinity GPU NUMA ID
GPU0 X NODE NODE NODE SYS SYS SYS SYS SYS 0-23,96-119 0 N/A
GPU1 NODE X NODE NODE SYS SYS SYS SYS SYS 0-23,96-119 0 N/A
GPU2 NODE NODE X NODE SYS SYS SYS SYS SYS 0-23,96-119 0 N/A
GPU3 NODE NODE NODE X SYS SYS SYS SYS SYS 0-23,96-119 0 N/A
GPU4 SYS SYS SYS SYS X NODE NODE NODE NODE 24-47,120-143 1 N/A
GPU5 SYS SYS SYS SYS NODE X NODE NODE NODE 24-47,120-143 1 N/A
GPU6 SYS SYS SYS SYS NODE NODE X NODE NODE 24-47,120-143 1 N/A
GPU7 SYS SYS SYS SYS NODE NODE NODE X NODE 24-47,120-143 1 N/A
NIC0 SYS SYS SYS SYS NODE NODE NODE NODE X
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
NIC Legend:
NIC0: mlx5_bond_0
==============================
Environment Variables
==============================
LD_LIBRARY_PATH=/usr/local/cuda-12.4/lib64:
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
TORCHINDUCTOR_CACHE_DIR=/tmp/torchinductor_Yifei
🐛 Describe the bug
-
Bug summary: When running a Mamba-hybrid model (Qwen3_5MoeForConditionalGeneration) with --enable-prefix-caching, vLLM auto-scales the block_size to 2096 to align Mamba and Attention pages. However, it crashes during VllmConfig validation because the default max_num_batched_tokens for chunked prefill is 2048.
-
Suggested fix: The framework should either automatically scale max_num_batched_tokens to be at least block_size when Mamba align mode forces it, or provide a clearer warning to the user to manually increase --max-num-batched-tokens.
-
Full Traceback:
(vllm) Yifei@lyg1058:~/models$ CUDA_VISIBLE_DEVICES=2,3 vllm serve Qwen3.5-122B-A10B-GPTQ-Int4 --served-model-name Qwen3.5-27B-AWQ-4bit --gpu-memory-utilization 0.9 --port 8848 -tp 2 --max-model-len 131072 --enable-auto-tool-choice --tool-call-parser qwen3_coder --reasoning-parser qwen3 --max-num-seqs 16 --enable-prefix-caching
(APIServer pid=913608) INFO 03-11 03:44:50 [utils.py:302]
(APIServer pid=913608) INFO 03-11 03:44:50 [utils.py:302] █ █ █▄ ▄█
(APIServer pid=913608) INFO 03-11 03:44:50 [utils.py:302] ▄▄ ▄█ █ █ █ ▀▄▀ █ version 0.17.0
(APIServer pid=913608) INFO 03-11 03:44:50 [utils.py:302] █▄█▀ █ █ █ █ model Qwen3.5-122B-A10B-GPTQ-Int4
(APIServer pid=913608) INFO 03-11 03:44:50 [utils.py:302] ▀▀ ▀▀▀▀▀ ▀▀▀▀▀ ▀ ▀
(APIServer pid=913608) INFO 03-11 03:44:50 [utils.py:302]
(APIServer pid=913608) INFO 03-11 03:44:50 [utils.py:238] non-default args: {'model_tag': 'Qwen3.5-122B-A10B-GPTQ-Int4', 'enable_auto_tool_choice': True, 'tool_call_parser': 'qwen3_coder', 'port': 8848, 'model': 'Qwen3.5-122B-A10B-GPTQ-Int4', 'max_model_len': 131072, 'served_model_name': ['Qwen3.5-27B-AWQ-4bit'], 'reasoning_parser': 'qwen3', 'tensor_parallel_size': 2, 'enable_prefix_caching': True, 'max_num_seqs': 16}
(APIServer pid=913608) INFO 03-11 03:44:59 [model.py:531] Resolved architecture: Qwen3_5MoeForConditionalGeneration
(APIServer pid=913608) INFO 03-11 03:44:59 [model.py:1554] Using max model len 131072
(APIServer pid=913608) INFO 03-11 03:44:59 [gptq_marlin.py:229] The model is convertible to gptq_marlin during runtime. Using gptq_marlin kernel.
(APIServer pid=913608) INFO 03-11 03:45:00 [scheduler.py:231] Chunked prefill is enabled with max_num_batched_tokens=2048.
(APIServer pid=913608) WARNING 03-11 03:45:00 [config.py:381] Mamba cache mode is set to 'align' for Qwen3_5MoeForConditionalGeneration by default when prefix caching is enabled
(APIServer pid=913608) INFO 03-11 03:45:00 [config.py:401] Warning: Prefix caching in Mamba cache 'align' mode is currently enabled. Its support for Mamba layers is experimental. Please report any issues you may observe.
(APIServer pid=913608) INFO 03-11 03:45:00 [config.py:544] Setting attention block size to 2096 tokens to ensure that attention page size is >= mamba page size.
(APIServer pid=913608) INFO 03-11 03:45:00 [config.py:575] Padding mamba page size by 0.58% to ensure that mamba page size and attention page size are exactly equal.
(APIServer pid=913608) INFO 03-11 03:45:01 [vllm.py:747] Asynchronous scheduling is enabled.
(APIServer pid=913608) Traceback (most recent call last):
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/bin/vllm", line 10, in <module>
(APIServer pid=913608) sys.exit(main())
(APIServer pid=913608) ^^^^^^
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/site-packages/vllm/entrypoints/cli/main.py", line 73, in main
(APIServer pid=913608) args.dispatch_function(args)
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/site-packages/vllm/entrypoints/cli/serve.py", line 112, in cmd
(APIServer pid=913608) uvloop.run(run_server(args))
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/site-packages/uvloop/__init__.py", line 92, in run
(APIServer pid=913608) return runner.run(wrapper())
(APIServer pid=913608) ^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/asyncio/runners.py", line 118, in run
(APIServer pid=913608) return self._loop.run_until_complete(task)
(APIServer pid=913608) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=913608) File "uvloop/loop.pyx", line 1518, in uvloop.loop.Loop.run_until_complete
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/site-packages/uvloop/__init__.py", line 48, in wrapper
(APIServer pid=913608) return await main
(APIServer pid=913608) ^^^^^^^^^^
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/site-packages/vllm/entrypoints/openai/api_server.py", line 471, in run_server
(APIServer pid=913608) await run_server_worker(listen_address, sock, args, **uvicorn_kwargs)
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/site-packages/vllm/entrypoints/openai/api_server.py", line 490, in run_server_worker
(APIServer pid=913608) async with build_async_engine_client(
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/contextlib.py", line 210, in __aenter__
(APIServer pid=913608) return await anext(self.gen)
(APIServer pid=913608) ^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/site-packages/vllm/entrypoints/openai/api_server.py", line 96, in build_async_engine_client
(APIServer pid=913608) async with build_async_engine_client_from_engine_args(
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/contextlib.py", line 210, in __aenter__
(APIServer pid=913608) return await anext(self.gen)
(APIServer pid=913608) ^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/site-packages/vllm/entrypoints/openai/api_server.py", line 122, in build_async_engine_client_from_engine_args
(APIServer pid=913608) vllm_config = engine_args.create_engine_config(usage_context=usage_context)
(APIServer pid=913608) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/site-packages/vllm/engine/arg_utils.py", line 1890, in create_engine_config
(APIServer pid=913608) config = VllmConfig(
(APIServer pid=913608) ^^^^^^^^^^^
(APIServer pid=913608) File "/data/home/Yifei/miniconda3/envs/vllm/lib/python3.11/site-packages/pydantic/_internal/_dataclasses.py", line 121, in __init__
(APIServer pid=913608) s.__pydantic_validator__.validate_python(ArgsKwargs(args, kwargs), self_instance=s)
(APIServer pid=913608) pydantic_core._pydantic_core.ValidationError: 1 validation error for VllmConfig
(APIServer pid=913608) Assertion failed, In Mamba cache align mode, block_size (2096) must be <= max_num_batched_tokens (2048). [type=assertion_error, input_value=ArgsKwargs((), {'model_co...transfer_config': None}), input_type=ArgsKwargs]
(APIServer pid=913608) For further information visit https://errors.pydantic.dev/2.12/v/assertion_error
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.