[Bug]: Pixtral inference not working correctly with LLMEngine/AsyncEngine

[larme]

Your current environment

The output of `python collect_env.py`

Collecting environment information...
PyTorch version: 2.4.0+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A

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: version 3.22.1
Libc version: glibc-2.35

Python version: 3.10.12 (main, Jul 29 2024, 16:56:48) [GCC 11.4.0] (64-bit runtime)
Python platform: Linux-5.15.0-119-generic-x86_64-with-glibc2.35
Is CUDA available: True
CUDA runtime version: 12.2.140
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration: GPU 0: NVIDIA A100 80GB PCIe
Nvidia driver version: 535.183.01
cuDNN version: Probably one of the following:
/usr/lib/x86_64-linux-gnu/libcudnn.so.8.9.6
/usr/lib/x86_64-linux-gnu/libcudnn_adv_infer.so.8.9.6
/usr/lib/x86_64-linux-gnu/libcudnn_adv_train.so.8.9.6
/usr/lib/x86_64-linux-gnu/libcudnn_cnn_infer.so.8.9.6
/usr/lib/x86_64-linux-gnu/libcudnn_cnn_train.so.8.9.6
/usr/lib/x86_64-linux-gnu/libcudnn_ops_infer.so.8.9.6
/usr/lib/x86_64-linux-gnu/libcudnn_ops_train.so.8.9.6
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True

CPU:
Architecture:                         x86_64
CPU op-mode(s):                       32-bit, 64-bit
Address sizes:                        46 bits physical, 48 bits virtual
Byte Order:                           Little Endian
CPU(s):                               12
On-line CPU(s) list:                  0-11
Vendor ID:                            GenuineIntel
Model name:                           Intel(R) Xeon(R) Gold 6342 CPU @ 2.80GHz
CPU family:                           6
Model:                                106
Thread(s) per core:                   1
Core(s) per socket:                   12
Socket(s):                            1
Stepping:                             6
BogoMIPS:                             5600.12
Flags:                                fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush acpi mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl cpuid tsc_known_freq pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch cpuid_fault invpcid_single intel_ppin ssbd ibrs ibpb stibp ibrs_enhanced fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves umip pku ospke gfni vaes vpclmulqdq rdpid md_clear flush_l1d arch_capabilities
Hypervisor vendor:                    Xen
Virtualization type:                  full
L1d cache:                            576 KiB (12 instances)
L1i cache:                            384 KiB (12 instances)
L2 cache:                             15 MiB (12 instances)
L3 cache:                             432 MiB (12 instances)
NUMA node(s):                         1
NUMA node0 CPU(s):                    0-11
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:        Mitigation; Clear CPU buffers; SMT Host state unknown
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 and seccomp
Vulnerability Spectre v1:             Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:             Mitigation; Enhanced / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI SW loop, KVM SW loop
Vulnerability Srbds:                  Not affected
Vulnerability Tsx async abort:        Not affected

Versions of relevant libraries:
[pip3] numpy==1.26.4
[pip3] nvidia-cublas-cu12==12.1.3.1
[pip3] nvidia-cuda-cupti-cu12==12.1.105
[pip3] nvidia-cuda-nvrtc-cu12==12.1.105
[pip3] nvidia-cuda-runtime-cu12==12.1.105
[pip3] nvidia-cudnn-cu12==9.1.0.70
[pip3] nvidia-cufft-cu12==11.0.2.54
[pip3] nvidia-curand-cu12==10.3.2.106
[pip3] nvidia-cusolver-cu12==11.4.5.107
[pip3] nvidia-cusparse-cu12==12.1.0.106
[pip3] nvidia-ml-py==11.525.150
[pip3] nvidia-nccl-cu12==2.20.5
[pip3] nvidia-nvjitlink-cu12==12.6.68
[pip3] nvidia-nvtx-cu12==12.1.105
[pip3] pyzmq==26.2.0
[pip3] torch==2.4.0
[pip3] torchvision==0.19.0
[pip3] transformers==4.44.2
[pip3] triton==3.0.0
[conda] Could not collect
ROCM Version: Could not collect
Neuron SDK Version: N/A
vLLM Version: 0.6.1@3fd2b0d21cd9ec78de410fdf8aa1de840e9ad77a
vLLM Build Flags:
CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled
GPU Topology:
[4mGPU0	CPU Affinity	NUMA Affinity	GPU NUMA ID[0m
GPU0	 X 	0-11	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

Model Input Dumps

No response

🐛 Describe the bug

This code snippets does not work

import PIL.Image
import uuid
from vllm import EngineArgs, LLMEngine
from vllm import SamplingParams, TextPrompt
from vllm.multimodal import MultiModalDataBuiltins

MODEL_ID = "mistral-community/pixtral-12b-240910"
ENGINE_ARGS = EngineArgs(
    model=MODEL_ID,
    tokenizer_mode="mistral",
    enable_prefix_caching=True,
    limit_mm_per_prompt=dict(image=4),
    max_num_batched_tokens=16384,
)
SAMPLING_PARAM = SamplingParams()

engine = LLMEngine.from_engine_args(ENGINE_ARGS)
prompt = "describe the image"

engine_inputs = TextPrompt(prompt=prompt)

image = PIL.Image.open("demo.jpg")
mm_data = MultiModalDataBuiltins(image=[image])
engine_inputs["multi_modal_data"] = mm_data

engine.add_request(uuid.uuid4().hex, engine_inputs, SAMPLING_PARAM)

while True:
    out = engine.step()
    for request_output in request_outputs:
        if request_output.finished:
            print(request_output)
    if not engine.has_unfinished_requests():
        break

This will give an output message like:

 File "/workspace/codes/example/vllm/pixtral-12b/venv/lib/python3.10/site-packages/vllm/model_executor/models/pixtral.py", line 117, in merge_multimodal_embeddings
    assert (seq_len == N_txt +
AssertionError: seq_len 7 should be equal to N_txt + N_img (7, 1200, 0)

It seems that I need to manually padding the input token ids with image_token_ids like this:

tokenizer = engine.get_tokenizer()
token_ids = tokenizer(prompt).input_ids
image_token_id = 10
token_ids = [image_token_id] * image_token_num + token_ids

engine_inputs = TokensPrompt(prompt_token_ids=token_ids)

mm_data = MultiModalDataBuiltins(image=[image])
engine_inputs["multi_modal_data"] = mm_data

To make it work.

AsyncLLMEngine also see the same limitation and I need similar modification to make it works like: https://github.com/bentoml/BentoVLLM/pull/71/files#diff-357f77bce00e63217bb5ec382293bca653276e58af9bdfb6e7c50ca9487e27aeR84-R94

Is this behavior intended or a bug?

Before submitting a new issue...

• [X] 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.