search

vllm-project / vllm

A high-throughput and memory-efficient inference and serving engine for LLMs
https://docs.vllm.ai
Apache License 2.0
29.08k stars 4.34k forks source link

[Usage]: guided_json fails on pixtral when using OpenAI API #9359

Closed ktrapeznikov closed 1 week ago

ktrapeznikov commented 1 week ago

Your current environment

The output of `python collect_env.py` ```text No module named 'vllm._version' from vllm.version import __version__ as VLLM_VERSION 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: Could not collect Libc version: glibc-2.35 Python version: 3.12.7 (main, Oct 1 2024, 08:52:12) [GCC 11.4.0] (64-bit runtime) Python platform: Linux-5.4.0-196-generic-x86_64-with-glibc2.35 Is CUDA available: True CUDA runtime version: Could not collect CUDA_MODULE_LOADING set to: LAZY GPU models and configuration: GPU 0: NVIDIA RTX A6000 Nvidia driver version: 550.107.02 cuDNN version: Could not collect 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: 43 bits physical, 48 bits virtual Byte Order: Little Endian CPU(s): 256 On-line CPU(s) list: 0-254 Off-line CPU(s) list: 255 Vendor ID: AuthenticAMD Model name: AMD EPYC 7H12 64-Core Processor CPU family: 23 Model: 49 Thread(s) per core: 2 Core(s) per socket: 64 Socket(s): 2 Stepping: 0 Frequency boost: enabled CPU max MHz: 2600.0000 CPU min MHz: 0.0000 BogoMIPS: 5199.91 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 nopl nonstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 x2apic 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 ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr wbnoinvd arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif umip rdpid overflow_recov succor smca sme sev sev_es Virtualization: AMD-V L1d cache: 4 MiB (128 instances) L1i cache: 4 MiB (128 instances) L2 cache: 64 MiB (128 instances) L3 cache: 512 MiB (32 instances) NUMA node(s): 2 NUMA node0 CPU(s): 0-63,128-191 NUMA node1 CPU(s): 64-127,192-254 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: Vulnerable 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; Retpolines; IBPB conditional; IBRS_FW; STIBP conditional; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected Vulnerability Srbds: Not affected Vulnerability Tsx async abort: Not affected Versions of relevant libraries: [pip3] flashinfer==0.1.6+cu121torch2.4 [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==12.560.30 [pip3] nvidia-nccl-cu12==2.20.5 [pip3] nvidia-nvjitlink-cu12==12.6.77 [pip3] nvidia-nvtx-cu12==12.1.105 [pip3] pyzmq==26.2.0 [pip3] torch==2.4.0 [pip3] torchvision==0.19.0 [pip3] transformers==4.45.2 [pip3] triton==3.0.0 [conda] Could not collect ROCM Version: Could not collect Neuron SDK Version: N/A vLLM Version: N/A (dev) vLLM Build Flags: CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled GPU Topology: GPU0 NIC0 NIC1 CPU Affinity NUMA Affinity GPU NUMA ID GPU0 X SYS SYS 0-63,128-191 0 N/A NIC0 SYS X PIX NIC1 SYS PIX 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_0 NIC1: mlx5_1 ```

Model Input Dumps

No response

🐛 Describe the bug

I have OpenAI compatible server running inside docker with the following arguments:

--model mistralai/Pixtral-12B-2409 --dtype bfloat16 --uvicorn-log-level debug --tensor-parallel-size 1 --max-num-seqs 32 --gpu-memory-utilization=.98 --disable-log-stats --tokenizer_pool_size=4 --distributed-executor-backend=ray --tokenizer_mode mistral --limit_mm_per_prompt 'image=4' --max-model-len 16000

If I make a request:


guided_json = {'properties': {'image_caption': {'title': 'Image Caption', 'type': 'string'}},
 'required': ['image_caption'],
 'title': 'Caption',
 'type': 'object'}

kwargs = dict(max_tokens=512,
                      stream = False,
                        top_p=.9,
                        temperature=.5, model = model_id,
                        guided_json =guided_json,
                        messages=[{"role":"system",
                                   "content": "Format response as JSON"},
                                  {"role": "user",
                                    "content": [
                                        {"type": "text", "text": "caption the image"},
                                        {
                                            "type": "image_url",
                                            "image_url": {
                                                "url": get_image_url(f),
                                            },
                                        },
                                    ],
                        }],
                    )

a = requests.post(url, json= kwargs)

I get <Response [500]> .

In the vllm logs I also get:

WARNING 10-14 20:36:36 chat_utils.py:570] 'add_generation_prompt' is not supported for mistral tokenizer, so it will be ignored.
WARNING 10-14 20:36:36 chat_utils.py:574] 'continue_final_message' is not supported for mistral tokenizer, so it will be ignored.

If remove guided_json argument then everything works.

guided_json argument works for other VLMs such as llava or nvidia/NVLM-D-72B

Before submitting a new issue...

DarkLight1337 commented 1 week ago

I believe Mistral tokenizer currently doesn't support guided decoding yet. cc @patrickvonplaten

ktrapeznikov commented 1 week ago

oh is that because outlines package does not have support for mistral_inference