search

vllm-project / vllm

A high-throughput and memory-efficient inference and serving engine for LLMs
https://docs.vllm.ai
Apache License 2.0
29.84k stars 4.51k forks source link

[Bug]: Online serving failing for Phi-3-vision-128k-instruct #8553

Closed Muhtasham closed 2 weeks ago

Muhtasham commented 1 month ago

Your current environment

The output of `python collect_env.py` ```text collect_env.py 100%[============================================================================================================================>] 24.67K --.-KB/s in 0.001s 2024-09-18 04:29:30 (38.0 MB/s) - ‘collect_env.py’ saved [25265/25265] 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: 3.10.12 (main, Sep 11 2024, 15:47:36) [GCC 11.4.0] (64-bit runtime) Python platform: Linux-5.15.0-118-generic-x86_64-with-glibc2.35 Is CUDA available: True CUDA runtime version: 11.5.119 CUDA_MODULE_LOADING set to: LAZY GPU models and configuration: GPU 0: NVIDIA H100 80GB HBM3 GPU 1: NVIDIA H100 80GB HBM3 GPU 2: NVIDIA H100 80GB HBM3 GPU 3: NVIDIA H100 80GB HBM3 GPU 4: NVIDIA H100 80GB HBM3 GPU 5: NVIDIA H100 80GB HBM3 GPU 6: NVIDIA H100 80GB HBM3 GPU 7: NVIDIA H100 80GB HBM3 Nvidia driver version: 550.90.07 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: 46 bits physical, 57 bits virtual Byte Order: Little Endian CPU(s): 208 On-line CPU(s) list: 0-207 Vendor ID: GenuineIntel Model name: Intel(R) Xeon(R) Platinum 8470 CPU family: 6 Model: 143 Thread(s) per core: 2 Core(s) per socket: 52 Socket(s): 2 Stepping: 8 BogoMIPS: 4000.00 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 Virtualization: VT-x L1d cache: 4.9 MiB (104 instances) L1i cache: 3.3 MiB (104 instances) L2 cache: 208 MiB (104 instances) L3 cache: 210 MiB (2 instances) NUMA node(s): 8 NUMA node0 CPU(s): 0-12,104-116 NUMA node1 CPU(s): 13-25,117-129 NUMA node2 CPU(s): 26-38,130-142 NUMA node3 CPU(s): 39-51,143-155 NUMA node4 CPU(s): 52-64,156-168 NUMA node5 CPU(s): 65-77,169-181 NUMA node6 CPU(s): 78-90,182-194 NUMA node7 CPU(s): 91-103,195-207 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 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 BHI_DIS_S 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==12.560.30 [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.post2@9ba0817ff1eb514f51cc6de9cb8e16c98d6ee44f vLLM Build Flags: CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled GPU Topology: GPU0 GPU1 GPU2 GPU3 GPU4 GPU5 GPU6 GPU7 NIC0 NIC1 NIC2 NIC3 NIC4 NIC5 NIC6 NIC7 NIC8 NIC9 NIC10 NIC11 CPU Affinity NUMA Affinity GPU NUMA ID GPU0 X NV18 NV18 NV18 NV18 NV18 NV18 NV18 PIX PIX PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS 0-12,104-116 0 N/A GPU1 NV18 X NV18 NV18 NV18 NV18 NV18 NV18 SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS 26-38,130-142 2 N/A GPU2 NV18 NV18 X NV18 NV18 NV18 NV18 NV18 SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS 39-51,143-155 3 N/A GPU3 NV18 NV18 NV18 X NV18 NV18 NV18 NV18 SYS SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS 13-25,117-129 1 N/A GPU4 NV18 NV18 NV18 NV18 X NV18 NV18 NV18 SYS SYS SYS SYS SYS SYS PIX PIX PIX SYS SYS SYS 52-64,156-168 4 N/A GPU5 NV18 NV18 NV18 NV18 NV18 X NV18 NV18 SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX SYS SYS 78-90,182-194 6 N/A GPU6 NV18 NV18 NV18 NV18 NV18 NV18 X NV18 SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX SYS 91-103,195-207 7 N/A GPU7 NV18 NV18 NV18 NV18 NV18 NV18 NV18 X SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX 65-77,169-181 5 N/A NIC0 PIX SYS SYS SYS SYS SYS SYS SYS X PIX PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS NIC1 PIX SYS SYS SYS SYS SYS SYS SYS PIX X PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS NIC2 PIX SYS SYS SYS SYS SYS SYS SYS PIX PIX X SYS SYS SYS SYS SYS SYS SYS SYS SYS NIC3 SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS X SYS SYS SYS SYS SYS SYS SYS SYS NIC4 SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS X SYS SYS SYS SYS SYS SYS SYS NIC5 SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS X SYS SYS SYS SYS SYS SYS NIC6 SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS X PIX PIX SYS SYS SYS NIC7 SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX X PIX SYS SYS SYS NIC8 SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX PIX X SYS SYS SYS NIC9 SYS SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS X SYS SYS NIC10 SYS SYS SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS X SYS NIC11 SYS SYS SYS SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS 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 NIC2: mlx5_2 NIC3: mlx5_3 NIC4: mlx5_4 NIC5: mlx5_5 NIC6: mlx5_6 NIC7: mlx5_7 NIC8: mlx5_8 NIC9: mlx5_9 NIC10: mlx5_10 NIC11: mlx5_11 ```

🐛 Describe the bug

online inference falls for this model


vllm serve microsoft/Phi-3-vision-128k-instruct --trust-remote-code  --max-num-seqs 5 --disable_sliding_window 

...
_pickle.PicklingError: Can't pickle <class 'transformers_modules.microsoft.Phi-3-vision-128k-instruct.c45209e90a4c4f7d16b2e9d48503c7f3e83623ed.configuration_phi3_v.Phi3VConfig'>: it's not the same object as transformers_modules.microsoft.Phi-3-vision-128k-instruct.c45209e90a4c4f7d16b2e9d48503c7f3e83623ed.configuration_phi3_v.Phi3VConfig

Doesn't seem to be transformers (4.44.2) or weights issue I did clear hf cache and redownloaded Also doesn't seem to support TP

same params for the offline inference works 

# Phi-3-Vision
def run_phi3v(question):
    prompt = f"<|user|>\n<|image_1|>\n{question}<|end|>\n<|assistant|>\n"  # noqa: E501
    # Note: The default setting of max_num_seqs (256) and
    # max_model_len (128k) for this model may cause OOM.
    # You may lower either to run this example on lower-end GPUs.

    # In this example, we override max_num_seqs to 5 while
    # keeping the original context length of 128k.
    llm = LLM(
        model="microsoft/Phi-3-vision-128k-instruct",
        trust_remote_code=True,
        max_num_seqs=5,
        disable_sliding_window=True,
    )
    stop_token_ids = None
    return llm, prompt, stop_token_ids
python offline_inference_vision_language.py
INFO 09-18 04:49:57 llm_engine.py:223] Initializing an LLM engine (v0.6.1.post2) with config: model='microsoft/Phi-3-vision-128k-instruct', speculative_config=None, tokenizer='microsoft/Phi-3-vision-128k-instruct', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, override_neuron_config=None, rope_scaling=None, rope_theta=None, tokenizer_revision=None, trust_remote_code=True, dtype=torch.bfloat16, max_seq_len=131072, download_dir=None, load_format=LoadFormat.AUTO, tensor_parallel_size=1, pipeline_parallel_size=1, disable_custom_all_reduce=False, quantization=None, enforce_eager=False, kv_cache_dtype=auto, quantization_param_path=None, device_config=cuda, decoding_config=DecodingConfig(guided_decoding_backend='outlines'), observability_config=ObservabilityConfig(otlp_traces_endpoint=None, collect_model_forward_time=False, collect_model_execute_time=False), seed=0, served_model_name=microsoft/Phi-3-vision-128k-instruct, use_v2_block_manager=False, num_scheduler_steps=1, enable_prefix_caching=False, use_async_output_proc=True)
INFO 09-18 04:50:31 model_runner.py:997] Starting to load model microsoft/Phi-3-vision-128k-instruct...

INFO 09-18 04:50:32 weight_utils.py:242] Using model weights format ['*.safetensors']
Loading safetensors checkpoint shards:   0% Completed | 0/2 [00:00<?, ?it/s]
Loading safetensors checkpoint shards:  50% Completed | 1/2 [00:00<00:00,  2.21it/s]
Loading safetensors checkpoint shards: 100% Completed | 2/2 [00:01<00:00,  1.69it/s]
Loading safetensors checkpoint shards: 100% Completed | 2/2 [00:01<00:00,  1.76it/s]

I
INFO 09-18 04:50:38 gpu_executor.py:122] # GPU blocks: 8681, # CPU blocks: 682
INFO 09-18 04:50:39 model_runner.py:1311] Capturing the model for CUDA graphs. This may lead to unexpected consequences if the model is not static. To run the model in eager mode, set 'enforce_eager=True' or use '--enforce-eager' in the CLI.
INFO 09-18 04:50:39 model_runner.py:1315] CUDA graphs can take additional 1~3 GiB memory per GPU. If you are running out of memory, consider decreasing `gpu_memory_utilization` or enforcing eager mode. You can also reduce the `max_num_seqs` as needed to decrease memory usage.
INFO 09-18 04:50:41 model_runner.py:1430] Graph capturing finished in 1 secs.
Processed prompts: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████| 4/4 [00:00<00:00,  4.54it/s, est. speed input: 8794.91 toks/s, output: 290.44 toks/s]
 The image captures a breathtaking view of the Tokyo Skytree, a renowned landmark in Japan. The Skytree, a tall, white tower with a gold dome, stands majestically against the backdrop of a clear blue sky. The tower is surrounded by a multitude of cherry
 The image captures a serene scene of a tall, white tower standing majestically against a clear blue sky. The tower, which is the focal point of the image, is surrounded by a multitude of cherry blossom trees. These trees, in full bloom, are adorned with
 The image captures a serene scene of a tall, white tower standing majestically against a clear blue sky. The tower, which appears to be a lighthouse, is adorned with a golden dome that gleams under the sunlight. The perspective of the image is from a low angle,
 The image captures the iconic Tokyo Skytree, a towering structure that stands tall against the backdrop of a clear blue sky. The Skytree, painted in a pristine white, is adorned with a golden dome that adds a touch of elegance to its imposing presence.

Before submitting a new issue...

Muhtasham commented 1 month ago

Also microsoft/Phi-3.5-vision-instruct fails with TP

vllm serve microsoft/Phi-3.5-vision-instruct --max-model-len 4096 --trust-remote-code --tensor-parallel-size 8 
ModuleNotFoundError: No module named 'transformers_modules.microsoft.Phi-3'
Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "/usr/lib/python3.10/multiprocessing/spawn.py", line 116, in spawn_main
    exitcode = _main(fd, parent_sentinel)
  File "/usr/lib/python3.10/multiprocessing/spawn.py", line 126, in _main
    self = reduction.pickle.load(from_parent)
ModuleNotFoundError: No module named 'transformers_modules.microsoft.Phi-3'
DarkLight1337 commented 1 month ago

See #8288

russellb commented 1 month ago

fwiw, i was able to reproduce this, but it was fixed for me as of #8157

IdoAmit198 commented 1 month ago

I am using version 0.6.1.post2 and having the same issue with microsoft/Phi-3.5-mini-instruct and microsoft/Phi-3.5-MoE-instruct models. I have followed a previous work-around for current bug and added the following prior to running:

export VLLM_WORKER_MULTIPROC_METHOD=spawn

When using those models through HuggingFace I am not facing that issue.