[Bug] Using pipeline inference CogVLM2 works fine but server fails

xiangqi1997 commented 3 weeks ago

Checklist

[X] 1. I have searched related issues but cannot get the expected help.
[X] 2. The bug has not been fixed in the latest version.

Describe the bug

参考https://github.com/InternLM/lmdeploy/blob/main/docs/zh_cn/multi_modal/cogvlm.md，使用pipeline可以推理得到结果，但使用api_server时，可以启动服务，但调用时卡住，服务端只有GET，没有POST，显示GPU100%利用率卡住，请问是什么原因？

Reproduction

lmdeploy serve api_server ~/.cache/huggingface/hub/models--THUDM--cogvlm2-llama3-chinese-chat-19B/snapshots/d88b352bce5ee58a289b1ac8328553eb31efa2ef/ --backend pytorch --tp 2 --cache-max-entry-count 0.1 --session-len 4096

Environment

ys.platform: linux
Python: 3.10.14 (main, May  6 2024, 19:42:50) [GCC 11.2.0]
CUDA available: True
MUSA available: False
numpy_random_seed: 2147483648
GPU 0,1,2: NVIDIA GeForce RTX 3090
CUDA_HOME: /usr/local/cuda-12
NVCC: Cuda compilation tools, release 12.4, V12.4.99
GCC: gcc (GCC) 9.3.0
PyTorch: 2.2.2+cu121
PyTorch compiling details: PyTorch built with:
  - GCC 9.3
  - C++ Version: 201703
  - Intel(R) oneAPI Math Kernel Library Version 2022.2-Product Build 20220804 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v3.3.2 (Git Hash 2dc95a2ad0841e29db8b22fbccaf3e5da7992b01)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - LAPACK is enabled (usually provided by MKL)
  - NNPACK is enabled
  - CPU capability usage: AVX512
  - CUDA Runtime 12.1
  - NVCC architecture flags: -gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_90,code=sm_90
  - CuDNN 8.9.2
  - Magma 2.6.1
  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=12.1, CUDNN_VERSION=8.9.2, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wsuggest-override -Wno-psabi -Wno-error=pedantic -Wno-error=old-style-cast -Wno-missing-braces -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=2.2.2, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=1, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, USE_ROCM_KERNEL_ASSERT=OFF, 

TorchVision: 0.17.2+cu121
LMDeploy: 0.4.2+8cc7f43
transformers: 4.41.2
gradio: Not Found
fastapi: 0.111.0
pydantic: 2.7.3
triton: 2.2.0

Error traceback

No response

RunningLeon commented 3 weeks ago

@xiangqi1997 hi, do use use same settings for pipeline and api_server？ For instance, --cache-max-entry-count 0.1 is used to set kv cache mem after loading model and it's too small.

xiangqi1997 commented 3 weeks ago

sorry my problem, 现在server可以正常使用🙏

InternLM / lmdeploy