[Bug] Why tensorRT not support int8?

[pedroHuang123]

Checklist

• I have searched related issues but cannot get the expected help.
• 2. I have read the FAQ documentation but cannot get the expected help.
• 3. The bug has not been fixed in the latest version.

Describe the bug

If i want to support tensorRT PTQ,what can I do for it.

Reproduction

deploy.py

Environment

2023-01-11 15:39:02,658 - mmdeploy - INFO - **********Environmental information**********
2023-01-11 15:39:21,260 - mmdeploy - INFO - sys.platform: win32
2023-01-11 15:39:21,260 - mmdeploy - INFO - Python: 3.9.13 | packaged by conda-forge | (main, May 27 2022, 16:51:29) [MSC v.1929 64 bit (AMD64)]
2023-01-11 15:39:21,261 - mmdeploy - INFO - CUDA available: True
2023-01-11 15:39:21,261 - mmdeploy - INFO - GPU 0: NVIDIA GeForce RTX 3090
2023-01-11 15:39:21,261 - mmdeploy - INFO - CUDA_HOME: C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.1
2023-01-11 15:39:21,261 - mmdeploy - INFO - NVCC: Cuda compilation tools, release 11.1, V11.1.74
2023-01-11 15:39:21,261 - mmdeploy - INFO - MSVC: 用于 x64 的 Microsoft (R) C/C++ 优化编译器 19.33.31629 版
2023-01-11 15:39:21,261 - mmdeploy - INFO - GCC: n/a
2023-01-11 15:39:21,261 - mmdeploy - INFO - PyTorch: 1.9.0+cu111
2023-01-11 15:39:21,261 - mmdeploy - INFO - PyTorch compiling details: PyTorch built with:
  - C++ Version: 199711
  - MSVC 192829337
  - Intel(R) Math Kernel Library Version 2020.0.2 Product Build 20200624 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v2.1.2 (Git Hash 98be7e8afa711dc9b66c8ff3504129cb82013cdb)
  - OpenMP 2019
  - CPU capability usage: AVX2
  - CUDA Runtime 11.1
  - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-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_37,code=compute_37
  - CuDNN 8.0.5
  - Magma 2.5.4
  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.1, CUDNN_VERSION=8.0.5, CXX_COMPILER=C:/w/b/windows/tmp_bin/sccache-cl.exe, CXX_FLAGS=/DWIN32 /D_WINDOWS /GR /EHsc /w /bigobj -DUSE_PTHREADPOOL -openmp:experimental -IC:/w/b/windows/mkl/include -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOCUPTI -DUSE_FBGEMM -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.9.0, 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=OFF, USE_NNPACK=OFF, USE_OPENMP=ON, 

2023-01-11 15:39:21,261 - mmdeploy - INFO - TorchVision: 0.10.0+cu111
2023-01-11 15:39:21,262 - mmdeploy - INFO - OpenCV: 4.6.0
2023-01-11 15:39:21,262 - mmdeploy - INFO - MMCV: 1.5.1
2023-01-11 15:39:21,262 - mmdeploy - INFO - MMCV Compiler: MSVC 192930137
2023-01-11 15:39:21,262 - mmdeploy - INFO - MMCV CUDA Compiler: 11.1
2023-01-11 15:39:21,262 - mmdeploy - INFO - MMDeploy: 0.11.0+ca3c68b
2023-01-11 15:39:21,262 - mmdeploy - INFO - 

2023-01-11 15:39:21,262 - mmdeploy - INFO - **********Backend information**********
2023-01-11 15:39:23,381 - mmdeploy - INFO - onnxruntime: 1.8.1	ops_is_avaliable : True
2023-01-11 15:39:23,530 - mmdeploy - INFO - tensorrt: 8.2.3.0	ops_is_avaliable : True
2023-01-11 15:39:23,643 - mmdeploy - INFO - ncnn: None	ops_is_avaliable : False
2023-01-11 15:39:23,649 - mmdeploy - INFO - pplnn_is_avaliable: False
2023-01-11 15:39:23,697 - mmdeploy - INFO - openvino_is_avaliable: False
2023-01-11 15:39:23,787 - mmdeploy - INFO - snpe_is_available: False
2023-01-11 15:39:23,795 - mmdeploy - INFO - ascend_is_available: False
2023-01-11 15:39:23,801 - mmdeploy - INFO - coreml_is_available: False
2023-01-11 15:39:23,801 - mmdeploy - INFO - 

2023-01-11 15:39:23,801 - mmdeploy - INFO - **********Codebase information**********
2023-01-11 15:39:23,862 - mmdeploy - INFO - mmdet:	None
2023-01-11 15:39:23,863 - mmdeploy - INFO - mmseg:	None
2023-01-11 15:39:23,863 - mmdeploy - INFO - mmcls:	0.25.0
2023-01-11 15:39:23,863 - mmdeploy - INFO - mmocr:	None
2023-01-11 15:39:23,863 - mmdeploy - INFO - mmedit:	None
2023-01-11 15:39:23,863 - mmdeploy - INFO - mmdet3d:	None
2023-01-11 15:39:23,863 - mmdeploy - INFO - mmpose:	None
2023-01-11 15:39:23,863 - mmdeploy - INFO - mmrotate:	None
2023-01-11 15:39:23,863 - mmdeploy - INFO - mmaction:	None
2023-01-11 15:39:23,864 - mmdeploy - INFO - mmflow:	0.5.1

Error traceback

No response

[hanrui1sensetime]

Seems that you want to support the trt int8 deployment for mmflow. You can imitate the trt int8 config for mmflow codebase. If it failed, you can post the failed log on this issue.