DenseMatchingBenchmark环境配置及复现

DenseMatchingBenchmark环境配置及复现

Requirements:

PyTorch1.7.1+, Python3.8+, Cuda11.0+
    mmcv==1.3.5

 Option 1: Step-by-step installation

a. Create a conda virtual environment and activate it. Then install required packages.
   conda create --name dm-b python=3.8
   //source activate(importent)
   conda activate dense_matching_benchmark

   # this installs the right pip and dependencies for the fresh python
   conda install ipython
   conda install pip

   # install required packages from requirements.txt
   pip install -r /(自己的路经）/requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple
   注：repuirement.txt文件里面有pytorch,删掉。

  torch>=1.7.1   torchvision>=0.8.2    torchaudio>=0.7.2 自己下载

b.Install PyTorch stable or nightly and torchvision following the  https://pytorch.org/
   conda install pytorch==1.7.1 torchvision==0.8.2 torchaudio==0.7.2 cudatoolkit=10.1 -c pytorch
 
   conda 安装报错 specified in the package manifest cannot be found.
      解决：  conda clean --packages --tarballs    删除缓存的包重新下载
   离线安装：
   conda install --offline pytorch-1.3.1-py3.7_cuda10.1.243_cudnn7.6.3_0.tar.bz2
   conda install --offline torchvision-0.4.2-py37_cu101.tar.bz2
   conda install --offline cudatoolkit-10.1.243-h6bb024c_0.tar.bz2