Skip to content

Latest commit

 

History

History
59 lines (47 loc) · 2.04 KB

get_started.md

File metadata and controls

59 lines (47 loc) · 2.04 KB
Raw

Prerequisites

  • Linux or macOS (Windows is in experimental support)
  • Python 3.6+
  • PyTorch 1.3+
  • CUDA 9.2+ (If you build PyTorch from source, CUDA 9.0 is also compatible)
  • GCC 5+
  • MMCV

Installation

I ran experiments with PyTorch 1.8.0, CUDA 11.1, Python 3.7, and Ubuntu 20.04. Other settings that satisfact the requirement would work.

If you have a similar environment

You can simply follow our settings:

Use Anaconda to create a conda environment:

conda create -n MDE python=3.7
conda activate MDE

Install Pytorch:

conda install pytorch==1.8.0 torchvision==0.9.0 torchaudio==0.8.0 cudatoolkit=11.1 -c pytorch -c conda-forge

Then, install MMCV and install our toolbox:

pip install mmcv-full==1.3.13 -f https://download.openmmlab.com/mmcv/dist/cu111/torch1.8.0/index.html

git clone https://github.com/zhyever/Monocular-Depth-Estimation-Toolbox.git
cd Monocular-Depth-Estimation-Toolbox
pip install -e .

If training, you should install the tensorboard:

pip install future tensorboard

If you have a different environment,

You only need to install PyTorch and MMCV (1.3.1<= version <=1.4.0) correspondingly, and then build our codebase:

git clone https://github.com/zhyever/Monocular-Depth-Estimation-Toolbox.git
cd Monocular-Depth-Estimation-Toolbox
pip install -e .

More information about installation can be found in docs of MMSegmentation (see get_started.md).

Other Dependence

When reproducing Adabins, Pytorch3d is needed to calculate the chamfer loss. You can annotate the import in depth/models/losses/chamferloss.py if you don't wanna train Adabins, or you should install Pytorch3d:

conda install -c fvcore -c iopath -c conda-forge fvcore iopath
conda install -c bottler nvidiacub
conda install pytorch3d -c pytorch3d

In the future, I will try to remove this dependence.