KPConv-PyTorch/INSTALL.md

57 lines
2 KiB
Markdown
Raw Normal View History

2020-04-27 22:43:55 +00:00
# Installation instructions
## Ubuntu 18.04
2020-04-27 22:21:49 +00:00
2020-04-27 22:43:55 +00:00
* Make sure <a href="https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html">CUDA</a> and <a href="https://docs.nvidia.com/deeplearning/sdk/cudnn-install/index.html">cuDNN</a> are installed. One configuration has been tested:
- PyTorch 1.4.0, CUDA 10.1 and cuDNN 7.6
2020-04-27 22:21:49 +00:00
* Ensure all python packages are installed :
sudo apt update
sudo apt install python3-dev python3-pip python3-tk
2020-04-27 22:43:55 +00:00
* Follow <a href="https://pytorch.org/get-started/locally/">PyTorch installation procedure</a>.
2020-04-27 22:21:49 +00:00
* Install the other dependencies with pip:
- numpy
- scikit-learn
2020-04-27 22:43:55 +00:00
- PyYAML
2020-04-27 22:21:49 +00:00
- matplotlib (for visualization)
- mayavi (for visualization)
- PyQt5 (for visualization)
2020-04-27 22:43:55 +00:00
* Compile the C++ extension modules for python located in `cpp_wrappers`. Open a terminal in this folder, and run:
2020-04-27 22:21:49 +00:00
sh compile_wrappers.sh
You should now be able to train Kernel-Point Convolution models
2020-04-27 22:43:55 +00:00
## Windows 10
* Make sure <a href="https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html">CUDA</a> and <a href="https://docs.nvidia.com/deeplearning/sdk/cudnn-install/index.html">cuDNN</a> are installed. One configuration has been tested:
- PyTorch 1.4.0, CUDA 10.1 and cuDNN 7.5
* Follow <a href="https://pytorch.org/get-started/locally/">PyTorch installation procedure</a>.
2020-04-27 22:46:34 +00:00
* We used the PyCharm IDE to pip install all python dependencies (including PyTorch) in a venv:
- torch
- torchvision
2020-04-27 22:43:55 +00:00
- numpy
- scikit-learn
- PyYAML
- matplotlib (for visualization)
- mayavi (for visualization)
- PyQt5 (for visualization)
* Compile the C++ extension modules for python located in `cpp_wrappers`. You just have to execute two .bat files:
cpp_wrappers/cpp_neighbors/build.bat
2020-04-27 22:46:34 +00:00
and
2020-04-27 22:43:55 +00:00
cpp_wrappers/cpp_subsampling/build.bat
You should now be able to train Kernel-Point Convolution models
2020-04-27 22:21:49 +00:00