Corrections
This commit is contained in:
parent
4dc6b7a225
commit
5b64cc5394
39
INSTALL.md
Normal file
39
INSTALL.md
Normal file
|
@ -0,0 +1,39 @@
|
||||||
|
### Installation instructions for Ubuntu 16.04
|
||||||
|
|
||||||
|
* 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. Three configurations have been tested:
|
||||||
|
- TensorFlow 1.4.1, CUDA 8.0 and cuDNN 6.0
|
||||||
|
- TensorFlow 1.12.0, CUDA 9.0 and cuDNN 7.4
|
||||||
|
- ~~TensorFlow 1.13.0, CUDA 10.0 and cuDNN 7.5~~ (bug found only with this version).
|
||||||
|
|
||||||
|
* Ensure all python packages are installed :
|
||||||
|
|
||||||
|
sudo apt update
|
||||||
|
sudo apt install python3-dev python3-pip python3-tk
|
||||||
|
|
||||||
|
* Follow <a href="https://www.tensorflow.org/install/pip">Tensorflow installation procedure</a>.
|
||||||
|
|
||||||
|
* Install the other dependencies with pip:
|
||||||
|
- numpy
|
||||||
|
- scikit-learn
|
||||||
|
- psutil
|
||||||
|
- matplotlib (for visualization)
|
||||||
|
- mayavi (for visualization)
|
||||||
|
- PyQt5 (for visualization)
|
||||||
|
|
||||||
|
* Compile the customized Tensorflow operators located in `tf_custom_ops`. Open a terminal in this folder, and run:
|
||||||
|
|
||||||
|
sh compile_op.sh
|
||||||
|
|
||||||
|
N.B. If you installed Tensorflow in a virtual environment, it needs to be activated when running these scripts
|
||||||
|
|
||||||
|
* Compile the C++ extension module for python located in `cpp_wrappers`. Open a terminal in this folder, and run:
|
||||||
|
|
||||||
|
sh compile_wrappers.sh
|
||||||
|
|
||||||
|
You should now be able to train Kernel-Point Convolution models
|
||||||
|
|
||||||
|
### Installation instructions for Ubuntu 18.04 (Thank to @noahtren)
|
||||||
|
|
||||||
|
* Remove the `-D_GLIBCXX_USE_CXX11_ABI=0` flag for each line in `tf_custom_ops/compile_op.sh` (problem with the version of gcc). One configuration has been tested:
|
||||||
|
|
||||||
|
- TensorFlow 1.12.0, CUDA 9.0 and cuDNN 7.3.1
|
Loading…
Reference in a new issue