Skip to content

Instantly share code, notes, and snippets.

@philippmuench
Last active August 13, 2019 13:35
Show Gist options
  • Save philippmuench/6cb8a5904560972b682a452eba8d414e to your computer and use it in GitHub Desktop.
Save philippmuench/6cb8a5904560972b682a452eba8d414e to your computer and use it in GitHub Desktop.
# requirement: anaconda
conda install virtualenv
python3 -m venv ~/deepG_env
source ~/deepG_env/bin/activate
# deepG is tested with tensorflow 1.14
pip3 install https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow_gpu-1.14.0-cp37-cp37m-linux_x86_64.whl
pip3 install keras
# setup cuda, required is cuda10 and cudnn 7.4
conda install -c anaconda cudatoolkit
conda install -c fragcolor cuda10.0
conda install pytorch
export LD_LIBRARY_PATH=LD_LIBRARY_PATH:/usr/local/cuda-9.0/lib64/
# requirement: devtools, reticulate_1.12
# cave: previous version of reticulate cannot make use of the pip3 environment and will lead to an error when executing line 7
# python3 is not supported on reticulate_1.10 !!! So you may want to update packages using update.packages(ask=F)
devtools::install_github("rstudio/tensorflow")
library(tensorflow)
library(keras)
use_virtualenv("~/deepG_env/", required = TRUE)
use_condaenv(condaenv = "myenv", required=TRUE)
keras::load_model_hdf5("example_files/GenomeNet_v1.hdf5")
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment