What is this repo about?

1. Quickly set-up Anaconda based Python (3.6) environment for deep learning (Keras, PyTorch) in Ubuntu 16.04 (xenial) or 18.04 (bionic)
2. Quickly set-up a Deep Learning instance in AWS, GCP or Paperspace Gradient (coming soon)

The Setup (Ubuntu 16.04 or 18.04 in local workstation)

1. Before you go through these installation steps:
   • UEFI
   • These scripts have been regularly used on systems with fresh Ubuntu installations
2. Clone this repo
   • Need to have git installed for cloning (forking+cloning) this repo
   • Alternatively, you can download the repo here and extract the content
3. Change your current directory to the cloned/downloaded repo: cd python-deeplearning-setup
4. Install the following scripts one-by-one
   • Make the scripts executable: run chmod +x *.sh
   • Install CUDA: run ./cuda.sh
   • Install cuDNN: run ./cudnn.sh
   • Install Miniconda: run ./conda.sh
   • Create the cv environment: conda env create -f cv.yml
   • (Optional) Setup remote access to Jupyter Notebooks:
     • Edit Line no 10 of condaremote.sh to set the correct path (will remove this step in a later iteration)
     • run ./condaremote.sh