BlackBird Linux Onboard

Getting Started

Prerequisites

• At least 1024 MB RAM
• Python 3.5+
• pip 9.0.1+
• git

Requirements for running BlackBird on Linux are minimal. However, please note...

• It is recommended that you develop with a CUDA-capable GPU, with tensorflow-gpu installed. This is not a requirement, but you may find that large networks take a long time to run in the default version of tensorflow. You can find help getting the prerequisites working in Ubuntu here.
• Compatibility with Python 2.x is not guaranteed, nor is it tested for. It is strongly recommended that you develop in Python 3.5+. Should you be interested in incorporating backwards compatibility into BlackBird, please consult one of the project managers (jordan-singer, kevdog25).

Setup Steps

1. Clone the master repository
   1. git clone https://www.github.com/jordan-singer/BlackBird
   2. cd BlackBird
2. Set up a virtual environment
   1. pip install virtualenv
   2. python -m virtualenv BlackBirdEnvironment
   3. source BlackBirdEnvironment/bin/activate
   4. pip install -r requirements.txt