Skip to content

Latest commit

 

History

History
28 lines (18 loc) · 935 Bytes

README.md

File metadata and controls

28 lines (18 loc) · 935 Bytes

Quadcopter Training Project for Udacity Deep Learning Nanodedree

Project overview

This is Quadcopter Training project for Udacity's Deep Learning Nanodegree program. In this project, I build and train Reinforcement Learning agent in Jupiter Notebook using Keras. This project contains two files:

  • task.py
  • agent.py

Prerequisites

The code is Python in a Jupyter Notebook and it uses:

You can install everything you need to run this project with Anaconda.

Installing

Clone this repository

git clone https://github.com/rauf-mifteev/DLND_Quadcopter_Training.git

Run the Code

Navigate to the cloned directories location and start jupyter notebook with Quadcopter_Project.ipynb

jupyter notebook Quadcopter_Project.ipynb