This is my attempt to create a coherent sounding Elon Musk bot by training a recurrent neural network on his tweets from 2010-2017.
Although improvement is observable, it is still in the early stages.
He can be found at https://twitter.com/ElonDeepAI.
Check the license file for the license.
- Upload pre-trained data
- Explain setup
- Make it a clean dataset
This bot is built on top of Hunkim's Word RNN Tensorflow repo. It's very versatile in how you can use it to train models.
Install Tweepy with this tutorial.
Replace the sample.py and train.py from Hunkim's repo with mine and add twitterbot_ai.py to the folder. Add the elonmusk folder to the data folder.
Run twitterbot_ai.py and it will generate and send a tweet to your account.
I will add my pre-trained data as soon as I clean it up.