Text Generation with Keras and TensorFlow (10.3)
After a viral blog post by Andrej Karpathy demonstrated that recurrent neural networks are capable of producing very realistic looking (but fake) text, C source code, and even LaTex, there has been considerable interested in this technology. This video demonstrates the use of LSTM, in Keras/TensorFlow, to generate text based on a sample corpus.
Code for This Video:
https://github.com/jeffheaton/t81_558_deep_learning/blob/master/t81_558_class_10_3_text_generation.ipynb
Course Homepage: https://sites.wustl.edu/jeffheaton/t81-558/
Follow Me/Subscribe:
https://www.youtube.com/user/HeatonResearch
https://github.com/jeffheaton
https://twitter.com/jeffheaton
Support Me on Patreon: https://www.patreon.com/jeffheaton
After a viral blog post by Andrej Karpathy demonstrated that recurrent neural networks are capable of producing very realistic looking (but fake) text, C source code, and even LaTex, there has been considerable interested in this technology. This video demonstrates the use of LSTM, in Keras/TensorFlow, to generate text based on a sample corpus.
Code for This Video:
https://github.com/jeffheaton/t81_558_deep_learning/blob/master/t81_558_class_10_3_text_generation.ipynb
Course Homepage: https://sites.wustl.edu/jeffheaton/t81-558/
Follow Me/Subscribe:
https://www.youtube.com/user/HeatonResearch
https://github.com/jeffheaton
https://twitter.com/jeffheaton
Support Me on Patreon: https://www.patreon.com/jeffheaton
