In order to study about Machine Learning for Natural Language Processing, I decided to create a chatbot by study the theory and code from MLT (machine lerning community in Tokyo)
I also got a help from by brother to make a Web Application for this project.
I've learned the basic how these cool virtual assistants work, Google Assistant, Siri and my favorite of all time Jarvis (lol)
The train script is quite long because it also include the text preprocessing process (tokenization, counting, convert word to vector, etc.).
I added the comment in the train.py code so it might help you not get lost in the code.
- install NodeJS dependencies: cd to los_bot folder ad run
npm install - run the application:
npm start.
- MLT seq2seq_bot (https://github.com/Machine-Learning-Tokyo/seq2seq_bot)
- los_bot (https://github.com/flukeiedong/dumpling_chat)