Description🔎

This repository contain notebook that used to train Chatbot Model that deployed using flask on Chatbot Deploy with Flask

Main Library 📚

• Pytorch

Dataset💾

•Chatbot Model are trained with preprocessed Cornell Movie-Dialogs Corpus
•All dataset version (raw and formatted) can be found in this Drive

Model 🤖

This chatbot used sequence-to-sequence architecture wich have Encoder and Decoder in it. Bidirectional Gated Recurrent Unit (GRU) are implemented in Encoder while Vanilla GRU in Decoder. In additon to improve chatbot performance i'm using Luong Attention Mechanism that implemented in Decoder.

How to Install🔧

There are several tools that i used on this notebook:
• Anaconda Installation
• Chatbot Environment Installation -> Follow this Guide with using chatbot_env.yml from this repository

How to Run🏃‍♂️

•Open notebook on local or GoogleCollabs
•You can skip this step if using GoogleCollabs. Activate the imported environment with
conda activate chatbot_env in terminal
•Run each cell step by step
•On the mounting cell you need to connect your notebook with Google Drive account which has the required dataset on correctly path