RAF: A RNN-based Chatbot Development Tool

Please read "To Do List" before use. The tool is under development and not all features are available

The purpose of this tool is to help develop an RNN-based chatbot.

Development continues. The final version hasn't been released yet.

Features

1. Data preprocessing for different datasets
2. Chatbot training with an rnn-based neural network
3. Adaptability to different datasets
4. Easy control with UI interface

To Do List

1. Data Prepocessing Tools (Status: The first version is ready. Updates keep coming)
2. Dictionary Class (Status: Final version)
3. Tool Helpers (Status: The first version is ready. Updates keep coming)
4. RNN Architecture (Status: Final version)
5. Trainer (Status: The first version is ready. Updates keep coming)
6. Example Main Code (Status: Under development)
7. UI (Status: Development has not started)
8. User Manual (Status: Development has not started)

Run

To run, download the tool kit and add it to your project folder.

Source

This tool is sourced from https://github.com/ywk991112/pytorch-chatbot, https://github.com/spro/practical-pytorch/tree/master/seq2seq-translation and https://github.com/floydhub/textutil-preprocess-cornell-movie-corpus