README.md

Chitchatting Chatbo - A seq2seq model with attention.

Motivation

- Gain deepened understanding on NLP
- Hands-on experience implementing seq2seq models with attention with TensorFlow
- Familiarize model training using TPUs

Steps

- Theoretical understanding on Encoder-Decoder based seq2seq conversational agent with attentions
- Explore and preprocess a variety of conversational datasets
- Build encoder-decoder model with attention:
	1. Shared embedding layer for encoder and decoder
	2. Two GRU layer for both encoder and decoder
	3. Bidirectional encoder
	3. Decoder with Luong's attention
- Define training loop with teacher-forcing
- Inference with both greedy and beam search
- Deploy the model with Flask

How to use:

- Clone the repo
- Create an virtual environment and install the packeges using: pip install -r requirements.txt
- run flask app: python app.py