easybot

Learn to implement a chat bot with deep learning.

The first version is based on: tensorflow_chatbot from llSourcell.

usage

Put the dataset into easybot/data/

How to get the sample dataset

Change seq2seq.ini to switch to the training mode：

mode = train

Start training

python execute.py

Testing

You can stop the training anytime, the trained model will be saved in easybot/working_dir/. Change seq2seq.ini to switch to the testing mode, then start execute.py：

mode = test python execute.py

All the code for serve mode is in easybot/webui/

• Main app:

python webui/app.py

• Startup/Shutdown:

sh webui/startup.sh sh webui/shutdown.sh

• Background runing:

sh webui/startup.sh

• App Daemon:

sh webui/app-daemon.sh

• Archive log:

sh webui/archivelog.sh

Demonstration snapshot:

Easybot 聊天机器人