An Python implementation of BotZoo. A powerful chatbot development framework that requires no hardcore programing. Design a bot in 10 minutes and continuously improve its performance via interactive machine learning to reach the state-of-art performance.
pip install -r requirments.txt
sh docker.shdocker pull convmind/flowbotdocker run -p 7000:7000 -d convmind/botzooUse as a local variable
from botzoo import FlowBot
bot = BotZoo()
# create a new conversation
bot.create_session("{user_id}", "{graph_name}")
# send a message to the bot
bot.receive_event(session_id, usr_input, sync=True)
# terminate a conversation
bot.remove_session(session_id)
Use as a Restful API service using Flask
from botzoo import RestfulFlowBot
bot = RestfulFlowBot()
bot.run(port=3000)
See Restful API documentations for details.
For English/Chinese (LUIS name -> Internal Name)
Two database will be used:
flowmaker
ml_core
Flowbot uses Redis as its dialog state memory. Run Redis via:
redis-server &
Open terminal client:
redis-cli
Check all keys
Set new value
Read spcific keys
The body of the POST call will contains the follow example JSON:
data = {'session': session ( a JSON dump of mongo document),
'global_contexts': max_slot_vals ([{'name': 'location',
'value': 'pittsburgh'} ...],
'registered_inputs': [{'sys': 'hi', 'usr': 'how are you'}...]}
Content type
Image
Digest
Size
369.2 MB
Last updated
almost 7 years ago
docker pull convmind/botzoo