Bloom Chatbot

A chatbot built with the Bloom open-source model inspired by sentdex work and re-used later for a class project

Setup

1. git clone
2. cd bloom-chatbot
3. pip install -r requirements.txt
4. Edit src/config.py to your liking
5. python main.py

Configuration

You can use the src/config.py file to change the settings of the chatbot.

The SCENARIO variable is the initial prompt that is given to the model for it to get some context. It should describe in which context does the conversation occurs and who are the different actors.

BOT_NAME and USER_NAME respectively represent the name of the bot and the name of the user within the conversation. They will be prefixed before every message of the discussion.

AUDIO_OUTPUT is a boolean that indicates whether the bot should use text to speech for its response.

AUDIO_INPUT is a boolean that indicates whether the user should input using speech recognition -> This setting is useful if you try to practice pronounciation in a different language for example.

LANGUAGE and AUDIO_LANGUAGE are the language of the conversation and the language of the audio output. They should be set to the same value if you want to use both audio features.

• AUDIO_LANGUAGE should be set to a BCP-47 from Google documentation
• LANGUAGE should be set to a language code like 'en', 'fr', ...

Usage

Text mode

At each step the user writes a response to the bot using its keyboard

Audio mode

• The user can start and stop the recording using SPACEBAR. When the speech recognition has finished processing the user can re-record to correct any mistake by pressing SPACEBAR again.
• The user can validate the response by pressing ENTER

License

Bloom is an open-source model distributed under its specific license, your use of this software should align with the license of the model.

I am not responsible for any results and/or damages caused by the use of this software.