A discord bot wrapper for the Sayori-Bot main project
This code repo utilizes a heavily modified version of MokouBot
Repo: https://github.com/benoxoft/MokouBot
Install python >= 3.5
pip install -r requirements.txt
To utilize the voice response, install ffmpeg. Edit your PATH environment variable if needed.
Create a new application here: https://discordapp.com/developers/applications/me#top
Find the api key in the new application under Token and copy-paste it into a new file named API_KEY.txt located in the project root directory.
Edit BOTID
Edit OWNERID
Edit ChatBot name
Edit your bot's name to the one set in config.py
python3 train.py
-----OR-----
The bot grabs conversations from threads on 4chan. You can configure from which board to get the data in config.py
By default it grabs conversations from those boards:
SFW_BOARDS = ['jp', 'a', 'v',]
NSFW_BOARDS = ['b', 'pol']
The training script will also grab images from the SFW boards and place them in the images/ directory. If you do not want to download the images (1-2 GB), edit config.py and remove all the boards from SFW_BOARDS and place them in NSFW_BOARDS like so:
SFW_BOARDS = []
NSFW_BOARDS = ['jp', 'a', 'v', 'b', 'pol']
When ready start training:
python3 4chan_train.py
python3 bot.py
Go there and invite your bot: https://discordapp.com/oauth2/authorize?client_id={YOUR_CLIENT_ID}&scope=bot&permissions=0
Change {YOUR_CLIENT_ID} to the client ID of your bot
Type "help" starting with your command prefix
Ex:
~help
This command will return a PM message to the person who sent it. The list can be found in the features directory.
The bot picks up messages that include its Discord name.
The bot will not respond to the name you included in the config.py or bot.py files
To get the bot to respond verbally, join a voice channel and type in join
starting with your command prefix stated in config.py (Default COMMANDPREFIX = '~').