blest-bot

Installing dependencies

Run the following to install the required pip packages

pip install -r requirements.txt

Setup environnement file

You need to create a file named .env and place the following inside

# .env
DISCORD_TOKEN={YOUR_DISCORD_TOKEN_HERE}
DISCORD_GUILDS={TEXT_CHANNEL_IDS}
TRIGGER_WORD={TRIGGER_WORD}
SUFFIX_WORD={SUFFIX_WORD}

Example:

# .env
DISCORD_TOKEN=488494944994
DISCORD_GUILDS=437289473289,43829042389,4738294279
TRIGGER_WORD=blest
SUFFIX_WORD=kappo

Watch out ! The text channel ids must be separated by a comma ,

Create the training file

The script training_set_generator.py creates the train_1.txt file containing the 4000 last messages from the channels in the .env file

To start the script simply execute the following line

python3 training_set_generator.py

If you want to make your own training file, simply create a file named train_X.txt where X is any integer. Make sure each sentences are separeted by a new line, like the following :

This is a sentence
This is another sentence

Start the bot

You can either start the bot in your terminal with

python3 bot.py

or as a background process with

python3 bot.py > logs.txt &

Using docker, you can simply run docker compose run blest-bot-training to run the training program and docker compose run blest-bot-svc to run the actual bot

Stop the bot

To stop the bot, you can simply kill the process with Ctrl + C if it runs inside a terminal or you can find its PID with pgrep -lf python3 if it's running in the background.