BazelBot

This is a discord bot that generates non-sensical gibberish for entertainment purposes. The goal of this project is to use the power of LLM's to give my friends a good chuckle.

The main idea is that the bot will pull a certain amount of messages from a discord text channel. He will store those messages in a database. When the user wants to use the !bazel command, the bot will take 10 random messages from the database and generate a gibberish sentence, based on those messages.

Deploy

• Clone the code in a directory of your choosing.

• Navigate to the root directory of this project.

• Create and fill in the required details in the .env file. You can copy this from .env.template

  • You will have to create a discord bot by following these steps: https://discordpy.readthedocs.io/en/stable/discord.html (Make sure to give him admin privileges) In the developer portal, you can retrieve the bot token
  • Activate developer mode in discord: https://support-dev.discord.com/hc/en-us/articles/360028717192-Where-can-I-find-my-Application-Team-Server-ID and copy the ID of a text channel of your choice. This channel will be used as the inspiration for the bot.
  • Also make sure that the DB_CONNECTION_URL in the .env file matches the sqlalchemy.url parameter in the alembic.ini file:

    sqlalchemy.url = postgresql+psycopg2://postgres:postgres@bazelbot-db-1:5432/bazelbot_db

• To fascilitate the bot, we will use ollama and have the gurubot/llama3-guru-uncensored:latest model installed. This is a specific model that is uncensored, so it is able to generate the NSFW content that we need. The model will be pulled using the ollama-pull service in docker compose. The only thing that is needed, is to enable the NVIDIA Container Toolkit to be able to use the GPU's power. You can install it by following the steps in this link: https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html

• If you do not have an Nvidia GPU or a GPU installed you can comment out the following lines in the docker-compose.yaml. Note that this will tremendously slow down the bot!

  services:
  server:
  build:
    context: .
  env_file:
    - .env
  ports:
    - 8000:8000
  volumes:
    - ./src/data:/app/src/data
  depends_on:
    ollama-pull:
        condition: service_completed_successfully
  ollama-pull:
  image: docker/genai:ollama-pull
  env_file:
    - .env
  ollama:
    image: ollama/ollama:latest
    ports:
      - "11434:11434"
    volumes:
      - ollama_volume:/root/.ollama
    # deploy:
    #  resources:
    #    reservations:
    #      devices:
    #        - driver: nvidia
    #          count: all
    #          capabilities: [gpu]
  volumes:
  ollama_volume:

• Finally, execute docker compose up -d in the root directory of the project. Now you should be good to go!

Dev

• Clone the repo and make sure you have docker and poetry installed

• Navigate to the root directory of the project

• Run poetry install

• Create and fill in the required details in the .env file. You can copy this from .env.template

  • You will have to create a discord both by following these steps: https://discordpy.readthedocs.io/en/stable/discord.html (Make sure to give him admin privileges) In the developer portal, you can retrieve the bot token
  • Activate developer mode in discord: https://support-dev.discord.com/hc/en-us/articles/360028717192-Where-can-I-find-my-Application-Team-Server-ID and copy the ID of a text channel of your choice. This channel will be used as the inspiration for the bot.
  • Also make sure that the DB_CONNECTION_URL in the .env file matches the sqlalchemy.url parameter in the alembic.ini file:

    sqlalchemy.url = postgresql+psycopg2://postgres:postgres@localhost:5433/bazelbot_db

• Make sure you have ollama installed on your system: curl -fsSL https://ollama.com/install.sh | sh and make sure that the ollama service is running: systemctl start ollama

• Install this model for ollama: gurubot/llama3-guru-uncensored:latest. Do this my running: ollama pull gurubot/llama3-guru-uncensored:latest

• Run make test-up

• Run make run to run the bot locally

• If you want to run the test suite, run make test

ENJOY BAZELING!