Skip to content

Latest commit

 

History

History
76 lines (68 loc) · 4.35 KB

setup-docker.md

File metadata and controls

76 lines (68 loc) · 4.35 KB
Raw

Docker Setup

Important

Currently, it seems like wsl does not like Nvidia Container Toolkit. It will work initially then reset it for some odd reason. For now, it is advised to use an actually Linux machine to run using Docker. If you do not care about utilizing your GPU or don't even have a Nvidia GPU then disregard this.

Nvidia Container Toolkit Setup

Installation with Apt

  • Instructions can be found here in Steps to reproduce or below:
    • Step 1. Configure the production repository on machine:
curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg \
  && curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list | \
    sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \
    sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list
  • Step 2. Update the packages list from the repository:
sudo apt-get update
  • Step 3. Install the Nvidia Container Toolkit:
sudo apt-get install -y nvidia-container-toolkit

Configurating with Docker

Step 1.Configure the container runtime by using the nvidia-ctk command:

sudo nvidia-ctk runtime configure --runtime=docker

The nvidia-ctk command modifies the /etc/docker/daemon.json file on the host. The file is updated so that Docker can use the NVIDIA Container Runtime.

Step 2. Restart the Docker daemon:

sudo systemctl restart docker

References for setup

To Run (with Docker and Docker Compose)

  • With the inclusion of subnets in the docker-compose.yml, you will need to set the SUBNET_ADDRESS, OLLAMA_IP, OLLAMA_PORT, and DISCORD_IP. Here are some default values if you don't care:
    • OLLAMA_IP = 172.18.0.2
    • OLLAMA_PORT = 11434
    • DISCORD_IP = 172.18.0.3
    • SUBNET_ADDRESS = 172.18.0.0
    • Don't understand any of this? watch a Networking video to understand subnetting.
    • You also need all environment variables shown in .env.sample
  • Otherwise, there is no need to install any npm packages for this, you just need to run npm run start to pull the containers and spin them up.
  • For cleaning up on Linux (or Windows), run the following commands:
    • docker compose stop
    • docker compose rm
    • docker ps to check if containers have been removed.
    • This may not work if the nvidia installation was done incorrectly. If this is the case, please utilize the Manual "Clean-up" shown below.
  • You can also use npm run clean to clean up the containers and remove the network to address a possible Address already in use problem. This script does not have to end without error to work.

Manual Run (with Docker)

  • Run the following commands:
    • npm run docker:build
    • npm run docker:ollama
    • npm run docker:client
    • docker ps to see if the containers are there!
      • Names should be discord and ollama.
    • You can also just run npm run docker:start now for the above commands.
  • Clean-up:
    • docker ps for the conatiner id's. Use -a flag as necessary.
    • docker rm -f discord && docker rm -f ollama to remove the containers.
      • docker rm -f CONTAINER_ID do for both containers if naming issues arise.
    • docker network rm ollama-net removes the network.
      • docker network prune will also work so long as the network is unused.
  • Remove Image:
    • If you need to remove the image run docker image rm IMAGE_ID. You can get the image id by running docker images.