This is a containerized sandpit application that contains the necessary basic services for AI and data dev.
- Docker. Make sure this is installed on whatever platform you are running on.
- Go to root directory of this project in terminal, and execute the following. This should bring up all containers.
- If any changes were made to config files, this will also trigger a rebuild of containers.
docker compose -f sandpit.yaml up -d
- Verify all containers are running with
docker ps
For a full list, see the docker yaml file.
-
Mongo (noSQL) and mongo express (management interface)
- Access via mongo-express
- Can also be access via mongosh at localhost:27017
-
Mariadb (SQL) and adminer (management interface)
- Access via adminer
-
Jupyterlab
- Access via link
-
Ollama based DeepSeek R1 model
- Based on guide
- Runs in ollama service
- Exposes the API service by ollama-web, accessible at: http://localhost:3001/
docker compose -f sandpit.yaml build
- Example, check the base OS on a docker image, get the container id, by running
docker ps
. Suppose the ID is 0d307873619c, you can execute the following:
docker container exec 0d bash -c "cat /etc/os-release"
This tells docker to exec the command pwd
on container id starting with 0d
(no need to enter the full id). To get an interactive terninal:
docker container exec -it 0d bash -c "cat /etc/os-release"
More docs (here)[https://docs.docker.com/reference/cli/docker/container/exec/]