ccc0168 / docker-open-interpreter

Docker Open Interpreter

Run open-interpreter in Docker.

DockerHub Image Repo: https://hub.docker.com/r/guoqiao/open-interpreter

GitHub Code Repo: https://github.com/guoqiao/docker-open-interpreter

Run Open Interpreter with Docker Image

Run with gpt-4 model

OPENAI_API_KEY=sk-xxx
docker run -it -e OPENAI_API_KEY=${OPENAI_API_KEY} guoqiao/open-interpreter:latest

NOTE:

• your api key must have gpt-4 access
• subscription to chatgpt plus doesn't mean your api key has gpt-4 access

Run with gpt-3.5-turbo model

OPENAI_API_KEY=sk-xxx
docker run -it -e OPENAI_API_KEY=${OPENAI_API_KEY} guoqiao/open-interpreter:latest interpreter --fast

Run with code-llama models

OPEN_INTERPRETER_MODELS_DIR=/mnt/storage/open-interpreter-models

docker run -it \
    -v ${OPEN_INTERPRETER_MODELS_DIR}:/root/.local/share/"Open Interpreter"/models  \
    guoqiao/open-interpreter:latest \
    interpreter --local

NOTE:

• models are large(e.g.: 2.7G ~ 34G), here we define a folder to download and store them.
• ideally that folder is on your storage drive

Run with falcon models

OPEN_INTERPRETER_MODELS_DIR=/mnt/storage/open-interpreter-models

docker run -it \
    -v ${OPEN_INTERPRETER_MODELS_DIR}:/root/.local/share/"Open Interpreter"/models  \
    guoqiao/open-interpreter:latest \
    interpreter --falcon