Steps to Create an AI Project Step 1: Initialize the Project Create a project using Poetry:
poetry new your_project_name --name app
cd your_project_name
poetry shell
-
Note: If you want to run your project locally, you can select the kernel (virtual environment) that you created by using the Poetry shell. Inside the project directory, create a Dockerfile. Step 2: Setup Docker Compose Create a file named compose.yaml with the necessary configuration. Run the Docker Compose command to set up the environment:
docker compose up
Add necessary packages to pyproject.toml in the devcontainer. To add the OpenAI package, use:
poetry add openai
Step 3: Write and Run Code Write all the code in a Jupyter Notebook file (.ipynb). Run the cells in the Jupyter Notebook to execute the code.
-
API Keys Security: OpenAI keys must be deleted after the project is complete.
-
Environment Variables:
-
Store the keys in a .env file.
-
Create a file named .backupenv. If working on a client project, the client should rename .backupenv to .env and write their API keys in this file.
-
Git Ignore: Ensure the .env file is included in .gitignore to prevent sensitive information from being committed to the repository:
.env