Image Generation Agent

Image Generation Agent is an open source project aimed at helping you produce stunning images aligned with your prompts, through the automation of the prompt refinement-generation-visual feedback loop.

Installation

This is a uv project, thus you have to make sure uv is installed on your machine - if not, you can get it with:

pip install uv

Or follow the installation guidelines on uv docs.

Once uv is on your machine, you can clone this repository:

git clone https://github.com/run-llama/image-generation-agent
cd image-generation-agent

And run:

uv sync
source .venv/bin/activate

Congrats, you successfully installed this project and its dependencies!

Setting up

Access the scripts sub-folder, and modify the .env.example file so that it contains a valid GOOGLE_API_KEY and OPENAI_API_KEY. After that, make sure to rename it to .env:

cd scripts/
mv .env.example .env

Alternatively, you can export the keys as environmental variables from your terminal:

export GOOGLE_API_KEY="***"
export OPENAI_API_KEY="sk-***"

Launching

While still being in the scripts sub-folder, you can launch the backend of the project with:

python3 server.py

You will se a log message telling you that the websocket server has been started on port 8765.

After the websocket has connected, launch the frontend from another terminal window (remaining in the scripts sub-folder and with the virtual environment active), using:

python3 client.py

With this command, you'll have the application running on http://localhost:7860 and you will be able to interact with it!

How does it work?

The agent uses two tools under the hood:

• generate_image: this exploits OpenAI image generation API to create images starting from textual prompts.
• evaluate_generated_image: this uses the advanced vision capabilities of Gemini, employing the model as a judge of the faithfulness and quality of the image

Whenever you submit a request, the agent first activates the generate_image tool, then it assess the fit of the image with your prompt using the evaluate_generated_image tool. If the image is deemed suitable, it is returned to the user, whereas otherwise the prompt is refined and the generate-evaluate loop is resumed.

Contributing

Contributions are more than welcome! Follow the contribution guidelines to make sure your contribution is compliant with the repo's requirements :)

License and rights of usage

This is an open-source project distributed under an MIT License.