| title | emoji | colorFrom | colorTo | sdk | pinned | app_port |
|---|---|---|---|---|---|---|
GenAI Detection |
🖼️ |
red |
green |
docker |
false |
7860 |
This is GenAI Image Classification project built with Kedro library.
We are trying to discriminate between images that are taken in the real world or created by hand and images generated using generative models like DALLE, Midjourney, or Stable Diffusion.
From our world:
AI generated:
Docker is required to run the project. To build the Docker image: Clone the repo and run the following command:
docker build -f enivronments/Dockerfile.kedro -t deep-learning-project .
or just run:
docker pull kabanosk/deep-learning-project
There are other Dockerfiles in the environments directory that can be used to build images for different purposes.
Dockerfile.kedro- image with entrypointkedro runDockerfile.api- image with api and gradio dependecy groupDockerfile.test- image with test dependecy groupDockerfile- image with dev dependecy group
To run the training and evaluation pipeline, you need to run the Docker container with the following command
# training
docker run -v ./data:/data --env WANDB_API_KEY=<WANDB_API_KEY> deep-learning-project --pipeline training
# evaluation
docker run -v ./data:/data --env WANDB_API_KEY=<WANDB_API_KEY> deep-learning-project --pipeline evaluation
If you want to run Kedro manualy, you can run:
docker run -v ./data:/data --expose 4242 -it --entrypoint /bin/bash deep-learning-project
Then you can run Kedro commands:
kedro run
To visualize project run:
poetry install --with dev
kedro viz run --port 4242
Then you can open browser and go to http://localhost:4242/
To run API see: API
This project is licensed under the MIT License.
For any questions or concerns, please contact us at email.