Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/ashakoen/replicate-image-gen-python
A simple Python image generation API intended to be deployed to Google Cloud Run. The script uses the Replicate API to generate images using FLUX.1
https://github.com/ashakoen/replicate-image-gen-python
Last synced: about 22 hours ago
JSON representation
A simple Python image generation API intended to be deployed to Google Cloud Run. The script uses the Replicate API to generate images using FLUX.1
- Host: GitHub
- URL: https://github.com/ashakoen/replicate-image-gen-python
- Owner: ashakoen
- License: mit
- Created: 2024-08-28T23:44:57.000Z (3 months ago)
- Default Branch: main
- Last Pushed: 2024-08-29T21:48:19.000Z (3 months ago)
- Last Synced: 2024-08-30T01:03:10.608Z (3 months ago)
- Language: Python
- Size: 9.77 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Replicate Image Gen
Replicate Image Gen is a Python application that leverages the Replicate API to generate images using the FLUX.1 models based on text prompts. It's the backend of a bigger app I'm building to interact with the Replicate APIs.
## Features
- Generates images from text prompts using FLUX.1 models and LoRAs.
- Wide customization with environment variables for model parameters.
- Secure API access with header-based authentication (`x-api-key`).
- Designed for deployment on Google Cloud Run with Docker support.
- Environment variable management via `.env` files for easy configuration.
## Installation
### Clone the Repository
First, clone the repository to your local machine:
```
git clone https://github.com/ashakoen/replicate-image-gen-python.git
cd replicate-image-gen-python
```
### Python Environment
Ensure you have Python 3.10 or later installed on your system.
### Install Dependencies
Either install with pip (assuming a virtual environment is set up):
```
pip install -r requirements.txt
```
Or if using Docker, you don't need to manually install dependencies; the Dockerfile will handle it.
## Usage
### Running the Docker Container
Build the Docker image:
```
docker build -t flask-image-generator .
```
Run the Docker container with the `.env` file:
```
docker run -p 8080:8080 --env-file .env flask-image-generator
```
### Interacting with the API
After starting the container, you can interact with the API. The service provides an endpoint to generate images based on the user-provided text prompt.
#### Endpoint: `/generate-images`
- **Method:** POST
- **Headers:**
- `Content-Type: application/json`
- `x-api-key: `
- **Body:** JSON object containing the field:
- `prompt`: A string representing the text input for generating images.
Example usage with `curl`:
```
curl -X POST http://localhost:8080/generate-images \
-H "x-api-key: your_header_api_key_here" \
-H "Content-Type: application/json" \
-d '{"prompt": "A cute photo of a dog"}'
```
### Docker Compose
You can use Docker Compose as well for managing the container:
```
docker-compose up --build
```
## Configuration
The application uses a `.env` file to configure the parameters used by the Replicate API. See below for the complete list of configurable environment variables:
```
- REPLICATE_API_KEY: Your API key for Replicate.
- MODEL: The model to use (e.g., "dev").
- NUM_OUTPUTS: The number of output images (default: 1).
- GUIDANCE_SCALE: The guidance scale for the generation process.
- NUM_INFERENCE_STEPS: The number of inference steps.
- OUTPUT_FORMAT: The format for the output image (e.g., "png").
- OUTPUT_QUALITY: The quality of the output image (e.g., 100).
- DISABLE_SAFETY_CHECKER: Disables the safety checker (boolean: true/false).
- LORA_SCALE: Scale for LoRA fine-tuning (default: 1).
- ASPECT_RATIO: Aspect ratio of output image (e.g., "9:16").
- LORA: The LoRA model to apply during image generation.
- X_API_KEY: The key required for accessing the API.
```
### Google Cloud Run Integration
This repository can be easily integrated with Google Cloud Run for continuous deployment:
1. **Link Your GitHub Repository:** Use the Google Cloud Console to set up a continuous integration pipeline with your GitHub repository.
2. **Environment Variables:** Ensure the required environment variables are set either in the Google Cloud Run console or by using Google Cloud's Secret Manager for secure storage.
3. **GitHub Deployment:**
- Push your changes to GitHub.
- Google Cloud Run will automatically build and deploy your service.
## The Story Behind Flask Image Generator
The Flask Image Generator was developed to simplify the process of generating high-quality images based on detailed text prompts. With the rise of AI models capable of creating images from textual descriptions, this project aimed to provide a simple, yet powerful, platform for testing and integrating such features into various applications.
By using the Replicate API, the project reduces the complexity of model management and leverages pre-trained models for quick deployment and testing. It’s designed to be easily customizable and ready to be integrated into cloud environments, making it ideal for prototyping and production use alike.
## File Structure
- app.py: The main Flask application for processing image generation requests.
- requirements.txt: Python dependencies.
- Dockerfile: Instructions for building the Docker image.
- docker-compose.yml: Simplifies local container management.
- .env: Environment variable configurations (not included in repository).
- .dockerignore: Excludes unnecessary files from the Docker context.
- README.md: This file, providing information about the project.
## Dependencies
This project uses several Python libraries detailed in the `requirements.txt` file:
- Flask: Micro web framework for routing and handling API requests.
- Flask-CORS: Handles cross-origin requests.
- Loguru: Logging library for structured logs.
- Replicate: Python client for integrating with the Replicate API.
- python-dotenv: Loads environment variables from a `.env` file.
## License
This project is licensed under the MIT License. See the LICENSE file for details.
## Contributing
Contributions are welcome! Feel free to open an issue or submit a pull request with any improvements or bug fixes.
## Contact
For any questions or suggestions, open a Github issue.