text2image

This is a Library to generate images from text.

License
Environment
Technology

Demo

Example of generated images.

A cat in a hat	The photo of Tokyo Tower

Requirements

• Docker and docker-compose are required. The versions are as follows.

  • Docker: v26.0.2
  • Docker Compose: v2.22.0

• Nvidia Driver is required for CUDA 12.5.

Getting Started

1. Clone Repository

$ git clone git@github.com:nglcobdai/text2image.git
$ cd text2image

2. Create .env file

Copy .env.example to .env

$ cp .env.example .env.dev

Edit .env.dev file

Key	Description
DATADRIVE	The path to the directory where generated image save.
HUGGINGFACE_API_TOKEN	The API token for Hugging Face. This token is required to access the FLUX.1-dev model.

3. Docker Build & Run

$ docker-compose build --no-cache
$ docker-compose run --rm dev

4. Run Python Script

$ python src/__main__.py