tonykipkemboi / streamlit-replicate-img-app

Using Replicate to build Streamlit app for image generations!

✨ Image Generation App ✨

Image Generator App: where art meets algorithms and dreams meet pixels! 🚀

Overview

Powered by cutting-edge AI models running on Replicate and wrapped in a Streamlit interface, this app lets you transform plain text prompts into mesmerizing visual masterpieces.

Technical Features

• Neural Model: Leverages the power of the replicate.run model for image generation, providing detailed and accurate depictions.
• Streamlit Framework: Built atop the versatile Streamlit library, ensuring a smooth and responsive UI/UX.
• Dynamic Customization: You can peek "under the hood", tune hyperparameters like guidance_scale, prompt_strength, and more for fine-grained control.
• Gallery: A curated gallery for inspiration, showcasing the prowess of the underlying model.

Getting Started

1. Clone the repository:

   git clone https://github.com/tonykipkemboi/streamlit-replicate-img-app.git

2. Navigate to the project directory:

   cd streamlit-replicate-img-app

3. Install the dependencies:

   pip install -r requirements.txt

4. Rename the .streamlit/example_secrets.toml file to .streamlit/secrets.toml.

5. Paste your Replicate API token in the secrets.toml file:

   REPLICATE_API_TOKEN = "paste-your-replicate-api-token-here"

Usage

1. Run the Streamlit app:

   streamlit run streamlit_app.py

2. Navigate to the provided local URL, and voila! Start crafting your visual narratives.

Contributions

Your insights can make this tool even better! Feel free to fork, make enhancements, and raise a PR.

Attribution

• Developed by: The wizards over at Stability AI 🧙‍♂️

• Model type: Diffusion-based text-to-image generative model

• License: CreativeML Open RAIL++-M License

• Model Description: This is a model that can be used to generate and modify images based on text prompts. It is a Latent Diffusion Model that uses two fixed, pretrained text encoders (OpenCLIP-ViT/G and CLIP-ViT/L).

• Resources for more information: Check out our GitHub Repository and the SDXL report on arXiv.