This repository contains resources and code for utilizing Latent CLIP for zero-shot prediction and reward-based noise optimization.

To set up the environment, use the provided environment.yml file:

conda env create -f environment.yml
conda activate latentclipenv

The starting point for understanding and utilizing Latent CLIP is the Jupyter Notebook:

minimal_usage_latent_clip.ipynb

This notebook demonstrates:

• Zero-shot prediction using Latent CLIP.
• Reward-based noise optimization using Latent CLIP-based rewards .

The file workflow_sdxl_turbo.json is a workflow designed for use with ComfyUI.

Steps to use the workflow:

1. Clone ComfyUI from GitHub:

   git clone https://github.com/comfyanonymous/ComfyUI.git
   cd ComfyUI

2. Run ComfyUI:

   python main.py

3. Load the Workflow:

   • In the ComfyUI interface, press Ctrl + O (or click "Load").
   • Select the file workflow_sdxl_turbo.json from the assets/ folder.

4. For more information on ComfyUI, visit:
   ➡️ ComfyUI GitHub Repository

supplementary/
│
├── assets/            # Additional resources
│   └── workflow_sdxl_turbo.json  # ComfyUI workflow file (https://github.com/comfyanonymous/ComfyUI)
│
├── Latent_ReNO/       # Implementation for reward-based noise optimization
├── environment.yml    # Conda environment setup file
├── helper.py          # Utility functions for supporting the notebook
└── minimal_usage_latent_clip.ipynb  # Main notebook for starting with Latent CLIP