flux_dev_scaled_fp8_test

Introduction

The FLUX_DEV_SCALED_FP8_TEST is a test model developed by Comfyanonymous, designed to deliver improved results compared to the regular FP8 model. It is intended for use with the latest version of ComfyUI.

Architecture

This model utilizes the "scaled FP8" technology, which is expected to offer enhancements over the traditional FP8 models. It integrates seamlessly with the ComfyUI ecosystem, specifically with diffusion models.

Training

Details on the specific training process for the FLUX_DEV_SCALED_FP8_TEST model are not provided. However, the emphasis is on its compatibility with ComfyUI and its expected superior performance over traditional models.

Guide: Running Locally

1. Download the Model: Obtain the FLUX_DEV_SCALED_FP8_TEST model file from the repository.
2. Install ComfyUI: Ensure you have the newest version of ComfyUI installed from GitHub.
3. Place Model in Directory: Move the model file into your ComfyUI/models/diffusion_models/ directory.
4. Load the Model: Use the "Load Diffusion Model" node within ComfyUI to load and start utilizing the model.

Cloud GPUs

For optimal performance, consider using cloud-based GPU services such as AWS, Google Cloud Platform, or Azure to handle intensive computational requirements.

License

The FLUX_DEV_SCALED_FP8_TEST model is released under the flux-1-dev-non-commercial-license. For more details, review the license agreement.