Wan2.1 T2V 14B FusionX FP8 GGUF DEPRECATED

🚀 vrgamedevgirl84/Wan14BT2VFusionX GGUF Conversion

This repository provides a direct GGUF conversion of the vrgamedevgirl84/Wan14BT2VFusionX model, facilitating its use in relevant applications.

🚀 Quick Start

This repository contains a direct GGUF conversion of the vrgamedevgirl84/Wan14BT2VFusionX model, originally sourced from vrgamedevgirl84/Wan14BT2VFusionX.

All quantized versions were created from the base FP8 model WanT2V_MasterModel.safetensors using the conversion scripts provided by city96, available at the ComfyUI-GGUF GitHub repository.

For FP16 quantized version, please see QuantStack/Wan2.1-14B-T2V-FusionX-GGUF

The process involved first converting the safetensors model to a FP16 GGUF, then quantizing it, and finally applying the 5D fix.

✨ Features

• Compatibility: The model files are compatible with the ComfyUI - GGUF custom node.
• Quantization: All quantized versions are derived from the base FP8 model using specific conversion scripts.

📦 Installation

• Place the model files in the directory:
  ComfyUI/models/unet
• For detailed installation instructions, please refer to the ComfyUI - GGUF GitHub repository.

📚 Documentation

Additional Resources

• The VAE can be downloaded from Kijai’s repository on Hugging Face.

Reference

• For an overview of quantization types, please see the LLaMA 3 8B Scoreboard quantization chart.

📄 License

This project is licensed under the Apache - 2.0 license.

⚠️ Important Note

This is based on a lower - precision FP8 model and is now deprecated.

It has been superseded by an FP16 - based quantized version, available here:
QuantStack/Wan2.1-14B-T2V-FusionX-GGUF