HunyuanVideo-1.5-720p-fp8

HunyuanVideo-1.5-720p-fp8

This is a repackaged version of <a href="https://huggingface.co/tencent/HunyuanVideo-1.5"><b>Tencent's HunyuanVideo-1.5</b></a>, containing the 720p T2V transformer quantized to fp8 along with essential components required for inference. This optimized package offers reduced memory footprint and faster inference while maintaining excellent generation quality.

Features

• FP8 Quantization: Reduced memory usage and faster inference without significant quality loss
• Optimized Package: Contains only the 720p_t2v transformer (fp8) and essential components
• Lower VRAM Requirements: Enables deployment on GPUs with less memory
• Efficient Inference: Faster loading times and reduced model size
• High-Quality 720p Generation: Maintains the quality of the original HunyuanVideo-1.5 model at 720p resolution
• Based on State-of-the-Art: Built on Tencent's advanced video generation technology

Usage

This model is planned to work with Aquiles-Image, providing an OpenAI-compatible API for video generation:

pip install aquiles-image
aquiles-image serve --model "hunyuanVideo-1.5-720p-fp8"

Learn more in the full documentation.

Example of a Video Generated with This Model

<div align="center"> <video width="80%" controls> <source src="https://cdn-uploads.huggingface.co/production/uploads/63cb46191b705cc951e88e6c/RHZ9Le60ov6wUS2N6o-8t.mp4" type="video/mp4"> Your browser does not support the video tag. </video> </div>

Generated with prompt: A direct continuation of the existing shot of a chameleon crawling slowly along a mossy branch. Begin with the chameleon already mid-step, camera tracking right at the same close, eye-level angle. After three seconds, its eyes swivel independently, one pausing to glance toward the lens before it resumes moving forward. Maintain the 100 mm anamorphic lens with shallow depth of field, dappled rainforest light, faint humidity haze, and subtle film grain. The moss texture and background greenery should remain consistent, with the chameleon's deliberate gait flowing naturally as if no cut occurred.

Acknowledgments

We extend our gratitude to:

• The Tencent Hunyuan Team for developing and open-sourcing HunyuanVideo-1.5
• The lightx2v team (<a href="https://huggingface.co/lightx2v">@lightx2v</a>) for providing the high-quality fp8 quantized weights

This repackaged version is made possible by their excellent work and commitment to the open-source community.

For more information about the original model, please visit the official HunyuanVideo-1.5 repository.