defaults/hunyuan_custom_audio.json

{
	"model":
	{
		"name": "Hunyuan Video Custom Audio 720p 13B",
		"architecture" : "hunyuan_custom_audio",
		"description": "The Hunyuan Video Custom Audio model can be used to generate scenes of a person speaking given a Reference Image and a Recorded Voice or Song. The reference image is not a start image and therefore one can represent the person in a different context.The video length can be anything up to 10s. It is also quite good to generate no sound Video based on a person.",
		"URLs": [
			"https://huggingface.co/DeepBeepMeep/HunyuanVideo/resolve/main/hunyuan_video_custom_audio_720_bf16.safetensors",
			"https://huggingface.co/DeepBeepMeep/HunyuanVideo/resolve/main/hunyuan_video_custom_audio_720_quanto_bf16_int8.safetensors"
		]
	}
}