Hunyuan video with TeaCache speed up🚀! 混元視頻模型+TeaCache_200%加速工作流
5.0
0 reviews10
4.5K
1.4K
1
Description
Hunyuan video model with Teacache speed up 200%🚀!
混元視頻模型+TeaCache_200%加速工作流
- 顯卡: 4090d
- 生成ai視頻長度：3秒視頻 （可以調節視頻的長度和分辨率）
- 無TeaCache： 153秒
- TeaCache 1.6X加速: 105秒
- TeaCache 2.1X加速: 39秒
- 網紅小姐姐跳舞作品展示： https://kevin-machine-learning.notion.site/AI-Hunyuan-200-17930ccdc7828082b37ef35249d65fee
- 保姆級教程視頻地址： https://youtu.be/NLukkCU292I?si=QFePNfiAfc4VvcUx
混元視頻模型 (Hunyuan video model)+ Lora模型，工作流下載地址：
Discussion
(No comments yet)
Loading...
Reviews
No reviews yet
Versions (1)
- latest (a year ago)
Node Details
Primitive Nodes (7)
EmptyHunyuanLatentVideo (1)
FluxGuidance (1)
ModelSamplingSD3 (1)
Note (1)
Reroute (2)
TeaCacheHunyuanVideoSampler_FOK (1)
Custom Nodes (11)
ComfyUI
- VAELoader (1)
- UNETLoader (1)
- DualCLIPLoader (1)
- CLIPTextEncode (1)
- RandomNoise (1)
- BasicGuider (1)
- KSamplerSelect (1)
- BasicScheduler (1)
- VAEDecodeTiled (1)
- VAEDecode (1)
- VHS_VideoCombine (1)
Model Details
Checkpoints (0)
LoRAs (0)