Sorry, due to company policy we will not be making the training code public for Hyper-SD. But feel free to follow what our team present next, we'll do our best to make it happen next time.
Yanzuo Lu
oliveryanzuolu
AI & ML interests
None yet
Recent Activity
upvoted
a
paper
18 days ago
Distribution Matching Variational AutoEncoder
upvoted
a
paper
18 days ago
Unified Video Editing with Temporal Reasoner
upvoted
a
paper
21 days ago
Live Avatar: Streaming Real-time Audio-Driven Avatar Generation with Infinite Length
Organizations
replied to
their
post
over 1 year ago
posted
an
update
over 1 year ago
Post
2710
Thrilled to introduce our Hyper-SD, offering hyper-fast⚡️ and hyper-quality✨ text-to-image generation.
Our model achieves SINGLE-STEP inference on both SD1.5 and SDXL architecture without evident losses of aesthetics, styles and structures.
Project page: https://hyper-sd.github.io
HuggingFace repo: ByteDance/Hyper-SD
Our model achieves SINGLE-STEP inference on both SD1.5 and SDXL architecture without evident losses of aesthetics, styles and structures.
Project page: https://hyper-sd.github.io
HuggingFace repo: ByteDance/Hyper-SD
-
- 3 replies