Hugging Face
Models
Datasets
Spaces
Buckets
new
Docs
Enterprise
Pricing
Website
Tasks
HuggingChat
Collections
Languages
Organizations
Community
Blog
Posts
Daily Papers
Hardware
Learn
Discord
Forum
GitHub
Solutions
Team & Enterprise
Hugging Face PRO
Enterprise Support
Inference Providers
Inference Endpoints
Storage Buckets
Log In
Sign Up
Spaces:
multimodalart
/
MiniMax-H3-images
like
2
Running
on
Zero
App
Files
Files
Community
Fetching metadata from the HF Docker repository...
main
MiniMax-H3-images
52.5 kB
Ctrl+K
Ctrl+K
1 contributor
History:
10 commits
multimodalart
HF Staff
Drop the NCII guard: it scores edit prompts and there is no image input
ac50538
verified
about 12 hours ago
.gitattributes
Safe
1.52 kB
initial commit
about 16 hours ago
README.md
Safe
10.9 kB
Document the reference decode, the slice-0 finding and the measured latencies
about 14 hours ago
app.py
Safe
27 kB
Drop the NCII guard: it scores edit prompts and there is no image input
about 12 hours ago
h3_image_blocks.py
Safe
2.78 kB
MiniMax-H3 as an image model: 5-frame floor, single-slice decode
about 16 hours ago
h3_single_frame_vae.py
Safe
6.65 kB
Decode single-frame slices through 512 px tiles
about 12 hours ago
load_decoder.py
Safe
2.2 kB
Reference decode recipe verbatim; length + latent_index controls; vendored strict loader
about 14 hours ago
requirements.txt
Safe
1.46 kB
Add the missing peft pin; without it the turbo LoRA cannot attach
about 16 hours ago
