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
apozz
/
WorldStereo-fp8
like
0
Image-to-Video
Safetensors
fp8
quantized
video
wan
worldstereo
License:
mit
Model card
Files
Files and versions
xet
Community
Copy to bucket
new
main
WorldStereo-fp8
20.4 GB
Ctrl+K
Ctrl+K
1 contributor
History:
10 commits
apozz
fix: re-cook (Linear-only) -- previous version cast 2 Conv3d weights as fp8 which broke load_state_dict
af612e1
verified
3 months ago
.gitattributes
Safe
1.52 kB
initial commit
3 months ago
LICENSE
1.36 kB
add LICENSE
3 months ago
NOTICE
1.21 kB
add NOTICE
3 months ago
README.md
6.28 kB
update: 549 fp8 (Linear-only), Conv weights stay bf16
3 months ago
WorldStereo-memory-dmd-fp8_e4m3fn_scaled.safetensors
20.4 GB
xet
fix: re-cook (Linear-only) -- previous version cast 2 Conv3d weights as fp8 which broke load_state_dict
3 months ago
config.json
3.91 kB
add config.json
3 months ago
cook_worldstereo_fp8.py
7.48 kB
fix: is_castable_weight now requires len(shape)==2 (Linear only); Conv weights stay bf16
3 months ago
