114 GB

Ctrl+K

1 contributor

card: field notes from discussions #1/#2/#4 — 24 GB frame budgets, keyframe VAE encode on CPU, kill disarmed during prompt encode (0.5.80), P-core QoS

41b5099 verified 1 day ago

images
images: 4x face-crop A/B panels, clipproj v1 vs v2 2 days ago
samples
samples: ClipProj 4B variant, same prompt and seed as the q4tp/q2tp A-B 4 days ago
.gitattributes

3.27 kB
clipproj4b-fl2va v2: split projection — vision rows through a vision-fitted map, faces hold 2 days ago
README.md

31 kB
card: field notes from discussions #1/#2/#4 — 24 GB frame budgets, keyframe VAE encode on CPU, kill disarmed during prompt encode (0.5.80), P-core QoS 1 day ago
mmh3-turbo-clipproj4b-fl2va-q4tp.cmf
14.5 GB
xet

clipproj + fl2va in one 14.5 GB file: the 4B tower and the VAE encoder join the compact build; first-frame conditioning verified frame-for-frame 3 days ago
mmh3-turbo-clipproj4b-fl2va-v2-q4tp.cmf

14.5 GB
xet

clipproj4b-fl2va v2: split projection — vision rows through a vision-fitted map, faces hold 2 days ago
mmh3-turbo-clipproj4b-fl2va-v2-q4tp.cmf.sha256

106 Bytes
clipproj4b-fl2va v2: split projection — vision rows through a vision-fitted map, faces hold 2 days ago
mmh3-turbo-clipproj4b-q4tp.cmf

14.1 GB
xet

ClipProj 4B prompt encoder: 25.20 GB -> 14.13 GB, everything still q4tp 4 days ago
mmh3-turbo-fl2va-q2tp.cmf

20.1 GB
xet

Upload mmh3-turbo-fl2va-q2tp.cmf with huggingface_hub 10 days ago
mmh3-turbo-fl2va-q4tp.cmf

25.7 GB
xet

Upload mmh3-turbo-fl2va-q4tp.cmf with huggingface_hub 10 days ago
mmh3-turbo-q4tp.cmf

25.2 GB
xet

Upload mmh3-turbo-q4tp.cmf with huggingface_hub 11 days ago