kijai
974dd656da
some offloading for multitalk
2025-07-03 21:56:19 +03:00
kijai and Rudra-ai-coder
475f371016
Multiple talkers
...
Initial commit, works but needs more utility for the mask creation.
Based mostly on Rudra-ai-coder's modifications.
Co-Authored-By: Rudra-ai-coder <177262225+rudra-ai-coder@users.noreply.github.com >
2025-07-03 19:03:12 +03:00
kijai
f621d2a5fc
Update model.py
2025-06-30 20:31:39 +03:00
kijai
2da36b7fda
context window progress bar
...
better than nothing, and allow cancelling mid context
2025-06-20 17:29:27 +03:00
kijai
157fbb1b7f
Merge branch 'main' into multitalk
2025-06-20 13:53:31 +03:00
kijai
8387a64b15
bump version
2025-06-20 10:57:13 +03:00
kijai
09b4c3a865
support context windows, torch compile fixes
2025-06-19 20:29:32 +03:00
kijai
b5cae8bf1a
Fix FLF2V model
2025-06-18 15:49:01 +03:00
kijai
58104b620f
init
2025-06-18 15:41:53 +03:00
kijai
b322192d93
Support minimaxremover
...
https://minimax-remover.github.io/
2025-06-17 09:45:12 +03:00
kijai
ad43eed0d0
Add MagCache and consolidate cache_args to support both
...
Not really getting great results yet with MagCache, but at least pretty much on bar with TeaCache so it can be an option that hopefully improves in time
2025-06-13 12:31:13 +03:00
kijai
e7c39757f6
Allow compiling crossattn
...
I don't remember why I even disabled this... seems to work, big effect with NAG
2025-06-12 19:04:15 +03:00
kijai
d34cabc985
Cleanup, create own node for NAG
2025-06-12 18:18:28 +03:00
kijai
23796233df
Some fixes and adjustments
2025-06-12 17:26:54 +03:00
kabachuha
803ba1cd72
rm commented out code
2025-06-12 15:34:11 +03:00
kabachuha
e8d3f1cca5
fixup when context lens are none
2025-06-12 15:30:30 +03:00
kabachuha
bdf7a3d02a
pass nag scale into the model
2025-06-12 15:26:59 +03:00
kabachuha
0ac366ba03
add nag to attention internal parts
2025-06-12 14:43:51 +03:00
kijai
a290f75b5e
Update model.py
2025-06-05 23:41:17 +03:00
kijai
e2f42b5773
realisdance
2025-06-03 08:02:59 +03:00
kijai
cd2884d88a
Initial ATI support
...
https://github.com/bytedance/ATI
2025-05-31 01:17:18 +03:00
kijai
fb73022a06
controlnet device fix
2025-05-29 18:42:46 +03:00
kijai
129f368380
Fix uni3c on fp16 and add strength parameters
2025-05-29 13:25:52 +03:00
kijai
23e3368de3
Initial Dilated Controlnet support
...
https://huggingface.co/TheDenk/wan2.1-t2v-1.3b-controlnet-hed-v1
https://huggingface.co/TheDenk/wan2.1-t2v-14b-controlnet-hed-v1
2025-05-26 21:54:48 +03:00
kijai
ef1ed29178
partial Uni3C implementation
2025-05-26 12:35:36 +03:00
kijai
7875031efe
Support list input for VACE strength
...
Allows finer control of VACE strength over timesteps
2025-05-21 19:04:25 +03:00
kijai
fa7015315a
Possibly reduce TeaCache VRAM use a bit
...
Uses 500-700MB less at 1280x720
2025-05-20 21:07:15 +03:00
kijai
abdac5ec3a
Update model.py
2025-05-14 10:31:44 +03:00
kijai
bfed2afc2e
Add flex_attention for CausVid
2025-05-13 21:35:34 +03:00
kijai
f2bc29b931
Fix TeaCache when not using coefficients
2025-05-05 21:58:15 +03:00
kijai
77f78ceded
Support Fun Control-Camera
2025-04-29 20:41:27 +03:00
kijai
ca0974d5f0
cleanup
2025-04-29 01:46:39 +03:00
kijai
8117c6f033
Context windows with FantasyTalking
2025-04-28 20:24:41 +03:00
kijai
5683d8306a
Support FantasyTalking
2025-04-28 19:10:32 +03:00
kijai
4fc4159c88
fix up default rope dtype
2025-04-27 17:22:02 +03:00
kijai
fc7ab666a2
support loading Fun camera model
...
Input not functional yet
2025-04-25 17:02:10 +03:00
kijai
33b8f1c082
Support Fun control 1.1
...
Supports the new reference image input to the 1.1 control model
2025-04-25 15:44:18 +03:00
kijai
863829d083
dtype fixes and maybe allow fp8_fast to work
...
Unsure when this has changed but seems that _scaled_mm now works with same dtype? This fixes the quality degradation of fp8_fast in initial tests.
2025-04-25 03:12:10 +03:00
kijai
c7b2635f01
Update model.py
2025-04-23 20:37:48 +03:00
kijai
b3f3b0cd93
Update model.py
2025-04-23 20:11:08 +03:00
kijai
e0e5fcf713
Add start/end percent for Phantom and other small fixes
2025-04-23 01:00:41 +03:00
kijai
6fadcbd957
Support Phantom, refactor model dtypes, reduce DF model memory use
2025-04-22 21:13:27 +03:00
kijai
949c887e0c
Fix FLF2V when using vram management node
2025-04-22 09:44:46 +03:00
kijai
f93e54d21b
Allow unianimate with DF
2025-04-22 01:01:22 +03:00
kijai
f371c0a736
Allow TeaCache for DF
...
Guess it works
2025-04-21 18:30:40 +03:00
kijai
65f5505fca
Initial Skyreels DiffusionForcing model support
...
On it's own sampler at least for now while I figure out how it all works.
2025-04-21 18:13:51 +03:00
kijai
2cfae9fa8f
Make UniAnimate ref pose optional
2025-04-19 17:41:04 +03:00
kijai
421e375f13
Strength controls for Unianimate
2025-04-18 23:06:49 +03:00
kijai
7462356743
Support UniAnimate-DiT
...
https://github.com/ali-vilab/UniAnimate-DiT
2025-04-18 19:43:40 +03:00
kijai
8257cd1f8a
Fix TeaCache end step
2025-04-17 01:18:42 +03:00
kijai