kijai
97da701379
Fix DF sampler
2025-08-11 18:46:46 +03:00
kijai
7c93aea182
rename fp8 optimization in code for clarity as it does more than that with loras
2025-08-11 14:32:57 +03:00
kijai
48fa904ad8
fp8 matmul for scaled models
...
Fp8 matmul (fp8_fast) doesn't seem feasible with unmerged LoRAs as you'd need to first upcast, then apply LoRA, then downcast back to fp8 and that is too slow. Direct adding in fp8 is also not possible since that's just not something fp8 dtypes support.
2025-08-09 10:17:11 +03:00
kijai
9edab74562
Default non_blocking to False
...
too many RAM issues
2025-08-01 02:29:07 +03:00
kijai
3567202e75
Fix compile and support radial attn in the DF sampler
2025-07-27 14:19:31 +03:00
kijai
c08296747f
more refactoring
2025-07-18 16:31:34 +03:00
kijai
6bc53b771d
refactor scheduler import
2025-07-17 15:56:55 +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
ef40577b70
fixes
2025-05-30 12:29:38 +03:00
kijai
031d8ac817
disable enhance always with DF
2025-04-25 15:47:25 +03:00
kijai
6fadcbd957
Support Phantom, refactor model dtypes, reduce DF model memory use
2025-04-22 21:13:27 +03:00
kijai
6ba7bc811e
Update nodes.py
2025-04-22 02:05:08 +03:00
kijai
f93e54d21b
Allow unianimate with DF
2025-04-22 01:01:22 +03:00
kijai
e3ea2bf392
Allow VACE with DF
2025-04-21 20:01:20 +03:00
kijai
884069121f
Fix latent preview fallback if disabled
2025-04-21 19:11:16 +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