b85216a3c0
Lower T5 memory usage by a few hundred MB.
7 months ago
25853d0be8
Use common function for casting weights to input.
7 months ago
bb1969cab7
Initial support for the stable audio open model.
8 months ago
6c23854f54
Fix OSX latent2rgb previews.
9 months ago
448d9263a2
Fix control loras breaking.
12 months ago
db8b59ecff
Lower memory usage for loras in lowvram mode at the cost of perf.
12 months ago
667c92814e
Stable Cascade Stage B.
1 year ago
78a70fda87
Remove useless import.
1 year ago
36a7953142
Greatly improve lowvram sampling speed by getting rid of accelerate.
...
Let me know if this breaks anything.
1 year ago
77755ab8db
Refactor comfy.ops
...
comfy.ops -> comfy.ops.disable_weight_init
This should make it more clear what they actually do.
Some unused code has also been removed.
1 year ago
ba07cb748e
Use faster manual cast for fp8 in unet.
1 year ago
57926635e8
Switch text encoder to manual cast.
...
Use fp16 text encoder weights for CPU inference to lower memory usage.
1 year ago
af365e4dd1
All the unet ops with weights are now handled by comfy.ops
1 year ago
412d3ff57d
Refactor.
1 year ago
00c0b2c507
Initialize text encoder to target dtype.
2 years ago
d6e4b342e6
Support for Control Loras.
...
Control loras are controlnets where some of the weights are stored in
"lora" format: an up and a down low rank matrice that when multiplied
together and added to the unet weight give the controlnet weight.
This allows a much smaller memory footprint depending on the rank of the
matrices.
These controlnets are used just like regular ones.
2 years ago
bb1f45d6e8
Properly disable weight initialization in clip models.
2 years ago
21f04fe632
Disable default weight values in unet conv2d for faster loading.
2 years ago
6971646b8b
Speed up model loading a bit.
...
Default pytorch Linear initializes the weights which is useless and slow.
2 years ago
comfyanonymous