logtd
|
e1cb93c383
|
Fix model and cond transformer options merge
|
9 months ago |
comfyanonymous
|
10847dfafe
|
Cleanup uni_pc inpainting.
This causes some small changes to the uni pc inpainting behavior but it
seems to improve results slightly.
|
9 months ago |
comfyanonymous
|
18c151b3e3
|
Add some latent2rgb matrices for previews.
|
9 months ago |
comfyanonymous
|
0d0fbabd1d
|
Pass pooled CLIP to stage b.
|
9 months ago |
comfyanonymous
|
c6b7a157ed
|
Align simple scheduling closer to official stable cascade scheduler.
|
9 months ago |
comfyanonymous
|
88f300401c
|
Enable fp16 by default on mps.
|
9 months ago |
comfyanonymous
|
e93cdd0ad0
|
Remove print.
|
9 months ago |
comfyanonymous
|
3711b31dff
|
Support Stable Cascade in checkpoint format.
|
9 months ago |
comfyanonymous
|
d91f45ef28
|
Some cleanups to how the text encoders are loaded.
|
9 months ago |
comfyanonymous
|
a7b5eaa7e3
|
Forgot to commit this.
|
9 months ago |
comfyanonymous
|
3b2e579926
|
Support loading the Stable Cascade effnet and previewer as a VAE.
The effnet can be used to encode images for img2img with Stage C.
|
9 months ago |
comfyanonymous
|
dccca1daa5
|
Fix gligen lowvram mode.
|
9 months ago |
comfyanonymous
|
8b60d33bb7
|
Add ModelSamplingStableCascade to control the shift sampling parameter.
shift is 2.0 by default on Stage C and 1.0 by default on Stage B.
|
9 months ago |
comfyanonymous
|
6bcf57ff10
|
Fix attention masks properly for multiple batches.
|
9 months ago |
comfyanonymous
|
11e3221f1f
|
fp8 weight support for Stable Cascade.
|
9 months ago |
comfyanonymous
|
f8706546f3
|
Fix attention mask batch size in some attention functions.
|
9 months ago |
comfyanonymous
|
3b9969c1c5
|
Properly fix attention masks in CLIP with batches.
|
9 months ago |
comfyanonymous
|
5b40e7a5ed
|
Implement shift schedule for cascade stage C.
|
9 months ago |
comfyanonymous
|
929e266f3e
|
Manual cast for bf16 on older GPUs.
|
9 months ago |
comfyanonymous
|
6c875d846b
|
Fix clip attention mask issues on some hardware.
|
9 months ago |
comfyanonymous
|
805c36ac9c
|
Make Stable Cascade work on old pytorch 2.0
|
9 months ago |
comfyanonymous
|
f2d1d16f4f
|
Support Stable Cascade Stage B lite.
|
9 months ago |
comfyanonymous
|
0b3c50480c
|
Make --force-fp32 disable loading models in bf16.
|
9 months ago |
comfyanonymous
|
97d03ae04a
|
StableCascade CLIP model support.
|
9 months ago |
comfyanonymous
|
667c92814e
|
Stable Cascade Stage B.
|
9 months ago |
comfyanonymous
|
f83109f09b
|
Stable Cascade Stage C.
|
9 months ago |
comfyanonymous
|
5e06baf112
|
Stable Cascade Stage A.
|
9 months ago |
comfyanonymous
|
aeaeca10bd
|
Small refactor of is_device_* functions.
|
9 months ago |
comfyanonymous
|
38b7ac6e26
|
Don't init the CLIP model when the checkpoint has no CLIP weights.
|
9 months ago |
Jedrzej Kosinski
|
f44225fd5f
|
Fix infinite while loop being possible in ddim_scheduler
|
9 months ago |
comfyanonymous
|
25a4805e51
|
Add a way to set different conditioning for the controlnet.
|
9 months ago |
blepping
|
a352c021ec
|
Allow custom samplers to request discard penultimate sigma
|
9 months ago |
comfyanonymous
|
c661a8b118
|
Don't use numpy for calculating sigmas.
|
9 months ago |
comfyanonymous
|
236bda2683
|
Make minimum tile size the size of the overlap.
|
10 months ago |
comfyanonymous
|
66e28ef45c
|
Don't use is_bf16_supported to check for fp16 support.
|
10 months ago |
comfyanonymous
|
24129d78e6
|
Speed up SDXL on 16xx series with fp16 weights and manual cast.
|
10 months ago |
comfyanonymous
|
4b0239066d
|
Always use fp16 for the text encoders.
|
10 months ago |
comfyanonymous
|
da7a8df0d2
|
Put VAE key name in model config.
|
10 months ago |
comfyanonymous
|
89507f8adf
|
Remove some unused imports.
|
10 months ago |
Dr.Lt.Data
|
05cd00695a
|
typo fix - calculate_sigmas_scheduler (#2619)
self.scheduler -> scheduler_name
Co-authored-by: Lt.Dr.Data <lt.dr.data@gmail.com>
|
10 months ago |
comfyanonymous
|
4871a36458
|
Cleanup some unused imports.
|
10 months ago |
comfyanonymous
|
78a70fda87
|
Remove useless import.
|
10 months ago |
comfyanonymous
|
d76a04b6ea
|
Add unfinished ImageOnlyCheckpointSave node to save a SVD checkpoint.
This node is unfinished, SVD checkpoints saved with this node will
work with ComfyUI but not with anything else.
|
10 months ago |
comfyanonymous
|
f9e55d8463
|
Only auto enable bf16 VAE on nvidia GPUs that actually support it.
|
10 months ago |
comfyanonymous
|
2395ae740a
|
Make unclip more deterministic.
Pass a seed argument note that this might make old unclip images different.
|
10 months ago |
comfyanonymous
|
53c8a99e6c
|
Make server storage the default.
Remove --server-storage argument.
|
10 months ago |
comfyanonymous
|
977eda19a6
|
Don't round noise mask.
|
10 months ago |
comfyanonymous
|
10f2609fdd
|
Add InpaintModelConditioning node.
This is an alternative to VAE Encode for inpaint that should work with
lower denoise.
This is a different take on #2501
|
10 months ago |
comfyanonymous
|
1a57423d30
|
Fix issue when using multiple t2i adapters with batched images.
|
10 months ago |
comfyanonymous
|
6a7bc35db8
|
Use basic attention implementation for small inputs on old pytorch.
|
10 months ago |