 AUTOMATIC1111
|
806ea639e6
Merge pull request #11066 from aljungberg/patch-1
|
il y a 2 ans |
 Alexander Ljungberg
|
d9cc0910c8
Fix upcast attention dtype error.
|
il y a 2 ans |
|
AUTOMATIC1111
|
56bf522913
Merge pull request #10990 from vkage/sd_hijack_optimizations_bugfix
|
il y a 2 ans |
|
AUTOMATIC
|
2e23c9c568
fix the broken line for #10990
|
il y a 2 ans |
 Vivek K. Vasishtha
|
b1a72bc7e2
torch.cuda.is_available() check for SdOptimizationXformers
|
il y a 2 ans |
|
AUTOMATIC
|
3ee1238630
revert default cross attention optimization to Doggettx
|
il y a 2 ans |
|
AUTOMATIC
|
36888092af
revert default cross attention optimization to Doggettx
|
il y a 2 ans |
|
AUTOMATIC
|
05933840f0
rename print_error to report, use it with together with package name
|
il y a 2 ans |
 Aarni Koskela
|
00dfe27f59
Add & use modules.errors.print_error where currently printing exception info by hand
|
il y a 2 ans |
|
Aarni Koskela
|
df004be2fc
Add a couple `from __future__ import annotations`es for Py3.9 compat
|
il y a 2 ans |
|
AUTOMATIC1111
|
1e5afd4fa9
Apply suggestions from code review
|
il y a 2 ans |
|
AUTOMATIC
|
8a3d232839
fix linter issues
|
il y a 2 ans |
|
AUTOMATIC
|
2582a0fd3b
make it possible for scripts to add cross attention optimizations
|
il y a 2 ans |
|
Aarni Koskela
|
49a55b410b
Autofix Ruff W (not W605) (mostly whitespace)
|
il y a 2 ans |
|
AUTOMATIC
|
028d3f6425
ruff auto fixes
|
il y a 2 ans |
|
AUTOMATIC
|
762265eab5
autofixes from ruff
|
il y a 2 ans |
 brkirch
|
7aab389d6f
Fix for Unet NaNs
|
il y a 2 ans |
 FNSpd
|
280ed8f00f
Update sd_hijack_optimizations.py
|
il y a 2 ans |
|
FNSpd
|
c84c9df737
Update sd_hijack_optimizations.py
|
il y a 2 ans |
 Pam
|
8d7fa2f67c
sdp_attnblock_forward hijack
|
il y a 2 ans |
|
Pam
|
37acba2633
argument to disable memory efficient for sdp
|
il y a 2 ans |
|
Pam
|
fec0a89511
scaled dot product attention
|
il y a 2 ans |
|
brkirch
|
e3b53fd295
Add UI setting for upcasting attention to float32
|
il y a 2 ans |
|
AUTOMATIC
|
59146621e2
better support for xformers flash attention on older versions of torch
|
il y a 2 ans |
 Takuma Mori
|
3262e825cc
add --xformers-flash-attention option & impl
|
il y a 2 ans |
|
AUTOMATIC
|
40ff6db532
extra networks UI
|
il y a 2 ans |
|
brkirch
|
c18add68ef
Added license
|
il y a 2 ans |
|
brkirch
|
b95a4c0ce5
Change sub-quad chunk threshold to use percentage
|
il y a 2 ans |
|
brkirch
|
d782a95967
Add Birch-san's sub-quadratic attention implementation
|
il y a 2 ans |
|
brkirch
|
35b1775b32
Use other MPS optimization for large q.shape[0] * q.shape[1]
|
il y a 2 ans |
