1bfec873fa
add an experimental option to apply loras to outputs rather than inputs
2023-01-25 11:29:46 +03:00
e3b53fd295
Add UI setting for upcasting attention to float32
...
Adds "Upcast cross attention layer to float32" option in Stable Diffusion settings. This allows for generating images using SD 2.1 models without --no-half or xFormers.
In order to make upcasting cross attention layer optimizations possible it is necessary to indent several sections of code in sd_hijack_optimizations.py so that a context manager can be used to disable autocast. Also, even though Stable Diffusion (and Diffusers) only upcast q and k, unfortunately my findings were that most of the cross attention layer optimizations could not function unless v is upcast also.
2023-01-25 01:13:04 -05:00
84d9ce30cb
Add option for float32 sampling with float16 UNet
...
This also handles type casting so that ROCm and MPS torch devices work correctly without --no-half. One cast is required for deepbooru in deepbooru_model.py, some explicit casting is required for img2img and inpainting. depth_model can't be converted to float16 or it won't work correctly on some systems (it's known to have issues on MPS) so in sd_models.py model.depth_model is removed for model.half().
2023-01-25 01:13:02 -05:00
48a15821de
remove the pip install stuff because it does not work as i hoped it would
2023-01-25 00:49:16 +03:00
bef1931895
add fastapi to requirements
2023-01-24 23:50:04 +03:00
93fad28a97
print progress when installing torch
...
add PIP_INSTALLER_LOCATION env var to install pip if it's not installed
remove accidental call to accelerate when venv is disabled
add another env var to skip venv - SKIP_VENV
2023-01-24 21:13:05 +03:00
5228ec8bda
remove fairscale requirement, add fake fairscale to make BLIP not complain about it mk2
2023-01-24 20:30:43 +03:00
28189985e6
remove fairscale requirement, add fake fairscale to make BLIP not complain about it
2023-01-24 20:24:27 +03:00
dac45299dd
make git commands not fail for extensions when you have spaces in webui directory
2023-01-24 20:22:19 +03:00
ec8774729e
swaps xyz axes internally if one costs more
2023-01-24 02:53:35 -05:00
e46bfa5a9e
handling sub grids and merging into one
2023-01-24 02:24:32 -05:00
9fc354e130
implements most of xyz grid script
2023-01-24 02:22:40 -05:00
d30ac02f28
renamed xy to xyz grid
...
this is mostly just so git can detect it properly
2023-01-24 02:21:32 -05:00
602a1864b0
also return the removed field to sdapi/v1/upscalers because someone might have relied on it existing
2023-01-24 10:09:30 +03:00
42a70d7477
repair sdapi/v1/upscalers returning bogus results
2023-01-24 10:05:45 +03:00
8b903322e6
Merge pull request #7140 from vladmandic/api-decode-image
...
Add exception handling to API image decode
2023-01-24 09:54:20 +03:00
848ef919b3
Merge pull request #7144 from catboxanon/swinir-interrupt
...
Make SwinIR upscaler interruptible and skippable
2023-01-24 09:51:53 +03:00
393e09c1c3
Merge pull request #7148 from acncagua/improvement_launch.py
...
Set Linux xformers 0.0.16RC425
2023-01-24 09:38:10 +03:00
f64af77adc
Fix different first gen with Approx NN previews
...
The loading of the model for approx nn live previews can change the internal state of PyTorch, resulting in a different image. This can be avoided by preloading the approx nn model in advance.
2023-01-23 22:49:20 -05:00
078e16e4d3
Set Linux xformers 0.0.16RC425
2023-01-24 12:21:07 +09:00
3c47b05036
Also make SwinIR skippable
2023-01-23 22:00:27 -05:00
f993525820
Make SwinIR interruptible
2023-01-23 21:50:59 -05:00
45e270dfc8
add image decod exception handling
2023-01-23 17:11:22 -05:00
82a28bfe35
Fix extra network thumbs label color
...
Added white color for labels.
2023-01-23 22:36:27 +01:00
5c1cb9263f
fix BLIP failing to import depending on configuration
2023-01-24 00:24:17 +03:00
7ba7f4ed6e
Merge pull request #7113 from vladmandic/interrogate
...
Add selector to interrogate categories
2023-01-24 00:09:14 +03:00
7b1c7ba87b
add support for apostrophe in extra network names
2023-01-23 23:11:34 +03:00
865af20d8a
suppress A matching Triton is not available message
...
you can all now stop worrying about it
2023-01-23 21:28:59 +03:00
04a561c11c
add option to skip interrogate categories
2023-01-23 12:29:23 -05:00
efa7287be0
Merge branch 'AUTOMATIC1111:master' into interrogate
2023-01-23 12:25:07 -05:00
c6f20f7262
make loras before 0.4.0 ALSO work
2023-01-23 18:52:55 +03:00
171a5b3bb9
Merge pull request #7032 from gmq/extra-network-styles
...
Extra network view style
2023-01-23 18:46:37 +03:00
756a2c3c0f
Merge pull request #7116 from vladmandic/api-image-format
...
API should use same image format as specified in WebUI settings
2023-01-23 18:37:48 +03:00
dbcb6fac77
feat(extra-networks): replace icon background with border
2023-01-23 12:14:01 -03:00
e407d1af89
add support for loras trained on kohya's scripts 0.4.0 (alphas)
2023-01-23 18:12:51 +03:00
6e1b296baf
api-image-format
2023-01-23 10:10:59 -05:00
e8c3d03f7d
a possible fix for broken image upscaling
2023-01-23 17:59:58 +03:00
7ff1ef77dd
add a message about new torch/xformers version and a way to upgrade by specifying a commandline flag
2023-01-23 17:17:31 +03:00
56f63cd498
Merge pull request #5939 from petalas/petalas/torch-upgrade
...
upgrading torch, torchvision, xformers (windows), to use cu117
2023-01-23 17:15:51 +03:00
925dd09c91
improve interrogate
2023-01-23 09:03:17 -05:00
59146621e2
better support for xformers flash attention on older versions of torch
2023-01-23 16:40:20 +03:00
3fa482076a
Merge remote-tracking branch 'takuma104/xformers-flash-attention'
2023-01-23 16:01:53 +03:00
194cbd065e
fix open directory button failing
2023-01-23 15:50:32 +03:00
97ba01a213
Merge pull request #7081 from EllangoK/xy-hires
...
Adds Hires Steps to X/Y Plot, and updates step calculation
2023-01-23 15:24:11 +03:00
663353098e
Merge pull request #7031 from EllangoK/master
...
Fixes various button overflowing UI and compact checkbox
2023-01-23 15:22:06 +03:00
74608300d1
Merge pull request #7093 from Shondoit/fix-dark-mode
...
Fix dark mode
2023-01-23 15:09:26 +03:00
41265a026d
third time's the charm
2023-01-23 14:50:20 +03:00
fabdae089e
add missing import to previous commit
2023-01-23 14:42:49 +03:00
669dbd9725
Fix dark mode
...
Fixes #7048
Co-Authored-By: J.J. Tolton <jjtolton@gmail.com>
2023-01-23 09:54:42 +01:00
b5230197a6
rework extras tab to use script system
2023-01-23 09:24:43 +03:00
AUTOMATIC