a5c5635968
detailer validate if model is compatible
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-10 07:53:45 -05:00
6f33ec3357
SDNQ use the model quant params instead of user settings on Lora
2025-11-10 00:12:38 +03:00
c536637908
add kanvas integration
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-09 15:41:20 -05:00
b045d72670
remove obsolete video scripts
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-09 09:54:17 -05:00
c6593ac180
fix detailer with lora
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-09 09:14:26 -05:00
3a3c984411
Merge pull request #4388 from vladmandic/kanvas
...
merge kanvas to dev
2025-11-09 07:57:54 -05:00
7cf5f8b45d
Revert Python >= 3.10 type syntax
2025-11-08 14:59:10 -08:00
19f9fc95fd
Improve extension url handling
2025-11-08 14:56:04 -08:00
0bb0bf4681
Fix url formatting in update module
2025-11-08 14:54:34 -08:00
113c582171
IPEX fix early init failure
2025-11-09 00:30:03 +03:00
f4ee9c7052
Add Flex attention
2025-11-09 00:14:38 +03:00
f491955991
Merge pull request #4383 from vladmandic/dev
...
refresh branch
2025-11-08 15:43:28 -05:00
69180202d3
kanvas integration
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-08 15:41:52 -05:00
2bbbb684cc
Rename CK Flash attention to just Flash attention
2025-11-08 23:24:40 +03:00
a93715e0da
Don't expose AMD Triton Flash Atten for non AMD
2025-11-08 23:20:55 +03:00
93797dff8e
ROCm enable Dynamic atten only for RDNA2 and older GPUs
2025-11-08 22:53:26 +03:00
e7e41d30cf
Merge pull request #4381 from vladmandic/dev
...
refresh branch
2025-11-08 11:59:06 -05:00
d292c85df0
strong-enforce auth check on all api endpoints
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-08 11:41:24 -05:00
ba270db6ad
separate settings for lora fuse
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-08 11:08:06 -05:00
56026c4e61
refactor attention handling
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-08 10:55:41 -05:00
155ee7f84c
fix sage-attention checks on sm86
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-08 08:47:21 -05:00
3bfe3a65b3
Merge pull request #4378 from awsr/hires-fix2
...
More Hires/denoising fixes
2025-11-08 08:15:15 -05:00
b13ec726b8
Merge pull request #4379 from awsr/img2img-scale
...
Make selected_scale_tab match default UI behavior in img2img
2025-11-08 08:14:26 -05:00
1c95960fb0
fix detailer with face-restorers
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-08 08:12:33 -05:00
6bcc509234
Remove unused value
2025-11-08 04:37:00 -08:00
d40a4f44a9
Rename variable to be less confusing
2025-11-08 04:36:18 -08:00
b26d999991
Make selected_scale_tab match default UI behavior
...
The fact that Gradio didn't even have a simple way to update States from within code until later versions makes my head hurt. Actually... a lot of things about Gradio make my head hurt.
2025-11-08 04:14:52 -08:00
165ea3bf97
Update ui_img2img.py
2025-11-08 04:00:20 -08:00
a4aef23f39
Merge pull request #4371 from vladmandic/dev
...
refresh branch
2025-11-07 17:56:14 -05:00
90e800a587
Fix Hires info text handling
2025-11-07 12:58:49 -08:00
7f4e151a13
cleanup
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-07 14:14:41 -05:00
f2835499b1
kanvas bindings
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-07 12:21:48 -05:00
0e8429dbd8
Cleanup
2025-11-07 18:49:29 +03:00
d287bed69f
Merge branch 'kanvas' into dev
2025-11-07 10:31:41 -05:00
93f28f07ac
Make SDNQ not depended on quantization_config.json and fix invalid quantization_config getting attached to the model on load
2025-11-07 18:11:21 +03:00
5dde890cb6
initial kanvas integration
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-06 22:05:19 -05:00
afa68b6dac
recursive inline wildcards
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-06 14:11:55 -05:00
a5cefc96b6
add nvidia-sana-video-2b
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-06 08:40:10 -05:00
5ffbca9377
cleanup and update changelog
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-05 12:57:30 -05:00
0c2da8e178
Merge pull request #4361 from CalamitousFelicitousness/patch-2
...
Sage Attention 2 + Triton workaround Qwen-Image
2025-11-05 12:44:34 -05:00
bdc477d252
Refactor GPU backend selection for sage attention
...
Removed hot path, now everything is defined at setup
Also, passing device to get_device_capability so that it works properly with multi-gpu setups.
2025-11-05 16:31:13 +00:00
4c791fb795
Remove model check logic for SA2 workaround
2025-11-05 10:53:07 +00:00
2c8cbf70a2
inline wildcard obey seed
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-04 20:50:13 -05:00
18676996d0
Sage Attention 2 + Triton workaround Qwen-Image
...
Workaround to prevent black images generated with Qwen-Image models when Sage Attention 2 is enabled with Triton as backend on devices with compute capability 8.0 and 8.6.
Simply switches back to Cuda backend for these models only.
Proof of concept, feel free to close if this is not appropriate.
2025-11-04 23:31:14 +00:00
65dfc9b4d0
fix ui_disabled
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-04 12:06:32 -05:00
db3f5ff01b
detailer expert mode enhance parsing
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-04 10:41:23 -05:00
e265f1aace
networks info details include preview image metadata
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-04 10:17:52 -05:00
3527fac3a0
keep exif on thumbnail generation
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-04 10:03:22 -05:00
9927b3e62c
controlnet allow processor to keep aspect-ratio for override images based on i2i or t2i resolution
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-04 09:37:18 -05:00
4a48695918
fix postprocess
...
Signed-off-by: Vladimir Mandic <mandic00@live.com >
2025-11-04 08:23:45 -05:00
Vladimir Mandic
Disty0
awsr
CalamitousFelicitousness