3eaa908086
Fixes
2024-03-06 10:58:27 -06:00
b58a14d231
Cleanup
2024-03-06 10:21:48 -06:00
f51d2874ab
Update clip_interrogator.py
2024-03-05 16:53:43 -06:00
6ab4d297b7
Fix subject swap, cleanup clip interrogator
2024-03-05 16:50:32 -06:00
36cbadf39a
Fix clip output
2024-03-05 11:08:38 -06:00
8602e93209
Updates
2024-03-05 10:49:58 -06:00
cc418436d0
Use other elements to influence prompts from LLAVA2
2024-02-24 16:38:58 -06:00
34dac6859f
Remove OWLV1
2024-02-24 15:07:17 -06:00
0a09c50665
Update llava_interrogator.py
2024-02-24 14:10:22 -06:00
f7e982c9f6
Cleanup, fix caption output.
2024-02-24 14:08:38 -06:00
81dcb53ccb
Fix Mplug owl2
2024-02-24 14:07:35 -06:00
2f728a0dc8
WIP
2023-12-30 22:06:26 -06:00
4705d6a870
FILEMANAGER
2023-12-24 18:43:29 -06:00
fad26f09a7
Work sync
2023-12-22 12:44:04 -06:00
ea0a879558
Cleanup
2023-12-19 16:26:50 -06:00
23ea333fa4
Add BLIP prompt (mostly for testing)
2023-12-19 16:11:42 -06:00
005c5355bb
V2?
2023-12-19 15:57:37 -06:00
1ceda17086
Worksave.
2023-02-02 15:29:39 -06:00
330f47f859
Better Caption split/join
...
Split caps to tags by commas, append/prepend/replace with a list..
Use proper "max caption length" by counting tags, versus characters like a dumb person.
Add "tags to ignore" field.
Adjust some default params.
2023-01-03 11:39:09 -06:00
cffe3c5bb4
Unload SD model on process start.
2023-01-03 10:47:54 -06:00
0342a88068
Merge pull request #12 from Zuxier/main
...
fix for filename issues and max flavors
2023-01-02 20:59:44 -06:00
1def06d8a4
Add min/max CLIP length, don't adjust image size unless we tell it to
2023-01-02 20:59:28 -06:00
e0bbc36a75
fix for max flavors
2022-12-20 02:07:08 +01:00
551c9bea0d
fix filename issues
2022-12-20 01:01:36 +01:00
5fa74185f5
Code cleanup, fixes
2022-12-15 13:48:36 -06:00
e4c638c7b8
Fix cross-extension import, update reallysafe.
...
And some code cleanup.
2022-12-15 12:03:50 -06:00
d1fe158903
Update smartprocess.py
2022-12-10 19:17:26 -06:00
1416a2a07a
Update smartprocess.py
2022-12-10 18:50:20 -06:00
cdfc93b8a4
Dur
2022-12-10 18:00:02 -06:00
bead6660ef
Don't append ratings
2022-12-10 17:58:15 -06:00
798549e164
Duplicating extension...
2022-12-10 17:56:54 -06:00
da90277e5e
Moar fun
2022-12-10 17:54:02 -06:00
58bcff7f78
Super Update
...
Add WD14 tagger.
Add CLIP v2.1 interrogator.
Allow filtering wd14, booru tags by score.
Add don't rename option.
Update ReallySafe
Bump BLIP version?
Remove split image options.
Completely overhaul smartprocess...
2022-12-10 17:33:34 -06:00
93380435c0
Updates, Improvements
...
Update yolov5 model version.
Select largest target if multiple targets found when smart cropping.
Pad smart cropped images before cropping.
Add pad only mode.
Set our input dimensions to one square value.
Don't use hashes for new filenames, it makes diffing batches impossible.
2022-12-07 15:57:45 -06:00
c840807b84
Cleanup UI, add "pad image" option.
2022-12-07 13:03:54 -06:00
a5dcbc42b6
Update install.py
2022-12-06 08:22:34 -06:00
c01aa302b7
Merge pull request #1 from ClashSAN/main
...
adds picture to README.md
2022-11-29 12:57:25 -06:00
a2afe6c7b9
Merge pull request #5 from knoopx/fix-caption-length
...
fix max caption length bug
2022-11-29 12:57:13 -06:00
b5b71f99e4
Merge pull request #4 from knoopx/fix-deepbooru
...
fix deepbooru tagging
2022-11-29 12:57:00 -06:00
483e461dec
fix max caption length bug
2022-11-28 16:15:37 +01:00
a3dba522ec
fix deepbooru tagging
2022-11-28 15:53:01 +01:00
2d48ebe170
Build caption after cropping
2022-11-12 10:19:05 -06:00
0344636c36
adds picture to README.md
2022-11-12 08:33:52 +00:00
9f1f88077f
Update README.md
2022-11-11 15:31:32 -06:00
646ac975e0
Update README.md
2022-11-11 15:18:49 -06:00
2b79465d99
Initial Commit
2022-11-11 14:35:46 -06:00
6f1a6b23de
Initial commit
2022-11-11 14:34:40 -06:00
d8ahazard
Zuxier
Victor Martinez
ClashSAN