Damian at mba
f7cd98c238
tweak default cross-attention values
2022-10-23 20:38:28 +02:00
Damian at mba
8e7d744c60
fix bad math
2022-10-23 19:43:35 +02:00
Damian at mba
9210bf7d3a
also parse shape_freedom keyword
2022-10-23 19:40:00 +02:00
Damian at mba
8f35819ddf
add shape_freedom arg to .swap()
2022-10-23 19:38:31 +02:00
Damian at mba
04d93f0445
for k* samplers, estimate step_index from sigma
2022-10-23 16:26:50 +02:00
Damian at mba
7d677a63b8
cross attention control options
2022-10-23 14:58:25 +02:00
Damian at mba
8273c04575
wip implementing options in diffuse step
2022-10-22 12:15:34 +02:00
Damian at mba
ee7d4d712a
parsing CrossAttentionControlSubstitute options works
2022-10-22 11:27:56 +02:00
Damian at mba
cdb664f6e5
Merge branch 'development' into fix-prompts
2022-10-21 21:34:09 +02:00
psychedelicious
a127eeff20
Fixes gallery bugs & adds gallery context menu
2022-10-22 07:54:39 +13:00
Lincoln Stein
1ca517d73b
Merge branch 'fix-high-step-count' of https://github.com/holstvoogd/InvokeAI into holstvoogd-fix-high-step-count
2022-10-21 13:58:00 -04:00
Lincoln Stein
38b1dce7c3
Merge branch 'development' of github.com:invoke-ai/InvokeAI into development
2022-10-21 12:58:51 -04:00
Lincoln Stein
c9f9eed04e
resolve numerous small merge bugs
...
- This merges PR #882
Coauthor: ArDiouscuros
2022-10-21 12:57:15 -04:00
Lincoln Stein
fbea657eff
fix a number of bugs in textual inversion
...
- remove unsupported testtubelogger, use csvlogger instead
- fix logic for parsing --gpus option so that it won't crash if
trailing comma absent
- change trainer accelerator from unsupported 'ddp' to 'auto'
2022-10-21 16:35:35 +02:00
Lincoln Stein
55db9dba0a
Merge branch 'Improved-fetch-and-option-to-replay-commands-from-file' of https://github.com/ArDiouscuros/stable-diffusion into ArDiouscuros-Improved-fetch-and-option-to-replay-commands-from-file
...
- various small conflicts fixed
2022-10-21 10:12:35 -04:00
Damian at mba
64051d081c
cleanup
2022-10-21 15:07:11 +02:00
Lincoln Stein
ddb007af65
Merge branch 'development' into fix-high-step-count
2022-10-21 06:55:17 -04:00
Damian at mba
e574a1574f
txt2mask.py now tracking development again
2022-10-21 12:42:07 +02:00
Damian at mba
2bf9f1f0d8
rename StrcuturedConditioning to ExtraConditioningInfo
2022-10-21 12:18:40 +02:00
Damian at mba
8142b72bcd
Merge remote-tracking branch 'upstream/development' into fix-prompts
2022-10-21 11:59:44 +02:00
Damian at mba
dc2f30a34e
put back txt2mask import
2022-10-21 11:59:42 +02:00
Damian at mba
b385fdd7de
non-normalized blend
2022-10-21 04:34:53 +02:00
Damian at mba
d965540103
more blend fixes
2022-10-21 04:23:19 +02:00
Damian at mba
404d59b1b8
fix blend
2022-10-21 04:18:17 +02:00
Damian at mba
4c1267338b
bring in attention etc.
2022-10-21 03:54:13 +02:00
Damian at mba
2e0b1c4c8b
ok now we're cooking
2022-10-21 03:29:50 +02:00
Damian at mba
da75876639
better support for word.swap(otherWord) without parantheses or quotes
2022-10-21 00:08:28 +02:00
Jan Skurovec
d4d1014c9f
fix for 'model is not defined' when loading embedding
2022-10-20 17:31:46 -04:00
psychedelicious
213e12fe13
Filters existing images when adding new images; Fixes #1085 ; Builds fresh bundle
2022-10-20 16:53:48 -04:00
wfng92
3e0a7b6229
Correct color channels in upscale using array slicing
2022-10-20 16:52:07 -04:00
Damian at mba
da88097aba
fix prompt handling in conditioning.py
2022-10-20 21:41:32 +02:00
Damian at mba
3f13dd3ae8
prompt parsing is now much more robust
2022-10-20 21:05:36 +02:00
Damian at mba
79b4afeae7
parser working with basic escapes
2022-10-20 16:56:34 +02:00
Damian at mba
da223dfe81
wip re-writing parts of prompt parser
2022-10-20 15:56:46 +02:00
psychedelicious
899ba975a6
Improves logic to determine if clipseg weights should be downloaded
2022-10-20 06:56:50 -04:00
psychedelicious
bfa65560eb
Fixes torch.load() for MPS/CPU
2022-10-20 06:56:50 -04:00
psychedelicious
ed9307f469
Fix typo
2022-10-20 06:56:50 -04:00
Lincoln Stein
ff87239fb0
fix broken image in docs
2022-10-20 06:56:50 -04:00
Lincoln Stein
a357bf4f19
add !mask command to view output of clipseg
...
- The !mask command takes an image path, a text prompt, and
(optionally) a masking threshold. It creates a mask over the region
indicated by the prompt, and outputs several files that show which
regions will be masked by the chosen prompt and threshold.
- The mask images should not be passed directly to img2img because
they are designed for visualization only. Instead, use the
--text_mask option to pass the selected prompt and threshold.
- See docs/features/INPAINTING.md for details.
2022-10-20 06:56:50 -04:00
Lincoln Stein
63f274f6df
adjust environment & requirements files
2022-10-20 06:56:50 -04:00
Lincoln Stein
2ca4242f5f
fix clipseg loading problems
...
- The directory "models" in the main InvokeAI directory was conflicting
with loading "models.clipseg". To fix this issue, I have renamed the
models.clipseg to clipseg_models.clipseg, and applied this change to
the 'models-rename' branch of invoke-ai's fork of clipseg.
2022-10-20 06:56:50 -04:00
Damian at mba
c9d27634b4
bring in prompt parser from fix-prompts branch
...
attention is parsed but ignored, blends old syntax doesn't work,
conjunctions are parsed but ignored, the only part that's used
here is the new .blend() syntax and cross-attention control
using .swap()
2022-10-20 12:01:48 +02:00
noodlebox
027990928e
Fix typo in docs: s/Formally/Formerly
2022-10-20 02:44:16 -04:00
psychedelicious
87469a5fdd
Flips channels using array slicing instead of using OpenCV
2022-10-19 23:44:47 -04:00
psychedelicious
4101127011
Corrects color channels in face restoration; Fixes #1167
2022-10-19 23:32:57 -04:00
psychedelicious
f6191a4f12
Builds fresh bundle
2022-10-19 20:12:19 -04:00
psychedelicious
8c5d614c38
Increases max CFG Scale to 200
2022-10-19 20:12:19 -04:00
Damian at mba
42883545f9
add prompt language support for cross-attention .swap
2022-10-20 01:42:04 +02:00
Damian at mba
61357e4e6e
be less verbose when assembling prompt
2022-10-19 21:12:07 +02:00
Damian at mba
c6ae9f1176
remove unnecessary assertion
2022-10-19 21:12:07 +02:00
