Commit Graph

104 Commits

Author SHA1 Message Date
Qing
0ddc1f105c pix2pix to instruct_pix2pix 2023-01-28 21:24:51 +08:00
Qing
05e82598de add pix2pix 2023-01-28 21:13:21 +08:00
Qing
205170e1e5 add DiffusionInpaintModel 2023-01-27 20:59:22 +08:00
Qing
148e97e8da optimize sd/paint_by_example modle VRAM usage 2023-01-18 18:34:10 +08:00
Qing
a7240eedb5 lots of update 2 2023-01-07 08:52:11 +08:00
Qing
a22536becc lots of updates 2023-01-05 22:07:39 +08:00
Qing
d2cec0d415 fix sd --no-half 2023-01-04 21:27:37 +08:00
Qing
59ee89bd34 add --no-half arg 2023-01-03 21:30:33 +08:00
Qing
221d44c037 fix test 2022-12-11 22:03:36 +08:00
Qing
203f2bc9c7 add paint by example 2022-12-11 09:14:28 +08:00
Qing
6e9d3d8442 update setup.py 2022-12-10 21:31:51 +08:00
Qing
4a739aaae7 fix CPUTextEncoder 2022-12-04 22:07:33 +08:00
Qing
6a0ffdc96e update diffusers to 0.9; add SD2 2022-12-04 20:44:59 +08:00
Qing
af87cca643 add forward_post_process function 2022-11-30 21:53:56 +08:00
Anders Haglund
a0240721e3 Add argument for enabling xFormers optimizations 2022-11-28 17:54:16 -08:00
Anders Haglund
8e408640a4 Match stable diffusion result's histogram to image's 2022-11-23 14:50:58 -08:00
Qing
c02d918368 0.27.1 2022-11-21 11:49:35 +08:00
Qing
4686c7cb78 fix manga mask 2022-11-18 23:06:41 +08:00
Qing
08c295a70d add Manga Model 2022-11-18 22:21:52 +08:00
Qing
d7c3149f67 add euler sampler 2022-11-15 21:11:45 +08:00
Qing
4f42dda2cf fix skimage import color
https://github.com/Sanster/lama-cleaner/issues/121
2022-11-13 12:45:29 +08:00
Qing
8c2904c9c8 add negative prompt 2022-11-13 12:45:29 +08:00
Anders Haglund
543473bac5 Add gaussian blur to mask before blending original image with result 2022-11-10 13:02:51 -08:00
Qing
b3cb11f804 remvoe sd amp.autocast 2022-11-04 15:33:59 +08:00
Qing
34d0327ff0 sd1.5: fix a bug when run --device=cpu on a host has gpu 2022-11-04 14:02:17 +08:00
Qing
e2e2f5f853 fix sd1.5: pass height/width to pipeline 2022-10-21 10:28:17 +08:00
Qing
6ccb6cd291 add sd1.5 2022-10-20 21:01:14 +08:00
Qing
3c87b050d9 update sd inpainting pipeline 2022-10-15 22:34:53 +08:00
Qing
f4fcece180 fix cv2 params 2022-10-09 21:32:13 +08:00
Qing
dba7b01da7 enable text_encoder cpu 2022-09-30 22:44:03 +08:00
Qing
4d6cec0ca2 add run-sd-local arg 2022-09-29 13:13:09 +08:00
Qing
ec7b2d8e2d add sd-cpu-textencoder args 2022-09-29 12:20:55 +08:00
Qing
0d57e552cf add sd-disable-nsfw arg 2022-09-29 09:42:19 +08:00
Qing
36b078123c fix sd pad_mod to 64
diffusers bug: https://github.com/huggingface/diffusers/pull/505
2022-09-27 22:37:42 +08:00
Qing
35cbbd653c Add OpenCV2 model 2022-09-25 21:27:12 +08:00
Qing
5bdc5c1526 fix model init 2022-09-22 22:45:24 +08:00
Qing
26d938839c make mask blur work 2022-09-22 21:50:41 +08:00
Qing
e1fb0030d1 sd make change sampler work 2022-09-22 12:38:32 +08:00
Qing
32854d40da wip 2022-09-19 22:37:42 +08:00
Qing
3cc9521221 update README 2022-09-04 21:23:58 +08:00
Qing
2119a5f905 FcF use unique resize strategy 2022-09-04 16:00:42 +08:00
Qing
9ccc577be2 fix resize mask bug 2022-09-02 11:13:15 +08:00
Qing
38c8837af7 Add FcF 2022-09-02 10:37:30 +08:00
Qing
6d2b24ed6b add MAT model 2022-08-24 21:30:20 +08:00
Qing
a5e840765e make crop mode use more context 2022-08-22 23:23:48 +08:00
Qing
cfcaf82a21 zits use structure_upsample_model 2022-07-19 21:47:21 +08:00
Qing
e11aed0b1e add github model url; fix cpu tensor 2022-07-18 20:35:55 +08:00
Qing
01c7f3b77d fix zits 2022-07-15 17:51:27 +08:00
Qing
a94f7e4ffe add ZITS 2022-07-14 16:49:03 +08:00
Qing
35b92ba9de ldm add plms sampler 2022-06-12 21:21:04 +08:00
Qing
e4a6c91f4a enable fp16 for ldm by default 2022-05-30 10:51:34 +08:00
Sanster
a219da27f7 add setup.py 2022-04-18 15:15:51 +08:00
Sanster
f7e1e073dc make model switch work with toast 2022-04-18 08:46:19 +08:00
Sanster
205286a414 big update 2022-04-16 21:29:18 +08:00