lllyasviel
ef35383b4a
VAE patcher and more types of unet patches
2024-02-29 22:37:34 -08:00
lllyasviel
5166a723c2
bring back tiling
...
close #215
2024-02-25 21:29:35 -08:00
lllyasviel
6287c73d98
rework memory management for extras
...
now face post-processing uses gpu
close #312
2024-02-25 20:37:14 -08:00
lllyasviel
437c348926
Add build-in extension "NeverOOM"
...
see also discussions
2024-02-24 19:09:06 -08:00
lllyasviel
434ca2169f
Add optimization --cuda-stream
...
See also the readme for more details
2024-02-24 14:00:48 -08:00
lllyasviel
a844834193
safer stream initialization
2024-02-23 20:28:27 -08:00
lllyasviel
88f395091b
add two optimizations
...
--pin-shared-memory and --cuda-malloc
See also the updates in Readme for more details
2024-02-23 18:39:32 -08:00
lllyasviel
539bc5035d
safe cleanup to avoid potential problems
2024-02-22 01:28:38 -08:00
lllyasviel
638ee43bf1
Merge upstream PR 14855
2024-02-21 23:59:40 -08:00
lllyasviel
4ce60f2176
support SDXL-Lightning
...
added samplers: Euler SGMUniform, Euler A SGMUniform, DPM++ 2M SGMUniform, DPM++ 2M SDE SGMUniform
ByteDance official recommended one is Euler SGMUniform
Remember to use CFG=1.0
2024-02-21 15:42:38 -08:00
lllyasviel
43c9e3b5ce
begin to use double versioning
...
now you will see things like f0.0.14v1.8.0rc. It means forge 0.0.14 on webui 1.8.0rc
2024-02-17 08:32:15 -08:00
lllyasviel
07659efd88
fix import order
2024-02-17 08:03:17 -08:00
lllyasviel
d81e353d89
fix outpaint in inpaint_only+lama
...
the problem in inpaint_only+lama+"Resize and Fill" is fixed now.
SDXL+inpaint_only+lama+"Resize and Fill"+inpaint_fooocus_v26+"stop at 0.5 or 0.6" can be used as one SOTA outpaint method now
2024-02-14 02:45:23 -08:00
lllyasviel
fa8be06613
fix tile-colorfix problems
...
This will completely solve problems related to tile-colorfix and now tile colorfix/colorfix+sharp give same results to webui-cn
2024-02-14 02:19:04 -08:00
Chenlei Hu
9c5038c766
Add --forge-ref-a1111-home cmd arg to reference existing A1111 checkout ( #203 )
...
* Add --forge-ref-a1111-home cmd arg to reference existing A1111 checkout
* nit
2024-02-11 21:22:31 -05:00
lllyasviel
6e71d97478
Update forge_version.py
2024-02-10 03:25:27 -08:00
lllyasviel
ac94f38d9a
Update forge_version.py
2024-02-09 22:21:53 -08:00
lllyasviel
371687a1da
fix degradation
2024-02-09 20:58:38 -08:00
lllyasviel
e9459b6c4a
support sd1.5 model with v prediction #123
2024-02-09 20:51:18 -08:00
lllyasviel
d6f2e5bdd9
Disable Legacy Multidiffusion
...
Both Multidiffusion and Tiled VAE are integrated and this extension can be disabled
2024-02-09 19:07:50 -08:00
lllyasviel
54edd29725
integrate latent modifier
2024-02-09 14:38:57 -08:00
lllyasviel
8059533eaf
fix lora not loaded for text encoder #142
2024-02-09 13:50:50 -08:00
Chenlei Hu
ed60a99826
Fix import error ( #146 )
2024-02-08 23:02:33 -05:00
Chenlei Hu
388ca351f4
Revert "Fix ruff linter ( #137 )" ( #143 )
...
This reverts commit 6b3ad64388
.
2024-02-08 21:24:04 -05:00
Chenlei Hu
6b3ad64388
Fix ruff linter ( #137 )
...
* Fix ruff linter
* Remove unused imports
* Remove unused imports
2024-02-08 20:35:20 -05:00
lllyasviel
f06ba8e60b
Significantly reduce thread abuse for faster model moving
...
This will move all major gradio calls into the main thread rather than random gradio threads.
This ensures that all torch.module.to() are performed in main thread to completely possible avoid GPU fragments.
In my test now model moving is 0.7 ~ 1.2 seconds faster, which means all 6GB/8GB VRAM users will get 0.7 ~ 1.2 seconds faster per image on SDXL.
2024-02-08 10:13:59 -08:00
lllyasviel
291ec743b6
use better context manager to fix potential problems
2024-02-08 02:00:54 -08:00
lllyasviel
760f727eb9
use better context manager to fix potential problems
2024-02-08 01:51:18 -08:00
lllyasviel
a1670c536d
Allow controlnet_model_function_wrapper
...
for animatediff to manage controlnet batch slicing window
2024-02-07 20:06:23 -08:00
lllyasviel
e62631350a
Update forge_version.py
2024-02-06 17:55:14 -08:00
lllyasviel
402b7beb87
fix repeated model loading bug on 4090
2024-02-06 08:02:03 -08:00
lllyasviel
a578da074b
fix repeated model loading bug on 4090
2024-02-06 07:59:19 -08:00
lllyasviel
d76b830add
reduce prints
2024-02-06 07:56:15 -08:00
lllyasviel
7359740f36
Revert "safer device"
...
This reverts commit 1204d490d9
.
2024-02-06 05:27:07 -08:00
lllyasviel
1204d490d9
safer device
2024-02-06 05:01:58 -08:00
lllyasviel
9c31b0ddcb
try fix #56
2024-02-06 04:51:08 -08:00
lllyasviel
6aee7a2032
Update forge_version.py
2024-02-05 23:55:27 -08:00
lllyasviel
74ff4a9ba9
set CUDA_VISIBLE_DEVICES with args
2024-02-05 23:53:11 -08:00
lllyasviel
b8cd6d2e21
Update forge_version.py
2024-02-05 20:12:37 -08:00
lllyasviel
53057f33ed
Update forge_version.py
2024-02-05 15:30:45 -08:00
lllyasviel
88f6df4dcd
add_sampler_pre_cfg_function
2024-02-05 14:48:43 -08:00
lllyasviel
2c95b09a73
Euler A Turbo
2024-02-05 14:23:42 -08:00
lllyasviel
0ba407fd9c
new samplers
...
DDPM
DDPM Karras
DPM++ 2M Turbo
DPM++ 2M SDE Turbo
LCM Karras
2024-02-05 03:58:06 -08:00
lllyasviel
af017121d2
Update forge_version.py
2024-02-05 01:31:24 -08:00
lllyasviel
858bfc7a32
Update unet_patcher.py
2024-02-04 19:40:21 -08:00
lllyasviel
fdbe5b7aa7
memory_peak_estimation_modifier
2024-02-04 19:36:58 -08:00
lllyasviel
a3245e3eb1
Update unet_patcher.py
2024-02-03 23:05:06 -08:00
lllyasviel
8be7b347aa
Update unet_patcher.py
2024-02-03 23:03:21 -08:00
lllyasviel
dc082be7f9
Update unet_patcher.py
2024-02-03 22:54:42 -08:00
lllyasviel
c11490d560
add_alphas_cumprod_modifier
2024-02-03 22:33:18 -08:00