I'm now at 41.2 it/s, up from 39.2

[aifartist]

In addition to #6954, which tripled my performance, I just found that adding:
torch.backends.cudnn.benchmark = True
to the main py file in A1111 got me up to 41.2 it/s.
I had seen this option before but hadn't noticed any speedup, but I was less savvy at that time.

A question for our NN experts here: Do 'input sizes' change 'often' in any of the various and many A1111 functional pieces?
If that does occur there can be negative impacts from using this option depending on how often the sizes change.

FYI, Today I helped another windows user go from 10 it/s to about 30 it/s. When I upgraded the cuDNN dll's to v8.7 it only went to 16 it/s. But then I discovered that bat file was setting --precision=full and when that was removed it then went to 30.

Is precision=full needed for anything other than training or to prevent black images when using sd v2?

[aifartist]

Actually closer to 41.4

100%|███████████| 20/20 [00:00<00:00, 41.53it/s]
100%|███████████| 20/20 [00:00<00:00, 41.47it/s]
100%|███████████| 20/20 [00:00<00:00, 41.29it/s]
100%|███████████| 20/20 [00:00<00:00, 41.42it/s]
100%|███████████| 20/20 [00:00<00:00, 41.36it/s]
100%|███████████| 20/20 [00:00<00:00, 41.43it/s]
100%|███████████| 20/20 [00:00<00:00, 41.40it/s]

[cibernicola]

I have many questions about this topic.
And I have them because I haven't just had any effective boosts in terms of performance.
I have a 3090 (compute_86) and a Ryzen 53600XT + 48GB RAM, running Windows 10.

Here goes the questions:

What model are you using for testing?
What Sampling method?
What image size?
What prompt?

In my case testing:
sd-1.5-pruned
Euler a
512*512
test

| 1/20 [00:00<00:04, 4.39it/s] Total progress: 10%|██████▋ | 2/20 [00:00<00:01, 13.33it/s]← 15%|████████████▍ | 3/20 [00:00<00:02, 5.92it/s] Total progress: 20%|█████████████▍ | 4/20 [00:00<00:02, 7.71it/s]← 20%|████████████████▌ | 4/20 [00:00<00:02, 5.70it/s] Total progress: 25%|████████████████▊ | 5/20 [00:00<00:02, 6.62it/s]← 25%|████████████████████▊ | 5/20 [00:00<00:02, 5.42it/s] Total progress: 30%|████████████████████ | 6/20 [00:00<00:02, 6.45it/s]← 30%|████████████████████████▉ | 6/20 [00:01<00:02, 5.63it/s] Total progress: 35%|███████████████████████▍ | 7/20 [00:01<00:02, 6.26it/s]← 35%|█████████████████████████████ | 7/20 [00:01<00:02, 5.69it/s] Total progress: 40%|██████████████████████████▊ | 8/20 [00:01<00:01, 6.20it/s]← 40%|█████████████████████████████████▏ | 8/20 [00:01<00:02, 5.81it/s] Total progress: 45%|██████████████████████████████▏ | 9/20 [00:01<00:01, 6.25it/s]← 45%|█████████████████████████████████████▎ | 9/20 [00:01<00:01, 5.99it/s] Total progress: 50%|█████████████████████████████████ | 10/20 [00:01<00:01, 6.20it/s]← 50%|█████████████████████████████████████████ | 10/20 [00:01<00:01, 6.01it/s] Total progress: 55%|████████████████████████████████████▎ | 11/20 [00:01<00:01, 5.05it/s]← 55%|█████████████████████████████████████████████ | 11/20 [00:02<00:01, 4.94it/s] Total progress: 60%|███████████████████████████████████████▌ | 12/20 [00:01<00:01, 5.17it/s]← 60%|█████████████████████████████████████████████████▏ | 12/20 [00:02<00:01, 5.09it/s] Total progress: 65%|██████████████████████████████████████████▉ | 13/20 [00:02<00:01, 5.22it/s]← 65%|█████████████████████████████████████████████████████▎ | 13/20 [00:02<00:01, 5.16it/s] Total progress: 70%|██████████████████████████████████████████████▏ | 14/20 [00:02<00:01, 5.19it/s]← 70%|█████████████████████████████████████████████████████████▍ | 14/20 [00:02<00:01, 5.16it/s] Total progress: 75%|█████████████████████████████████████████████████▌ | 15/20 [00:02<00:00, 5.36it/s]← 75%|█████████████████████████████████████████████████████████████▌ | 15/20 [00:02<00:00, 5.32it/s] Total progress: 80%|████████████████████████████████████████████████████▊ | 16/20 [00:02<00:00, 5.48it/s]← 80%|█████████████████████████████████████████████████████████████████▌ | 16/20 [00:02<00:00, 5.45it/s] Total progress: 85%|████████████████████████████████████████████████████████ | 17/20 [00:02<00:00, 5.57it/s]← 85%|█████████████████████████████████████████████████████████████████████▋ | 17/20 [00:03<00:00, 5.56it/s] Total progress: 90%|███████████████████████████████████████████████████████████▍ | 18/20 [00:03<00:00, 5.63it/s]← 90%|█████████████████████████████████████████████████████████████████████████▊ | 18/20 [00:03<00:00, 5.28it/s] Total progress: 95%|██████████████████████████████████████████████████████████████▋ | 19/20 [00:03<00:00, 4.57it/s]← 95%|█████████████████████████████████████████████████████████████████████████████▉ | 19/20 [00:03<00:00, 4.64it/s] Total progress: 100%|██████████████████████████████████████████████████████████████████| 20/20 [00:03<00:00, 4.84it/s]←100%|██████████████████████████████████████████████████████████████████████████████████| 20/20 [00:03<00:00, 5.30it/s] Total progress: 100%|██████████████████████████████████████████████████████████████████| 20/20 [00:03<00:00, 5.36it/s]

On the one hand, there is some volatility in terms of speed, on the other hand, the way of representation is different?

I have gone through the various steps you mention and have not gone any further.

Any suggestions?

[ghost]

Lately I have seen so many people with performance problems of the 30xx or 40xx cards. They are slower than my card (2080).

I read a comment somewhere mentioning that an older xformers version might solve it, but its a rumour I can't test it. Something is going on

[cibernicola]

Maybe part of my problem comes from OS settings... My previous GPU was a 2060, its the same OS, have no idea how to check it.

[VictorZakharov]

Is there a solution for Windows? I tried many things, most of them have no effect, while some make it worse.

[aifartist]

Even better, with --opt-channelslast I get 42.1 it/s

[bbecausereasonss]

Sorry but which .py file are you adding the original change to?

[aifartist]

Sorry but which .py file are you adding the original change to?

What "original change"? I'm sure I've mentioned several things.

[VictorZakharov]

Has no effect on my setup.

[bbecausereasonss]

I meant which 'main py file'?

[tuangd]

Where did you add that torch.backends.cudnn.benchmark = True to?

[hithereai]

this made my 4090 slower

[VictorZakharov]

Has no effect on my setup. Window 10, 4070 Ti.

[mmhan2008]

@aifartist Can you tell me which py file to modify torch.compile(mode='max-autotune') and batchsize=4 in? thank you very much

[aifartist]

modules/sd_hijack.py. However, don't expect it to actually work. When PYT 2.0 came out something got broken because I use to have been successful using it. And I do not have time to debug that now.

Vlado's fork has the modification to sd_hijack.py but, as I just said, unless you are a hard core hacker that can try a bunch of old versions of pytorch to find one where it barely worked you WILL NOT be able to use this.

[mmhan2008]

thank you very much，I'll do some research

[CHollman82]

Thank you, this increased my speed as well with a 4080

[aifartist]

@CHollman82 This is the numbers I get with torch.compile() but I have to make fixes to the pytorch code to get it to work.
Also, it takes about 20 minutes to compile a model the first time.
100%|████████████████| 20/20 [00:00<00:00, 45.03it/s]
100%|████████████████| 20/20 [00:00<00:00, 45.06it/s]
100%|████████████████| 20/20 [00:00<00:00, 45.53it/s]
100%|████████████████| 20/20 [00:00<00:00, 45.13it/s]
100%|████████████████| 20/20 [00:00<00:00, 45.42it/s]
100%|████████████████| 20/20 [00:00<00:00, 45.20it/s]
100%|████████████████| 20/20 [00:00<00:00, 45.48it/s]
100%|████████████████| 20/20 [00:00<00:00, 45.19it/s]
100%|████████████████| 20/20 [00:00<00:00, 45.28it/s]

[ghost]

Cant wait to get a 4090 too. I probably wont reach these top scores because my CPU is a bit older (Ryzen 3600X)

If TensorRT and distilled diffusion models will work together... this is going to get crazy. Real time prompt rendering and living in an animation wonderland

[kou201]

How did you get this crazy speed? My 4090 can only get a maximum speed of 23it/s under the full default setting in webui. And I have replaced the latest cudnn file and used the latest xformers.

[aifartist]

If you have a slow CPU and a uber fast GPU you can easily see only 23 it/s. I assume you aren't using a very long prompt when just benchmarking.

[FurkanGozukara]

24.) Automatic1111 Web UI - PC - Free
For downgrade to older version if you don't like Torch 2 : first delete venv, let it reinstall, then activate venv and run this command pip install -r "path_of_SD_Extension\requirements.txt"
How To Install New DREAMBOOTH & Torch 2 On Automatic1111 Web UI PC For Epic Performance Gains Guide

[VictorZakharov]

Pls add more info, what exactly you did, which OS, video card, what were the gains etc. AFAIK, torch 2.0 does not improve performance.

[FurkanGozukara]

Pls add more info, what exactly you did, which OS, video card, what were the gains etc. AFAIK, torch 2.0 does not improve performance.

windows 10
RTX 3060

i felt increase but maybe not much . i think new cards getting more

[VictorZakharov]

I have 4070 Ti. The only thing that increased performance by +40% for me was replacing CUDA dlls, from 5it/sec to 7, not a lot.
Nothing else did anything, some steps made it worse.

[quasiblob]

Which exact versions of CUDA files you used? I was looking into this thing and was also looking to replace CUDA files and see if it made any difference.

[Evilander]

I get 1.51 it/s with my 4070. Ha. Even with your addition to main.py and the command args.

[VictorZakharov]

Try SD.Next with Stability Matrix. I am getting 25 it/sec on 4070 Ti. Installs in a few clicks. No parameter tweaking.

[Evilander]

Try SD.Next with Stability Matrix. I am getting 25 it/sec on 4070 Ti. Installs in a few clicks. No parameter tweaking.

Cool - looking into that now. Just spent two days (after working fine) fighting with Automatic1111 - cuda memory errors instantly. Torch not compiled with cuda/gpu. Nonetype errors. Mat1 and Mat2 errors. Hahaha!

[chaseforks]

Is anyone running SD.NEXT with a 4070 Ti 12GB and cu121? I checked the published benchmarks and couldn't find any that matched my setup. My benchmark below

timestamp 2024-01-14 04:33:03.298575
performance error / error / error
version app:SD.next updated:2023-12-30 hash:ab7b78cc url:https://github.com/vladmandic/automatic/tree/master
system arch:x86_64 cpu:x86_64 system:Linux release:6.6.6-76060606-generic python:3.10.12
libraries torch:2.1.2+cu121 Full half xformers:0.0.24.dev720 diffusers:0.25.0 transformers:4.36.2
gpu device:NVIDIA GeForce RTX 4070 Ti (1) (sm_90) (8, 9) cuda:12.1 cudnn:8902 driver:545.29.06 12GB
pipeline diffusers xFormers none
model Diffusers/diffusers/stable-diffusion-xl-1.0-inpainting-0.1 [models/Dif]

I feel like I should be getting better speeds but can't seem to find the right settings. What I have now at least is stable. Trying some changes leads to crash.

[ruichang555]

I only got 12-14s/it Running SD A1111 (v1.10.1) on RTX A6000. Loaded XL checkpoint (realvisxlv40_v20Bakedvae), ControlNet XL ip-(adaptor-plus-sdxl-vit-h) to process 1080 x 1920 images. Using --xformer and --medvram, live preview disabled. Batch sizee=4, Batch count=1 (but I do enabled "Batch Loopback"-since my batch count=1, it shouldn't do anything?) Not using Loopback in Script. GPU temp is 81 Celsius.

Thanks!
Michael

++++++++Run Report+++++++
Python 3.10.6 (tags/v3.10.6:9c7b4bd, Aug 1 2022, 21:53:49) [MSC v.1932 64 bit (AMD64)]
Version: v1.10.1
Commit hash: 82a973c
Installing requirements
current transparent-background 1.3.2
Installing sd-webui-controlnet requirement: changing opencv-python version from 4.7.0.72 to 4.8.0
CUDA 12.1
Requirement already satisfied: insightface==0.7.3 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from -r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (0.7.3)
Requirement already satisfied: onnx==1.14.0 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from -r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 2)) (1.14.0)
Requirement already satisfied: onnxruntime==1.15.0 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from -r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 3)) (1.15.0)
Collecting opencv-python==4.7.0.72 (from -r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 4))
Using cached opencv_python-4.7.0.72-cp37-abi3-win_amd64.whl.metadata (18 kB)
Requirement already satisfied: ifnude in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from -r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 5)) (0.0.3)
Requirement already satisfied: cython in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from -r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 6)) (3.0.11)
Requirement already satisfied: numpy in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (1.26.2)
Requirement already satisfied: tqdm in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (4.66.5)
Requirement already satisfied: requests in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (2.32.3)
Requirement already satisfied: matplotlib in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (3.9.2)
Requirement already satisfied: Pillow in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (9.5.0)
Requirement already satisfied: scipy in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (1.14.1)
Requirement already satisfied: scikit-learn in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (1.5.1)
Requirement already satisfied: scikit-image in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (0.21.0)
Requirement already satisfied: easydict in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (1.13)
Requirement already satisfied: albumentations in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (1.4.3)
Requirement already satisfied: prettytable in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (3.11.0)
Collecting protobuf>=3.20.2 (from onnx==1.14.0->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 2))
Using cached protobuf-5.28.0-cp310-abi3-win_amd64.whl.metadata (592 bytes)
Requirement already satisfied: typing-extensions>=3.6.2.1 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from onnx==1.14.0->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 2)) (4.12.2)
Requirement already satisfied: coloredlogs in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from onnxruntime==1.15.0->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 3)) (15.0.1)
Requirement already satisfied: flatbuffers in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from onnxruntime==1.15.0->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 3)) (24.3.25)
Requirement already satisfied: packaging in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from onnxruntime==1.15.0->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 3)) (23.2)
Requirement already satisfied: sympy in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from onnxruntime==1.15.0->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 3)) (1.13.2)
Requirement already satisfied: opencv-python-headless>=4.5.1.48 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from ifnude->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 5)) (4.10.0.84)
Requirement already satisfied: PyYAML in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from albumentations->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (6.0.2)
Requirement already satisfied: networkx>=2.8 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from scikit-image->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (3.3)
Requirement already satisfied: imageio>=2.27 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from scikit-image->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (2.35.1)
Requirement already satisfied: tifffile>=2022.8.12 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from scikit-image->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (2024.8.24)
Requirement already satisfied: PyWavelets>=1.1.1 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from scikit-image->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (1.7.0)
Requirement already satisfied: lazy_loader>=0.2 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from scikit-image->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (0.4)
Requirement already satisfied: joblib>=1.2.0 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from scikit-learn->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (1.4.2)
Requirement already satisfied: threadpoolctl>=3.1.0 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from scikit-learn->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (3.5.0)
Requirement already satisfied: humanfriendly>=9.1 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from coloredlogs->onnxruntime==1.15.0->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 3)) (10.0)
Requirement already satisfied: contourpy>=1.0.1 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from matplotlib->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (1.2.1)
Requirement already satisfied: cycler>=0.10 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from matplotlib->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (0.12.1)
Requirement already satisfied: fonttools>=4.22.0 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from matplotlib->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (4.53.1)
Requirement already satisfied: kiwisolver>=1.3.1 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from matplotlib->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (1.4.5)
Requirement already satisfied: pyparsing>=2.3.1 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from matplotlib->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (3.1.4)
Requirement already satisfied: python-dateutil>=2.7 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from matplotlib->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (2.9.0.post0)
Requirement already satisfied: wcwidth in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from prettytable->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (0.2.13)
Requirement already satisfied: charset-normalizer<4,>=2 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from requests->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (3.3.2)
Requirement already satisfied: idna<4,>=2.5 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from requests->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (3.8)
Requirement already satisfied: urllib3<3,>=1.21.1 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from requests->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (2.2.2)
Requirement already satisfied: certifi>=2017.4.17 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from requests->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (2024.7.4)
Requirement already satisfied: mpmath<1.4,>=1.1.0 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from sympy->onnxruntime==1.15.0->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 3)) (1.3.0)
Requirement already satisfied: colorama in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from tqdm->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (0.4.6)
Requirement already satisfied: pyreadline3 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from humanfriendly>=9.1->coloredlogs->onnxruntime==1.15.0->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 3)) (3.4.1)
Requirement already satisfied: six>=1.5 in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (from python-dateutil>=2.7->matplotlib->insightface==0.7.3->-r C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-roop\requirements.txt (line 1)) (1.16.0)
Using cached opencv_python-4.7.0.72-cp37-abi3-win_amd64.whl (38.2 MB)
Using cached protobuf-5.28.0-cp310-abi3-win_amd64.whl (431 kB)
Installing collected packages: protobuf, opencv-python
Attempting uninstall: protobuf
Found existing installation: protobuf 3.20.0
Uninstalling protobuf-3.20.0:
Successfully uninstalled protobuf-3.20.0
Attempting uninstall: opencv-python
Found existing installation: opencv-python 4.10.0.84
Uninstalling opencv-python-4.10.0.84:
Successfully uninstalled opencv-python-4.10.0.84
Successfully installed opencv-python-4.7.0.72 protobuf-5.28.0
Launching Web UI with arguments: --xformers --medvram
[-] ADetailer initialized. version: 24.9.0, num models: 10
[AddNet] Updating model hashes...
0it [00:00, ?it/s]
[AddNet] Updating model hashes...
0it [00:00, ?it/s]
ControlNet preprocessor location: C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-controlnet\annotator\downloads
2024-09-02 22:27:50,373 - ControlNet - INFO - ControlNet v1.1.455
Requirement already satisfied: toolz in c:\users\user\desktop\sd.webui\system\python\lib\site-packages (0.12.1)
sd-webui-prompt-all-in-one background API service started successfully.
22:27:52 - ReActor - STATUS - Running v0.7.1-a2 on Device: CUDA
2024-09-02 22:27:53,141 - roop - INFO - roop v0.0.2
2024-09-02 22:27:53,166 - roop - INFO - roop v0.0.2
[ControlNet-Travel] extension Mikubill/sd-webui-controlnet found, ControlNet-Travel loaded :)
Loading weights [74dda471cc] from C:\Users\User\Desktop\sd.webui\webui\models\Stable-diffusion\realvisxlV40_v20Bakedvae.safetensors
Creating model from config: C:\Users\User\Desktop\sd.webui\webui\repositories\generative-models\configs\inference\sd_xl_base.yaml
Applying attention optimization: xformers... done.
Model loaded in 17.1s (load weights from disk: 0.7s, create model: 2.4s, apply weights to model: 6.6s, apply half(): 0.1s, calculate empty prompt: 7.2s).
2024-09-02 22:28:11,527 - ControlNet - INFO - ControlNet UI callback registered.
Running on local URL: http://127.0.0.1:7860/

To create a public link, set share=True in launch().
Startup time: 65.6s (prepare environment: 27.3s, import torch: 3.9s, import gradio: 1.2s, setup paths: 1.2s, initialize shared: 2.0s, other imports: 0.8s, load scripts: 6.0s, scripts before_ui_callback: 0.5s, create ui: 20.1s, gradio launch: 1.6s, app_started_callback: 0.8s).
2024-09-02 22:32:15,421 - ControlNet - INFO - Preview Resolution = 1920
2024-09-02 22:32:15,421 - ControlNet - INFO - ip-adapter-auto => ip-adapter_clip_h
Starting job task(lyuq9l9hufqa290)
INFO:modules.shared_state:Starting job task(lyuq9l9hufqa290)

Inpaint batch is enabled. 3 masks found.
Will process 3 images, creating 4 new images for each.
2024-09-02 22:33:19,097 - ControlNet - INFO - unit_separate = False, style_align = True
2024-09-02 22:33:19,371 - ControlNet - INFO - Loading model: ip-adapter-plus_sdxl_vit-h [bc449f62]
2024-09-02 22:33:19,458 - ControlNet - INFO - Loaded state_dict from [C:\Users\User\Desktop\sd.webui\webui\extensions\sd-webui-controlnet\models\ip-adapter-plus_sdxl_vit-h.safetensors]
2024-09-02 22:33:21,331 - ControlNet - INFO - ControlNet model ip-adapter-plus_sdxl_vit-h bc449f62 loaded.
2024-09-02 22:33:21,336 - ControlNet - INFO - Using preprocessor: ip-adapter-auto
2024-09-02 22:33:21,336 - ControlNet - INFO - preprocessor resolution = 1920
2024-09-02 22:33:21,337 - ControlNet - INFO - ip-adapter-auto => ip-adapter_clip_h
2024-09-02 22:33:21,399 - ControlNet - INFO - IP-Adapter using advanced weighting [0.0, 0.0, 0.0, 0.5, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
2024-09-02 22:33:24,852 - ControlNet - INFO - ControlNet Hooked - Time = 5.762352466583252
100%|██████████████████████████████████████████████████████████████████████████████████| 40/40 [09:34<00:00, 14.36s/it]
2024-09-02 22:43:33,834 - ControlNet - INFO - unit_separate = False, style_align = True 40/120 [09:17<19:05, 14.32s/it]
2024-09-02 22:43:33,835 - ControlNet - INFO - Loading model from cache: ip-adapter-plus_sdxl_vit-h [bc449f62]
2024-09-02 22:43:33,838 - ControlNet - INFO - Using preprocessor: ip-adapter-auto
2024-09-02 22:43:33,838 - ControlNet - INFO - preprocessor resolution = 1920
2024-09-02 22:43:33,839 - ControlNet - INFO - IP-Adapter using advanced weighting [0.0, 0.0, 0.0, 0.5, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
2024-09-02 22:43:33,936 - ControlNet - INFO - ControlNet Hooked - Time = 0.10921502113342285
100%|██████████████████████████████████████████████████████████████████████████████████| 40/40 [09:22<00:00, 14.06s/it]
2024-09-02 22:53:08,790 - ControlNet - INFO - unit_separate = False, style_align = True 80/120 [18:53<09:25, 14.14s/it]
2024-09-02 22:53:08,790 - ControlNet - INFO - Loading model from cache: ip-adapter-plus_sdxl_vit-h [bc449f62]
2024-09-02 22:53:08,794 - ControlNet - INFO - Using preprocessor: ip-adapter-auto
2024-09-02 22:53:08,794 - ControlNet - INFO - preprocessor resolution = 1920
2024-09-02 22:53:08,795 - ControlNet - INFO - IP-Adapter using advanced weighting [0.0, 0.0, 0.0, 0.5, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
2024-09-02 22:53:08,895 - ControlNet - INFO - ControlNet Hooked - Time = 0.1121377944946289
100%|██████████████████████████████████████████████████████████████████████████████████| 40/40 [09:24<00:00, 14.10s/it]
Total progress: 100%|████████████████████████████████████████████████████████████████| 120/120 [28:38<00:00, 14.32s/it]
Ending job 3 out of 3 (1766.75 seconds)██████████████████████████████████████████████| 120/120 [28:38<00:00, 13.74s/it]
INFO:modules.shared_state:Ending job 3 out of 3 (1766.75 seconds)
Starting job task(7x7xxkckemjhqcr)
INFO:modules.shared_state:Starting job task(7x7xxkckemjhqcr)