TRT is the future and the future is Now

#aiart #A1111 #nvidia #tensorRT #ai #StableDiffusion

Install nvidia TensorRT on A1111

▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬

SOCIAL MEDIA LINKS!

✨ Support my (*・‿・)ノ⌒*:･ﾟ✧

character available at https://ko-fi.com/impactframes/shop

SD related https://civitai.com/user/impactframes

▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬

VIDEO LINKS📄🖍️o(≧o≦)o🔥

▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬

https://github.com/AUTOMATIC1111/stable-diffusion-webui

https://nvidia.custhelp.com/app/answers/detail/a_id/5487/~/tensorrt-extension-for-stable-diffusion-web-ui

https://github.com/NVIDIA/Stable-Diffusion-WebUI-TensorRT/tree/main

https://www.python.org/downloads/release/python-31011/

https://git-scm.com/downloads

https://civitai.com/models/125703/protovision-xl-high-fidelity-3d-photorealism-anime-hyperrealism-no-refiner-needed

▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬

🔥NOTES

▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬

This only works on Nvidia GPUs

will only work on dev branch for now

git clone https://github.com/AUTOMATIC1111/stable-diffusion-webui.git

cd stable-diffusion-webui

git checkout dev

git pull

run the webui.bat to instal the venv

cd models

delete the folder inside

make your models symlinks

for example you could ls to list the folders inside your original models instalation copy the list names

and ask bard, chatgpt or clade ai to make a command for each but be sure to read and make sure is correct b4 you execute it

mklink /D E:\stable-diffusion-webui\models \\wsl.localhost\Ubuntu\home\impactframes\automatic\models

mklink /D E:\stable-diffusion-webui\models C:\ComfyUI\models

cd ..

cd ./venv/scripts

activate

python.exe -m pip install --upgrade pip

python.exe -m pip install nvidia-cudnn-cu11==8.9.4.25 --no-cache-dir

python.exe -m pip install --pre --extra-index-url https://pypi.nvidia.com/ tensorrt==9.0.1.post11.dev4 --no-cache-dir

python.exe -m pip install polygraphy --extra-index-url https://pypi.ngc.nvidia.com

python.exe -m pip install onnxruntime

python.exe -m pip install colored

python.exe -m pip install protobuf==3.20.2

python.exe -m pip install onnx-graphsurgeon --extra-index-url https://pypi.ngc.nvidia.com

python.exe -m pip install xformers==0.0.21

python.exe -m pip uninstall -y nvidia-cudnn-cu11

Move into the extensions folder and

and tyme CDM in the search bar

x:\stable-diffusion-webui\extensions>

git clone https://github.com/NVIDIA/Stable-Diffusion-WebUI-TensorRT.git

now add the variables to the webui-user.bat

@echo off

set POLYGRAPHY_AUTOINSTALL_DEPS=1

set CUDA_MODULE_LOADING=LAZY

set PYTHON=

set GIT=

set VENV_DIR=

set COMMANDLINE_ARGS= --api --xformers

call webui.bat

save it then double click and wait a long time until it starts at least give it 3 min it should load

Go to settings > User Interface > quicklist > add sd_unet

select a model

TensorRT tab export a default enginne or with the settings you like