Installing AI for drawing

Preparation

Let's assume that Ubuntu 26.04 is ready.

sudo apt update
sudo apt install software-properties-common
sudo add-apt-repository ppa:deadsnakes/ppa

Downloaad stable diffusion-webui-forge

git clone https://github.com/lllyasviel/stable-diffusion-webui-forge-git
sudo apt install python3.10 python3.10-venv python3.10-dev

Edito i files

Edito webui-user.sh e setto

python_cmd="python3.10"

Resetto il venv

rm -rf venv

e lo rigenero

./webui.sh

Poi eseguo

./venv/bin/python -m pip install --upgrade pip setuptools wheel
./venv/bin/python -m pip install torch==2.3.1 torchvision==0.18.1 torchaudio==2.3.1 --index.url https://download.pytorch.org/whl/cu127 --extra-extra-index-url https://pypi.org/simple
./venv/bin/python -m pip install setuptools==69.5.1
./venv/bin/python -m pip install --index-url https://pypi.org/simple "typing-extensions==4.12.2"

Edito modules/launch_utils.py

Alla riga 443 sostituisco

run-pip (f"install {clip_package}", "clip")

con

run-pip (f"install --no-build-isolation {clip_package}", "clip")

Poi continuo con le modifiche a venv

./venv/bin/python -m pip uninstall -y numpy scikit-image==0.21.0
./venv/bin/python -m pip install numpy==1.26.4 scikit-image==0.21.0

Come ultima cosa

curl -L -o ~/stable-diffusion-webui-forge/models/Stable-diffusion/v1_5-pruned-emaonly.safetensors "https://huggingface.co/stable-diffusion-v1-5/stable-diffusion-v1-5/resolve/main/v1-5-pruned-emaonly.safetensors?download=true"

A questo punto basta lanciare ./webui.sh per ottenere la pagina richiesta.