sdxs-cli

---

Setup

To use SDXS you'll need to have Termux, this app allows you to use a linux distro in your mobile phone. Download Termux and install it on your phone.

Install a linux distro

pkg install proot-distro -y
proot-distro install debian

Note: the distro will be located at /data/data/com.termux/files/usr/var/lib/proot-distro/installed-rootfs/debian/root.

Allow Termux to make a storage mount next to $HOME (wiki).

apt upgrade -y
termux-setup-storage

Login to the linux distro

proot-distro login debian --shared-tmp

Install sdxs-cli inside the linux distro

apt update -y
apt install git git-lfs vim python3 python3-pip python3-venv -y
git clone https://github.com/self-destruction/sdxs-cli
cd sdxs-cli/

Install deps of sdxs-cli

python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Run cli (it will take longer to run the first time, cause model is being downloaded)

python app.py --prompt "A majestic goat with a realistic sunset and a blue lake, very detailed, a masterpiece"

Usage params:

Params:
  --prompt     # prompt string
  --large_vae  # (optional) not exist by default, takes 4 times longer

To get access the generated pictures, you have to exit the linux dist (exit command) and copy the contents of the generation folder to mount with the following command:

cp -r /data/data/com.termux/files/usr/var/lib/proot-distro/installed-rootfs/debian/root/sdxs-cli/output $HOME/storage/downloads

Note: now the output folder with generations can be found in the file manager in the root folder Download.

Launching on next times when everything is installed

proot-distro login debian --shared-tmp
cd sdxs-cli/ && source venv/bin/activate
python app.py --prompt "A majestic goat with a realistic sunset and a blue lake, very detailed, a masterpiece"

Don't forget to sync output folder to the mounted one.

Results

Testing was done on the Snapdragon 8 gen 3. 12+8 gb ram.

Image	Command	Result
	python app.py --prompt "A majestic goat with a realistic sunset and a blue lake, very detailed, a masterpiece"	1/1 [00:03<00:00, 3.94s/it]. Execution time: 7.472 seconds
	python app.py --prompt "A majestic goat with a realistic sunset and a blue lake, very detailed, a masterpiece" --large_vae	1/1 [00:03<00:00, 3.63s/it]. Execution time: 35.191 seconds