ThreeStudio Stable Zero123 Install

Table of contents

Requirements

Threestudio Repo

Stable Zero123 Model Download

Photoshop or GIMP

Anaconda Download

Images

Video

Instructions

WSL Setup

Press the Windows key and type in ‘Turn Windows features on and off’

It will ask for you to restart your computer

Restart your computer

Press the Windows key and type in ‘powershell’ and run as administrator

wsl --install

Then fill in the username and password.

Python and Environment Setup

sudo apt-get update
sudo apt install python3-pip python3-dev libgomp1 libegl1-mesa-dev
sudo apt update

CUDA toolkit 12.1 Setup

sudo apt-key del 7fa2af80
wget https://developer.download.nvidia.com/compute/cuda/repos/wsl-ubuntu/x86_64/cuda-wsl-ubuntu.pin
sudo mv cuda-wsl-ubuntu.pin /etc/apt/preferences.d/cuda-repository-pin-600
wget https://developer.download.nvidia.com/compute/cuda/12.1.0/local_installers/cuda-repo-wsl-ubuntu-12-1-local_12.1.0-1_amd64.deb
sudo dpkg -i cuda-repo-wsl-ubuntu-12-1-local_12.1.0-1_amd64.deb

It will tell you to install key run this command: copy and paste it

sudo cp /var/cuda-repo-wsl-ubuntu-12-1-local/cuda-*-keyring.gpg /usr/share/keyrings/
sudo apt-get update
sudo apt-get -y install cuda
export PATH=/usr/local/cuda-12.1/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda-12.1/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

Then to check your version, make sure it’s 12.1

nvcc --version 

Anaconda Setup

cd /mnt/c/Users/jorda/Downloads/
bash Anaconda3-2023.09-0-Linux-x86_64.sh -b -u -p ~/anaconda3
cd ~
~/anaconda3/bin/conda init bash
~/anaconda3/bin/conda init zsh

To check bash profile:

nano .bashrc

Check bottom of .bashrc for >>> conda initialize >>>

Open a new terminal ubuntu tab

Run conda to check if it works:

conda

ThreeStudio StableZero123 Setup

Download the ThreeStudio repo:

git clone https://github.com/threestudio-project/threestudio.git

We will change directory to copy Stable Zero123 ckpt file from windows to WSL:

cd ~/threestudio/load/zero123
cp /mnt/c/Users/jorda/Downloads/stable_zero123.ckpt .

Speed up 3d Generation times by configuring stable-zero123.yaml file:

cd ~/threestudio/configs
nano stable-zero123.yaml

Change batch_size = [12, 8, 4] ~ 5 hours, [8, 4, 2] ~50 mins, [4, 2, 1] ~5mins, [1, 1, 1] ~3mins

press down arrow until at batch_size and change it to [4, 2, 1]

press ctrl+x to quit

Y to save

press enter to keep same file name

Creating Conda Environment

cd ~/threestudio
conda create --name threestudio python=3.10.13
conda activate threestudio 

Install ThreeStudio Dependencies

pip install torch==2.1.2+cu121 torchvision torchaudio -f https://download.pytorch.org/whl/cu121/torch_stable.html
export PATH="/usr/local/cuda/bin:$PATH"
export LD_LIBRARY_PATH="/usr/local/cuda/lib64:$LD_LIBRARY_PATH"
export LIBRARY_PATH="/usr/local/cuda/lib64/stubs:$LIBRARY_PATH"
pip install ninja
pip install -r requirements.txt

Version Check

nvcc --version
python3 --version
conda list

Image to 3D Model Workflow

Create a new dir in C:\Users\jorda called threestudio_exports

Create a new dir in C:\Users\jorda\threestudio_exports called export

Go to either my website to get some samples or go to bing image generator and generate a front and back view

Use something similar to this prompt but keep it simple: T pose of a simple character design front and back solid background

Move your 512x512 png file in your downloads folder

initiate the WSL transfer:

cd ~/threestudio/load/images
cp /mnt/c/Users/jorda/Downloads/CuteMiniFront.png .

change back to main dir:

cd ~/threestudio

3d Training:

python launch.py --config configs/stable-zero123.yaml --train --gpu 0 data.image_path=load/images/CuteMiniFront.png system.prompt_processor.prompt="T pose of a 3d adventure character"

Export the 3d Object:

TRIAL_DIR='outputs/zero123-sai/[64, 128, 256]_CuteMiniFront.png@20240111-181839'
CONFIG_FILE="${TRIAL_DIR}/configs/parsed.yaml"
CHECKPOINT_FILE="${TRIAL_DIR}/ckpts/last.ckpt"
python launch.py --config "${CONFIG_FILE}" --export --gpu 0 resume="${CHECKPOINT_FILE}" system.exporter_type=mesh-exporter system.exporter.context_type=cuda

Change dir to saved export file:

cd ~/threestudio/outputs/zero123-sai
cd [64 (press tab to keep autofilling until you need to type) This will be a lot easier to try and open the dir. The name of it will be in the last output in the CLI
cd save

Copy from WSL to windows:

cp -r * /mnt/c/Users/jorda/threestudio_exports/export

Go back to windows and go to the C:\Users\jorda\threestudio_exports dir and rename export to StableZero123_NameOfObject_BS_4-2-1

Repeat Image to 3D Model Workflow to generate more models.