Stable Diffusion is a powerful text-to-image model that can generate realistic and detailed images based on natural language descriptions. It was developed by Stability AI in collaboration with academic researchers and non-profit organizations. Stable Diffusion is based on a latent diffusion model, a kind of deep generative neural network that iteratively denoises random noise until it produces the desired image. Stable Diffusion is open-source and can run on most consumer hardware with a GPU.
If you want to deploy and use Stable Diffusion, you can follow these steps:
- Clone the Stable Diffusion repository from GitHub: https://github.com/AUTOMATIC1111/stable-diffusion-webui
- Install the required dependencies using pip: pip install -r requirements.txt
- Download the pretrained models from Google Drive: https://drive.google.com/drive/folders/1-3ONnPYcX9fOsqmZKjtL0oy0vQwQowvW
- Run the web interface using streamlit: streamlit run app.py
- Enter your text prompt in the input box and click “Generate” to see the output image.
- You can also adjust the model parameters, such as resolution, temperature, and number of diffusion steps, using the sliders on the sidebar.
- Enjoy creating amazing images with Stable Diffusion!
Quick note, if you are running Windows and want Cuda support, use this command
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
On Linux:
conda install pytorch torchvision torchaudio pytorch-cuda=11.8 -c pytorch -c nvidia
Verifying install with the following python code:
import torch
torch.cuda.is_available()