BRAND NEW Stable Diffusion 2 Model | Local Installation | Tutorial | Manual Command Line Prompting

Want to learn how to train Stable Diffusion with your own subjects instead? Watch my past tutorial which uses the past Stable Diffusion 1.4 model and AUTOMATIC1111's awesome Web UI here: https://www.youtube.com/watch?v=Ho5cjrNAXrs

THE NEW STABLE DIFFUSION 2 MODELS ARE HOT AND READY!

WALKTHROUGH:

Get Python version 3.9:
https://www.python.org/downloads/release/python-390/

Get Git:
https://git-scm.com/downloads

=========================================

Get Stable Diffusion dependencies:
1. Get OpenCV with PIP:
pip install opencv-python

2. Get NumPy with PIP:
pip install numpy

~~~~~~~~~~~~~~

3. Get PyTorch with PIP:
----- 3A. (need to modify registry first to enable long paths):
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\FileSystem
Set LongPathsEnabled to 1

----- 3B. (check CUDA version to get the correct PyTorch):
NVIDIA Control Panel - Help - System Information - Components

https://pytorch.org/get-started/locally/

~~~~~~~~~~~~~~

4. Get PyTorch Lighting with PIP:
pip install pytorch-lightning

5. Get OmegaConf with PIP:
pip install omegaconf

6. Get TQDM with PIP:
pip install TQDM

7. Get EINOPS with PIP:
pip install einops

8. Get watermarking modules with PIP:
pip install imwatermark
pip install invisible-watermark

9. Get Transformers with PIP:
pip install transformers

10. Get Open Clip with PIP:
pip install open_clip_torch

=========================================

Get Stable Diffusion base files (by cloning the repository):
https://github.com/Stability-AI/stablediffusion

Get the new Stable Diffusion 2 model checkpoints (via download link):
512 - https://huggingface.co/stabilityai/stable-diffusion-2-base/blob/main/512-base-ema.ckpt
768 - https://huggingface.co/stabilityai/stable-diffusion-2/blob/main/768-v-ema.ckpt

=========================================

11. Initiate the LDM environment:
pip install -e .

=========================================

GENERATE ART WITH THE NEW MODELS:

PROMPTING GUIDE: https://www.assemblyai.com/blog/how-to-run-stable-diffusion-locally-to-generate-images/

512px generation sample prompt:
python scripts/txt2img.py --prompt "a professional photograph of an astronaut riding a horse" --ckpt C:\STABLEDIFFUSION2\stablediffusion\models\512model.ckpt --H 512 --W 512 --steps 30 --seed -1 --n_samples 1 --n_iter 1

768px generation sample prompt:
python scripts/txt2img.py --prompt "a professional photograph of an astronaut riding a horse" --ckpt C:\STABLEDIFFUSION2\stablediffusion\models\768model.ckpt --config configs/stable-diffusion/v2-inference-v.yaml --H 768 --W 768

=========================================

Thank you so much for watching, listening to, and supporting my content!

- Mike Yeah

© YebbaProductions