I've been running DD for a few months now... I tend to just edit the python script or use e.g. entmike's fork which can read config files to make changes to the 50+ parameters (basically everything is better than having to use Jupyter Notebooks IMO), granted if you don't have a GPU with 6+ GB of VRAM, you can often get a decent enough GPU for free from Google Colab. For running locally, I can also highly recommend Visions of Chaos, which includes multiple versions/forks of Disco Diffusion, as well as a ton of other latent diffusion scripts, not to mention many many many other generation features such as fractals and even music. They also recently added the ability to train your own diffusion models which I've been doing the last few days using thousands of my own photographs. It also has a pretty nice GUI and the dev is extremely responsive on Discord. Also after you do the setup for VoC it handles running all the python venv setup stuff otherwise necessary with local DD installs. In any case, check out DD Discord and/or VoC discord for lots of info, tips, help, examples, and support.