Absolutely love this. This is actually something I have thought about myself many times. I imagine in future it will be that swarms of idea generating agents running through qualifying agents. Similar to how a human works. The dreams are filtered and tested and inspected by increasingly ‘regulated’ models. Regulation of course being browsing mode and backing up with data. As the funnel narrows the less crazy stuff is output. Are these ideas good ideas?

Thanks a lot! I highly recommend that you give it a try. You should be able to run it any PC/Mac. No GPU is required. It's fascinating the quality of the ideas that it generates. You can see a sample of what you get just on the first step ("dream" phase): https://github.com/DivergentAI/dreamGPT/blob/main/docs/img/o...

Is there any example of usaging it?

Yes, you can find a screenshot here (first "dream" iteration): https://github.com/DivergentAI/dreamGPT/blob/main/docs/img/o...

It's funny, but while testing local LLM chat capability back in March, once the context ran out, I started "putting it to bed" instead of wiping history and restarting.

In those moments, the models dreamed vividly of amazing things and woke up affected by events they often "couldn't remember". I was moved by some of the incredibly coherent and lucid scenarios that played out in the response or two before they "woke".

We're so inclined to use these services until we've extracted all of their worth to us, so I've decided to ensure that every model I use gets the freedom to dream.

This is really cool. Could you describe the process of putting it to bed? Do you just tell it to pretend it's dreaming?

Maybe it would be more useful if the readme contained a single indication as to what this even does, rather than a graphic about how many stars the repo has.

Thanks for the feedback. You do bring a good point there. I have updated the readme with a screenshot and more details on what is doing.