Asklemmy
A loosely moderated place to ask open-ended questions
Search asklemmy π
If your post meets the following criteria, it's welcome here!
- Open-ended question
- Not offensive: at this point, we do not have the bandwidth to moderate overtly political discussions. Assume best intent and be excellent to each other.
- Not regarding using or support for Lemmy: context, see the list of support communities and tools for finding communities below
- Not ad nauseam inducing: please make sure it is a question that would be new to most members
- An actual topic of discussion
Looking for support?
Looking for a community?
- Lemmyverse: community search
- sub.rehab: maps old subreddits to fediverse options, marks official as such
- !lemmy411@lemmy.ca: a community for finding communities
~Icon~ ~by~ ~@Double_A@discuss.tchncs.de~
Please don't, if someone wants that,, they can just go do it themselves.
Hey, I didn't mean for the LLM to write the response to the WP publicly, but for my own satisfaction for reading a good story. I wouldn't be posting such responses online anyway
I'd rather see no responses than any "written" by an LLM. I say that having come across several interesting prompts without any responses.
Copied from the other comment:
Hey, I didn't mean for the LLM to write the response to the WP publicly, but for my own satisfaction for reading a good story. I wouldn't be posting such responses online anyway
GPT 4 can write decently but it's either going to be a bit corny or very derivative. Really a LLM is the wrong type of AI for this task. A GAN (a type of Generative AI) built for writing is going to produce a lot better text in terms of creativity and bringing something new to the table. If it's too out there you could then pass that through GPT to fix any potential language issues or coherency issues (in terms of language, GPT can't fix plot holes well).
Thanks. I'll take a look if I can run such a GaN privately. I didn't mean for the LLM to write the response to the WP publicly, but for my own satisfaction for reading a good story. I wouldn't be posting such responses online anyway.
Thanks again for pointing out the differences between them
Iβve had good results running models like:
https://huggingface.co/ehartford/Wizard-Vicuna-13B-Uncensored
In KoboldAI.
You have to tweak the generation parameters substantially though if you want anything original, but if you go too far the grammar starts suffering.
Thanks, will take a look