Cemmy

joined 1 month ago
[–] Cemmy@lemmy.world 1 points 2 weeks ago* (last edited 2 weeks ago)

The strange thing is, in the Brave browser, even it says to use Firefox or Chrome, I don't know if it is using the lore or not, but I haven't seen that message and I can actually, unlike either with Firefox or Chrome, save new lore. Does Brave work otherwise? Cos it seems to NOT have the problems Firefox or Chrome have.

EDIT: After a test. I think the Brave version is managing to use the lore.

[–] Cemmy@lemmy.world 1 points 2 weeks ago

OK I tried Chrome, and when I tried saving new lore, which I can't do Firefox either, it just fails after a while, it failed but I got this error message:

There was an error while saving the lore entries: TypeError: window.textEmbedderFunction is not a function at embedTextWithLocalModel (:4560:19) at async window.embedTexts (:4594:20) at async HTMLButtonElement.sendButtonClickHandler (:8184:35)

3
submitted 2 weeks ago* (last edited 2 weeks ago) by Cemmy@lemmy.world to c/perchance@lemmy.world
 

I just got a message in AI Character Chat saying "For some reason your web browser hasn't correctly loaded the files needed for your character's memory storage/retrieval to work properly. Please use Chrome or Firefox if possible, and ensure you're using the latest version. In the meantime, character memories/lore is disabled. This is fine - it just means that the AI won't be quite as 'smart' as it could be."

I am using Firefox and it is up-to-date. I am on a Windows 10 PC. Is it because I'm using it in on a private tab? Or I have an antivirus internet security program, or that I always have vpn on? Or does it matter that I only have 7 GBs free on my device? What could be causing this error?

OK, it doesn't seem to be the private tab, it came up again on a non-private tab.

And I told the internet security extension for the browser to allow banners and data collecting on the site, and still got the error.

Could it related to this: https://lemmy.world/post/33186393

[–] Cemmy@lemmy.world 1 points 2 weeks ago (1 children)

Oh, OK, well thanks for answering.

2
submitted 2 weeks ago* (last edited 2 weeks ago) by Cemmy@lemmy.world to c/perchance@lemmy.world
 

The error says An unhandled promise rejection occurred: "Content-Length header of network response exceeds response Body." What will this error do? The generator is just the default character chat.

 

Hey I know on your main page for the AI Character Chat, you can create new folders for organising your characters, but is there any way, if you have some that you want all in the same folder (say they're all going to be used in the same chat), once you have them in that folder to be able to reorder what order they are in?

[–] Cemmy@lemmy.world 1 points 3 weeks ago

Cool, I hadn't really been bothered about which direction my images were being seen from, but now I will, since it will allow other instructions for sizes, thanks!

[–] Cemmy@lemmy.world 1 points 3 weeks ago (2 children)

Thanks for the advice and linking me to that site! When you say if you have frame/camera angle descriptions can they be added when you are trying to get pictures as cartoons, anime, comics, etc instead of realistic photo images?

 

When generating images, is there way to get it to understand the relative size of objects in the picture? Like you want an object A, and then you want an object B and you want Object B to x times the size of Object A. Can you go "Object B is 3 the size of Object A"? Or does it get measurements? "Object A is 1 meter long, Object B is 3 meters long."?

[–] Cemmy@lemmy.world 1 points 3 weeks ago

That's a pity, thanks though.

 

In the AI Story Generator, I see it talks about about extra paragraphs being added as special summary paragraphs, and yeah I see them, with things like "SUMMARY^1:" in front. It says to not move or delete them, but what about adding new ones to help it out by manually typing out paragraphs in the main story box and also typing that label in front?

[–] Cemmy@lemmy.world 1 points 1 month ago (2 children)

Oh sorry I should have specified I meant for creating an image.

2
submitted 1 month ago* (last edited 1 month ago) by Cemmy@lemmy.world to c/perchance@lemmy.world
 

Hey, is there a way to make the AI not to mix up different character attributes? What I mean is, say I am trying to get characters from a show, but not ones it seems to know, but also make changes to them. I write ""First Characters Name" has "Attribute 1". "First Characters Name" has "Attribute 2". "Second Characters Name" has "Attribute 3"." and then a lot of the time it creates 2 characters, one with "Attribute 1" and "Attribute 3" and the other with "Attribute 2" (this is simplified, I am trying more than 3 attributes).

EDIT: Sorry, new here, only really used the image generator so far, so that's what I'm talking about.

[–] Cemmy@lemmy.world 1 points 1 month ago

Ah right, got you, thanks.

[–] Cemmy@lemmy.world 2 points 1 month ago (6 children)

Wait, are you saying everything here https://perchance.org/tutorial is not used anymore?

[–] Cemmy@lemmy.world 1 points 1 month ago

Oh, so the brackets, including one or two, do nothing. Right, got you, thanks.

 

On the Description box of the image generator there is a little i which says "Add brackets around text (like this) or ((even more)) to get the AI to pay more attention to it". Is this limited to 2 sets of brackets or does it continue, that having (((something))) makes the AI pay even more attention and then 4 sets even more and so on?