Click here for holiday cheer!
AvaniJB - Total Latte (and Gemini) Love
Now also for gemini-exp-1206
JB made to be easy to edit/customise and extremely lightweight at less than 200 tokens for all essentials.
Everything you need to know is in the FAQ and JB. If something doesn't work, @ on discord, check the thread on Scylla. Burner is littledoodler@proton.me. I'm on /aicg/ sometimes. All art is by myself, not genned. Check https://rentry.org/latteguide for more latte info. Not necessary, but good to know. Gallery of art for past versions: https://postimg.cc/gallery/vyG3MSG
- Credits
- DOWNLOAD HERE
- FAQ
- "I hate bloat presets."
- "Who the fuck is this catboy who keeps making things jolly?"
- "Where does logit bias go?"
- "Does this work on Claude?"
- "How do I edit?"
- "How do I edit (on Gemini)?
- "I don't know what (x) prompt does!"
- "I read the Readme's and this FAQ and I still have questions/want to contribute!"
- Old versions
- My Bots
Credits
- all my testers and helpers!!! (permanent credits: @kira####, @pur####, @holyw####, @sopade####, @hal####, @guil####)
- karukaru, knottydog, mongo, and whoever else for the new del prompt for 2.4. I didn't come up with the idea, i merely adapted it (because frankly for this latte version it works the best)
- koalache & karukaru for exchanging ideas and helping me change dumb stuff. Also being fellow jbmakies
- @rarest#### for being an absolute regex entity, and giving me a few ideas
- @Sir#### for access to his proxy, as well as donating a regex and his info board :lattelove:
- Mr. Regex :salute:
- @minsk#### for his amazing JED jb, and exchanging ideas with
- @Smiley####, for inspiring previous presets!
- the lattelover cult on scylla :lattelove:
- and someone who doesnt want to be named but got me into making jbs now, i guess. theyre also to be credited for at least half of the logit bias.
DOWNLOAD HERE
2.5 - Padoru Edition | Updated 23.12.24
Written for both chatgpt-4o-latest and the gpt-4o-1120 snapshot. Yes, it also works for other GPT-4o versions, you may need to tweak the logit, though.
Alternate Sampler choices/recommendations: 0.89 to 0.91 Temp, 0.10 Freq. Pen., 0.05 Pres. Pen., 1.0 Top P. (recommended by @holy#### @icarus)
1.1 Temp, 0 Freq. Pen., 0 Pres. Pen, 0.95 Top P.
Samplers can depend on preference. Keep in mind that Logit Bias values need to be higher for the same effect if Top P is reduced from 1.
Alternate </style> prompt that I use for myself, but haven't included to avoid too many prompts:
# The narrator is autodiegetic. The narration will be written as if from first-person perspective, with third-person pronouns.
</style>
Just copypaste it into </style> if you want to use this.
Logit Bias (v13) — last updated 24.12.24 — NOT optional, 100% required
catbox fileditch yes you can use this with whatever other preset you want. not stopping you
Regexes
Your required regexes depend on what you want to use. The non-End regexes are optional QoL. Credit for new delete error and delete last paragraph regex goes to Mr. Regex himself. Credit for the remove del regex goes to KnottyDog (i think).
End #1 delete del | catbox delete del | fileditch End #2 delete error | catbox delete error | fileditch End #4 & #5 delete repeat | catbox delete repeat | fileditch End #5 & #6 delete curry | catbox delete curry | fileditch
QoL shitty last paragraph deleter | catbox fix quotes | catbox fix quotes | fileditch
Gemini Version [updated: 12.12, 1:00] (updated for 1206 & 2.0 Flash EXP)
catbox | normal version (WILL NOT GET UPDATED FOR NOW) catbox | schizokino version UPDATED No regexes needed. Normal version is more normal. Schizokino version is... a very special boy. Incredibly quirky. Try both, yknow. For fun. Gemini stuff is ATM in development. Don't expect all toggles to work. I recommend setting top_k to 0 and using 2.0 Flash EXP. Been having a good time that way.
Patch Notes (2.5)
- changed the way style prompt is structured. Several prompts are now wrapped in <style> tags, which makes them more effective. due to this, squash system messages is also turned on now.
- removed all mentions of {{char}} and {{user}}, giving the entire jb far better compatibility with non-{{char}} cards. due to this, character names behaviour is now changed to completion object
- end#0 has been made jolly. presents for everyone!
- style 1 and 2 recieved several changes. style 1 is the quirkier one, style 2 is more normal.
- structure (aka the random structure prompt) has been moved out of style into its own prompt that you can toggle on or off.
- new prompt for randomising narration in "</style>". apparently insanely effective, you should probably end up choosing one, but the roulette is pretty fun for making swipes a lot more diverse.
- the database tag and the prompts within have been consolidated into one prompt named "Database". card personality is NOT included in the prompt by default because it usually just makes writing worse.
- included Database (Alt) prompt if writing becomes sloppy. Will give the model more amnesia, but reinforces character traits better and makes writing fresher. disabled by default.
- added "anti-filter" toggle. may reduce amount of rejects you get with end 1 and 2.
- improved homosexuality
FAQ
"I hate bloat presets."
That's not a question, but yeah, me too. This is why this JB is extremely lightweight, coming in at 200 tokens for basic functionalities, +400 tokens for an additional, strong Style prompt, and no optional toggles that exceed 200 tokens. Most toggles are around 50-100 tokens at the very most and have been heavily tested to ensure they actually do something. Adding your own toggles is incredibly easy.
"Who the fuck is this catboy who keeps making things jolly?"
Skill Check :)
"Where does logit bias go?"
Under chat completion presets, above the prompt manager. Your endpoint needs to be openai. Doesn't work on Azure or Claude or Gemini.
"Does this work on Claude?"
This preset has been tested on Claude and works 100%. How well? IdDK.
"How do I edit?"
Small guide on how Latte works:
Latte prioritises tokens that are lower down. Meaning that the lower down something is, the stronger it is represented. You don't need to write it in a gaslighty way, just be concise in what you want and you will get the exact same results. Aka ao3 style prompting doesn't really work here. Anyone who says otherwise is huffing placeboium.
Use as little tokens as possible. Be direct and concise in your instructions. Cut down whatever yap you're thinking about and just say directly what you want from Latte. Start any prompt with a hashtag like this:
You can also give your instructions a title you can call back to.
Anything you want to call back to, or if you want to group instructions, you want to do like this:
Example of slop:
Example of good:
There are several special prompts marked 🥤 on the JB specifically made for you to work with. They contain instructions on how to edit them on top of that.
"How do I edit (on Gemini)?
Gemini preset is built differently. Instead of System, every role is AI Assistant. For whatever reason, this works far better. This means Avi is also gone. You write prompts the same way like on Latte, with one crucial difference:
An actual example:
Everything else works the same. It doesn't matter how you structure the prompt manager to Gemini (so far), it will read and weigh all of it the same amount.
"I don't know what (x) prompt does!"
Open it and read.
"I read the Readme's and this FAQ and I still have questions/want to contribute!"
Scroll up to check the latteguide rentry or to contact me.
Old versions
2.3.5: Curry Edition catbox fileditch
2.2: Spooky Edition catbox fileditch
2.1: Avi Edition catbox
2.0: Voodoo Edition catbox fileditch
1.0.3: The one before the big one Edition catbox
1.0.0: I have no idea what I'm doing Edition fileditch
My Bots
Once a year, I upload bots. I'll probably upload more of mine. Link to my profile here.