I just got access to this model and couldn't wait to try out something today.

I just got access to this model and couldn't wait to try out something today. As a personal project, I've been building an experience point rating bot for discord. Most work off of numbers of posts or timeframes, but I've been working on my own classification model using KNN and a Hugging Face text analyzer. I did a preliminary test in Jev with some sample data. It didn't take me long, was able to use plain english, and it did pretty darn well out of the box. Well done. I'm going to experiment more, but if you want my very first attempt, here is what I put: ``` { "level_of_post": { "type": "choice", "instructions": "How good is this post from discord `text`", "criteria": { "Very Detailed": "The post contains a lot of description as well as some dialog. It advances the story in a significant way and includes descriptors about the environment and possibly internal dialog or debate.", "Somewhat Detailed": "The post contains some description and adjectives, but largely about the subject themselves. May have some dialog amidst the decriptions.", "Si