Interesting; would you have any specific recommendations for an intelligent (high parameter count? Something that can hold a conversation ...), uncensored, specific model that is able to do this?
If anyone is looking for a cloud option (e.g. don't have the hardware, just want to pay someone else to handle the compute while still writing about generally uncensored things like a violent RPG adventure or something explicit in the adult sense), then I guess NovelAI could fit the bill: https://novelai.net/
I guess most people focus on their image generation which is pretty cool for anime characters and things like placeholder art (say, in gamedev), but their text generation options are also pretty nice and can be used either through their web UI, or something like SillyTavern: https://github.com/SillyTavern/SillyTavern
If you do want to run something locally and have the hardware, I think people discuss models in a variety of places, like https://old.reddit.com/r/LocalLLaMA/comments/16425mk/best_ns... (quick search, was curious, NSFW discussion) and it seems like there's plenty of models on HuggingFace, though I haven't particularly looked into them.
I will absolutely say that even when I was trying to run a general purpose chatbot LLM, it was kind of impressive how quickly things are moving ahead in regards to LLMs: all of these GGML, GGUF and other formats, being able to use even consumer GPUs (though I couldn't get ROCm working with my RX580) and experimental attempts like using Vulkan.
I think most custom fine-tunes and merges on HuggingFace will do this unless they specifically mention it being censored. Even the lower param models have been surprisingly good, with relatively fast progress being made in the 7b and 11b models.
My "daily driver" is Fimbulvetr v2 11b, surprisingly slapped together by an EMT. Kunoichi 7b seems to be a pretty popular model too. These can be run locally with as little as 8 GB free RAM (preferably VRAM) with an easy install solution like LMStudio or Faraday.
You can generally find a lot of recommendations in places like SillyTavernAI or LocalLLaMa on reddit:
This one works with GGUF-compatible llama.cpp wrappers like the fully-local, fully-private Layla app or the other examples tossed around. With some prompt tweaking it's capable of a broad variety of... tasks.
This model is very compliant and doesn't tend to go to "I'm sorry, as a large language model..." type replies.
Somethings I can immediately think of, but definitely not enough to address the issue.
1. Anonymous student ratings, if the solutions are obviously unclear or unhelpful to the student. A history feature would need to be there so the student can go back and mark a question as unhelpful or wrong even if he realises this after a week.
2. You'd need to separate unclear or unhelpful answers from plain wrong maybe? I don't know what's the line you'd need the draw but if an answer is deemed to be false, the student would be eligible for some in-app credit that he can redeem towards his next question/subscription.
3. Random cross verification of answers? Other teachers would get to judge a random answer from another teacher and provide feedback. Students would be notified automatically if an answer is revised in this way, while also providing the correct solution. Teachers can have an infraction counter or something for the same.
Even if they do absolutely nothing with the number, I'm not putting it in. It's not a question of their intention but my convenience and peace of mind.
100% fair. Sorry about that quantxx. The mobile login was only meant to facilitate the login for mobile users through quick Twilio verification. Will move that to username or email instead. Really appreciate the feedback.
reddit has the worst dark pattern in this case... I think most people think they need an email address to signup... yet you can just click next and skip email input... I have about 40 reddit accounts with no emails associated with them... first one was probably created 15 years ago, and last one yesterday
Nice topic to build a project but there's one immediate issue I see with the implementation- The books. It attempts at trying to "facilitate" or "drive away" your current mood by recommending books. Most people can't read books because they have an attention problem. Also imagine someone who does read them, and decide to pick up a book reminiscing about the youth just because he felt nostalgic at a particular moment in their life. What's next? How's he to find the drive to actually finish what he starts? If someone's emotionally pretty turbulent, what happens if he keeps starting new books and can't finish any, leading to feel a certain dread when coming back to the reading table because of all the unfinished quotas he has. This kind of feature will go well with someone more like Blinkist or Deepstash in my opinion with overall summary and immediate conclusion available. Then if the user likes the content at a glance, that would be the fuel they need to actually complete a book after they pick it up. Still, a cool project. I do read a lot and I'd probably pick some book recommendations as well.
MultiUserDomain - a precursor to today's MMO. You could connect to an ascii based adventure/roleplaying game that others could connect as well and communicate. I played astaria.net back in the late 90s.
Multi-User Dungeon; Cf. MUCK for multi-user chat kingdom and friends. Essentially a shared text-only set of chat rooms with a theme and some built-in game logic.
