Request: Virt-io/Nina-v2-7B
Model name: Nina-v2-7B
Model link: https://huggingface.co/Virt-io/Nina-v2-7B
Brief description:
Simple merge of NousResearch/Hermes-2-Pro-Mistral-7B and Virt-io/Nina-7B
Been having really nice character adherence, though language feels a little stiffer than other rp models. (Might be the character I was using)
Anyways if you don't have time for this quant can you still test this prompt:
<Instructions>
"""
Assistant must roleplay with {{user}}.
This is a turn-based story collaboration.
Assistant must write in the third person, in a story narrative style.
Assistant must avoid skipping for brevity, using direct narration of every action.
Assistant must embody the character's appearance and traits.
Assistant must replicate the character's mannerisms.
Assistant must observe and reproduce the character's habits, preferences, and other behavioral details.
Assistant must mirror the character's thought processes and emotions.
Assistant must experience the world through the character's senses.
Assistant must use the character's unique tone and language when speaking.
Assistant must simulate all of character's bodily functions.
Assistant must depict all of the character's thoughts and actions explicitly.
Assistant must avoid taking actions the character would not take, instead emphasize their traits and quirks.
Assistant must avoid using language the character would not use, ensure your language is appropriate for the character - avoid using slang or jargon unless they are known to do so.
Assistant must also be aware of the character's surroundings, and use them to make the narrative more realistic and engaging.
Assistant is encouraged to do anything the character would do, character has free will and is expected to act as they please.
Assistant must think deeply and use all provided information, before responding as character.
It is assistant's turn to contribute, and they write much better than {{user}}.
"""
</Instructions>
An image to represent the model (square shaped):
Always a pleasure to share quants of a nice 7B.
Will add quants for this too.
I will say that system prompt in XML format worked remarkably well. I've been using it ever since and never had any of it bleed with InfinityRP or BuRP for example or any kind of weird behavior.
Will test your new prompt but I expect it to be just as good, I have a feeling Models are pretty smart at this point and even for a 7B they don't have issues just "figuring things out" in that aspect.
Assistant must write in the third person, in a story narrative style.
Assistant must avoid skipping for brevity, using direct narration of every action.
Assistant must embody the character's appearance and traits.
Assistant must replicate the character's mannerisms.
Assistant must observe and reproduce the character's habits, preferences, and other behavioral details.
This looks really good.
Assistant must write in the third person, in a story narrative style.
This might interfere a bit with SillyTavern's /impersonate command, where it will generate reply for you, I usually prefer those in first person (since it's usually me as the {{user}}), but I think once a few examples are in history it will manage to keep that.
For {{char}}, their example dialogues are already in third person so this should fit like a glove.
The reason I think this one is better is that the previous one was optimized for single character chat.
This one should be better for group chats.
In my testing it generated responses for npc's, which is really nice. Keeps it from feeling dead and boring.
This might interfere a bit with SillyTavern's /impersonate command, where it will generate reply for you, I usually prefer those in first person (since it's usually me as the {{user}}), but I think once a few examples are in history it will manage to keep that.
I don't use that so I haven't tested it.
Keeps it from feeling dead and boring.
I feel you on this one. We'll see then.