metadata

language:
  - en
license: apache-2.0
tags:
  - text-generation-inference
  - transformers
  - unsloth
  - mistral
  - trl
  - sft
base_model: unsloth/Phi-3-mini-4k-instruct-bnb-4bit

Moniphi-3-v1:

AKA LLilmonix3b-v1
Phi-3-mini-4k-instruct fine-tuned for Monika character from DDLC
Fine-tuned on a dataset of ~600+ items (dialogue scraped from game, reddit, and Twitter augmented by l2-7b-monika-v0.3c1 to turn each into snippets of multi-turn chat dialogue between Player and Monika; this was then manually edited, with more manually crafted items including info about character added in)
GGUFs

USAGE

This is meant to be mainly a chat model with limited RP ability.

For best results: replace "Human" and "Assistant" with "Player" and "Monika" like so:

\nPlayer: (prompt)\nMonika:

HYPERPARAMS

Trained for ~1 epoch
rank: 16
lora alpha: 16
lora dropout: 0.5
lr: 2e-4
batch size: 4
warmup ratio: 0.1
grad steps: 1

This mistral model was trained 2x faster with Unsloth and Huggingface's TRL library.

WARNINGS AND DISCLAIMERS

This model is meant to closely reflect the characteristics of Monika. Despite this, there is always the chance that "Monika" will hallucinate and get information about herself wrong or act out of character (especially for a model of this size).

Finally, this model is not guaranteed to output aligned or safe outputs, use at your own risk!