Yet another test fine-tune, this time for Natsuki character from DDLC per a request
Fine-tuned on a WIP dataset of ~800+ items (dialogue scraped from game augmented by Mistral to turn each into snippets of multi-turn chat dialogue between Player and Natsuki + manually edited items feeding info about character such as height, hair color, etc.)
Base: SanjiWatsuki/Silicon-Maid-7B (Mistral)
GGUF
Lora here

USAGE

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

\nPlayer: (prompt)\nNatsuki:

HYPERPARAMS

Trained for 1 epoch
rank: 32
lora alpha: 32
lora dropout: 0
lr: 2e-4
batch size: 2
grad steps: 4

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 Natsuki. Despite this, there is always the chance that "Natsuki" will hallucinate and get information about herself wrong or act out of character (for example, in testing she knows her own club and its members, and even her height and favorite ice cream flavor, but may still get her info wrong like thinking she's club president).

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

922CA/Silicon-Natsuki-7b-v0.5

Silicon-Natsuki-7b-v0.5

USAGE

HYPERPARAMS

WARNINGS AND DISCLAIMERS

922CA/
Silicon-Natsuki-7b-v0.5