This repo contains GGUF quants of
kuno-kunoichi-v1-DPO-v2-SLERP-7B.
kuno-kunoichi-v1-DPO-v2-SLERP-7B is a merge of pre-trained language models created using
mergekit.
The expectation is that the result is more robust against errors or when merging due to "denseness", as the two models likely implement comparable reasoning at least somewhat differently.
Tested lightly with ChatML instruct format prompting using temperature=1.1 and minP=0.03. The model natively supports Alpaca format prompts.
This model was merged using the SLERP merge method.
1slices:
2 - sources:
3 - model: SanjiWatsuki/Kunoichi-7B
4 layer_range: [0,32]
5 - model: SanjiWatsuki/Kunoichi-DPO-v2-7B
6 layer_range: [0,32]
7merge_method: slerp
8base_model: SanjiWatsuki/Kunoichi-7B
9parameters:
10 t:
11 - value: 0.5
12dtype: float16
13
