Another test model.
This is a merge of pre-trained language models created using
mergekit.
This model was merged using the SLERP merge method.
1base_model: wzhouad/gemma-2-9b-it-WPO-HB
2dtype: bfloat16
3merge_method: slerp
4parameters:
5 t:
6 - filter: self_attn
7 value: [0.0, 0.5, 0.3, 0.7, 1.0]
8 - filter: mlp
9 value: [1.0, 0.5, 0.7, 0.3, 0.0]
10 - value: 0.5
11slices:
12- sources:
13 - layer_range: [0, 42]
14 model: nbeerbower/Gemma2-Gutenberg-Doppel-9B
15 - layer_range: [0, 42]
16 model: wzhouad/gemma-2-9b-it-WPO-HB
