This is a merge of pre-trained language models created using
mergekit.
This model was merged using the
TIES merge method using
Orion-zhen/Meissa-Qwen2.5-7B-Instruct as a base.
1models:
2 - model: Orion-zhen/Meissa-Qwen2.5-7B-Instruct
3 parameters:
4 density: 1
5 weight: 1
6
7 - model: win10/high-speed-mixing-7B-V1
8 parameters:
9 density: 1
10 weight: 1
11
12 - model: Orion-zhen/Meissa-Qwen2.5-7B-Instruct
13 parameters:
14 density: 1
15 weight: 1
16
17 - model: win10/high-speed-mixing-7B-V1
18 parameters:
19 density: 1
20 weight: 1
21
22merge_method: ties
23base_model: Orion-zhen/Meissa-Qwen2.5-7B-Instruct
24parameters:
25 normalize: true
26 int8_mask: true
27dtype: bfloat16
28
