This is a merge of pre-trained language models created using
mergekit.
This model was merged using the
DARE TIES merge method using
Zyphra/ZR1-1.5B as a base.
1models:
2 - model: Zyphra/ZR1-1.5B
3 parameters:
4 weight: 5
5 - model: nvidia/AceMath-1.5B-Instruct
6 parameters:
7 weight: 4
8 - model: Spestly/AwA-1.5B
9 parameters:
10 weight: 3
11 - model: MadeAgents/Hammer2.1-1.5b
12 parameters:
13 weight: 3
14 - model: SakanaAI/TinySwallow-1.5B-Instruct
15 parameters:
16 weight: 2
17 - model: Amu/t1-1.5B
18 parameters:
19 weight: 2
20tokenizer:
21 source: union
22merge_method: dare_ties
23base_model: Zyphra/ZR1-1.5B
24dtype: bfloat16
25
