This is a merge of pre-trained language models created using
mergekit.
This model was merged using the
DELLA merge method using
marcuscedricridia/Springer-32B as a base.
1merge_method: della
2base_model: marcuscedricridia/Springer-32B
3models:
4 - model: TIGER-Lab/Qwen2.5-32B-Instruct-CFT
5 parameters:
6 density: 1
7 weight: 1
8 lambda: 0.9
9parameters:
10 density: 1
11 weight: 1
12 lambda: 0.9
13 normalize: true
14 int8_mask: true
15dtype: bfloat16
16tokenizer_source: base
17name: Springer-32B-19
18
