This is a merge of pre-trained language models created using
mergekit.
This model was merged using the
TIES merge method using
meta-llama/Llama-3.2-1B as a base.
1base_model: meta-llama/Llama-3.2-1B
2dtype: float16
3merge_method: ties
4modules:
5 default:
6 slices:
7 - sources:
8 - layer_range: [0, 16]
9 model: Alelcv27/llama3-1b-math-dpo
10 parameters:
11 density: 0.5
12 weight: 0.8
13 - layer_range: [0, 16]
14 model: Alelcv27/llama3-1b-code-dpo
15 parameters:
16 density: 0.3
17 weight: 0.5
18 - layer_range: [0, 16]
19 model: meta-llama/Llama-3.2-1B
20parameters:
21 normalize: 1.0
