This is a merge of pre-trained language models created using
mergekit.
This model was merged using the Passthrough merge method.
1
2merge_method: passthrough
3dtype: bfloat16
4tokenizer_source: EpistemeAI/ReasoningCore-Llama-3.2-3B-r1-v1_2
5
6#base_model: huihui-ai/Hermes-3-Llama-3.2-3B-abliterated
7
8slices:
9 - sources:
10 - model: NeuraLakeAi/iSA-03-Mini-3B-Hybrid-Preview
11 layer_range: [0, 16]
12 parameters:
13 weight: 0.7
14
15 - sources:
16 - model: merge
17 layer_range: [12, 22]
18 parameters:
19 weight: 0.65
20
21 - sources:
22 - model: EpistemeAI/ReasoningCore-Llama-3.2-3B-r1-v1_2
23 layer_range: [18, 28]
24 parameters:
25 weight: 0.8
26
27
