This is a merge of pre-trained language models created using
mergekit.
This model was merged using the
Task Arithmetic merge method using
prithivMLmods/QwQ-LCoT2-7B-Instruct as a base.
1models:
2 - model: happzy2633/qwen2.5-7b-ins-v3
3 parameters:
4 density: [1, 0.7, 0.1] # density gradient
5 weight: 1.0
6 - model: prithivMLmods/QwQ-LCoT2-7B-Instruct
7 parameters:
8 density: 0.5
9 weight: [0, 0.3, 0.7, 1] # weight gradient
10 - model: happzy2633/qwen2.5-7b-ins-v3
11 parameters:
12 density: 0.33
13 weight:
14 - filter: mlp
15 value: 0.5
16 - value: 0
17merge_method: task_arithmetic
18base_model: prithivMLmods/QwQ-LCoT2-7B-Instruct
19parameters:
20 normalize: true
21 int8_mask: true
22dtype: float16
23
