Views
No views yet
1
2slices:
3 - sources:
4 - model: Qwen/Qwen2.5-Coder-1.5B-Instruct
5 layer_range: [0, 28]
6 weight: 0.6
7 - model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
8 layer_range: [0, 28]
9 weight: 0.4
10merge_method: slerp
11base_model: Qwen/Qwen2.5-Coder-1.5B-Instruct
12parameters:
13 t:
14 - filter: self_attn
15 value: [0, 0.5, 0.3, 0.7, 1]
16 - filter: mlp
17 value: [1, 0.5, 0.7, 0.3, 0]
18 - value: 0.5
19dtype: bfloat16
20