This is a merge of pre-trained language models created using
mergekit.
This model was merged using the
SLERP merge method.
1base_model: YOYO-AI/Qwen2.5-14B-1M-YOYO-V3
2dtype: bfloat16
3merge_method: slerp
4parameters:
5 t:
6 - filter: self_attn
7 value: [0.7, 0.6, 0.5, 0.4, 0.3] #
8 - filter: mlp
9 value: [0.3, 0.4, 0.5, 0.6, 0.7]
10 - filter: input_layernorm|post_attention_layernorm
11 value: 0.6
12 - value: 0.5
13slices:
14- sources:
15 - layer_range: [0, 48]
16 model: YOYO-AI/Qwen2.5-14B-1M-YOYO-V3
17 - layer_range: [0, 48]
18 model: Lunzima/NQLSG-Qwen2.5-14B-MegaFusion-v8.7
19
