| base_model: YOYO-AI/Qwen2.5-14B-1M-YOYO-V3 | |
| dtype: bfloat16 | |
| merge_method: slerp | |
| parameters: | |
| t: | |
| - filter: self_attn | |
| value: [0.7, 0.6, 0.5, 0.4, 0.3] # | |
| - filter: mlp | |
| value: [0.3, 0.4, 0.5, 0.6, 0.7] | |
| - filter: input_layernorm|post_attention_layernorm | |
| value: 0.6 | |
| - value: 0.5 | |
| slices: | |
| - sources: | |
| - layer_range: [0, 48] | |
| model: YOYO-AI/Qwen2.5-14B-1M-YOYO-V3 | |
| - layer_range: [0, 48] | |
| model: Lunzima/NQLSG-Qwen2.5-14B-MegaFusion-v8.7 | |