Views
No views yet
1base_model: models/saiga_nemo_12b_sft_m9_d16_simpo_m21_d36_doestoevsky_orpo_m1
2dtype: bfloat16
3merge_method: slerp
4parameters:
5 t:
6 - filter: self_attn
7 value: [0.0, 0.5, 0.3, 0.7, 1.0]
8 - filter: mlp
9 value: [1.0, 0.5, 0.7, 0.3, 0.0]
10 - value: 0.5
11slices:
12- sources:
13 - layer_range: [0, 40]
14 model: models/saiga_nemo_12b_sft_m9_d16_simpo_m21_d36
15 - layer_range: [0, 40]
16 model: models/saiga_nemo_12b_sft_m9_d16_simpo_m21_d36_doestoevsky_orpo_m1
17