Views
No views yet
1dtype: bfloat16
2merge_method: passthrough
3slices:
4 - sources:
5 - layer_range: [0, 4]
6 model: allenai/OLMoE-1B-7B-0125-Instruct
7 - sources:
8 - layer_range: [4, 6]
9 model: allenai/OLMoE-1B-7B-0125-Instruct
10 - sources:
11 - layer_range: [4, 6]
12 model: allenai/OLMoE-1B-7B-0125-Instruct
13 parameters:
14 scale:
15 - filter: o_proj
16 value: 0.0
17 - filter: down_proj
18 value: 0.0
19 - value: 1.0
20 - sources:
21 - layer_range: [6, 8]
22 model: allenai/OLMoE-1B-7B-0125-Instruct
23 - sources:
24 - layer_range: [6, 8]
25 model: allenai/OLMoE-1B-7B-0125-Instruct
26 parameters:
27 scale:
28 - filter: o_proj
29 value: 0.0
30 - filter: down_proj
31 value: 0.0
32 - value: 1.0
33 - sources:
34 - layer_range: [8, 10]
35 model: allenai/OLMoE-1B-7B-0125-Instruct
36 - sources:
37 - layer_range: [8, 10]
38 model: allenai/OLMoE-1B-7B-0125-Instruct
39 parameters:
40 scale:
41 - filter: o_proj
42 value: 0.0
43 - filter: down_proj
44 value: 0.0
45 - value: 1.0
46 - sources:
47 - layer_range: [10, 12]
48 model: allenai/OLMoE-1B-7B-0125-Instruct
49 - sources:
50 - layer_range: [10, 12]
51 model: allenai/OLMoE-1B-7B-0125-Instruct
52 parameters:
53 scale:
54 - filter: o_proj
55 value: 0.0
56 - filter: down_proj
57 value: 0.0
58 - value: 1.0
59 - sources:
60 - layer_range: [12, 16]
61 model: allenai/OLMoE-1B-7B-0125-Instruct
62