Views
No views yet
1models:
2 - model: FelixChao/WestSeverus-7B-DPO-v2
3 # No parameters necessary for base model
4 - model: FelixChao/WestSeverus-7B-DPO-v2
5 parameters:
6 density: [1, 0.7, 0.1]
7 weight: [0, 0.3, 0.7, 1]
8 - model: CultriX/Wernicke-7B-v9
9 parameters:
10 density: [1, 0.7, 0.3]
11 weight: [0, 0.25, 0.5, 1]
12 - model: mlabonne/NeuralBeagle14-7B
13 parameters:
14 density: 0.25
15 weight:
16 - filter: mlp
17 value: 0.5
18 - value: 0
19merge_method: ties
20base_model: FelixChao/WestSeverus-7B-DPO-v2
21parameters:
22 int8_mask: true
23 normalize: true
24 sparsify:
25 - filter: mlp
26 value: 0.5
27 - filter: self_attn
28 value: 0.5
29dtype: float16
30| Metric | Value |
|---|---|
| Avg. | 75.36 |
| AI2 Reasoning Challenge (25-Shot) | 73.38 |
| HellaSwag (10-Shot) | 88.50 |
| MMLU (5-Shot) | 64.94 |
| TruthfulQA (0-shot) | 71.50 |
| Winogrande (5-shot) | 83.58 |
| GSM8k (5-shot) | 70.28 |