Update README.md
Browse files
README.md
CHANGED
@@ -2,41 +2,10 @@
|
|
2 |
base_model:
|
3 |
- Quazim0t0/Alien-CoT-14B-sce
|
4 |
- Quazim0t0/SuperNova14b
|
5 |
-
tags:
|
6 |
-
- merge
|
7 |
-
- mergekit
|
8 |
-
- lazymergekit
|
9 |
-
- Quazim0t0/Alien-CoT-14B-sce
|
10 |
-
- Quazim0t0/SuperNova14b
|
11 |
---
|
12 |
|
13 |
# Dyson-14b
|
14 |
|
15 |
-
Dyson-14b is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
16 |
-
* [Quazim0t0/Alien-CoT-14B-sce](https://huggingface.co/Quazim0t0/Alien-CoT-14B-sce)
|
17 |
-
* [Quazim0t0/SuperNova14b](https://huggingface.co/Quazim0t0/SuperNova14b)
|
18 |
-
|
19 |
-
## 🧩 Configuration
|
20 |
-
|
21 |
-
```yaml
|
22 |
-
base_model: Quazim0t0/SuperNova14b
|
23 |
-
dtype: bfloat16
|
24 |
-
merge_method: slerp
|
25 |
-
parameters:
|
26 |
-
t:
|
27 |
-
- filter: self_attn
|
28 |
-
value: [0.0, 0.5, 0.3, 0.7, 1.0]
|
29 |
-
- filter: mlp
|
30 |
-
value: [1.0, 0.5, 0.7, 0.3, 0.0]
|
31 |
-
- value: 0.5
|
32 |
-
slices:
|
33 |
-
- sources:
|
34 |
-
- layer_range: [0, 40]
|
35 |
-
model: Quazim0t0/Alien-CoT-14B-sce
|
36 |
-
- layer_range: [0, 40]
|
37 |
-
model: Quazim0t0/SuperNova14b
|
38 |
-
|
39 |
-
```
|
40 |
|
41 |
## 💻 Usage
|
42 |
|
|
|
2 |
base_model:
|
3 |
- Quazim0t0/Alien-CoT-14B-sce
|
4 |
- Quazim0t0/SuperNova14b
|
|
|
|
|
|
|
|
|
|
|
|
|
5 |
---
|
6 |
|
7 |
# Dyson-14b
|
8 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
9 |
|
10 |
## 💻 Usage
|
11 |
|