Update README.md
Browse files
README.md
CHANGED
@@ -10,37 +10,12 @@ tags:
|
|
10 |
- rinna/japanese-gpt-1b
|
11 |
---
|
12 |
|
13 |
-
# jp-gpt-
|
14 |
|
15 |
jp-gpt-ab-dareties2 is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
16 |
* [oshizo/qa-refine-japanese-gpt-1b](https://huggingface.co/oshizo/qa-refine-japanese-gpt-1b)
|
17 |
* [rinna/japanese-gpt-1b](https://huggingface.co/rinna/japanese-gpt-1b)
|
18 |
|
19 |
-
## 🧩 Configuration
|
20 |
-
|
21 |
-
```yaml
|
22 |
-
slices:
|
23 |
-
- sources:
|
24 |
-
- layer_range: [0, 24]
|
25 |
-
model: oshizo/qa-refine-japanese-gpt-1b
|
26 |
-
parameters:
|
27 |
-
density: [1, 0.7, 0.1]
|
28 |
-
weight: 1.0
|
29 |
-
- layer_range: [0, 24]
|
30 |
-
model: rinna/japanese-gpt-1b
|
31 |
-
parameters:
|
32 |
-
density: 0.33
|
33 |
-
weight:
|
34 |
-
- filter: mlp
|
35 |
-
value: 0.5
|
36 |
-
- value: 0
|
37 |
-
merge_method: dare_ties
|
38 |
-
base_model: rinna/japanese-gpt-1b
|
39 |
-
parameters:
|
40 |
-
normalize: true
|
41 |
-
int8_mask: true
|
42 |
-
dtype: bfloat16
|
43 |
-
```
|
44 |
|
45 |
## 💻 Usage
|
46 |
|
|
|
10 |
- rinna/japanese-gpt-1b
|
11 |
---
|
12 |
|
13 |
+
# jp-gpt-1b-dareties2
|
14 |
|
15 |
jp-gpt-ab-dareties2 is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
16 |
* [oshizo/qa-refine-japanese-gpt-1b](https://huggingface.co/oshizo/qa-refine-japanese-gpt-1b)
|
17 |
* [rinna/japanese-gpt-1b](https://huggingface.co/rinna/japanese-gpt-1b)
|
18 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
|
20 |
## 💻 Usage
|
21 |
|