Commit
•
0fa07e7
1
Parent(s):
8edbcc4
Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- merge
|
4 |
+
- mergekit
|
5 |
+
- GGUF
|
6 |
+
---
|
7 |
+
# neural-Kunoichi2-7B-slerp-GGUF quantized version
|
8 |
+
this is the quantized version of [neural-Kunoichi2-7B-slerp](https://huggingface.co/seyf1elislam/neural-Kunoichi2-7B-slerp)
|
9 |
+
|
10 |
+
## [neural-Kunoichi2-7B-slerp](https://huggingface.co/seyf1elislam/neural-Kunoichi2-7B-slerp)
|
11 |
+
|
12 |
+
neural-Kunoichi2-7B-slerp is a merge of the following models using LazyMergekit:
|
13 |
+
* [SanjiWatsuki/Kunoichi-DPO-v2-7B](https://huggingface.co/SanjiWatsuki/Kunoichi-DPO-v2-7B)
|
14 |
+
* [mlabonne/NeuralPipe-7B-ties](https://huggingface.co/mlabonne/NeuralPipe-7B-ties)
|
15 |
+
|
16 |
+
## 🧩 Configuration
|
17 |
+
|
18 |
+
```yaml
|
19 |
+
merge_method: slerp
|
20 |
+
base_model: SanjiWatsuki/Kunoichi-DPO-v2-7B
|
21 |
+
slices:
|
22 |
+
- sources:
|
23 |
+
- model: SanjiWatsuki/Kunoichi-DPO-v2-7B
|
24 |
+
layer_range: [0, 32]
|
25 |
+
- model: mlabonne/NeuralPipe-7B-ties
|
26 |
+
layer_range: [0, 32]
|
27 |
+
parameters:
|
28 |
+
t:
|
29 |
+
- filter: self_attn
|
30 |
+
value: [0, 0.5, 0.3, 0.7, 1]
|
31 |
+
- filter: mlp
|
32 |
+
value: [1, 0.5, 0.7, 0.3, 0]
|
33 |
+
- value: 0.5
|
34 |
+
dtype: bfloat16
|
35 |
+
```
|