Update README.md
Browse files
README.md
CHANGED
@@ -22,8 +22,6 @@ license_link: LICENSE
|
|
22 |
|
23 |
Same idea as previous merge, but with [saishf's](https://huggingface.co/saishf) SOVL merge line.
|
24 |
|
25 |
-
Neko-Maid-SLERP is using the same config, using [saishf/SOVLish-Maid-L3-8B](https://huggingface.co/saishf/SOVLish-Maid-L3-8B) and [saishf/Kitty-Cat-SOVL-8B-L3-V1](https://huggingface.co/saishf/Kitty-Cat-SOVL-8B-L3-V1) as base.
|
26 |
-
|
27 |
---
|
28 |
|
29 |
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
@@ -60,3 +58,19 @@ parameters:
|
|
60 |
dtype: bfloat16
|
61 |
|
62 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
22 |
|
23 |
Same idea as previous merge, but with [saishf's](https://huggingface.co/saishf) SOVL merge line.
|
24 |
|
|
|
|
|
25 |
---
|
26 |
|
27 |
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
|
|
58 |
dtype: bfloat16
|
59 |
|
60 |
```
|
61 |
+
|
62 |
+
```yaml
|
63 |
+
slices:
|
64 |
+
- sources:
|
65 |
+
- model: saishf/Kitty-Cat-SOVL-8B-L3-V1
|
66 |
+
layer_range: [0, 32]
|
67 |
+
- model: saishf/SOVLish-Maid-L3-8B
|
68 |
+
layer_range: [0, 32]
|
69 |
+
merge_method: slerp
|
70 |
+
base_model: saishf/Kitty-Cat-SOVL-8B-L3-V1
|
71 |
+
parameters:
|
72 |
+
t:
|
73 |
+
- value: [0.55, 0.15, 0.55, 0.15, 0.55, 0.15, 0.55, 0.15, 0.35, 0.15, 0.55, 0.15, 0.55, 0.15, 0.55, 0.15, 0.55]
|
74 |
+
dtype: bfloat16
|
75 |
+
|
76 |
+
```
|