Update README.md
Browse files
README.md
CHANGED
|
@@ -4,10 +4,23 @@ language:
|
|
| 4 |
- en
|
| 5 |
---
|
| 6 |
|
| 7 |
-
A generalist / roleplaying model merge based on Llama 3.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
|
| 9 |
---
|
| 10 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
Mergekit Config
|
| 12 |
```
|
| 13 |
models:
|
|
|
|
| 4 |
- en
|
| 5 |
---
|
| 6 |
|
| 7 |
+
A generalist / roleplaying model merge based on Llama 3. Models are selected from my personal experience while using them.
|
| 8 |
+
|
| 9 |
+
I personally think this is an improvement over Stheno v3.2, considering the other models helped balance out its creativity and at the same time improving its logic.
|
| 10 |
+
|
| 11 |
+
Settings:
|
| 12 |
+
```
|
| 13 |
+
Instruct // Context Template: Llama-3-Instruct
|
| 14 |
+
Temperature: 1.4
|
| 15 |
+
min_p: 0.1
|
| 16 |
+
```
|
| 17 |
|
| 18 |
---
|
| 19 |
|
| 20 |
+
Merging seems to be a black box magic though? In my personal experience merging multiple models from different datasets / data works better than combining them all in one.
|
| 21 |
+
|
| 22 |
+
*Values chosen are from long-running personal experimentation since Llama-2 Merging Era. I have tweaked them to fit this recipe.*
|
| 23 |
+
|
| 24 |
Mergekit Config
|
| 25 |
```
|
| 26 |
models:
|