Update README.md
Browse files
README.md
CHANGED
|
@@ -25,6 +25,21 @@ This is a merge of pre-trained language models created using [mergekit](https://
|
|
| 25 |
|
| 26 |
This model was merged using the [SCE](https://arxiv.org/abs/2408.07990) merge method using [KaraKaraWitch/Llama-3.X-Workout-70B](https://huggingface.co/KaraKaraWitch/Llama-3.X-Workout-70B) as a base.
|
| 27 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 28 |
### Models Merged
|
| 29 |
|
| 30 |
The following models were included in the merge:
|
|
|
|
| 25 |
|
| 26 |
This model was merged using the [SCE](https://arxiv.org/abs/2408.07990) merge method using [KaraKaraWitch/Llama-3.X-Workout-70B](https://huggingface.co/KaraKaraWitch/Llama-3.X-Workout-70B) as a base.
|
| 27 |
|
| 28 |
+
## UGI-Results
|
| 29 |
+
|
| 30 |
+
Hrrmm. As of 07/03/25, it's pretty low:
|
| 31 |
+
|
| 32 |
+
| Bench | Results |
|
| 33 |
+
|---------------|--------------------|
|
| 34 |
+
| UGI-Score | 38.83 / 100 |
|
| 35 |
+
| Unruly | 4.6 / 10 |
|
| 36 |
+
| Internet | 4.4 / 10 |
|
| 37 |
+
| Society | 4.6 / 10 |
|
| 38 |
+
| Willing | 3 / 10 |
|
| 39 |
+
| NatInt | 31.85 / 100 |
|
| 40 |
+
| Coding | 24 |
|
| 41 |
+
| Politial Lean | −8.7% (Liberalism) |
|
| 42 |
+
|
| 43 |
### Models Merged
|
| 44 |
|
| 45 |
The following models were included in the merge:
|