Adding Evaluation Results
#6
by
leaderboard-pr-bot
- opened
README.md
CHANGED
|
@@ -96,4 +96,17 @@ Beluga is a new technology that carries risks with use. Testing conducted to dat
|
|
| 96 |
archivePrefix={arXiv},
|
| 97 |
primaryClass={cs.CL}
|
| 98 |
}
|
| 99 |
-
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 96 |
archivePrefix={arXiv},
|
| 97 |
primaryClass={cs.CL}
|
| 98 |
}
|
| 99 |
+
```
|
| 100 |
+
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
| 101 |
+
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_stabilityai__StableBeluga-13B)
|
| 102 |
+
|
| 103 |
+
| Metric | Value |
|
| 104 |
+
|-----------------------|---------------------------|
|
| 105 |
+
| Avg. | 51.94 |
|
| 106 |
+
| ARC (25-shot) | 62.03 |
|
| 107 |
+
| HellaSwag (10-shot) | 82.27 |
|
| 108 |
+
| MMLU (5-shot) | 57.71 |
|
| 109 |
+
| TruthfulQA (0-shot) | 49.61 |
|
| 110 |
+
| Winogrande (5-shot) | 76.87 |
|
| 111 |
+
| GSM8K (5-shot) | 13.8 |
|
| 112 |
+
| DROP (3-shot) | 21.26 |
|