Adding Evaluation Results
#2
by
leaderboard-pr-bot
- opened
README.md
CHANGED
|
@@ -135,3 +135,17 @@ Qualitative evaluation suggests that the evol-instruct-70k fine-tuned Galactica
|
|
| 135 |
primaryClass={cs.CL}
|
| 136 |
}
|
| 137 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 135 |
primaryClass={cs.CL}
|
| 136 |
}
|
| 137 |
```
|
| 138 |
+
|
| 139 |
+
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
| 140 |
+
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_GeorgiaTechResearchInstitute__galactica-6.7b-evol-instruct-70k)
|
| 141 |
+
|
| 142 |
+
| Metric | Value |
|
| 143 |
+
|-----------------------|---------------------------|
|
| 144 |
+
| Avg. | 33.68 |
|
| 145 |
+
| ARC (25-shot) | 42.58 |
|
| 146 |
+
| HellaSwag (10-shot) | 49.3 |
|
| 147 |
+
| MMLU (5-shot) | 32.96 |
|
| 148 |
+
| TruthfulQA (0-shot) | 42.1 |
|
| 149 |
+
| Winogrande (5-shot) | 56.27 |
|
| 150 |
+
| GSM8K (5-shot) | 0.38 |
|
| 151 |
+
| DROP (3-shot) | 12.16 |
|