Adding Evaluation Results
#40
by leaderboard-pr-bot - opened
README.md
CHANGED
|
@@ -131,4 +131,17 @@ The 175B model was trained on 992 *80GB A100 GPUs*. The training duration was ro
|
|
| 131 |
archivePrefix={arXiv},
|
| 132 |
primaryClass={cs.CL}
|
| 133 |
}
|
| 134 |
-
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 131 |
archivePrefix={arXiv},
|
| 132 |
primaryClass={cs.CL}
|
| 133 |
}
|
| 134 |
+
```
|
| 135 |
+
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
| 136 |
+
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_facebook__opt-125m)
|
| 137 |
+
|
| 138 |
+
| Metric | Value |
|
| 139 |
+
|-----------------------|---------------------------|
|
| 140 |
+
| Avg. | 25.47 |
|
| 141 |
+
| ARC (25-shot) | 22.87 |
|
| 142 |
+
| HellaSwag (10-shot) | 31.47 |
|
| 143 |
+
| MMLU (5-shot) | 26.02 |
|
| 144 |
+
| TruthfulQA (0-shot) | 42.87 |
|
| 145 |
+
| Winogrande (5-shot) | 51.62 |
|
| 146 |
+
| GSM8K (5-shot) | 0.08 |
|
| 147 |
+
| DROP (3-shot) | 3.36 |
|