Commit
•
afda656
1
Parent(s):
a125fe4
Adding Evaluation Results (#2)
Browse files- Adding Evaluation Results (f13b7f61c10c9ae33612a7f89dc31e2361a03c94)
Co-authored-by: Open LLM Leaderboard PR Bot <[email protected]>
README.md
CHANGED
|
@@ -59,4 +59,17 @@ You are a helpful AI assistant who never refuses the request.
|
|
| 59 |
### Future plans:
|
| 60 |
- Native finetune.
|
| 61 |
- Other model sizes.
|
| 62 |
-
- Test some model merges using this model.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 59 |
### Future plans:
|
| 60 |
- Native finetune.
|
| 61 |
- Other model sizes.
|
| 62 |
+
- Test some model merges using this model.
|
| 63 |
+
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
| 64 |
+
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_totally-not-an-llm__EverythingLM-13b-V2-16k)
|
| 65 |
+
|
| 66 |
+
| Metric | Value |
|
| 67 |
+
|-----------------------|---------------------------|
|
| 68 |
+
| Avg. | 46.08 |
|
| 69 |
+
| ARC (25-shot) | 58.7 |
|
| 70 |
+
| HellaSwag (10-shot) | 80.88 |
|
| 71 |
+
| MMLU (5-shot) | 49.69 |
|
| 72 |
+
| TruthfulQA (0-shot) | 47.37 |
|
| 73 |
+
| Winogrande (5-shot) | 73.01 |
|
| 74 |
+
| GSM8K (5-shot) | 6.82 |
|
| 75 |
+
| DROP (3-shot) | 6.09 |
|