Commit
•
5ff3ddb
1
Parent(s):
a1e85ba
Adding Evaluation Results (#1)
Browse files- Adding Evaluation Results (e29cf23d36c1f15b0e6b801a26baaaec42430105)
Co-authored-by: Open LLM Leaderboard PR Bot <[email protected]>
README.md
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
| 3 |
+
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_dhmeltzer__llama-7b-SFT_ds_wiki65k_1024_r_64_alpha_16_merged)
|
| 4 |
+
|
| 5 |
+
| Metric | Value |
|
| 6 |
+
|-----------------------|---------------------------|
|
| 7 |
+
| Avg. | 42.74 |
|
| 8 |
+
| ARC (25-shot) | 54.35 |
|
| 9 |
+
| HellaSwag (10-shot) | 78.06 |
|
| 10 |
+
| MMLU (5-shot) | 45.35 |
|
| 11 |
+
| TruthfulQA (0-shot) | 37.11 |
|
| 12 |
+
| Winogrande (5-shot) | 73.4 |
|
| 13 |
+
| GSM8K (5-shot) | 4.62 |
|
| 14 |
+
| DROP (3-shot) | 6.28 |
|