Commit
•
baf1d07
1
Parent(s):
e9a972b
Adding Evaluation Results (#8)
Browse files- Adding Evaluation Results (a7b533b3ba5d76c3fb0e8b6a526964cf1de16d83)
Co-authored-by: Open LLM Leaderboard PR Bot <leaderboard-pr-bot@users.noreply.huggingface.co>
README.md
CHANGED
|
@@ -45,3 +45,17 @@ python train.py configs/llama2_7b_chat_uncensored.yaml
|
|
| 45 |
|
| 46 |
# Fine-tuning guide
|
| 47 |
https://georgesung.github.io/ai/qlora-ift/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 45 |
|
| 46 |
# Fine-tuning guide
|
| 47 |
https://georgesung.github.io/ai/qlora-ift/
|
| 48 |
+
|
| 49 |
+
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
| 50 |
+
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_georgesung__llama2_7b_chat_uncensored)
|
| 51 |
+
|
| 52 |
+
| Metric | Value |
|
| 53 |
+
|-----------------------|---------------------------|
|
| 54 |
+
| Avg. | 43.39 |
|
| 55 |
+
| ARC (25-shot) | 53.58 |
|
| 56 |
+
| HellaSwag (10-shot) | 78.66 |
|
| 57 |
+
| MMLU (5-shot) | 44.49 |
|
| 58 |
+
| TruthfulQA (0-shot) | 41.34 |
|
| 59 |
+
| Winogrande (5-shot) | 74.11 |
|
| 60 |
+
| GSM8K (5-shot) | 5.84 |
|
| 61 |
+
| DROP (3-shot) | 5.69 |
|