Adding Evaluation Results
#9
by
leaderboard-pr-bot
- opened
README.md
CHANGED
|
@@ -45,4 +45,17 @@ A preliminary evaluation of the model quality is conducted by creating a set of
|
|
| 45 |
|
| 46 |
## Major updates of weights v1.1
|
| 47 |
- Refactor the tokenization and separator. In Vicuna v1.1, the separator has been changed from `"###"` to the EOS token `"</s>"`. This change makes it easier to determine the generation stop criteria and enables better compatibility with other libraries.
|
| 48 |
-
- Fix the supervised fine-tuning loss computation for better model quality.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 45 |
|
| 46 |
## Major updates of weights v1.1
|
| 47 |
- Refactor the tokenization and separator. In Vicuna v1.1, the separator has been changed from `"###"` to the EOS token `"</s>"`. This change makes it easier to determine the generation stop criteria and enables better compatibility with other libraries.
|
| 48 |
+
- Fix the supervised fine-tuning loss computation for better model quality.
|
| 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_eachadea__vicuna-13b-1.1)
|
| 51 |
+
|
| 52 |
+
| Metric | Value |
|
| 53 |
+
|-----------------------|---------------------------|
|
| 54 |
+
| Avg. | 47.0 |
|
| 55 |
+
| ARC (25-shot) | 52.73 |
|
| 56 |
+
| HellaSwag (10-shot) | 80.13 |
|
| 57 |
+
| MMLU (5-shot) | 51.94 |
|
| 58 |
+
| TruthfulQA (0-shot) | 52.08 |
|
| 59 |
+
| Winogrande (5-shot) | 74.19 |
|
| 60 |
+
| GSM8K (5-shot) | 8.64 |
|
| 61 |
+
| DROP (3-shot) | 9.31 |
|