Adding Evaluation Results
#2
by leaderboard-pr-bot - opened
README.md
CHANGED
|
@@ -17,4 +17,17 @@ Javalion-R is a research artefact with dual purpose for entertainment as well as
|
|
| 17 |
|
| 18 |
Mileage mat vary. No refunds best wishes. Mainly intended to be utilized with Open Source KoboldAI software. Optimal sampler and settings not determined. Feedback Welcome!
|
| 19 |
|
| 20 |
-
https://github.com/KoboldAI/KoboldAI-Client
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
|
| 18 |
Mileage mat vary. No refunds best wishes. Mainly intended to be utilized with Open Source KoboldAI software. Optimal sampler and settings not determined. Feedback Welcome!
|
| 19 |
|
| 20 |
+
https://github.com/KoboldAI/KoboldAI-Client
|
| 21 |
+
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
| 22 |
+
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_digitous__Javalion-R)
|
| 23 |
+
|
| 24 |
+
| Metric | Value |
|
| 25 |
+
|-----------------------|---------------------------|
|
| 26 |
+
| Avg. | 35.42 |
|
| 27 |
+
| ARC (25-shot) | 41.72 |
|
| 28 |
+
| HellaSwag (10-shot) | 68.02 |
|
| 29 |
+
| MMLU (5-shot) | 30.81 |
|
| 30 |
+
| TruthfulQA (0-shot) | 34.44 |
|
| 31 |
+
| Winogrande (5-shot) | 65.43 |
|
| 32 |
+
| GSM8K (5-shot) | 2.65 |
|
| 33 |
+
| DROP (3-shot) | 4.85 |
|