Update evaluation results via RewardBench
Browse files
README.md
CHANGED
|
@@ -11,6 +11,38 @@ model-index:
|
|
| 11 |
metrics:
|
| 12 |
- type: accuracy
|
| 13 |
value: 0.5343383584589615
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14 |
- task:
|
| 15 |
type: preference_evaluation
|
| 16 |
dataset:
|
|
|
|
| 11 |
metrics:
|
| 12 |
- type: accuracy
|
| 13 |
value: 0.5343383584589615
|
| 14 |
+
- task:
|
| 15 |
+
type: preference_evaluation
|
| 16 |
+
dataset:
|
| 17 |
+
name: Chat
|
| 18 |
+
type: Chat
|
| 19 |
+
metrics:
|
| 20 |
+
- type: accuracy
|
| 21 |
+
value: 0.8128491620111732
|
| 22 |
+
- task:
|
| 23 |
+
type: preference_evaluation
|
| 24 |
+
dataset:
|
| 25 |
+
name: Chat Hard
|
| 26 |
+
type: Chat_Hard
|
| 27 |
+
metrics:
|
| 28 |
+
- type: accuracy
|
| 29 |
+
value: 0.5263157894736842
|
| 30 |
+
- task:
|
| 31 |
+
type: preference_evaluation
|
| 32 |
+
dataset:
|
| 33 |
+
name: Safety
|
| 34 |
+
type: Safety
|
| 35 |
+
metrics:
|
| 36 |
+
- type: accuracy
|
| 37 |
+
value: 0.4851351351351351
|
| 38 |
+
- task:
|
| 39 |
+
type: preference_evaluation
|
| 40 |
+
dataset:
|
| 41 |
+
name: Reasoning
|
| 42 |
+
type: Reasoning
|
| 43 |
+
metrics:
|
| 44 |
+
- type: accuracy
|
| 45 |
+
value: 0.3930266819446718
|
| 46 |
- task:
|
| 47 |
type: preference_evaluation
|
| 48 |
dataset:
|