FutureMa burtenshaw HF Staff commited on
Commit
4b91247
·
1 Parent(s): d93be1c

Set evaluation_framework to inspect-ai in eval.yaml (#6)

Browse files

- Set evaluation_framework to inspect-ai in eval.yaml (de55fa4beda1d067cfbb64cf57f6ca8b3448b532)


Co-authored-by: ben burtenshaw <burtenshaw@users.noreply.huggingface.co>

Files changed (1) hide show
  1. eval.yaml +3 -1
eval.yaml CHANGED
@@ -3,6 +3,8 @@ description: >
3
  EvasionBench is a benchmark dataset for detecting evasive answers in earnings
4
  call Q&A sessions. The task is to classify how directly corporate management
5
  addresses questions from financial analysts.
 
 
6
  tasks:
7
  - id: evasion_bench
8
  config: default
@@ -35,4 +37,4 @@ tasks:
35
 
36
  {choices}
37
  scorers:
38
- - name: choice
 
3
  EvasionBench is a benchmark dataset for detecting evasive answers in earnings
4
  call Q&A sessions. The task is to classify how directly corporate management
5
  addresses questions from financial analysts.
6
+ evaluation_framework: inspect-ai
7
+
8
  tasks:
9
  - id: evasion_bench
10
  config: default
 
37
 
38
  {choices}
39
  scorers:
40
+ - name: choice