Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
10K - 100K
ArXiv:
License:
Update eval.yaml
Browse files
eval.yaml
CHANGED
|
@@ -20,19 +20,20 @@ tasks:
|
|
| 20 |
|
| 21 |
solvers:
|
| 22 |
- name: prompt_template
|
| 23 |
-
|
| 24 |
-
|
|
|
|
| 25 |
|
| 26 |
-
|
| 27 |
- name: multiple_choice
|
| 28 |
-
|
| 29 |
-
|
| 30 |
-
|
|
|
|
| 31 |
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
{choices}
|
| 35 |
|
|
|
|
| 36 |
scorers:
|
| 37 |
- name: choice
|
| 38 |
|
|
|
|
| 20 |
|
| 21 |
solvers:
|
| 22 |
- name: prompt_template
|
| 23 |
+
args:
|
| 24 |
+
template: |
|
| 25 |
+
Question: {prompt}
|
| 26 |
|
| 27 |
+
Answer: {answer}
|
| 28 |
- name: multiple_choice
|
| 29 |
+
args:
|
| 30 |
+
template: |
|
| 31 |
+
You are a financial analyst. Your task is to Detect Evasive Answers in Financial Q&A.
|
| 32 |
+
The entire content of your response should be of the following format: 'ANSWER: $LETTER' (without quotes) where LETTER is one of {letters}.
|
| 33 |
|
| 34 |
+
{question}
|
|
|
|
|
|
|
| 35 |
|
| 36 |
+
{choices}
|
| 37 |
scorers:
|
| 38 |
- name: choice
|
| 39 |
|