Datasets:
mahirlabibdihan
commited on
Update README.md
Browse files
README.md
CHANGED
@@ -13,4 +13,36 @@ configs:
|
|
13 |
path: dataset.json
|
14 |
paperswithcode_id: mapeval-api
|
15 |
---
|
16 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
13 |
path: dataset.json
|
14 |
paperswithcode_id: mapeval-api
|
15 |
---
|
16 |
+
|
17 |
+
# MapEval-API
|
18 |
+
|
19 |
+
[MapEval](https://arxiv.org/abs/2501.00316)-API is created using [MapQaTor](https://arxiv.org/abs/2412.21015).
|
20 |
+
|
21 |
+
# Usage
|
22 |
+
|
23 |
+
```python
|
24 |
+
from datasets import load_dataset
|
25 |
+
|
26 |
+
# Load dataset
|
27 |
+
ds = load_dataset("MapEval/MapEval-API", name="benchmark")
|
28 |
+
|
29 |
+
# Generate better prompts
|
30 |
+
for item in ds["test"]:
|
31 |
+
# Start with a clear task description
|
32 |
+
prompt = (
|
33 |
+
"You are a highly intelligent assistant. "
|
34 |
+
"Answer the multiple-choice question by selecting the correct option.\n\n"
|
35 |
+
"Question:\n" + item["question"] + "\n\n"
|
36 |
+
"Options:\n"
|
37 |
+
)
|
38 |
+
|
39 |
+
# List the options more clearly
|
40 |
+
for i, option in enumerate(item["options"], start=1):
|
41 |
+
prompt += f"{i}. {option}\n"
|
42 |
+
|
43 |
+
# Add a concluding sentence to encourage selection of the answer
|
44 |
+
prompt += "\nSelect the best option by choosing its number."
|
45 |
+
|
46 |
+
# Use the prompt as needed
|
47 |
+
print(prompt) # Replace with your processing logic
|
48 |
+
```
|