mirror of
https://github.com/open-compass/opencompass.git
synced 2025-05-30 16:03:24 +08:00
20 lines
751 B
Python
20 lines
751 B
Python
|
from ..utils.function_utils import compute_rouge
|
||
|
|
|
||
|
|
#法条记忆问答
|
||
|
|
def compute_ftcs(data_dict):
|
||
|
|
"""
|
||
|
|
Compute the ROUGE-L score between the prediction and the reference
|
||
|
|
"""
|
||
|
|
references, predictions = [], []
|
||
|
|
for example in data_dict:
|
||
|
|
question, prediction, answer = example["origin_prompt"], example["prediction"], example["refr"]
|
||
|
|
answer = answer.replace("答案:", "")
|
||
|
|
predictions.append(prediction)
|
||
|
|
references.append(answer)
|
||
|
|
|
||
|
|
# compute the accuracy of score_list
|
||
|
|
rouge_scores = compute_rouge(predictions, references)
|
||
|
|
rouge_ls = [score["rouge-l"]["f"] for score in rouge_scores]
|
||
|
|
average_rouge_l = sum(rouge_ls) / len(rouge_ls)
|
||
|
|
return {"score": average_rouge_l}
|