qasper.yaml 1.13 KB
Newer Older
Baber Abbasi's avatar
Baber Abbasi committed
1
2
3
4
5
6
7

tag:
  - longbench
task: longbench_qasper
dataset_path: THUDM/LongBench
test_split: test
dataset_name: qasper
Timur Aysin's avatar
Timur Aysin committed
8
doc_to_text: "You are given a scientific article and a question. Answer the question as concisely as you can, using a single phrase or sentence if possible. If the question cannot be answered based on the information in the article, write \"unanswerable\". If the question is a yes/no question, answer \"yes\", \"no\", or \"unanswerable\". Do not provide any explanation.\n\nArticle: {{context}}\n\n Answer the question based on the above article as concisely as you can, using a single phrase or sentence if possible. If the question cannot be answered based on the information in the article, write \"unanswerable\". If the question is a yes/no question, answer \"yes\", \"no\", or \"unanswerable\". Do not provide any explanation.\n\nQuestion: {{input}}\n\nAnswer:"
9
10
doc_to_target: '{{answers}}'
process_results: !function metrics.get_qa_f1_score
Baber Abbasi's avatar
Baber Abbasi committed
11
12
13
generation_kwargs:
  max_gen_toks: 128
  temperature: 1
Timur Aysin's avatar
Timur Aysin committed
14
  do_sample: False
Baber Abbasi's avatar
Baber Abbasi committed
15
  until: []
Baber Abbasi's avatar
Baber Abbasi committed
16
metric_list:
17
  - metric: "qa_f1_score"
Baber Abbasi's avatar
Baber Abbasi committed
18
19
20
    aggregation: mean
    higher_is_better: True
metadata:
Timur Aysin's avatar
Timur Aysin committed
21
  version: 4.0