Make citations module-level constants

lm_eval/tasks/unscramble.py

@@ -7,7 +7,17 @@ involves giving the model a word distorted by some combination of scrambling,
 addition, or deletion of characters, and asking it to recover the original word.
 
 Homepage: https://github.com/openai/gpt-3/tree/master/data
+"""
+import gzip
+import json
+import shutil
+from pathlib import Path
+from best_download import download_file
+from lm_eval.base import Task, rf
+from lm_eval.metrics import mean
 
+
+_CITATION = """
 @inproceedings{NEURIPS2020_1457c0d6,
     author = {Brown, Tom and Mann, Benjamin and Ryder, Nick and Subbiah, Melanie and Kaplan, Jared D and Dhariwal, Prafulla and Neelakantan, Arvind and Shyam, Pranav and Sastry, Girish and Askell, Amanda and Agarwal, Sandhini and Herbert-Voss, Ariel and Krueger, Gretchen and Henighan, Tom and Child, Rewon and Ramesh, Aditya and Ziegler, Daniel and Wu, Jeffrey and Winter, Clemens and Hesse, Chris and Chen, Mark and Sigler, Eric and Litwin, Mateusz and Gray, Scott and Chess, Benjamin and Clark, Jack and Berner, Christopher and McCandlish, Sam and Radford, Alec and Sutskever, Ilya and Amodei, Dario},
     booktitle = {Advances in Neural Information Processing Systems},
@@ -20,13 +30,6 @@ Homepage: https://github.com/openai/gpt-3/tree/master/data
     year = {2020}
 }
 """
-import gzip
-import json
-import shutil
-from pathlib import Path
-from best_download import download_file
-from lm_eval.base import Task, rf
-from lm_eval.metrics import mean
 
 
 def extract_gzip(gz, to):

lm_eval/tasks/webqs.py

@@ -8,7 +8,13 @@ large knowledge graph. The questions are mostly centered around a single named e
 The questions are popular ones asked on the web (at least in 2013).
 
 Homepage: https://worksheets.codalab.org/worksheets/0xba659fe363cb46e7a505c5b6a774dc8a
+"""
+from . common import HFTask
+from lm_eval.base import rf
+from ..metrics import mean
 
+
+_CITATION = """
 @inproceedings{berant-etal-2013-semantic,
     title = "Semantic Parsing on {F}reebase from Question-Answer Pairs",
     author = "Berant, Jonathan  and
@@ -24,9 +30,6 @@ Homepage: https://worksheets.codalab.org/worksheets/0xba659fe363cb46e7a505c5b6a7
     pages = "1533--1544",
 }
 """
-from . common import HFTask
-from lm_eval.base import rf
-from ..metrics import mean
 
 
 class WebQs(HFTask):

lm_eval/tasks/wikitext.py

@@ -8,7 +8,15 @@ extracted from the set of verified Good and Featured articles on Wikipedia.
 NOTE: This `Task` is based on WikiText-2.
 
 Homepage: https://www.salesforce.com/products/einstein/ai-research/the-wikitext-dependency-language-modeling-dataset/
+"""
+import os
+import re
+from lm_eval.base import rf, PerplexityTask
+from lm_eval.utils import sh
+from best_download import download_file
+
 
+_CITATION = """
 @misc{merity2016pointer,
     title={Pointer Sentinel Mixture Models}, 
     author={Stephen Merity and Caiming Xiong and James Bradbury and Richard Socher},
@@ -18,12 +26,6 @@ Homepage: https://www.salesforce.com/products/einstein/ai-research/the-wikitext-
     primaryClass={cs.CL}
 }
 """
-import os
-import re
-from lm_eval.base import rf, PerplexityTask
-from lm_eval.utils import sh
-
-from best_download import download_file
 
 
 def wikitext_detokenizer(string):

lm_eval/tasks/winogrande.py

@@ -13,7 +13,14 @@ Trinh & Le in Simple Method for Commonsense Reasoning (2018).
 See: https://arxiv.org/abs/1806.02847
 
 Homepage: https://leaderboard.allenai.org/winogrande/submissions/public
+"""
+import numpy as np
+from . common import HFTask
+from lm_eval.base import rf
+from ..metrics import mean
 
+
+_CITATION = """
 @article{sakaguchi2019winogrande,
     title={WinoGrande: An Adversarial Winograd Schema Challenge at Scale},
     author={Sakaguchi, Keisuke and Bras, Ronan Le and Bhagavatula, Chandra and Choi, Yejin},
@@ -21,10 +28,6 @@ Homepage: https://leaderboard.allenai.org/winogrande/submissions/public
     year={2019}
 }
 """
-import numpy as np
-from . common import HFTask
-from lm_eval.base import rf
-from ..metrics import mean
 
 
 class Winogrande(HFTask):

lm_eval/tasks/wsc273.py

@@ -12,7 +12,15 @@ as described by Trinh & Le in Simple Method for Commonsense Reasoning (2018).
 See: https://arxiv.org/abs/1806.0
 
 Homepage: https://cs.nyu.edu/~davise/papers/WinogradSchemas/WS.html
+"""
+import numpy as np
+import random
+from lm_eval.base import rf
+from ..metrics import mean
+from . common import HFTask
 
+
+_CITATION = """
 @inproceedings{ea01b9c0db064caca6986b925d75f2bb,
     title = "The winograd schema challenge",
     abstract = "In this paper, we present an alternative to the Turing Test that has some conceptual and practical advantages. A Wino-grad schema is a pair of sentences that differ only in one or two words and that contain a referential ambiguity that is resolved in opposite directions in the two sentences. We have compiled a collection of Winograd schemas, designed so that the correct answer is obvious to the human reader, but cannot easily be found using selectional restrictions or statistical techniques over text corpora. A contestant in the Winograd Schema Challenge is presented with a collection of one sentence from each pair, and required to achieve human-level accuracy in choosing the correct disambiguation.",
@@ -27,11 +35,6 @@ Homepage: https://cs.nyu.edu/~davise/papers/WinogradSchemas/WS.html
     note = "13th International Conference on the Principles of Knowledge Representation and Reasoning, KR 2012 ; Conference date: 10-06-2012 Through 14-06-2012",
 }
 """
-import numpy as np
-import random
-from lm_eval.base import rf
-from ..metrics import mean
-from . common import HFTask
 
 
 class WinogradSchemaChallenge273(HFTask):