MCPcopy Create free account
hub / github.com/GanjinZero/BioBART / parse_args

Function parse_args

downstream_src/summarization.py:70–282  ·  view source on GitHub ↗
()

Source from the content-addressed store, hash-verified

68
69
70def parse_args():
71 parser = argparse.ArgumentParser(description="Finetune a transformers model on a summarization task")
72 parser.add_argument(
73 "--dataset_name",
74 type=str,
75 default=None,
76 help="The name of the dataset to use (via the datasets library).",
77 )
78 parser.add_argument(
79 "--dataset_config_name",
80 type=str,
81 default=None,
82 help="The configuration name of the dataset to use (via the datasets library).",
83 )
84 parser.add_argument(
85 "--train_file", type=str, default=None, help="A csv or a json file containing the training data."
86 )
87 parser.add_argument(
88 "--test_file", type=str, default=None, help="A csv or a json file containing the training data."
89 )
90 parser.add_argument(
91 "--validation_file", type=str, default=None, help="A csv or a json file containing the validation data."
92 )
93 parser.add_argument(
94 "--ignore_pad_token_for_loss",
95 type=bool,
96 default=True,
97 help="Whether to ignore the tokens corresponding to " "padded labels in the loss computation or not.",
98 )
99 parser.add_argument(
100 "--max_source_length",
101 type=int,
102 default=1024,
103 help="The maximum total input sequence length after "
104 "tokenization.Sequences longer than this will be truncated, sequences shorter will be padded.",
105 )
106 parser.add_argument(
107 "--source_prefix",
108 type=str,
109 default=None,
110 help="A prefix to add before every source text " "(useful for T5 models).",
111 )
112 parser.add_argument(
113 "--preprocessing_num_workers",
114 type=int,
115 default=None,
116 help="The number of processes to use for the preprocessing.",
117 )
118 parser.add_argument(
119 "--overwrite_cache", type=bool, default=None, help="Overwrite the cached training and evaluation sets"
120 )
121 parser.add_argument(
122 "--max_target_length",
123 type=int,
124 default=128,
125 help="The maximum total sequence length for target text after "
126 "tokenization. Sequences longer than this will be truncated, sequences shorter will be padded."
127 "during ``evaluate`` and ``predict``.",

Callers 1

mainFunction · 0.70

Calls

no outgoing calls

Tested by

no test coverage detected