Welcome to mirror list, hosted at ThFree Co, Russian Federation.

github.com/stanfordnlp/stanza.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat (limited to 'stanza/utils/datasets/ner/prepare_ner_dataset.py')
-rw-r--r--stanza/utils/datasets/ner/prepare_ner_dataset.py5
1 files changed, 2 insertions, 3 deletions
diff --git a/stanza/utils/datasets/ner/prepare_ner_dataset.py b/stanza/utils/datasets/ner/prepare_ner_dataset.py
index 9d7e089a..54a2c7e3 100644
--- a/stanza/utils/datasets/ner/prepare_ner_dataset.py
+++ b/stanza/utils/datasets/ner/prepare_ner_dataset.py
@@ -316,10 +316,9 @@ def process_bsnlp(paths, short_name):
output_filename = os.path.join(base_output_path, '%s.%s.json' % (short_name, shard))
prepare_ner_file.process_dataset(csv_file, output_filename)
-def main():
+def main(dataset_name):
paths = default_paths.get_default_paths()
- dataset_name = sys.argv[1]
random.seed(1234)
if dataset_name == 'fi_turku':
@@ -344,4 +343,4 @@ def main():
raise ValueError(f"dataset {dataset_name} currently not handled")
if __name__ == '__main__':
- main()
+ main(sys.argv[1])