Sebastien Montella
commited on
Commit
•
053885b
1
Parent(s):
af9b304
json to jsonl
Browse files- train_val_test_split.py +31 -12
train_val_test_split.py
CHANGED
@@ -35,10 +35,29 @@ def save_json(data, path):
|
|
35 |
None
|
36 |
|
37 |
"""
|
38 |
-
|
39 |
json.dump(data, open(path, 'w+', encoding='utf-8'))
|
40 |
|
41 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
42 |
def get_types(xml):
|
43 |
""" Extract the different transformation types.
|
44 |
|
@@ -300,28 +319,28 @@ def split(args):
|
|
300 |
if args.split_criteria == 'random':
|
301 |
train, val, test = random_split(pairs, args.training_ratio)
|
302 |
os.makedirs(f'{args.out_dir}/{version}/random_split/', exist_ok=True)
|
303 |
-
|
304 |
-
|
305 |
-
|
306 |
|
307 |
elif args.split_criteria == 'transformations':
|
308 |
seen_transformations_train, seen_transformations_val, seen_transformations_test, unseen_transformations_test = challenge_seen_unseen_transformation_split(pairs_by_transformation, args.training_ratio)
|
309 |
|
310 |
os.makedirs(f'{args.out_dir}/{version}/transformations_split/', exist_ok=True)
|
311 |
-
|
312 |
-
|
313 |
-
|
314 |
-
|
315 |
|
316 |
elif args.split_criteria == 'source_dataset':
|
317 |
|
318 |
itwiki_train, itwiki_val, itwiki_test, tn_test = challenge_seen_unseen_source_dataset_split(pairs_by_source_dataset, args.training_ratio)
|
319 |
|
320 |
os.makedirs(f'{args.out_dir}/{version}/source_dataset_split/', exist_ok=True)
|
321 |
-
|
322 |
-
|
323 |
-
|
324 |
-
|
325 |
|
326 |
|
327 |
|
|
|
35 |
None
|
36 |
|
37 |
"""
|
38 |
+
|
39 |
json.dump(data, open(path, 'w+', encoding='utf-8'))
|
40 |
|
41 |
|
42 |
+
def save_jsonl(data, path):
|
43 |
+
""" Save data as a json file where each line is a dictionary
|
44 |
+
|
45 |
+
Args:
|
46 |
+
data (:obj:`list`): Data to store as json.
|
47 |
+
path (:obj:`str`): Path where data will be stored.
|
48 |
+
|
49 |
+
Returns:
|
50 |
+
None
|
51 |
+
|
52 |
+
"""
|
53 |
+
|
54 |
+
|
55 |
+
with open(path, 'w+', encoding='utf-8') as f:
|
56 |
+
for sample_dict in data:
|
57 |
+
f.write(json.dumps(sample_dict))
|
58 |
+
f.write('\n')
|
59 |
+
|
60 |
+
|
61 |
def get_types(xml):
|
62 |
""" Extract the different transformation types.
|
63 |
|
|
|
319 |
if args.split_criteria == 'random':
|
320 |
train, val, test = random_split(pairs, args.training_ratio)
|
321 |
os.makedirs(f'{args.out_dir}/{version}/random_split/', exist_ok=True)
|
322 |
+
save_jsonl(train, f'{args.out_dir}/{version}/random_split/train.jsonl')
|
323 |
+
save_jsonl(val, f'{args.out_dir}/{version}/random_split/val.jsonl')
|
324 |
+
save_jsonl(test, f'{args.out_dir}/{version}/random_split/test.jsonl')
|
325 |
|
326 |
elif args.split_criteria == 'transformations':
|
327 |
seen_transformations_train, seen_transformations_val, seen_transformations_test, unseen_transformations_test = challenge_seen_unseen_transformation_split(pairs_by_transformation, args.training_ratio)
|
328 |
|
329 |
os.makedirs(f'{args.out_dir}/{version}/transformations_split/', exist_ok=True)
|
330 |
+
save_jsonl(seen_transformations_train, f'{args.out_dir}/{version}/transformations_split/train.jsonl')
|
331 |
+
save_jsonl(seen_transformations_val, f'{args.out_dir}/{version}/transformations_split/val.jsonl')
|
332 |
+
save_jsonl(seen_transformations_test, f'{args.out_dir}/{version}/transformations_split/seen_transformations_test.jsonl')
|
333 |
+
save_jsonl(unseen_transformations_test, f'{args.out_dir}/{version}/transformations_split/unseen_transformations_test.jsonl')
|
334 |
|
335 |
elif args.split_criteria == 'source_dataset':
|
336 |
|
337 |
itwiki_train, itwiki_val, itwiki_test, tn_test = challenge_seen_unseen_source_dataset_split(pairs_by_source_dataset, args.training_ratio)
|
338 |
|
339 |
os.makedirs(f'{args.out_dir}/{version}/source_dataset_split/', exist_ok=True)
|
340 |
+
save_jsonl(itwiki_train, f'{args.out_dir}/{version}/source_dataset_split/itwiki_train.jsonl')
|
341 |
+
save_jsonl(itwiki_val, f'{args.out_dir}/{version}/source_dataset_split/itwiki_val.jsonl')
|
342 |
+
save_jsonl(itwiki_test, f'{args.out_dir}/{version}/source_dataset_split/itwiki_test.jsonl')
|
343 |
+
save_jsonl(tn_test, f'{args.out_dir}/{version}/source_dataset_split/tn_test.jsonl')
|
344 |
|
345 |
|
346 |
|