Remove vw files

This commit is contained in:
Iwona Christop 2022-05-28 16:51:52 +02:00
parent 483fb9837c
commit 4ac818e536
8 changed files with 3 additions and 1861730 deletions

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

Binary file not shown.

View File

@ -15,7 +15,7 @@ def generateChooChoo(data, target, categories):
def predictFuture(test): def predictFuture(test):
data = [x[2].replace('\n', '').replace(':', '') for x in io.read(test + '/in.tsv')] data = [x[2].replace('\n', '').replace(':', '') for x in io.read(test + '/in.tsv')]
with open('vw_ireland_news_' + test, 'w', encoding='utf-8') as f: with open('vw-' + test, 'w', encoding='utf-8') as f:
for text in data: for text in data:
f.write('1 |text ' + text + '\n') f.write('1 |text ' + text + '\n')
@ -24,11 +24,11 @@ if __name__ == '__main__':
ireland_news_train, categories = generateChooChoo('train/in.tsv.xz', 'train/expected.tsv.xz', categories={}) ireland_news_train, categories = generateChooChoo('train/in.tsv.xz', 'train/expected.tsv.xz', categories={})
ireland_news_dev, _ = generateChooChoo('dev-0/in.tsv', 'dev-0/expected.tsv', categories) ireland_news_dev, _ = generateChooChoo('dev-0/in.tsv', 'dev-0/expected.tsv', categories)
with open('vw_ireland_news_train', 'w', encoding='utf-8') as f: with open('vw-train', 'w', encoding='utf-8') as f:
for target, text in zip(ireland_news_train['target'], ireland_news_train['data']): for target, text in zip(ireland_news_train['target'], ireland_news_train['data']):
f.write(str(categories[target] + 1) + ' |text ' + text + '\n') f.write(str(categories[target] + 1) + ' |text ' + text + '\n')
with open('vw_ireland_news_dev0', 'w', encoding='utf-8') as f, open('ireland_news_dev0_targets', 'w', encoding='utf-8') as f_targets: with open('vw-dev0', 'w', encoding='utf-8') as f, open('vw-dev0-targets', 'w', encoding='utf-8') as f_targets:
for target, text in zip(ireland_news_dev['target'], ireland_news_dev['data']): for target, text in zip(ireland_news_dev['target'], ireland_news_dev['data']):
f.write('1 |text ' + text + '\n') f.write('1 |text ' + text + '\n')
f_targets.write(str(categories[target] + 1) + '\n') f_targets.write(str(categories[target] + 1) + '\n')

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff