unit_test / data /preprocess /cc12m /cc12m_train_download_list.py
herrius's picture
Upload 259 files
32b542e
raw
history blame contribute delete
441 Bytes
import os
captions = []
urls = []
with open('cc12m.tsv') as fp:
for cnt, line in enumerate(fp):
s = line.split('\t')
captions.append(s[0].split(' '))
urls.append(s[1][:-1])
with open('train4download.txt', 'w') as fp:
for cnt, url in enumerate(urls):
fp.write("../train_image/{:08d}.jpg\t\"{}\"\n".format(cnt, url))
if not os.path.exists('../train_image'):
os.makedirs('../train_image')