def download_file(self, embeddings_folder, logger):
logger.info('Downloading word embeddings file for "{}"...'.format(
self.alias))
out_path = os.path.join(embeddings_folder, self.embeddings_filename)
if os.path.exists(out_path):
logger.info('Already downloaded -> skipping!')
else:
url = '{}/{}'.format(self.base_url, self.embeddings_filename)
r = requests.get(url, stream=True)
total_size = int(
r.headers.get('content-length', self.approximate_filesize)
) # size of the embeddings file (bytes)
chunk_size = 4 * 1024 * 1024 # 4 MB
bar = DataTransferBar(max_value=total_size).start()
completed_bytes = 0
try:
with open(out_path, 'wb') as f:
for chunk in r.iter_content(chunk_size=chunk_size):
if chunk:
f.write(chunk)
completed_bytes += len(chunk)
if completed_bytes > bar.max_value:
bar.max_value = completed_bytes
bar.update(completed_bytes)
except BaseException as e:
os.unlink(out_path)
raise e
bar.finish()
logger.info('Done!')
