未验证 提交 836e1e0b 编写于 作者: Q Qiyang Min 提交者: GitHub

Merge pull request #12910 from velconia/fix_flowers_dataset

Fix flowers dataset download problem
...@@ -19,6 +19,7 @@ import hashlib ...@@ -19,6 +19,7 @@ import hashlib
import os import os
import errno import errno
import shutil import shutil
import six
import sys import sys
import importlib import importlib
import paddle.dataset import paddle.dataset
...@@ -94,6 +95,8 @@ def download(url, module_name, md5sum, save_name=None): ...@@ -94,6 +95,8 @@ def download(url, module_name, md5sum, save_name=None):
dl = 0 dl = 0
total_length = int(total_length) total_length = int(total_length)
for data in r.iter_content(chunk_size=4096): for data in r.iter_content(chunk_size=4096):
if six.PY2:
data = six.b(data)
dl += len(data) dl += len(data)
f.write(data) f.write(data)
done = int(50 * dl / total_length) done = int(50 * dl / total_length)
......
...@@ -35,6 +35,7 @@ import itertools ...@@ -35,6 +35,7 @@ import itertools
import functools import functools
from .common import download from .common import download
import tarfile import tarfile
import six
import scipy.io as scio import scipy.io as scio
from paddle.dataset.image import * from paddle.dataset.image import *
from paddle.reader import * from paddle.reader import *
...@@ -45,10 +46,10 @@ from six.moves import cPickle as pickle ...@@ -45,10 +46,10 @@ from six.moves import cPickle as pickle
from six.moves import zip from six.moves import zip
__all__ = ['train', 'test', 'valid'] __all__ = ['train', 'test', 'valid']
DATA_URL = 'http://www.robots.ox.ac.uk/~vgg/data/flowers/102/102flowers.tgz' DATA_URL = 'http://paddlemodels.cdn.bcebos.com/flowers/102flowers.tgz'
LABEL_URL = 'http://www.robots.ox.ac.uk/~vgg/data/flowers/102/imagelabels.mat' LABEL_URL = 'http://paddlemodels.cdn.bcebos.com/flowers/imagelabels.mat'
SETID_URL = 'http://www.robots.ox.ac.uk/~vgg/data/flowers/102/setid.mat' SETID_URL = 'http://paddlemodels.cdn.bcebos.com/flowers/setid.mat'
DATA_MD5 = '33bfc11892f1e405ca193ae9a9f2a118' DATA_MD5 = '52808999861908f626f3c1f4e79d11fa'
LABEL_MD5 = 'e0620be6f572b9609742df49c70aed4d' LABEL_MD5 = 'e0620be6f572b9609742df49c70aed4d'
SETID_MD5 = 'a5357ecc9cb78c4bef273ce3793fc85c' SETID_MD5 = 'a5357ecc9cb78c4bef273ce3793fc85c'
# In official 'readme', tstid is the flag of test data # In official 'readme', tstid is the flag of test data
...@@ -120,7 +121,10 @@ def reader_creator(data_file, ...@@ -120,7 +121,10 @@ def reader_creator(data_file,
file = file.strip() file = file.strip()
batch = None batch = None
with open(file, 'rb') as f: with open(file, 'rb') as f:
batch = pickle.load(f) if six.PY2:
batch = pickle.load(f)
else:
batch = pickle.load(f, encoding='bytes')
data = batch['data'] data = batch['data']
labels = batch['label'] labels = batch['label']
for sample, label in zip(data, batch['label']): for sample, label in zip(data, batch['label']):
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册