类型错误:预期的字节，找到描述符

Question

如何修复以下有关tf.record的错误回溯？

(tensorflow1) c:tensorflow1modelsresearchobject_detection>python generate_tfrecord.py --csv_input=data/train_labels.csv --output_path=data/train.record --image_dir=(image directory)
Traceback (most recent call last):
File "generate_tfrecord.py", line 17, in <module>
import tensorflow as tf
File "C:UsersDell-OguzAnaconda3envstensorflow1libsite-packagestensorflow__init__.py", line 24, in <module>
from tensorflow.python import pywrap_tensorflow  # pylint: disable=unused-import
File "C:UsersDell-OguzAnaconda3envstensorflow1libsite-packagestensorflowpython__init__.py", line 52, in <module>
from tensorflow.core.framework.graph_pb2 import *
File "C:UsersDell-OguzAnaconda3envstensorflow1libsite-packagestensorflowcoreframeworkgraph_pb2.py", line 15, in <module>
from tensorflow.core.framework import node_def_pb2 as tensorflow_dot_core_dot_framework_dot_node__def__pb2
File "C:UsersDell-OguzAnaconda3envstensorflow1libsite-packagestensorflowcoreframeworknode_def_pb2.py", line 15, in <module>
from tensorflow.core.framework import attr_value_pb2 as tensorflow_dot_core_dot_framework_dot_attr__value__pb2
File "C:UsersDell-OguzAnaconda3envstensorflow1libsite-packagestensorflowcoreframeworkattr_value_pb2.py", line 15, in <module>
from tensorflow.core.framework import tensor_pb2 as tensorflow_dot_core_dot_framework_dot_tensor__pb2
File "C:UsersDell-OguzAnaconda3envstensorflow1libsite-packagestensorflowcoreframeworktensor_pb2.py", line 15, in <module>
from tensorflow.core.framework import resource_handle_pb2 as tensorflow_dot_core_dot_framework_dot_resource__handle__pb2
File "C:UsersDell-OguzAnaconda3envstensorflow1libsite-packagestensorflowcoreframeworkresource_handle_pb2.py", line 91, in <module>
__module__ = 'tensorflow.core.framework.resource_handle_pb2'
TypeError: expected bytes, Descriptor found

Answer 1

试试这个，为我工作

pip install protobuf-py3
pip install --upgrade protobuf

Answer 2

当我在现有的tenforflow上安装带有conda install -c anaconda tensorflow-gpu的张量流时，我遇到了同样的错误。

删除现有环境并创建一个新环境。然后运行conda install -c anaconda tensorflow-gpu.

Answer 3

可能是由于不匹配的张量流和张量板版本。

我在为我的代码创建一个单独的 conda 环境后解决了它。

Answer 4

就我而言，

pip install --ignore-installed --upgrade tensorflow-gpu==1.15.0

做了这项工作:)。