Question

训练初始模型后。在推断时我得到了这个错误。（我正在使用https://github.com/tensorflow/tensorflow/tree/master/tensorflow/examples/image_retraining）

TypeError：无法将feed_dict键解释为Tensor：名称 'DecodeJpeg / contents：0'指的是不存在的Tensor。该操作'DecodeJpeg / contents'，图中不存在。

omer@omer-HP-EliteBook-8540w:~/Downloads/tensorflow-
master/tensorflow/examples/image_retraining$ python label_image.py --graph=/tmp/output_graph.pb --labels=/tmp/output_labels.txt --image=/home/omer/Downloads/tensorflow-master/tensorflow/examples/image_retraining/pwd/glass/glass52.jpg --output_layer=final_result


2017-08-13 17:10:05.642446: W 
tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow 
library wasn't compiled to use SSE4.1 instructions, but these are 
available on your machine and could speed up CPU computations.
2017-08-13 17:10:05.642507: W 
tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow 
library wasn't compiled to use SSE4.2 instructions, but these are 
available on your machine and could speed up CPU computations.

Traceback (most recent call last):
  File "retraining-example.py", line 86, in <module>
run_inference_on_image()
  File "retraining-example.py", line 69, in run_inference_on_image
{'DecodeJpeg/contents:0': image_data})
  File "/home/omer/installed/acaconda2/lib/python2.7/site-
packages/tensorflow/python/client/session.py", line 778, in run
run_metadata_ptr)
File "/home/omer/installed/acaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 933, in _run
+ e.args[0])

TypeError: Cannot interpret feed_dict key as Tensor: The name 
'DecodeJpeg/contents:0' refers to a Tensor which does not exist. The 
operation, 'DecodeJpeg/contents', does not exist in the graph.

请帮助，我是这个主题的新手

Answer 1

Google已在retrain.py更新了他们的image_retraining。他们删除了op:DecodeJpeg，因此您无法获得它。

名称'DecodeJpeg / contents：0'指的是不存在的Tensor

1 个答案: