cifar10_eval.py issue: cannot infer Tensor's rank

[zhisong]

I ran a unmodified version of cifar10_eval.py, here is the error message

Traceback (most recent call last):
File "cifar10_eval.py", line 157, in
tf.app.run()
File "/home/zhisong/tensorflow/tensorflow/python/platform/app.py", line 44, in run
_sys.exit(main(_sys.argv[:1] + flags_passthrough))
File "cifar10_eval.py", line 153, in main
evaluate()
File "cifar10_eval.py", line 121, in evaluate
images, labels = cifar10.inputs(eval_data=eval_data)
File "/home/zhisong/tensorflow/models/tutorials/image/cifar10/cifar10.py", line 183, in inputs
batch_size=FLAGS.batch_size)
File "/home/zhisong/tensorflow/models/tutorials/image/cifar10/cifar10_input.py", line 257, in inputs
shuffle=False)
File "/home/zhisong/tensorflow/models/tutorials/image/cifar10/cifar10_input.py", line 132, in _generate_image_and_label_batch
capacity=min_queue_examples + 3 * batch_size)
File "/home/zhisong/tensorflow/tensorflow/python/training/input.py", line 872, in batch
name=name)
File "/home/zhisong/tensorflow/tensorflow/python/training/input.py", line 655, in _batch
shapes = _shapes([tensor_list], shapes, enqueue_many)
File "/home/zhisong/tensorflow/tensorflow/python/training/input.py", line 598, in _shapes
raise ValueError("Cannot infer Tensor's rank: %s" % tl[i])
ValueError: Cannot infer Tensor's rank: Tensor("Cast:0", dtype=int32)

By comparing to distorted_inputs() in cifar10_input.py, I suspect the following lines are missing in inputs()

float_image.set_shape([height, width, 3])
read_input.label.set_shape([1])

After adding the above two lines, the script works just fine

[jart]

Thank you for bringing this to our attention. It looks like @nealwu forgot to update the inputs() function in cl/141805344.