Tensorflow InvalidArgumentError: You must feed a value for placeholder tensor

Question

I come across this error when I run the following code, trying to get prediction on test data. By printing out operation names, I see no operation named Placeholder_4, and can see an op named prediction_output

with tf.Session() as sess:
    imported_meta = tf.train.import_meta_graph("test-model-220.meta")
    imported_meta.restore(sess, tf.train.latest_checkpoint('./'))

    graph = tf.get_default_graph()

    for op in graph.get_operations():
        print(op.name)

    batch_size = tf.placeholder(tf.int64)
    a, b = tf.placeholder(tf.float32, shape=[None, len(X)]), tf.placeholder(tf.float32, shape=[None, 8])

    train_dataset = tf.data.Dataset.from_tensor_slices((a, b)).batch(batch_size).repeat()
    test_dataset = tf.data.Dataset.from_tensor_slices((a, b)).batch(batch_size)

    iterator = tf.data.Iterator.from_structure(train_dataset.output_types, train_dataset.output_shapes)

    features, labels = iterator.get_next()

    # create the initialisation operations
    training_init_op = iterator.make_initializer(train_dataset)
    test_init_op = iterator.make_initializer(test_dataset)
    # a = tf.placeholder(tf.float32, shape=[None, len(X)])

    sess.run(test_init_op)

    pred = graph.get_tensor_by_name("prediction_output:0")
    prediction = sess.run(pred, feed_dict={a: X, b: [[1, 1, 1, 1, 1, 1, 1, 1]], batch_size: 1})

    # print(prediction)

Stack trace of error:

Traceback (most recent call last):
  File "C:\Users\akshar.b\PycharmProjects\hello\venv\lib\site-packages\tensorflow\python\client\session.py", line 1322, in _do_call
    return fn(*args)
  File "C:\Users\akshar.b\PycharmProjects\hello\venv\lib\site-packages\tensorflow\python\client\session.py", line 1307, in _run_fn
    options, feed_dict, fetch_list, target_list, run_metadata)
  File "C:\Users\akshar.b\PycharmProjects\hello\venv\lib\site-packages\tensorflow\python\client\session.py", line 1409, in _call_tf_sessionrun
    run_metadata)
tensorflow.python.framework.errors_impl.InvalidArgumentError: You must feed a value for placeholder tensor 'Placeholder_4' with dtype float and shape [?,1]
     [[Node: Placeholder_4 = Placeholder[dtype=DT_FLOAT, shape=[?,1], _device="/job:localhost/replica:0/task:0/device:CPU:0"]()]]

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "C:/Users/akshar.b/pythonfiles/Extract_Coordinates/restore.py", line 70, in <module>
    sess.run(test_init_op)
  File "C:\Users\akshar.b\PycharmProjects\hello\venv\lib\site-packages\tensorflow\python\client\session.py", line 900, in run
    run_metadata_ptr)
  File "C:\Users\akshar.b\PycharmProjects\hello\venv\lib\site-packages\tensorflow\python\client\session.py", line 1135, in _run
    feed_dict_tensor, options, run_metadata)
  File "C:\Users\akshar.b\PycharmProjects\hello\venv\lib\site-packages\tensorflow\python\client\session.py", line 1316, in _do_run
    run_metadata)
  File "C:\Users\akshar.b\PycharmProjects\hello\venv\lib\site-packages\tensorflow\python\client\session.py", line 1335, in _do_call
    raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.InvalidArgumentError: You must feed a value for placeholder tensor 'Placeholder_4' with dtype float and shape [?,1]
     [[Node: Placeholder_4 = Placeholder[dtype=DT_FLOAT, shape=[?,1], _device="/job:localhost/replica:0/task:0/device:CPU:0"]()]]

Caused by op 'Placeholder_4', defined at:
  File "C:/Users/akshar.b/pythonfiles/Extract_Coordinates/restore.py", line 56, in <module>
    a, b = tf.placeholder(tf.float32, shape=[None, len(X)]), tf.placeholder(tf.float32, shape=[None, 8])
  File "C:\Users\akshar.b\PycharmProjects\hello\venv\lib\site-packages\tensorflow\python\ops\array_ops.py", line 1808, in placeholder
    return gen_array_ops.placeholder(dtype=dtype, shape=shape, name=name)
  File "C:\Users\akshar.b\PycharmProjects\hello\venv\lib\site-packages\tensorflow\python\ops\gen_array_ops.py", line 5835, in placeholder
    "Placeholder", dtype=dtype, shape=shape, name=name)
  File "C:\Users\akshar.b\PycharmProjects\hello\venv\lib\site-packages\tensorflow\python\framework\op_def_library.py", line 787, in _apply_op_helper
    op_def=op_def)
  File "C:\Users\akshar.b\PycharmProjects\hello\venv\lib\site-packages\tensorflow\python\framework\ops.py", line 3392, in create_op
    op_def=op_def)
  File "C:\Users\akshar.b\PycharmProjects\hello\venv\lib\site-packages\tensorflow\python\framework\ops.py", line 1718, in __init__
    self._traceback = self._graph._extract_stack()  # pylint: disable=protected-access

InvalidArgumentError (see above for traceback): You must feed a value for placeholder tensor 'Placeholder_4' with dtype float and shape [?,1]
     [[Node: Placeholder_4 = Placeholder[dtype=DT_FLOAT, shape=[?,1], _device="/job:localhost/replica:0/task:0/device:CPU:0"]()]]

Answer 1

You're getting a placeholder exception because you are not feeding a concrete value to place_holder. It looks like a is not holding a concrete value when passed to feed_dict.

Answer 2

A few mistakes that I noticed in your code:

• a, b = tf.placeholder(tf.float32, shape=[None, len(X)]), tf.placeholder(tf.float32, shape=[None, 8]). Python does not support assignments of a, b= c, d form.
• I am positively sure you are feeding wrong dimensions for X in the placeholder.
• I do not think creating a placeholder within a session is a good idea, since it is a part of the computational graph and must be present when the graph is created i.e. before the session is even run.