python - tf.data.Dataset + tf.lookup.index_table_from_file 导致 "Table not initialized"错误

Question

我使用的是 TensorFlow 版本 1.8.0。尽管使用了 make_initialized_iterator() 并运行了 iterator.initializer，但当我尝试在映射到tf.data.Dataset。这是一个崩溃的简化示例:

import tensorflow as tf

words = tf.contrib.lookup.index_table_from_file("words.txt", num_oov_buckets=1, key_column_index=0)

sentences = tf.data.TextLineDataset("sentences.txt")
sentences = sentences.map(lambda string: tf.string_split([string]).values)
dataset = sentences.map(lambda tokens: words.lookup(tokens))

iterator = dataset.make_initializable_iterator()
next_element = iterator.get_next()

tf.add_to_collection(tf.GraphKeys.TABLE_INITIALIZERS, iterator.initializer)

with tf.Session() as sess:
  for i in range(2):
    sess.run(iterator.initializer)
    print(sess.run(next_element))

这是错误:

2018-07-06 10:33:23.371736: W tensorflow/core/framework/op_kernel.cc:1318] OP_REQUIRES failed at lookup_table_op.cc:675 : Failed precondition: Table not initialized.
Traceback (most recent call last):
  File "/anaconda2/envs/py36/lib/python3.6/site-packages/tensorflow/python/client/session.py", line 1322, in _do_call
    return fn(*args)
  File "/anaconda2/envs/py36/lib/python3.6/site-packages/tensorflow/python/client/session.py", line 1307, in _run_fn
    options, feed_dict, fetch_list, target_list, run_metadata)
  File "/anaconda2/envs/py36/lib/python3.6/site-packages/tensorflow/python/client/session.py", line 1409, in _call_tf_sessionrun
    run_metadata)
tensorflow.python.framework.errors_impl.FailedPreconditionError: Table not initialized.
     [[Node: string_to_index_Lookup/hash_table_Lookup = LookupTableFindV2[Tin=DT_STRING, Tout=DT_INT64](string_to_index_lookup_hash_table_lookup_placeholder, arg0, string_to_index_lookup_hash_table_lookup_placeholder_1)]]
     [[Node: IteratorGetNext = IteratorGetNext[output_shapes=[[?]], output_types=[DT_INT64], _device="/job:localhost/replica:0/task:0/device:CPU:0"](Iterator)]]

Answer 1

您必须添加sess.run(tf.tables_initializer())添加with tf.Session()-block的开头