tf.lookup.experimental.DatasetInitializer

View source on GitHub

Creates a table initializer from a tf.data.Dataset.

tf.lookup.experimental.DatasetInitializer(
    dataset
)

Sample usage:

  keys = tf.data.Dataset.range(100)
  values = tf.data.Dataset.range(100).map(
      lambda x: string_ops.as_string(x * 2))
  ds = tf.data.Dataset.zip((keys, values))
  init = tf.lookup.experimental.DatasetInitializer(ds)
  table = tf.lookup.StaticHashTable(init, "")
  output = table.lookup([0, 1, 2])
  assertEquals(outputs, ["0", "2", "4"])

Raises: ValueError if dataset doesn't conform to specifications.

dataset A tf.data.Dataset object that produces tuples of scalars. The first scalar is treated as a key and the second as value.

dataset A tf.data.Dataset object that produces tuples of scalars. The first scalar is treated as a key and the second as value.
key_dtype The expected table key dtype.
value_dtype The expected table value dtype.

Methods

initialize

View source

initialize(
    table
)

Returns the table initialization op.