*   `tf.data`

    *   `tf.data.Dataset.zip` now supports Python-style zipping, i.e. `Dataset.zip(a, b, c)`.
    * `tf.data.Dataset.shuffle` now supports `tf.data.UNKNOWN_CARDINALITY` When doing a "full shuffle" using  `dataset = dataset.shuffle(dataset.cardinality())`. But remember, a "full shuffle" will load the full dataset into memory so that it can be shuffled, so make sure to only use this with small datasets or datasets of small objects (like filenames).

		{"(*Rand).Int63n", Method, 0},
		{"(*Rand).Intn", Method, 0},
		{"(*Rand).NormFloat64", Method, 0},
		{"(*Rand).Perm", Method, 0},
		{"(*Rand).Read", Method, 6},
		{"(*Rand).Seed", Method, 0},
		{"(*Rand).Shuffle", Method, 10},
		{"(*Rand).Uint32", Method, 0},
		{"(*Rand).Uint64", Method, 8},
		{"(*Zipf).Uint64", Method, 0},
		{"ExpFloat64", Func, 0},
		{"Float32", Func, 0},
		{"Float64", Func, 0},

      return ArraysAreCastCompatible(inferred, actual);
    }
  }];
}

def TF_ConjugateTransposeOp : TF_Op<"ConjugateTranspose", [Pure]> {
  let summary = [{
Shuffle dimensions of x according to a permutation and conjugate the result.
  }];

  let description = [{
The output `y` has the same rank as `x`. The shapes of `x` and `y` satisfy: