boundaries.

*   XLA compilation:

    *   `tf.function(experimental_compile=True)` has become a stable API,
        renamed `tf.function(jit_compile=True)`.
    *   XLA can now compile MirroredStrategy: the step function passed
        to`strategy.run` can now be annoted with `jit_compile=True`.

*   `tf.distribute`:

    *   Rename `experimental_prefetch_to_device` in `tf.distribute.InputOptions`

{{/* 2 extra rounds for 256-bit keys. */}} Lenc256: {{- template "load_keys" (load_keys_args 0 2) }} {{- template "enc" (enc_args 0 $N true) }} {{- template "enc" (enc_args 1 $N true) }} {{/* 2 extra rounds for 192-bit keys. */}} Lenc192: {{- template "load_keys" (load_keys_args 2 2) }} {{- template "enc" (enc_args 2 $N true) }} {{- template "enc" (enc_args 3 $N true) }} {{/* 10 rounds for 128-bit (with special handling for final). */}} Lenc128: {{- template "load_keys" (load_keys_args 4 11) }} {{-...