.size());

    // Almost force an overflow in the binomial coefficient calculation
    assertEquals(
        1391975640 /*C(34,14)*/,
        Collections2.orderedPermutations(concat(nCopies(20, 1), nCopies(14, 2))).size());
    // Do force an overflow in the binomial coefficient calculation
    assertEquals(
        Integer.MAX_VALUE,

    1 * 2 * 3 * 4 * 5 * 6 * 7 * 8 * 9 * 10 * 11,
    1 * 2 * 3 * 4 * 5 * 6 * 7 * 8 * 9 * 10 * 11 * 12
  };

  /**
   * Returns {@code n} choose {@code k}, also known as the binomial coefficient of {@code n} and
   * {@code k}, or {@link Integer#MAX_VALUE} if the result does not fit in an {@code int}.
   *
   * @throws IllegalArgumentException if {@code n < 0}, {@code k < 0} or {@code k > n}
   */

    1L * 2 * 3 * 4 * 5 * 6 * 7 * 8 * 9 * 10 * 11 * 12 * 13 * 14 * 15 * 16 * 17 * 18 * 19 * 20
  };

  /**
   * Returns {@code n} choose {@code k}, also known as the binomial coefficient of {@code n} and
   * {@code k}, or {@link Long#MAX_VALUE} if the result does not fit in a {@code long}.
   *
   * @throws IllegalArgumentException if {@code n < 0}, {@code k < 0}, or {@code k > n}
   */

        `tf.contrib.data.sloppy_interleave()`.
*   Add `train_and_evaluate` for simple distributed `Estimator` training.
*   Add `tf.spectral.dct` for computing the DCT-II.
*   Add Mel-Frequency Cepstral Coefficient support to `tf.contrib.signal` (with
    GPU and gradient support).
*   Add a self-check on `import tensorflow` for Windows DLL issues.
*   Add NCHW support to `tf.depth_to_space` on GPU.

JSON can only contain UTF-8 encoded strings, so it can't contain raw bytes.

Base64 can encode binary data in strings, but to do it, it needs to use more characters than the original binary data, so it would normally be less efficient than regular files.

Use base64 only if you definitely need to include binary data in JSON, and you can't use files for that.

## Pydantic `bytes` { #pydantic-bytes }

          @Override
          protected char @Nullable [] escape(int cp) {
            return ('a' <= cp && cp <= 'z') ? new char[] {Character.toUpperCase((char) cp)} : null;
          }

          // Inefficient implementation that defines all letters as escapable.
          @Override
          protected int nextEscapeIndex(CharSequence csq, int index, int end) {

        | ((source[offset + 1] & 0xFF) << 8)
        | ((source[offset + 2] & 0xFF) << 16)
        | ((source[offset + 3] & 0xFF) << 24);
  }

  /**
   * Indicates that the load and store operations will be very efficient because of use of VarHandle
   * or Unsafe. May be useful for calling code to fall back on an alternative implementation that is
   * slower than those implementations but faster than the pure-Java mask-and-shift.
   */

   * if and only if {@code stream} was efficiently splittable and its underlying spliterator
   * reported {@link Spliterator#SUBSIZED}. This is generally the case if the underlying stream
   * comes from a data structure supporting efficient indexed random access, typically an array or
   * list.
   *
   * <p>The order of the resulting stream is defined if and only if the order of the original stream
   * was defined.
   */

 * the return type. For this case, we provide a more expensive workaround in {@code
 * JdkFutureAdapters}. However, when possible, it is more efficient and reliable to create a {@code
 * ListenableFuture} directly.
 *
 * @author Sven Mawson
 * @author Nishant Thakkar
 * @since 1.0
 */
/*

   * if and only if {@code stream} was efficiently splittable and its underlying spliterator
   * reported {@link Spliterator#SUBSIZED}. This is generally the case if the underlying stream
   * comes from a data structure supporting efficient indexed random access, typically an array or
   * list.
   *
   * <p>The order of the resulting stream is defined if and only if the order of the original stream
   * was defined.
   */