Method stop_recording

tensorflow/python/eager/backprop.py:864–893

Temporarily stops recording operations on this tape. Operations executed while this context manager is active will not be recorded on the tape. This is useful for reducing the memory used by tracing all computations. For example: ``` with tf.GradientTape(persistent=True)

(self)

  @tf_contextlib.contextmanager
  def stop_recording(self):
    """Temporarily stops recording operations on this tape.

    Operations executed while this context manager is active will not be
    recorded on the tape. This is useful for reducing the memory used by tracing
    all computations.

    For example:

    ```
      with tf.GradientTape(persistent=True) as t:
        loss = compute_loss(model)
        with t.stop_recording():
          # The gradient computation below is not traced, saving memory.
          grads = t.gradient(loss, model.variables)
    ```

    Yields:
      None
    Raises:
      RuntimeError: if the tape is not currently recording.
    """
    if self._tape is None:
      raise RuntimeError(
          "Trying to stop recording a tape which is not recording.")
    self._pop_tape()
    try:
      yield
    finally:
      self._push_tape()

  def reset(self):
    """Clears all information stored in this tape.