squeeze¶

dragon.vm.tensorflow.squeeze(
  input,
  axis=None,
  copy=True,
  name=None
)[source]¶

Remove the dimensions of input with size 1.

axis could be negative or None:

x = tf.ones((1, 2, 2, 1))

# Remove all matched dimensions if axis is None
# Otherwise, only the specified axes will be removed
print(tf.squeeze(x).shape)  # (1, 2, 2, 1) -> (2, 2)
print(tf.squeeze(x, axis=0).shape)  # (1, 2, 2, 1) -> (2, 2, 1)

# A negative axis is the last-k axis
print(tf.squeeze(x, axis=3).shape)  # (1, 2, 2, 1) -> (1, 2, 2)
print(tf.squeeze(x, axis=-1).shape)  # Equivalent

# Also, axis could be a sequence of integers
print(tf.squeeze(x, axis=[0, 3]).shape)  # (1, 2, 2, 1) -> (2, 2)

Parameters:

• input (dragon.Tensor) – The input tensor.
• axis (Union[int, Sequence[int]], optional) – The axis to remove.
• copy (bool, optional, default=True) – Return a new tensor or call in-place.
• name (str, optional) – The operation name.

Returns:

dragon.Tensor – The output tensor.