tf.raw_ops.DataFormatVecPermute  |  TensorFlow v2.16.1 (original) (raw)

tf.raw_ops.DataFormatVecPermute

Stay organized with collections Save and categorize content based on your preferences.

Permute input tensor from src_format to dst_format.

View aliases

Compat aliases for migration

SeeMigration guide for more details.

tf.compat.v1.raw_ops.DataFormatVecPermute

tf.raw_ops.DataFormatVecPermute(
    x, src_format='NHWC', dst_format='NCHW', name=None
)

Given source and destination format strings of length n=4 or 5, the input tensor must be a vector of size n or n-2, or a 2D tensor of shape (n, 2) or (n-2, 2).

If the first dimension of the input tensor is n-2, it is assumed that non-spatial dimensions are omitted (i.e N, C).

For example, with src_format of NHWC, dst_format of NCHW, and input:

[1, 2, 3, 4]

, the output will be:

[1, 4, 2, 3]

With src_format of NDHWC, dst_format of NCDHW, and input:

[[1, 6], [2, 7], [3, 8], [4, 9], [5, 10]]

, the output will be:

[[1, 6], [5, 10], [2, 7], [3, 8], [4, 9]]

With src_format of NHWC, dst_format of NCHW, and input:

[1, 2]

, the output will be:

[1, 2]