experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Fix bugs in docstring of conv3d_transpose.

Browse Source 
Seems like depth was missing and NCHW and NHWC were mentioned
instead of NCDHW and NDHWC.

PiperOrigin-RevId: 313320657
Change-Id: Id599e6e7b91a18f193b952153c50c27839c5693f
This commit is contained in:
Andrew Selle 2020-05-26 20:51:52 -07:00 committed by TensorFlower Gardener
parent 0dda89c61e
commit 0a1449a983
1 changed files with 6 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
tensorflow/python/ops/nn_ops.py
View File

@ -3061,12 +3061,12 @@ def conv3d_transpose_v2(input, # pylint: disable=redefined-builtin
rather than an actual deconvolution. rather than an actual deconvolution.
Args: Args:
input: A 5-D `Tensor` of type `float` and shape `[batch, height, width, input: A 5-D `Tensor` of type `float` and shape `[batch, depth, height,
in_channels]` for `NHWC` data format or `[batch, in_channels, height, width, in_channels]` for `NDHWC` data format or `[batch, in_channels,
width]` for `NCHW` data format. depth, height, width]` for `NCDHW` data format.
filters: A 5-D `Tensor` with the same type as `value` and shape `[height, filters: A 5-D `Tensor` with the same type as `value` and shape `[depth,
width, output_channels, in_channels]`. `filter`'s `in_channels` dimension height, width, output_channels, in_channels]`. `filter`'s `in_channels`
must match that of `value`. dimension must match that of `value`.
output_shape: A 1-D `Tensor` representing the output shape of the output_shape: A 1-D `Tensor` representing the output shape of the
deconvolution op. deconvolution op.
strides: An int or list of `ints` that has length `1`, `3` or `5`. The strides: An int or list of `ints` that has length `1`, `3` or `5`. The