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

Merge pull request #18565 from alexpantyukhin/ganhead_constructor_validate

Browse Source 
add checking for input values in GANHead constructor
This commit is contained in:
Martin Wicke 2018-06-27 11:58:44 -07:00 committed by GitHub
commit 85cb8d48b7
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 12 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
tensorflow/contrib/gan/python/estimator/python/head_impl.py
View File

@ -103,9 +103,20 @@ class GANHead(head._Head): # pylint: disable=protected-access
name: name of the head. If provided, summary and metrics keys will be
suffixed by `"/" + name`.
"""
if not callable(generator_loss_fn):
raise TypeError('generator_loss_fn must be callable.')
if not callable(discriminator_loss_fn):
raise TypeError('discriminator_loss_fn must be callable.')
if not use_loss_summaries in [True, False, None]:
raise ValueError('use_loss_summaries must be True, False or None.')
if get_hooks_fn is not None and not callable(get_hooks_fn):
raise TypeError('get_hooks_fn must be callable.')
if name is not None and not isinstance(name, str):
raise TypeError('name must be string.')
if get_hooks_fn is None:
get_hooks_fn = tfgan_train.get_sequential_train_hooks()
# TODO(joelshor): Validate inputs.
if use_loss_summaries in [True, False]:
generator_loss_fn = functools.partial(