tf.compat.v1.flags.mark_flags_as_required
Ensures that flags are not None during program execution.
tf.compat.v1.flags.mark_flags_as_required(
flag_names, flag_values=_flagvalues.FLAGS
)
Recommended usage:
if name == 'main': flags.mark_flags_as_required(['flag1', 'flag2', 'flag3']) app.run()
| Args | |
|---|---|
flag_names |
Sequence[str], names of the flags. |
flag_values |
flags.FlagValues, optional FlagValues instance where the flags are defined. |
| Raises | |
|---|---|
AttributeError |
If any of flag name has not already been defined as a flag. |
© 2020 The TensorFlow Authors. All rights reserved.
Licensed under the Creative Commons Attribution License 3.0.
Code samples licensed under the Apache 2.0 License.
https://www.tensorflow.org/versions/r2.3/api_docs/python/tf/compat/v1/flags/mark_flags_as_required