RSVP for your your local TensorFlow Everywhere event today!


Ensures that flag is not None during program execution.

Registers a flag validator, which will follow usual validator rules. Important note: validator will pass for any non-None value, such as False, 0 (zero), '' (empty string) and so on.

It is recommended to call this method like this:

if name == 'main': flags.mark_flag_as_required('your_flag_name')

Because validation happens at we want to ensure required-ness is enforced at that time. You generally do not want to force users who import your code to have additional required flags for their own binaries or tests.

flag_name str, name of the flag
flag_values flags.FlagValues, optional FlagValues instance where the flag is defined.

AttributeError Raised when flag_name is not registered as a valid flag name.