View source on GitHub |
A structure representing the program state.
tff.learning.programs.ProgramState(
state, round_number, next_evaluation_timestamp_seconds, data_iterator
)
Attributes | |
---|---|
state
|
The state of the training process. |
round_number
|
The current round number. |
next_evaluation_timestamp_seconds
|
The timestamp of the next evaluation in seconds. |
data_iterator
|
The tff.program.FederatedDataSourceIterator used for
training.
|