def test_raises_no_repeat_and_no_take(self):
with self.assertRaisesRegex(
ValueError, 'Argument client_epochs_per_round is set to -1'):
emnist_dataset.get_emnist_datasets(
client_batch_size=10,
client_epochs_per_round=-1,
max_batches_per_client=-1)
def run_experiment():
"""Data preprocessing and experiment execution."""
emnist_train, emnist_test = emnist_dataset.get_emnist_datasets(
FLAGS.client_batch_size,
FLAGS.client_epochs_per_round,
only_digits=FLAGS.only_digits)
example_dataset = emnist_train.create_tf_dataset_for_client(
emnist_train.client_ids[0])
input_spec = example_dataset.element_spec
client_datasets_fn = training_utils.build_client_datasets_fn(
emnist_train, FLAGS.clients_per_round)
evaluate_fn = training_utils.build_evaluate_fn(
eval_dataset=emnist_test,
model_builder=model_builder,
loss_builder=loss_builder,
metrics_builder=metrics_builder,
assign_weights_to_keras_model=assign_weights_fn)
client_optimizer_fn = functools.partial(
utils_impl.create_optimizer_from_flags, 'client')
server_optimizer_fn = functools.partial(
utils_impl.create_optimizer_from_flags, 'server')
def tff_model_fn():
keras_model = model_builder()
return tff.learning.from_keras_model(keras_model,
input_spec=input_spec,
loss=loss_builder(),
metrics=metrics_builder())
if FLAGS.use_compression:
# We create a `MeasuredProcess` for broadcast process and a
# `MeasuredProcess` for aggregate process by providing the
# `_broadcast_encoder_fn` and `_mean_encoder_fn` to corresponding utilities.
# The fns are called once for each of the model weights created by
# tff_model_fn, and return instances of appropriate encoders.
encoded_broadcast_process = (
tff.learning.framework.build_encoded_broadcast_process_from_model(
tff_model_fn, _broadcast_encoder_fn))
encoded_mean_process = (
tff.learning.framework.build_encoded_mean_process_from_model(
tff_model_fn, _mean_encoder_fn))
else:
encoded_broadcast_process = None
encoded_mean_process = None
iterative_process = tff.learning.build_federated_averaging_process(
model_fn=tff_model_fn,
client_optimizer_fn=client_optimizer_fn,
server_optimizer_fn=server_optimizer_fn,
aggregation_process=encoded_mean_process,
broadcast_process=encoded_broadcast_process)
training_loop.run(iterative_process=iterative_process,
client_datasets_fn=client_datasets_fn,
validation_fn=evaluate_fn)
def main(argv):
if len(argv) > 1:
raise app.UsageError('Expected no command-line arguments, '
'got: {}'.format(argv))
emnist_train, emnist_test = emnist_dataset.get_emnist_datasets(
FLAGS.client_batch_size,
FLAGS.client_epochs_per_round,
only_digits=False)
input_spec = emnist_train.create_tf_dataset_for_client(
emnist_train.client_ids[0]).element_spec
if FLAGS.model == 'cnn':
model_builder = functools.partial(
emnist_models.create_conv_dropout_model, only_digits=False)
elif FLAGS.model == '2nn':
model_builder = functools.partial(
emnist_models.create_two_hidden_layer_model, only_digits=False)
else:
raise ValueError('Cannot handle model flag [{!s}].'.format(
FLAGS.model))
loss_builder = tf.keras.losses.SparseCategoricalCrossentropy
metrics_builder = lambda: [tf.keras.metrics.SparseCategoricalAccuracy()]
training_process = iterative_process_builder.from_flags(
input_spec=input_spec,
model_builder=model_builder,
loss_builder=loss_builder,
metrics_builder=metrics_builder)
client_datasets_fn = training_utils.build_client_datasets_fn(
train_dataset=emnist_train,
train_clients_per_round=FLAGS.clients_per_round,
random_seed=FLAGS.client_datasets_random_seed)
assign_weights_fn = fed_avg_schedule.ServerState.assign_weights_to_keras_model
evaluate_fn = training_utils.build_evaluate_fn(
eval_dataset=emnist_test,
model_builder=model_builder,
loss_builder=loss_builder,
metrics_builder=metrics_builder,
assign_weights_to_keras_model=assign_weights_fn)
test_fn = evaluate_fn
logging.info('Training model:')
logging.info(model_builder().summary())
training_loop.run(training_process,
client_datasets_fn,
evaluate_fn,
test_fn=test_fn)
def test_take_with_repeat(self):
emnist_train, _ = emnist_dataset.get_emnist_datasets(
client_batch_size=10,
client_epochs_per_round=-1,
max_batches_per_client=10,
only_digits=True)
self.assertEqual(len(emnist_train.client_ids), 3383)
for i in range(10):
client_ds = emnist_train.create_tf_dataset_for_client(
emnist_train.client_ids[i])
self.assertEqual(_compute_length_of_dataset(client_ds), 10)
def test_emnist_dataset_structure(self):
emnist_train, emnist_test = emnist_dataset.get_emnist_datasets(
client_batch_size=10, client_epochs_per_round=1, only_digits=True)
self.assertEqual(len(emnist_train.client_ids), 3383)
sample_train_ds = emnist_train.create_tf_dataset_for_client(
emnist_train.client_ids[0])
train_batch = next(iter(sample_train_ds))
train_batch_shape = train_batch[0].shape
test_batch = next(iter(emnist_test))
test_batch_shape = test_batch[0].shape
self.assertEqual(train_batch_shape.as_list(), [10, 28, 28, 1])
self.assertEqual(test_batch_shape.as_list(), [TEST_BATCH_SIZE, 28, 28, 1])
def main(argv):
if len(argv) > 1:
raise app.UsageError('Expected no command-line arguments, '
'got: {}'.format(argv))
emnist_train, emnist_test = emnist_dataset.get_emnist_datasets(
FLAGS.client_batch_size, FLAGS.client_epochs_per_round, only_digits=False)
if FLAGS.model == 'cnn':
model_builder = functools.partial(
emnist_models.create_conv_dropout_model, only_digits=False)
elif FLAGS.model == '2nn':
model_builder = functools.partial(
emnist_models.create_two_hidden_layer_model, only_digits=False)
else:
raise ValueError('Cannot handle model flag [{!s}].'.format(FLAGS.model))
loss_builder = tf.keras.losses.SparseCategoricalCrossentropy
metrics_builder = lambda: [tf.keras.metrics.SparseCategoricalAccuracy()]
if FLAGS.uniform_weighting:
def client_weight_fn(local_outputs):
del local_outputs
return 1.0
else:
client_weight_fn = None # Defaults to the number of examples per client.
def model_fn():
return tff.learning.from_keras_model(
model_builder(),
loss_builder(),
input_spec=emnist_test.element_spec,
metrics=metrics_builder())
if FLAGS.noise_multiplier is not None:
if not FLAGS.uniform_weighting:
raise ValueError(
'Differential privacy is only implemented for uniform weighting.')
dp_query = tff.utils.build_dp_query(
clip=FLAGS.clip,
noise_multiplier=FLAGS.noise_multiplier,
expected_total_weight=FLAGS.clients_per_round,
adaptive_clip_learning_rate=FLAGS.adaptive_clip_learning_rate,
target_unclipped_quantile=FLAGS.target_unclipped_quantile,
clipped_count_budget_allocation=FLAGS.clipped_count_budget_allocation,
expected_num_clients=FLAGS.clients_per_round,
per_vector_clipping=FLAGS.per_vector_clipping,
model=model_fn())
weights_type = tff.learning.framework.weights_type_from_model(model_fn)
aggregation_process = tff.utils.build_dp_aggregate_process(
weights_type.trainable, dp_query)
else:
aggregation_process = None
server_optimizer_fn = optimizer_utils.create_optimizer_fn_from_flags('server')
client_optimizer_fn = optimizer_utils.create_optimizer_fn_from_flags('client')
iterative_process = tff.learning.build_federated_averaging_process(
model_fn=model_fn,
server_optimizer_fn=server_optimizer_fn,
client_weight_fn=client_weight_fn,
client_optimizer_fn=client_optimizer_fn,
aggregation_process=aggregation_process)
client_datasets_fn = training_utils.build_client_datasets_fn(
emnist_train, FLAGS.clients_per_round)
evaluate_fn = training_utils.build_evaluate_fn(
eval_dataset=emnist_test,
model_builder=model_builder,
loss_builder=loss_builder,
metrics_builder=metrics_builder,
assign_weights_to_keras_model=dp_utils.assign_weights_to_keras_model)
logging.info('Training model:')
logging.info(model_builder().summary())
training_loop.run(
iterative_process=iterative_process,
client_datasets_fn=client_datasets_fn,
validation_fn=evaluate_fn,
)
def run_federated(
iterative_process_builder: Callable[...,
tff.templates.IterativeProcess],
assign_weights_fn: Callable[[Any, tf.keras.Model], None],
client_epochs_per_round: int,
client_batch_size: int,
clients_per_round: int,
client_datasets_random_seed: Optional[int] = None,
emnist_model: Optional[str] = 'cnn'):
"""Runs an iterative process on the EMNIST character recognition task.
This method will load and pre-process dataset and construct a model used for
the task. It then uses `iterative_process_builder` to create an iterative
process that it applies to the task, using
`tensorflow_federated.python.research.utils.training_loop`.
We assume that the iterative process has the following functional type
signatures:
* `initialize`: `( -> S@SERVER)` where `S` represents the server state.
* `next`: `<S@SERVER, {B*}@CLIENTS> -> <S@SERVER, T@SERVER>` where `S`
represents the server state, `{B*}` represents the client datasets,
and `T` represents a python `Mapping` object.
Args:
iterative_process_builder: A function that accepts a no-arg `model_fn`, and
returns a `tff.templates.IterativeProcess`. The `model_fn` must return a
`tff.learning.Model`.
assign_weights_fn: A function that accepts the server state `S` and a
`tf.keras.Model`, and updates the weights in the Keras model. This is used
to do evaluation using Keras.
client_epochs_per_round: An integer representing the number of epochs of
training performed per client in each training round.
client_batch_size: An integer representing the batch size used on clients.
clients_per_round: An integer representing the number of clients
participating in each round.
client_datasets_random_seed: An optional int used to seed which clients are
sampled at each round. If `None`, no seed is used.
emnist_model: A string specifying the model used for character recognition.
Can be one of `cnn` and `2nn`, corresponding to a CNN model and a densely
connected 2-layer model (respectively).
"""
emnist_train, emnist_test = emnist_dataset.get_emnist_datasets(
client_batch_size, client_epochs_per_round, only_digits=False)
input_spec = emnist_train.create_tf_dataset_for_client(
emnist_train.client_ids[0]).element_spec
if emnist_model == 'cnn':
model_builder = functools.partial(
emnist_models.create_conv_dropout_model, only_digits=False)
elif emnist_model == '2nn':
model_builder = functools.partial(
emnist_models.create_two_hidden_layer_model, only_digits=False)
else:
raise ValueError(
'Cannot handle model flag [{!s}], must be one of {!s}.'.format(
emnist_model, EMNIST_MODELS))
loss_builder = tf.keras.losses.SparseCategoricalCrossentropy
metrics_builder = lambda: [tf.keras.metrics.SparseCategoricalAccuracy()]
def tff_model_fn() -> tff.learning.Model:
return tff.learning.from_keras_model(keras_model=model_builder(),
input_spec=input_spec,
loss=loss_builder(),
metrics=metrics_builder())
training_process = iterative_process_builder(tff_model_fn)
client_datasets_fn = training_utils.build_client_datasets_fn(
train_dataset=emnist_train,
train_clients_per_round=clients_per_round,
random_seed=client_datasets_random_seed)
evaluate_fn = training_utils.build_evaluate_fn(
eval_dataset=emnist_test,
model_builder=model_builder,
loss_builder=loss_builder,
metrics_builder=metrics_builder,
assign_weights_to_keras_model=assign_weights_fn)
logging.info('Training model:')
logging.info(model_builder().summary())
training_loop.run(iterative_process=training_process,
client_datasets_fn=client_datasets_fn,
validation_fn=evaluate_fn,
test_fn=evaluate_fn)
def run_federated(
iterative_process_builder: Callable[..., tff.templates.IterativeProcess],
client_epochs_per_round: int,
client_batch_size: int,
clients_per_round: int,
max_batches_per_client: Optional[int] = -1,
client_datasets_random_seed: Optional[int] = None,
emnist_model: Optional[str] = 'cnn',
total_rounds: Optional[int] = 1500,
experiment_name: Optional[str] = 'federated_emnist_cr',
root_output_dir: Optional[str] = '/tmp/fed_opt',
max_eval_batches: Optional[int] = None,
**kwargs):
"""Runs an iterative process on the EMNIST character recognition task.
This method will load and pre-process dataset and construct a model used for
the task. It then uses `iterative_process_builder` to create an iterative
process that it applies to the task, using
`tensorflow_federated.python.research.utils.training_loop`.
We assume that the iterative process has the following functional type
signatures:
* `initialize`: `( -> S@SERVER)` where `S` represents the server state.
* `next`: `<S@SERVER, {B*}@CLIENTS> -> <S@SERVER, T@SERVER>` where `S`
represents the server state, `{B*}` represents the client datasets,
and `T` represents a python `Mapping` object.
Moreover, the server state must have an attribute `model` of type
`tff.learning.ModelWeights`.
Args:
iterative_process_builder: A function that accepts a no-arg `model_fn`, and
returns a `tff.templates.IterativeProcess`. The `model_fn` must return a
`tff.learning.Model`.
client_epochs_per_round: An integer representing the number of epochs of
training performed per client in each training round.
client_batch_size: An integer representing the batch size used on clients.
clients_per_round: An integer representing the number of clients
participating in each round.
max_batches_per_client: An optional int specifying the number of batches
taken by each client at each round. If `-1`, the entire client dataset is
used.
client_datasets_random_seed: An optional int used to seed which clients are
sampled at each round. If `None`, no seed is used.
emnist_model: A string specifying the model used for character recognition.
Can be one of `cnn` and `2nn`, corresponding to a CNN model and a densely
connected 2-layer model (respectively).
total_rounds: The number of federated training rounds.
experiment_name: The name of the experiment being run. This will be appended
to the `root_output_dir` for purposes of writing outputs.
root_output_dir: The name of the root output directory for writing
experiment outputs.
max_eval_batches: If set to a positive integer, evaluation datasets are
capped to at most that many batches. If set to None or a nonpositive
integer, the full evaluation datasets are used.
**kwargs: Additional arguments configuring the training loop. For details
on supported arguments, see
`tensorflow_federated/python/research/utils/training_utils.py`.
"""
emnist_train, _ = emnist_dataset.get_emnist_datasets(
client_batch_size,
client_epochs_per_round,
max_batches_per_client=max_batches_per_client,
only_digits=False)
_, emnist_test = emnist_dataset.get_centralized_datasets(
train_batch_size=client_batch_size,
max_test_batches=max_eval_batches,
only_digits=False)
input_spec = emnist_train.create_tf_dataset_for_client(
emnist_train.client_ids[0]).element_spec
if emnist_model == 'cnn':
model_builder = functools.partial(
emnist_models.create_conv_dropout_model, only_digits=False)
elif emnist_model == '2nn':
model_builder = functools.partial(
emnist_models.create_two_hidden_layer_model, only_digits=False)
else:
raise ValueError(
'Cannot handle model flag [{!s}], must be one of {!s}.'.format(
emnist_model, EMNIST_MODELS))
loss_builder = tf.keras.losses.SparseCategoricalCrossentropy
metrics_builder = lambda: [tf.keras.metrics.SparseCategoricalAccuracy()]
def tff_model_fn() -> tff.learning.Model:
return tff.learning.from_keras_model(
keras_model=model_builder(),
input_spec=input_spec,
loss=loss_builder(),
metrics=metrics_builder())
training_process = iterative_process_builder(tff_model_fn)
client_datasets_fn = training_utils.build_client_datasets_fn(
train_dataset=emnist_train,
train_clients_per_round=clients_per_round,
random_seed=client_datasets_random_seed)
evaluate_fn = training_utils.build_evaluate_fn(
eval_dataset=emnist_test,
model_builder=model_builder,
loss_builder=loss_builder,
metrics_builder=metrics_builder)
logging.info('Training model:')
logging.info(model_builder().summary())
training_loop.run(
iterative_process=training_process,
client_datasets_fn=client_datasets_fn,
validation_fn=evaluate_fn,
test_fn=evaluate_fn,
total_rounds=total_rounds,
experiment_name=experiment_name,
root_output_dir=root_output_dir,
**kwargs)
def main(argv):
if len(argv) > 1:
raise app.UsageError('Expected no command-line arguments, '
'got: {}'.format(argv))
emnist_train, emnist_test = emnist_dataset.get_emnist_datasets(
FLAGS.client_batch_size,
FLAGS.client_epochs_per_round,
max_batches_per_client=FLAGS.max_batches_per_client,
only_digits=FLAGS.only_digits)
central_emnist_train, _ = emnist_dataset.get_centralized_emnist_datasets(
batch_size=100, only_digits=FLAGS.only_digits, shuffle_train=False)
input_spec = emnist_train.create_tf_dataset_for_client(
emnist_train.client_ids[0]).element_spec
if FLAGS.model == 'cnn':
model_builder = functools.partial(
emnist_models.create_conv_dropout_model,
only_digits=FLAGS.only_digits)
elif FLAGS.model == 'orig_cnn':
model_builder = functools.partial(
emnist_models.create_original_fedavg_cnn_model,
only_digits=FLAGS.only_digits)
elif FLAGS.model == '2nn':
model_builder = functools.partial(
emnist_models.create_two_hidden_layer_model,
only_digits=FLAGS.only_digits)
else:
raise ValueError('Cannot handle model flag [{!s}].'.format(
FLAGS.model))
loss_builder = tf.keras.losses.SparseCategoricalCrossentropy
metrics_builder = lambda: [tf.keras.metrics.SparseCategoricalAccuracy()]
if FLAGS.client_weight == 'uniform':
def client_weight_fn(local_outputs):
del local_outputs
return 1.0
elif FLAGS.client_weight == 'num_samples':
client_weight_fn = None
else:
raise ValueError(
'Unsupported client_weight flag [{!s}]. Currently only '
'`uniform` and `num_samples` are supported.'.format(
FLAGS.client_weight))
training_process = decay_iterative_process_builder.from_flags(
input_spec=input_spec,
model_builder=model_builder,
loss_builder=loss_builder,
metrics_builder=metrics_builder,
client_weight_fn=client_weight_fn)
client_datasets_fn = training_utils.build_client_datasets_fn(
emnist_train,
FLAGS.clients_per_round,
random_seed=FLAGS.client_datasets_random_seed)
assign_weights_fn = adaptive_fed_avg.ServerState.assign_weights_to_keras_model
evaluate_fn = training_utils.build_evaluate_fn(
eval_dataset=emnist_test,
model_builder=model_builder,
loss_builder=loss_builder,
metrics_builder=metrics_builder,
assign_weights_to_keras_model=assign_weights_fn)
train_evaluate_fn = training_utils.build_evaluate_fn(
eval_dataset=central_emnist_train,
model_builder=model_builder,
loss_builder=loss_builder,
metrics_builder=metrics_builder,
assign_weights_to_keras_model=assign_weights_fn)
logging.info('Training model:')
logging.info(model_builder().summary())
training_loop.run(
iterative_process=training_process,
client_datasets_fn=client_datasets_fn,
validation_fn=evaluate_fn,
train_eval_fn=train_evaluate_fn,
)
