model module

Interfaces for ClientModel and ServerModel.

class model.Model(lr)

Bases: abc.ABC

close()

create_model()

Creates the model for the task.

Returns:features: A placeholder for the samples’ features. labels: A placeholder for the samples’ labels. train_op: A Tensorflow operation that, when run with the features and

the labels, trains the model.

eval_metric_ops: A Tensorflow operation that, when run with features and labels,

returns the accuracy of the model.

Return type:A 4-tuple consisting of

optimizer

Optimizer to be used by the model.

process_x(raw_x_batch)

Pre-processes each batch of features before being fed to the model.

process_y(raw_y_batch)

Pre-processes each batch of labels before being fed to the model.

test(data)

Tests the current model on the given data.

Parameters:data – dict of the form {‘x’: [list], ‘y’: [list]} Returns:dict of metrics that will be recorded by the simulation.

train(data, num_epochs=1, batch_size=10)

Trains the client model.

Parameters:

• data – Dict of the form {‘x’: [list], ‘y’: [list]}.
• num_epochs – Number of epochs to train.
• batch_size – Size of training batches.

Returns:

Number of FLOPs computed while training given data update: List of np.ndarray weights, with each weight array

corresponding to a variable in the resulting graph

Return type:

comp

class model.ServerModel(model)

Bases: object

close()

cur_model

save(path='checkpoints/model.ckpt')

send_to(clients)

Copies server model variables to each of the given clients

Parameters:clients – list of Client objects

size

update(updates)

Updates server model using given client updates.

Parameters:updates – list of (num_samples, update), where num_samples is the number of training samples corresponding to the update, and update is a list of variable weights