tff.learning.ModelWeights

View source on GitHub

A container for the trainable and non-trainable variables of a Model.

View aliases

Main aliases

tff.learning.models.ModelWeights

tff.learning.ModelWeights(
    trainable, non_trainable
)

Used in the notebooks

Used in the tutorials
Composing Learning Algorithms Federated Learning for Text Generation Random noise generation in TFF

Note this does not include the model's local variables.

It may also be used to hold other values that are parallel to these variables, e.g., tensors corresponding to variable values, or updates to model variables.

Attributes
trainable	A namedtuple alias for field number 0
non_trainable	A namedtuple alias for field number 1

Methods

assign_weights_to

View source

assign_weights_to(
    model
)

Assign these TFF model weights to the weights of a model.

Args
model	A tf.keras.Model or tff.learning.models.VariableModel instance to assign the weights to.

convert_variables_to_arrays

View source

convert_variables_to_arrays() -> 'ModelWeights'

Converts any internal tf.Variables to numpy arrays.

from_model

View source

@classmethod
from_model(
    model
)

from_tff_result

View source

@classmethod
from_tff_result(
    struct
)