tfa.callbacks.AverageModelCheckpoint

The callback that saves average model weights.

tfa.callbacks.AverageModelCheckpoint(
    update_weights: bool,
    filepath: str,
    monitor: str = 'val_loss',
    verbose: int = 0,
    save_best_only: bool = False,
    save_weights_only: bool = False,
    mode: str = 'auto',
    save_freq: str = 'epoch',
    **kwargs
)

Used in the notebooks

Used in the tutorials

The callback that should be used with optimizers that extend tfa.optimizers.AveragedOptimizerWrapper, i.e., tfa.optimizers.MovingAverage and tfa.optimizers.StochasticAverage optimizers. It saves and, optionally, assigns the averaged weights.

update_weights If True, assign the moving average weights to the model, and save them. If False, keep the old non-averaged weights, but the saved model uses the average weights.

See tf.keras.callbacks.ModelCheckpoint for the other args.

Methods

set_model

View source

set_model(
    model
)

set_params

set_params(
    params
)