update

UpdateOp

Bases: TensorOp

This class performs updates to a model's weights based on the loss.

Parameters:

Name	Type	Description	Default
model	Union[tf.keras.Model, torch.nn.Module]	Model instance compiled by fe.build.	required
loss_name	str	The name of loss.	required
mode	Union[None, str, Iterable[str]]	What mode(s) to execute this Op in. For example, "train", "eval", "test", or "infer". To execute regardless of mode, pass None. To execute in all modes except for a particular one, you can pass an argument like "!infer" or "!train".	'train'

Source code in fastestimator\fastestimator\op\tensorop\model\update.py

class UpdateOp(TensorOp):
    """This class performs updates to a model's weights based on the loss.

    Args:
        model: Model instance compiled by fe.build.
        loss_name: The name of loss.
        mode: What mode(s) to execute this Op in. For example, "train", "eval", "test", or "infer". To execute
            regardless of mode, pass None. To execute in all modes except for a particular one, you can pass an argument
            like "!infer" or "!train".
    """
    def __init__(self,
                 model: Union[tf.keras.Model, torch.nn.Module],
                 loss_name: str,
                 mode: Union[None, str, Iterable[str]] = "train"):
        super().__init__(inputs=loss_name, outputs=None, mode=mode)
        self.model = model
        self.retain_graph = False
        self.weight_decay = isinstance(self.model, tf.keras.Model) and self.model.losses
        if not hasattr(self.model, "loss_name"):
            self.model.loss_name = {loss_name}
        else:
            self.model.loss_name.add(loss_name)

    def forward(self, data: Union[Tensor, List[Tensor]], state: Dict[str, Any]):
        if not state["warmup"]:
            if self.weight_decay:
                data = data + tf.reduce_sum(self.model.losses)
            update_model(self.model, data, tape=state['tape'], retain_graph=self.retain_graph)