finetuner.tuner.state module

class finetuner.tuner.state.TunerState(num_epochs=0, epoch=0, num_batches_train=0, num_batches_val=0, batch_index=0, current_loss=0.0, eval_metrics=<factory>, learning_rates=<factory>)[source]

Bases: object

A data container representing information on the current run.

num_epochs: int = 0
epoch: int = 0
num_batches_train: int = 0
num_batches_val: int = 0
batch_index: int = 0
current_loss: float = 0.0
eval_metrics: Dict[str, float]
learning_rates: Dict[str, float]