def decode_output(self, output: Union[torch.Tensor, Dict[str,
torch.Tensor],
Iterable[torch.Tensor], Any],
mask: torch.BoolTensor, batch: Dict[str, Any], decoder,
**kwargs) -> Union[Dict[str, Any], Any]:
return TransformerNamedEntityRecognizer.decode_output(
self, output, batch['mask'], batch, decoder)
def compute_loss(
self, batch: Dict[str, Any],
output: Union[torch.Tensor, Dict[str, torch.Tensor],
Iterable[torch.Tensor], Any], criterion
) -> Union[torch.FloatTensor, Dict[str, torch.FloatTensor]]:
return TransformerNamedEntityRecognizer.compute_loss(
self, criterion, output, batch['tag_id'], batch['mask'])
def update_metrics(self, batch: Dict[str, Any],
output: Union[torch.Tensor, Dict[str, torch.Tensor],
Iterable[torch.Tensor], Any],
prediction: Dict[str, Any], metric: Union[MetricDict,
Metric]):
return TransformerNamedEntityRecognizer.update_metrics(
self, metric, output, batch['tag_id'], batch['mask'], batch,
prediction)
def prediction_to_result(self, prediction: Dict[str, Any],
batch: Dict[str, Any]) -> Union[List, Dict]:
return TransformerNamedEntityRecognizer.prediction_to_human(
self, prediction, self.vocabs['tag'].idx_to_token, batch)
def input_is_flat(self, data) -> bool:
return TransformerNamedEntityRecognizer.input_is_flat(self, data)
def build_metric(self, **kwargs):
return TransformerNamedEntityRecognizer.build_metric(self, **kwargs)