gqa
allennlp_models.vision.dataset_readers.gqa
GQAReader#
@DatasetReader.register("gqa")
class GQAReader(VisionReader):
| def __init__(
| self,
| image_dir: Union[str, PathLike],
| *, image_loader: Optional[ImageLoader] = None,
| *, image_featurizer: Optional[Lazy[GridEmbedder]] = None,
| *, region_detector: Optional[Lazy[RegionDetector]] = None,
| *, answer_vocab: Optional[Union[str, Vocabulary]] = None,
| *, feature_cache_dir: Optional[Union[str, PathLike]] = None,
| *, data_dir: Optional[Union[str, PathLike]] = None,
| *, tokenizer: Tokenizer = None,
| *, token_indexers: Dict[str, TokenIndexer] = None,
| *, cuda_device: Optional[Union[int, torch.device]] = None,
| *, max_instances: Optional[int] = None,
| *, image_processing_batch_size: int = 8,
| *, write_to_cache: bool = True
| ) -> None
Parametersimage_dir: `str`¶
Path to directory containing `png` image files.
image_loader : ImageLoader
image_featurizer: Lazy[GridEmbedder]
The backbone image processor (like a ResNet), whose output will be passed to the region
detector for finding object boxes in the image.
region_detector: Lazy[RegionDetector]
For pulling out regions of the image (both coordinates and features) that will be used by
downstream models.
data_dir: str
Path to directory containing text files for each dataset split. These files contain
the sentences and metadata for each task instance.
tokenizer: Tokenizer
, optional
token_indexers: Dict[str, TokenIndexer]
text_to_instance#
class GQAReader(VisionReader):
| ...
| def text_to_instance(
| self,
| question: str,
| image: Optional[Union[str, Tuple[Tensor, Tensor, Optional[Tensor], Optional[Tensor]]]],
| answer: Optional[Dict[str, str]] = None,
| *, use_cache: bool = True
| ) -> Optional[Instance]
apply_token_indexers#
class GQAReader(VisionReader):
| ...
| def apply_token_indexers(self, instance: Instance) -> None