add DOCUMENT_CONVERTERS

tacred.py

@@ -3,8 +3,10 @@ from typing import Any, Callable, Dict, List, Optional, Tuple
 
 import datasets
 import pytorch_ie.data.builder
+from pytorch_ie import token_based_document_to_text_based
 from pytorch_ie.annotations import BinaryRelation, LabeledSpan, _post_init_single_label
 from pytorch_ie.core import Annotation, AnnotationList, Document, annotation_field
+from pytorch_ie.documents import TextDocumentWithLabeledEntitiesAndRelations, TokenBasedDocument
 
 
 @dataclass(eq=True, frozen=True)
@@ -36,6 +38,12 @@ class TacredDocument(Document):
     dependency_relations: AnnotationList[TokenRelation] = annotation_field(target="tokens")
 
 
+@dataclass
+class SimpleTacredDocument(TokenBasedDocument):
+    entities: AnnotationList[LabeledSpan] = annotation_field(target="tokens")
+    relations: AnnotationList[BinaryRelation] = annotation_field(target="entities")
+
+
 def example_to_document(
     example: Dict[str, Any],
     relation_int2str: Callable[[int], str],
@@ -135,6 +143,18 @@ def document_to_example(
     }
 
 
+def convert_to_text_document_with_labeled_entities_and_relations(
+    document: TacredDocument,
+) -> TextDocumentWithLabeledEntitiesAndRelations:
+    doc_simplified = document.as_type(SimpleTacredDocument)
+    result = token_based_document_to_text_based(
+        doc_simplified,
+        result_document_type=TextDocumentWithLabeledEntitiesAndRelations,
+        join_tokens_with=" ",
+    )
+    return result
+
+
 class TacredConfig(datasets.BuilderConfig):
     """BuilderConfig for Tacred."""
 
@@ -149,6 +169,10 @@ class TacredConfig(datasets.BuilderConfig):
 class Tacred(pytorch_ie.data.builder.GeneratorBasedBuilder):
     DOCUMENT_TYPE = TacredDocument
 
+    DOCUMENT_CONVERTERS = {
+        TextDocumentWithLabeledEntitiesAndRelations: convert_to_text_document_with_labeled_entities_and_relations,
+    }
+
     BASE_DATASET_PATH = "DFKI-SLT/tacred"
 
     BUILDER_CONFIGS = [