add prepare_documents

src/pytorch_ie/taskmodules/taskmodule.py

@@ -101,7 +101,9 @@ def encode(
 
     @abstractmethod
     def encode_input(
-        self, documents: List[Document]
+        self,
+        documents: List[Document],
+        is_training: bool = False,
     ) -> Tuple[List[InputEncoding], List[Metadata], Optional[List[Document]]]:
         raise NotImplementedError()
 

src/pytorch_ie/taskmodules/transformer_re_text_classification.py

@@ -266,7 +266,9 @@ def _encode_text(
         return encoding
 
     def encode_input(
-        self, documents: List[Document]
+        self,
+        documents: List[Document],
+        is_training: bool = False,
     ) -> Tuple[
         List[TransformerReTextClassificationInputEncoding],
         List[Metadata],

src/pytorch_ie/taskmodules/transformer_seq2seq.py

@@ -81,7 +81,9 @@ def encode_input_strings(self, inputs: List[str]) -> List[TransformerSeq2SeqInpu
         ]
 
     def encode_input(
-        self, documents: List[Document]
+        self,
+        documents: List[Document],
+        is_training: bool = False,
     ) -> Tuple[List[TransformerSeq2SeqInputEncoding], List[Metadata], Optional[List[Document]]]:
         input_strings = [self.document_to_input_string(document) for document in documents]
         return (

src/pytorch_ie/taskmodules/transformer_span_classification.py

@@ -99,7 +99,9 @@ def prepare(self, documents: List[Document]) -> None:
         self.id_to_label = {v: k for k, v in self.label_to_id.items()}
 
     def encode_input(
-        self, documents: List[Document]
+        self,
+        documents: List[Document],
+        is_training: bool = False,
     ) -> Tuple[
         List[TransformerSpanClassificationInputEncoding],
         List[Metadata],

src/pytorch_ie/taskmodules/transformer_text_classification.py

@@ -119,7 +119,9 @@ def prepare(self, documents: List[Document]) -> None:
         self.id_to_label = {v: k for k, v in self.label_to_id.items()}
 
     def encode_input(
-        self, documents: List[Document]
+        self,
+        documents: List[Document],
+        is_training: bool = False,
     ) -> Tuple[
         List[TransformerTextClassificationInputEncoding],
         List[Metadata],

src/pytorch_ie/taskmodules/transformer_token_classification.py

@@ -127,7 +127,9 @@ def encode_text(
         )
 
     def encode_input(
-        self, documents: List[Document]
+        self,
+        documents: List[Document],
+        is_training: bool = False,
     ) -> Tuple[
         List[TransformerTokenClassificationInputEncoding],
         List[Metadata],