DataCollatorForTokenClassification numpy fix (#13609)

* Fix issue when labels are supplied as Numpy array instead of list

* Fix issue when labels are supplied as Numpy array instead of list

* Fix same issue in the `TokenClassification` data collator

* Style pass

src/transformers/data/data_collator.py

@@ -291,11 +291,11 @@ class DataCollatorForTokenClassification(DataCollatorMixin):
         padding_side = self.tokenizer.padding_side
         if padding_side == "right":
             batch[label_name] = [
-                label + [self.label_pad_token_id] * (sequence_length - len(label)) for label in labels
+                list(label) + [self.label_pad_token_id] * (sequence_length - len(label)) for label in labels
             ]
         else:
             batch[label_name] = [
-                [self.label_pad_token_id] * (sequence_length - len(label)) + label for label in labels
+                [self.label_pad_token_id] * (sequence_length - len(label)) + list(label) for label in labels
             ]
 
         batch = {k: torch.tensor(v, dtype=torch.int64) for k, v in batch.items()}
@@ -321,9 +321,13 @@ class DataCollatorForTokenClassification(DataCollatorMixin):
         sequence_length = tf.convert_to_tensor(batch["input_ids"]).shape[1]
         padding_side = self.tokenizer.padding_side
         if padding_side == "right":
-            batch["labels"] = [label + [self.label_pad_token_id] * (sequence_length - len(label)) for label in labels]
+            batch["labels"] = [
+                list(label) + [self.label_pad_token_id] * (sequence_length - len(label)) for label in labels
+            ]
         else:
-            batch["labels"] = [[self.label_pad_token_id] * (sequence_length - len(label)) + label for label in labels]
+            batch["labels"] = [
+                [self.label_pad_token_id] * (sequence_length - len(label)) + list(label) for label in labels
+            ]
 
         batch = {k: tf.convert_to_tensor(v, dtype=tf.int64) for k, v in batch.items()}
         return batch
@@ -348,9 +352,13 @@ class DataCollatorForTokenClassification(DataCollatorMixin):
         sequence_length = np.array(batch["input_ids"]).shape[1]
         padding_side = self.tokenizer.padding_side
         if padding_side == "right":
-            batch["labels"] = [label + [self.label_pad_token_id] * (sequence_length - len(label)) for label in labels]
+            batch["labels"] = [
+                list(label) + [self.label_pad_token_id] * (sequence_length - len(label)) for label in labels
+            ]
         else:
-            batch["labels"] = [[self.label_pad_token_id] * (sequence_length - len(label)) + label for label in labels]
+            batch["labels"] = [
+                [self.label_pad_token_id] * (sequence_length - len(label)) + list(label) for label in labels
+            ]
 
         batch = {k: np.array(v, dtype=np.int64) for k, v in batch.items()}
         return batch