Fixing #13381 (#13400)

* Fixing #13381

* Enabling automatic LED models.

src/transformers/pipelines/zero_shot_classification.py

@@ -88,7 +88,7 @@ def _parse_and_tokenize(
         hypothesis_template,
         padding=True,
         add_special_tokens=True,
-        truncation=TruncationStrategy.DO_NOT_TRUNCATE,
+        truncation=TruncationStrategy.ONLY_FIRST,
         **kwargs
     ):
         """
@@ -113,13 +113,31 @@ def _parse_and_tokenize(
                 )
                 inputs.append(model_input)
         else:
-            inputs = self.tokenizer(
-                sequence_pairs,
-                add_special_tokens=add_special_tokens,
-                return_tensors=return_tensors,
-                padding=padding,
-                truncation=truncation,
-            )
+            try:
+                inputs = self.tokenizer(
+                    sequence_pairs,
+                    add_special_tokens=add_special_tokens,
+                    return_tensors=return_tensors,
+                    padding=padding,
+                    truncation=truncation,
+                )
+            except Exception as e:
+                if "too short" in str(e):
+                    # tokenizers might yell that we want to truncate
+                    # to a value that is not even reached by the input.
+                    # In that case we don't want to truncate.
+                    # It seems there's not a really better way to catch that
+                    # exception.
+
+                    inputs = self.tokenizer(
+                        sequence_pairs,
+                        add_special_tokens=add_special_tokens,
+                        return_tensors=return_tensors,
+                        padding=padding,
+                        truncation=TruncationStrategy.DO_NOT_TRUNCATE,
+                    )
+                else:
+                    raise e
 
         return inputs
 

tests/test_pipelines_zero_shot.py

@@ -105,6 +105,20 @@ def run_entailment_id(self, zero_shot_classifier: Pipeline):
         zero_shot_classifier.model.config.label2id = original_label2id
         self.assertEqual(original_entailment, zero_shot_classifier.entailment_id)
 
+    @require_torch
+    def test_truncation(self):
+        zero_shot_classifier = pipeline(
+            "zero-shot-classification",
+            model="sshleifer/tiny-distilbert-base-cased-distilled-squad",
+            framework="pt",
+        )
+        # There was a regression in 4.10 for this
+        # Adding a test so we don't make the mistake again.
+        # https://github.com/huggingface/transformers/issues/13381#issuecomment-912343499
+        zero_shot_classifier(
+            "Who are you voting for in 2020?" * 100, candidate_labels=["politics", "public health", "science"]
+        )
+
     @require_torch
     def test_small_model_pt(self):
         zero_shot_classifier = pipeline(