diff --git a/setup.py b/setup.py
index 96cd1c16aed2..ba64455a5870 100644
--- a/setup.py
+++ b/setup.py
@@ -146,7 +146,7 @@
     "tomli",
     "tiktoken",
     "timm>=1.0.23",
-    "tokenizers>=0.22.0,<=0.23.0",
+    "tokenizers>=0.23.1,<0.24.0",
     "torch>=2.4",
     "torchaudio",
     "torchvision",
diff --git a/src/transformers/convert_slow_tokenizer.py b/src/transformers/convert_slow_tokenizer.py
index 2637db428a11..5f7c80d1a9f2 100644
--- a/src/transformers/convert_slow_tokenizer.py
+++ b/src/transformers/convert_slow_tokenizer.py
@@ -481,8 +481,8 @@ def converted(self) -> Tokenizer:
         tokenizer.pre_tokenizer = pre_tokenizers.BertPreTokenizer()
         tokenizer.decoder = decoders.BPEDecoder(suffix=token_suffix)
         tokenizer.post_processor = processors.BertProcessing(
-            sep=(self.original_tokenizer.sep_token, self.original_tokenizer.sep_token_id),
-            cls=(self.original_tokenizer.cls_token, self.original_tokenizer.cls_token_id),
+            (self.original_tokenizer.sep_token, self.original_tokenizer.sep_token_id),
+            (self.original_tokenizer.cls_token, self.original_tokenizer.cls_token_id),
         )
 
         return tokenizer
@@ -552,10 +552,10 @@ def converted(self) -> Tokenizer:
         tokenizer.pre_tokenizer = pre_tokenizers.ByteLevel(add_prefix_space=ot.add_prefix_space)
         tokenizer.decoder = decoders.ByteLevel()
         tokenizer.post_processor = processors.RobertaProcessing(
-            sep=(ot.sep_token, ot.sep_token_id),
-            cls=(ot.cls_token, ot.cls_token_id),
-            add_prefix_space=ot.add_prefix_space,
-            trim_offsets=True,  # True by default on Roberta (historical)
+            (ot.sep_token, ot.sep_token_id),
+            (ot.cls_token, ot.cls_token_id),
+            True,  # trim_offsets: True by default on Roberta (historical)
+            ot.add_prefix_space,
         )
 
         return tokenizer
@@ -1459,10 +1459,10 @@ def converted(self) -> Tokenizer:
 
         # Hack to have a ByteLevel and TemplateProcessor
         tokenizer.post_processor = processors.RobertaProcessing(
-            sep=(self.original_tokenizer.eos_token, self.original_tokenizer.eos_token_id),
-            cls=(self.original_tokenizer.bos_token, self.original_tokenizer.bos_token_id),
-            add_prefix_space=False,
-            trim_offsets=False,
+            (self.original_tokenizer.eos_token, self.original_tokenizer.eos_token_id),
+            (self.original_tokenizer.bos_token, self.original_tokenizer.bos_token_id),
+            False,  # trim_offsets
+            False,  # add_prefix_space
         )
         return tokenizer
 
diff --git a/src/transformers/dependency_versions_table.py b/src/transformers/dependency_versions_table.py
index 1c9d55d7507e..a70583dfe28d 100644
--- a/src/transformers/dependency_versions_table.py
+++ b/src/transformers/dependency_versions_table.py
@@ -73,7 +73,7 @@
     "tomli": "tomli",
     "tiktoken": "tiktoken",
     "timm": "timm>=1.0.23",
-    "tokenizers": "tokenizers>=0.22.0,<=0.23.0",
+    "tokenizers": "tokenizers>=0.23.1,<0.24.0",
     "torch": "torch>=2.4",
     "torchaudio": "torchaudio",
     "torchvision": "torchvision",
diff --git a/src/transformers/models/bertweet/tokenization_bertweet.py b/src/transformers/models/bertweet/tokenization_bertweet.py
index 95e9f46787e8..5d639a705609 100644
--- a/src/transformers/models/bertweet/tokenization_bertweet.py
+++ b/src/transformers/models/bertweet/tokenization_bertweet.py
@@ -15,6 +15,7 @@
 """Tokenization classes for BERTweet"""
 
 import html
+import html.entities
 import os
 import re
 
diff --git a/src/transformers/models/clip/tokenization_clip.py b/src/transformers/models/clip/tokenization_clip.py
index 018c630afbce..cb0c02667d94 100644
--- a/src/transformers/models/clip/tokenization_clip.py
+++ b/src/transformers/models/clip/tokenization_clip.py
@@ -115,10 +115,10 @@ def __init__(
         )
 
         self._tokenizer.post_processor = processors.RobertaProcessing(
-            sep=(str(eos_token), self.eos_token_id),
-            cls=(str(bos_token), self.bos_token_id),
-            add_prefix_space=False,
-            trim_offsets=False,
+            (str(eos_token), self.eos_token_id),
+            (str(bos_token), self.bos_token_id),
+            False,  # trim_offsets
+            False,  # add_prefix_space
         )
 
         # Very ugly hack to enable padding to have a correct decoding see https://github.com/huggingface/tokenizers/issues/872
diff --git a/src/transformers/models/herbert/tokenization_herbert.py b/src/transformers/models/herbert/tokenization_herbert.py
index eb0543187b62..97d7f12347af 100644
--- a/src/transformers/models/herbert/tokenization_herbert.py
+++ b/src/transformers/models/herbert/tokenization_herbert.py
@@ -103,8 +103,8 @@ def __init__(
         )
 
         self._tokenizer.post_processor = processors.BertProcessing(
-            sep=(self.sep_token, 2),
-            cls=(self.cls_token, 0),
+            (self.sep_token, 2),
+            (self.cls_token, 0),
         )
 
 
diff --git a/src/transformers/models/layoutlmv2/tokenization_layoutlmv2.py b/src/transformers/models/layoutlmv2/tokenization_layoutlmv2.py
index d8677d06facf..619efcd23c40 100644
--- a/src/transformers/models/layoutlmv2/tokenization_layoutlmv2.py
+++ b/src/transformers/models/layoutlmv2/tokenization_layoutlmv2.py
@@ -249,7 +249,7 @@ def __call__(
         word_labels: list[int] | list[list[int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
@@ -401,7 +401,7 @@ def batch_encode_plus(
         word_labels: list[int] | list[list[int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
@@ -466,7 +466,7 @@ def encode_plus(
         word_labels: list[int] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
diff --git a/src/transformers/models/layoutlmv3/tokenization_layoutlmv3.py b/src/transformers/models/layoutlmv3/tokenization_layoutlmv3.py
index cda7c0b8f324..2fc53565b745 100644
--- a/src/transformers/models/layoutlmv3/tokenization_layoutlmv3.py
+++ b/src/transformers/models/layoutlmv3/tokenization_layoutlmv3.py
@@ -226,10 +226,10 @@ def __init__(
         sep_token_id = self.sep_token_id
 
         self._tokenizer.post_processor = processors.RobertaProcessing(
-            sep=(sep, sep_token_id),
-            cls=(cls, cls_token_id),
-            add_prefix_space=add_prefix_space,
-            trim_offsets=True,
+            (sep, sep_token_id),
+            (cls, cls_token_id),
+            True,  # trim_offsets
+            add_prefix_space,
         )
         self.cls_token_box = cls_token_box
         self.sep_token_box = sep_token_box
@@ -246,7 +246,7 @@ def __call__(
         word_labels: list[int] | list[list[int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
@@ -398,7 +398,7 @@ def batch_encode_plus(
         word_labels: list[int] | list[list[int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
@@ -463,7 +463,7 @@ def encode_plus(
         word_labels: list[int] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
diff --git a/src/transformers/models/layoutxlm/tokenization_layoutxlm.py b/src/transformers/models/layoutxlm/tokenization_layoutxlm.py
index b4d51c9bf3dc..25c8cad665ce 100644
--- a/src/transformers/models/layoutxlm/tokenization_layoutxlm.py
+++ b/src/transformers/models/layoutxlm/tokenization_layoutxlm.py
@@ -300,7 +300,7 @@ def encode_plus(
         word_labels: list[int] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
@@ -359,7 +359,7 @@ def batch_encode_plus(
         word_labels: list[list[int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
@@ -419,7 +419,7 @@ def __call__(
         word_labels: list[int] | list[list[int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
diff --git a/src/transformers/models/luke/tokenization_luke.py b/src/transformers/models/luke/tokenization_luke.py
index bed415c79291..6b43b08f9a04 100644
--- a/src/transformers/models/luke/tokenization_luke.py
+++ b/src/transformers/models/luke/tokenization_luke.py
@@ -433,7 +433,7 @@ def __call__(
         entities_pair: EntityInput | list[EntityInput] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         max_entity_length: int | None = None,
         stride: int = 0,
@@ -1133,7 +1133,7 @@ def prepare_for_model(
         pair_entity_token_spans: list[tuple[int, int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         max_entity_length: int | None = None,
         stride: int = 0,
diff --git a/src/transformers/models/markuplm/tokenization_markuplm.py b/src/transformers/models/markuplm/tokenization_markuplm.py
index 9ebfdde292f4..6af2438fffd9 100644
--- a/src/transformers/models/markuplm/tokenization_markuplm.py
+++ b/src/transformers/models/markuplm/tokenization_markuplm.py
@@ -284,7 +284,7 @@ def __call__(
         node_labels: list[int] | list[list[int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         is_split_into_words: bool = False,
@@ -459,7 +459,7 @@ def batch_encode_plus(
         node_labels: list[int] | list[list[int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
@@ -524,7 +524,7 @@ def encode_plus(
         node_labels: list[int] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
diff --git a/src/transformers/models/mluke/tokenization_mluke.py b/src/transformers/models/mluke/tokenization_mluke.py
index abe9ff85d190..77f6dbe4387f 100644
--- a/src/transformers/models/mluke/tokenization_mluke.py
+++ b/src/transformers/models/mluke/tokenization_mluke.py
@@ -471,7 +471,7 @@ def __call__(
         entities_pair: EntityInput | list[EntityInput] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         max_entity_length: int | None = None,
         stride: int = 0,
@@ -1171,7 +1171,7 @@ def prepare_for_model(
         pair_entity_token_spans: list[tuple[int, int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         max_entity_length: int | None = None,
         stride: int = 0,
diff --git a/src/transformers/models/mpnet/tokenization_mpnet.py b/src/transformers/models/mpnet/tokenization_mpnet.py
index 3400e31e5ea6..e7ac345fe15f 100644
--- a/src/transformers/models/mpnet/tokenization_mpnet.py
+++ b/src/transformers/models/mpnet/tokenization_mpnet.py
@@ -156,10 +156,10 @@ def __init__(
         sep_token_id = self.sep_token_id if self.sep_token_id is not None else 2
 
         self._tokenizer.post_processor = processors.RobertaProcessing(
-            sep=(sep_str, sep_token_id),
-            cls=(cls_str, cls_token_id),
-            trim_offsets=True,
-            add_prefix_space=False,
+            (sep_str, sep_token_id),
+            (cls_str, cls_token_id),
+            True,  # trim_offsets
+            False,  # add_prefix_space
         )
 
     @property
diff --git a/src/transformers/models/pop2piano/tokenization_pop2piano.py b/src/transformers/models/pop2piano/tokenization_pop2piano.py
index 7c58edaed35e..1964291129f8 100644
--- a/src/transformers/models/pop2piano/tokenization_pop2piano.py
+++ b/src/transformers/models/pop2piano/tokenization_pop2piano.py
@@ -305,7 +305,7 @@ def relative_tokens_ids_to_notes(self, tokens: np.ndarray, start_idx: float, cut
             notes = notes[note_order.argsort()]
             return notes
 
-    def notes_to_midi(self, notes: np.ndarray, beatstep: np.ndarray, offset_sec: int = 0.0):
+    def notes_to_midi(self, notes: np.ndarray, beatstep: np.ndarray, offset_sec: float = 0.0):
         """
         Converts notes to Midi.
 
@@ -474,7 +474,7 @@ def __call__(
         self,
         notes: np.ndarray | list[pretty_midi.Note] | list[list[pretty_midi.Note]],
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         pad_to_multiple_of: int | None = None,
         return_attention_mask: bool | None = None,
diff --git a/src/transformers/models/roberta/tokenization_roberta.py b/src/transformers/models/roberta/tokenization_roberta.py
index 40b4e78426f6..0ab78d76f931 100644
--- a/src/transformers/models/roberta/tokenization_roberta.py
+++ b/src/transformers/models/roberta/tokenization_roberta.py
@@ -168,10 +168,10 @@ def __init__(
             **kwargs,
         )
         self._tokenizer.post_processor = processors.RobertaProcessing(
-            sep=(str(sep_token), self.sep_token_id),
-            cls=(str(cls_token), self.cls_token_id),
-            add_prefix_space=add_prefix_space,
-            trim_offsets=trim_offsets,
+            (str(sep_token), self.sep_token_id),
+            (str(cls_token), self.cls_token_id),
+            trim_offsets,
+            add_prefix_space,
         )
 
 
diff --git a/src/transformers/models/roc_bert/tokenization_roc_bert.py b/src/transformers/models/roc_bert/tokenization_roc_bert.py
index 3663ee35f21b..a9641c93f1b8 100644
--- a/src/transformers/models/roc_bert/tokenization_roc_bert.py
+++ b/src/transformers/models/roc_bert/tokenization_roc_bert.py
@@ -175,7 +175,7 @@ def __call__(
         text_target: TextInput | list[TextInput] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         max_target_length: int | None = None,
         stride: int = 0,
@@ -299,7 +299,7 @@ def encode_plus(
         text_pair: TextInput | PreTokenizedInput | EncodedInput | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         is_split_into_words: bool = False,
@@ -356,7 +356,7 @@ def batch_encode_plus(
         | list[EncodedInputPair],
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         is_split_into_words: bool = False,
@@ -535,7 +535,7 @@ def prepare_for_model(
         pair_pronunciation_ids: list[int] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
diff --git a/src/transformers/models/tapas/tokenization_tapas.py b/src/transformers/models/tapas/tokenization_tapas.py
index e3af74b816df..feacbce4134a 100644
--- a/src/transformers/models/tapas/tokenization_tapas.py
+++ b/src/transformers/models/tapas/tokenization_tapas.py
@@ -2539,7 +2539,7 @@ def parse_text(text):
 
 
 _PrimitiveNumericValue = float | tuple[float | None]
-_SortKeyFn = Callable[[NumericValue], tuple[float, Ellipsis]]
+_SortKeyFn = Callable[[NumericValue], tuple[float, ...]]
 
 _DATE_TUPLE_SIZE = 3
 
diff --git a/src/transformers/models/udop/tokenization_udop.py b/src/transformers/models/udop/tokenization_udop.py
index cf49ebfd7e55..a1bf65af5c31 100644
--- a/src/transformers/models/udop/tokenization_udop.py
+++ b/src/transformers/models/udop/tokenization_udop.py
@@ -260,11 +260,11 @@ def __init__(
     @add_end_docstrings(UDOP_ENCODE_KWARGS_DOCSTRING)
     def __call__(
         self,
-        text: TextInput | PreTokenizedInput | list[TextInput] | list[PreTokenizedInput] = None,
+        text: TextInput | PreTokenizedInput | list[TextInput] | list[PreTokenizedInput] | None = None,
         text_pair: PreTokenizedInput | list[PreTokenizedInput] | None = None,
         boxes: list[list[int]] | list[list[list[int]]] | None = None,
         word_labels: list[int] | list[list[int]] | None = None,
-        text_target: TextInput | PreTokenizedInput | list[TextInput] | list[PreTokenizedInput] = None,
+        text_target: TextInput | PreTokenizedInput | list[TextInput] | list[PreTokenizedInput] | None = None,
         text_pair_target: TextInput | PreTokenizedInput | list[TextInput] | list[PreTokenizedInput] | None = None,
         **kwargs,
     ) -> BatchEncoding:
@@ -305,7 +305,7 @@ def call_boxes(
         word_labels: list[int] | list[list[int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         pad_to_multiple_of: int | None = None,
@@ -467,7 +467,7 @@ def batch_encode_plus_boxes(
         word_labels: list[list[int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         is_split_into_words: bool = False,
@@ -767,7 +767,7 @@ def encode_boxes(
         word_labels: list[list[int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         return_tensors: str | TensorType | None = None,
@@ -810,7 +810,7 @@ def encode_plus_boxes(
         word_labels: list[list[int]] | None = None,
         add_special_tokens: bool = True,
         padding: bool | str | PaddingStrategy = False,
-        truncation: bool | str | TruncationStrategy = None,
+        truncation: bool | str | TruncationStrategy | None = None,
         max_length: int | None = None,
         stride: int = 0,
         is_split_into_words: bool = False,
diff --git a/tests/test_tokenization_common.py b/tests/test_tokenization_common.py
index 4335a8fd920c..64640f5d9f51 100644
--- a/tests/test_tokenization_common.py
+++ b/tests/test_tokenization_common.py
@@ -1713,14 +1713,9 @@ def test_maximum_encoding_length_single_input(self):
         # Overflowing tokens are handled quite differently in slow and fast tokenizers
         if isinstance(tokenizer, TokenizersBackend):
             truncated_sequence = information["input_ids"][0]
-            overflowing_tokens = information["input_ids"][1]
-            self.assertEqual(len(information["input_ids"]), 2)
 
             self.assertEqual(len(truncated_sequence), total_length - 2)
             self.assertEqual(truncated_sequence, sequence[:-2])
-
-            self.assertEqual(len(overflowing_tokens), 2 + stride)
-            self.assertEqual(overflowing_tokens, sequence[-(2 + stride) :])
         else:
             truncated_sequence = information["input_ids"]
             overflowing_tokens = information["overflowing_tokens"]
@@ -1752,8 +1747,6 @@ def test_maximum_encoding_length_pair_input(self):
 
         self.assertGreater(len(seq1_tokens), 2 + stride)
 
-        smallest = seq1_tokens if len(seq0_tokens) > len(seq1_tokens) else seq0_tokens
-
         # We are not using the special tokens - a bit too hard to test all the tokenizers with this
         # TODO try this again later
         sequence = tokenizer.encode(seq_0, seq_1, add_special_tokens=False)  # , add_prefix_space=False)
@@ -1835,9 +1828,6 @@ def test_maximum_encoding_length_pair_input(self):
             tokenizer.encode(seq_0, add_special_tokens=False)
             + tokenizer.encode(seq_1, add_special_tokens=False)[-(2 + stride) :]
         )
-        overflow_longest_sequence = (
-            overflow_first_sequence if len(seq0_tokens) > len(seq1_tokens) else overflow_second_sequence
-        )
 
         # Overflowing tokens are handled quite differently in slow and fast tokenizers
         if isinstance(tokenizer, TokenizersBackend):
@@ -1852,14 +1842,9 @@ def test_maximum_encoding_length_pair_input(self):
                 # add_prefix_space=False,
             )
             truncated_sequence = information["input_ids"][0]
-            overflowing_tokens = information["input_ids"][1]
-            self.assertEqual(len(information["input_ids"]), 2)
 
             self.assertEqual(len(truncated_sequence), len(sequence) - 2)
             self.assertEqual(truncated_sequence, truncated_longest_sequence)
-
-            self.assertEqual(len(overflowing_tokens), 2 + stride + len(smallest))
-            self.assertEqual(overflowing_tokens, overflow_longest_sequence)
         else:
             # No overflowing tokens when using 'longest' in python tokenizers
             with self.assertRaises(ValueError) as context:
@@ -1895,14 +1880,9 @@ def test_maximum_encoding_length_pair_input(self):
                 # add_prefix_space=False,
             )
             truncated_sequence = information["input_ids"][0]
-            overflowing_tokens = information["input_ids"][1]
-            self.assertEqual(len(information["input_ids"]), 2)
 
             self.assertEqual(len(truncated_sequence), len(sequence) - 2)
             self.assertEqual(truncated_sequence, truncated_longest_sequence)
-
-            self.assertEqual(len(overflowing_tokens), 2 + stride + len(smallest))
-            self.assertEqual(overflowing_tokens, overflow_longest_sequence)
         else:
             # No overflowing tokens when using 'longest' in python tokenizers
             with self.assertRaises(ValueError) as context: