From 45fc8c7951f978c0f8f13c8bab52c744cd5c4784 Mon Sep 17 00:00:00 2001
From: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Date: Fri, 9 Apr 2021 11:57:44 -0400
Subject: [PATCH] Make `get_special_tokens_mask` consider all tokens (#11163)

---
 docs/source/model_doc/convbert.rst            |  3 +-
 docs/source/model_doc/led.rst                 |  3 +-
 .../models/albert/tokenization_albert.py      |  9 ++----
 .../models/albert/tokenization_albert_fast.py | 31 ------------------
 .../models/barthez/tokenization_barthez.py    |  9 ++----
 .../barthez/tokenization_barthez_fast.py      | 30 -----------------
 .../models/bert/tokenization_bert.py          |  9 ++----
 .../models/bertweet/tokenization_bertweet.py  |  9 ++----
 .../models/big_bird/tokenization_big_bird.py  |  9 ++----
 .../camembert/tokenization_camembert.py       |  9 ++----
 .../camembert/tokenization_camembert_fast.py  | 30 -----------------
 .../models/deberta/tokenization_deberta.py    |  9 ++----
 .../deberta_v2/tokenization_deberta_v2.py     | 12 ++-----
 .../models/fsmt/tokenization_fsmt.py          | 12 ++-----
 .../herbert/tokenization_herbert_fast.py      |  9 ++----
 .../models/m2m_100/tokenization_m2m_100.py    | 10 +++---
 .../models/mbart/tokenization_mbart.py        | 10 +++---
 .../models/mbart/tokenization_mbart50.py      | 10 +++---
 .../models/mbart/tokenization_mbart50_fast.py | 32 -------------------
 .../models/mbart/tokenization_mbart_fast.py   | 32 -------------------
 .../models/mpnet/tokenization_mpnet.py        |  9 ++----
 .../models/phobert/tokenization_phobert.py    |  9 ++----
 .../prophetnet/tokenization_prophetnet.py     |  9 ++----
 .../models/roberta/tokenization_roberta.py    |  9 ++----
 .../tokenization_speech_to_text.py            | 10 +++---
 src/transformers/models/t5/tokenization_t5.py | 10 +++---
 .../models/tapas/tokenization_tapas.py        |  9 ++----
 .../models/xlm/tokenization_xlm.py            | 12 ++-----
 .../tokenization_xlm_prophetnet.py            |  9 ++----
 .../xlm_roberta/tokenization_xlm_roberta.py   |  9 ++----
 .../tokenization_xlm_roberta_fast.py          | 31 ------------------
 .../models/xlnet/tokenization_xlnet.py        |  9 ++----
 .../models/xlnet/tokenization_xlnet_fast.py   | 31 ------------------
 src/transformers/tokenization_utils.py        | 10 ++++++
 ...on_{{cookiecutter.lowercase_modelname}}.py |  9 ++----
 .../{{cookiecutter.lowercase_modelname}}.rst  |  3 +-
 36 files changed, 90 insertions(+), 385 deletions(-)

diff --git a/docs/source/model_doc/convbert.rst b/docs/source/model_doc/convbert.rst
index 80ed9ebc37..69f7473354 100644
--- a/docs/source/model_doc/convbert.rst
+++ b/docs/source/model_doc/convbert.rst
@@ -56,8 +56,7 @@ ConvBertTokenizerFast
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 .. autoclass:: transformers.ConvBertTokenizerFast
-    :members: build_inputs_with_special_tokens, get_special_tokens_mask,
-        create_token_type_ids_from_sequences, save_vocabulary
+    :members:
 
 
 ConvBertModel
diff --git a/docs/source/model_doc/led.rst b/docs/source/model_doc/led.rst
index 4dbdbbaeb3..83a9386165 100644
--- a/docs/source/model_doc/led.rst
+++ b/docs/source/model_doc/led.rst
@@ -73,8 +73,7 @@ LEDTokenizerFast
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 .. autoclass:: transformers.LEDTokenizerFast
-    :members: build_inputs_with_special_tokens, get_special_tokens_mask,
-        create_token_type_ids_from_sequences, save_vocabulary
+    :members:
 
 
 LED specific outputs
diff --git a/src/transformers/models/albert/tokenization_albert.py b/src/transformers/models/albert/tokenization_albert.py
index a271f86064..92c06bbcde 100644
--- a/src/transformers/models/albert/tokenization_albert.py
+++ b/src/transformers/models/albert/tokenization_albert.py
@@ -267,12 +267,9 @@ class AlbertTokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is not None:
             return [1] + ([0] * len(token_ids_0)) + [1] + ([0] * len(token_ids_1)) + [1]
diff --git a/src/transformers/models/albert/tokenization_albert_fast.py b/src/transformers/models/albert/tokenization_albert_fast.py
index 1d6e82b12d..cb817ddcc0 100644
--- a/src/transformers/models/albert/tokenization_albert_fast.py
+++ b/src/transformers/models/albert/tokenization_albert_fast.py
@@ -184,37 +184,6 @@ class AlbertTokenizerFast(PreTrainedTokenizerFast):
             return cls + token_ids_0 + sep
         return cls + token_ids_0 + sep + token_ids_1 + sep
 
-    def get_special_tokens_mask(
-        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
-    ) -> List[int]:
-        """
-        Retrieves sequence ids from a token list that has no special tokens added. This method is called when adding
-        special tokens using the tokenizer ``prepare_for_model`` method.
-
-        Args:
-            token_ids_0 (:obj:`List[int]`):
-                List of ids.
-            token_ids_1 (:obj:`List[int]`, `optional`, defaults to :obj:`None`):
-                Optional second list of IDs for sequence pairs.
-            already_has_special_tokens (:obj:`bool`, `optional`, defaults to :obj:`False`):
-                Set to True if the token list is already formatted with special tokens for the model
-
-        Returns:
-            :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
-        """
-
-        if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
-
-        if token_ids_1 is not None:
-            return [1] + ([0] * len(token_ids_0)) + [1] + ([0] * len(token_ids_1)) + [1]
-        return [1] + ([0] * len(token_ids_0)) + [1]
-
     def create_token_type_ids_from_sequences(
         self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
     ) -> List[int]:
diff --git a/src/transformers/models/barthez/tokenization_barthez.py b/src/transformers/models/barthez/tokenization_barthez.py
index 428f6fec65..641cc80c1d 100644
--- a/src/transformers/models/barthez/tokenization_barthez.py
+++ b/src/transformers/models/barthez/tokenization_barthez.py
@@ -180,12 +180,9 @@ class BarthezTokenizer(PreTrainedTokenizer):
             :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
         """
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formated with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return [1] + ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/barthez/tokenization_barthez_fast.py b/src/transformers/models/barthez/tokenization_barthez_fast.py
index 1a9610c556..224bfb6453 100644
--- a/src/transformers/models/barthez/tokenization_barthez_fast.py
+++ b/src/transformers/models/barthez/tokenization_barthez_fast.py
@@ -164,36 +164,6 @@ class BarthezTokenizerFast(PreTrainedTokenizerFast):
         sep = [self.sep_token_id]
         return cls + token_ids_0 + sep + sep + token_ids_1 + sep
 
-    def get_special_tokens_mask(
-        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
-    ) -> List[int]:
-        """
-        Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
-        special tokens using the tokenizer ``prepare_for_model`` method.
-
-        Args:
-            token_ids_0 (:obj:`List[int]`):
-                List of IDs.
-            token_ids_1 (:obj:`List[int]`, `optional`):
-                Optional second list of IDs for sequence pairs.
-            already_has_special_tokens (:obj:`bool`, `optional`, defaults to :obj:`False`):
-                Whether or not the token list is already formatted with special tokens for the model.
-
-        Returns:
-            :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
-        """
-        if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formated with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
-
-        if token_ids_1 is None:
-            return [1] + ([0] * len(token_ids_0)) + [1]
-        return [1] + ([0] * len(token_ids_0)) + [1, 1] + ([0] * len(token_ids_1)) + [1]
-
     def create_token_type_ids_from_sequences(
         self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
     ) -> List[int]:
diff --git a/src/transformers/models/bert/tokenization_bert.py b/src/transformers/models/bert/tokenization_bert.py
index 8f3ecfabf6..fbb2cfc029 100644
--- a/src/transformers/models/bert/tokenization_bert.py
+++ b/src/transformers/models/bert/tokenization_bert.py
@@ -290,12 +290,9 @@ class BertTokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is not None:
             return [1] + ([0] * len(token_ids_0)) + [1] + ([0] * len(token_ids_1)) + [1]
diff --git a/src/transformers/models/bertweet/tokenization_bertweet.py b/src/transformers/models/bertweet/tokenization_bertweet.py
index aaeffd7380..bf110274da 100644
--- a/src/transformers/models/bertweet/tokenization_bertweet.py
+++ b/src/transformers/models/bertweet/tokenization_bertweet.py
@@ -220,12 +220,9 @@ class BertweetTokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return [1] + ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/big_bird/tokenization_big_bird.py b/src/transformers/models/big_bird/tokenization_big_bird.py
index 3cafcda189..4d03a0a3ac 100644
--- a/src/transformers/models/big_bird/tokenization_big_bird.py
+++ b/src/transformers/models/big_bird/tokenization_big_bird.py
@@ -219,12 +219,9 @@ class BigBirdTokenizer(PreTrainedTokenizer):
             :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
         """
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return [1] + ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/camembert/tokenization_camembert.py b/src/transformers/models/camembert/tokenization_camembert.py
index eb57acec89..8337d6826c 100644
--- a/src/transformers/models/camembert/tokenization_camembert.py
+++ b/src/transformers/models/camembert/tokenization_camembert.py
@@ -178,12 +178,9 @@ class CamembertTokenizer(PreTrainedTokenizer):
             :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
         """
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return [1] + ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/camembert/tokenization_camembert_fast.py b/src/transformers/models/camembert/tokenization_camembert_fast.py
index 648da8be70..a6333b98d0 100644
--- a/src/transformers/models/camembert/tokenization_camembert_fast.py
+++ b/src/transformers/models/camembert/tokenization_camembert_fast.py
@@ -162,36 +162,6 @@ class CamembertTokenizerFast(PreTrainedTokenizerFast):
         sep = [self.sep_token_id]
         return cls + token_ids_0 + sep + sep + token_ids_1 + sep
 
-    def get_special_tokens_mask(
-        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
-    ) -> List[int]:
-        """
-        Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
-        special tokens using the tokenizer ``prepare_for_model`` method.
-
-        Args:
-            token_ids_0 (:obj:`List[int]`):
-                List of IDs.
-            token_ids_1 (:obj:`List[int]`, `optional`):
-                Optional second list of IDs for sequence pairs.
-            already_has_special_tokens (:obj:`bool`, `optional`, defaults to :obj:`False`):
-                Whether or not the token list is already formatted with special tokens for the model.
-
-        Returns:
-            :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
-        """
-        if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
-
-        if token_ids_1 is None:
-            return [1] + ([0] * len(token_ids_0)) + [1]
-        return [1] + ([0] * len(token_ids_0)) + [1, 1] + ([0] * len(token_ids_1)) + [1]
-
     def create_token_type_ids_from_sequences(
         self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
     ) -> List[int]:
diff --git a/src/transformers/models/deberta/tokenization_deberta.py b/src/transformers/models/deberta/tokenization_deberta.py
index 089c6dc509..ddd08e5286 100644
--- a/src/transformers/models/deberta/tokenization_deberta.py
+++ b/src/transformers/models/deberta/tokenization_deberta.py
@@ -174,12 +174,9 @@ class DebertaTokenizer(GPT2Tokenizer):
             :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
         """
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return [1] + ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/deberta_v2/tokenization_deberta_v2.py b/src/transformers/models/deberta_v2/tokenization_deberta_v2.py
index a0e80f6b00..78509f88d7 100644
--- a/src/transformers/models/deberta_v2/tokenization_deberta_v2.py
+++ b/src/transformers/models/deberta_v2/tokenization_deberta_v2.py
@@ -187,16 +187,8 @@ class DebertaV2Tokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(
-                map(
-                    lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0,
-                    token_ids_0,
-                )
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
             )
 
         if token_ids_1 is not None:
diff --git a/src/transformers/models/fsmt/tokenization_fsmt.py b/src/transformers/models/fsmt/tokenization_fsmt.py
index 124a9541d7..226d18cc3e 100644
--- a/src/transformers/models/fsmt/tokenization_fsmt.py
+++ b/src/transformers/models/fsmt/tokenization_fsmt.py
@@ -437,16 +437,8 @@ class FSMTTokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(
-                map(
-                    lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0,
-                    token_ids_0,
-                )
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
             )
         # no bos used in fairseq
         if token_ids_1 is not None:
diff --git a/src/transformers/models/herbert/tokenization_herbert_fast.py b/src/transformers/models/herbert/tokenization_herbert_fast.py
index 2e5ba1d17a..7a67d5e737 100644
--- a/src/transformers/models/herbert/tokenization_herbert_fast.py
+++ b/src/transformers/models/herbert/tokenization_herbert_fast.py
@@ -126,12 +126,9 @@ class HerbertTokenizerFast(PreTrainedTokenizerFast):
             :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
         """
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return [1] + ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/m2m_100/tokenization_m2m_100.py b/src/transformers/models/m2m_100/tokenization_m2m_100.py
index cbd8a0aa0d..3d2f273d72 100644
--- a/src/transformers/models/m2m_100/tokenization_m2m_100.py
+++ b/src/transformers/models/m2m_100/tokenization_m2m_100.py
@@ -207,12 +207,10 @@ class M2M100Tokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
+
         prefix_ones = [1] * len(self.prefix_tokens)
         suffix_ones = [1] * len(self.suffix_tokens)
         if token_ids_1 is None:
diff --git a/src/transformers/models/mbart/tokenization_mbart.py b/src/transformers/models/mbart/tokenization_mbart.py
index c256132d7e..a38aaf7ef3 100644
--- a/src/transformers/models/mbart/tokenization_mbart.py
+++ b/src/transformers/models/mbart/tokenization_mbart.py
@@ -149,12 +149,10 @@ class MBartTokenizer(XLMRobertaTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
+
         prefix_ones = [1] * len(self.prefix_tokens)
         suffix_ones = [1] * len(self.suffix_tokens)
         if token_ids_1 is None:
diff --git a/src/transformers/models/mbart/tokenization_mbart50.py b/src/transformers/models/mbart/tokenization_mbart50.py
index f5f1a2f60f..5afd9b215f 100644
--- a/src/transformers/models/mbart/tokenization_mbart50.py
+++ b/src/transformers/models/mbart/tokenization_mbart50.py
@@ -241,12 +241,10 @@ class MBart50Tokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
+
         prefix_ones = [1] * len(self.prefix_tokens)
         suffix_ones = [1] * len(self.suffix_tokens)
         if token_ids_1 is None:
diff --git a/src/transformers/models/mbart/tokenization_mbart50_fast.py b/src/transformers/models/mbart/tokenization_mbart50_fast.py
index bda4b7cf36..f22d02e59b 100644
--- a/src/transformers/models/mbart/tokenization_mbart50_fast.py
+++ b/src/transformers/models/mbart/tokenization_mbart50_fast.py
@@ -160,38 +160,6 @@ class MBart50TokenizerFast(PreTrainedTokenizerFast):
         self._src_lang = new_src_lang
         self.set_src_lang_special_tokens(self._src_lang)
 
-    def get_special_tokens_mask(
-        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
-    ) -> List[int]:
-        """
-        Retrieves sequence ids from a token list that has no special tokens added. This method is called when adding
-        special tokens using the tokenizer ``prepare_for_model`` method.
-
-        Args:
-            token_ids_0 (:obj:`List[int]`):
-                List of ids.
-            token_ids_1 (:obj:`List[int]`, `optional`):
-                Optional second list of IDs for sequence pairs.
-            already_has_special_tokens (:obj:`bool`, `optional`, defaults to :obj:`False`):
-                Whether or not the token list is already formatted with special tokens for the model.
-
-        Returns:
-            :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
-        """
-
-        if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
-        prefix_ones = [1] * len(self.prefix_tokens)
-        suffix_ones = [1] * len(self.suffix_tokens)
-        if token_ids_1 is None:
-            return prefix_ones + ([0] * len(token_ids_0)) + suffix_ones
-        return prefix_ones + ([0] * len(token_ids_0)) + ([0] * len(token_ids_1)) + suffix_ones
-
     def build_inputs_with_special_tokens(
         self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
     ) -> List[int]:
diff --git a/src/transformers/models/mbart/tokenization_mbart_fast.py b/src/transformers/models/mbart/tokenization_mbart_fast.py
index e690218315..bbe9ed7d5d 100644
--- a/src/transformers/models/mbart/tokenization_mbart_fast.py
+++ b/src/transformers/models/mbart/tokenization_mbart_fast.py
@@ -131,38 +131,6 @@ class MBartTokenizerFast(XLMRobertaTokenizerFast):
         self._src_lang = new_src_lang
         self.set_src_lang_special_tokens(self._src_lang)
 
-    def get_special_tokens_mask(
-        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
-    ) -> List[int]:
-        """
-        Retrieves sequence ids from a token list that has no special tokens added. This method is called when adding
-        special tokens using the tokenizer ``prepare_for_model`` method.
-
-        Args:
-            token_ids_0 (:obj:`List[int]`):
-                List of ids.
-            token_ids_1 (:obj:`List[int]`, `optional`):
-                Optional second list of IDs for sequence pairs.
-            already_has_special_tokens (:obj:`bool`, `optional`, defaults to :obj:`False`):
-                Whether or not the token list is already formatted with special tokens for the model.
-
-        Returns:
-            :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
-        """
-
-        if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
-        prefix_ones = [1] * len(self.prefix_tokens)
-        suffix_ones = [1] * len(self.suffix_tokens)
-        if token_ids_1 is None:
-            return prefix_ones + ([0] * len(token_ids_0)) + suffix_ones
-        return prefix_ones + ([0] * len(token_ids_0)) + ([0] * len(token_ids_1)) + suffix_ones
-
     def build_inputs_with_special_tokens(
         self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
     ) -> List[int]:
diff --git a/src/transformers/models/mpnet/tokenization_mpnet.py b/src/transformers/models/mpnet/tokenization_mpnet.py
index 125fde68a5..8041ec4ec5 100644
--- a/src/transformers/models/mpnet/tokenization_mpnet.py
+++ b/src/transformers/models/mpnet/tokenization_mpnet.py
@@ -266,12 +266,9 @@ class MPNetTokenizer(PreTrainedTokenizer):
             :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
         """
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formated with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return [1] + ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/phobert/tokenization_phobert.py b/src/transformers/models/phobert/tokenization_phobert.py
index e99e58002e..3caca9012d 100644
--- a/src/transformers/models/phobert/tokenization_phobert.py
+++ b/src/transformers/models/phobert/tokenization_phobert.py
@@ -201,12 +201,9 @@ class PhobertTokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return [1] + ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/prophetnet/tokenization_prophetnet.py b/src/transformers/models/prophetnet/tokenization_prophetnet.py
index cd51662b55..25df78162e 100644
--- a/src/transformers/models/prophetnet/tokenization_prophetnet.py
+++ b/src/transformers/models/prophetnet/tokenization_prophetnet.py
@@ -203,12 +203,9 @@ class ProphetNetTokenizer(PreTrainedTokenizer):
             :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
         """
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/roberta/tokenization_roberta.py b/src/transformers/models/roberta/tokenization_roberta.py
index 9a037d1d15..696868fdfc 100644
--- a/src/transformers/models/roberta/tokenization_roberta.py
+++ b/src/transformers/models/roberta/tokenization_roberta.py
@@ -215,12 +215,9 @@ class RobertaTokenizer(GPT2Tokenizer):
             :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
         """
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return [1] + ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/speech_to_text/tokenization_speech_to_text.py b/src/transformers/models/speech_to_text/tokenization_speech_to_text.py
index bf3402295a..502021d535 100644
--- a/src/transformers/models/speech_to_text/tokenization_speech_to_text.py
+++ b/src/transformers/models/speech_to_text/tokenization_speech_to_text.py
@@ -199,12 +199,10 @@ class Speech2TextTokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.bos_token_id, self.eos_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
+
         prefix_ones = [1] * len(self.prefix_tokens)
         suffix_ones = [1]
         if token_ids_1 is None:
diff --git a/src/transformers/models/t5/tokenization_t5.py b/src/transformers/models/t5/tokenization_t5.py
index 74dc811c6e..90a0159aef 100644
--- a/src/transformers/models/t5/tokenization_t5.py
+++ b/src/transformers/models/t5/tokenization_t5.py
@@ -157,12 +157,10 @@ class T5Tokenizer(PreTrainedTokenizer):
             :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
         """
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
+
         # normal case: some special tokens
         if token_ids_1 is None:
             return ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/tapas/tokenization_tapas.py b/src/transformers/models/tapas/tokenization_tapas.py
index 9716193951..3d1e82ac51 100644
--- a/src/transformers/models/tapas/tokenization_tapas.py
+++ b/src/transformers/models/tapas/tokenization_tapas.py
@@ -510,12 +510,9 @@ class TapasTokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is not None:
             return [1] + ([0] * len(token_ids_0)) + [1] + ([0] * len(token_ids_1))
diff --git a/src/transformers/models/xlm/tokenization_xlm.py b/src/transformers/models/xlm/tokenization_xlm.py
index d861ccc0ed..95730451fd 100644
--- a/src/transformers/models/xlm/tokenization_xlm.py
+++ b/src/transformers/models/xlm/tokenization_xlm.py
@@ -906,16 +906,8 @@ class XLMTokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(
-                map(
-                    lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0,
-                    token_ids_0,
-                )
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
             )
 
         if token_ids_1 is not None:
diff --git a/src/transformers/models/xlm_prophetnet/tokenization_xlm_prophetnet.py b/src/transformers/models/xlm_prophetnet/tokenization_xlm_prophetnet.py
index ba1d160ee2..188292ed17 100644
--- a/src/transformers/models/xlm_prophetnet/tokenization_xlm_prophetnet.py
+++ b/src/transformers/models/xlm_prophetnet/tokenization_xlm_prophetnet.py
@@ -200,12 +200,9 @@ class XLMProphetNetTokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/xlm_roberta/tokenization_xlm_roberta.py b/src/transformers/models/xlm_roberta/tokenization_xlm_roberta.py
index 4549d212ec..8ecec6dffe 100644
--- a/src/transformers/models/xlm_roberta/tokenization_xlm_roberta.py
+++ b/src/transformers/models/xlm_roberta/tokenization_xlm_roberta.py
@@ -206,12 +206,9 @@ class XLMRobertaTokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return [1] + ([0] * len(token_ids_0)) + [1]
diff --git a/src/transformers/models/xlm_roberta/tokenization_xlm_roberta_fast.py b/src/transformers/models/xlm_roberta/tokenization_xlm_roberta_fast.py
index 9426d6c4aa..fbdeca2e1a 100644
--- a/src/transformers/models/xlm_roberta/tokenization_xlm_roberta_fast.py
+++ b/src/transformers/models/xlm_roberta/tokenization_xlm_roberta_fast.py
@@ -172,37 +172,6 @@ class XLMRobertaTokenizerFast(PreTrainedTokenizerFast):
         sep = [self.sep_token_id]
         return cls + token_ids_0 + sep + sep + token_ids_1 + sep
 
-    def get_special_tokens_mask(
-        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
-    ) -> List[int]:
-        """
-        Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
-        special tokens using the tokenizer ``prepare_for_model`` method.
-
-        Args:
-            token_ids_0 (:obj:`List[int]`):
-                List of IDs.
-            token_ids_1 (:obj:`List[int]`, `optional`):
-                Optional second list of IDs for sequence pairs.
-            already_has_special_tokens (:obj:`bool`, `optional`, defaults to :obj:`False`):
-                Whether or not the token list is already formatted with special tokens for the model.
-
-        Returns:
-            :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
-        """
-
-        if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
-
-        if token_ids_1 is None:
-            return [1] + ([0] * len(token_ids_0)) + [1]
-        return [1] + ([0] * len(token_ids_0)) + [1, 1] + ([0] * len(token_ids_1)) + [1]
-
     def create_token_type_ids_from_sequences(
         self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
     ) -> List[int]:
diff --git a/src/transformers/models/xlnet/tokenization_xlnet.py b/src/transformers/models/xlnet/tokenization_xlnet.py
index 4980f450cb..97fd542c28 100644
--- a/src/transformers/models/xlnet/tokenization_xlnet.py
+++ b/src/transformers/models/xlnet/tokenization_xlnet.py
@@ -270,12 +270,9 @@ class XLNetTokenizer(PreTrainedTokenizer):
         """
 
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is not None:
             return ([0] * len(token_ids_0)) + [1] + ([0] * len(token_ids_1)) + [1, 1]
diff --git a/src/transformers/models/xlnet/tokenization_xlnet_fast.py b/src/transformers/models/xlnet/tokenization_xlnet_fast.py
index f3a46c2d78..364dccf3d6 100644
--- a/src/transformers/models/xlnet/tokenization_xlnet_fast.py
+++ b/src/transformers/models/xlnet/tokenization_xlnet_fast.py
@@ -190,37 +190,6 @@ class XLNetTokenizerFast(PreTrainedTokenizerFast):
             return token_ids_0 + sep + cls
         return token_ids_0 + sep + token_ids_1 + sep + cls
 
-    def get_special_tokens_mask(
-        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
-    ) -> List[int]:
-        """
-        Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
-        special tokens using the tokenizer ``prepare_for_model`` method.
-
-        Args:
-            token_ids_0 (:obj:`List[int]`):
-                List of IDs.
-            token_ids_1 (:obj:`List[int]`, `optional`):
-                Optional second list of IDs for sequence pairs.
-            already_has_special_tokens (:obj:`bool`, `optional`, defaults to :obj:`False`):
-                Whether or not the token list is already formatted with special tokens for the model.
-
-        Returns:
-            :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
-        """
-
-        if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
-
-        if token_ids_1 is not None:
-            return ([0] * len(token_ids_0)) + [1] + ([0] * len(token_ids_1)) + [1, 1]
-        return ([0] * len(token_ids_0)) + [1, 1]
-
     def create_token_type_ids_from_sequences(
         self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
     ) -> List[int]:
diff --git a/src/transformers/tokenization_utils.py b/src/transformers/tokenization_utils.py
index 0606c3f9c1..fafe8a5597 100644
--- a/src/transformers/tokenization_utils.py
+++ b/src/transformers/tokenization_utils.py
@@ -670,6 +670,16 @@ class PreTrainedTokenizer(PreTrainedTokenizerBase):
         Returns:
             A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
         """
+        if already_has_special_tokens:
+            if token_ids_1 is not None:
+                raise ValueError(
+                    "You should not supply a second sequence if the provided sequence of "
+                    "ids is already formatted with special tokens for the model."
+                )
+
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
         return [0] * ((len(token_ids_1) if token_ids_1 else 0) + len(token_ids_0))
 
     @overload
diff --git a/templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/tokenization_{{cookiecutter.lowercase_modelname}}.py b/templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/tokenization_{{cookiecutter.lowercase_modelname}}.py
index 7973c1e1dd..ec154a9b1c 100644
--- a/templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/tokenization_{{cookiecutter.lowercase_modelname}}.py
+++ b/templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/tokenization_{{cookiecutter.lowercase_modelname}}.py
@@ -225,12 +225,9 @@ class {{cookiecutter.camelcase_modelname}}Tokenizer(PreTrainedTokenizer):
             :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
         """
         if already_has_special_tokens:
-            if token_ids_1 is not None:
-                raise ValueError(
-                    "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formatted with special tokens for the model."
-                )
-            return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
 
         if token_ids_1 is None:
             return [1] + ([0] * len(token_ids_0)) + [1]
diff --git a/templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/{{cookiecutter.lowercase_modelname}}.rst b/templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/{{cookiecutter.lowercase_modelname}}.rst
index 67384736e7..7510fe44e9 100644
--- a/templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/{{cookiecutter.lowercase_modelname}}.rst
+++ b/templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/{{cookiecutter.lowercase_modelname}}.rst
@@ -46,8 +46,7 @@ Tips:
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 .. autoclass:: transformers.{{cookiecutter.camelcase_modelname}}TokenizerFast
-    :members: build_inputs_with_special_tokens, get_special_tokens_mask,
-        create_token_type_ids_from_sequences, save_vocabulary
+    :members:
 
 
 {% if "PyTorch" in cookiecutter.generate_tensorflow_and_pytorch -%}