Add a check regarding the number of occurrences of ``` (#18389)

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

src/transformers/modeling_tf_utils.py

@@ -1879,7 +1879,7 @@ class TFPreTrainedModel(tf.keras.Model, TFModelUtilsMixin, TFGenerationMixin, Pu
 
                 Increasing the size will add newly initialized vectors at the end. Reducing the size will remove
                 vectors from the end. If not provided or `None`, just returns a pointer to the input tokens
-                ``tf.Variable``` module of the model without doing anything.
+                `tf.Variable` module of the model without doing anything.
 
         Return:
             `tf.Variable`: Pointer to the resized Embedding Module or the old Embedding Module if `new_num_tokens` is

src/transformers/modeling_utils.py

@@ -1221,7 +1221,7 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, GenerationMixin, PushToHubMix
 
                 Increasing the size will add newly initialized vectors at the end. Reducing the size will remove
                 vectors from the end. If not provided or `None`, just returns a pointer to the input tokens
-                ``torch.nn.Embedding``` module of the model without doing anything.
+                `torch.nn.Embedding` module of the model without doing anything.
 
         Return:
             `torch.nn.Embedding`: Pointer to the resized Embedding Module or the old Embedding Module if
@@ -1285,9 +1285,9 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, GenerationMixin, PushToHubMix
 
                 Increasing the size will add newly initialized vectors at the end. Reducing the size will remove
                 vectors from the end. If not provided or `None`, just returns a pointer to the input tokens
-                ``torch.nn.Linear``` module of the model without doing anything. transposed (`bool`, *optional*,
-                defaults to `False`): Whether `old_lm_head` is transposed or not. If True `old_lm_head.size()` is
-                `lm_head_dim, vocab_size` else `vocab_size, lm_head_dim`.
+                `torch.nn.Linear` module of the model without doing anything. transposed (`bool`, *optional*, defaults
+                to `False`): Whether `old_lm_head` is transposed or not. If True `old_lm_head.size()` is `lm_head_dim,
+                vocab_size` else `vocab_size, lm_head_dim`.
 
         Return:
             `torch.nn.Linear`: Pointer to the resized Linear Module or the old Linear Module if `new_num_tokens` is

src/transformers/models/bart/modeling_tf_bart.py

@@ -910,11 +910,11 @@ class TFBartDecoder(tf.keras.layers.Layer):
 
                 If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                 that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
-                all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of
-                shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
-                `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more
-                control over how to convert `input_ids` indices into associated vectors than the model's internal
-                embedding lookup matrix.
+                all `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of shape
+                `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids`
+                you can choose to directly pass an embedded representation. This is useful if you want more control
+                over how to convert `input_ids` indices into associated vectors than the model's internal embedding
+                lookup matrix.
             output_attentions (`bool`, *optional*):
                 Whether or not to return the attentions tensors of all attention layers. See `attentions` under
                 returned tensors for more detail.

src/transformers/models/blenderbot/modeling_tf_blenderbot.py

@@ -894,11 +894,11 @@ class TFBlenderbotDecoder(tf.keras.layers.Layer):
 
                 If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                 that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
-                all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of
-                shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
-                `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more
-                control over how to convert `input_ids` indices into associated vectors than the model's internal
-                embedding lookup matrix.
+                all `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of shape
+                `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids`
+                you can choose to directly pass an embedded representation. This is useful if you want more control
+                over how to convert `input_ids` indices into associated vectors than the model's internal embedding
+                lookup matrix.
             output_attentions (`bool`, *optional*):
                 Whether or not to return the attentions tensors of all attention layers. See `attentions` under
                 returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value

src/transformers/models/blenderbot_small/modeling_tf_blenderbot_small.py

@@ -898,11 +898,11 @@ class TFBlenderbotSmallDecoder(tf.keras.layers.Layer):
 
                 If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                 that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
-                all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of
-                shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
-                `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more
-                control over how to convert `input_ids` indices into associated vectors than the model's internal
-                embedding lookup matrix.
+                all `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of shape
+                `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids`
+                you can choose to directly pass an embedded representation. This is useful if you want more control
+                over how to convert `input_ids` indices into associated vectors than the model's internal embedding
+                lookup matrix.
             output_attentions (`bool`, *optional*):
                 Whether or not to return the attentions tensors of all attention layers. See `attentions` under
                 returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value

src/transformers/models/deberta/modeling_deberta.py

@@ -825,7 +825,7 @@ DEBERTA_START_DOCSTRING = r"""
 
     This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
     Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
-    and behavior.```
+    and behavior.
 
 
     Parameters:

src/transformers/models/deberta_v2/modeling_deberta_v2.py

@@ -920,7 +920,7 @@ DEBERTA_START_DOCSTRING = r"""
 
     This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
     Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
-    and behavior.```
+    and behavior.
 
 
     Parameters:

src/transformers/models/dpr/tokenization_dpr.py

@@ -297,7 +297,7 @@ class CustomDPRReaderTokenizerMixin:
               spans in the same passage. It corresponds to the sum of the start and end logits of the span.
             - **relevance_score**: `float` that corresponds to the score of the each passage to answer the question,
               compared to all the other passages. It corresponds to the output of the QA classifier of the DPRReader.
-            - **doc_id**: ``int``` the id of the passage. - **start_index**: `int` the start index of the span
+            - **doc_id**: `int` the id of the passage. - **start_index**: `int` the start index of the span
               (inclusive). - **end_index**: `int` the end index of the span (inclusive).
 
         Examples:

src/transformers/models/dpr/tokenization_dpr_fast.py

@@ -297,7 +297,7 @@ class CustomDPRReaderTokenizerMixin:
               spans in the same passage. It corresponds to the sum of the start and end logits of the span.
             - **relevance_score**: `float` that corresponds to the score of the each passage to answer the question,
               compared to all the other passages. It corresponds to the output of the QA classifier of the DPRReader.
-            - **doc_id**: ``int``` the id of the passage. - ***start_index**: `int` the start index of the span
+            - **doc_id**: `int` the id of the passage. - ***start_index**: `int` the start index of the span
               (inclusive). - **end_index**: `int` the end index of the span (inclusive).
 
         Examples:

src/transformers/models/led/modeling_led.py

@@ -2009,8 +2009,8 @@ class LEDDecoder(LEDPreTrainedModel):
 
                 If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                 that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
-                all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor`
-                of shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
+                all `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor` of
+                shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
                 `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more
                 control over how to convert `input_ids` indices into associated vectors than the model's internal
                 embedding lookup matrix.

src/transformers/models/led/modeling_tf_led.py

@@ -1991,7 +1991,7 @@ class TFLEDDecoder(tf.keras.layers.Layer):
                 Contains precomputed key and value hidden-states of the attention blocks. Can be used to speed up
                 decoding. If `past_key_values` are used, the user can optionally input only the last
                 `decoder_input_ids` (those that don't have their past key value states given to this model) of shape
-                `(batch_size, 1)` instead of all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`.
+                `(batch_size, 1)` instead of all `decoder_input_ids` of shape `(batch_size, sequence_length)`.
                 inputs_embeds (`tf.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
                 Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation.
                 This is useful if you want more control over how to convert `input_ids` indices into associated vectors

src/transformers/models/m2m_100/modeling_m2m_100.py

@@ -646,11 +646,10 @@ M2M_100_INPUTS_DOCSTRING = r"""
 
             If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that
             don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
-            ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor` of
-            shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids`
-            you can choose to directly pass an embedded representation. This is useful if you want more control over
-            how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup
-            matrix.
+            `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor` of shape
+            `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids` you
+            can choose to directly pass an embedded representation. This is useful if you want more control over how to
+            convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix.
         decoder_inputs_embeds (`torch.FloatTensor` of shape `(batch_size, target_sequence_length, hidden_size)`, *optional*):
             Optionally, instead of passing `decoder_input_ids` you can choose to directly pass an embedded
             representation. If `past_key_values` is used, optionally only the last `decoder_inputs_embeds` have to be
@@ -952,8 +951,8 @@ class M2M100Decoder(M2M100PreTrainedModel):
 
                 If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                 that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
-                all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor`
-                of shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
+                all `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor` of
+                shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
                 `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more
                 control over how to convert `input_ids` indices into associated vectors than the model's internal
                 embedding lookup matrix.

src/transformers/models/marian/modeling_tf_marian.py

@@ -937,11 +937,11 @@ class TFMarianDecoder(tf.keras.layers.Layer):
 
                 If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                 that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
-                all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of
-                shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
-                `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more
-                control over how to convert `input_ids` indices into associated vectors than the model's internal
-                embedding lookup matrix.
+                all `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of shape
+                `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids`
+                you can choose to directly pass an embedded representation. This is useful if you want more control
+                over how to convert `input_ids` indices into associated vectors than the model's internal embedding
+                lookup matrix.
             output_attentions (`bool`, *optional*):
                 Whether or not to return the attentions tensors of all attention layers. See `attentions` under
                 returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value

src/transformers/models/mbart/modeling_tf_mbart.py

@@ -927,11 +927,11 @@ class TFMBartDecoder(tf.keras.layers.Layer):
 
                 If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                 that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
-                all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of
-                shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
-                `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more
-                control over how to convert `input_ids` indices into associated vectors than the model's internal
-                embedding lookup matrix.
+                all `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of shape
+                `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids`
+                you can choose to directly pass an embedded representation. This is useful if you want more control
+                over how to convert `input_ids` indices into associated vectors than the model's internal embedding
+                lookup matrix.
             output_attentions (`bool`, *optional*):
                 Whether or not to return the attentions tensors of all attention layers. See `attentions` under
                 returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value

src/transformers/models/mbart/tokenization_mbart.py

@@ -57,8 +57,8 @@ class MBartTokenizer(PreTrainedTokenizer):
     Adapted from [`RobertaTokenizer`] and [`XLNetTokenizer`]. Based on
     [SentencePiece](https://github.com/google/sentencepiece).
 
-    The tokenization method is `<tokens> <eos> <language code>` for source language documents, and ``<language code>
-    <tokens> <eos>``` for target language documents.
+    The tokenization method is `<tokens> <eos> <language code>` for source language documents, and `<language code>
+    <tokens> <eos>` for target language documents.
 
     Examples:
 

src/transformers/models/mbart/tokenization_mbart_fast.py

@@ -68,8 +68,8 @@ class MBartTokenizerFast(PreTrainedTokenizerFast):
     This tokenizer inherits from [`PreTrainedTokenizerFast`] which contains most of the main methods. Users should
     refer to this superclass for more information regarding those methods.
 
-    The tokenization method is `<tokens> <eos> <language code>` for source language documents, and ``<language code>
-    <tokens> <eos>``` for target language documents.
+    The tokenization method is `<tokens> <eos> <language code>` for source language documents, and `<language code>
+    <tokens> <eos>` for target language documents.
 
     Examples:
 

src/transformers/models/opt/modeling_tf_opt.py

@@ -598,7 +598,7 @@ class TFOPTDecoder(tf.keras.layers.Layer):
 
                 If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                 that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
-                all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`.
+                all `decoder_input_ids` of shape `(batch_size, sequence_length)`.
             inputs_embeds (`tf.Tensor` of
                 shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
                 `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more

src/transformers/models/pegasus/modeling_tf_pegasus.py

@@ -943,11 +943,11 @@ class TFPegasusDecoder(tf.keras.layers.Layer):
 
                 If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                 that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
-                all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of
-                shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
-                `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more
-                control over how to convert `input_ids` indices into associated vectors than the model's internal
-                embedding lookup matrix.
+                all `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of shape
+                `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids`
+                you can choose to directly pass an embedded representation. This is useful if you want more control
+                over how to convert `input_ids` indices into associated vectors than the model's internal embedding
+                lookup matrix.
             output_attentions (`bool`, *optional*):
                 Whether or not to return the attentions tensors of all attention layers. See `attentions` under
                 returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value

src/transformers/models/plbart/tokenization_plbart.py

@@ -100,8 +100,8 @@ class PLBartTokenizer(PreTrainedTokenizer):
     Adapted from [`RobertaTokenizer`] and [`XLNetTokenizer`]. Based on
     [SentencePiece](https://github.com/google/sentencepiece).
 
-    The tokenization method is `<tokens> <eos> <language code>` for source language documents, and ``<language code>
-    <tokens> <eos>``` for target language documents.
+    The tokenization method is `<tokens> <eos> <language code>` for source language documents, and `<language code>
+    <tokens> <eos>` for target language documents.
 
     Args:
         vocab_file (`str`):

src/transformers/models/retribert/modeling_retribert.py

@@ -201,7 +201,7 @@ class RetriBertModel(RetriBertPreTrainedModel):
                 Indices of input sequence tokens in the vocabulary for the documents in a batch.
             attention_mask_doc (`torch.FloatTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Mask to avoid performing attention on documents padding token indices.
-            checkpoint_batch_size (`int`, *optional*, defaults to ```-1`):
+            checkpoint_batch_size (`int`, *optional*, defaults to `-1`):
                 If greater than 0, uses gradient checkpointing to only compute sequence representation on
                 `checkpoint_batch_size` examples at a time on the GPU. All query representations are still compared to
                 all document representations in the batch.

src/transformers/models/speech_to_text/modeling_speech_to_text.py

@@ -663,8 +663,8 @@ SPEECH_TO_TEXT_INPUTS_DOCSTRING = r"""
 
             If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that
             don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
-            ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. decoder_inputs_embeds (`torch.FloatTensor`
-            of shape `(batch_size, target_sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
+            `decoder_input_ids` of shape `(batch_size, sequence_length)`. decoder_inputs_embeds (`torch.FloatTensor` of
+            shape `(batch_size, target_sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
             `decoder_input_ids` you can choose to directly pass an embedded representation. If `past_key_values` is
             used, optionally only the last `decoder_inputs_embeds` have to be input (see `past_key_values`). This is
             useful if you want more control over how to convert `decoder_input_ids` indices into associated vectors
@@ -965,8 +965,8 @@ class Speech2TextDecoder(Speech2TextPreTrainedModel):
 
                 If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                 that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
-                all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor`
-                of shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
+                all `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor` of
+                shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
                 `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more
                 control over how to convert `input_ids` indices into associated vectors than the model's internal
                 embedding lookup matrix.

src/transformers/models/speech_to_text/modeling_tf_speech_to_text.py

@@ -1002,11 +1002,11 @@ class TFSpeech2TextDecoder(tf.keras.layers.Layer):
 
                 If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                 that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
-                all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of
-                shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
-                `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more
-                control over how to convert `input_ids` indices into associated vectors than the model's internal
-                embedding lookup matrix.
+                all `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`tf.Tensor` of shape
+                `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids`
+                you can choose to directly pass an embedded representation. This is useful if you want more control
+                over how to convert `input_ids` indices into associated vectors than the model's internal embedding
+                lookup matrix.
             output_attentions (`bool`, *optional*):
                 Whether or not to return the attentions tensors of all attention layers. See `attentions` under
                 returned tensors for more detail.

src/transformers/models/speech_to_text_2/modeling_speech_to_text_2.py

@@ -572,8 +572,8 @@ class Speech2Text2Decoder(Speech2Text2PreTrainedModel):
 
                 If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                 that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
-                all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor`
-                of shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
+                all `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor` of
+                shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing
                 `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more
                 control over how to convert `input_ids` indices into associated vectors than the model's internal
                 embedding lookup matrix.

src/transformers/models/xglm/modeling_xglm.py

@@ -90,11 +90,11 @@ XGLM_INPUTS_DOCSTRING = r"""
             blocks) that can be used (see `past_key_values` input) to speed up sequential decoding.
 
             If `past_key_values` are used, the user can optionally input only the last `input_ids` (those that don't
-            have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
-            ``input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor` of shape
-            `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids` you
-            can choose to directly pass an embedded representation. This is useful if you want more control over how to
-            convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix.
+            have their past key value states given to this model) of shape `(batch_size, 1)` instead of all `input_ids`
+            of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor` of shape `(batch_size,
+            sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids` you can choose to
+            directly pass an embedded representation. This is useful if you want more control over how to convert
+            `input_ids` indices into associated vectors than the model's internal embedding lookup matrix.
         inputs_embeds (`torch.FloatTensor` of shape `(batch_size, target_sequence_length, hidden_size)`, *optional*):
             Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. If
             `past_key_values` is used, optionally only the last `inputs_embeds` have to be input (see

templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/modeling_{{cookiecutter.lowercase_modelname}}.py

@@ -2136,7 +2136,7 @@ class {{cookiecutter.camelcase_modelname}}PreTrainedModel(PreTrainedModel):
 
             If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids`
             (those that don't have their past key value states given to this model) of shape `(batch_size, 1)`
-            instead of all ``decoder_input_ids``` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated
+            instead of all `decoder_input_ids` of shape `(batch_size, sequence_length)`. inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated
             vectors than the model's internal embedding lookup matrix.
         decoder_inputs_embeds (`torch.FloatTensor` of shape `(batch_size, target_sequence_length, hidden_size)`, *optional*):
             Optionally, instead of passing `decoder_input_ids` you can choose to directly pass an embedded
@@ -2483,7 +2483,7 @@ class {{cookiecutter.camelcase_modelname}}Decoder({{cookiecutter.camelcase_model
 
                 If `past_key_values` are used, the user can optionally input only the last
                 `decoder_input_ids` (those that don't have their past key value states given to this model) of
-                shape `(batch_size, 1)` instead of all ``decoder_input_ids``` of shape `(batch_size,
+                shape `(batch_size, 1)` instead of all `decoder_input_ids` of shape `(batch_size,
                 sequence_length)`. inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*): Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices
                 into associated vectors than the model's internal embedding lookup matrix.
             output_attentions (`bool`, *optional*):

utils/prepare_for_doc_test.py

@@ -92,6 +92,9 @@ def process_doc_file(code_file, add_new_line=True):
 
     # fmt: off
     splits = code.split("```")
+    if len(splits) % 2 != 1:
+        raise ValueError("The number of occurrences of ``` should be an even number.")
+
     splits = [s if i % 2 == 0 else process_code_block(s, add_new_line=add_new_line) for i, s in enumerate(splits)]
     clean_code = "```".join(splits)
     # fmt: on