[Docs] More general docstrings (#14028)

* up

* finish

* up

* up

* finish

src/transformers/models/longformer/modeling_tf_longformer.py

@@ -2088,7 +2088,7 @@ class TFLongformerForMaskedLM(TFLongformerPreTrainedModel, TFMaskedLanguageModel
 
     @add_start_docstrings_to_model_forward(LONGFORMER_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFLongformerMaskedLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -2197,7 +2197,7 @@ class TFLongformerForQuestionAnswering(TFLongformerPreTrainedModel, TFQuestionAn
 
     @add_start_docstrings_to_model_forward(LONGFORMER_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="allenai/longformer-large-4096-finetuned-triviaqa",
         output_type=TFLongformerQuestionAnsweringModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -2366,7 +2366,7 @@ class TFLongformerForSequenceClassification(TFLongformerPreTrainedModel, TFSeque
 
     @add_start_docstrings_to_model_forward(LONGFORMER_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFLongformerSequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -2492,7 +2492,7 @@ class TFLongformerForMultipleChoice(TFLongformerPreTrainedModel, TFMultipleChoic
         LONGFORMER_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length")
     )
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFLongformerMultipleChoiceModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -2645,7 +2645,7 @@ class TFLongformerForTokenClassification(TFLongformerPreTrainedModel, TFTokenCla
 
     @add_start_docstrings_to_model_forward(LONGFORMER_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFLongformerTokenClassifierOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/lxmert/modeling_lxmert.py

@@ -901,7 +901,7 @@ class LxmertModel(LxmertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(LXMERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=LxmertModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1384,7 +1384,7 @@ class LxmertForQuestionAnswering(LxmertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(LXMERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=LxmertForQuestionAnsweringOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/lxmert/modeling_tf_lxmert.py

@@ -950,7 +950,7 @@ class TFLxmertModel(TFLxmertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(LXMERT_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFLxmertModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/m2m_100/modeling_m2m_100.py

@@ -1123,7 +1123,7 @@ class M2M100Model(M2M100PreTrainedModel):
 
     @add_start_docstrings_to_model_forward(M2M_100_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=Seq2SeqModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/marian/modeling_tf_marian.py

@@ -1224,7 +1224,7 @@ class TFMarianModel(TFMarianPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MARIAN_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFSeq2SeqModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/mbart/modeling_mbart.py

@@ -1134,7 +1134,7 @@ class MBartModel(MBartPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MBART_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=Seq2SeqModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1405,7 +1405,7 @@ class MBartForSequenceClassification(MBartPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MBART_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=Seq2SeqSequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1518,7 +1518,7 @@ class MBartForQuestionAnswering(MBartPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MBART_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=Seq2SeqQuestionAnsweringModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/mbart/modeling_tf_mbart.py

@@ -1208,7 +1208,7 @@ class TFMBartModel(TFMBartPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MBART_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFSeq2SeqModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/megatron_bert/modeling_megatron_bert.py

@@ -873,7 +873,7 @@ class MegatronBertModel(MegatronBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MEGATRON_BERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=BaseModelOutputWithPoolingAndCrossAttentions,
         config_class=_CONFIG_FOR_DOC,
@@ -1282,7 +1282,7 @@ class MegatronBertForMaskedLM(MegatronBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MEGATRON_BERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=MaskedLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1480,7 +1480,7 @@ class MegatronBertForSequenceClassification(MegatronBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MEGATRON_BERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=SequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1566,7 +1566,7 @@ class MegatronBertForMultipleChoice(MegatronBertPreTrainedModel):
         MEGATRON_BERT_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length")
     )
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=MultipleChoiceModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1661,7 +1661,7 @@ class MegatronBertForTokenClassification(MegatronBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MEGATRON_BERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TokenClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1751,7 +1751,7 @@ class MegatronBertForQuestionAnswering(MegatronBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MEGATRON_BERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=QuestionAnsweringModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/mobilebert/modeling_mobilebert.py

@@ -817,7 +817,7 @@ class MobileBertModel(MobileBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MOBILEBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=BaseModelOutputWithPooling,
         config_class=_CONFIG_FOR_DOC,
@@ -1032,7 +1032,7 @@ class MobileBertForMaskedLM(MobileBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MOBILEBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=MaskedLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1222,7 +1222,7 @@ class MobileBertForSequenceClassification(MobileBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MOBILEBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=SequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1322,7 +1322,7 @@ class MobileBertForQuestionAnswering(MobileBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MOBILEBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=QuestionAnsweringModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1427,7 +1427,7 @@ class MobileBertForMultipleChoice(MobileBertPreTrainedModel):
         MOBILEBERT_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length")
     )
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=MultipleChoiceModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1526,7 +1526,7 @@ class MobileBertForTokenClassification(MobileBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MOBILEBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TokenClassifierOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/mobilebert/modeling_tf_mobilebert.py

@@ -933,7 +933,7 @@ class TFMobileBertModel(TFMobileBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MOBILEBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFBaseModelOutputWithPooling,
         config_class=_CONFIG_FOR_DOC,
@@ -1124,7 +1124,7 @@ class TFMobileBertForMaskedLM(TFMobileBertPreTrainedModel, TFMaskedLanguageModel
 
     @add_start_docstrings_to_model_forward(MOBILEBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFMaskedLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1348,7 +1348,7 @@ class TFMobileBertForSequenceClassification(TFMobileBertPreTrainedModel, TFSeque
 
     @add_start_docstrings_to_model_forward(MOBILEBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFSequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1456,7 +1456,7 @@ class TFMobileBertForQuestionAnswering(TFMobileBertPreTrainedModel, TFQuestionAn
 
     @add_start_docstrings_to_model_forward(MOBILEBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFQuestionAnsweringModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1591,7 +1591,7 @@ class TFMobileBertForMultipleChoice(TFMobileBertPreTrainedModel, TFMultipleChoic
         MOBILEBERT_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length")
     )
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFMultipleChoiceModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1742,7 +1742,7 @@ class TFMobileBertForTokenClassification(TFMobileBertPreTrainedModel, TFTokenCla
 
     @add_start_docstrings_to_model_forward(MOBILEBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFTokenClassifierOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/mpnet/modeling_mpnet.py

@@ -511,7 +511,7 @@ class MPNetModel(MPNetPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=BaseModelOutputWithPooling,
         config_class=_CONFIG_FOR_DOC,
@@ -593,7 +593,7 @@ class MPNetForMaskedLM(MPNetPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=MaskedLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -695,7 +695,7 @@ class MPNetForSequenceClassification(MPNetPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=SequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -777,7 +777,7 @@ class MPNetForMultipleChoice(MPNetPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=MultipleChoiceModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -869,7 +869,7 @@ class MPNetForTokenClassification(MPNetPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TokenClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -977,7 +977,7 @@ class MPNetForQuestionAnswering(MPNetPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=QuestionAnsweringModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/mpnet/modeling_tf_mpnet.py

@@ -684,7 +684,7 @@ class TFMPNetModel(TFMPNetPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFBaseModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -813,7 +813,7 @@ class TFMPNetForMaskedLM(TFMPNetPreTrainedModel, TFMaskedLanguageModelingLoss):
 
     @add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFMaskedLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -934,7 +934,7 @@ class TFMPNetForSequenceClassification(TFMPNetPreTrainedModel, TFSequenceClassif
 
     @add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFSequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1040,7 +1040,7 @@ class TFMPNetForMultipleChoice(TFMPNetPreTrainedModel, TFMultipleChoiceLoss):
 
     @add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFMultipleChoiceModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1172,7 +1172,7 @@ class TFMPNetForTokenClassification(TFMPNetPreTrainedModel, TFTokenClassificatio
 
     @add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFTokenClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1271,7 +1271,7 @@ class TFMPNetForQuestionAnswering(TFMPNetPreTrainedModel, TFQuestionAnsweringLos
 
     @add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFQuestionAnsweringModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/openai/modeling_openai.py

@@ -433,7 +433,7 @@ class OpenAIGPTModel(OpenAIGPTPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(OPENAI_GPT_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=BaseModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -552,7 +552,7 @@ class OpenAIGPTLMHeadModel(OpenAIGPTPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(OPENAI_GPT_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=CausalLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -756,7 +756,7 @@ class OpenAIGPTForSequenceClassification(OpenAIGPTPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(OPENAI_GPT_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=SequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/openai/modeling_tf_openai.py

@@ -522,7 +522,7 @@ class TFOpenAIGPTModel(TFOpenAIGPTPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(OPENAI_GPT_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFBaseModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -598,7 +598,7 @@ class TFOpenAIGPTLMHeadModel(TFOpenAIGPTPreTrainedModel, TFCausalLanguageModelin
 
     @add_start_docstrings_to_model_forward(OPENAI_GPT_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFCausalLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -856,7 +856,7 @@ class TFOpenAIGPTForSequenceClassification(TFOpenAIGPTPreTrainedModel, TFSequenc
 
     @add_start_docstrings_to_model_forward(OPENAI_GPT_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFSequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/pegasus/modeling_tf_pegasus.py

@@ -1233,7 +1233,7 @@ class TFPegasusModel(TFPegasusPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(PEGASUS_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFSeq2SeqModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/reformer/modeling_reformer.py

@@ -1992,7 +1992,7 @@ class ReformerModel(ReformerPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(REFORMER_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=ReformerModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -2198,7 +2198,7 @@ class ReformerModelWithLMHead(ReformerPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(REFORMER_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=CausalLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -2313,7 +2313,7 @@ class ReformerForMaskedLM(ReformerPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(REFORMER_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=MaskedLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -2394,7 +2394,7 @@ class ReformerForSequenceClassification(ReformerPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(REFORMER_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=SequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -2512,7 +2512,7 @@ class ReformerForQuestionAnswering(ReformerPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(REFORMER_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=QuestionAnsweringModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/rembert/modeling_rembert.py

@@ -781,7 +781,7 @@ class RemBertModel(RemBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(REMBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=BaseModelOutputWithPastAndCrossAttentions,
         config_class=_CONFIG_FOR_DOC,
@@ -933,7 +933,7 @@ class RemBertForMaskedLM(RemBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(REMBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=MaskedLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1175,7 +1175,7 @@ class RemBertForSequenceClassification(RemBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(REMBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=SequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1259,7 +1259,7 @@ class RemBertForMultipleChoice(RemBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(REMBERT_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=MultipleChoiceModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1351,7 +1351,7 @@ class RemBertForTokenClassification(RemBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(REMBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=TokenClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1439,7 +1439,7 @@ class RemBertForQuestionAnswering(RemBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(REMBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=QuestionAnsweringModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/rembert/modeling_tf_rembert.py

@@ -956,7 +956,7 @@ class TFRemBertModel(TFRemBertPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(REMBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=TFBaseModelOutputWithPoolingAndCrossAttentions,
         config_class=_CONFIG_FOR_DOC,
@@ -1078,7 +1078,7 @@ class TFRemBertForMaskedLM(TFRemBertPreTrainedModel, TFMaskedLanguageModelingLos
 
     @add_start_docstrings_to_model_forward(REMBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=TFMaskedLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1186,7 +1186,7 @@ class TFRemBertForCausalLM(TFRemBertPreTrainedModel, TFCausalLanguageModelingLos
         }
 
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=TFCausalLMOutputWithCrossAttentions,
         config_class=_CONFIG_FOR_DOC,
@@ -1329,7 +1329,7 @@ class TFRemBertForSequenceClassification(TFRemBertPreTrainedModel, TFSequenceCla
 
     @add_start_docstrings_to_model_forward(REMBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=TFSequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1435,7 +1435,7 @@ class TFRemBertForMultipleChoice(TFRemBertPreTrainedModel, TFMultipleChoiceLoss)
 
     @add_start_docstrings_to_model_forward(REMBERT_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=TFMultipleChoiceModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1579,7 +1579,7 @@ class TFRemBertForTokenClassification(TFRemBertPreTrainedModel, TFTokenClassific
 
     @add_start_docstrings_to_model_forward(REMBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=TFTokenClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1675,7 +1675,7 @@ class TFRemBertForQuestionAnswering(TFRemBertPreTrainedModel, TFQuestionAnswerin
 
     @add_start_docstrings_to_model_forward(REMBERT_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint="rembert",
         output_type=TFQuestionAnsweringModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/roberta/modeling_roberta.py

@@ -739,7 +739,7 @@ class RobertaModel(RobertaPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=BaseModelOutputWithPoolingAndCrossAttentions,
         config_class=_CONFIG_FOR_DOC,
@@ -1058,7 +1058,7 @@ class RobertaForMaskedLM(RobertaPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=MaskedLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1171,7 +1171,7 @@ class RobertaForSequenceClassification(RobertaPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=SequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1267,7 +1267,7 @@ class RobertaForMultipleChoice(RobertaPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=MultipleChoiceModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1364,7 +1364,7 @@ class RobertaForTokenClassification(RobertaPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TokenClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1476,7 +1476,7 @@ class RobertaForQuestionAnswering(RobertaPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=QuestionAnsweringModelOutput,
         config_class=_CONFIG_FOR_DOC,

src/transformers/models/roberta/modeling_tf_roberta.py

@@ -933,7 +933,7 @@ class TFRobertaModel(TFRobertaPreTrainedModel):
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFBaseModelOutputWithPoolingAndCrossAttentions,
         config_class=_CONFIG_FOR_DOC,
@@ -1108,7 +1108,7 @@ class TFRobertaForMaskedLM(TFRobertaPreTrainedModel, TFMaskedLanguageModelingLos
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFMaskedLMOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1222,7 +1222,7 @@ class TFRobertaForCausalLM(TFRobertaPreTrainedModel, TFCausalLanguageModelingLos
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFCausalLMOutputWithCrossAttentions,
         config_class=_CONFIG_FOR_DOC,
@@ -1392,7 +1392,7 @@ class TFRobertaForSequenceClassification(TFRobertaPreTrainedModel, TFSequenceCla
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFSequenceClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1503,7 +1503,7 @@ class TFRobertaForMultipleChoice(TFRobertaPreTrainedModel, TFMultipleChoiceLoss)
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFMultipleChoiceModelOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1641,7 +1641,7 @@ class TFRobertaForTokenClassification(TFRobertaPreTrainedModel, TFTokenClassific
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFTokenClassifierOutput,
         config_class=_CONFIG_FOR_DOC,
@@ -1742,7 +1742,7 @@ class TFRobertaForQuestionAnswering(TFRobertaPreTrainedModel, TFQuestionAnswerin
 
     @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
     @add_code_sample_docstrings(
-        tokenizer_class=_TOKENIZER_FOR_DOC,
+        processor_class=_TOKENIZER_FOR_DOC,
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TFQuestionAnsweringModelOutput,
         config_class=_CONFIG_FOR_DOC,