Split autoclasses on modality (#20559)

* split autoclasses on modality

* apply review

* auto classes

docs/source/en/model_doc/auto.mdx

@@ -74,226 +74,250 @@ Likewise, if your `NewModel` is a subclass of [`PreTrainedModel`], make sure its
 
 [[autodoc]] AutoProcessor
 
-## AutoModel
+## Generic model classes
+
+The following auto classes are available for instantiating a base model class without a specific head.
+
+### AutoModel
 
 [[autodoc]] AutoModel
 
-## AutoModelForPreTraining
+### TFAutoModel
+
+[[autodoc]] TFAutoModel
+
+### FlaxAutoModel
+
+[[autodoc]] FlaxAutoModel
+
+## Generic pretraining classes
+
+The following auto classes are available for instantiating a model with a pretraining head.
+
+### AutoModelForPreTraining
 
 [[autodoc]] AutoModelForPreTraining
 
-## AutoModelForCausalLM
+### TFAutoModelForPreTraining
 
-[[autodoc]] AutoModelForCausalLM
+[[autodoc]] TFAutoModelForPreTraining
 
-## AutoModelForDepthEstimation
+### FlaxAutoModelForPreTraining
 
-[[autodoc]] AutoModelForDepthEstimation
+[[autodoc]] FlaxAutoModelForPreTraining
 
-## AutoModelForMaskedLM
+## Natural Language Processing
 
-[[autodoc]] AutoModelForMaskedLM
+The following auto classes are available for the following natural language processing tasks.
 
-## AutoModelForSeq2SeqLM
+### AutoModelForCausalLM
 
-[[autodoc]] AutoModelForSeq2SeqLM
+[[autodoc]] AutoModelForCausalLM
 
-## AutoModelForSequenceClassification
+### TFAutoModelForCausalLM
 
-[[autodoc]] AutoModelForSequenceClassification
+[[autodoc]] TFAutoModelForCausalLM
 
-## AutoModelForMultipleChoice
+### FlaxAutoModelForCausalLM
 
-[[autodoc]] AutoModelForMultipleChoice
+[[autodoc]] FlaxAutoModelForCausalLM
 
-## AutoModelForNextSentencePrediction
+### AutoModelForMaskedLM
 
-[[autodoc]] AutoModelForNextSentencePrediction
+[[autodoc]] AutoModelForMaskedLM
 
-## AutoModelForTokenClassification
+### TFAutoModelForMaskedLM
 
-[[autodoc]] AutoModelForTokenClassification
+[[autodoc]] TFAutoModelForMaskedLM
 
-## AutoModelForQuestionAnswering
+### FlaxAutoModelForMaskedLM
 
-[[autodoc]] AutoModelForQuestionAnswering
+[[autodoc]] FlaxAutoModelForMaskedLM
 
-## AutoModelForTableQuestionAnswering
+### AutoModelForSeq2SeqLM
 
-[[autodoc]] AutoModelForTableQuestionAnswering
+[[autodoc]] AutoModelForSeq2SeqLM
 
-## AutoModelForDocumentQuestionAnswering
+### TFAutoModelForSeq2SeqLM
 
-[[autodoc]] AutoModelForDocumentQuestionAnswering
+[[autodoc]] TFAutoModelForSeq2SeqLM
 
-## AutoModelForImageClassification
+### FlaxAutoModelForSeq2SeqLM
 
-[[autodoc]] AutoModelForImageClassification
+[[autodoc]] FlaxAutoModelForSeq2SeqLM
 
-## AutoModelForVideoClassification
+### AutoModelForSequenceClassification
 
-[[autodoc]] AutoModelForVideoClassification
+[[autodoc]] AutoModelForSequenceClassification
 
-## AutoModelForVision2Seq
+### TFAutoModelForSequenceClassification
 
-[[autodoc]] AutoModelForVision2Seq
+[[autodoc]] TFAutoModelForSequenceClassification
 
-## AutoModelForVisualQuestionAnswering
+### FlaxAutoModelForSequenceClassification
 
-[[autodoc]] AutoModelForVisualQuestionAnswering
+[[autodoc]] FlaxAutoModelForSequenceClassification
 
-## AutoModelForAudioClassification
+### AutoModelForMultipleChoice
 
-[[autodoc]] AutoModelForAudioClassification
+[[autodoc]] AutoModelForMultipleChoice
 
-## AutoModelForAudioFrameClassification
+### TFAutoModelForMultipleChoice
 
-[[autodoc]] AutoModelForAudioFrameClassification
+[[autodoc]] TFAutoModelForMultipleChoice
 
-## AutoModelForCTC
+### FlaxAutoModelForMultipleChoice
 
-[[autodoc]] AutoModelForCTC
+[[autodoc]] FlaxAutoModelForMultipleChoice
 
-## AutoModelForSpeechSeq2Seq
+### AutoModelForNextSentencePrediction
 
-[[autodoc]] AutoModelForSpeechSeq2Seq
+[[autodoc]] AutoModelForNextSentencePrediction
 
-## AutoModelForAudioXVector
+### TFAutoModelForNextSentencePrediction
 
-[[autodoc]] AutoModelForAudioXVector
+[[autodoc]] TFAutoModelForNextSentencePrediction
 
-## AutoModelForMaskedImageModeling
+### FlaxAutoModelForNextSentencePrediction
 
-[[autodoc]] AutoModelForMaskedImageModeling
+[[autodoc]] FlaxAutoModelForNextSentencePrediction
 
-## AutoModelForObjectDetection
+### AutoModelForTokenClassification
 
-[[autodoc]] AutoModelForObjectDetection
+[[autodoc]] AutoModelForTokenClassification
 
-## AutoModelForImageSegmentation
+### TFAutoModelForTokenClassification
 
-[[autodoc]] AutoModelForImageSegmentation
+[[autodoc]] TFAutoModelForTokenClassification
 
-## AutoModelForSemanticSegmentation
+### FlaxAutoModelForTokenClassification
 
-[[autodoc]] AutoModelForSemanticSegmentation
+[[autodoc]] FlaxAutoModelForTokenClassification
 
-## AutoModelForInstanceSegmentation
+### AutoModelForQuestionAnswering
 
-[[autodoc]] AutoModelForInstanceSegmentation
+[[autodoc]] AutoModelForQuestionAnswering
 
-## AutoModelForZeroShotObjectDetection
+### TFAutoModelForQuestionAnswering
 
-[[autodoc]] AutoModelForZeroShotObjectDetection
+[[autodoc]] TFAutoModelForQuestionAnswering
 
-## TFAutoModel
+### FlaxAutoModelForQuestionAnswering
 
-[[autodoc]] TFAutoModel
+[[autodoc]] FlaxAutoModelForQuestionAnswering
 
-## TFAutoModelForPreTraining
+## Computer vision
 
-[[autodoc]] TFAutoModelForPreTraining
+The following auto classes are available for the following computer vision tasks.
 
-## TFAutoModelForCausalLM
+### AutoModelForDepthEstimation
 
-[[autodoc]] TFAutoModelForCausalLM
+[[autodoc]] AutoModelForDepthEstimation
 
-## TFAutoModelForImageClassification
+### AutoModelForImageClassification
+
+[[autodoc]] AutoModelForImageClassification
+
+### TFAutoModelForImageClassification
 
 [[autodoc]] TFAutoModelForImageClassification
 
-## TFAutoModelForSemanticSegmentation
+### FlaxAutoModelForImageClassification
 
-[[autodoc]] TFAutoModelForSemanticSegmentation
+[[autodoc]] FlaxAutoModelForImageClassification
 
-## TFAutoModelForMaskedLM
+### AutoModelForVideoClassification
 
-[[autodoc]] TFAutoModelForMaskedLM
+[[autodoc]] AutoModelForVideoClassification
 
-## TFAutoModelForSeq2SeqLM
+### AutoModelForMaskedImageModeling
 
-[[autodoc]] TFAutoModelForSeq2SeqLM
+[[autodoc]] AutoModelForMaskedImageModeling
 
-## TFAutoModelForSequenceClassification
+### AutoModelForObjectDetection
 
-[[autodoc]] TFAutoModelForSequenceClassification
+[[autodoc]] AutoModelForObjectDetection
 
-## TFAutoModelForMultipleChoice
+### AutoModelForImageSegmentation
 
-[[autodoc]] TFAutoModelForMultipleChoice
+[[autodoc]] AutoModelForImageSegmentation
 
-## TFAutoModelForNextSentencePrediction
+### AutoModelForSemanticSegmentation
 
-[[autodoc]] TFAutoModelForNextSentencePrediction
+[[autodoc]] AutoModelForSemanticSegmentation
 
-## TFAutoModelForTableQuestionAnswering
+### TFAutoModelForSemanticSegmentation
 
-[[autodoc]] TFAutoModelForTableQuestionAnswering
+[[autodoc]] TFAutoModelForSemanticSegmentation
 
-## TFAutoModelForDocumentQuestionAnswering
+### AutoModelForInstanceSegmentation
 
-[[autodoc]] TFAutoModelForDocumentQuestionAnswering
+[[autodoc]] AutoModelForInstanceSegmentation
 
-## TFAutoModelForTokenClassification
+### AutoModelForZeroShotObjectDetection
 
-[[autodoc]] TFAutoModelForTokenClassification
+[[autodoc]] AutoModelForZeroShotObjectDetection
 
-## TFAutoModelForQuestionAnswering
+## Audio
 
-[[autodoc]] TFAutoModelForQuestionAnswering
+The following auto classes are available for the following audio tasks.
 
-## TFAutoModelForVision2Seq
+### AutoModelForAudioClassification
 
-[[autodoc]] TFAutoModelForVision2Seq
+[[autodoc]] AutoModelForAudioClassification
 
-## TFAutoModelForSpeechSeq2Seq
+### AutoModelForAudioFrameClassification
 
-[[autodoc]] TFAutoModelForSpeechSeq2Seq
+[[autodoc]] AutoModelForAudioFrameClassification
 
-## FlaxAutoModel
+### AutoModelForCTC
 
-[[autodoc]] FlaxAutoModel
+[[autodoc]] AutoModelForCTC
 
-## FlaxAutoModelForCausalLM
+### AutoModelForSpeechSeq2Seq
 
-[[autodoc]] FlaxAutoModelForCausalLM
+[[autodoc]] AutoModelForSpeechSeq2Seq
 
-## FlaxAutoModelForPreTraining
+### TFAutoModelForSpeechSeq2Seq
 
-[[autodoc]] FlaxAutoModelForPreTraining
+[[autodoc]] TFAutoModelForSpeechSeq2Seq
 
-## FlaxAutoModelForMaskedLM
+### AutoModelForAudioXVector
 
-[[autodoc]] FlaxAutoModelForMaskedLM
+[[autodoc]] AutoModelForAudioXVector
 
-## FlaxAutoModelForSeq2SeqLM
+## Multimodal
 
-[[autodoc]] FlaxAutoModelForSeq2SeqLM
+The following auto classes are available for the following multimodal tasks.
 
-## FlaxAutoModelForSequenceClassification
+### AutoModelForTableQuestionAnswering
 
-[[autodoc]] FlaxAutoModelForSequenceClassification
+[[autodoc]] AutoModelForTableQuestionAnswering
 
-## FlaxAutoModelForQuestionAnswering
+### TFAutoModelForTableQuestionAnswering
 
-[[autodoc]] FlaxAutoModelForQuestionAnswering
+[[autodoc]] TFAutoModelForTableQuestionAnswering
 
-## FlaxAutoModelForTokenClassification
+### AutoModelForDocumentQuestionAnswering
 
-[[autodoc]] FlaxAutoModelForTokenClassification
+[[autodoc]] AutoModelForDocumentQuestionAnswering
 
-## FlaxAutoModelForMultipleChoice
+### TFAutoModelForDocumentQuestionAnswering
 
-[[autodoc]] FlaxAutoModelForMultipleChoice
+[[autodoc]] TFAutoModelForDocumentQuestionAnswering
 
-## FlaxAutoModelForNextSentencePrediction
+### AutoModelForVisualQuestionAnswering
 
-[[autodoc]] FlaxAutoModelForNextSentencePrediction
+[[autodoc]] AutoModelForVisualQuestionAnswering
 
-## FlaxAutoModelForImageClassification
+### AutoModelForVision2Seq
 
-[[autodoc]] FlaxAutoModelForImageClassification
+[[autodoc]] AutoModelForVision2Seq
+
+### TFAutoModelForVision2Seq
+
+[[autodoc]] TFAutoModelForVision2Seq
 
-## FlaxAutoModelForVision2Seq
+### FlaxAutoModelForVision2Seq
 
 [[autodoc]] FlaxAutoModelForVision2Seq