[v5] 🚨Refactor subprocessors handling in processors

src/transformers/models/align/processing_align.py

@@ -59,9 +59,6 @@ class AlignProcessor(ProcessorMixin):
 
     """
 
-    attributes = ["image_processor", "tokenizer"]
-    image_processor_class = "EfficientNetImageProcessor"
-    tokenizer_class = ("BertTokenizer", "BertTokenizerFast")
     valid_processor_kwargs = AlignProcessorKwargs
 
     def __init__(self, image_processor, tokenizer):

src/transformers/models/altclip/processing_altclip.py

@@ -35,10 +35,6 @@ class AltCLIPProcessor(ProcessorMixin):
             The tokenizer is a required input.
     """
 
-    attributes = ["image_processor", "tokenizer"]
-    image_processor_class = ("CLIPImageProcessor", "CLIPImageProcessorFast")
-    tokenizer_class = ("XLMRobertaTokenizer", "XLMRobertaTokenizerFast")
-
     @deprecate_kwarg(old_name="feature_extractor", version="5.0.0", new_name="image_processor")
     def __init__(self, image_processor=None, tokenizer=None):
         super().__init__(image_processor, tokenizer)

src/transformers/models/aria/modular_aria.py

@@ -941,10 +941,6 @@ class AriaProcessor(ProcessorMixin):
             A dictionary indicating size conversions for images.
     """
 
-    attributes = ["image_processor", "tokenizer"]
-    image_processor_class = "AriaImageProcessor"
-    tokenizer_class = "AutoTokenizer"
-
     def __init__(
         self,
         image_processor=None,

src/transformers/models/aria/processing_aria.py

@@ -67,10 +67,6 @@ class AriaProcessor(ProcessorMixin):
             A dictionary indicating size conversions for images.
     """
 
-    attributes = ["image_processor", "tokenizer"]
-    image_processor_class = "AriaImageProcessor"
-    tokenizer_class = "AutoTokenizer"
-
     def __init__(
         self,
         image_processor=None,

src/transformers/models/auto/feature_extraction_auto.py

@@ -42,6 +42,7 @@
         ("audio-spectrogram-transformer", "ASTFeatureExtractor"),
         ("clap", "ClapFeatureExtractor"),
         ("clvp", "ClvpFeatureExtractor"),
+        ("csm", "EncodecFeatureExtractor"),
         ("dac", "DacFeatureExtractor"),
         ("data2vec-audio", "Wav2Vec2FeatureExtractor"),
         ("dia", "DiaFeatureExtractor"),
@@ -50,14 +51,21 @@
         ("granite_speech", "GraniteSpeechFeatureExtractor"),
         ("hubert", "Wav2Vec2FeatureExtractor"),
         ("kyutai_speech_to_text", "KyutaiSpeechToTextFeatureExtractor"),
+        ("markuplm", "MarkupLMFeatureExtractor"),
         ("mctct", "MCTCTFeatureExtractor"),
         ("mimi", "EncodecFeatureExtractor"),
         ("moonshine", "Wav2Vec2FeatureExtractor"),
         ("moshi", "EncodecFeatureExtractor"),
+        ("musicgen", "EncodecFeatureExtractor"),
+        ("musicgen_melody", "MusicgenMelodyFeatureExtractor"),
+        ("parakeet", "ParakeetFeatureExtractor"),
         ("parakeet_ctc", "ParakeetFeatureExtractor"),
         ("parakeet_encoder", "ParakeetFeatureExtractor"),
         ("phi4_multimodal", "Phi4MultimodalFeatureExtractor"),
         ("pop2piano", "Pop2PianoFeatureExtractor"),
+        ("qwen2_5_omni", "WhisperFeatureExtractor"),
+        ("qwen2_audio", "WhisperFeatureExtractor"),
+        ("qwen3_omni_moe", "WhisperFeatureExtractor"),
         ("seamless_m4t", "SeamlessM4TFeatureExtractor"),
         ("seamless_m4t_v2", "SeamlessM4TFeatureExtractor"),
         ("sew", "Wav2Vec2FeatureExtractor"),
@@ -67,6 +75,7 @@
         ("unispeech", "Wav2Vec2FeatureExtractor"),
         ("unispeech-sat", "Wav2Vec2FeatureExtractor"),
         ("univnet", "UnivNetFeatureExtractor"),
+        ("voxtral", "WhisperFeatureExtractor"),
         ("wav2vec2", "Wav2Vec2FeatureExtractor"),
         ("wav2vec2-bert", "Wav2Vec2FeatureExtractor"),
         ("wav2vec2-conformer", "Wav2Vec2FeatureExtractor"),

src/transformers/models/auto/image_processing_auto.py

@@ -62,7 +62,9 @@
             ("aimv2", ("CLIPImageProcessor", "CLIPImageProcessorFast")),
             ("aimv2_vision_model", ("CLIPImageProcessor", "CLIPImageProcessorFast")),
             ("align", ("EfficientNetImageProcessor", "EfficientNetImageProcessorFast")),
+            ("altclip", ("CLIPImageProcessor", "CLIPImageProcessorFast")),
             ("aria", ("AriaImageProcessor", None)),
+            ("aya_vision", ("GotOcr2ImageProcessor", "GotOcr2ImageProcessorFast")),
             ("beit", ("BeitImageProcessor", "BeitImageProcessorFast")),
             ("bit", ("BitImageProcessor", "BitImageProcessorFast")),
             ("blip", ("BlipImageProcessor", "BlipImageProcessorFast")),
@@ -73,6 +75,8 @@
             ("clip", ("CLIPImageProcessor", "CLIPImageProcessorFast")),
             ("clipseg", ("ViTImageProcessor", "ViTImageProcessorFast")),
             ("cohere2_vision", (None, "Cohere2VisionImageProcessorFast")),
+            ("colpali", ("SiglipImageProcessor", "SiglipImageProcessorFast")),
+            ("colqwen2", ("Qwen2VLImageProcessor", "Qwen2VLImageProcessorFast")),
             ("conditional_detr", ("ConditionalDetrImageProcessor", "ConditionalDetrImageProcessorFast")),
             ("convnext", ("ConvNextImageProcessor", "ConvNextImageProcessorFast")),
             ("convnextv2", ("ConvNextImageProcessor", "ConvNextImageProcessorFast")),
@@ -94,9 +98,11 @@
             ("edgetam", (None, "Sam2ImageProcessorFast")),
             ("efficientformer", ("EfficientFormerImageProcessor", None)),
             ("efficientloftr", ("EfficientLoFTRImageProcessor", "EfficientLoFTRImageProcessorFast")),
+            ("emu3", ("Emu3ImageProcessor", None)),
             ("efficientnet", ("EfficientNetImageProcessor", "EfficientNetImageProcessorFast")),
             ("eomt", ("EomtImageProcessor", "EomtImageProcessorFast")),
             ("flava", ("FlavaImageProcessor", "FlavaImageProcessorFast")),
+            ("florence2", ("CLIPImageProcessor", "CLIPImageProcessorFast")),
             ("focalnet", ("BitImageProcessor", "BitImageProcessorFast")),
             ("fuyu", ("FuyuImageProcessor", None)),
             ("gemma3", ("Gemma3ImageProcessor", "Gemma3ImageProcessorFast")),
@@ -115,11 +121,13 @@
             ("imagegpt", ("ImageGPTImageProcessor", "ImageGPTImageProcessorFast")),
             ("instructblip", ("BlipImageProcessor", "BlipImageProcessorFast")),
             ("instructblipvideo", ("InstructBlipVideoImageProcessor", None)),
+            ("internvl", ("GotOcr2ImageProcessor", "GotOcr2ImageProcessorFast")),
             ("janus", ("JanusImageProcessor", "JanusImageProcessorFast")),
             ("kosmos-2", ("CLIPImageProcessor", "CLIPImageProcessorFast")),
             ("kosmos-2.5", ("Kosmos2_5ImageProcessor", "Kosmos2_5ImageProcessorFast")),
             ("layoutlmv2", ("LayoutLMv2ImageProcessor", "LayoutLMv2ImageProcessorFast")),
             ("layoutlmv3", ("LayoutLMv3ImageProcessor", "LayoutLMv3ImageProcessorFast")),
+            ("layoutxlm", ("LayoutLMv2ImageProcessor", "LayoutLMv2ImageProcessor")),
             ("levit", ("LevitImageProcessor", "LevitImageProcessorFast")),
             ("lfm2_vl", (None, "Lfm2VlImageProcessorFast")),
             ("lightglue", ("LightGlueImageProcessor", None)),
@@ -142,6 +150,7 @@
             ("mobilevitv2", ("MobileViTImageProcessor", "MobileViTImageProcessorFast")),
             ("nat", ("ViTImageProcessor", "ViTImageProcessorFast")),
             ("nougat", ("NougatImageProcessor", "NougatImageProcessorFast")),
+            ("omdet-turbo", ("DetrImageProcessor", "DetrImageProcessorFast")),
             ("oneformer", ("OneFormerImageProcessor", "OneFormerImageProcessorFast")),
             ("ovis2", ("Ovis2ImageProcessor", "Ovis2ImageProcessorFast")),
             ("owlv2", ("Owlv2ImageProcessor", "Owlv2ImageProcessorFast")),
@@ -156,14 +165,17 @@
             ("prompt_depth_anything", ("PromptDepthAnythingImageProcessor", "PromptDepthAnythingImageProcessorFast")),
             ("pvt", ("PvtImageProcessor", "PvtImageProcessorFast")),
             ("pvt_v2", ("PvtImageProcessor", "PvtImageProcessorFast")),
+            ("qwen2_5_omni", ("Qwen2VLImageProcessor", "Qwen2VLImageProcessorFast")),
             ("qwen2_5_vl", ("Qwen2VLImageProcessor", "Qwen2VLImageProcessorFast")),
             ("qwen2_vl", ("Qwen2VLImageProcessor", "Qwen2VLImageProcessorFast")),
+            ("qwen3_omni_moe", ("Qwen2VLImageProcessor", "Qwen2VLImageProcessorFast")),
             ("qwen3_vl", ("Qwen2VLImageProcessor", "Qwen2VLImageProcessorFast")),
             ("regnet", ("ConvNextImageProcessor", "ConvNextImageProcessorFast")),
             ("resnet", ("ConvNextImageProcessor", "ConvNextImageProcessorFast")),
             ("rt_detr", ("RTDetrImageProcessor", "RTDetrImageProcessorFast")),
             ("sam", ("SamImageProcessor", "SamImageProcessorFast")),
             ("sam2", (None, "Sam2ImageProcessorFast")),
+            ("sam2_video", (None, "Sam2ImageProcessorFast")),
             ("sam_hq", ("SamImageProcessor", "SamImageProcessorFast")),
             ("segformer", ("SegformerImageProcessor", "SegformerImageProcessorFast")),
             ("seggpt", ("SegGptImageProcessor", None)),
@@ -181,12 +193,14 @@
             ("textnet", ("TextNetImageProcessor", "TextNetImageProcessorFast")),
             ("timesformer", ("VideoMAEImageProcessor", None)),
             ("timm_wrapper", ("TimmWrapperImageProcessor", None)),
+            ("trocr", ("ViTImageProcessor", "ViTImageProcessorFast")),
             ("tvlt", ("TvltImageProcessor", None)),
             ("tvp", ("TvpImageProcessor", "TvpImageProcessorFast")),
             ("udop", ("LayoutLMv3ImageProcessor", "LayoutLMv3ImageProcessorFast")),
             ("upernet", ("SegformerImageProcessor", "SegformerImageProcessorFast")),
             ("van", ("ConvNextImageProcessor", "ConvNextImageProcessorFast")),
             ("video_llama_3", ("VideoLlama3ImageProcessor", "VideoLlama3ImageProcessorFast")),
+            ("video_llava", ("VideoLlavaImageProcessor", None)),
             ("videomae", ("VideoMAEImageProcessor", None)),
             ("vilt", ("ViltImageProcessor", "ViltImageProcessorFast")),
             ("vipllava", ("CLIPImageProcessor", "CLIPImageProcessorFast")),

src/transformers/models/auto/processing_auto.py

@@ -108,6 +108,7 @@
         ("mllama", "MllamaProcessor"),
         ("mm-grounding-dino", "GroundingDinoProcessor"),
         ("moonshine", "Wav2Vec2Processor"),
+        ("omdet-turbo", "OmDetTurboProcessor"),
         ("oneformer", "OneFormerProcessor"),
         ("ovis2", "Ovis2Processor"),
         ("owlv2", "Owlv2Processor"),

src/transformers/models/auto/tokenization_auto.py

@@ -73,6 +73,7 @@
             ),
         ),
         ("align", ("BertTokenizer", "BertTokenizerFast" if is_tokenizers_available() else None)),
+        ("altclip", ("XLMRobertaTokenizer", "XLMRobertaTokenizerFast" if is_tokenizers_available() else None)),
         ("arcee", ("LlamaTokenizer", "LlamaTokenizerFast" if is_tokenizers_available() else None)),
         ("aria", ("LlamaTokenizer", "LlamaTokenizerFast" if is_tokenizers_available() else None)),
         ("aya_vision", (None, "CohereTokenizerFast" if is_tokenizers_available() else None)),
@@ -157,6 +158,7 @@
         ("codegen", ("CodeGenTokenizer", "CodeGenTokenizerFast" if is_tokenizers_available() else None)),
         ("cohere", (None, "CohereTokenizerFast" if is_tokenizers_available() else None)),
         ("cohere2", (None, "CohereTokenizerFast" if is_tokenizers_available() else None)),
+        ("cohere2_vision", (None, "CohereTokenizerFast" if is_tokenizers_available() else None)),
         ("colpali", ("LlamaTokenizer", "LlamaTokenizerFast" if is_tokenizers_available() else None)),
         ("colqwen2", ("Qwen2Tokenizer", "Qwen2TokenizerFast" if is_tokenizers_available() else None)),
         ("convbert", ("ConvBertTokenizer", "ConvBertTokenizerFast" if is_tokenizers_available() else None)),
@@ -225,6 +227,7 @@
             ),
         ),
         ("distilbert", ("DistilBertTokenizer", "DistilBertTokenizerFast" if is_tokenizers_available() else None)),
+        ("donut", ("XLMRobertaTokenizer", "XLMRobertaTokenizerFast" if is_tokenizers_available() else None)),
         (
             "dpr",
             (
@@ -239,6 +242,7 @@
         ("ernie4_5_moe", (None, "LlamaTokenizerFast" if is_tokenizers_available() else None)),
         ("ernie_m", ("ErnieMTokenizer" if is_sentencepiece_available() else None, None)),
         ("esm", ("EsmTokenizer", None)),
+        ("evolla", (None, "LlamaTokenizerFast" if is_tokenizers_available() else None)),
         (
             "exaone4",
             (
@@ -253,10 +257,13 @@
             ("FastSpeech2ConformerTokenizer" if is_g2p_en_available() else None, None),
         ),
         ("flaubert", ("FlaubertTokenizer", None)),
+        ("flava", ("BertTokenizer", "BertTokenizerFast" if is_tokenizers_available() else None)),
         ("flex_olmo", (None, "GPT2TokenizerFast" if is_tokenizers_available() else None)),
+        ("florence2", ("BertTokenizer", "BertTokenizerFast" if is_tokenizers_available() else None)),
         ("fnet", ("FNetTokenizer", "FNetTokenizerFast" if is_tokenizers_available() else None)),
         ("fsmt", ("FSMTTokenizer", None)),
         ("funnel", ("FunnelTokenizer", "FunnelTokenizerFast" if is_tokenizers_available() else None)),
+        ("fuyu", (None, "LlamaTokenizerFast" if is_tokenizers_available() else None)),
         (
             "gemma",
             (
@@ -305,6 +312,7 @@
         ("glm4_moe", (None, "PreTrainedTokenizerFast" if is_tokenizers_available() else None)),
         ("glm4v", (None, "PreTrainedTokenizerFast" if is_tokenizers_available() else None)),
         ("glm4v_moe", (None, "PreTrainedTokenizerFast" if is_tokenizers_available() else None)),
+        ("got_ocr2", (None, "PreTrainedTokenizerFast" if is_tokenizers_available() else None)),
         ("gpt-sw3", ("GPTSw3Tokenizer" if is_sentencepiece_available() else None, None)),
         ("gpt2", ("GPT2Tokenizer", "GPT2TokenizerFast" if is_tokenizers_available() else None)),
         ("gpt_bigcode", ("GPT2Tokenizer", "GPT2TokenizerFast" if is_tokenizers_available() else None)),
@@ -315,6 +323,7 @@
         ("gptj", ("GPT2Tokenizer", "GPT2TokenizerFast" if is_tokenizers_available() else None)),
         ("gptsan-japanese", ("GPTSanJapaneseTokenizer", None)),
         ("granite", ("GPT2Tokenizer", None)),
+        ("granite_speech", ("GPT2Tokenizer", None)),
         ("granitemoe", ("GPT2Tokenizer", None)),
         ("granitemoehybrid", ("GPT2Tokenizer", None)),
         ("granitemoeshared", ("GPT2Tokenizer", None)),
@@ -354,11 +363,14 @@
             ),
         ),
         ("kosmos-2.5", (None, "PreTrainedTokenizerFast" if is_tokenizers_available() else None)),
+        ("kyutai_speech_to_text", (None, "PretrainedTokenizerFast" if is_tokenizers_available() else None)),
         ("layoutlm", ("LayoutLMTokenizer", "LayoutLMTokenizerFast" if is_tokenizers_available() else None)),
         ("layoutlmv2", ("LayoutLMv2Tokenizer", "LayoutLMv2TokenizerFast" if is_tokenizers_available() else None)),
         ("layoutlmv3", ("LayoutLMv3Tokenizer", "LayoutLMv3TokenizerFast" if is_tokenizers_available() else None)),
         ("layoutxlm", ("LayoutXLMTokenizer", "LayoutXLMTokenizerFast" if is_tokenizers_available() else None)),
         ("led", ("LEDTokenizer", "LEDTokenizerFast" if is_tokenizers_available() else None)),
+        ("lfm2", (None, "PreTrainedTokenizerFast" if is_tokenizers_available() else None)),
+        ("lfm2_vl", (None, "PreTrainedTokenizerFast" if is_tokenizers_available() else None)),
         ("lilt", ("LayoutLMv3Tokenizer", "LayoutLMv3TokenizerFast" if is_tokenizers_available() else None)),
         (
             "llama",
@@ -399,6 +411,7 @@
         ("mamba", (None, "GPTNeoXTokenizerFast" if is_tokenizers_available() else None)),
         ("mamba2", (None, "GPTNeoXTokenizerFast" if is_tokenizers_available() else None)),
         ("marian", ("MarianTokenizer" if is_sentencepiece_available() else None, None)),
+        ("markuplm", ("MarkupLMTokenizer", "MarkupLMTokenizerFast" if is_tokenizers_available() else None)),
         (
             "mbart",
             (
@@ -485,6 +498,7 @@
                 "NllbTokenizerFast" if is_tokenizers_available() else None,
             ),
         ),
+        ("nougat", (None, "NougatTokenizerFast" if is_tokenizers_available() else None)),
         (
             "nystromformer",
             (
@@ -506,6 +520,7 @@
             ("OpenAIGPTTokenizer", "OpenAIGPTTokenizerFast" if is_tokenizers_available() else None),
         ),
         ("opt", ("GPT2Tokenizer", "GPT2TokenizerFast" if is_tokenizers_available() else None)),
+        ("ovis2", (None, "Qwen2TokenizerFast" if is_tokenizers_available() else None)),
         ("owlv2", ("CLIPTokenizer", "CLIPTokenizerFast" if is_tokenizers_available() else None)),
         ("owlvit", ("CLIPTokenizer", "CLIPTokenizerFast" if is_tokenizers_available() else None)),
         ("paligemma", ("LlamaTokenizer", "LlamaTokenizerFast" if is_tokenizers_available() else None)),
@@ -531,6 +546,7 @@
                 None,
             ),
         ),
+        ("perception_lm", (None, "LlamaTokenizerFast" if is_tokenizers_available() else None)),
         (
             "persimmon",
             (
@@ -540,6 +556,7 @@
         ),
         ("phi", ("CodeGenTokenizer", "CodeGenTokenizerFast" if is_tokenizers_available() else None)),
         ("phi3", ("LlamaTokenizer", "LlamaTokenizerFast" if is_tokenizers_available() else None)),
+        ("phi4_multimodal", (None, "GPT2TokenizerFast" if is_tokenizers_available() else None)),
         ("phimoe", ("LlamaTokenizer", "LlamaTokenizerFast" if is_tokenizers_available() else None)),
         ("phobert", ("PhobertTokenizer", None)),
         ("pix2struct", ("T5Tokenizer", "T5TokenizerFast" if is_tokenizers_available() else None)),
@@ -553,6 +570,7 @@
             ),
         ),
         ("plbart", ("PLBartTokenizer" if is_sentencepiece_available() else None, None)),
+        ("pop2piano", ("Pop2PianoTokenizer", None)),
         ("prophetnet", ("ProphetNetTokenizer", None)),
         ("qdqbert", ("BertTokenizer", "BertTokenizerFast" if is_tokenizers_available() else None)),
         (
@@ -659,6 +677,7 @@
             ),
         ),
         ("smollm3", (None, "PreTrainedTokenizerFast" if is_tokenizers_available() else None)),
+        ("smolvlm", ("PreTrainedTokenizer", "PreTrainedTokenizerFast" if is_tokenizers_available() else None)),
         ("speech_to_text", ("Speech2TextTokenizer" if is_sentencepiece_available() else None, None)),
         ("speech_to_text_2", ("Speech2Text2Tokenizer", None)),
         ("speecht5", ("SpeechT5Tokenizer" if is_sentencepiece_available() else None, None)),
@@ -693,6 +712,7 @@
         ("tapas", ("TapasTokenizer", None)),
         ("tapex", ("TapexTokenizer", None)),
         ("transfo-xl", ("TransfoXLTokenizer", None)),
+        ("trocr", ("XLMRobertaTokenizer", "XLMRobertaTokenizerFast" if is_tokenizers_available() else None)),
         ("tvp", ("BertTokenizer", "BertTokenizerFast" if is_tokenizers_available() else None)),
         (
             "udop",
@@ -708,9 +728,14 @@
                 "T5TokenizerFast" if is_tokenizers_available() else None,
             ),
         ),
+        ("video_llama_3", ("Qwen2Tokenizer", "Qwen2TokenizerFast" if is_tokenizers_available() else None)),
         ("video_llava", ("LlamaTokenizer", "LlamaTokenizerFast" if is_tokenizers_available() else None)),
         ("vilt", ("BertTokenizer", "BertTokenizerFast" if is_tokenizers_available() else None)),
         ("vipllava", ("LlamaTokenizer", "LlamaTokenizerFast" if is_tokenizers_available() else None)),
+        (
+            "vision_text_dual_encoder",
+            ("PreTrainedTokenizer", "PreTrainedTokenizerFast" if is_tokenizers_available() else None),
+        ),
         ("visual_bert", ("BertTokenizer", "BertTokenizerFast" if is_tokenizers_available() else None)),
         ("vits", ("VitsTokenizer", None)),
         (
@@ -726,6 +751,7 @@
         ("wav2vec2-bert", ("Wav2Vec2CTCTokenizer", None)),
         ("wav2vec2-conformer", ("Wav2Vec2CTCTokenizer", None)),
         ("wav2vec2_phoneme", ("Wav2Vec2PhonemeCTCTokenizer", None)),
+        ("wav2vec2_with_lm", ("Wav2Vec2CTCTokenizer", None)),
         ("whisper", ("WhisperTokenizer", "WhisperTokenizerFast" if is_tokenizers_available() else None)),
         ("xclip", ("CLIPTokenizer", "CLIPTokenizerFast" if is_tokenizers_available() else None)),
         (

src/transformers/models/auto/video_processing_auto.py

@@ -64,6 +64,7 @@
             ("video_llava", "VideoLlavaVideoProcessor"),
             ("videomae", "VideoMAEVideoProcessor"),
             ("vjepa2", "VJEPA2VideoProcessor"),
+            ("video_llama_3", "VideoLlama3VideoProcessor"),  # PLACEHOLDER - needs proper video processor class
         ]
     )