o
i @ s U d Z ddlZddlZddlZddlZddlmZ ddlmZm Z m
Z
ddl m Z ddl
mZmZ ddlmZ dd lmZ dd
lmZmZmZmZmZmZmZm Z dd l m Z d d
l m Z d dl!m"Z"m#Z#m$Z$m%Z% e &e'Z(dgZ)ere Z*ee+e,e e+ e e+ f f e-d< neg dZ*e*. D ]\Z/\Z0Z1e sdZ0e sdZ1e0e1fe*e/< qe e"e*Z2de+fddZ3 d'de
e+ej4f de e
e+ej4f de5de e5 de e6e+e+f de e
e5e+f d e e+ d e5fd d Z7d d! Z8e d"d#G d$d% d%Z9d&d%gZ:dS )(zAutoImageProcessor class. N) OrderedDict)
TYPE_CHECKINGOptionalUnion )PretrainedConfig) get_class_from_dynamic_moduleresolve_trust_remote_code)ImageProcessingMixin)BaseImageProcessorFast) CONFIG_NAMEIMAGE_PROCESSOR_NAME cached_fileis_timm_config_dictis_timm_local_checkpointis_torchvision_availableis_vision_availablelogging)requires )_LazyAutoMapping)CONFIG_MAPPING_NAMES
AutoConfigmodel_type_to_module_name!replace_list_option_in_docstringsQwen2VLImageProcessor IMAGE_PROCESSOR_MAPPING_NAMES))aimv2CLIPImageProcessorCLIPImageProcessorFast)aimv2_vision_modelr )alignEfficientNetImageProcessor EfficientNetImageProcessorFast)aria)AriaImageProcessorN)beitBeitImageProcessorBeitImageProcessorFast)bitBitImageProcessorBitImageProcessorFast)blipBlipImageProcessorBlipImageProcessorFast)zblip-2r1 ) bridgetower)BridgeTowerImageProcessor BridgeTowerImageProcessorFast) chameleon)ChameleonImageProcessorChameleonImageProcessorFast) chinese_clip)ChineseCLIPImageProcessor ChineseCLIPImageProcessorFast)clipr )clipsegViTImageProcessorViTImageProcessorFast)cohere2_vision)N Cohere2VisionImageProcessorFast)conditional_detr) ConditionalDetrImageProcessor!ConditionalDetrImageProcessorFast)convnextConvNextImageProcessorConvNextImageProcessorFast)
convnextv2rH )cvtrH )zdata2vec-visionr) ) deepseek_vl)DeepseekVLImageProcessor DeepseekVLImageProcessorFast)deepseek_vl_hybrid) DeepseekVLHybridImageProcessor"DeepseekVLHybridImageProcessorFast)deformable_detr) DeformableDetrImageProcessor DeformableDetrImageProcessorFast)deit)DeiTImageProcessorDeiTImageProcessorFast)depth_anythingDPTImageProcessorDPTImageProcessorFast) depth_pro)DepthProImageProcessorDepthProImageProcessorFast)deta)DetaImageProcessorN)detrDetrImageProcessorDetrImageProcessorFast)dinatr? )dinov2r- )
dinov3_vit)NDINOv3ViTImageProcessorFast)z
donut-swin)DonutImageProcessorDonutImageProcessorFast)dptrZ )edgetamNSam2ImageProcessorFast)efficientformer) EfficientFormerImageProcessorN)efficientloftr) EfficientLoFTRImageProcessor EfficientLoFTRImageProcessorFast) efficientnetr# )eomt)EomtImageProcessorEomtImageProcessorFast)flava)FlavaImageProcessorFlavaImageProcessorFast)focalnetr- )fuyu)FuyuImageProcessorN)gemma3Gemma3ImageProcessorGemma3ImageProcessorFast)gemma3nSiglipImageProcessorSiglipImageProcessorFast)gitr )glm4v)Glm4vImageProcessorGlm4vImageProcessorFast)glpn)GLPNImageProcessorN)got_ocr2)GotOcr2ImageProcessorGotOcr2ImageProcessorFast)zgrounding-dinoGroundingDinoImageProcessor GroundingDinoImageProcessorFast)groupvitr )hierar- )idefics)IdeficsImageProcessorN)idefics2)Idefics2ImageProcessorIdefics2ImageProcessorFast)idefics3)Idefics3ImageProcessorIdefics3ImageProcessorFast)ijepar? )imagegpt)ImageGPTImageProcessorImageGPTImageProcessorFast) instructblipr1 )instructblipvideo) InstructBlipVideoImageProcessorN)janus)JanusImageProcessorJanusImageProcessorFast)zkosmos-2r )z
kosmos-2.5)Kosmos2_5ImageProcessorKosmos2_5ImageProcessorFast)
layoutlmv2)LayoutLMv2ImageProcessor LayoutLMv2ImageProcessorFast)
layoutlmv3LayoutLMv3ImageProcessor LayoutLMv3ImageProcessorFast)levit)LevitImageProcessorLevitImageProcessorFast)lfm2_vl)NLfm2VlImageProcessorFast) lightglue)LightGlueImageProcessorN)llama4)Llama4ImageProcessorLlama4ImageProcessorFast)llava)LlavaImageProcessorLlavaImageProcessorFast)
llava_next)LlavaNextImageProcessorLlavaNextImageProcessorFast)llava_next_video) LlavaNextVideoImageProcessorN)llava_onevision) LlavaOnevisionImageProcessor LlavaOnevisionImageProcessorFast) mask2former)Mask2FormerImageProcessor Mask2FormerImageProcessorFast)
maskformer)MaskFormerImageProcessor MaskFormerImageProcessorFast)
metaclip_2r )zmgp-strr? )mistral3PixtralImageProcessorPixtralImageProcessorFast)mlcdr )mllama)MllamaImageProcessorN)zmm-grounding-dinor ) mobilenet_v1)MobileNetV1ImageProcessor MobileNetV1ImageProcessorFast) mobilenet_v2)MobileNetV2ImageProcessor MobileNetV2ImageProcessorFast) mobilevitMobileViTImageProcessorMobileViTImageProcessorFast) mobilevitv2r )natr? )nougat)NougatImageProcessorNougatImageProcessorFast) oneformer)OneFormerImageProcessorOneFormerImageProcessorFast)ovis2)Ovis2ImageProcessorOvis2ImageProcessorFast)owlv2)Owlv2ImageProcessorOwlv2ImageProcessorFast)owlvit)OwlViTImageProcessorOwlViTImageProcessorFast) paligemmar ) perceiver)PerceiverImageProcessorPerceiverImageProcessorFast)
perception_lm)N PerceptionLMImageProcessorFast)phi4_multimodal)N Phi4MultimodalImageProcessorFast)
pix2struct)Pix2StructImageProcessorN)pixtralr )
poolformer)PoolFormerImageProcessor PoolFormerImageProcessorFast)prompt_depth_anything)!PromptDepthAnythingImageProcessor%PromptDepthAnythingImageProcessorFast)pvtPvtImageProcessorPvtImageProcessorFast)pvt_v2r )
qwen2_5_vlr Qwen2VLImageProcessorFast)qwen2_vlr )qwen3_vlr )regnetrH )resnetrH )rt_detr)RTDetrImageProcessorRTDetrImageProcessorFast)samSamImageProcessorSamImageProcessorFast)sam2rn )sam_hqr ) segformerSegformerImageProcessorSegformerImageProcessorFast)seggpt)SegGptImageProcessorN) shieldgemma2r )siglipr )siglip2)Siglip2ImageProcessorSiglip2ImageProcessorFast)smolvlm)SmolVLMImageProcessorSmolVLMImageProcessorFast) superglue)SuperGlueImageProcessorN)
superpoint)SuperPointImageProcessor SuperPointImageProcessorFast) swiftformerr? )swinr? )swin2sr)Swin2SRImageProcessorSwin2SRImageProcessorFast)swinv2r? )ztable-transformerrc )textnet)TextNetImageProcessorTextNetImageProcessorFast) timesformerVideoMAEImageProcessorN) timm_wrapper)TimmWrapperImageProcessorN)tvlt)TvltImageProcessorN)tvp)TvpImageProcessorTvpImageProcessorFast)udopr )upernetr )vanrH )videomaer1 )vilt)ViltImageProcessorViltImageProcessorFast)vipllavar )vitr? )
vit_hybrid)ViTHybridImageProcessorN)vit_maer? )vit_msnr? )vitmatte)VitMatteImageProcessorVitMatteImageProcessorFast)xclipr )yolos)YolosImageProcessorYolosImageProcessorFast)zoedepth)ZoeDepthImageProcessorZoeDepthImageProcessorFast
class_namec C s | dkrt S t D ]'\}}| |v r1t|}td| d}zt|| W S ty0 Y q
w q
tj
D ]}|D ]}t|dd | krK| S q;q7td}t || r\t|| S d S )Nr .ztransformers.models__name__ transformers) r r itemsr importlib
import_modulegetattrAttributeErrorIMAGE_PROCESSOR_MAPPING_extra_contentvalueshasattr)rQ module_name
extractorsmodule extractor main_module rc b/home/ubuntu/.local/lib/python3.10/site-packages/transformers/models/auto/image_processing_auto.py#get_image_processor_class_from_name s,
re F pretrained_model_name_or_path cache_dirforce_downloadresume_downloadproxiestokenrevisionlocal_files_onlyc K s | dd} | durtdt |durtd| }t| t|||||||dddd }
|
du r5td i S t |
dd
} t
| W d S 1 sKw Y dS )
a
Loads the image processor configuration from a pretrained model image processor configuration.
Args:
pretrained_model_name_or_path (`str` or `os.PathLike`):
This can be either:
- a string, the *model id* of a pretrained model configuration hosted inside a model repo on
huggingface.co.
- a path to a *directory* containing a configuration file saved using the
[`~PreTrainedTokenizer.save_pretrained`] method, e.g., `./my_model_directory/`.
cache_dir (`str` or `os.PathLike`, *optional*):
Path to a directory in which a downloaded pretrained model configuration should be cached if the standard
cache should not be used.
force_download (`bool`, *optional*, defaults to `False`):
Whether or not to force to (re-)download the configuration files and override the cached versions if they
exist.
resume_download:
Deprecated and ignored. All downloads are now resumed by default when possible.
Will be removed in v5 of Transformers.
proxies (`dict[str, str]`, *optional*):
A dictionary of proxy servers to use by protocol or endpoint, e.g., `{'http': 'foo.bar:3128',
'http://hostname': 'foo.bar:4012'}.` The proxies are used on each request.
token (`str` or *bool*, *optional*):
The token to use as HTTP bearer authorization for remote files. If `True`, will use the token generated
when running `hf auth login` (stored in `~/.huggingface`).
revision (`str`, *optional*, defaults to `"main"`):
The specific model version to use. It can be a branch name, a tag name, or a commit id, since we use a
git-based system for storing models and other artifacts on huggingface.co, so `revision` can be any
identifier allowed by git.
local_files_only (`bool`, *optional*, defaults to `False`):
If `True`, will only try to load the image processor configuration from local files.
Passing `token=True` is required when you want to use a private model.
Returns:
`Dict`: The configuration of the image processor.
Examples:
```python
# Download configuration from huggingface.co and cache.
image_processor_config = get_image_processor_config("google-bert/bert-base-uncased")
# This model does not have a image processor config so the result will be an empty dict.
image_processor_config = get_image_processor_config("FacebookAI/xlm-roberta-base")
# Save a pretrained image processor locally and you can reload its config
from transformers import AutoTokenizer
image_processor = AutoImageProcessor.from_pretrained("google/vit-base-patch16-224-in21k")
image_processor.save_pretrained("image-processor-test")
image_processor_config = get_image_processor_config("image-processor-test")
```use_auth_tokenNrThe `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.V`token` and `use_auth_token` are both specified. Please set only the argument `token`.F)
rg rh ri rj rk rl rm _raise_exceptions_for_gated_repo%_raise_exceptions_for_missing_entries'_raise_exceptions_for_connection_errorszbCould not locate the image processor configuration file, will try to use the model config instead.zutf-8)encoding) popwarningswarn
FutureWarning
ValueErrorr r
loggerinfoopenjsonload) rf rg rh ri rj rk rl rm kwargsrn resolved_config_filereaderrc rc rd get_image_processor_config s> E$r c C s t d| d d S )NzFast image processor class zz is available for this model. Using slow image processor class. To use the fast image processor class set `use_fast=True`.)rz warning)
fast_classrc rc rd '_warning_fast_image_processor_availableZ s
r )vision)backendsc @ sB e Zd ZdZdd Zeeedd Ze d
dd Z
dS ) AutoImageProcessora%
This is a generic image processor class that will be instantiated as one of the image processor classes of the
library when created with the [`AutoImageProcessor.from_pretrained`] class method.
This class cannot be instantiated directly using `__init__()` (throws an error).
c C s t d)NzAutoImageProcessor is designed to be instantiated using the `AutoImageProcessor.from_pretrained(pretrained_model_name_or_path)` method.)OSError)selfrc rc rd __init__j s zAutoImageProcessor.__init__c O s | dd}|dur tdt |ddurtd||d< | dd}| dd}| dd}d |d
< d |v r?| d }n t|rFt}nt}zt j
|fd |i|\} }
W n1 t y } z%zt j
|fd ti|\} }
W n t yw | w t | s~| W Y d} ~ nd} ~ ww | d d} d}
d
| di v r| d d
}
| du r|
du r| dd}|dur|
dd} d| di v r| d d }|
dd}
| du r|
du rt|tstj|fd|i|}t|d d} t|drd
|jv r|jd
}
d}| dur|du r(| d}|s | tv r t r d }td| d |s(td |r5| ds5| d7 } |rXt sXt| dd }|du rQtd| dtd d}|r|t D ]
}| |v rh nq_| dd } d}td t| }n | d}t|}|du r| drtd| d |
du}|dupt |t v }|r|
durt|
t s|
df}
|r|
d dur|
d }n|
d }d |v r|!d d }nd}t"|||||}|r|r|s|
d durt#|
d t$||fi |}| d!d}
|% |j&| fi |S |dur#|j&| fi |S t |t v rlt t | }|\}}|s@|dur@t#| |rW|sK|du rW|j|g|R i |S |durh|j|g|R i |S td"td#| d$t d%t d&t d'd('d)d* tD
)+aI
Instantiate one of the image processor classes of the library from a pretrained model vocabulary.
The image processor class to instantiate is selected based on the `model_type` property of the config object
(either passed as an argument or loaded from `pretrained_model_name_or_path` if possible), or when it's
missing, by falling back to using pattern matching on `pretrained_model_name_or_path`:
List options
Params:
pretrained_model_name_or_path (`str` or `os.PathLike`):
This can be either:
- a string, the *model id* of a pretrained image_processor hosted inside a model repo on
huggingface.co.
- a path to a *directory* containing a image processor file saved using the
[`~image_processing_utils.ImageProcessingMixin.save_pretrained`] method, e.g.,
`./my_model_directory/`.
- a path or url to a saved image processor JSON *file*, e.g.,
`./my_model_directory/preprocessor_config.json`.
cache_dir (`str` or `os.PathLike`, *optional*):
Path to a directory in which a downloaded pretrained model image processor should be cached if the
standard cache should not be used.
force_download (`bool`, *optional*, defaults to `False`):
Whether or not to force to (re-)download the image processor files and override the cached versions if
they exist.
resume_download:
Deprecated and ignored. All downloads are now resumed by default when possible.
Will be removed in v5 of Transformers.
proxies (`dict[str, str]`, *optional*):
A dictionary of proxy servers to use by protocol or endpoint, e.g., `{'http': 'foo.bar:3128',
'http://hostname': 'foo.bar:4012'}.` The proxies are used on each request.
token (`str` or *bool*, *optional*):
The token to use as HTTP bearer authorization for remote files. If `True`, will use the token generated
when running `hf auth login` (stored in `~/.huggingface`).
revision (`str`, *optional*, defaults to `"main"`):
The specific model version to use. It can be a branch name, a tag name, or a commit id, since we use a
git-based system for storing models and other artifacts on huggingface.co, so `revision` can be any
identifier allowed by git.
use_fast (`bool`, *optional*, defaults to `False`):
Use a fast torchvision-base image processor if it is supported for a given model.
If a fast image processor is not available for a given model, a normal numpy-based image processor
is returned instead.
return_unused_kwargs (`bool`, *optional*, defaults to `False`):
If `False`, then this function returns just the final image processor object. If `True`, then this
functions returns a `Tuple(image_processor, unused_kwargs)` where *unused_kwargs* is a dictionary
consisting of the key/value pairs whose keys are not image processor attributes: i.e., the part of
`kwargs` which has not been used to update `image_processor` and is otherwise ignored.
trust_remote_code (`bool`, *optional*, defaults to `False`):
Whether or not to allow for custom models defined on the Hub in their own modeling files. This option
should only be set to `True` for repositories you trust and in which you have read the code, as it will
execute code present on the Hub on your local machine.
image_processor_filename (`str`, *optional*, defaults to `"config.json"`):
The name of the file in the model directory to use for the image processor config.
kwargs (`dict[str, Any]`, *optional*):
The values in kwargs of any keys which are image processor attributes will be used to override the
loaded values. Behavior concerning key/value pairs whose keys are *not* image processor attributes is
controlled by the `return_unused_kwargs` keyword parameter.
Passing `token=True` is required when you want to use a private model.
Examples:
```python
>>> from transformers import AutoImageProcessor
>>> # Download image processor from huggingface.co and cache.
>>> image_processor = AutoImageProcessor.from_pretrained("google/vit-base-patch16-224-in21k")
>>> # If image processor files are in a directory (e.g. image processor was saved using *save_pretrained('./test/saved_model/')*)
>>> # image_processor = AutoImageProcessor.from_pretrained("./test/saved_model/")
```rn Nro rk rp configuse_fasttrust_remote_codeT
_from_autoimage_processor_filenameimage_processor_typer auto_mapfeature_extractor_typeFeatureExtractorImageProcessorAutoFeatureExtractorFastz The image processor of type `aS ` is now loaded as a fast processor by default, even if the model checkpoint was saved with a slow processor. This is a breaking change and may produce slightly different outputs. To continue using the slow processor, instantiate this class with `use_fast=False`. Note that this behavior will be extended to all models in a future release.aC Using a slow image processor as `use_fast` is unset and a slow processor was saved with this model. `use_fast=True` will be the default behavior in v4.52, even if the model was saved with a slow processor. This will result in minor differences in outputs. You'll still be able to use a slow processor with `use_fast=False`.`zU` requires `torchvision` to be installed. Please install `torchvision` and try again.zcUsing `use_fast=True` but `torchvision` is not available. Falling back to the slow image processor.Fzz`use_fast` is set to `True` but the image processor class does not have a fast version. Falling back to the slow version.z\` does not have a slow version. Please set `use_fast=True` when instantiating the processor.r r z--
code_revisionzZThis image processor cannot be instantiated. Please make sure you have `Pillow` installed.z Unrecognized image processor in z2. Should have a `image_processor_type` key in its z of z3, or one of the following `model_type` keys in its z: z, c s s | ]}|V qd S )Nrc ).0crc rc rd q s z5AutoImageProcessor.from_pretrained..)(ru rv rw rx getry r r r
r
get_image_processor_dict Exceptionr replace
isinstancer r from_pretrainedrX r] r endswithFORCE_FAST_IMAGE_PROCESSORr rz warning_oncere r r\ removesuffixtyperZ tuplesplitr r r register_for_auto_class from_dictjoin)clsrf inputsr rn r r r r config_dict_initial_exceptionr image_processor_auto_mapfeature_extractor_classfeature_extractor_auto_mapimage_processor_classimage_processorsimage_processor_type_slowhas_remote_codehas_local_code class_ref
upstream_repoimage_processor_tupleimage_processor_class_pyimage_processor_class_fastrc rc rd r p s8 O
z"AutoImageProcessor.from_pretrainedNFc C s |dur|dur t dtdt |}|du r |du r t d|dur-t|tr-t d|dur:t|ts:t d|durX|durXt|trX|j|krXt d|j d| d | tjv rot| \}}|du ri|}|du ro|}tj | ||f|d
dS ) a)
Register a new image processor for this class.
Args:
config_class ([`PretrainedConfig`]):
The configuration corresponding to the model to register.
image_processor_class ([`ImageProcessingMixin`]): The image processor to register.
NzHCannot specify both image_processor_class and slow_image_processor_classzThe image_processor_class argument is deprecated and will be removed in v4.42. Please use `slow_image_processor_class`, or `fast_image_processor_class` insteadzSYou need to specify either slow_image_processor_class or fast_image_processor_classzIYou passed a fast image processor in as the `slow_image_processor_class`.zNThe `fast_image_processor_class` should inherit from `BaseImageProcessorFast`.zThe fast processor class you are passing has a `slow_image_processor_class` attribute that is not consistent with the slow processor class you passed (fast tokenizer has z and you passed z!. Fix one of those so they match!)exist_ok)
ry rv rw rx
issubclassr slow_image_processor_classrZ r[ register) config_classr r fast_image_processor_classr
existing_slow
existing_fastrc rc rd r t sJ
zAutoImageProcessor.register)NNNF) rS
__module__ __qualname____doc__r classmethodr r r staticmethodr rc rc rc rd r a s r rZ )NFNNNNF);r rV r} osrv collectionsr typingr r r configuration_utilsr dynamic_module_utilsr r image_processing_utilsr
image_processing_utils_fastr utilsr r
r r r r r r utils.import_utilsr auto_factoryr configuration_autor r r r
get_loggerrS rz r r strr __annotations__rU
model_type
slow_classr rZ re PathLikebooldictr r r __all__rc rc rc rd sz (
(
g P