\jQ U d Z ddlZddlZddlmZ ddlmZ ddlmZ ddl m
Z
m Z ddl m
Z
mZmZmZmZmZmZmZ dd lmZ dd
lmZ d d lmZ d d
lm Z d dl m Z m Z m Z m!Z! ej" e# Z$er& e Z ee%e&e%dz e%dz f f e'd< n eg d Z( e j) e( e j* D ]\ Z+Z,e,Z- e sdZ-e-e e+< ee e Z.de%fdZ/ d!de%ej0 z de%ej0 z dz de1de2e%e%f dz de1e%z dz de%dz de1fdZ3 ed G d d Z4d d gZ5dS )"zAutoVideoProcessor class. N) OrderedDict)
TYPE_CHECKING )PreTrainedConfig) get_class_from_dynamic_moduleresolve_trust_remote_code) CONFIG_NAMEIMAGE_PROCESSOR_NAMEPROCESSOR_NAMEVIDEO_PROCESSOR_NAME cached_fileis_torchvision_availableloggingsafe_load_json_file)requires)BaseVideoProcessor )_LazyAutoMapping) VIDEO_PROCESSOR_MAPPING_NAMES)CONFIG_MAPPING_NAMES
AutoConfigmodel_type_to_module_name!replace_list_option_in_docstringsr ) ) cosmos3_omniQwen3VLVideoProcessor) exaone4_5Qwen2VLVideoProcessor) instructblip InstructBlipVideoVideoProcessor)pe_audio_videoPeVideoVideoProcessor) qwen2_5_omnir )
qwen2_5_vlr )qwen3_5r ) qwen3_5_moer )qwen3_omni_moer ) qwen3_vl_moer )
videoprism LlavaOnevisionVideoProcessor
class_namec t j D ]U\ }}| |k rJt | }t j d| d } t || c S # t $ r Y Qw xY wVt j D ]}t |dd | k r|c S t j d }t || rt || S d S )N.ztransformers.models__name__ transformers) r itemsr importlib
import_modulegetattrAttributeErrorVIDEO_PROCESSOR_MAPPING_extra_contentvalueshasattr)r* module_name extractormodule main_modules \C:\Projects\DSCons\.venv\Lib\site-packages\transformers\models\auto\video_processing_auto.py video_processor_class_from_namer= T s "?"E"G"G Y " "3K@@K ,->->->@UVVF
vz22222!
# -;BBDD 9j$ / /: = = >
).99K{J'' 0{J/// 4s A
A&%A&F pretrained_model_name_or_path cache_dirforce_downloadproxiestokenrevisionlocal_files_onlyc T
t t dd
}
fdt t fD } | r| d nd}
|
s |s t d i S d} |t
| } d| v r| d } |
| t
|
} | pi S )a
Loads the video processor configuration from a pretrained model video 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
[`~BaseVideoProcessor.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.
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 video processor configuration from local files.
Passing `token=True` is required when you want to use a private model.
Returns:
`Dict`: The configuration of the video processor.
Examples:
```python
# Download configuration from huggingface.co and cache.
video_processor_config = get_video_processor_config("llava-hf/llava-onevision-qwen2-0.5b-ov-hf")
# This model does not have a video processor config so the result will be an empty dict.
video_processor_config = get_video_processor_config("FacebookAI/xlm-roberta-base")
# Save a pretrained video processor locally and you can reload its config
from transformers import AutoVideoProcessor
video_processor = AutoVideoProcessor.from_pretrained("llava-hf/llava-onevision-qwen2-0.5b-ov-hf")
video_processor.save_pretrained("video-processor-test")
video_processor = get_video_processor_config("video-processor-test")
```F) filenamer? r@ rA rB rC rD _raise_exceptions_for_gated_repo%_raise_exceptions_for_missing_entriesc N g | ]!}t | d d d x "S )F)
rF r? r@ rA rB rC rD rG rH '_raise_exceptions_for_connection_errors)r
)
.0rF r? r@ rD r> rA
resolved_filerC rB s
r<
z.get_video_processor_config.. sm & & & (-!# - !!1166;8=
M r Nz8Could not locate the video processor configuration file.video_processor)r
r r r
loggerinfor )r> r? r@ rA rB rC rD kwargsresolved_processor_file resolved_video_processor_files resolved_video_processor_filevideo_processor_dictprocessor_dictrL s ``````` @r< get_video_processor_configrX l s0 D *% %))..3 & & & & & & & & & & &-/CD& & &"( Jh$q$B1$E$Emq! ) 1H NOOO
*,-DEE . .#12C#D $05I5Q 23P Q Q %2%rN )vision torchvision)backendsc j e Zd ZdZd Ze ee d Ze dd Z
dS )AutoVideoProcessora%
This is a generic video processor class that will be instantiated as one of the video processor classes of the
library when created with the [`AutoVideoProcessor.from_pretrained`] class method.
This class cannot be instantiated directly using `__init__()` (throws an error).
c t d )NzAutoVideoProcessor is designed to be instantiated using the `AutoVideoProcessor.from_pretrained(pretrained_model_name_or_path)` method.)OSError)selfs r< __init__zAutoVideoProcessor.__init__ s
d
rN c V | dd }| dd }d|d< t j |fi |\ }}| dd }d} d| di v r|d d } |}| {| d d }
|
'|
d
d } t | | }d | di v r$|d d } | d
d } |r t
|t st j |fd|i|}t |dd }t |d rd|j v r
|j d } n# t $ r Y nw xY w|t | }| du}
|dupt | t v }|o4|pt t | j d
}|
r5d| v r | d d }nd}t' ||||
| }|
rO|rM|sK| }t) ||fi |}| dd }| |j |g|R i |S | |j |g|R i |S t | t v r-t t | }| |j |g|R i |S t- st | d t d| dt. dt0 dt0 dd d t4 D
)a
Instantiate one of the video processor classes of the library from a pretrained model vocabulary.
The video 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 video_processor hosted inside a model repo on
huggingface.co.
- a path to a *directory* containing a video processor file saved using the
[`~video_processing_utils.BaseVideoProcessor.save_pretrained`] method, e.g.,
`./my_model_directory/`.
- a path to a saved video 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 video 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 video processor files and override the cached versions if
they exist.
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.
return_unused_kwargs (`bool`, *optional*, defaults to `False`):
If `False`, then this function returns just the final video processor object. If `True`, then this
functions returns a `Tuple(video_processor, unused_kwargs)` where *unused_kwargs* is a dictionary
consisting of the key/value pairs whose keys are not video processor attributes: i.e., the part of
`kwargs` which has not been used to update `video_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.
kwargs (`dict[str, Any]`, *optional*):
The values in kwargs of any keys which are video processor attributes will be used to override the
loaded values. Behavior concerning key/value pairs whose keys are *not* video 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 AutoVideoProcessor
>>> # Download video processor from huggingface.co and cache.
>>> video_processor = AutoVideoProcessor.from_pretrained("llava-hf/llava-onevision-qwen2-0.5b-ov-hf")
>>> # If video processor files are in a directory (e.g. video processor was saved using *save_pretrained('./test/saved_model/')*)
>>> # video_processor = AutoVideoProcessor.from_pretrained("./test/saved_model/")
```configNtrust_remote_codeT
_from_autovideo_processor_typer] auto_mapimage_processor_typeImageProcessorVideoProcessorAutoImageProcessorz
transformers.z--r
code_revisionzT requires `torchvision` to be installed. Please install `torchvision` and try again.z Unrecognized video processor in z2. Should have a `video_processor_type` key in its z of z3, or one of the following `model_type` keys in its z: z, c 3 K | ]}|V d S )N )rK cs r< z5AutoVideoProcessor.from_pretrained.. s" @j@jq@j@j@j@j@j@jrN )popr get_video_processor_dictgetreplacer=
isinstancer r from_pretrainedr2 r7 rg
ValueErrortyper4
__module__
startswithsplitr r register_for_auto_classr r r joinr )clsr> inputsrR rc rd config_dict_video_processor_classvideo_processor_auto_mapimage_processor_class video_processor_class_inferredimage_processor_auto_maphas_remote_codehas_local_codeexplicit_local_code
upstream_repo class_refs r< rv z"AutoVideoProcessor.from_pretrained sd L Hd++ "JJ':D A A #| +DEbmmflmm Q +0F M M #' ;??:r#B#B B B'2:'>?S'T $ ! (-E-M$/OO4JD$Q$Q !$01F1N1NO_aq1r1r. 33QRR^,J)#{z2'F'FFF+6z+BCW+X(+C+K+KL\^n+o+o( ! (
!&*:;; ' 75 IZ ^d F
)08NPT(U(U%6:.. U3G6?3Z3Z/5?S/T,
! ,$CDY$Z$Z !2$>.d:ed6llNe>e , 1 ! J% >t D DQ G
$
9!#@.Racp! ! o 0 o9L o0I$A)Mj$u$unt$u$u !
?D11A ! 9 9 ; ; ;8(89VjY_jjjcijj j
"
.8(89VjY_jjjcijj j
&\\ 4
4
4$;DLL$I !$0<,<=Zn]cnnngmnnn ()) 0 G G G
m/L
m
m1E
m
mKV
m
m(3
m
m7;yy@j@jLi@j@j@j7j7j
m
m
s A E( (
E54E5Fc @ t | || dS )a7
Register a new video processor for this class.
Args:
config_class ([`PreTrainedConfig`]):
The configuration corresponding to the model to register.
video_processor_class ([`BaseVideoProcessor`]):
The video processor to register.
)exist_okN)r4 register) config_classr r s r< r zAutoVideoProcessor.register s' (( 7LW_(`````rN N)F) r- ry __qualname____doc__ra classmethodr r rv staticmethodr rn rN r< r] r] s
&&'DEEW
W
FE [W
r a a a \a a arN r] r4 )NFNNNF)6r r0 os collectionsr typingr configuration_utilsr dynamic_module_utilsr r utilsr r
r r r
r r r utils.import_utilsr video_processing_utilsr auto_factoryr
auto_mappingsr configuration_autor r r r
get_loggerr- rP strtuple__annotations__%MISSING_VIDEO_PROCESSOR_MAPPING_NAMESupdater/
model_typevideo_processorsfast_video_processor_classr4 r= PathLikebooldictrX r] __all__rn rN r< r s # # # # # # 4 3 3 3 3 3 \ \ \ \ \ \ \ \ + * * * * * 8 8 8 8 8 8 * * * * * * 8 8 8 8 8 8
H % % P VaU`UbUb!;sE#*cDj:P4Q/Q#Rbbbb -8K
- -) )!()NOOO$G$A$G$I$I K K J !1 $ # % % *%)"0J!*--**+?A^__ 4 +/ %) # "u& u&#u&R[ 4'u& u& #s(^d
" u&
#: u& Dj
u& u& u& u& u&p
,---xa xa xa xa xa xa xa .-xav %&:
;rN