codekingpro/portable-devtools
114k
1�
2 ��j
� � �� � S SK r S SKrS SKJr S SKJrJr S SKrS SKJ r S SK3J r S SK J
r
JrJrJrJrJrJr S SKJrJrJrJrJr SS KJr SS4K J r SS K J r J r J!r! \ RD " \#5 r$\" \%5 RL S - S
- r'\" \%5 RL S - S- r(\RR " S5 r* " S S5 r+ " S S\+5 r, " S S\+5 r- " S S\+5 r.S\/S\S 4S jr0S \/\- S\1S- 4S jr2S \/\- S \1SS4S jr3SS SSSSS!.S"\/S#\/S$\/S%\/S&\/S'\S(\/S)\/S*\/S- S+\4S,\/S- S-\/S- S.\/S- S/\/S- S\14 S0 jjr5\!SS SSSSS SS1.S2\/S3\1S4\/S- S5\4S6\/S- S7\/S- S8\/S- S9\/S- S:\4S;\/S- S\/4S< jj5 r6g)=� N)�Path)�Any�Literal)�hf_hub_download)� upload_file)�CardData�DatasetCardData�5EvalResult�
ModelCardData�
SpaceCardData�eval_results_to_model_index�model_index_to_eval_results)�HfHubHTTPError� get_session�hf_raise_for_status�is_jinja_available� yaml_dump� )� constants)�EntryNotFoundError)�SoftTemporaryDirectory�logging�validate_hf_hub_args� templateszmodelcard_template.mdzdatasetcard_template.mdz1^(\s*---[\r\n]+)([\S\s]*?)([\r\n]+---(\r\n|\n|$))c �~ � \ rS rSr\r\rSrS S\ S\64S jjr \ S 5 r
\
R S\ 4S j5 r
S rS \\ - 4S7 jr\ S S \ \- S
\ S - S\ S - S\84S jj5 rS S
\ S - 4S jjr S S\ S\ S - S
\ S - S\ S - S\ S - S\ S - S\9S - S\ S - 4S jjr\ S!S\S\ S - S\ S - 4S jj5 rS rg )"�RepoCard�% �model�content�ignore_metadata_errorsc � � X l Xl g)a� Initialize a RepoCard from string content. The content should be a10Markdown file with a YAML block at the beginning and a Markdown body.11 12Args:13 content (`str`): The content of the Markdown file.14 15Example:16 ```python17 >>> from huggingface_hub.repocard import RepoCard18 >>> text = '''19 ... ---20 ... language: en21 ... license: mit22 ... ---23 ...24 ... # My repo25 ... '''26 >>> card = RepoCard(text)27 >>> card.data.to_dict()28 {'language': 'en', 'license': 'mit'}29 >>> card.text30 '\n# My repo\n'31 32 ```33> [!TIP]34> Raises the following error:35>36> - [`ValueError`](https://docs.python.org/3/library/exceptions.html#ValueError)37> when the content of the repo card metadata is not a dictionary.38N)r r )�selfr r s �^D:\code\apps\devtools\python\user_packages\Python313\site-packages\huggingface_hub/repocard.py�__init__�RepoCard.__init__* s � �D '=�#� � � c � � [ U R 5 =( d SnSU U R R XR S9 U SU U R39 3$ )zLThe content of the RepoCard, including the YAML block and the Markdown body.�40�---)�41line_break�original_order)�_detect_line_ending�_content�data�to_yaml�_original_order�text)r" r* s r# r �RepoCard.contentO sz � � )�� � �7�?�4�42��Z�L����!2�!2�j�Ym�Ym�!2�!n� o�pz�o{�{~� @J� K� LP� LU� LU� KV� W� Wr& c �� � Xl [ R U5 nU( ad UR S5 nXR 5 S U l [ R " U5 nUc 0 n[ U[ 5 ( d [ S5 eO [ R S5 0 nXl U R " S0 UDSU R 0D6U l [! UR# 5 5 U l g)z Set the content of the RepoCard.� N�)repo card metadata block should be a dictzBRepo card metadata block was not found. Setting CardData to empty.r � )r- �REGEX_YAML_BLOCK�search�group�endr1 �yaml� safe_load�43isinstance�dict�44ValueError�logger�warning�card_data_classr r. �list�keysr0 )r" r �match�45yaml_block� data_dicts r# r r2 U s� � � �
� �'�'��0�� � ���Q� �J� � � � �
�.�D�I� ���z�2�I�� � � � �i��.�.� �!L�M�M� /�
�N�N�_� `��I� �I��(�(�i�9�i�T�Mh�Mh�i�� � #�I�N�N�$4� 5�� r& c � � U R $ �N)r )r" s r# �__str__�RepoCard.__str__p s � ��|�|�r& �filepathc �� � [ U5 nUR R SSS9 [ USSSS9 nUR [ U 5 5 SSS5 g! , ( d f g= f)a3 Save a RepoCard to a file.46 47Args:48 filepath (`Union[Path, str]`): Filepath to the markdown file to save.49 50Example:51 ```python52 >>> from huggingface_hub.repocard import RepoCard53 >>> card = RepoCard("---\nlanguage: en\n---\n# This is a test repo card")54 >>> card.save("/tmp/test.md")55 56 ```57T)�parents�exist_ok�w� �utf-8��mode�newline�encodingN)r �parent�mkdir�open�write�str)r" rL �fs r# �save�
RepoCard.saves sQ � � ��>����� � �d�T� �:�
�(��b�7�
C�q�
�G�G�C��I� � D�
C�
C�s �A�58A$N�repo_id_or_path� repo_type�tokenc � � [ U5 R 5 ( a [ U5 nO[[ U[ 5 ( a7 [ [ U[59 R U=( d U R US95 nO[ SU S35 eUR SSSS9 nU " UR 5 US9sS S S 5 $ ! , ( d f g = f)60a� Initialize a RepoCard from a Hugging Face Hub repo's README.md or a local filepath.61 62Args:63 repo_id_or_path (`Union[str, Path]`):64 The repo ID associated with a Hugging Face Hub repo or a local filepath.65 repo_type (`str`, *optional*):66 The type of Hugging Face repo to push to. Defaults to None, which will use "model". Other options67 are "dataset" and "space". Not used when loading from a local filepath. If this is called from a child68 class, the default value will be the child class's `repo_type`.69 token (`str`, *optional*):70 Authentication token, obtained with `huggingface_hub.HfApi.login` method. Will default to the stored token.71 ignore_metadata_errors (`str`):72 If True, errors while parsing the metadata section will be ignored. Some information might be lost during73 the process. Use it at your own risk.74 75Returns:76 [`huggingface_hub.repocard.RepoCard`]: The RepoCard (or subclass) initialized from the repo's77 README.md file or filepath.78 79Example:80 ```python81 >>> from huggingface_hub.repocard import RepoCard82 >>> card = RepoCard.load("nateraw/food")83 >>> assert card.data.tags == ["generated_from_trainer", "image-classification", "pytorch"]84 85 ```86)r` ra z.Cannot load RepoCard: path not found on disk (z).�rrQ rR rS )r N) r �is_filer= r[ r r �
REPOCARD_NAMEr` r? rY �read)�clsr_ r` ra r � card_pathr\ s r# �load�
RepoCard.load� s� � �H �� � (� (� *� *� �_�-�I�
���
-�
-� � �#� �+�+� '� 8�3�=�=� � ��I� � M�o�M^�^`� a�b� b� �^�^��b�7�^�
C�q��q�v�v�x�8N�O� D�
C�
C�s �B6�687Cc � � U=( d U R nU[ U 5 S.nSS0n [ 5 R SX#S9n[ U5 g! [88 a, nWR S:X a [ UR 5 eUeSnAff = f)a Validates card against Hugging Face Hub's card validation logic.89Using this function requires access to the internet, so it is only called90internally by [`huggingface_hub.repocard.RepoCard.push_to_hub`].91 92Args:93 repo_type (`str`, *optional*, defaults to "model"):94 The type of Hugging Face repo to push to. Options are "model", "dataset", and "space".95 If this function is called from a child class, the default will be the child class's `repo_type`.96 97> [!TIP]98> Raises the following errors:99>100> - [`ValueError`](https://docs.python.org/3/library/exceptions.html#ValueError)101> if the card fails validation checks.102> - [`HTTPError`](https://requests.readthedocs.io/en/latest/api/#requests.HTTPError)103> if the request to the Hub API fails for any other reason.104)�repoTyper �Acceptz105text/plainz(https://huggingface.co/api/validate-yaml)�json�headersi� N) r` r[ r �postr r � status_coder? r1 )r" r` �bodyro �response�excs r# �validate�RepoCard.validate� s� � �( �/�� � � � "��4�y�106�� �\�*�� �"�}�)�)�*T�[_�)�q�H� �� )�� � ��#�#�s�*� �� � �/�/�� �� �s �#A
�
107B�'A>�>B�repo_id�commit_message�commit_description�revision� create_pr�
parent_commitc �T � U=( d U R nU R US9 [ 5 n [ U 5 [ R108 - n109U110R
[ U 5 SS9 [ [ U1115 [ R112 UUUUUUUUS9113n SSS5 U $ ! , ( d f W $ = f)ar Push a RepoCard to a Hugging Face Hub repo.114 115Args:116 repo_id (`str`):117 The repo ID of the Hugging Face Hub repo to push to. Example: "nateraw/food".118 token (`str`, *optional*):119 Authentication token, obtained with `huggingface_hub.HfApi.login` method. Will default to120 the stored token.121 repo_type (`str`, *optional*, defaults to "model"):122 The type of Hugging Face repo to push to. Options are "model", "dataset", and "space". If this123 function is called by a child class, it will default to the child class's `repo_type`.124 commit_message (`str`, *optional*):125 The summary / title / first line of the generated commit.126 commit_description (`str`, *optional*)127 The description of the generated commit.128 revision (`str`, *optional*):129 The git revision to commit from. Defaults to the head of the `"main"` branch.130 create_pr (`bool`, *optional*):131 Whether or not to create a Pull Request with this commit. Defaults to `False`.132 parent_commit (`str`, *optional*):133 The OID / SHA of the parent commit, as a hexadecimal string. Shorthands (7 first characters) are also supported.134 If specified and `create_pr` is `False`, the commit will fail if `revision` does not point to `parent_commit`.135 If specified and `create_pr` is `True`, the pull request will be created from `parent_commit`.136 Specifying `parent_commit` ensures the repo has not changed before committing the changes, and can be137 especially useful if the repo is updated / committed too concurrently.138Returns:139 `str`: URL of the commit which updated the card metadata.140)r` rR )rV )141�path_or_fileobj� path_in_reporw ra r` rx ry r{ rz r| N) r` ru r r r re �142write_textr[ r ) r" rw ra r` rx ry rz r{ r| �tmpdir�tmp_path�urls r# � push_to_hub�RepoCard.push_to_hub� s� � �R �/�� � � �
�
�
� �
�*�
#�
%���F�|�i�&=�&=�=�H� � � ��D� �G� � <� � #�H�
� &� 4� 4� ��#� -�#5�#�!� +� �C� &� �143� &�
%� �144�s �A B�145B'� card_data�
template_path� template_strc � � [ 5 ( a SSKnO [ S5 eUR 5 R 5 nUR U5 Ub [
U5 R 5 nUc# [
U R 5 R 5 nUR U5 nUR " SSUR 5 0UD6nU " U5 $ )a� Initialize a RepoCard from a template. By default, it uses the default template.146 147Templates are Jinja2 templates that can be customized by passing keyword arguments.148 149Args:150 card_data (`huggingface_hub.CardData`):151 A huggingface_hub.CardData instance containing the metadata you want to include in the YAML152 header of the repo card on the Hugging Face Hub.153 template_path (`str`, *optional*):154 A path to a markdown file with optional Jinja template variables that can be filled155 in with `template_kwargs`. Defaults to the default template.156 157Returns:158 [`huggingface_hub.repocard.RepoCard`]: A RepoCard instance with the specified card data and content from the159 template.160r NzjUsing RepoCard.from_template requires Jinja2 to be installed. Please install it with `pip install Jinja2`.r� r6 ) r �jinja2� ImportError�to_dict�copy�updater � read_text�default_template_path�Template�renderr/ ) rg r� r� r� �template_kwargsr� �kwargs�templater s r# �
from_template�RepoCard.from_template! s� � �0 � � � � �9��
�161 �"�"�$�)�)�+���
�
�o�&� � $� �
�.�8�8�:�L� � � �� 9� 9�:�D�D�F�L��?�?�<�0���/�/�J�I�,=�,=�,?�J�6�J���7�|�r& )r- r0 r r. r r1 )F)NNFrI )NNNNNNN�NN)�__name__�162__module__� __qualname__�__firstlineno__r rB �TEMPLATE_MODELCARD_PATHr� r` r[ �boolr$ �propertyr �setterrJ r r] � classmethodri ru r� r� �__static_attributes__r6 r& r# r r % s� � � �O� 3���I�# �� # �T� # �J �W� �W�163
�^�^�6�s� 6� �6�4 � �T�C�Z� �( � !%� �',� 3P� �t��3P� ��:�3P� �T�z� 3P�164 !%� 3P� �3P�j#�#��*� #�P !� $�%)�)-� #�!%�$(�=��=� �T�z�=� ��:� =�165 �d�166� =� �$�J�
=� ��*�=� �$�;�=� �T�z�=�~ � %)�#'� ( ��( � �T�z�( � �D�j� ( � �( r& r c �b ^ � \ rS rSr\r\rSr\ S S\S\167S- S\168S- 4U 4S jjj5 r Sr U =r
$ )169� ModelCardiM r Nr� r� r� c �( >� [ TU ] " XU40 UD6$ )a5 Initialize a ModelCard from a template. By default, it uses the default template, which can be found here:170https://github.com/huggingface/huggingface_hub/blob/main/src/huggingface_hub/templates/modelcard_template.md171 172Templates are Jinja2 templates that can be customized by passing keyword arguments.173 174Args:175 card_data (`huggingface_hub.ModelCardData`):176 A huggingface_hub.ModelCardData instance containing the metadata you want to include in the YAML177 header of the model card on the Hugging Face Hub.178 template_path (`str`, *optional*):179 A path to a markdown file with optional Jinja template variables that can be filled180 in with `template_kwargs`. Defaults to the default template.181 182Returns:183 [`huggingface_hub.ModelCard`]: A ModelCard instance with the specified card data and content from the184 template.185 186Example:187 ```python188 >>> from huggingface_hub import ModelCard, ModelCardData, EvalResult189 190 >>> # Using the Default Template191 >>> card_data = ModelCardData(192 ... language='en',193 ... license='mit',194 ... library_name='timm',195 ... tags=['image-classification', 'resnet'],196 ... datasets=['beans'],197 ... metrics=['accuracy'],198 ... )199 >>> card = ModelCard.from_template(200 ... card_data,201 ... model_description='This model does x + y...'202 ... )203 204 >>> # Including Evaluation Results205 >>> card_data = ModelCardData(206 ... language='en',207 ... tags=['image-classification', 'resnet'],208 ... eval_results=[209 ... EvalResult(210 ... task_type='image-classification',211 ... dataset_type='beans',212 ... dataset_name='Beans',213 ... metric_type='accuracy',214 ... metric_value=0.9,215 ... ),216 ... ],217 ... model_name='my-cool-model',218 ... )219 >>> card = ModelCard.from_template(card_data)220 221 >>> # Using a Custom Template222 >>> card_data = ModelCardData(223 ... language='en',224 ... tags=['image-classification', 'resnet']225 ... )226 >>> card = ModelCard.from_template(227 ... card_data=card_data,228 ... template_path='./src/huggingface_hub/templates/modelcard_template.md',229 ... custom_template_var='custom value', # will be replaced in template if it exists230 ... )231 232 ```233��superr� �rg r� r� r� r� � __class__s �r# r� �ModelCard.from_templateR s �� �R �w�$�Y�|�_��_�_r& r6 r� )r� r� r� r� r rB r� r� r` r� r[ r� r� �
__classcell__�r� s @r# r� r� M s[ �� �#�O� 3���I�� %)�#'� H`� �H`� �T�z�H`� �D�j� H`� �H`r& r� c �b ^ � \ rS rSr\r\rSr\ S S\S\234S- S\235S- 4U 4S jjj5 r Sr U =r
$ )236� DatasetCardi� �datasetNr� r� r� c �( >� [ TU ] " XU40 UD6$ )ad Initialize a DatasetCard from a template. By default, it uses the default template, which can be found here:237https://github.com/huggingface/huggingface_hub/blob/main/src/huggingface_hub/templates/datasetcard_template.md238 239Templates are Jinja2 templates that can be customized by passing keyword arguments.240 241Args:242 card_data (`huggingface_hub.DatasetCardData`):243 A huggingface_hub.DatasetCardData instance containing the metadata you want to include in the YAML244 header of the dataset card on the Hugging Face Hub.245 template_path (`str`, *optional*):246 A path to a markdown file with optional Jinja template variables that can be filled247 in with `template_kwargs`. Defaults to the default template.248 249Returns:250 [`huggingface_hub.DatasetCard`]: A DatasetCard instance with the specified card data and content from the251 template.252 253Example:254 ```python255 >>> from huggingface_hub import DatasetCard, DatasetCardData256 257 >>> # Using the Default Template258 >>> card_data = DatasetCardData(259 ... language='en',260 ... license='mit',261 ... annotations_creators='crowdsourced',262 ... task_categories=['text-classification'],263 ... task_ids=['sentiment-classification', 'text-scoring'],264 ... multilinguality='monolingual',265 ... pretty_name='My Text Classification Dataset',266 ... )267 >>> card = DatasetCard.from_template(268 ... card_data,269 ... pretty_name=card_data.pretty_name,270 ... )271 272 >>> # Using a Custom Template273 >>> card_data = DatasetCardData(274 ... language='en',275 ... license='mit',276 ... )277 >>> card = DatasetCard.from_template(278 ... card_data=card_data,279 ... template_path='./src/huggingface_hub/templates/datasetcard_template.md',280 ... custom_template_var='custom value', # will be replaced in template if it exists281 ... )282 283 ```284r� r� s �r# r� �DatasetCard.from_template� s �� �r �w�$�Y�|�_��_�_r& r6 r� )r� r� r� r� r rB �TEMPLATE_DATASETCARD_PATHr� r` r� r[ r� r� r� r� s @r# r� r� � sV �� �%�O� 5���I�� %)�#'� 8`�"�8`� �T�z�8`� �D�j� 8`� �8`r& r� c � � \ rS rSr\r\rSrSr g)� SpaceCardi� �spacer6 N)285r� r� r� r� r rB r� r� r` r� r6 r& r# r� r� � s � �#�O� 3���Ir& r� r �return)�
r( �
286Nc � � U R S5 nU R S5 nU R S5 nX- S:X a gX1:X a X2:X a gX:� a gg)z�Detect the line ending of a string. Used by RepoCard to avoid making huge diff on newlines.287 288Uses same implementation as in Hub server, keep it in sync.289 290Returns:291 str: The detected line ending of the string.292r� r( r� r N)�count)r �cr�lf�crlfs r# r, r, � sQ � � 293� � �t� �B� � � �t� �B� �=�=�� �D� �w�!�|�� �z�d�j�� �w��r& �294local_pathc � � [ U 5 R 5 n[ R U5 nU( aL UR S5 n[295 R " U5 nUb [ U[ 5 ( a U$ [ S5 eg )Nr4 r5 )296r r� r7 r8 r9 r; r< r= r> r? )r� r rE rF r. s r# �
metadata_loadr� � sh � ��:� �(�(�*�G� � #� #�G� ,�E� ��[�[��^�297��~�~�j�)�� �<�:�d�D�1�1��K��D�E�E�r& r. c �� � SnSn[ R R U 5 ( a~ [ U SSS9 nUR 5 n[ UR [ 5 ( a UR S nO+[ UR [ 5 ( a UR nSSS5 [ U SSSS9 n[ USUS 9n[ R U5 nU( a2 USUR 5 S298U U S299U 3- X6R 5 S - nO S300U U S301U U 3nUR U5 UR 5 SSS5 g! , ( d f N�= f! , ( d f g= f) a 302Save the metadata dict in the upper YAML part Trying to preserve newlines as303in the existing file. Docs about open() with newline="" parameter:304https://docs.python.org/3/library/functions.html?highlight=open#open Does305not work with "^M" linebreaks, which are replaced by 306 307r( rQ �utf8)rU rV r NrP F)� sort_keysr* r) )�os�path�existsrY rf r= �newlines�tupler[ r r7 r8 �startr: rZ �close)r� r. r* r �readme� data_yamlrE �outputs r# �
metadata_saver� s2 � � �J��G� �w�w�~�~�j�!�!�
�*�b�6�
:�f� �k�k�m�G��&�/�/�5�1�1� #�_�_�Q� /�308��F�O�O�S�1�1� #�_�_�309� ;� 310�j�#�r�F� ;�v� �d�e�311�K� � �'�'��0�� � �_�u�{�{�}�-�#�j�\�)��C�PZ�|�0\�\�_f�gp�gp�gr�gt�_u�u�F��:�,�y�k��Z�L�� �J�F�� � �V� �� � �� 312<� ;� ;�
:�� 313<� ;�s �A+E�3BE�314E�315E"F)�metrics_config�metrics_verified�dataset_config�
dataset_split�dataset_revision�metrics_verification_token�model_pretty_name�task_pretty_name�task_id�metrics_pretty_name�316metrics_id�
metrics_value�dataset_pretty_name�317dataset_idr� r� r� r� r� r� c �B � S[ U [ UUUUUUUUU U
U318U U S9
/S90$ )u� 319Creates a metadata dict with the result from a model evaluated on a dataset.320 321Args:322 model_pretty_name (`str`):323 The name of the model in natural language.324 task_pretty_name (`str`):325 The name of a task in natural language.326 task_id (`str`):327 Example: automatic-speech-recognition. A task id.328 metrics_pretty_name (`str`):329 A name for the metric in natural language. Example: Test WER.330 metrics_id (`str`):331 Example: wer. A metric id from https://hf.co/metrics.332 metrics_value (`Any`):333 The value from the metric. Example: 20.0 or "20.0 ± 1.2".334 dataset_pretty_name (`str`):335 The name of the dataset in natural language.336 dataset_id (`str`):337 Example: common_voice. A dataset id from https://hf.co/datasets.338 metrics_config (`str`, *optional*):339 The name of the metric configuration used in `load_metric()`.340 Example: bleurt-large-512 in `load_metric("bleurt", "bleurt-large-512")`.341 metrics_verified (`bool`, *optional*, defaults to `False`):342 Indicates whether the metrics originate from Hugging Face's [evaluation service](https://huggingface.co/spaces/autoevaluate/model-evaluator) or not. Automatically computed by Hugging Face, do not set.343 dataset_config (`str`, *optional*):344 Example: fr. The name of the dataset configuration used in `load_dataset()`.345 dataset_split (`str`, *optional*):346 Example: test. The name of the dataset split used in `load_dataset()`.347 dataset_revision (`str`, *optional*):348 Example: 5503434ddd753f426f4b38109466949a1217c2bb. The name of the dataset dataset revision349 used in `load_dataset()`.350 metrics_verification_token (`bool`, *optional*):351 A JSON Web Token that is used to verify whether the metrics originate from Hugging Face's [evaluation service](https://huggingface.co/spaces/autoevaluate/model-evaluator) or not.352 353Returns:354 `dict`: a metadata dict with the result from a model evaluated on a dataset.355 356Example:357 ```python358 >>> from huggingface_hub import metadata_eval_result359 >>> results = metadata_eval_result(360 ... model_pretty_name="RoBERTa fine-tuned on ReactionGIF",361 ... task_pretty_name="Text Classification",362 ... task_id="text-classification",363 ... metrics_pretty_name="Accuracy",364 ... metrics_id="accuracy",365 ... metrics_value=0.2662102282047272,366 ... dataset_pretty_name="ReactionJPEG",367 ... dataset_id="julien-c/reactionjpeg",368 ... dataset_config="default",369 ... dataset_split="test",370 ... )371 >>> results == {372 ... 'model-index': [373 ... {374 ... 'name': 'RoBERTa fine-tuned on ReactionGIF',375 ... 'results': [376 ... {377 ... 'task': {378 ... 'type': 'text-classification',379 ... 'name': 'Text Classification'380 ... },381 ... 'dataset': {382 ... 'name': 'ReactionJPEG',383 ... 'type': 'julien-c/reactionjpeg',384 ... 'config': 'default',385 ... 'split': 'test'386 ... },387 ... 'metrics': [388 ... {389 ... 'type': 'accuracy',390 ... 'value': 0.2662102282047272,391 ... 'name': 'Accuracy',392 ... 'verified': False393 ... }394 ... ]395 ... }396 ... ]397 ... }398 ... ]399 ... }400 True401 402 ```403� model-index)
� task_name� task_type� metric_name� metric_type� metric_value� dataset_name� dataset_type�
metric_config�verified� verify_tokenr� r� r� )�404model_name� eval_results)r
r405 )r� r� r� r� r� r� r� r� r� r� r� r� r� r� s r# �metadata_eval_resultr� ' sQ � �R �2�(�� .� %� 3� *�!.�!4�!+�"0� -�!;�#1�"/�%5���406� � r& )r` � overwritera rx ry rz r{ r| rw �metadatar` r� ra rx ry rz r{ r| c �� � Ub UOSnUb US:X a [ n407O(US:X a [ n408OUS:X a [ n409O[ SU 35 e U410R XUS9n UR 5 GH� u p�U S:X Ga= S U
S411 ; a [ U S U 5 U
S412 S ' [ U
5 u p�U R R c" X�R l X�R l
Mo U R R nU H� nS nU H� nUR U5 ( d M UU:w a- U( d& [ S
UR SUR S35 eSnUR" Ul UR$ SL d Mr UR& Ul M� U( a M� U R R R) U5 M� GMJ U R R+ U 5 b5 U( d. U R R+ U 5 U
:w a [ SU S35 eX�R U ' GM� U R- U UUUUUUU S9$ ! [413 a. US:X a [ S5 eU414R
[ 5 5 n GN f = f)a 415Updates the metadata in the README.md of a repository on the Hugging Face Hub.416If the README.md file doesn't exist yet, a new one is created with metadata and417the default ModelCard or DatasetCard template. For `space` repo, an error is thrown418as a Space cannot exist without a `README.md` file.419 420Args:421 repo_id (`str`):422 The name of the repository.423 metadata (`dict`):424 A dictionary containing the metadata to be updated.425 repo_type (`str`, *optional*):426 Set to `"dataset"` or `"space"` if updating to a dataset or space,427 `None` or `"model"` if updating to a model. Default is `None`.428 overwrite (`bool`, *optional*, defaults to `False`):429 If set to `True` an existing field can be overwritten, otherwise430 attempting to overwrite an existing field will cause an error.431 token (`str`, *optional*):432 The Hugging Face authentication token.433 commit_message (`str`, *optional*):434 The summary / title / first line of the generated commit. Defaults to435 `f"Update metadata with huggingface_hub"`436 commit_description (`str` *optional*)437 The description of the generated commit438 revision (`str`, *optional*):439 The git revision to commit from. Defaults to the head of the440 `"main"` branch.441 create_pr (`boolean`, *optional*):442 Whether or not to create a Pull Request from `revision` with that commit.443 Defaults to `False`.444 parent_commit (`str`, *optional*):445 The OID / SHA of the parent commit, as a hexadecimal string. Shorthands (7 first characters) are also supported.446 If specified and `create_pr` is `False`, the commit will fail if `revision` does not point to `parent_commit`.447 If specified and `create_pr` is `True`, the pull request will be created from `parent_commit`.448 Specifying `parent_commit` ensures the repo has not changed before committing the changes, and can be449 especially useful if the repo is updated / committed too concurrently.450Returns:451 `str`: URL of the commit which updated the card metadata.452 453Example:454 ```python455 >>> from huggingface_hub import metadata_update456 >>> metadata = {'model-index': [{'name': 'RoBERTa fine-tuned on ReactionGIF',457 ... 'results': [{'dataset': {'name': 'ReactionGIF',458 ... 'type': 'julien-c/reactiongif'},459 ... 'metrics': [{'name': 'Recall',460 ... 'type': 'recall',461 ... 'value': 0.7762102282047272}],462 ... 'task': {'name': 'Text Classification',463 ... 'type': 'text-classification'}}]}]}464 >>> url = metadata_update("hf-internal-testing/reactiongif-roberta-card", metadata)465 466 ```467z$Update metadata with huggingface_hubr r� r� zUnknown repo_type: )ra r` zJCannot update metadata on a Space that doesn't contain a `README.md` file.r� �namer r� Fz6You passed a new value for the existing metric 'name: z, type: z6'. Set `overwrite=True` to overwrite existing metrics.Tz9You passed a new value for the existing meta data field 'z7'. Set `overwrite=True` to overwrite existing metadata.)ra r` rx ry r{ rz r| )r� r� r r? ri r r� r �items�getattrr r. r� r� �is_equal_except_valuer� r� r� r� r� �append�getr� )rw r� r` r� ra rx ry rz r{ r| �468card_class�card�key�valuer� � new_results�existing_results�469new_result� result_found�existing_results r# �metadata_updater� � ss � �H (6�'A�^�Gm�N� ��I�� 0� �470� �i� � �471� �g� � �472��.�y�k�:�;�;�4����w�y��I�� �n�n�&�473�� �-� ��U�1�X�%�#*�4� �w�#G��a��� �&A�%�&H� #�J��y�y�%�%�-�)4� � �&�'1� � �$�#'�9�9�#9�#9� � #.�J�#(�L�+;��%�;�;�O�L�L� )�_� <�Y�&0�%/�/9�/E�/E�.F�h�'1�'=�'=�&>� ?F�%F�'"� !"� ,0�L�;E�;R�;R�O� 8� .� 7� 7�4� ?�?I�?V�?V�� <� ,<� (�<� � � �.�.�5�5�j�A�! #.�& �y�y�}�}�S�!�-�i�D�I�I�M�M�RU�DV�Z_�D_� �O�PS�u� UL� L�� �474 "'� � �#��W '�Z � � ����%�-���#� � � ��k � 4� �� � � i�j� j� �'�'��475�3��
4�s �H% �%4I � I )7r� �re�pathlibr �typingr r r; � huggingface_hub.file_downloadr �huggingface_hub.hf_apir � huggingface_hub.repocard_datar r r476 r r r
r �huggingface_hub.utilsr r r r r rQ r �errorsr �utilsr r r �477get_loggerr� r@ �__file__rW r� r� �compiler7 r r� r� r� r[ r, r> r� r� r� r� r� r6 r& r# �<module>r s� �� � � � � � 9� .�� � � r� q� � &� H� H� 478� � �H� %�� �x�.�/�/�+�=�@W�W� � ��N� 1� 1�K� ?�B[� [� � �:�:� R�S� �e � e �P N`�� N`�b>`�(� >`�B�� � �� ��1I�)J� �*479�c�D�j� 480�T�D�[� 481� �c�D�j� �� �� �T "&� "�!%� $�#'�-1� }��}� �}� � }�482 � }� �
}� �}� �}� �}� �$�J�}� �}� �$�J�}� ��:�}� �D�j� }� !$�d�483� }� 484�!}�@ �485 !�� �!%�%)� �� $�R�
�R��R� �T�z� R�486 � R� ��:�
R� �$�J�R� �d�487�R� �D�j�R� �R� ��:�R� �R� �Rr&