U -ez@sdZddlZddlZddlZddlZddlZddlZddlZddlZddl Z ddl m Z ddl m Z mZddlmZddlmZmZmZmZmZmZmZmZmZddlZddlZddlmZmZm Z dd l!m"Z"dd l#m$Z$dd l%m&Z&m'Z'dd l(m)Z)m*Z*m+Z+m,Z,m-Z-m.Z.m/Z/dd l0m1Z1m2Z2ddl3m4Z4ddl5m6Z6ddl7m8Z8m9Z9m:Z:m;Z;ddlm?Z?m@Z@ddlAmBZBddlCmDZDmEZEddlFmGZGddlHmIZIddlJmKZKmLZLddlMmNZNmOZOmPZPmQZQmRZRddlSmTZTddlUmVZVddlWmXZXmYZYmZZZm[Z[m\Z\m]Z]m^Z^m_Z_m`Z`ddlambZbddlcmdZdddlemfZfmgZgddlhmiZidd ljmkZkdd!llmmZmdd"lnmoZoeiepZqereNsd#gZte"judfeveeeevfd$d%d&Zwdeeee'eeIfd(d)d*ZxGd+d,d,Zyee'ee&eevevee'd-d.d/Zzdd0eevee'd1d2d3Z{eevevd4d5d6Z|evevd7d8d9Z}eveveevevevevfee4eeevevfd:d;d<Z~eveveveveeevevfeeevevfeee6evfevd=d>d?Zeveeevevfeeevevfeveveve6eevevfd@dAdBZde+ee4eeevevfdCdDdEZde+ee4eeevevfdCdFdGZde*eevee4eeeveeveffdHdIdJZeveevdKdLdMZdevekeeeveevefee4eee&evfdNdOdPZe GdQdRdRZe GdSd0d0Ze GdTdUdUZGdVdWdWZGdXdYdYZGdZd[d[eZGd\d]d]eZGd^d_d_eZGd`dadaeZGdbdcdceZGdddedeeZGdfdgdgeZGdhdidieZGdjdkdkeZdeveeeveofee4eee6evfeeveeveeeeeve*fedldmdnZeVdodeveeeveofee4eee6evfeevedpdqdrZeVdsdeveeveeeeveeveee4eee6evfeeeveofeIdu dvdwZdeveeveeveeeveeveeveeveevfffeevee=ee4eee6evfeeeveofeeeevfeee'dy dzd{ZdeveeveeveeeveeveeveeveevfffeeeveTfeevee=ee4eee6evfeeefevfeeeeeeveofeeeevfeeeeeee1e$e2eGfd|d}d~Zdeveeeeee$e1fdddZdS)zAccess datasets.N)Counter) dataclassfield)Path) AnyDictListMappingOptionalSequenceTupleTypeUnion) DatasetCardDatasetCardDataHfApi)config)Dataset) BuilderConfigDatasetBuilder)DEFAULT_PATTERNS_ALL DataFilesDict DataFilesListEmptyDatasetErrorget_data_patternsget_metadata_patternssanitize_patterns) DatasetDictIterableDatasetDict)DownloadConfig) DownloadMode)StreamingDownloadManager xbasenamexglobxjoin)Features)extract_path_from_uriis_remote_filesystem)Hasher) DatasetInfoDatasetInfosDict)IterableDataset)Metric)camelcase_to_snakecasesnakecase_to_camelcase)_EXTENSION_TO_MODULE_MODULE_SUPPORTS_METADATA_MODULE_TO_EXTENSIONS_PACKAGED_DATASETS_MODULES_hash_python_lines)Split) deprecated) OfflineModeIsEnabled!_raise_if_offline_mode_is_enabled cached_path head_hf_s3 hf_github_urlinit_hf_modulesis_relative_pathrelative_to_absolute_pathurl_or_path_join)FileLock) hf_hub_url)VerificationModeis_small_dataset) get_logger)MetadataConfigs) get_imports)Version.zip)namehf_modules_cachec Cs^t|}tj||}tj|ddtjtj|dsZttj|ddW5QRX|S)a^ Create a module with name `name` in which you can add dynamic modules such as metrics or datasets. The module can be imported using its name. The module is created in the HF_MODULE_CACHE directory by default (~/.cache/huggingface/modules) but it can be overridden by specifying a path to another directory in `hf_modules_cache`. Texist_ok __init__.pyw)r<ospathjoinmakedirsexistsopen)rIrJdynamic_modules_pathrVN/var/www/html/Darija-Ai-Train/env/lib/python3.8/site-packages/datasets/load.pyinit_dynamic_modules\s  rXTreturncCsvt|}|rt}nt}d}|jD]J\}}t|r&t||r&t |rNq&|}t |}|dk r&||kr&qrq&|S)zImport a module at module_path and return its main class: - a DatasetBuilder if dataset is True - a Metric if dataset is False N) importlib import_modulerr-__dict__itemsinspectisclass issubclass isabstract getmodule) module_pathdatasetmoduleZ main_cls_typeZmodule_main_clsrIobjZ obj_modulerVrVrWimport_main_classns   rhc@seZdZdZddZdS)#_InitializeConfiguredDatasetBuilderaL From https://stackoverflow.com/questions/4647566/pickle-a-dynamically-parameterized-sub-class See also ConfiguredDatasetBuilder.__reduce__ When called with the param value as the only argument, returns an un-initialized instance of the parameterized class. Subsequent __setstate__ will be called by pickle. cCst}t||||d|_|S)N)default_config_name dataset_name)riconfigure_builder_class __class__)self builder_clsmetadata_configsrjrIrgrVrVrW__call__sz,_InitializeConfiguredDatasetBuilder.__call__N)__name__ __module__ __qualname____doc__rqrVrVrVrWrisri)robuilder_configsrjrkrZcsTGfddd}jt||_jt||_|S)z Dynamically create a builder class with custom builder configs parsed from README.md file, i.e. set BUILDER_CONFIGS class variable of a builder class to custom configs list. cs"eZdZZZjZddZdS)z9configure_builder_class..ConfiguredDatasetBuildercSs,|jjd}t||j|j|jf|jfS)Nr)rm__mro__riBUILDER_CONFIGSDEFAULT_CONFIG_NAMErkr]copy)rnZparent_builder_clsrVrVrW __reduce__s zDconfigure_builder_class..ConfiguredDatasetBuilder.__reduce__N)rrrsrtrxryr{rVrorvrjrVrWConfiguredDatasetBuildersr})rrlower capitalizer/rt)rorvrjrkr}rVr|rWrls rl DatasetModule)dataset_modulerkrZcCs.t|j}|jjr*t||jj|jj|d}|S)N)rvrjrk)rhrdbuilder_configs_parametersrvrlrj)rrkrorVrVrWget_dataset_builder_classs r) file_pathsrZc Cszg}|D]4}tj|r2|tt|dq||qg}|D]*}t|dd}|| W5QRXqFt |S)zt Convert a list of scripts or text files provided in file_paths into a hashed filename in a repeatable way. z *.[pP][yY]utf-8encoding) rOrPisdirextendlistrrglobappendrT readlinesr4)rZ to_use_files file_pathlinesfrVrVrW files_to_hashs  rrI resource_typecCs@tjsdnd}t | dkrTdnd}d| krnd| d<t d|d|dd| d|dd| d |S)a Download additional module for a module .py at URL (or local path) /.py The imports must have been parsed first using ``get_imports``. If some modules need to be installed with pip, an error is raised showing how to install them. This function return the list of downloaded modules as tuples (import_name, module_file_path). The downloaded modules can then be moved into an importable directory with ``_copy_script_and_other_resources_in_importable_dir``. NzDownloading extra modulesZlibraryz Error in the z script, importing relative z module but z: is the name of the script. Please change relative import zl to another name and add a '# From: URL_OR_PATH' comment pointing to the original relative import file path.ZinternalrZexternalzWrong import_typerr dependencies dependencyZthemitZsklearnz scikit-learnzTo be able to use z$, you need to install the following : z, z. Please install z using 'pip install  z' for instance.)rz download_descr ValueErrorr?r9rOrPrQr[r\ ImportErrorlenkeysvalues)rIrrr local_importsZlibrary_importsZ import_type import_name import_pathZ sub_directoryZurl_or_filenameZlocal_import_pathZneeds_to_be_installedZlibrary_import_nameZlibrary_import_pathlibZ_dependencies_strZ _them_strrVrVrW_download_additional_modulessR   2r)rIimportable_directory_pathsubdirectory_nameoriginal_local_pathradditional_files download_moderZc Cs,tj||}tj||d}|d} t| |tjkrTtj|rTt|tj |ddtj|d} tj| st | dW5QRXtj |ddtj|d} tj| st | dW5QRXtj|st ||tj |dd} tj| s:||d } t | dd d } t | | W5QRX|D]\}}tj|rtj||d}tj|st ||nFtj|rtj||}tj|st||ntd |q>|D]@\}}tj||}tj|rt||st ||q|W5QRSQRXd S)aCopy a script and its required imports to an importable directory Args: name (str): name of the resource to load importable_directory_path (str): path to the loadable folder in the dynamic modules directory subdirectory_name (str): name of the subdirectory in importable_directory_path in which to place the script original_local_path (str): local path to the resource script local_imports (List[Tuple[str, str]]): list of (destination_filename, import_file_to_copy) additional_files (List[Tuple[str, str]]): list of (destination_filename, additional_file_to_copy) download_mode (Optional[Union[DownloadMode, str]]): download mode Return: importable_local_file: path to an importable module with importlib.import_module rz.lockTrKrMrNrz.json)zoriginal file pathzlocal file pathrrzError with local import at N)rOrPrQr@r!FORCE_REDOWNLOADrSshutilrmtreerRrTcopyfilesplitextjsondumpisfilercopytreerfilecmpcmp)rIrrrrrrZimportable_subdirectoryimportable_local_fileZ lock_pathZinit_file_path meta_pathmetaZ meta_filerrZfull_path_local_import file_nameZ original_pathZdestination_additional_pathrVrVrW2_copy_script_and_other_resources_in_importable_dir.sR             r) local_pathrrrUmodule_namespacerIrrZc Cstj|||dd}t|jdddt|jdjddt|gdd|D}t | dd ||||||d } t d | d tj |||dd|| dd g} | |fS) N/--T)parentsrLrMrKcSsg|] }|dqS)rrV).0locrVrVrW sz+_create_importable_file..)rIrrrrrrz#Created importable dataset file at .)rOrPrQreplacermkdirparenttouchrrsplitloggerdebugbasename) rrrrUrrIrrhashrrdrVrVrW_create_importable_files$   &r)data_files_listrrZcCstdd|dtjD}|rtttftttfddd}t||ddD]4\}}|t krnt |S|d krRt ||d SqRdifS) a*Infer module (and builder kwargs) from list of data files. It picks the module based on the most common file extension. In case of a draw ".parquet" is the favorite, and then alphabetical order. Args: data_files_list (DataFilesList): List of data files. download_config (bool or str, optional): mainly use use_auth_token or storage_options to support different platforms and auth types. Returns: tuple[str, dict[str, Any]]: Tuple with - inferred module name - dict of builder kwargs css6|].}t|dddD]}d|VqqdSrrNr#rr~rfilepathsuffixrVrVrW sz3infer_module_for_data_files_list..N) ext_countrZcSs|\}}||dk|fS)zBSort by count and set ".parquet" as the favorite in case of a drawz.parquetrV)rextcountrVrVrWsort_keysz2infer_module_for_data_files_list..sort_keyT)keyreverserHr) rrZ*DATA_FILES_MAX_NUMBER_FOR_MODULE_INFERENCEr strintboolsortedr^r0,infer_module_for_data_files_list_in_archives)rrextensions_counterrr_rVrVrW infer_module_for_data_files_lists   rcCsg}d}|D]`}t|dr |d7}|tjkr4qntt|d}|ddt|d|dd tjD7}q t d d |D}|r| ddd}|t krt |Sd ifS) aInfer module (and builder kwargs) from list of archive data files. Args: data_files_list (DataFilesList): List of data files. download_config (bool or str, optional): mainly use use_auth_token or storage_options to support different platforms and auth types. Returns: tuple[str, dict[str, Any]]: Tuple with - inferred module name - dict of builder kwargs rrHrz**cSsg|]}|ddqS)z::rr)rrrVrVrWrsz@infer_module_for_data_files_list_in_archives..T) recursiverNcss6|].}t|dddD]}d|VqqdSrrrrVrVrWrsz?infer_module_for_data_files_list_in_archives..) rendswithrZ2GLOBBED_DATA_FILES_MAX_NUMBER_FOR_MODULE_INFERENCEr%r"extractr$Z3ARCHIVED_DATA_FILES_MAX_NUMBER_FOR_MODULE_INFERENCEr most_commonr0)rrZarchived_filesZarchive_files_counterrZ extractedrrrVrVrWrs*  r) data_filesrPrrZcsfdd|D}tt|\tfdd|DrTtd|sz|rhd|dnd}td|fS) aInfer module (and builder kwargs) from data files. Raise if module names for different splits don't match. Args: data_files (DataFilesDict): List of data files. path (str, optional): Dataset name or path. DownloadConfig (bool or str, optional): for authenticate on the Hugging Face Hub for private remote files. Returns: tuple[str, dict[str, Any]]: Tuple with - inferred module name - builder kwargs csi|]\}}|t|dqS)r)rrrrrrVrW sz/infer_module_for_data_files..c3s|]}f|kVqdSNrV)rZ split_module)default_builder_kwargs module_namerVrWrsz.infer_module_for_data_files..z=Couldn't infer the same data file format for all splits. Got z in z. z1No (supported) data files or dataset script found)r^nextiterranyrFileNotFoundError)rrPrZ split_modulesrV)rrrrWinfer_module_for_data_filess r)rconfig_parametersrZcsFdddhfddt|D}t}|||||S)z Used to update hash of packaged modules which is used for creating unique cache directories to reflect different config parameters which are passed in metadata from readme. config_nameversion descriptioncsi|]\}}|kr||qSrVrVrparamvalueZparams_to_excluderVrWrsz6update_hash_with_config_parameters..)rr^r)update hexdigest)rrZparams_to_add_to_hashmrVrrW"update_hash_with_config_parameterss     r)rdrpsupports_metadatarrrrZc st|}|j|}g}|dk r&|nd}|D]p\} } | d} | d} | r`|d| n|} z.| dk rvt| nt| }tj|| t |d}Wn:t k r}zt d|d| d|W5d}~XYnX| dkrH|rH|t krHz t |}Wnt k rd}YnX|dk rHtj||d rHtfd d |D}fd d | D}|rrtd|d|f| || dfdd || Dq2||fS)Nrdata_dirrrallowed_extensionsrz Dataset at 'z?' doesn't contain data files matching the patterns for config 'zZ', check `data_files` and `data_fir` parameters in the `configs` YAML field in README.md. rcsi|]\}}||qSrVrVr)config_metadata_data_files_listrVrWrEsz@create_builder_configs_from_metadata_configs..cs"g|]}t|s|dkr|qS)defaulthasattr)rrbuilder_config_clsrVrWrJs z@create_builder_configs_from_metadata_configs..z#Some datasets params were ignored: zx. Make sure to use only valid params for the dataset builder and to have a up-to-date version of the `datasets` library.)rIrr cs(i|] \}}t|r|dkr||qS))rrr rrrrVrWrXs  )rhZBUILDER_CONFIG_CLASSZget_default_config_namer^getrrr from_patternsALL_ALLOWED_EXTENSIONSrrrrrrwarningr)rdrprrrrrorjrvrZ config_paramsZconfig_data_filesZconfig_data_dirZconfig_base_pathZconfig_patternsZconfig_data_files_dicteZconfig_metadata_patternsZignored_paramsrV)rr rW,create_builder_configs_from_metadata_configssz             rc@sFeZdZUdZdZeeed<dZee e ed<dZ ee ed<dS)BuilderConfigsParametersaDataclass containing objects related to creation of builder configurations from yaml's metadata content. Attributes: metadata_configs (`MetadataConfigs`, *optional*): Configs parsed from yaml's metadata. builder_configs (`list[BuilderConfig]`, *optional*): List of BuilderConfig objects created from metadata_configs above. default_config_name (`str`): Name of default config taken from yaml's metadata. Nrprvrj) rrrsrtrurpr rE__annotations__rvrrrjrrVrVrVrWrbs  rc@sHeZdZUeed<eed<eed<eedZeed<dZ e e ed<dS)rrdrbuilder_kwargs)default_factoryrN dataset_infos) rrrsrtrrdictrrrrr r+rVrVrVrWrts c@seZdZUeed<eed<dS) MetricModulerdrN)rrrsrtrrrVrVrVrWr}s rc@seZdZedddZdS)_DatasetModuleFactoryrYcCstdSrNotImplementedErrorrnrVrVrW get_modulesz _DatasetModuleFactory.get_moduleN)rrrsrtrr$rVrVrVrWr sr c@seZdZedddZdS)_MetricModuleFactoryrYcCstdSrr!r#rVrVrWr$sz_MetricModuleFactory.get_moduleN)rrrsrtrr$rVrVrVrWr%sr%c @sreZdZdZedd eeeeefee eee efeedddZ eeeddd Z e d d d ZdS)GithubMetricModuleFactoryuGet the module of a metric. The metric script is downloaded from GitHub. Use the new library 🤗 Evaluate instead: https://huggingface.co/docs/evaluate OUse the new library 🤗 Evaluate instead: https://huggingface.co/docs/evaluateNrIrevisionrrrUcCsd||_||_|r|nt|_|jjdkr4d|j_||_||_|jddksTt t |dddS)Nrrmetricr) rIr)rzr r max_retriesrrUrAssertionErrorrrnrIr)rrrUrVrVrW__init__s  z"GithubMetricModuleFactory.__init__)r)rZcCs>t|j|jd|dd}|j}|jdkr2d|_t||dS)NrFrPrIr)reDownloading builder scriptr)r;rIrrzrr9)rnr)rrrVrVrWdownload_loading_scripts   z1GithubMetricModuleFactory.download_loading_scriptrYc Cs|j}z||}|j}WnBtk r\|dk r6n"d}||}td|jdYnXt|}t|jt|jd|dd||j d}|j r|j nt }t ||g|d|j|j d \}}tt||S) Nmainz-Couldn't find a directory or a metric named 'zH' in this version. It was picked from the main branch on github instead.rFr1rIrrrmetricsrrrrUrrIr)r)r3rrrrIrFrr;rrUrXrrr[invalidate_cachesr)rnr)rrrrUrdrrVrVrWr$s>     z$GithubMetricModuleFactory.get_module)NNNN)rrrsrtrur6rr rrGr r!r0r3rr$rVrVrVrWr&s r&c@sPeZdZdZedd eeeeee efeedddZ e ddd Z dS) LocalMetricModuleFactoryuGet the module of a local metric. The metric script is loaded from a local script. Use the new library 🤗 Evaluate instead: https://huggingface.co/docs/evaluate r'NrPrrrUcCs.||_t|j|_|pt|_||_||_dSrrPrstemrIr rrrUrnrPrrrUrVrVrWr0s   z!LocalMetricModuleFactory.__init__rYc Csnt|j}t|jtt|jj||jd}|jr6|jnt }t |j|g|d|j|j d\}}t t||S)Nr5r6r7)rFrPrrIrrrrrUrXrrr[r8r)rnrrrUrdrrVrVrWr$s&  z#LocalMetricModuleFactory.get_module)NNN) rrrsrtrur6rr r rr!r0rr$rVrVrVrWr9s  r9c@sHeZdZdZd eeeeeeefeedddZ e dddZ dS) #LocalDatasetModuleFactoryWithScriptzTGet the module of a local dataset. The dataset script is loaded from a local script.Nr:cCs.||_t|j|_|pt|_||_||_dSrr;r=rVrVrWr0 s   z,LocalDatasetModuleFactoryWithScript.__init__rYc Cst|jjtj}t|jjd}t|j}t|jtt|jj||j d}g}| rl| tjt|f| r| d|f|j r|j nt }t|j|||d|j|jd\}}t|tt|jjd} t||| S)N README.mdr5datasetsr7)rr)rrPrrDATASETDICT_INFOS_FILENAMErFrrIrris_filerrUrXrrr[r8r) rndataset_infos_pathdataset_readme_pathrrrrUrdrrrVrVrWr$s6  z.LocalDatasetModuleFactoryWithScript.get_module)NNN) rrrsrtrurr r rr!r0rr$rVrVrVrWr> s r>c@sReZdZdZd eeeeeeeefeee efdddZ e dddZ dS) &LocalDatasetModuleFactoryWithoutScriptzGet the module of a dataset loaded from the user's data files. The dataset builder module to use is inferred from the data files extensions.N)rPr rrcCsN|rtj|rtd|t||_t|j|_||_||_ ||_ dS)Nz;`data_dir` must be relative to a dataset directory's root: ) rOrPisabsrras_posixr<rIrr r)rnrPr rrrVrVrWr0>s z/LocalDatasetModuleFactoryWithoutScript.__init__rYc stj|jd}tj|r(t|jnt}t |}t |}t |j|j pPd }|jdk rvt|j}n:|rdtt|krttt|d}nt|}tj||td}t||jd\}} |t|}|tk} |jdkrd| rd|tkrdz t|} Wntk r*d} YnX| dk rdt j| |drdtfdd|!D}t"|\} } |rt#| ||| | d \}}nd \}}| |jt$t |jj%d }|jdk s|s||d<|&| tjtj|jt'j(rht)tj|jt'j(d d J}t ddt*|!D}t+|dkrPtt|}|,||d<W5QRX|&||}|dkrt+|dkrtt|}t-| | ||t.|||ddS)Nr?rr)rr )rrPr csi|]\}}||qSrVrVrZmetadata_data_files_listrVrWrqszELocalDatasetModuleFactoryWithoutScript.get_module..)rrrNN)rrrkrrcSsi|]\}}|t|qSrVr* from_dictrrZdataset_info_dictrVrVrWrsrrrprvrjrr)/rOrPrQrrloaddatarrEfrom_dataset_card_datar+rr  expanduserresolverGrrrrrrrrrrfilter_extensionsr2r1rrrrr^r3rr.rIrrrArTrrpoprr)rnZ readme_pathdataset_card_datarprrpatternsrrrrmetadata_patternsrdrrvrjrrlegacy_dataset_infoslegacy_config_namerVrHrWr$Ns                z1LocalDatasetModuleFactoryWithoutScript.get_module)NNN) rrrsrtrurr rrrr!r0rr$rVrVrVrWrE:s rEc @sXeZdZdZd eeeeeeeefee eee efdddZ e dddZ dS) PackagedDatasetModuleFactoryz`Get the dataset builder module from the ones that are packaged with the library: csv, json, etc.N)rIr rrrcCs.||_||_||_||_||_t|dddSNrer,)rIrr rrr)rnrIr rrrrVrVrWr0s z%PackagedDatasetModuleFactory.__init__rYc st|jp d}|jdk r.t|jnt|}tj ||j |d}|j t k}|jdkr|r|t krz t|}Wntk rd}YnX|dk rtj ||j |drtfdd|D}t|j \}}|||j d}t|||S)Nrrrcsi|]\}}||qSrVrVrrHrVrWrsz;PackagedDatasetModuleFactory.get_module..)rrrk)rr rRrSrGrrrrrrrIr1rrrrr^r3r) rnrrWrrrXrdrrrVrHrWr$s>    z'PackagedDatasetModuleFactory.get_module)NNNN)rrrsrtrurr rrrr r!r0rr$rVrVrVrWr[s r[c @sfeZdZdZd eeeeefeeeeeee fee eee efdddZ e dddZdS) $HubDatasetModuleFactoryWithoutScriptz Get the module of a dataset loaded from data files of a dataset repository. The dataset builder module to use is inferred from the data files extensions. N)rIr)r rrrcCs:||_||_||_||_|p t|_||_t|dddSr\)rIr)rr r rrr)rnrIr)r rrrrVrVrWr0s  z-HubDatasetModuleFactoryWithoutScript.__init__rYc s\ttjj|j|j|jjdd}|j}d|jd|d|j p>d d}|j }|j dkrdd|_ z,t t|jd|d |d }tt|j}Wntk rt}YnXt|}t|}|jdk rt|j} nD|r d tt|kr ttt|d } nt||jd } tj| |t |jd } t!| |j|jd \} } | "t#| } | t$k} |jdkr| r| t%krz t&|}Wntk rd}YnX|dk rt'j||j|drtfdd| (D} t)| \}}|rt*|||| | |jd\}}nd\}}|t|jd|jd |jt+t|jjd}|jdk sL|s^| |d <|,| |j }|j dkrzd|_ zt t|jtj-|jd |d }t.|ddJ}tddt/|(D}t0|dkrtt|}|1||d<W5QRX|,||}Wntk rYnX|dkr@t0|dkr@tt|}t2||||t3|||ddS)NY@)r)tokentimeoutzhf://datasets/@rrDownloading readmer?r)rrr )rrPrr]csi|]\}}||qSrVrVrrHrVrWr1szCHubDatasetModuleFactoryWithoutScript.get_module..)rrrrrI)rrrepo_idrkDownloading metadatarrcSsi|]\}}|t|qSrVrJrLrVrVrWrWsrrrMrN)4rr HF_ENDPOINT dataset_inforIr)rr`shar rstriprzrr9rArrOrrPrrrErQr+rrrrrrrrrrrTr2r1rrrr^r3rr.rrArTrrrUrr)rnZhfh_dataset_infor)rrrDrVrprrWrrrrrXrdrrvrjrrCrrYrZrVrHrWr$s $                     z/HubDatasetModuleFactoryWithoutScript.get_module)NNNNN)rrrsrtrurr rrGrrr r!r0rr$rVrVrVrWr^s r^c@seZdZdZdeeeeefeeeee efeedddZ edddZ edd d Z edd d Z edd dZdS)!HubDatasetModuleFactoryWithScriptz Get the module of a dataset from a dataset repository. The dataset script comes from the script inside the dataset repository. Nr(cCs4||_||_|pt|_||_||_t|dddSr\)rIr)r rrrUrr/rVrVrWr0zs  z*HubDatasetModuleFactoryWithScript.__init__rYcCsHt|j|jddd|jd}|j}|jdkrHubDatasetModuleFactoryWithScript.download_dataset_readme_filec Cs|}|}|}t|}t|jt|jd|jd||jd}g}|rZ| t j |f|rl| d|f|j rx|j nt }t||||d|j|jd\}} t| t|jd|jd|jd} t|| | S) Nrrlr5r?r@r7rd)rrre)r3rnrorFrrIrAr)rrrrArUrXrrr[r8r) rnrrCrDrrrrUrdrrrVrVrWr$s> z,HubDatasetModuleFactoryWithScript.get_module)NNNN)rrrsrtrurr rrGr r!r0r3rnrorr$rVrVrVrWrkts  rkc@s4eZdZdZd eeedddZedddZdS) CachedDatasetModuleFactoryz Get the module of a dataset that has been loaded once already and cached. The script that is loaded from the cache is the most recent one with a matching name. NrIrUcCs$||_||_|jddks tdS)NrrrIrUrr.rnrIrUrVrVrWr0sz#CachedDatasetModuleFactory.__init__rYc s jr jnt}tj|djddtjrLddtDnd}|sjt djd|fdd }t ||d d }d tj|d t ||djd}t js|d7}t|dtj|djdd|jdd g}t|jd}t|||S)Nr@rrcSsg|]}t|dkr|qS@rrhrVrVrWrs z9CachedDatasetModuleFactory.get_module..zDataset  is not cached in cs&t|jdddjS)Nrrr)rrIrstatst_mtimeZ module_hashrrnrVrW_get_modification_timeszECachedDatasetModuleFactory.get_module.._get_modification_timerr3Using the latest cached version of the module from  (last modified on () since it couldn't be found locally at r&, or remotely on the Hugging Face Hub.)rre)rUrXrOrPrQrIrrlistdirrrtimectimerrrrrrr[r8r)rnrUhashesr~rZ warning_msgrdrrVr}rWr$s8 ,    z%CachedDatasetModuleFactory.get_module)N) rrrsrtrurr r0rr$rVrVrVrWrps rpc@s<eZdZdZedd eeedddZeddd Z dS) CachedMetricModuleFactoryu6 Get the module of a metric that has been loaded once already and cached. The script that is loaded from the cache is the most recent one with a matching name. Use the new library 🤗 Evaluate instead: https://huggingface.co/docs/evaluate r'NrqcCs$||_||_|jddks tdS)NrrrrrsrVrVrWr0sz"CachedMetricModuleFactory.__init__rYc sjr jnt}tj|djtjrDddtDnd}|sbtdjd|fdd}t ||dd }t d tj|d t ||d jd dtj|dj|jg}tt||S)Nr6cSsg|]}t|dkr|qSrtrvrwrVrVrWr%s z8CachedMetricModuleFactory.get_module..zMetric rycst|jdjS)Nr)rrIrzr{r|r}rVrWr~-szDCachedMetricModuleFactory.get_module.._get_modification_timerrrrrrr)rUrXrOrPrQrIrrrrrrrrrr[r8r)rnrUrr~rrdrVr}rWr$!s  , z$CachedMetricModuleFactory.get_module)N) rrrsrtrur6rr r0rr$rVrVrVrWr s  r)rPr)rrrUr rrZcKsn|dkrtf|}t|ptj}d|_d|_|tjk|_ttdd| t j d dd}| dsp|d}t j||} |tkrt|||||dS| |rt j|rt|||d Std t|nt j| rt| ||d St j|rt||||d St|rV|dd krVzRtttj} z| j|||j d d} Wnt!k r>} zt"| t#t$j%j&t$j%j'frt'd|dt(| j)dndt*| krd|d} t|r| d|dn| nFdt*| kr*d|d} |r| d|dn| } t| dn| W5d} ~ XYnX|dd| j+Dkrnt,|||||dWSt-||||||dWSWnt!k rR}zzt.||dWWYSt!k r@t"|t#rt'd|d|dt"|t/r|dt"|tr6td t| d |d!t(|j)d"|d|dYnXW5d}~XYnXntd t| d#dS)$a" Download/extract/cache a dataset module. Dataset codes are cached inside the dynamic modules cache to allow easy import (avoid ugly sys.path tweaks). Args: path (str): Path or name of the dataset. Depending on ``path``, the dataset builder that is used comes from a generic dataset script (JSON, CSV, Parquet, text etc.) or from the dataset script (a python file) inside the dataset directory. For local datasets: - if ``path`` is a local directory (containing data files only) -> load a generic dataset builder (csv, json, text etc.) based on the content of the directory e.g. ``'./path/to/directory/with/my/csv/data'``. - if ``path`` is a local dataset script or a directory containing a local dataset script (if the script has the same name as the directory): -> load the dataset builder from the dataset script e.g. ``'./dataset/squad'`` or ``'./dataset/squad/squad.py'``. For datasets on the Hugging Face Hub (list all available datasets with ``huggingface_hub.list_datasets()``) - if ``path`` is a dataset repository on the HF hub (containing data files only) -> load a generic dataset builder (csv, text etc.) based on the content of the repository e.g. ``'username/dataset_name'``, a dataset repository on the HF hub containing your data files. - if ``path`` is a dataset repository on the HF hub with a dataset script (if the script has the same name as the directory) -> load the dataset builder from the dataset script in the dataset repository e.g. ``glue``, ``squad``, ``'username/dataset_name'``, a dataset repository on the HF hub containing a dataset script `'dataset_name.py'`. revision (:class:`~utils.Version` or :obj:`str`, optional): Version of the dataset script to load. As datasets have their own git repository on the Datasets Hub, the default version "main" corresponds to their "main" branch. You can specify a different version than the default "main" by using a commit SHA or a git tag of the dataset repository. download_config (:class:`DownloadConfig`, optional): Specific download configuration parameters. download_mode (:class:`DownloadMode` or :obj:`str`, default ``REUSE_DATASET_IF_EXISTS``): Download/generate mode. dynamic_modules_path (Optional str, defaults to HF_MODULES_CACHE / "datasets_modules", i.e. ~/.cache/huggingface/modules/datasets_modules): Optional path to the directory in which the dynamic modules are saved. It must have been initialized with :obj:`init_dynamic_modules`. By default, the datasets and metrics are stored inside the `datasets_modules` module. data_dir (:obj:`str`, optional): Directory with the data files. Used only if `data_files` is not specified, in which case it's equal to pass `os.path.join(data_dir, "**")` as `data_files`. data_files (:obj:`Union[Dict, List, str]`, optional): Defining the data_files of the dataset configuration. **download_kwargs (additional keyword arguments): optional attributes for DownloadConfig() which will override the attributes in download_config if supplied. Returns: DatasetModule NTcSs|SrrVxrVrVrWzz(dataset_module_factory..rrr)r rrrrrUz"Couldn't find a dataset script at )r rrrr_)rer)r`razCouldn't reach 'z' on the Hub ()Z404z Dataset 'z' doesn't exist on the Hubz at revision ''Z401zT. If the repo is private or gated, make sure to log in with `huggingface-cli login`.cSsg|] }|jqSrV)Z rfilename)rZsiblingrVrVrWrsz*dataset_module_factory..r)rrrU)r)r rrrrUz1Couldn't reach the Hugging Face Hub for dataset 'z': z8 or any data file in the same directory. Couldn't find 'z"' on the Hugging Face Hub either: rz( or any data file in the same directory.)0r r!REUSE_DATASET_IF_EXISTSextract_compressed_file force_extractrZforce_downloadrfilterrrOseprrrPrQr3r[r$rr>rr>rrEr=rr8rrrgrhr`r isinstancer7requests exceptionsConnectTimeoutrmtyperrrZsiblingsrkr^rpr)rPr)rrrUr rdownload_kwargsr combined_pathZhf_apirhrmsge1rVrVrWdataset_module_factory<s7  &                $ rr')rPr)rrrUrZc Ksttjddtd|dkr.tf|}t|p8tj}d|_d|_t t dd| t j ddd }|d s|d }t j||}||rt j|rt|||d W5QRStd t|nt j|rt|||d W5QRSt|r|dd krz$t|||||dWW5QRStk r}zpz&t||dWWYVW5QRStk rt|ts|dtd t|d|ddYnXW5d}~XYnXntd t|dW5QRXdS)u Download/extract/cache a metric module. Use the new library 🤗 Evaluate instead: https://huggingface.co/docs/evaluate Metrics codes are cached inside the dynamic modules cache to allow easy import (avoid ugly sys.path tweaks). Args: path (str): Path or name of the metric script. - if ``path`` is a local metric script or a directory containing a local metric script (if the script has the same name as the directory): -> load the module from the metric script e.g. ``'./metrics/accuracy'`` or ``'./metrics/accuracy/accuracy.py'``. - if ``path`` is a metric on the Hugging Face Hub (ex: `glue`, `squad`) -> load the module from the metric script in the GitHub repository at huggingface/datasets e.g. ``'accuracy'`` or ``'rouge'``. revision (Optional ``Union[str, datasets.Version]``): If specified, the module will be loaded from the datasets repository at this version. By default: - it is set to the local version of the lib. - it will also try to load it from the main branch if it's not available at the local version of the lib. Specifying a version that is different from your local version of the lib might cause compatibility issues. download_config (:class:`DownloadConfig`, optional): Specific download configuration parameters. download_mode (:class:`DownloadMode` or :obj:`str`, default ``REUSE_DATASET_IF_EXISTS``): Download/generate mode. dynamic_modules_path (Optional str, defaults to HF_MODULES_CACHE / "datasets_modules", i.e. ~/.cache/huggingface/modules/datasets_modules): Optional path to the directory in which the dynamic modules are saved. It must have been initialized with :obj:`init_dynamic_modules`. By default, the datasets and metrics are stored inside the `datasets_modules` module. **download_kwargs (additional keyword arguments): optional attributes for DownloadConfig() which will override the attributes in download_config if supplied. Returns: MetricModule ignore'.*https://huggingface.co/docs/evaluate$messagecategoryNTcSs|SrrVrrVrVrWr)rz'metric_module_factory..rrrrz!Couldn't find a metric script at rrrz . Metric 'z/' doesn't exist on the Hugging Face Hub either.r)warningscatch_warningsfilterwarnings FutureWarningr r!rrrrrrrOrrrrPrQrr9r$rr>r=rr&rrr) rPr)rrrUrrrrrVrVrWmetric_module_factorys`0  &   & rueUse 'evaluate.load' instead, from the new library 🤗 Evaluate: https://huggingface.co/docs/evaluateF) rPr process_id num_process cache_dir experiment_idkeep_in_memoryrrr)rZc  Kstvtjddtdt|p$tj}t|| ||dj} t| dd} | f||||||d| } | j |d| W5QRSQRXd S) u( Load a `datasets.Metric`. Use `evaluate.load` instead, from the new library 🤗 Evaluate: https://huggingface.co/docs/evaluate Args: path (``str``): path to the metric processing script with the metric builder. Can be either: - a local path to processing script or the directory containing the script (if the script has the same name as the directory), e.g. ``'./metrics/rouge'`` or ``'./metrics/rogue/rouge.py'`` - a metric identifier on the HuggingFace datasets repo (list all available metrics with ``datasets.list_metrics()``) e.g. ``'rouge'`` or ``'bleu'`` config_name (:obj:`str`, optional): selecting a configuration for the metric (e.g. the GLUE metric has a configuration for each subset) process_id (:obj:`int`, optional): for distributed evaluation: id of the process num_process (:obj:`int`, optional): for distributed evaluation: total number of processes cache_dir (Optional str): path to store the temporary predictions and references (default to `~/.cache/huggingface/metrics/`) experiment_id (``str``): A specific experiment id. This is used if several distributed evaluations share the same file system. This is useful to compute metrics in distributed setups (in particular non-additive metrics like F1). keep_in_memory (bool): Whether to store the temporary results in memory (defaults to False) download_config (Optional ``datasets.DownloadConfig``: specific download configuration parameters. download_mode (:class:`DownloadMode` or :obj:`str`, default ``REUSE_DATASET_IF_EXISTS``): Download/generate mode. revision (Optional ``Union[str, datasets.Version]``): if specified, the module will be loaded from the datasets repository at this version. By default, it is set to the local version of the lib. Specifying a version that is different from your local version of the lib might cause compatibility issues. Returns: `datasets.Metric` Example: ```py >>> from datasets import load_metric >>> accuracy = load_metric('accuracy') >>> accuracy.compute(references=[1, 0], predictions=[1, 1]) {'accuracy': 0.5} ``` rrr)r)rrF)re)rrrrrrrN) rrrrr!rrrdrhdownload_and_prepare)rPrrrrrrrrr)Zmetric_init_kwargsZ metric_moduleZ metric_clsr+rVrVrW load_metricRs,7   rr6) rPrIr rrfeaturesrrr)r`storage_optionsrZc  s| dkr td| dt| } t|p*tj}| dk rN|rB|nt}| |_| dk rt|rb|nt}|j | t |||||d} | j }| d|}| d|}| d|p| j j}| d d}| d }| jr| j|nd}| j jr|| j jkrt|| j j|}tkrd|dkrdd d }fd dtD}|r\|d|dd7}t|t| |d}|f||||||||| | d || }|S)aBLoad a dataset builder from the Hugging Face Hub, or a local dataset. A dataset builder can be used to inspect general information that is required to build a dataset (cache directory, config, dataset info, etc.) without downloading the dataset itself. You can find the list of datasets on the [Hub](https://huggingface.co/datasets) or with [`huggingface_hub.list_datasets`]. A dataset is a directory that contains: - some data files in generic formats (JSON, CSV, Parquet, text, etc.) - and optionally a dataset script, if it requires some code to read the data files. This is used to load any kind of formats or structures. Note that dataset scripts can also download and read data files from anywhere - in case your data files already exist online. Args: path (`str`): Path or name of the dataset. Depending on `path`, the dataset builder that is used comes from a generic dataset script (JSON, CSV, Parquet, text etc.) or from the dataset script (a python file) inside the dataset directory. For local datasets: - if `path` is a local directory (containing data files only) -> load a generic dataset builder (csv, json, text etc.) based on the content of the directory e.g. `'./path/to/directory/with/my/csv/data'`. - if `path` is a local dataset script or a directory containing a local dataset script (if the script has the same name as the directory) -> load the dataset builder from the dataset script e.g. `'./dataset/squad'` or `'./dataset/squad/squad.py'`. For datasets on the Hugging Face Hub (list all available datasets with [`huggingface_hub.list_datasets`]) - if `path` is a dataset repository on the HF hub (containing data files only) -> load a generic dataset builder (csv, text etc.) based on the content of the repository e.g. `'username/dataset_name'`, a dataset repository on the HF hub containing your data files. - if `path` is a dataset repository on the HF hub with a dataset script (if the script has the same name as the directory) -> load the dataset builder from the dataset script in the dataset repository e.g. `glue`, `squad`, `'username/dataset_name'`, a dataset repository on the HF hub containing a dataset script `'dataset_name.py'`. name (`str`, *optional*): Defining the name of the dataset configuration. data_dir (`str`, *optional*): Defining the `data_dir` of the dataset configuration. If specified for the generic builders (csv, text etc.) or the Hub datasets and `data_files` is `None`, the behavior is equal to passing `os.path.join(data_dir, **)` as `data_files` to reference all the files in a directory. data_files (`str` or `Sequence` or `Mapping`, *optional*): Path(s) to source data file(s). cache_dir (`str`, *optional*): Directory to read/write data. Defaults to `"~/.cache/huggingface/datasets"`. features ([`Features`], *optional*): Set the features type to use for this dataset. download_config ([`DownloadConfig`], *optional*): Specific download configuration parameters. download_mode ([`DownloadMode`] or `str`, defaults to `REUSE_DATASET_IF_EXISTS`): Download/generate mode. revision ([`Version`] or `str`, *optional*): Version of the dataset script to load. As datasets have their own git repository on the Datasets Hub, the default version "main" corresponds to their "main" branch. You can specify a different version than the default "main" by using a commit SHA or a git tag of the dataset repository. token (`str` or `bool`, *optional*): Optional string or boolean to use as Bearer token for remote files on the Datasets Hub. If `True`, or not specified, will get token from `"~/.huggingface"`. use_auth_token (`str` or `bool`, *optional*): Optional string or boolean to use as Bearer token for remote files on the Datasets Hub. If `True`, or not specified, will get token from `"~/.huggingface"`. `use_auth_token` was deprecated in favor of `token` in version 2.14.0 and will be removed in 3.0.0. storage_options (`dict`, *optional*, defaults to `None`): **Experimental**. Key/value pairs to be passed on to the dataset file-system backend, if any. **config_kwargs (additional keyword arguments): Keyword arguments to be passed to the [`BuilderConfig`] and used in the [`DatasetBuilder`]. Returns: [`DatasetBuilder`] Example: ```py >>> from datasets import load_dataset_builder >>> ds_builder = load_dataset_builder('rotten_tomatoes') >>> ds_builder.info.features {'label': ClassLabel(num_classes=2, names=['neg', 'pos'], id=None), 'text': Value(dtype='string', id=None)} ``` r6'use_auth_token' was deprecated in favor of 'token' in version 2.14.0 and will be removed in 3.0.0. You can remove this warning by passing 'token= ' instead.N)r)rrr rr rrrkrz@Please specify the data files or data directory to load for the z dataset builder.csg|]}t|kr|qSrV)r0)r extensionrPrVrWr3s z(load_dataset_builder..z9 For example `data_files={"train": "path/to/data/train/*.rz"}`)rk) rrkrr rrinforr`r)rwarnrr!rrzr r`rrrrrUrrjrrrprr3r0rr)rPrIr rrrrrr)r`use_auth_tokenr config_kwargsrrrrkrr error_msgZexample_extensionsrobuilder_instancerVrrWload_dataset_buildersg              r)rPrIr rrrrrrverification_moder save_infosr)r` streamingnum_procrrZcKs|dkr td|dt|}| dkrN| r2tjntj} td| jdt|dkrdtdtnd}|dk r|std|dt|t j  rtd |r|dk rt d t |pt j}t| s| ptjntj} tf||||||||| ||d |}|r|j|d S|tk}|j||| |||d | dk r:| n t|jj} |j|| | d}|dk rttdt||}W5QRX| r||S)a>)Load a dataset from the Hugging Face Hub, or a local dataset. You can find the list of datasets on the [Hub](https://huggingface.co/datasets) or with [`huggingface_hub.list_datasets`]. A dataset is a directory that contains: - some data files in generic formats (JSON, CSV, Parquet, text, etc.). - and optionally a dataset script, if it requires some code to read the data files. This is used to load any kind of formats or structures. Note that dataset scripts can also download and read data files from anywhere - in case your data files already exist online. This function does the following under the hood: 1. Download and import in the library the dataset script from `path` if it's not already cached inside the library. If the dataset has no dataset script, then a generic dataset script is imported instead (JSON, CSV, Parquet, text, etc.) Dataset scripts are small python scripts that define dataset builders. They define the citation, info and format of the dataset, contain the path or URL to the original data files and the code to load examples from the original data files. You can find the complete list of datasets in the Datasets [Hub](https://huggingface.co/datasets). 2. Run the dataset script which will: * Download the dataset file from the original URL (see the script) if it's not already available locally or cached. * Process and cache the dataset in typed Arrow tables for caching. Arrow table are arbitrarily long, typed tables which can store nested objects and be mapped to numpy/pandas/python generic types. They can be directly accessed from disk, loaded in RAM or even streamed over the web. 3. Return a dataset built from the requested splits in `split` (default: all). It also allows to load a dataset from a local directory or a dataset repository on the Hugging Face Hub without dataset script. In this case, it automatically loads all the data files from the directory or the dataset repository. Args: path (`str`): Path or name of the dataset. Depending on `path`, the dataset builder that is used comes from a generic dataset script (JSON, CSV, Parquet, text etc.) or from the dataset script (a python file) inside the dataset directory. For local datasets: - if `path` is a local directory (containing data files only) -> load a generic dataset builder (csv, json, text etc.) based on the content of the directory e.g. `'./path/to/directory/with/my/csv/data'`. - if `path` is a local dataset script or a directory containing a local dataset script (if the script has the same name as the directory) -> load the dataset builder from the dataset script e.g. `'./dataset/squad'` or `'./dataset/squad/squad.py'`. For datasets on the Hugging Face Hub (list all available datasets with [`huggingface_hub.list_datasets`]) - if `path` is a dataset repository on the HF hub (containing data files only) -> load a generic dataset builder (csv, text etc.) based on the content of the repository e.g. `'username/dataset_name'`, a dataset repository on the HF hub containing your data files. - if `path` is a dataset repository on the HF hub with a dataset script (if the script has the same name as the directory) -> load the dataset builder from the dataset script in the dataset repository e.g. `glue`, `squad`, `'username/dataset_name'`, a dataset repository on the HF hub containing a dataset script `'dataset_name.py'`. name (`str`, *optional*): Defining the name of the dataset configuration. data_dir (`str`, *optional*): Defining the `data_dir` of the dataset configuration. If specified for the generic builders (csv, text etc.) or the Hub datasets and `data_files` is `None`, the behavior is equal to passing `os.path.join(data_dir, **)` as `data_files` to reference all the files in a directory. data_files (`str` or `Sequence` or `Mapping`, *optional*): Path(s) to source data file(s). split (`Split` or `str`): Which split of the data to load. If `None`, will return a `dict` with all splits (typically `datasets.Split.TRAIN` and `datasets.Split.TEST`). If given, will return a single Dataset. Splits can be combined and specified like in tensorflow-datasets. cache_dir (`str`, *optional*): Directory to read/write data. Defaults to `"~/.cache/huggingface/datasets"`. features (`Features`, *optional*): Set the features type to use for this dataset. download_config ([`DownloadConfig`], *optional*): Specific download configuration parameters. download_mode ([`DownloadMode`] or `str`, defaults to `REUSE_DATASET_IF_EXISTS`): Download/generate mode. verification_mode ([`VerificationMode`] or `str`, defaults to `BASIC_CHECKS`): Verification mode determining the checks to run on the downloaded/processed dataset information (checksums/size/splits/...). ignore_verifications (`bool`, defaults to `False`): Ignore the verifications of the downloaded/processed dataset information (checksums/size/splits/...). `ignore_verifications` was deprecated in version 2.9.1 and will be removed in 3.0.0. Please use `verification_mode` instead. keep_in_memory (`bool`, defaults to `None`): Whether to copy the dataset in-memory. If `None`, the dataset will not be copied in-memory unless explicitly enabled by setting `datasets.config.IN_MEMORY_MAX_SIZE` to nonzero. See more details in the [improve performance](../cache#improve-performance) section. save_infos (`bool`, defaults to `False`): Save the dataset information (checksums/size/splits/...). revision ([`Version`] or `str`, *optional*): Version of the dataset script to load. As datasets have their own git repository on the Datasets Hub, the default version "main" corresponds to their "main" branch. You can specify a different version than the default "main" by using a commit SHA or a git tag of the dataset repository. token (`str` or `bool`, *optional*): Optional string or boolean to use as Bearer token for remote files on the Datasets Hub. If `True`, or not specified, will get token from `"~/.huggingface"`. use_auth_token (`str` or `bool`, *optional*): Optional string or boolean to use as Bearer token for remote files on the Datasets Hub. If `True`, or not specified, will get token from `"~/.huggingface"`. `use_auth_token` was deprecated in favor of `token` in version 2.14.0 and will be removed in 3.0.0. task (`str`): The task to prepare the dataset for during training and evaluation. Casts the dataset's [`Features`] to standardized column names and types as detailed in `datasets.tasks`. `task` was deprecated in version 2.13.0 and will be removed in 3.0.0. streaming (`bool`, defaults to `False`): If set to `True`, don't download the data files. Instead, it streams the data progressively while iterating on the dataset. An [`IterableDataset`] or [`IterableDatasetDict`] is returned instead in this case. Note that streaming works for datasets that use data formats that support being iterated over like txt, csv, jsonl for example. Json files may be downloaded completely. Also streaming from remote zip or gzip files is supported but other compressed formats like rar and xz are not yet supported. The tgz format doesn't allow streaming. num_proc (`int`, *optional*, defaults to `None`): Number of processes when downloading and generating the dataset locally. Multiprocessing is disabled by default. storage_options (`dict`, *optional*, defaults to `None`): **Experimental**. Key/value pairs to be passed on to the dataset file-system backend, if any. **config_kwargs (additional keyword arguments): Keyword arguments to be passed to the `BuilderConfig` and used in the [`DatasetBuilder`]. Returns: [`Dataset`] or [`DatasetDict`]: - if `split` is not `None`: the dataset requested, - if `split` is `None`, a [`~datasets.DatasetDict`] with each split. or [`IterableDataset`] or [`IterableDatasetDict`]: if `streaming=True` - if `split` is not `None`, the dataset is requested - if `split` is `None`, a [`~datasets.streaming.IterableDatasetDict`] with each split. Example: Load a dataset from the Hugging Face Hub: ```py >>> from datasets import load_dataset >>> ds = load_dataset('rotten_tomatoes', split='train') # Map data files to splits >>> data_files = {'train': 'train.csv', 'test': 'test.csv'} >>> ds = load_dataset('namespace/your_dataset_name', data_files=data_files) ``` Load a local dataset: ```py # Load a CSV file >>> from datasets import load_dataset >>> ds = load_dataset('csv', data_files='path/to/local/my_dataset.csv') # Load a JSON file >>> from datasets import load_dataset >>> ds = load_dataset('json', data_files='path/to/local/my_dataset.json') # Load from a local loading script >>> from datasets import load_dataset >>> ds = load_dataset('path/to/local/loading_script/loading_script.py', split='train') ``` Load an [`~datasets.IterableDataset`]: ```py >>> from datasets import load_dataset >>> ds = load_dataset('rotten_tomatoes', split='train', streaming=True) ``` Load an image dataset with the `ImageFolder` dataset builder: ```py >>> from datasets import load_dataset >>> ds = load_dataset('imagefolder', data_dir='/path/to/images', split='train') ``` r6rrz'ignore_verifications' was deprecated in favor of 'verification_mode' in version 2.9.1 and will be removed in 3.0.0. You can remove this warning by passing 'verification_mode=zF'task' was deprecated in version 2.13.0 and will be removed in 3.0.0. NzEmpty 'data_files': 'z3'. It should be either non-empty or None (default).zjYou are trying to load a dataset that was saved using `save_to_disk`. Please use `load_from_disk` instead.zLoading a streaming dataset in parallel with `num_proc` is not implemented. To parallelize streaming, you can wrap the dataset with a PyTorch DataLoader using `num_workers` > 1 instead.) rPrIr rrrrrr)r`rr)rrrtry_from_hf_gcsrr)rrZ in_memoryr)rrrrBZ NO_CHECKSZ ALL_CHECKSrrrrDATASET_STATE_JSON_FILENAMErSr"r!rZ BASIC_CHECKSrZas_streaming_datasetr3rrCrZ dataset_sizeZ as_datasetr simplefilterZprepare_for_taskZ _save_infos)rPrIr rrrrrrrZignore_verificationsrrr)r`rtaskrrrrrrZdsrVrVrW load_datasetNs[          r) dataset_pathrrrZcCs|dkrtdt|j}tj||d}|d}t|rHt|}tj }nt d}|}t j j }| |sxtd|d|||tjr|||tjrtj|||dS|||tjrtj|||dStd|d d S) a Loads a dataset that was previously saved using [`~Dataset.save_to_disk`] from a dataset directory, or from a filesystem using any implementation of `fsspec.spec.AbstractFileSystem`. Args: dataset_path (`str`): Path (e.g. `"dataset/train"`) or remote URI (e.g. `"s3://my-bucket/dataset/train"`) of the [`Dataset`] or [`DatasetDict`] directory where the dataset will be loaded from. fs (`~filesystems.S3FileSystem` or `fsspec.spec.AbstractFileSystem`, *optional*): Instance of the remote filesystem used to download the files from. `fs` was deprecated in version 2.9.0 and will be removed in 3.0.0. Please use `storage_options` instead, e.g. `storage_options=fs.storage_options`. keep_in_memory (`bool`, defaults to `None`): Whether to copy the dataset in-memory. If `None`, the dataset will not be copied in-memory unless explicitly enabled by setting `datasets.config.IN_MEMORY_MAX_SIZE` to nonzero. See more details in the [improve performance](../cache#improve-performance) section. storage_options (`dict`, *optional*): Key/value pairs to be passed on to the file-system backend, if any. Returns: [`Dataset`] or [`DatasetDict`]: - If `dataset_path` is a path of a dataset directory: the dataset requested. - If `dataset_path` is a path of a dataset dict directory, a [`DatasetDict`] with each split. Example: ```py >>> from datasets import load_from_disk >>> ds = load_from_disk('path/to/dataset/directory') ``` r6z'fs' was deprecated in favor of 'storage_options' in version 2.9.0 and will be removed in 3.0.0. You can remove this warning by passing 'storage_options=fs.storage_options' instead.)rrfilez Directory z not found)rrz@ is neither a `Dataset` directory nor a `DatasetDict` directory.N)rrrrfsspecZget_fs_token_pathsr(r' posixpathrQ filesystemrOrPrSrrrZDATASET_INFO_FILENAMErrload_from_diskZDATASETDICT_JSON_FILENAMEr)rfsrrZfs_token_pathsZdest_dataset_pathZ path_joinrVrVrWrs2,    r)T)N)N)N)NN)NNN)NNNNNN)NNNN) NrrNNFNNN) NNNNNNNNNr6N)NNNNNNNNNr6NFNNr6r6FNN)r6NN)rurr[r_rrOrrrr collectionsr dataclassesrrpathlibrtypingrrrr r r r r rrrZhuggingface_hubrrrrrZ arrow_datasetrZbuilderrrrrrrrrrrZ dataset_dictrrZdownload.download_configr Zdownload.download_managerr!Z#download.streaming_download_managerr"r#r$r%rr&Z filesystemsr'r( fingerprintr)rr*r+Ziterable_datasetr,r+r-Znamingr.r/Zpackaged_modulesr0r1r2r3r4Zsplitsr5Zutils.deprecation_utilsr6Zutils.file_utilsr7r8r9r:r;r<r=r>r?Zutils.filelockr@Z utils.hubrAZutils.info_utilsrBrCZ utils.loggingrDZutils.metadatarEZutils.py_utilsrFZ utils.versionrGrrrrrrZMODULE_NAME_FOR_DYNAMIC_MODULESrrXrhrirlrrrrrrrrrrrrrrrrr r%r&r9r>rEr[r^rkrprrrrrrrrrVrVrVrWs  ,  $         ,       $ )   B W  & '   KK11o7^93 4bQ( /( 8