MCPcopy
hub / github.com/huggingface/datasets / get_module

Method get_module

src/datasets/load.py:536–556  ·  view source on GitHub ↗
(self)

Source from the content-addressed store, hash-verified

534 increase_load_count(name)
535
536 def get_module(self) -> DatasetModule:
537 base_path = Path(self.data_dir or "").expanduser().resolve().as_posix()
538 patterns = (
539 sanitize_patterns(self.data_files)
540 if self.data_files is not None
541 else get_data_patterns(base_path, download_config=self.download_config)
542 )
543 data_files = DataFilesDict.from_patterns(
544 patterns,
545 download_config=self.download_config,
546 base_path=base_path,
547 )
548
549 module_path, hash = _PACKAGED_DATASETS_MODULES[self.name]
550
551 builder_kwargs = {
552 "data_files": data_files,
553 "dataset_name": self.name,
554 }
555
556 return DatasetModule(module_path, hash, builder_kwargs)
557
558
559class HubDatasetModuleFactory(_DatasetModuleFactory):

Calls 5

sanitize_patternsFunction · 0.85
get_data_patternsFunction · 0.85
DatasetModuleClass · 0.85
resolveMethod · 0.45
from_patternsMethod · 0.45