Transformers

一般的なユーティリティ

このページには、ファイル utils.py にある Transformers の一般的なユーティリティ関数がすべてリストされています。

これらのほとんどは、ライブラリで一般的なコードを学習する場合にのみ役に立ちます。

列挙型と名前付きタプル

class transformers.utils.ExplicitEnum

< source >

( valuenames = Nonemodule = Nonequalname = Nonetype = Nonestart = 1 )

Enum with more explicit error message for missing values.

class transformers.utils.PaddingStrategy

< source >

( valuenames = Nonemodule = Nonequalname = Nonetype = Nonestart = 1 )

Possible values for the padding argument in PreTrainedTokenizerBase.call(). Useful for tab-completion in an IDE.

class transformers.TensorType

< source >

( valuenames = Nonemodule = Nonequalname = Nonetype = Nonestart = 1 )

Possible values for the return_tensors argument in PreTrainedTokenizerBase.call(). Useful for tab-completion in an IDE.

特別なデコレーター

transformers.add_start_docstrings

< source >

( *docstr )

transformers.utils.add_start_docstrings_to_model_forward

< source >

( *docstr )

transformers.add_end_docstrings

< source >

( *docstr )

transformers.utils.add_code_sample_docstrings

< source >

( *docstrprocessor_class = Nonecheckpoint = Noneoutput_type = Noneconfig_class = Nonemask = '[MASK]'qa_target_start_index = 14qa_target_end_index = 15model_cls = Nonemodality = Noneexpected_output = Noneexpected_loss = Nonereal_checkpoint = Nonerevision = None )

transformers.utils.replace_return_docstrings

< source >

( output_type = Noneconfig_class = None )

特殊なプロパティ

class transformers.utils.cached_property

< source >

( fget = Nonefset = Nonefdel = Nonedoc = None )

Descriptor that mimics @property but caches output in member variable.

From tensorflow_datasets

Built-in in functools from Python 3.8.

その他のユーティリティ

class transformers.utils._LazyModule

< source >

( namemodule_fileimport_structuremodule_spec = Noneextra_objects = None )

Module class that surfaces all objects but only performs associated imports when the objects are requested.