transformers 4.41.0__tar.gz → 4.41.2__tar.gz

{transformers-4.41.0/src/transformers.egg-info → transformers-4.41.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: transformers
-Version: 4.41.0
+Version: 4.41.2
 Summary: State-of-the-art Machine Learning for JAX, PyTorch and TensorFlow
 Home-page: https://github.com/huggingface/transformers
 Author: The Hugging Face team (past and future) with the help of all our contributors (https://github.com/huggingface/transformers/graphs/contributors)
@@ -20,48 +20,425 @@ Classifier: Programming Language :: Python :: 3.10
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Requires-Python: >=3.8.0
 Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: filelock
+Requires-Dist: huggingface-hub<1.0,>=0.23.0
+Requires-Dist: numpy>=1.17
+Requires-Dist: packaging>=20.0
+Requires-Dist: pyyaml>=5.1
+Requires-Dist: regex!=2019.12.17
+Requires-Dist: requests
+Requires-Dist: tokenizers<0.20,>=0.19
+Requires-Dist: safetensors>=0.4.1
+Requires-Dist: tqdm>=4.27
 Provides-Extra: ja
+Requires-Dist: fugashi>=1.0; extra == "ja"
+Requires-Dist: ipadic<2.0,>=1.0.0; extra == "ja"
+Requires-Dist: unidic_lite>=1.0.7; extra == "ja"
+Requires-Dist: unidic>=1.0.2; extra == "ja"
+Requires-Dist: sudachipy>=0.6.6; extra == "ja"
+Requires-Dist: sudachidict_core>=20220729; extra == "ja"
+Requires-Dist: rhoknp<1.3.1,>=1.1.0; extra == "ja"
 Provides-Extra: sklearn
+Requires-Dist: scikit-learn; extra == "sklearn"
 Provides-Extra: tf
+Requires-Dist: tensorflow<2.16,>2.9; extra == "tf"
+Requires-Dist: onnxconverter-common; extra == "tf"
+Requires-Dist: tf2onnx; extra == "tf"
+Requires-Dist: tensorflow-text<2.16; extra == "tf"
+Requires-Dist: keras-nlp>=0.3.1; extra == "tf"
 Provides-Extra: tf-cpu
+Requires-Dist: keras<2.16,>2.9; extra == "tf-cpu"
+Requires-Dist: tensorflow-cpu<2.16,>2.9; extra == "tf-cpu"
+Requires-Dist: onnxconverter-common; extra == "tf-cpu"
+Requires-Dist: tf2onnx; extra == "tf-cpu"
+Requires-Dist: tensorflow-text<2.16; extra == "tf-cpu"
+Requires-Dist: keras-nlp>=0.3.1; extra == "tf-cpu"
+Requires-Dist: tensorflow-probability<2.16; extra == "tf-cpu"
 Provides-Extra: torch
+Requires-Dist: torch; extra == "torch"
+Requires-Dist: accelerate>=0.21.0; extra == "torch"
 Provides-Extra: accelerate
+Requires-Dist: accelerate>=0.21.0; extra == "accelerate"
 Provides-Extra: retrieval
+Requires-Dist: faiss-cpu; extra == "retrieval"
+Requires-Dist: datasets!=2.5.0; extra == "retrieval"
 Provides-Extra: flax
+Requires-Dist: jax<=0.4.13,>=0.4.1; extra == "flax"
+Requires-Dist: jaxlib<=0.4.13,>=0.4.1; extra == "flax"
+Requires-Dist: flax<=0.7.0,>=0.4.1; extra == "flax"
+Requires-Dist: optax<=0.1.4,>=0.0.8; extra == "flax"
+Requires-Dist: scipy<1.13.0; extra == "flax"
 Provides-Extra: tokenizers
+Requires-Dist: tokenizers<0.20,>=0.19; extra == "tokenizers"
 Provides-Extra: ftfy
+Requires-Dist: ftfy; extra == "ftfy"
 Provides-Extra: onnxruntime
+Requires-Dist: onnxruntime>=1.4.0; extra == "onnxruntime"
+Requires-Dist: onnxruntime-tools>=1.4.2; extra == "onnxruntime"
 Provides-Extra: onnx
+Requires-Dist: onnxconverter-common; extra == "onnx"
+Requires-Dist: tf2onnx; extra == "onnx"
+Requires-Dist: onnxruntime>=1.4.0; extra == "onnx"
+Requires-Dist: onnxruntime-tools>=1.4.2; extra == "onnx"
 Provides-Extra: modelcreation
+Requires-Dist: cookiecutter==1.7.3; extra == "modelcreation"
 Provides-Extra: sagemaker
+Requires-Dist: sagemaker>=2.31.0; extra == "sagemaker"
 Provides-Extra: deepspeed
+Requires-Dist: deepspeed>=0.9.3; extra == "deepspeed"
+Requires-Dist: accelerate>=0.21.0; extra == "deepspeed"
 Provides-Extra: optuna
+Requires-Dist: optuna; extra == "optuna"
 Provides-Extra: ray
+Requires-Dist: ray[tune]>=2.7.0; extra == "ray"
 Provides-Extra: sigopt
+Requires-Dist: sigopt; extra == "sigopt"
 Provides-Extra: integrations
+Requires-Dist: optuna; extra == "integrations"
+Requires-Dist: ray[tune]>=2.7.0; extra == "integrations"
+Requires-Dist: sigopt; extra == "integrations"
 Provides-Extra: serving
+Requires-Dist: pydantic; extra == "serving"
+Requires-Dist: uvicorn; extra == "serving"
+Requires-Dist: fastapi; extra == "serving"
+Requires-Dist: starlette; extra == "serving"
 Provides-Extra: audio
+Requires-Dist: librosa; extra == "audio"
+Requires-Dist: pyctcdecode>=0.4.0; extra == "audio"
+Requires-Dist: phonemizer; extra == "audio"
+Requires-Dist: kenlm; extra == "audio"
 Provides-Extra: speech
+Requires-Dist: torchaudio; extra == "speech"
+Requires-Dist: librosa; extra == "speech"
+Requires-Dist: pyctcdecode>=0.4.0; extra == "speech"
+Requires-Dist: phonemizer; extra == "speech"
+Requires-Dist: kenlm; extra == "speech"
 Provides-Extra: torch-speech
+Requires-Dist: torchaudio; extra == "torch-speech"
+Requires-Dist: librosa; extra == "torch-speech"
+Requires-Dist: pyctcdecode>=0.4.0; extra == "torch-speech"
+Requires-Dist: phonemizer; extra == "torch-speech"
+Requires-Dist: kenlm; extra == "torch-speech"
 Provides-Extra: tf-speech
+Requires-Dist: librosa; extra == "tf-speech"
+Requires-Dist: pyctcdecode>=0.4.0; extra == "tf-speech"
+Requires-Dist: phonemizer; extra == "tf-speech"
+Requires-Dist: kenlm; extra == "tf-speech"
 Provides-Extra: flax-speech
+Requires-Dist: librosa; extra == "flax-speech"
+Requires-Dist: pyctcdecode>=0.4.0; extra == "flax-speech"
+Requires-Dist: phonemizer; extra == "flax-speech"
+Requires-Dist: kenlm; extra == "flax-speech"
 Provides-Extra: vision
+Requires-Dist: Pillow<=15.0,>=10.0.1; extra == "vision"
 Provides-Extra: timm
+Requires-Dist: timm; extra == "timm"
 Provides-Extra: torch-vision
+Requires-Dist: torchvision; extra == "torch-vision"
+Requires-Dist: Pillow<=15.0,>=10.0.1; extra == "torch-vision"
 Provides-Extra: natten
+Requires-Dist: natten<0.15.0,>=0.14.6; extra == "natten"
 Provides-Extra: codecarbon
+Requires-Dist: codecarbon==1.2.0; extra == "codecarbon"
 Provides-Extra: video
+Requires-Dist: decord==0.6.0; extra == "video"
+Requires-Dist: av==9.2.0; extra == "video"
 Provides-Extra: sentencepiece
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "sentencepiece"
+Requires-Dist: protobuf; extra == "sentencepiece"
 Provides-Extra: testing
+Requires-Dist: pytest<8.0.0,>=7.2.0; extra == "testing"
+Requires-Dist: pytest-rich; extra == "testing"
+Requires-Dist: pytest-xdist; extra == "testing"
+Requires-Dist: timeout-decorator; extra == "testing"
+Requires-Dist: parameterized; extra == "testing"
+Requires-Dist: psutil; extra == "testing"
+Requires-Dist: datasets!=2.5.0; extra == "testing"
+Requires-Dist: dill<0.3.5; extra == "testing"
+Requires-Dist: evaluate>=0.2.0; extra == "testing"
+Requires-Dist: pytest-timeout; extra == "testing"
+Requires-Dist: ruff==0.1.5; extra == "testing"
+Requires-Dist: sacrebleu<2.0.0,>=1.4.12; extra == "testing"
+Requires-Dist: rouge-score!=0.0.7,!=0.0.8,!=0.1,!=0.1.1; extra == "testing"
+Requires-Dist: nltk; extra == "testing"
+Requires-Dist: GitPython<3.1.19; extra == "testing"
+Requires-Dist: sacremoses; extra == "testing"
+Requires-Dist: rjieba; extra == "testing"
+Requires-Dist: beautifulsoup4; extra == "testing"
+Requires-Dist: tensorboard; extra == "testing"
+Requires-Dist: pydantic; extra == "testing"
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "testing"
+Requires-Dist: faiss-cpu; extra == "testing"
+Requires-Dist: datasets!=2.5.0; extra == "testing"
+Requires-Dist: cookiecutter==1.7.3; extra == "testing"
 Provides-Extra: deepspeed-testing
+Requires-Dist: deepspeed>=0.9.3; extra == "deepspeed-testing"
+Requires-Dist: accelerate>=0.21.0; extra == "deepspeed-testing"
+Requires-Dist: pytest<8.0.0,>=7.2.0; extra == "deepspeed-testing"
+Requires-Dist: pytest-rich; extra == "deepspeed-testing"
+Requires-Dist: pytest-xdist; extra == "deepspeed-testing"
+Requires-Dist: timeout-decorator; extra == "deepspeed-testing"
+Requires-Dist: parameterized; extra == "deepspeed-testing"
+Requires-Dist: psutil; extra == "deepspeed-testing"
+Requires-Dist: datasets!=2.5.0; extra == "deepspeed-testing"
+Requires-Dist: dill<0.3.5; extra == "deepspeed-testing"
+Requires-Dist: evaluate>=0.2.0; extra == "deepspeed-testing"
+Requires-Dist: pytest-timeout; extra == "deepspeed-testing"
+Requires-Dist: ruff==0.1.5; extra == "deepspeed-testing"
+Requires-Dist: sacrebleu<2.0.0,>=1.4.12; extra == "deepspeed-testing"
+Requires-Dist: rouge-score!=0.0.7,!=0.0.8,!=0.1,!=0.1.1; extra == "deepspeed-testing"
+Requires-Dist: nltk; extra == "deepspeed-testing"
+Requires-Dist: GitPython<3.1.19; extra == "deepspeed-testing"
+Requires-Dist: sacremoses; extra == "deepspeed-testing"
+Requires-Dist: rjieba; extra == "deepspeed-testing"
+Requires-Dist: beautifulsoup4; extra == "deepspeed-testing"
+Requires-Dist: tensorboard; extra == "deepspeed-testing"
+Requires-Dist: pydantic; extra == "deepspeed-testing"
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "deepspeed-testing"
+Requires-Dist: faiss-cpu; extra == "deepspeed-testing"
+Requires-Dist: datasets!=2.5.0; extra == "deepspeed-testing"
+Requires-Dist: cookiecutter==1.7.3; extra == "deepspeed-testing"
+Requires-Dist: optuna; extra == "deepspeed-testing"
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "deepspeed-testing"
+Requires-Dist: protobuf; extra == "deepspeed-testing"
 Provides-Extra: quality
+Requires-Dist: datasets!=2.5.0; extra == "quality"
+Requires-Dist: isort>=5.5.4; extra == "quality"
+Requires-Dist: ruff==0.1.5; extra == "quality"
+Requires-Dist: GitPython<3.1.19; extra == "quality"
+Requires-Dist: urllib3<2.0.0; extra == "quality"
 Provides-Extra: all
+Requires-Dist: tensorflow<2.16,>2.9; extra == "all"
+Requires-Dist: onnxconverter-common; extra == "all"
+Requires-Dist: tf2onnx; extra == "all"
+Requires-Dist: tensorflow-text<2.16; extra == "all"
+Requires-Dist: keras-nlp>=0.3.1; extra == "all"
+Requires-Dist: torch; extra == "all"
+Requires-Dist: accelerate>=0.21.0; extra == "all"
+Requires-Dist: jax<=0.4.13,>=0.4.1; extra == "all"
+Requires-Dist: jaxlib<=0.4.13,>=0.4.1; extra == "all"
+Requires-Dist: flax<=0.7.0,>=0.4.1; extra == "all"
+Requires-Dist: optax<=0.1.4,>=0.0.8; extra == "all"
+Requires-Dist: scipy<1.13.0; extra == "all"
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "all"
+Requires-Dist: protobuf; extra == "all"
+Requires-Dist: tokenizers<0.20,>=0.19; extra == "all"
+Requires-Dist: torchaudio; extra == "all"
+Requires-Dist: librosa; extra == "all"
+Requires-Dist: pyctcdecode>=0.4.0; extra == "all"
+Requires-Dist: phonemizer; extra == "all"
+Requires-Dist: kenlm; extra == "all"
+Requires-Dist: Pillow<=15.0,>=10.0.1; extra == "all"
+Requires-Dist: optuna; extra == "all"
+Requires-Dist: ray[tune]>=2.7.0; extra == "all"
+Requires-Dist: sigopt; extra == "all"
+Requires-Dist: timm; extra == "all"
+Requires-Dist: torchvision; extra == "all"
+Requires-Dist: Pillow<=15.0,>=10.0.1; extra == "all"
+Requires-Dist: codecarbon==1.2.0; extra == "all"
+Requires-Dist: accelerate>=0.21.0; extra == "all"
+Requires-Dist: decord==0.6.0; extra == "all"
+Requires-Dist: av==9.2.0; extra == "all"
 Provides-Extra: dev-torch
+Requires-Dist: pytest<8.0.0,>=7.2.0; extra == "dev-torch"
+Requires-Dist: pytest-rich; extra == "dev-torch"
+Requires-Dist: pytest-xdist; extra == "dev-torch"
+Requires-Dist: timeout-decorator; extra == "dev-torch"
+Requires-Dist: parameterized; extra == "dev-torch"
+Requires-Dist: psutil; extra == "dev-torch"
+Requires-Dist: datasets!=2.5.0; extra == "dev-torch"
+Requires-Dist: dill<0.3.5; extra == "dev-torch"
+Requires-Dist: evaluate>=0.2.0; extra == "dev-torch"
+Requires-Dist: pytest-timeout; extra == "dev-torch"
+Requires-Dist: ruff==0.1.5; extra == "dev-torch"
+Requires-Dist: sacrebleu<2.0.0,>=1.4.12; extra == "dev-torch"
+Requires-Dist: rouge-score!=0.0.7,!=0.0.8,!=0.1,!=0.1.1; extra == "dev-torch"
+Requires-Dist: nltk; extra == "dev-torch"
+Requires-Dist: GitPython<3.1.19; extra == "dev-torch"
+Requires-Dist: sacremoses; extra == "dev-torch"
+Requires-Dist: rjieba; extra == "dev-torch"
+Requires-Dist: beautifulsoup4; extra == "dev-torch"
+Requires-Dist: tensorboard; extra == "dev-torch"
+Requires-Dist: pydantic; extra == "dev-torch"
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "dev-torch"
+Requires-Dist: faiss-cpu; extra == "dev-torch"
+Requires-Dist: datasets!=2.5.0; extra == "dev-torch"
+Requires-Dist: cookiecutter==1.7.3; extra == "dev-torch"
+Requires-Dist: torch; extra == "dev-torch"
+Requires-Dist: accelerate>=0.21.0; extra == "dev-torch"
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "dev-torch"
+Requires-Dist: protobuf; extra == "dev-torch"
+Requires-Dist: tokenizers<0.20,>=0.19; extra == "dev-torch"
+Requires-Dist: torchaudio; extra == "dev-torch"
+Requires-Dist: librosa; extra == "dev-torch"
+Requires-Dist: pyctcdecode>=0.4.0; extra == "dev-torch"
+Requires-Dist: phonemizer; extra == "dev-torch"
+Requires-Dist: kenlm; extra == "dev-torch"
+Requires-Dist: Pillow<=15.0,>=10.0.1; extra == "dev-torch"
+Requires-Dist: optuna; extra == "dev-torch"
+Requires-Dist: ray[tune]>=2.7.0; extra == "dev-torch"
+Requires-Dist: sigopt; extra == "dev-torch"
+Requires-Dist: timm; extra == "dev-torch"
+Requires-Dist: torchvision; extra == "dev-torch"
+Requires-Dist: Pillow<=15.0,>=10.0.1; extra == "dev-torch"
+Requires-Dist: codecarbon==1.2.0; extra == "dev-torch"
+Requires-Dist: datasets!=2.5.0; extra == "dev-torch"
+Requires-Dist: isort>=5.5.4; extra == "dev-torch"
+Requires-Dist: ruff==0.1.5; extra == "dev-torch"
+Requires-Dist: GitPython<3.1.19; extra == "dev-torch"
+Requires-Dist: urllib3<2.0.0; extra == "dev-torch"
+Requires-Dist: fugashi>=1.0; extra == "dev-torch"
+Requires-Dist: ipadic<2.0,>=1.0.0; extra == "dev-torch"
+Requires-Dist: unidic_lite>=1.0.7; extra == "dev-torch"
+Requires-Dist: unidic>=1.0.2; extra == "dev-torch"
+Requires-Dist: sudachipy>=0.6.6; extra == "dev-torch"
+Requires-Dist: sudachidict_core>=20220729; extra == "dev-torch"
+Requires-Dist: rhoknp<1.3.1,>=1.1.0; extra == "dev-torch"
+Requires-Dist: scikit-learn; extra == "dev-torch"
+Requires-Dist: cookiecutter==1.7.3; extra == "dev-torch"
+Requires-Dist: onnxruntime>=1.4.0; extra == "dev-torch"
+Requires-Dist: onnxruntime-tools>=1.4.2; extra == "dev-torch"
 Provides-Extra: dev-tensorflow
+Requires-Dist: pytest<8.0.0,>=7.2.0; extra == "dev-tensorflow"
+Requires-Dist: pytest-rich; extra == "dev-tensorflow"
+Requires-Dist: pytest-xdist; extra == "dev-tensorflow"
+Requires-Dist: timeout-decorator; extra == "dev-tensorflow"
+Requires-Dist: parameterized; extra == "dev-tensorflow"
+Requires-Dist: psutil; extra == "dev-tensorflow"
+Requires-Dist: datasets!=2.5.0; extra == "dev-tensorflow"
+Requires-Dist: dill<0.3.5; extra == "dev-tensorflow"
+Requires-Dist: evaluate>=0.2.0; extra == "dev-tensorflow"
+Requires-Dist: pytest-timeout; extra == "dev-tensorflow"
+Requires-Dist: ruff==0.1.5; extra == "dev-tensorflow"
+Requires-Dist: sacrebleu<2.0.0,>=1.4.12; extra == "dev-tensorflow"
+Requires-Dist: rouge-score!=0.0.7,!=0.0.8,!=0.1,!=0.1.1; extra == "dev-tensorflow"
+Requires-Dist: nltk; extra == "dev-tensorflow"
+Requires-Dist: GitPython<3.1.19; extra == "dev-tensorflow"
+Requires-Dist: sacremoses; extra == "dev-tensorflow"
+Requires-Dist: rjieba; extra == "dev-tensorflow"
+Requires-Dist: beautifulsoup4; extra == "dev-tensorflow"
+Requires-Dist: tensorboard; extra == "dev-tensorflow"
+Requires-Dist: pydantic; extra == "dev-tensorflow"
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "dev-tensorflow"
+Requires-Dist: faiss-cpu; extra == "dev-tensorflow"
+Requires-Dist: datasets!=2.5.0; extra == "dev-tensorflow"
+Requires-Dist: cookiecutter==1.7.3; extra == "dev-tensorflow"
+Requires-Dist: tensorflow<2.16,>2.9; extra == "dev-tensorflow"
+Requires-Dist: onnxconverter-common; extra == "dev-tensorflow"
+Requires-Dist: tf2onnx; extra == "dev-tensorflow"
+Requires-Dist: tensorflow-text<2.16; extra == "dev-tensorflow"
+Requires-Dist: keras-nlp>=0.3.1; extra == "dev-tensorflow"
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "dev-tensorflow"
+Requires-Dist: protobuf; extra == "dev-tensorflow"
+Requires-Dist: tokenizers<0.20,>=0.19; extra == "dev-tensorflow"
+Requires-Dist: Pillow<=15.0,>=10.0.1; extra == "dev-tensorflow"
+Requires-Dist: datasets!=2.5.0; extra == "dev-tensorflow"
+Requires-Dist: isort>=5.5.4; extra == "dev-tensorflow"
+Requires-Dist: ruff==0.1.5; extra == "dev-tensorflow"
+Requires-Dist: GitPython<3.1.19; extra == "dev-tensorflow"
+Requires-Dist: urllib3<2.0.0; extra == "dev-tensorflow"
+Requires-Dist: scikit-learn; extra == "dev-tensorflow"
+Requires-Dist: cookiecutter==1.7.3; extra == "dev-tensorflow"
+Requires-Dist: onnxconverter-common; extra == "dev-tensorflow"
+Requires-Dist: tf2onnx; extra == "dev-tensorflow"
+Requires-Dist: onnxruntime>=1.4.0; extra == "dev-tensorflow"
+Requires-Dist: onnxruntime-tools>=1.4.2; extra == "dev-tensorflow"
+Requires-Dist: librosa; extra == "dev-tensorflow"
+Requires-Dist: pyctcdecode>=0.4.0; extra == "dev-tensorflow"
+Requires-Dist: phonemizer; extra == "dev-tensorflow"
+Requires-Dist: kenlm; extra == "dev-tensorflow"
 Provides-Extra: dev
+Requires-Dist: tensorflow<2.16,>2.9; extra == "dev"
+Requires-Dist: onnxconverter-common; extra == "dev"
+Requires-Dist: tf2onnx; extra == "dev"
+Requires-Dist: tensorflow-text<2.16; extra == "dev"
+Requires-Dist: keras-nlp>=0.3.1; extra == "dev"
+Requires-Dist: torch; extra == "dev"
+Requires-Dist: accelerate>=0.21.0; extra == "dev"
+Requires-Dist: jax<=0.4.13,>=0.4.1; extra == "dev"
+Requires-Dist: jaxlib<=0.4.13,>=0.4.1; extra == "dev"
+Requires-Dist: flax<=0.7.0,>=0.4.1; extra == "dev"
+Requires-Dist: optax<=0.1.4,>=0.0.8; extra == "dev"
+Requires-Dist: scipy<1.13.0; extra == "dev"
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "dev"
+Requires-Dist: protobuf; extra == "dev"
+Requires-Dist: tokenizers<0.20,>=0.19; extra == "dev"
+Requires-Dist: torchaudio; extra == "dev"
+Requires-Dist: librosa; extra == "dev"
+Requires-Dist: pyctcdecode>=0.4.0; extra == "dev"
+Requires-Dist: phonemizer; extra == "dev"
+Requires-Dist: kenlm; extra == "dev"
+Requires-Dist: Pillow<=15.0,>=10.0.1; extra == "dev"
+Requires-Dist: optuna; extra == "dev"
+Requires-Dist: ray[tune]>=2.7.0; extra == "dev"
+Requires-Dist: sigopt; extra == "dev"
+Requires-Dist: timm; extra == "dev"
+Requires-Dist: torchvision; extra == "dev"
+Requires-Dist: Pillow<=15.0,>=10.0.1; extra == "dev"
+Requires-Dist: codecarbon==1.2.0; extra == "dev"
+Requires-Dist: accelerate>=0.21.0; extra == "dev"
+Requires-Dist: decord==0.6.0; extra == "dev"
+Requires-Dist: av==9.2.0; extra == "dev"
+Requires-Dist: pytest<8.0.0,>=7.2.0; extra == "dev"
+Requires-Dist: pytest-rich; extra == "dev"
+Requires-Dist: pytest-xdist; extra == "dev"
+Requires-Dist: timeout-decorator; extra == "dev"
+Requires-Dist: parameterized; extra == "dev"
+Requires-Dist: psutil; extra == "dev"
+Requires-Dist: datasets!=2.5.0; extra == "dev"
+Requires-Dist: dill<0.3.5; extra == "dev"
+Requires-Dist: evaluate>=0.2.0; extra == "dev"
+Requires-Dist: pytest-timeout; extra == "dev"
+Requires-Dist: ruff==0.1.5; extra == "dev"
+Requires-Dist: sacrebleu<2.0.0,>=1.4.12; extra == "dev"
+Requires-Dist: rouge-score!=0.0.7,!=0.0.8,!=0.1,!=0.1.1; extra == "dev"
+Requires-Dist: nltk; extra == "dev"
+Requires-Dist: GitPython<3.1.19; extra == "dev"
+Requires-Dist: sacremoses; extra == "dev"
+Requires-Dist: rjieba; extra == "dev"
+Requires-Dist: beautifulsoup4; extra == "dev"
+Requires-Dist: tensorboard; extra == "dev"
+Requires-Dist: pydantic; extra == "dev"
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "dev"
+Requires-Dist: faiss-cpu; extra == "dev"
+Requires-Dist: datasets!=2.5.0; extra == "dev"
+Requires-Dist: cookiecutter==1.7.3; extra == "dev"
+Requires-Dist: datasets!=2.5.0; extra == "dev"
+Requires-Dist: isort>=5.5.4; extra == "dev"
+Requires-Dist: ruff==0.1.5; extra == "dev"
+Requires-Dist: GitPython<3.1.19; extra == "dev"
+Requires-Dist: urllib3<2.0.0; extra == "dev"
+Requires-Dist: fugashi>=1.0; extra == "dev"
+Requires-Dist: ipadic<2.0,>=1.0.0; extra == "dev"
+Requires-Dist: unidic_lite>=1.0.7; extra == "dev"
+Requires-Dist: unidic>=1.0.2; extra == "dev"
+Requires-Dist: sudachipy>=0.6.6; extra == "dev"
+Requires-Dist: sudachidict_core>=20220729; extra == "dev"
+Requires-Dist: rhoknp<1.3.1,>=1.1.0; extra == "dev"
+Requires-Dist: scikit-learn; extra == "dev"
+Requires-Dist: cookiecutter==1.7.3; extra == "dev"
 Provides-Extra: torchhub
+Requires-Dist: filelock; extra == "torchhub"
+Requires-Dist: huggingface-hub<1.0,>=0.23.0; extra == "torchhub"
+Requires-Dist: importlib_metadata; extra == "torchhub"
+Requires-Dist: numpy>=1.17; extra == "torchhub"
+Requires-Dist: packaging>=20.0; extra == "torchhub"
+Requires-Dist: protobuf; extra == "torchhub"
+Requires-Dist: regex!=2019.12.17; extra == "torchhub"
+Requires-Dist: requests; extra == "torchhub"
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "torchhub"
+Requires-Dist: torch; extra == "torchhub"
+Requires-Dist: tokenizers<0.20,>=0.19; extra == "torchhub"
+Requires-Dist: tqdm>=4.27; extra == "torchhub"
 Provides-Extra: agents
-License-File: LICENSE
+Requires-Dist: diffusers; extra == "agents"
+Requires-Dist: accelerate>=0.21.0; extra == "agents"
+Requires-Dist: datasets!=2.5.0; extra == "agents"
+Requires-Dist: torch; extra == "agents"
+Requires-Dist: sentencepiece!=0.1.92,>=0.1.91; extra == "agents"
+Requires-Dist: opencv-python; extra == "agents"
+Requires-Dist: Pillow<=15.0,>=10.0.1; extra == "agents"
 
 <!---
 Copyright 2020 The HuggingFace Team. All rights reserved.

{transformers-4.41.0 → transformers-4.41.2}/setup.py

@@ -426,7 +426,7 @@ install_requires = [
 
 setup(
     name="transformers",
-    version="4.41.0",  # expected format is one of x.y.z.dev0, or x.y.z.rc1 or x.y.z (no to dashes, yes to dots)
+    version="4.41.2",  # expected format is one of x.y.z.dev0, or x.y.z.rc1 or x.y.z (no to dashes, yes to dots)
     author="The Hugging Face team (past and future) with the help of all our contributors (https://github.com/huggingface/transformers/graphs/contributors)",
     author_email="transformers@huggingface.co",
     description="State-of-the-art Machine Learning for JAX, PyTorch and TensorFlow",

{transformers-4.41.0 → transformers-4.41.2}/src/transformers/__init__.py

@@ -18,7 +18,7 @@
 # to defer the actual importing for when the objects are requested. This way `import transformers` provides the names
 # in the namespace without actually importing anything (and especially none of the backends).
 
-__version__ = "4.41.0"
+__version__ = "4.41.2"
 
 from typing import TYPE_CHECKING
 

{transformers-4.41.0 → transformers-4.41.2}/src/transformers/dynamic_module_utils.py

@@ -197,7 +197,10 @@ def get_class_in_module(class_name: str, module_path: Union[str, os.PathLike]) -
     Returns:
         `typing.Type`: The class looked for.
     """
-    name = os.path.normpath(module_path).rstrip(".py").replace(os.path.sep, ".")
+    name = os.path.normpath(module_path)
+    if name.endswith(".py"):
+        name = name[:-3]
+    name = name.replace(os.path.sep, ".")
     module_spec = importlib.util.spec_from_file_location(name, location=Path(HF_MODULES_CACHE) / module_path)
     module = sys.modules.get(name)
     if module is None:

{transformers-4.41.0 → transformers-4.41.2}/src/transformers/generation/utils.py

@@ -1354,6 +1354,23 @@ class GenerationMixin:
             self._static_cache.reset()  # reset the cache for a new generation
         return self._static_cache
 
+    def _get_decoder_start_token_id(
+        self, decoder_start_token_id: Union[int, List[int]] = None, bos_token_id: int = None
+    ) -> int:
+        decoder_start_token_id = (
+            decoder_start_token_id
+            if decoder_start_token_id is not None
+            else self.generation_config.decoder_start_token_id
+        )
+        bos_token_id = bos_token_id if bos_token_id is not None else self.generation_config.bos_token_id
+
+        if decoder_start_token_id is not None:
+            return decoder_start_token_id
+        elif bos_token_id is not None:
+            return bos_token_id
+        else:
+            return
+
     def _prepare_special_tokens(
         self,
         generation_config: GenerationConfig,
@@ -1378,11 +1395,16 @@ class GenerationMixin:
                 return token
             return torch.tensor(token, device=device, dtype=torch.long)
 
+        # for BC we also try to get `decoder_start_token_id` from model's generation config (#30892)
+        if self.config.is_encoder_decoder:
+            generation_config.decoder_start_token_id = self._get_decoder_start_token_id(
+                generation_config.decoder_start_token_id, generation_config.bos_token_id
+            )
+
         bos_token_id = _tensor_or_none(generation_config.bos_token_id, device=device)
         eos_token_id = _tensor_or_none(generation_config.eos_token_id, device=device)
         pad_token_id = _tensor_or_none(generation_config.pad_token_id, device=device)
         decoder_start_token_id = _tensor_or_none(generation_config.decoder_start_token_id, device=device)
-        decoder_start_token_id = decoder_start_token_id if decoder_start_token_id is not None else bos_token_id
 
         # We can have more than one eos token. Always treat it as a 1D tensor (when it exists).
         if eos_token_id is not None and eos_token_id.ndim == 0: