File python-datasets.spec of Package python-datasets
#
# spec file for package python-datasets
#
# Copyright (c) 2024 SUSE LLC
#
# All modifications and additions to the file contributed by third parties
# remain the property of their copyright owners, unless otherwise agreed
# upon. The license for this file, and modifications and additions to the
# file, is the same license as for the pristine package itself (unless the
# license for the pristine package is not an Open Source License, in which
# case the license is the MIT License). An "Open Source License" is a
# license that conforms to the Open Source Definition (Version 1.9)
# published by the Open Source Initiative.
# Please submit bugfixes or comments via https://bugs.opensuse.org/
#
%{?sle15_python_module_pythons}
Name: python-datasets
Version: 3.6.0
Release: 0
Summary: HuggingFace community-driven open-source library of datasets
License: Apache-2.0
URL: https://github.com/huggingface/datasets
Source: https://files.pythonhosted.org/packages/source/d/datasets/datasets-%{version}.tar.gz
BuildRequires: python-rpm-macros
BuildRequires: %{python_module pip}
BuildRequires: %{python_module setuptools}
BuildRequires: %{python_module wheel}
# SECTION test requirements
BuildRequires: %{python_module aiohttp}
BuildRequires: %{python_module absl-py}
BuildRequires: %{python_module decorator}
BuildRequires: %{python_module dill >= 0.3.0}
BuildRequires: %{python_module filelock}
BuildRequires: %{python_module fsspec >= 2023.1.0}
BuildRequires: %{python_module huggingface-hub >= 0.24.0}
BuildRequires: %{python_module multiprocess <= 0.71.0}
BuildRequires: %{python_module numpy >= 1.17}
BuildRequires: %{python_module packaging}
BuildRequires: %{python_module pandas}
BuildRequires: %{python_module pyarrow >= 15.0.0}
BuildRequires: %{python_module Pillow}
BuildRequires: %{python_module PyYAML >= 5.1}
BuildRequires: %{python_module requests >= 2.32.2}
BuildRequires: %{python_module SQLAlchemy}
BuildRequires: %{python_module torch}
BuildRequires: %{python_module tqdm >= 4.66.3}
BuildRequires: %{python_module xxhash}
BuildRequires: %{python_module zstandard}
BuildRequires: %{python_module SoundFile}
BuildRequires: %{python_module pytest}
BuildRequires: %{python_module pytest-datadir}
BuildRequires: %{python_module sympy}
# /SECTION
BuildRequires: fdupes
Requires: python-aiohttp
Requires: python-dill >= 0.3.0
Requires: python-filelock
Requires: python-fsspec >= 2023.1.0
Requires: python-huggingface-hub >= 0.24.0
Requires: python-multiprocess <= 0.71.0
Requires: python-numpy >= 1.17
Requires: python-packaging
Requires: python-pandas
Requires: python-pyarrow >= 15.0.0
Requires: python-PyYAML >= 5.1
Requires: python-requests >= 2.32.2
Requires: python-tqdm >= 4.66.3
Requires: python-xxhash
Suggests: python-soundfile >= 0.12.1
Suggests: python-librosa
Suggests: python-soxr >= 0.4.0
Suggests: python-tensorflow == 2.12.0
Suggests: python-torch == 2.0.1
Suggests: python-transformers == 4.30.1
Suggests: python-absl-py
Suggests: python-decorator
Suggests: python-joblib < 1.3.0
Suggests: python-joblibspark
Suggests: python-pytest
Suggests: python-pytest-datadir
Suggests: python-pytest-xdist
Suggests: python-elasticsearch < 8.0.0
Suggests: python-faiss-cpu >= 1.8.0.post1
Suggests: python-jax >= 0.3.14
Suggests: python-jaxlib >= 0.3.14
Suggests: python-lz4
Suggests: python-moto
Suggests: python-py7zr
Suggests: python-rarfile >= 4.0
Suggests: python-sqlalchemy
Suggests: python-s3fs >= 2021.11.1
Suggests: python-protobuf < 4.0.0
Suggests: python-tensorflow >= 2.16.0
Suggests: python-tiktoken
Suggests: python-torch >= 2.0.0
Suggests: python-torchdata
Suggests: python-soundfile >= 0.12.1
Suggests: python-transformers >= 4.42.0
Suggests: python-zstandard
Suggests: python-polars >= 0.20.0
Suggests: python-Pillow >= 9.4.0
Suggests: python-soundfile >= 0.12.1
Suggests: python-librosa
Suggests: python-soxr >= 0.4.0
Suggests: python-ruff >= 0.3.0
Suggests: python-s3fs
Suggests: python-transformers
Suggests: python-torch
Suggests: python-tensorflow >= 2.6.0
Suggests: python-s3fs
Suggests: python-transformers
Suggests: python-torch
Suggests: python-tensorflow >= 2.6.0
Suggests: python-jax >= 0.3.14
Suggests: python-jaxlib >= 0.3.14
Suggests: python-ruff >= 0.3.0
Suggests: python-s3fs
Suggests: python-tensorflow >= 2.6.0
Suggests: python-tensorflow >= 2.6.0
Suggests: python-absl-py
Suggests: python-decorator
Suggests: python-joblib < 1.3.0
Suggests: python-joblibspark
Suggests: python-pytest
Suggests: python-pytest-datadir
Suggests: python-pytest-xdist
Suggests: python-elasticsearch < 8.0.0
Suggests: python-faiss-cpu >= 1.8.0.post1
Suggests: python-jax >= 0.3.14
Suggests: python-jaxlib >= 0.3.14
Suggests: python-lz4
Suggests: python-moto
Suggests: python-py7zr
Suggests: python-rarfile >= 4.0
Suggests: python-sqlalchemy
Suggests: python-s3fs >= 2021.11.1
Suggests: python-protobuf < 4.0.0
Suggests: python-tensorflow >= 2.16.0
Suggests: python-tiktoken
Suggests: python-torch >= 2.0.0
Suggests: python-torchdata
Suggests: python-soundfile >= 0.12.1
Suggests: python-transformers >= 4.42.0
Suggests: python-zstandard
Suggests: python-polars >= 0.20.0
Suggests: python-Pillow >= 9.4.0
Suggests: python-soundfile >= 0.12.1
Suggests: python-librosa
Suggests: python-soxr >= 0.4.0
Suggests: python-absl-py
Suggests: python-decorator
Suggests: python-joblib < 1.3.0
Suggests: python-joblibspark
Suggests: python-pytest
Suggests: python-pytest-datadir
Suggests: python-pytest-xdist
Suggests: python-elasticsearch < 8.0.0
Suggests: python-jax >= 0.3.14
Suggests: python-jaxlib >= 0.3.14
Suggests: python-lz4
Suggests: python-moto
Suggests: python-py7zr
Suggests: python-rarfile >= 4.0
Suggests: python-sqlalchemy
Suggests: python-s3fs >= 2021.11.1
Suggests: python-protobuf < 4.0.0
Suggests: python-tiktoken
Suggests: python-torch >= 2.0.0
Suggests: python-torchdata
Suggests: python-soundfile >= 0.12.1
Suggests: python-transformers >= 4.42.0
Suggests: python-zstandard
Suggests: python-polars >= 0.20.0
Suggests: python-Pillow >= 9.4.0
Suggests: python-soundfile >= 0.12.1
Suggests: python-soxr >= 0.4.0
Suggests: python-torch
Suggests: python-Pillow >= 9.4.0
BuildArch: noarch
Requires(post): update-alternatives
Requires(postun): update-alternatives
%python_subpackages
%description
HuggingFace community-driven open-source library of datasets.
%prep
%autosetup -p1 -n datasets-%{version}
sed -i -e '1{/#!.*env python/d;}' \
src/datasets/commands/datasets_cli.py \
src/datasets/utils/_filelock.py
%build
%pyproject_wheel
%install
%pyproject_install
%python_clone -a %{buildroot}%{_bindir}/datasets-cli
%python_expand %fdupes %{buildroot}%{$python_sitelib}
%check
# NOTE: disable Apache Spark tests since we don't have
# a python-pyspark package. That package depends on
# py4j which only supports upto Python 3.5.
# NOTE: disable tests which required the undefined "shared_datadir" fixture.
# * test_audio.py
# * test_image.py
# NOTE: disable tests which required internet connect.
# * test_json.py
# * test_file_utils.py
# * test_inspect.py
# * test_load.py
# * test_upstream_hub.py
# * test_cache.py
# * test_hub.py
# * test_iterable_dataset.py
# * test_folder_based_builder.py
# * test_test.py
# * test_parquet.py
# * test_offline_util.py
# NOTE: disable these tests as it requires AWS S3 access.
# * test_arrow_dataset.py
# FIXME: skip test_iterable_dataset_persists_epoch_in_torch_workers for now as it got stuck on a wait,
# possibly due to lack of GPU.
# NOTE: disable tests/test_fingerprint.py as it involves NVIDIA CUDA.
# NOTE: disable tests/test_distributed.py as coroutines is prohibited.
# NOTE: skip test_tensor_webdataset till this bugfix is included: https://github.com/python/cpython/issues/125631
%pytest --ignore tests/test_offline_util.py --ignore tests/test_distributed.py --ignore tests/test_fingerprint.py --ignore tests/test_data_files.py --ignore tests/test_arrow_dataset.py --ignore tests/io/test_parquet.py --ignore tests/commands/test_test.py --ignore tests/packaged_modules/test_spark.py --ignore tests/packaged_modules/test_folder_based_builder.py --ignore tests/features/test_audio.py --ignore tests/features/test_image.py --ignore tests/io/test_json.py --ignore tests/fixtures/hub.py --ignore tests/test_file_utils.py --ignore tests/test_inspect.py --ignore tests/test_load.py --ignore tests/test_upstream_hub.py --ignore tests/packaged_modules/test_cache.py --ignore tests/test_hub.py --ignore tests/test_iterable_dataset.py -k "not (test_iterable_dataset_persists_epoch_in_torch_workers or test_dataset_save_to_disk_and_load_from_disk_round_trip_with_large_list or test_tensor_webdataset)"
%post
%python_install_alternative datasets-cli
%postun
%python_uninstall_alternative datasets-cli
%files %{python_files}
%doc AUTHORS README.md
%license LICENSE
%python_alternative %{_bindir}/datasets-cli
%{python_sitelib}/datasets
%{python_sitelib}/datasets-%{version}.dist-info
%changelog
