Something went wrong. Try again.
What happens if I push all of nixpkgs
Something went wrong. Try again.
2.9 kB · 131 lines
Nix
at gcc-offload
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132{ lib, stdenv, buildPythonPackage, fetchFromGitHub,
# build-system setuptools,
# dependencies absl-py, h5py, ml-dtypes, namex, numpy, optree, packaging, rich, tensorflow, pythonAtLeast, distutils,
# tests dm-tree, jax, jaxlib, pandas, pydot, pytestCheckHook, tf-keras, torch,}:
buildPythonPackage rec { pname = "keras"; version = "3.7.0"; pyproject = true;
src = fetchFromGitHub { owner = "keras-team"; repo = "keras"; rev = "refs/tags/v${version}"; hash = "sha256-qidY1OmlOYPKVoxryx1bEukA7IS6rPV4jqlnuf3y39w="; };
build-system = [ setuptools ];
dependencies = [ absl-py h5py ml-dtypes namex numpy optree packaging rich tensorflow ] ++ lib.optionals (pythonAtLeast "3.12") [ distutils ];
pythonImportsCheck = [ "keras" "keras._tf_keras" ];
nativeCheckInputs = [ dm-tree jaxlib jax pandas pydot pytestCheckHook tf-keras torch ];
preCheck = '' export HOME=$(mktemp -d) '';
disabledTests = [ # Tries to install the package in the sandbox "test_keras_imports"
# TypeError: this __dict__ descriptor does not support '_DictWrapper' objects "test_reloading_default_saved_model" ] ++ lib.optionals stdenv.isDarwin [ # AttributeError: module 'numpy' has no attribute 'float128'. Did you mean: 'float16'? "test_spectrogram_error" ];
disabledTestPaths = [ # Datasets are downloaded from the internet "integration_tests/dataset_tests"
# TypeError: test_custom_fit.<locals>.CustomModel.train_step() missing 1 required positional argument: 'data' "integration_tests/jax_custom_fit_test.py"
# RuntimeError: Virtual devices cannot be modified after being initialized "integration_tests/tf_distribute_training_test.py"
# AttributeError: 'CustomModel' object has no attribute 'zero_grad' "integration_tests/torch_custom_fit_test.py"
# Fails for an unclear reason: # self.assertLen(list(net.parameters()), 2 # AssertionError: 0 != 2 "integration_tests/torch_workflow_test.py"
# Most tests require internet access "keras/src/applications/applications_test.py"
# TypeError: this __dict__ descriptor does not support '_DictWrapper' objects "keras/src/backend/tensorflow/saved_model_test.py" "keras/src/export/export_lib_test.py"
# KeyError: 'Unable to synchronously open object (bad object header version number)' "keras/src/saving/file_editor_test.py" ];
meta = { description = "Multi-backend implementation of the Keras API, with support for TensorFlow, JAX, and PyTorch"; homepage = "https://keras.io"; changelog = "https://github.com/keras-team/keras/releases/tag/v${version}"; license = lib.licenses.mit; maintainers = with lib.maintainers; [ GaetanLepage ]; };}