python3Packages.datasets: init at 1.0.0

This library provides open-access datasets an evaluation metrics for
natural language processing using PyTorch, TensorFlow, NumPy, and
Pandas.
This commit is contained in:
Daniël de Kok 2020-09-10 20:18:22 +02:00 committed by Jon
parent 9627ae4f9a
commit 1839c5cd7b
2 changed files with 54 additions and 0 deletions

View file

@ -0,0 +1,52 @@
{ lib
, buildPythonPackage
, fetchFromGitHub
, dill
, filelock
, numpy
, pandas
, pyarrow
, requests
, tqdm
, xxhash
}:
buildPythonPackage rec {
pname = "datasets";
version = "1.0.0";
src = fetchFromGitHub {
owner = "huggingface";
repo = pname;
rev = version;
sha256 = "13l52r7nhj2c1a10isy5309d2g6pmaivyqs5w6yjbjj4195jxya5";
};
propagatedBuildInputs = [
dill
filelock
numpy
pandas
pyarrow
requests
tqdm
xxhash
];
# Tests require pervasive internet access.
doCheck = false;
# Module import will attempt to create a cache directory.
postFixup = "export HF_MODULES_CACHE=$TMPDIR";
pythonImportsCheck = [ "datasets" ];
meta = with lib; {
homepage = "https://github.com/huggingface/datasets";
description = "Fast, efficient, open-access datasets and evaluation metrics for natural language processing";
changelog = "https://github.com/huggingface/datasets/releases/tag/${version}";
license = licenses.asl20;
platforms = platforms.unix;
maintainers = with maintainers; [ danieldk ];
};
}

View file

@ -1436,6 +1436,8 @@ in {
datamodeldict = callPackage ../development/python-modules/datamodeldict { };
datasets = callPackage ../development/python-modules/datasets { };
datasette = callPackage ../development/python-modules/datasette { };
datashader = callPackage ../development/python-modules/datashader { };