nixpkgs/pkgs/development/python-modules/pytorch/default.nix

71 lines
2.1 KiB
Nix
Raw Normal View History

{ buildPythonPackage, pythonOlder,
cudaSupport ? false, cudatoolkit ? null, cudnn ? null,
fetchFromGitHub, fetchpatch, lib, numpy, pyyaml, cffi, typing, cmake,
stdenv, linkFarm, symlinkJoin,
utillinux, which }:
2017-07-16 20:15:05 +01:00
assert cudnn == null || cudatoolkit != null;
assert !cudaSupport || cudatoolkit != null;
let
cudatoolkit_joined = symlinkJoin {
name = "${cudatoolkit.name}-unsplit";
paths = [ cudatoolkit.out cudatoolkit.lib ];
};
# Normally libcuda.so.1 is provided at runtime by nvidia-x11 via
# LD_LIBRARY_PATH=/run/opengl-driver/lib. We only use the stub
# libcuda.so from cudatoolkit for running tests, so that we dont have
# to recompile pytorch on every update to nvidia-x11 or the kernel.
cudaStub = linkFarm "cuda-stub" [{
name = "libcuda.so.1";
path = "${cudatoolkit}/lib/stubs/libcuda.so";
}];
cudaStubEnv = lib.optionalString cudaSupport
"LD_LIBRARY_PATH=${cudaStub}\${LD_LIBRARY_PATH:+:$LD_LIBRARY_PATH} ";
in buildPythonPackage rec {
version = "0.4.0";
2017-07-16 20:15:05 +01:00
pname = "pytorch";
name = "${pname}-${version}";
src = fetchFromGitHub {
2017-08-19 02:22:23 +01:00
owner = "pytorch";
repo = "pytorch";
rev = "v${version}";
fetchSubmodules = true;
sha256 = "12d5vqqaprk0igmih7fwa65ldmaawgijxl58h6dnw660wysc132j";
2017-07-16 20:15:05 +01:00
};
preConfigure = lib.optionalString cudaSupport ''
export CC=${cudatoolkit.cc}/bin/gcc CXX=${cudatoolkit.cc}/bin/g++
'' + lib.optionalString (cudaSupport && cudnn != null) ''
export CUDNN_INCLUDE_DIR=${cudnn}/include
2017-07-16 20:15:05 +01:00
'';
buildInputs = [
cmake
numpy.blas
utillinux
which
] ++ lib.optionals cudaSupport [cudatoolkit_joined cudnn];
2017-11-22 22:02:34 +00:00
2017-07-16 20:15:05 +01:00
propagatedBuildInputs = [
cffi
numpy
pyyaml
] ++ lib.optional (pythonOlder "3.5") typing;
2017-07-16 20:15:05 +01:00
checkPhase = ''
${cudaStubEnv}python test/run_test.py --exclude distributed
2017-07-16 20:15:05 +01:00
'';
2017-11-22 22:02:34 +00:00
2017-07-16 20:15:05 +01:00
meta = {
description = "Tensors and Dynamic neural networks in Python with strong GPU acceleration.";
homepage = http://pytorch.org/;
license = lib.licenses.bsd3;
platforms = lib.platforms.linux;
maintainers = with lib.maintainers; [ teh ];
};
}