summaryrefslogtreecommitdiff
path: root/pkgs/development/python-modules/pytorch (follow)
Commit message (Expand)AuthorAgeFilesLines
* Merge remote-tracking branch 'upstream/master' into structured-attrsorigin/structured-attrsRobin Gloster2021-05-242-7/+11
|\
| * python3Packages.pytorch-bin: use CUDA 11 wheelsDaniël de Kok2021-05-121-6/+6
| * python3Packages.pytorch-bin: add meta.changelogDaniël de Kok2021-05-111-0/+1
| * python3Packages.pytorch-bin: 1.8.0 -> 1.8.1•••Changelog: https://github.com/pytorch/pytorch/releases/tag/v1.8.1 Daniël de Kok2021-05-112-7/+10
* | Merge remote-tracking branch 'upstream/master' into structured-attrsRobin Gloster2021-04-291-0/+3
|\|
| * python3Packages.pytorch: require big-parallel•••This compiles in usually about 2h15m with a 2-core build, but about 10m on a big-parallel machine. Luke Granger-Brown2021-04-261-0/+3
* | Merge remote-tracking branch 'upstream/master' into structured-attrsRobin Gloster2021-04-231-8/+10
|\|
| * Merge pull request #118964 from danieldk/pytorch-no-tests•••python3Packages.pytorch: disable building of testsJörg Thalheim2021-04-111-6/+8
| |\
| | * python3Packages.pytorch: remove unused USE_MKL variable•••PyTorch will always try to find MKL, USE_MKL does not do anything anymore. So, we have to rely on MKL not being in the build environment. Daniël de Kok2021-04-101-2/+0
| | * python3Packages.pytorch: disable building of tests•••- We do not run them. - The test binaries are installed into site-packages by default. Daniël de Kok2021-04-101-0/+3
| | * python3Packages.pytorch: convert boolean environment variables•••We relied on the stdenv serialization for booleans. However, the PyTorch configure script expects 0/1. So convert the options to 0/1 values. Daniël de Kok2021-04-101-6/+7
| * | python3Packages.pytorch: 1.8.0 -> 1.8.1•••Changelog: https://github.com/pytorch/pytorch/releases/tag/v1.8.1 Daniël de Kok2021-03-251-2/+2
| |/
* | Merge remote-tracking branch 'upstream/master' into structured-attrsRobin Gloster2021-03-163-33/+39
|\|
| * pythonPackages.pytorch: remove ? null from inputsSandro Jäckel2021-03-101-3/+1
| * pythonPackages.pytorch: mark broken for darwinSandro Jäckel2021-03-101-5/+6
| * Merge pull request #115166 from danieldk/pytorch-bin-1.8.0•••python3Packages.pytorch-bin: 1.7.1 -> 1.8.0Daniël de Kok2021-03-082-5/+4
| |\
| | * python3Packages.pytorch-bin: 1.7.1 -> 1.8.0•••Changelog: https://github.com/pytorch/pytorch/releases/tag/v1.8.0 Daniël de Kok2021-03-062-5/+4
| * | python3Packages.pytorch: 1.7.1 -> 1.8.0•••Changelog: https://github.com/pytorch/pytorch/releases/tag/v1.8.0 Daniël de Kok2021-03-051-2/+2
| * | Merge pull request #113207 from danieldk/pytorch-cuda-11•••python3Packages.pytorch: add compute capabilities for CUDA 11Daniël de Kok2021-03-051-18/+26
| |\ \ | | |/ | |/|
| | * python3Packages.pytorch: add compute capabilities for CUDA 11•••CUDA 11 supports capabilities 8.0 and 8.6. The change adds these capabilities when CUDA 11 is used, enabling support for Ampere GPUs. Daniël de Kok2021-03-041-18/+26
* | | Merge remote-tracking branch 'upstream/master' into structured-attrsRobin Gloster2021-03-041-2/+1
|\| |
| * | python3Packages.pytorch-bin: do not depend on nvidia_x11•••libcuda.so should be provided through /run/opengl-driver/lib to use the system's version. This also avoids rebuilds when linuxPackages is updated. Daniël de Kok2021-02-151-2/+1
| |/
* | Merge remote-tracking branch 'upstream/master' into structured-attrsRobin Gloster2021-02-284-63/+257
|\|
| * pkgs/development/python-modules: stdenv.lib -> libPavol Rusnak2021-01-242-2/+2
| * mpi: use mpi attribute consistently as the default MPI implementations•••Use the attribute mpi to provide a system wide default MPI implementation. The default is openmpi (as before). This now allows for overriding the MPI implentation by using the overlay mechanism. Build all packages with mpich instead of the default openmpi can now be achived like this: self: super: { mpi = super.mpich; } All derivations that have been using "mpi ? null" to provide optional building with MPI have been change in the following way to allow for optional builds with MPI: { ... , mpi , useMpi ? false } Markus Kowalewski2021-01-231-5/+3
| * treewide: with stdenv.lib; in meta -> with lib;•••Part of: https://github.com/NixOS/nixpkgs/issues/108938 meta = with stdenv.lib; is a widely used pattern. We want to slowly remove the `stdenv.lib` indirection and encourage people to use `lib` directly. Thus let’s start with the meta field. This used a rewriting script to mostly automatically replace all occurances of this pattern, and add the `lib` argument to the package header if it doesn’t exist yet. The script in its current form is available at https://cs.tvl.fyi/depot@2f807d7f141068d2d60676a89213eaa5353ca6e0/-/blob/users/Profpatsch/nixpkgs-rewriter/default.nix Profpatsch2021-01-111-2/+2
| * Merge pull request #106585 from danieldk/pytorch-bin-1.7.1•••python3Packages.pytorch-bin: 1.7.0 -> 1.7.1Daniël de Kok2020-12-162-14/+9
| |\
| | * python3Packages.pytorch-bin: 1.7.0 -> 1.7.1•••This also adds Python 3.9 compatibility. Daniël de Kok2020-12-112-14/+9
| * | python3Packages.pytorch: 1.7.0 -> 1.7.1•••Changelog: https://github.com/pytorch/pytorch/releases/tag/v1.7.1 Daniël de Kok2020-12-111-2/+2
| |/
| * python3Packages.pytorch: add danieldk as a maintainerDaniël de Kok2020-12-041-1/+1
| * python3Packages.pytorch: Fix CMake files for split outputs•••Fixes #102146 Daniël de Kok2020-12-031-0/+9
| * utillinux: rename to util-linuxGraham Christensen2020-11-241-2/+2
| * Merge pull request #101917 from danieldk/pytorch-1.7.0•••python3Packages.pytorch: 1.6.0 -> 1.7.0Daniël de Kok2020-11-111-20/+26
| |\
| | * python3Packages.pytorch: 1.6.0 -> 1.7.0•••Changelog: https://github.com/pytorch/pytorch/releases/tag/v1.7.0 Daniël de Kok2020-10-281-20/+26
| * | python3Packages.pytorch-bin: 1.6.0 -> 1.7.0•••Changelog: https://github.com/pytorch/pytorch/releases/tag/v1.7.0 Daniël de Kok2020-10-282-7/+21
| |/
| * pytorch: update to use passthru's for checks•••This makes the checks a little easier to understand - we can just check it in the passthru. Also increase major check for cuda 11. Matthew Bauer2020-09-101-7/+3
| * pytorch-bin: init at 1.6.0Daniël de Kok2020-08-303-1/+81
| * python3Packages.pytorch: remove oneDNN dependency•••oneDNN was added as a dependency, but it is not actually used by PyTorch. PyTorch uses oneDNN from the vendored iDeep dependency. Using a system-provided oneDNN is currently not a supported build option. Daniël de Kok2020-08-261-5/+5
| * python3Packages.pytorch: fix library names in lib output on Darwin•••The library name fixup was attempted on the dev output, whereas it should be applied to the lib output. Daniël de Kok2020-08-081-14/+14
| * python3Packages.pytorch: fix Darwin build by disabling GCD•••PyTorch 1.6.0 has updated the vendored pthreadpool library, which has recently added support for Grand Central Dispatch. Unfortunately, it uses functionality (DISPATCH_APPLY_AUTO) that is only available since macOS 10.13, whereas we are still using 10.12 libraries. We can't directly pass through option to vendored libraries, since the setup.py scripts creates/filters the options that are passed to CMake. So, instead, this adds a small patch that disables the GCD functionality in pthreadpool. Daniël de Kok2020-08-082-0/+52
| * python3Packages.pytorch: 1.5.1 -> 1.6.0•••Changelog: https://github.com/pytorch/pytorch/releases/tag/v1.6.0 Daniël de Kok2020-07-311-3/+3
| * python3Packages.pytorch: create separate lib output•••At some point pytorch.dev was added to expose the libtorch headers and libraries to non-Python users of libtorch. However, this output currently has two disadvantages: 1. An application that compiles against the dev output will also have the libtorch header files in its closure. This is not so nice when e.g. building Docker images of applications that use libtorch. 2. The dev output has a large transitive closure with many dependencies that are not necessary when compiling against libtorch. This change adds the `lib` output so that applications that only link against libtorch libraries have a small closure. Before this change, the libtorch dependency adds 746MiB: % nix path-info -S `realpath result-dev` /nix/store/10rmy81bjk628sfpbj2szxlws6brq1xn-python3.8-pytorch-1.5.1-dev 782203848 With this change it is reduced to 196MiB: % nix path-info -S `realpath result-lib` /nix/store/bck65lf0z7gdhcf89w1zs5nz333lhgwa-python3.8-pytorch-1.5.1-lib 205865056 Daniël de Kok2020-07-121-2/+5
| * python3Packages.pytorch: fix AArch64 build•••aarch64-linux builds fail because of the use of opcodes in QNNPACK that the GNU assembpler does not support. This change cherry-picks an upstream patch that fixes this. See: https://github.com/pytorch/pytorch/issues/33124 https://github.com/pytorch/pytorch/pull/40584 Daniël de Kok2020-06-301-0/+15
| * python3Packages.pytorch: 1.5.0 -> 1.5.1•••Changelog: https://github.com/pytorch/pytorch/tree/v1.5.1 Daniël de Kok2020-06-301-2/+2
| * pythonPackages.pytorch: 1.4.1 -> 1.5.0•••Fixes previous bugs that required a patch Fixes CUDA build, see https://github.com/NixOS/nixpkgs/issues/89403 Guillaume Desforges2020-06-081-20/+2
| * python3Packages.pytorch: reference dnnl by new name oneDNNBenjamin Hipple2020-05-091-2/+2
| * python3Packages.pytorch: fixup builds with MKL•••- Pass `blas.provider` into `buildInputs`, so that CMake can find the actual `mkl` for inspection of its cmake files and headers. - Add `USE_MKL` correctly when the blas provider is `mkl`. - Use the MKLDNN and MKLDNN_CBLAS flags by default, since `mkldnn` is FOSS and always available.. - Remove a patch for MKL 2019, since we've moved to 2020. - Add a pythonImportsCheck for "torch" as a basic sanity-check - Removed some unused variables at the top of the file Benjamin Hipple2020-05-091-8/+17
| * pytorch: 1.2.0 -> 1.4.1•••Co-authored-by: Benjamin Hipple <bhipple@protonmail.com> Sam Stites2020-05-091-32/+67
* | Merge remote-tracking branch 'upstream/master' into structured-attrsRobin Gloster2020-04-151-1/+1
|\|
| * treewide: Per RFC45, remove all unquoted URLsMichael Reilly2020-04-101-1/+1
generated by cgit v1.2.3 (git 2.52.0) at 2026-08-20 11:17:04 +0000