Repository navigation
Package request: pytorch on darwin with GPU (MPS) support #243868
Description
Activity
Very cool! cc maintainers @tscholak @teh @thoughtpolice. Also cc @samuela.
curious if python3Packages.torch-bin works?
hmmm that's odd. aarch64-darwin is a supported platform
and there are srcs available here:
nixpkgs/pkgs/development/python-modules/torch/binary-hashes.nix
Lines 40 to 54 in 04f7903
what version of python are you using?
I made a
nixandnix-gpubranch at https://github.com/n8henrie/whisper;nix-gpuuses the wheel (and I've confirmed it enables the GPU to run whisper).However I'm not seeing any speed difference with a small test file (full of particularly witty content):
$ nix shell github:n8henrie/whisper/nix -c time whisper 20220922\ 084923.m4a /nix/store/0fyl0qi21ljiswg05qz5p2bpnl016k0l-python3.10-whisper/lib/python3.10/site-packages/whisper/transcribe.py:114: UserWarning: FP16 is not supported on CPU; using FP32 instead warnings.warn("FP16 is not supported on CPU; using FP32 instead") Detecting language using up to the first 30 seconds. Use `--language` to specify the language Detected language: English [00:00.000 --> 00:04.640] This is an example of voice recording. Something might actually say if I [00:04.640 --> 00:09.440] recording something on my watch. I think this is how it would turn out. 19.93user 11.08system 0:14.23elapsed 217%CPU (0avgtext+0avgdata 2062800maxresident)k 0inputs+0outputs (3246major+212332minor)pagefaults 0swaps $ $ nix shell github:n8henrie/whisper/nix-gpu -c time whisper 20220922\ 084923.m4a /nix/store/fvpkcahvngmsssm3yfchvgw11hdic6sx-python3.10-whisper/lib/python3.10/site-packages/whisper/transcribe.py:114: UserWarning: FP16 is not supported on CPU; using FP32 instead warnings.warn("FP16 is not supported on CPU; using FP32 instead") Detecting language using up to the first 30 seconds. Use `--language` to specify the language Detected language: English [00:00.000 --> 00:04.640] This is an example of voice recording. Something might actually say if I [00:04.640 --> 00:09.440] recording something on my watch. I think this is how it would turn out. 19.81user 11.16system 0:14.02elapsed 220%CPU (0avgtext+0avgdata 2177600maxresident)k 0inputs+0outputs (3955major+212007minor)pagefaults 0swaps
Not sure if that's going to be a training vs inference issue (or more likely whisper isn't configured to use the GPU, it has a
--deviceflag but seems to only look for cuda devices by default).Huh, looks like I just need to pass
--device mps(openai/whisper#382) but no luck:$ nix shell github:n8henrie/whisper/nix-gpu -c time whisper --device mps 20220922\ 084923.m4a Traceback (most recent call last): File "/nix/store/fvpkcahvngmsssm3yfchvgw11hdic6sx-python3.10-whisper/bin/.whisper-wrapped", line 9, in <module> sys.exit(cli()) File "/nix/store/fvpkcahvngmsssm3yfchvgw11hdic6sx-python3.10-whisper/lib/python3.10/site-packages/whisper/transcribe.py", line 444, in cli model = load_model(model_name, device=device, download_root=model_dir) File "/nix/store/fvpkcahvngmsssm3yfchvgw11hdic6sx-python3.10-whisper/lib/python3.10/site-packages/whisper/__init__.py", line 154, in load_model return model.to(device) File "/nix/store/0y2cidaighjvzwlw31k6fkcragba3jki-python3.10-torch-2.0.1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1145, in to return self._apply(convert) File "/nix/store/0y2cidaighjvzwlw31k6fkcragba3jki-python3.10-torch-2.0.1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 844, in _apply self._buffers[key] = fn(buf) File "/nix/store/0y2cidaighjvzwlw31k6fkcragba3jki-python3.10-torch-2.0.1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1143, in convert return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None, non_blocking) NotImplementedError: Could not run 'aten::empty.memory_format' with arguments from the 'SparseMPS' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::empty.memory_format' is only available for these backends: [CPU, MPS, Meta, QuantizedCPU, QuantizedMeta, MkldnnCPU, SparseCPU, SparseMeta, SparseCsrCPU, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMeta, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PythonDispatcher]. CPU: registered at /Users/runner/work/pytorch/pytorch/pytorch/build/aten/src/ATen/RegisterCPU.cpp:31034 [kernel] MPS: registered at /Users/runner/work/pytorch/pytorch/pytorch/build/aten/src/ATen/RegisterMPS.cpp:22748 [kernel] Meta: registered at /dev/null:241 [kernel] QuantizedCPU: registered at /Users/runner/work/pytorch/pytorch/pytorch/build/aten/src/ATen/RegisterQuantizedCPU.cpp:929 [kernel] QuantizedMeta: registered at /Users/runner/work/pytorch/pytorch/pytorch/build/aten/src/ATen/RegisterQuantizedMeta.cpp:105 [kernel] MkldnnCPU: registered at /Users/runner/work/pytorch/pytorch/pytorch/build/aten/src/ATen/RegisterMkldnnCPU.cpp:507 [kernel] SparseCPU: registered at /Users/runner/work/pytorch/pytorch/pytorch/build/aten/src/ATen/RegisterSparseCPU.cpp:1379 [kernel] SparseMeta: registered at /Users/runner/work/pytorch/pytorch/pytorch/build/aten/src/ATen/RegisterSparseMeta.cpp:249 [kernel] SparseCsrCPU: registered at /Users/runner/work/pytorch/pytorch/pytorch/build/aten/src/ATen/RegisterSparseCsrCPU.cpp:1128 [kernel] BackendSelect: registered at /Users/runner/work/pytorch/pytorch/pytorch/build/aten/src/ATen/RegisterBackendSelect.cpp:726 [kernel] Python: registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/core/PythonFallbackKernel.cpp:144 [backend fallback] FuncTorchDynamicLayerBackMode: registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/functorch/DynamicLayer.cpp:491 [backend fallback] Functionalize: registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/FunctionalizeFallbackKernel.cpp:280 [backend fallback] Named: registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/core/NamedRegistrations.cpp:7 [backend fallback] Conjugate: fallthrough registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/ConjugateFallback.cpp:21 [kernel] Negative: fallthrough registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/native/NegateFallback.cpp:23 [kernel] ZeroTensor: fallthrough registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/ZeroTensorFallback.cpp:90 [kernel] ADInplaceOrView: fallthrough registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/core/VariableFallbackKernel.cpp:63 [backend fallback] AutogradOther: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradCPU: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradCUDA: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradHIP: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradXLA: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradMPS: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradIPU: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradXPU: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradHPU: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradVE: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradLazy: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradMeta: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradMTIA: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradPrivateUse1: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradPrivateUse2: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradPrivateUse3: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] AutogradNestedTensor: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/VariableType_2.cpp:17484 [autograd kernel] Tracer: registered at /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/autograd/generated/TraceType_2.cpp:16726 [kernel] AutocastCPU: fallthrough registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/autocast_mode.cpp:487 [backend fallback] AutocastCUDA: fallthrough registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/autocast_mode.cpp:354 [backend fallback] FuncTorchBatched: registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/functorch/LegacyBatchingRegistrations.cpp:815 [backend fallback] FuncTorchVmapMode: fallthrough registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/functorch/VmapModeRegistrations.cpp:28 [backend fallback] Batched: registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/LegacyBatchingRegistrations.cpp:1073 [backend fallback] VmapMode: fallthrough registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/VmapModeRegistrations.cpp:33 [backend fallback] FuncTorchGradWrapper: registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/functorch/TensorWrapper.cpp:210 [backend fallback] PythonTLSSnapshot: registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/core/PythonFallbackKernel.cpp:152 [backend fallback] FuncTorchDynamicLayerFrontMode: registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/functorch/DynamicLayer.cpp:487 [backend fallback] PythonDispatcher: registered at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/core/PythonFallbackKernel.cpp:148 [backend fallback] Command exited with non-zero status 1 3.87user 1.23system 0:03.39elapsed 150%CPU (0avgtext+0avgdata 1210224maxresident)k 0inputs+0outputs (19major+112713minor)pagefaults 0swaps
timing based benchmarks are tricky to debug... does
import torch; print(torch.backends.mps.is_available())work for you in this flake? also ooc what's the equivalent tonvtopfor the MPS backend?NotImplementedError: Could not run 'aten::empty.memory_format' with arguments from the 'SparseMPS' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::empty.memory_format' is only available for these backends: [CPU, MPS, Meta, QuantizedCPU, QuantizedMeta, MkldnnCPU, SparseCPU, SparseMeta, SparseCsrCPU, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMeta, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PythonDispatcher].Yeah this error looks like it's an issue with whisper as opposed to nix's packaging of torch/torch-bin.
We can leave this open as a tracking issue for getting MPS support in our source build of torch however.
Reacted by Nathan Henriewhat version of python are you using?
Tried 39-311. I'm pinned to 23.05, I wonder if it's different in unstable?
Yup, that was it!
$ nix-shell -I nixpkgs=channel:nixpkgs-unstable -p 'python39.withPackages (ps: with ps; [ pytorch-bin ])' --command "python -c 'import torch; print(torch.backends.mps.is_available())'" True
import torch; print(torch.backends.mps.is_available())
@samuela yes, I just patched this into the whisper code (in the
cli()function) and confirmed mynixbranch showsFalseand mynix-gpubranch shows the device as expected.Reacted by Samuel Ainsworthyup, that makes for two TODOs
- Add MPS support to the source build
python3Packages.torch - Add python 3.11, 3.12 support to torch-bin
- Add MPS support to the source build
I can confirm that MPS support works with torch-bin and torchvision-bin now
I'm still trying to get it working. Tried adding the
USE_MPSflag but it still results without MPS support; I think the build script usesxcrunto sort out some SDK path business, which fails to find the MPSGraph framework, and disables MPS:bash: line 1: xcrun: command not found -- MPS: unable to get MacOS sdk version -- MPSGraph framework not foundhttps://github.com/pytorch/pytorch/blob/1da41157028ee8224e456f6fab18bc22fa2637fe/CMakeLists.txt#L99
Reacted by Mika TammiCan any of the aarch64-darwin folks help me build this locally via
nix develop?Currently builds fine with
nix build --offline(so I don't think it's a substituter issue) andnix shell.Steps I'm taking / have tried:
Attempt 1
nix develop .#python310Packages.pytorchsource $stdenv/setupgenericBuild
It looks like
/usr/bin/xcrunbleeds into the environment, setsUSE_MPSto true, and it fails to find the relevant SDK libraries (fixing this would hopefully help me close this issue, but I'd like to at least get it to build first).Attempt 2
nix develop -i .#python310Packages.pytorch --command bash --norc[link]source $stdenv/setupgenericBuild
USE_MPSis set to false, but the build fails looking for#include <google/protobuf/implicit_weak_message.h>After reading this issue, it looks like some of these are functions and some are variables, so I might need to e.g.
eval "${buildPhase}"instead of justbuildPhase.Attempt 3
nix develop -i .#python310Packages.pytorch --command bash --norcsource $stdenv/setupunpackPhaseeval $patchPhaseeval $preConfigurePhaseseval $configurePhasecd sourceeval $buildPhase
Here it fails with the same errors about protobuf.
I think these are relevant convos (for my future reference):
- nix develop --build and phases confusion nix#6202
- https://discourse.nixos.org/t/using-nix-develop-to-debug-builds/26024/4
- https://discourse.nixos.org/t/differences-between-nix-build-and-development-shell/24060
- https://discourse.nixos.org/t/unable-to-build-package-in-nix-shell-or-nix-develop-but-nix-build-or-nix-build-works-fine/15500
Can anyone point out what I'm getting wrong here?
nix developshould be able to build it ifnix buildcan, right? TIA for any help for a noob.2 remaining items
hmmm interesting... what's the process for adding a new
apple_sdkversion? cc @NixOS/darwin-maintainersPeople looking for a work-around might find this flake helpful. It's based on @n8henrie's wheel approach.
nix run "github:david-r-cox/pytorch-darwin-env#verificationScript"After #346043 lands, this should be doable. Just add
apple-sdk_13as a build input.Reacted by Samuel AinsworthI tried to enable this with:
diff --git a/pkgs/development/python-modules/torch/default.nix b/pkgs/development/python-modules/torch/default.nix index cd43c04d67d4..e7bbf4eb1b31 100644 --- a/pkgs/development/python-modules/torch/default.nix +++ b/pkgs/development/python-modules/torch/default.nix @@ -364,10 +364,14 @@ buildPythonPackage rec { # NB technical debt: building without NNPACK as workaround for missing `six` USE_NNPACK = 0; + # USE_PYTORCH_METAL = stdenv.hostPlatform.isDarwin; + USE_MPS = stdenv.hostPlatform.isDarwin; + cmakeFlags = [ # (lib.cmakeBool "CMAKE_FIND_DEBUG_MODE" true) (lib.cmakeFeature "CUDAToolkit_VERSION" cudaPackages.cudaVersion) + (lib.cmakeBool "USE_MPS" stdenv.hostPlatform.isDarwin) ] ++ lib.optionals cudaSupport [ # Unbreaks version discovery in enable_language(CUDA) when wrapping nvcc with ccache @@ -519,8 +523,12 @@ buildPythonPackage rec { ++ lib.optionals (cudaSupport || rocmSupport) [ effectiveMagma ] ++ lib.optionals stdenv.hostPlatform.isLinux [ numactl ] ++ lib.optionals stdenv.hostPlatform.isDarwin [ - darwin.apple_sdk.frameworks.Accelerate - darwin.apple_sdk.frameworks.CoreServices + darwin.apple_sdk_12_3.frameworks.Accelerate + darwin.apple_sdk_12_3.frameworks.CoreServices + darwin.apple_sdk_12_3.frameworks.Metal + darwin.apple_sdk_12_3.frameworks.MetalKit + darwin.apple_sdk_12_3.frameworks.MetalPerformanceShaders + darwin.apple_sdk_12_3.frameworks.MetalPerformanceShadersGraph darwin.libobjc ] ++ lib.optionals tritonSupport [ _tritonEffective ]I'm still getting the xcrun error:
python3.12-torch> bash: line 1: xcrun: command not found python3.12-torch> -- MPS: unable to get MacOS sdk version python3.12-torch> -- MPSGraph framework not foundI was waiting on @reckenrode's recent massive SDK refactor to land (at least in master, still staging as of this AM) and will retry then.
A couple sites helpful for tracking -- the best known is https://nixpk.gs/pr-tracker.html?pr=346043
(I really wish we could get an RSS feed for when user-specified PRs land in a specific branch.)
It’s in
staging-next, you can just target a PR to that. Many things have already been built.I rebased my hack on top of staging-next, now I got different errors:
-- CLANG_VERSION_STRING: clang version 16.0.6 Target: arm64-apple-darwin Thread model: posix InstalledDir: /nix/store/r0w268903vjlq1vrnajb3lvll4hs5z8g-clang-16.0.6/bin warning: unhandled Target key DefaultVariant warning: unhandled Target key SupportedTargets warning: unhandled Target key VersionMap warning: unhandled Target key Variants warning: unhandled Target key DebuggerOptions warning: unhandled Product key iOSSupportVersion warning: unhandled Target key DefaultVariant warning: unhandled Target key SupportedTargets warning: unhandled Target key VersionMap warning: unhandled Target key Variants warning: unhandled Target key DebuggerOptions warning: unhandled Product key iOSSupportVersion warning: unhandled Target key DefaultVariant warning: unhandled Target key SupportedTargets warning: unhandled Target key VersionMap warning: unhandled Target key Variants warning: unhandled Target key DebuggerOptions warning: unhandled Product key iOSSupportVersion -- sdk version: 11.3, mps supported: OFF warning: unhandled Target key DefaultVariant warning: unhandled Target key SupportedTargets warning: unhandled Target key VersionMap warning: unhandled Target key Variants warning: unhandled Target key DebuggerOptions warning: unhandled Product key iOSSupportVersion warning: unhandled Target key DefaultVariant warning: unhandled Target key SupportedTargets warning: unhandled Target key VersionMap warning: unhandled Target key Variants warning: unhandled Target key DebuggerOptions warning: unhandled Product key iOSSupportVersion warning: unhandled Target key DefaultVariant warning: unhandled Target key SupportedTargets warning: unhandled Target key VersionMap warning: unhandled Target key Variants warning: unhandled Target key DebuggerOptions warning: unhandled Product key iOSSupportVersion -- MPSGraph framework not found -- Could not find ccache. Consider installing ccache to speed up compilation.It now says SDK version is 11.3, maybe I need to change something else as well in that package, as I was trying to specify I want version 12.3
This was one of my test cases. Add
apple-sdk_13as abuildInput. Delete the old frameworks frombuildInputs. That should be it.It has to be the 13.3 SDK because MPS isn’t available in older SDKs, and the version of pytorch in nixpkgs does not build with the 14.4 (or presumably 15.0) SDK.
Edit: Apple’s article on MPS in PyTorch suggests the 12.3 SDK should work, so I could be wrong about the minimum supported SDK, but the requirements may have changed since it was posted. You can try the 12.3 SDK by adding
apple-sdk_12instead tobuildInputs.It now says SDK version is 11.3, maybe I need to change something else as well in that package, as I was trying to specify I want version 12.3
After the update, the old SDK frameworks are stubs that do nothing on their own . If you used
overrideSDK stdenv "12.3", you would get the 12.3 SDK, but using the old pattern is deprecated. After a tree-wide after 24.11, it will be moved todarwin-aliases.nixand warn on use.Reacted by Mika Tammi, Nathan Henrie and Joe GoldinLooks like success!
~/nixpkgs.git/ > nix-shell -p 'python3.withPackages (ps: with ps; [ torch ])' -I nixpkgs=. [nix-shell:~/nixpkgs.git]$ python Python 3.12.7 (main, Oct 1 2024, 02:05:46) [Clang 16.0.6 ] on darwin Type "help", "copyright", "credits" or "license" for more information. >>> import torch >>> print(f"PyTorch version: {torch.__version__}") PyTorch version: 2.5.0 >>> print(f"MPS available: {torch.backends.mps.is_available()}") MPS available: TrueReacted by Nathan HenrieReacted by Randy EckenrodeJust for tests I also tried to use apple-sdk_12 but I got these errors:
python3.12-torch> return MTLLanguageVersion3_0; python3.12-torch> ^~~~~~~~~~~~~~~~~~~~~ python3.12-torch> MTLLanguageVersion2_0 python3.12-torch> /nix/store/2vs624s5vsdnhralhwlin5gnid2dl1qy-apple-sdk-12.3/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/System/Library/Frameworks/Metal.framework/Headers/MTLLibrary.h:185:5: note: 'MTLLanguageVersion2_0' declared here python3.12-torch> MTLLanguageVersion2_0 API_AVAILABLE(macos(10.13), ios(11.0)) = (2 << 16), python3.12-torch> ^ python3.12-torch> 1 error generated. python3.12-torch> [2278/5844] Building CXX object caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/output-archive.cpp.o python3.12-torch> ninja: build stopped: subcommand failed. python3.12-torch> ERROR Backend subprocess exited when trying to invoke build_wheel error: builder for '/nix/store/n41ysm7qfk9azq8w09vs5l8iwvmwqvr3-python3.12-torch-2.5.0.drv' failed with exit code 1; last 10 log lines: > ^~~~~~~~~~~~~~~~~~~~~ > MTLLanguageVersion2_0 > /nix/store/2vs624s5vsdnhralhwlin5gnid2dl1qy-apple-sdk-12.3/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/System/Library/Frameworks/Metal.framework/Headers/MTLLibrary.h:185:5: note: 'MTLLanguageVersion2_0' declared here > MTLLanguageVersion2_0 API_AVAILABLE(macos(10.13), ios(11.0)) = (2 << 16), > ^ > 1 error generated. > [2278/5844] Building CXX object caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/output-archive.cpp.o > ninja: build stopped: subcommand failed. > > ERROR Backend subprocess exited when trying to invoke build_wheel For full logs, run 'nix log /nix/store/n41ysm7qfk9azq8w09vs5l8iwvmwqvr3-python3.12-torch-2.5.0.drv'.So I guess 13 is way to go
Reacted by Randy Eckenrode
Project description
Not sure if this is a "packaging request" but didn't seem like a bug report, just wanted a tracking issue, I might try to hack on this eventually.
Pytorch on aarch64-darwin now supports GPU:
https://pytorch.org/docs/stable/notes/mps.html
The currently packaged pytorch runs but doesn't use my Mac's GPU:
In contrast, using the pytorch-provided wheel:
Metadata
cc @NixOS/darwin-maintainers