Crash on Mac M1/M2 chip when using MPS support #55
Comments
|
yea, mps doesn't have lerp yet, i'll build in some fallback later this week |
|
@liangcl0928 do you want to try 1.6.12 and see if that fixes the issue? |
|
@liangcl0928 just recalled that i have a Mac M1 lol, so i was able to resolve it |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
My environment:
Mac OS 13.2.1 with M2 Pro
Python 3.9.16
pytorch: 2.0.1
vector_quantize_pytorch: 1.6.11
site-packages/vector_quantize_pytorch/vector_quantize_pytorch.py:444: UserWarning: The operator 'aten::lerp.Scalar_out' is not currently supported on the MPS backend and will fall back to run on the CPU. This may have performance implications. (Triggered internally at /Users/runner/work/temp/anaconda/conda-bld/pytorch_1682343673238/work/aten/src/ATen/mps/MPSFallback.mm:11.)
self.cluster_size.data.lerp(cluster_size, 1 - self.decay)
libc++abi: terminating with uncaught exception of type c10::TypeError: Trying to convert ComplexFloat to the MPS backend but it does not have support for that dtype.
Exception raised from getMPSScalarType at /Users/runner/work/_temp/anaconda/conda-bld/pytorch_1682343673238/work/aten/src/ATen/native/mps/OperationUtils.mm:91 (most recent call first):
frame #0: at::native::mps::getMPSScalarType(c10::ScalarType) + 180 (0x1369364ac in libtorch_cpu.dylib)
frame #1: at::native::mps::mpsGraphRankedPlaceHolder(MPSGraph*, at::Tensor const&) + 96 (0x136938654 in libtorch_cpu.dylib)
frame #2: invocation function for block in at::native::mps::unary_op(at::Tensor const&, at::Tensor const&, std::__1::basic_string<char, std::__1::char_traits, std::__1::allocator >, MPSGraphTensor* (MPSGraph*, MPSGraphTensor*) block_pointer, std::__1::function<bool (at::Tensor const&)>) + 104 (0x136a00504 in libtorch_cpu.dylib)
frame #3: invocation function for block in at::native::mps::MPSGraphCache::CreateCachedGraph(std::__1::basic_string<char, std::__1::char_traits, std::__1::allocator > const&, at::native::mps::MPSCachedGraph* () block_pointer) + 216 (0x13694c58c in libtorch_cpu.dylib)
frame #4: _dispatch_client_callout + 20 (0x199392504 in libdispatch.dylib)
frame #5: _dispatch_lane_barrier_sync_invoke_and_complete + 56 (0x1993a1a9c in libdispatch.dylib)
frame #6: at::native::mps::MPSGraphCache::CreateCachedGraph(std::__1::basic_string<char, std::__1::char_traits, std::__1::allocator > const&, at::native::mps::MPSCachedGraph* () block_pointer) + 160 (0x13693a5d0 in libtorch_cpu.dylib)
frame #7: at::native::mps::unary_op(at::Tensor const&, at::Tensor const&, std::__1::basic_string<char, std::__1::char_traits, std::__1::allocator >, MPSGraphTensor* (MPSGraph*, MPSGraphTensor*) block_pointer, std::_1::function<bool (at::Tensor const&)>) + 860 (0x1369fff9c in libtorch_cpu.dylib)
frame #8: at::native::abs_out_mps(at::Tensor const&, at::Tensor&) + 124 (0x136a02bd0 in libtorch_cpu.dylib)
frame #9: at::ops::abs_out::call(at::Tensor const&, at::Tensor&) + 276 (0x1330e54a8 in libtorch_cpu.dylib)
frame #10: at::native::abs(at::Tensor const&) + 232 (0x132adba88 in libtorch_cpu.dylib)
frame #11: c10::impl::wrap_kernel_functor_unboxed<c10::impl::detail::WrapFunctionIntoFunctor<c10::CompileTimeFunctionPointer<at::Tensor (c10::DispatchKeySet, at::Tensor const&), &(torch::autograd::VariableType::(anonymous namespace)::abs(c10::DispatchKeySet, at::Tensor const&))>, at::Tensor, c10::guts::typelist::typelist<c10::DispatchKeySet, at::Tensor const&> >, at::Tensor (c10::DispatchKeySet, at::Tensor const&)>::call(c10::OperatorKernel*, c10::DispatchKeySet, at::Tensor const&) + 920 (0x134e13950 in libtorch_cpu.dylib)
frame #12: at::_ops::abs::call(at::Tensor const&) + 264 (0x1330e4900 in libtorch_cpu.dylib)
frame #13: torch::autograd::THPVariable_abs(_object*, _object*) + 188 (0x104b006d4 in libtorch_python.dylib)
frame #14: method_vectorcall_NOARGS + 172 (0x102a964e4 in python3.9)
frame #15: call_function + 516 (0x102b793cc in python3.9)
frame #16: _PyEval_EvalFrameDefault + 26296 (0x102b75ba8 in python3.9)
frame #17: _PyEval_EvalCode + 2804 (0x102b6ee98 in python3.9)
frame #18: _PyFunction_Vectorcall + 220 (0x102a89fe4 in python3.9)
frame #19: call_function + 516 (0x102b793cc in python3.9)
frame #20: _PyEval_EvalFrameDefault + 26332 (0x102b75bcc in python3.9)
frame #21: function_code_fastcall + 116 (0x102a8a0a0 in python3.9)
frame #22: method_vectorcall + 516 (0x102a8cf24 in python3.9)
frame #23: _PyEval_EvalFrameDefault + 27140 (0x102b75ef4 in python3.9)
frame #24: _PyEval_EvalCode + 2804 (0x102b6ee98 in python3.9)
frame #25: _PyFunction_Vectorcall + 220 (0x102a89fe4 in python3.9)
frame #26: _PyObject_FastCallDictTstate + 320 (0x102a89674 in python3.9)
frame #27: _PyObject_Call_Prepend + 164 (0x102a8a464 in python3.9)
frame #28: slot_tp_call + 116 (0x102afa5b0 in python3.9)
frame #29: _PyObject_MakeTpCall + 616 (0x102a893c4 in python3.9)
frame #30: call_function + 668 (0x102b79464 in python3.9)
frame #31: _PyEval_EvalFrameDefault + 26332 (0x102b75bcc in python3.9)
frame #32: function_code_fastcall + 116 (0x102a8a0a0 in python3.9)
frame #33: method_vectorcall + 516 (0x102a8cf24 in python3.9)
frame #34: _PyEval_EvalFrameDefault + 27140 (0x102b75ef4 in python3.9)
frame #35: _PyEval_EvalCode + 2804 (0x102b6ee98 in python3.9)
frame #36: _PyFunction_Vectorcall + 220 (0x102a89fe4 in python3.9)
frame #37: _PyObject_FastCallDictTstate + 320 (0x102a89674 in python3.9)
frame #38: _PyObject_Call_Prepend + 164 (0x102a8a464 in python3.9)
frame #39: slot_tp_call + 116 (0x102afa5b0 in python3.9)
frame #40: _PyObject_MakeTpCall + 616 (0x102a893c4 in python3.9)
frame #41: call_function + 668 (0x102b79464 in python3.9)
frame #42: _PyEval_EvalFrameDefault + 26456 (0x102b75c48 in python3.9)
frame #43: _PyEval_EvalCode + 2804 (0x102b6ee98 in python3.9)
frame #44: _PyFunction_Vectorcall + 220 (0x102a89fe4 in python3.9)
frame #45: call_function + 516 (0x102b793cc in python3.9)
frame #46: _PyEval_EvalFrameDefault + 26456 (0x102b75c48 in python3.9)
frame #47: _PyEval_EvalCode + 2804 (0x102b6ee98 in python3.9)
frame #48: _PyFunction_Vectorcall + 220 (0x102a89fe4 in python3.9)
frame #49: call_function + 516 (0x102b793cc in python3.9)
frame #50: _PyEval_EvalFrameDefault + 26456 (0x102b75c48 in python3.9)
frame #51: _PyEval_EvalCode + 2804 (0x102b6ee98 in python3.9)
frame #52: _PyFunction_Vectorcall + 220 (0x102a89fe4 in python3.9)
frame #53: call_function + 516 (0x102b793cc in python3.9)
frame #54: _PyEval_EvalFrameDefault + 26456 (0x102b75c48 in python3.9)
frame #55: _PyEval_EvalCode + 2804 (0x102b6ee98 in python3.9)
frame #56: run_mod + 216 (0x102bc8bdc in python3.9)
frame #57: pyrun_file + 264 (0x102bc6728 in python3.9)
frame #58: PyRun_SimpleFileExFlags + 1332 (0x102bc5f50 in python3.9)
frame #59: Py_RunMain + 2148 (0x102be88c0 in python3.9)
frame #60: pymain_main + 1252 (0x102be9a40 in python3.9)
frame #61: main + 56 (0x102a3c770 in python3.9)
frame #62: start + 2544 (0x1991efe50 in dyld)
The text was updated successfully, but these errors were encountered: