Date: Fri, 21 Apr 2023 12:18:45 +0200 From: Mario Marietto <marietto2008@gmail.com> To: Aryeh Friedman <aryeh.friedman@gmail.com> Cc: FreeBSD Mailing List <freebsd-hackers@freebsd.org>, FreeBSD Mailing List <freebsd-questions@freebsd.org>, Yuri Victorovich <yuri@freebsd.org> Subject: Re: Installing openAI's GPT-2 Ada AI Language Model Message-ID: <CA%2B1FSijsSSpCFeKeaOt4gR36BAZ4J8j4QSJRJa-VF-a=J9e2uw@mail.gmail.com> In-Reply-To: <CAGBxaXkhC--ZppimDFabEwPhesjAJmrziNZm753eoyjy1sWzqg@mail.gmail.com> References: <CAGBxaXmhRLk9Lx_ZHeRdoN-K2fRLEhY3cBVtBymmAjd4bBh1OQ@mail.gmail.com> <CA%2B1FSihQ-f4uhiOjYH8Wo=AxFEkAKe3NRDJdopgT50J=_jY4fA@mail.gmail.com> <CAGBxaXnYojzQJqO62hkzUJvD2rzaNp%2Bem38FgCqVSBu%2BmkBi9A@mail.gmail.com> <CA%2B1FSijpiko%2B%2B%2BwJuXo2GVV6sz3yGVi7ig0X3037%2B1zE3n91hg@mail.gmail.com> <CAGBxaX=OcaHEZk3S7jQeYW64A_iRNTmJ%2Bab4U7h_hsrG%2BQqQPg@mail.gmail.com> <ZEEnZjzDCtR_ZG4P@graf.pompo.net> <CAGBxaXmU=Ja9EkoMyxQ0cNxYB4BeiktqQ3P64QcWg%2B=xijTiyQ@mail.gmail.com> <CA%2B1FSii6OOwi%2B%2Bau-_9ViU_SMZ%2BGbESG5H0McVTHQUwmMnOJGQ@mail.gmail.com> <CAGBxaXkhC--ZppimDFabEwPhesjAJmrziNZm753eoyjy1sWzqg@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
[-- Attachment #1 --] Can't you install pytorch using the linux miniconda installer like below ? # fetch https://gist.githubusercontent.com/shkhln/40ef290463e78fb2b0000c60f4ad797e/raw/f640983249607e38af405c95c457ce4afc85c608/uvm_ioctl_override.c # /compat/ubuntu/bin/gcc --sysroot=/compat/ubuntu -m64 -std=c99 -Wall -ldl -fPIC -shared -o dummy-uvm.so uvm_ioctl_override.c # pkg install linux-miniconda-installer # miniconda-installer # bash # source /home/marietto/miniconda3/etc/profile.d/conda.sh # conda activate (base) # conda activate pytorch On Fri, Apr 21, 2023 at 2:38 AM Aryeh Friedman <aryeh.friedman@gmail.com> wrote: > On Thu, Apr 20, 2023 at 12:24 PM Mario Marietto <marietto2008@gmail.com> > wrote: > > > > try to copy and paste the commands that you have issued on pastebin...i > need to understand the scenario > > After saving the patch from the bug report to PORT/files and running > portmaster -P misc/pytourch (brand new machine except for installing > portmaster): > > c/ATen/UfuncCPUKernel_add.cpp.AVX2.cpp.o -c > > /usr/ports/misc/pytorch/work/.build/aten/src/ATen/UfuncCPUKernel_add.cpp.AVX2.cpp > In file included from > > /usr/ports/misc/pytorch/work/.build/aten/src/ATen/UfuncCPUKernel_add.cpp.AVX2.cpp:1: > In file included from > /usr/ports/misc/pytorch/work/.build/aten/src/ATen/UfuncCPUKernel_add.cpp:3: > In file included from > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/native/ufunc/add.h:6: > In file included from > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/functional.h:3: > In file included from > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/functional_base.h:6: > In file included from > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec.h:6: > In file included from > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256.h:12: > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:253:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_acosf8_u10); > ^~~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:256:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_asinf8_u10); > ^~~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:259:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_atanf8_u10); > ^~~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:280:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_erff8_u10); > ^~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:283:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_erfcf8_u15); > ^~~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:300:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_expf8_u10); > ^~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:303:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_expm1f8_u10); > ^~~~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:393:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_logf8_u10); > ^~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:396:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_log2f8_u10); > ^~~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:399:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_log10f8_u10); > ^~~~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:402:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_log1pf8_u10); > ^~~~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:406:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_sinf8_u10); > ^~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:409:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_sinhf8_u10); > ^~~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:412:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_cosf8_u10); > ^~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:415:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_coshf8_u10); > ^~~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:447:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_tanf8_u10); > ^~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:450:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_tanhf8_u10); > ^~~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:460:16: > error: cannot initialize a parameter of type 'const __m256 > (*)(__m256)' with an lvalue of type '__m256 (__m256)': different > return type ('const __m256' (vector of 8 'float' values) vs '__m256' > (vector of 8 'float' values)) > return map(Sleef_lgammaf8_u10); > ^~~~~~~~~~~~~~~~~~ > > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49: > note: passing argument to parameter 'vop' here > Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const { > ^ > 18 errors generated. > [ 80% 1035/1283] /usr/bin/c++ -DAT_PER_OPERATOR_HEADERS > -DCPUINFO_SUPPORTED_PLATFORM=0 -DFMT_HEADER_ONLY=1 > -DHAVE_MALLOC_USABLE_SIZE=1 -DHAVE_MMAP=1 -DHAVE_SHM_OPEN=1 > -DHAVE_SHM_UNLINK=1 -DMINIZ_DISABLE_ZIP_READER_CRC32_CHECKS > -DONNXIFI_ENABLE_EXT=1 -DONNX_ML=1 -DONNX_NAMESPACE=onnx > -DUSE_EXTERNAL_MZCRC -D_FILE_OFFSET_BITS=64 -Dtorch_cpu_EXPORTS > -I/usr/ports/misc/pytorch/work/.build/aten/src > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src > -I/usr/ports/misc/pytorch/work/.build > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1 > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/foxi > -I/usr/ports/misc/pytorch/work/.build/third_party/foxi > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc/api > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc/api/include > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/caffe2/aten/src/TH > -I/usr/ports/misc/pytorch/work/.build/caffe2/aten/src/TH > -I/usr/ports/misc/pytorch/work/.build/caffe2/aten/src > -I/usr/ports/misc/pytorch/work/.build/caffe2/../aten/src > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/miniz-2.1.0 > > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/kineto/libkineto/include > > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/kineto/libkineto/src > > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/../third_party/catch/single_include > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/.. > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/c10/.. > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/cpuinfo/include > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/FP16/include > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/fmt/include > > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/flatbuffers/include > -isystem > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/cmake/../third_party/eigen > -isystem /usr/ports/misc/pytorch/work/pytorch-v1.13.1/caffe2 -O2 -pipe > -fstack-protector-strong -isystem /usr/local/include > -fno-strict-aliasing -isystem /usr/local/include -Wno-deprecated > -fvisibility-inlines-hidden -fopenmp=libomp -DNDEBUG -DUSE_KINETO > -DLIBKINETO_NOCUPTI -DSYMBOLICATE_MOBILE_DEBUG_HANDLE > -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra > -Werror=return-type -Werror=non-virtual-dtor > -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds > -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter > -Wno-unused-function -Wno-unused-result -Wno-strict-overflow > -Wno-strict-aliasing -Wno-error=deprecated-declarations > -Wvla-extension -Wno-range-loop-analysis -Wno-pass-failed > -Wno-error=pedantic -Wno-error=redundant-decls > -Wno-error=old-style-cast -Wconstant-conversion > -Wno-invalid-partial-specialization -Wno-typedef-redefinition > -Wno-unused-private-field -Wno-inconsistent-missing-override > -Wno-c++14-extensions -Wno-constexpr-not-const -Wno-missing-braces > -Wunused-lambda-capture -Wunused-local-typedef -Qunused-arguments > -fcolor-diagnostics -fdiagnostics-color=always > -Wno-unused-but-set-variable -fno-math-errno -fno-trapping-math > -Werror=format -Werror=cast-function-type -DHAVE_AVX512_CPU_DEFINITION > -DHAVE_AVX2_CPU_DEFINITION -O2 -pipe -fstack-protector-strong -isystem > /usr/local/include -fno-strict-aliasing -isystem /usr/local/include > -DNDEBUG -DNDEBUG -std=gnu++14 -fPIC -DTH_HAVE_THREAD -Wall -Wextra > -Wno-unused-parameter -Wno-unused-function -Wno-unused-result > -Wno-missing-field-initializers -Wno-write-strings > -Wno-unknown-pragmas -Wno-type-limits -Wno-array-bounds > -Wno-sign-compare -Wno-strict-overflow -Wno-strict-aliasing > -Wno-error=deprecated-declarations -Wno-missing-braces > -Wno-range-loop-analysis -fvisibility=hidden -O2 -fopenmp=libomp > -DCAFFE2_BUILD_MAIN_LIB -pthread -MD -MT > > caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/input-archive.cpp.o > -MF > caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/input-archive.cpp.o.d > -o > caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/input-archive.cpp.o > -c > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc/api/src/serialize/input-archive.cpp > [ 80% 1035/1283] /usr/bin/c++ -DAT_PER_OPERATOR_HEADERS > -DCPUINFO_SUPPORTED_PLATFORM=0 -DFMT_HEADER_ONLY=1 > -DHAVE_MALLOC_USABLE_SIZE=1 -DHAVE_MMAP=1 -DHAVE_SHM_OPEN=1 > -DHAVE_SHM_UNLINK=1 -DMINIZ_DISABLE_ZIP_READER_CRC32_CHECKS > -DONNXIFI_ENABLE_EXT=1 -DONNX_ML=1 -DONNX_NAMESPACE=onnx > -DUSE_EXTERNAL_MZCRC -D_FILE_OFFSET_BITS=64 -Dtorch_cpu_EXPORTS > -I/usr/ports/misc/pytorch/work/.build/aten/src > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src > -I/usr/ports/misc/pytorch/work/.build > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1 > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/foxi > -I/usr/ports/misc/pytorch/work/.build/third_party/foxi > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc/api > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc/api/include > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/caffe2/aten/src/TH > -I/usr/ports/misc/pytorch/work/.build/caffe2/aten/src/TH > -I/usr/ports/misc/pytorch/work/.build/caffe2/aten/src > -I/usr/ports/misc/pytorch/work/.build/caffe2/../aten/src > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/miniz-2.1.0 > > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/kineto/libkineto/include > > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/kineto/libkineto/src > > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/../third_party/catch/single_include > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/.. > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/c10/.. > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/cpuinfo/include > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/FP16/include > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/fmt/include > > -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/flatbuffers/include > -isystem > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/cmake/../third_party/eigen > -isystem /usr/ports/misc/pytorch/work/pytorch-v1.13.1/caffe2 -O2 -pipe > -fstack-protector-strong -isystem /usr/local/include > -fno-strict-aliasing -isystem /usr/local/include -Wno-deprecated > -fvisibility-inlines-hidden -fopenmp=libomp -DNDEBUG -DUSE_KINETO > -DLIBKINETO_NOCUPTI -DSYMBOLICATE_MOBILE_DEBUG_HANDLE > -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra > -Werror=return-type -Werror=non-virtual-dtor > -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds > -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter > -Wno-unused-function -Wno-unused-result -Wno-strict-overflow > -Wno-strict-aliasing -Wno-error=deprecated-declarations > -Wvla-extension -Wno-range-loop-analysis -Wno-pass-failed > -Wno-error=pedantic -Wno-error=redundant-decls > -Wno-error=old-style-cast -Wconstant-conversion > -Wno-invalid-partial-specialization -Wno-typedef-redefinition > -Wno-unused-private-field -Wno-inconsistent-missing-override > -Wno-c++14-extensions -Wno-constexpr-not-const -Wno-missing-braces > -Wunused-lambda-capture -Wunused-local-typedef -Qunused-arguments > -fcolor-diagnostics -fdiagnostics-color=always > -Wno-unused-but-set-variable -fno-math-errno -fno-trapping-math > -Werror=format -Werror=cast-function-type -DHAVE_AVX512_CPU_DEFINITION > -DHAVE_AVX2_CPU_DEFINITION -O2 -pipe -fstack-protector-strong -isystem > /usr/local/include -fno-strict-aliasing -isystem /usr/local/include > -DNDEBUG -DNDEBUG -std=gnu++14 -fPIC -DTH_HAVE_THREAD -Wall -Wextra > -Wno-unused-parameter -Wno-unused-function -Wno-unused-result > -Wno-missing-field-initializers -Wno-write-strings > -Wno-unknown-pragmas -Wno-type-limits -Wno-array-bounds > -Wno-sign-compare -Wno-strict-overflow -Wno-strict-aliasing > -Wno-error=deprecated-declarations -Wno-missing-braces > -Wno-range-loop-analysis -fvisibility=hidden -O2 -fopenmp=libomp > -DCAFFE2_BUILD_MAIN_LIB -pthread -MD -MT > > caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/output-archive.cpp.o > -MF > caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/output-archive.cpp.o.d > -o > caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/output-archive.cpp.o > -c > /usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc/api/src/serialize/output-archive.cpp > ninja: build stopped: subcommand failed. > ===> Compilation failed unexpectedly. > Try to set MAKE_JOBS_UNSAFE=yes and rebuild before reporting the failure to > the maintainer. > *** Error code 1 > > Stop. > make: stopped in /usr/ports/misc/pytorch > > > > > Il gio 20 apr 2023, 17:51 Aryeh Friedman <aryeh.friedman@gmail.com> ha > scritto: > >> > >> On Thu, Apr 20, 2023 at 7:52 AM Thierry Thomas <thierry@freebsd.org> > wrote: > >> > > >> > Le jeu. 20 avr. 23 à 12:53:05 +0200, Aryeh Friedman < > aryeh.friedman@gmail.com> > >> > écrivait : > >> > > >> > > Running without GPU (for now) on a bhyve vm (3 CPU, 2 GB RAM and 100 > >> > > GB of disk) which I intend for determining if it is worse going out > >> > > and getting the hardware to do GPU. The problem I had was getting > >> > > pytorch to work since it appears I have to build it from source and > it > >> > > blows up in that build. > >> > > >> > Have you seen > >> > <https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=269739>; ? > >> > >> This seems to be true for all OS's I guess I will have to find an > >> intel machine... this is as bad as the motivation that led me to do > >> petitecloud in the first place (openstack not running on AMD period). > >> Is there just no way to run a ANN in pytorch data format in any other > >> way that is not python (like Java?!!?) note the tensorflow port > >> required pytorch > >> > >> > >> -- > >> Aryeh M. Friedman, Lead Developer, http://www.PetiteCloud.org > >> > > > -- > Aryeh M. Friedman, Lead Developer, http://www.PetiteCloud.org > -- Mario. [-- Attachment #2 --] <div dir="ltr"><pre class="gmail-_3GnarIQX9tD_qsgXkfSDz1"><code class="gmail-_34q3PgLsx9zIU5BiSOjFoM">Can't you install pytorch using the linux miniconda installer like below ? <br><br># fetch <a href="https://gist.githubusercontent.com/shkhln/40ef290463e78fb2b0000c60f4ad797e/raw/f640983249607e38af405c95c457ce4afc85c608/uvm_ioctl_override.c">https://gist.githubusercontent.com/shkhln/40ef290463e78fb2b0000c60f4ad797e/raw/f640983249607e38af405c95c457ce4afc85c608/uvm_ioctl_override.c</a>; # /compat/ubuntu/bin/gcc --sysroot=/compat/ubuntu -m64 -std=c99 -Wall -ldl -fPIC -shared -o dummy-uvm.so uvm_ioctl_override.c # pkg install linux-miniconda-installer # miniconda-installer # bash # source /home/marietto/miniconda3/etc/profile.d/conda.sh # conda activate (base) # conda activate pytorch </code></pre></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Fri, Apr 21, 2023 at 2:38 AM Aryeh Friedman <<a href="mailto:aryeh.friedman@gmail.com">aryeh.friedman@gmail.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">On Thu, Apr 20, 2023 at 12:24 PM Mario Marietto <<a href="mailto:marietto2008@gmail.com" target="_blank">marietto2008@gmail.com</a>> wrote:<br> ><br> > try to copy and paste the commands that you have issued on pastebin...i need to understand the scenario<br> <br> After saving the patch from the bug report to PORT/files and running<br> portmaster -P misc/pytourch (brand new machine except for installing<br> portmaster):<br> <br> c/ATen/UfuncCPUKernel_add.cpp.AVX2.cpp.o -c<br> /usr/ports/misc/pytorch/work/.build/aten/src/ATen/UfuncCPUKernel_add.cpp.AVX2.cpp<br> In file included from<br> /usr/ports/misc/pytorch/work/.build/aten/src/ATen/UfuncCPUKernel_add.cpp.AVX2.cpp:1:<br> In file included from<br> /usr/ports/misc/pytorch/work/.build/aten/src/ATen/UfuncCPUKernel_add.cpp:3:<br> In file included from<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/native/ufunc/add.h:6:<br> In file included from<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/functional.h:3:<br> In file included from<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/functional_base.h:6:<br> In file included from<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec.h:6:<br> In file included from<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256.h:12:<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:253:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_acosf8_u10);<br> ^~~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:256:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_asinf8_u10);<br> ^~~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:259:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_atanf8_u10);<br> ^~~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:280:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_erff8_u10);<br> ^~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:283:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_erfcf8_u15);<br> ^~~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:300:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_expf8_u10);<br> ^~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:303:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_expm1f8_u10);<br> ^~~~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:393:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_logf8_u10);<br> ^~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:396:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_log2f8_u10);<br> ^~~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:399:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_log10f8_u10);<br> ^~~~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:402:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_log1pf8_u10);<br> ^~~~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:406:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_sinf8_u10);<br> ^~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:409:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_sinhf8_u10);<br> ^~~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:412:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_cosf8_u10);<br> ^~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:415:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_coshf8_u10);<br> ^~~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:447:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_tanf8_u10);<br> ^~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:450:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_tanhf8_u10);<br> ^~~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:460:16:<br> error: cannot initialize a parameter of type 'const __m256<br> (*)(__m256)' with an lvalue of type '__m256 (__m256)': different<br> return type ('const __m256' (vector of 8 'float' values) vs '__m256'<br> (vector of 8 'float' values))<br> return map(Sleef_lgammaf8_u10);<br> ^~~~~~~~~~~~~~~~~~<br> /usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/cpu/vec/vec256/vec256_bfloat16.h:209:49:<br> note: passing argument to parameter 'vop' here<br> Vectorized<BFloat16> map(const __m256 (*const vop)(__m256)) const {<br> ^<br> 18 errors generated.<br> [ 80% 1035/1283] /usr/bin/c++ -DAT_PER_OPERATOR_HEADERS<br> -DCPUINFO_SUPPORTED_PLATFORM=0 -DFMT_HEADER_ONLY=1<br> -DHAVE_MALLOC_USABLE_SIZE=1 -DHAVE_MMAP=1 -DHAVE_SHM_OPEN=1<br> -DHAVE_SHM_UNLINK=1 -DMINIZ_DISABLE_ZIP_READER_CRC32_CHECKS<br> -DONNXIFI_ENABLE_EXT=1 -DONNX_ML=1 -DONNX_NAMESPACE=onnx<br> -DUSE_EXTERNAL_MZCRC -D_FILE_OFFSET_BITS=64 -Dtorch_cpu_EXPORTS<br> -I/usr/ports/misc/pytorch/work/.build/aten/src<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src<br> -I/usr/ports/misc/pytorch/work/.build<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/foxi<br> -I/usr/ports/misc/pytorch/work/.build/third_party/foxi<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc/api<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc/api/include<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/caffe2/aten/src/TH<br> -I/usr/ports/misc/pytorch/work/.build/caffe2/aten/src/TH<br> -I/usr/ports/misc/pytorch/work/.build/caffe2/aten/src<br> -I/usr/ports/misc/pytorch/work/.build/caffe2/../aten/src<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/miniz-2.1.0<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/kineto/libkineto/include<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/kineto/libkineto/src<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/../third_party/catch/single_include<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/..<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/c10/..<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/cpuinfo/include<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/FP16/include<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/fmt/include<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/flatbuffers/include<br> -isystem /usr/ports/misc/pytorch/work/pytorch-v1.13.1/cmake/../third_party/eigen<br> -isystem /usr/ports/misc/pytorch/work/pytorch-v1.13.1/caffe2 -O2 -pipe<br> -fstack-protector-strong -isystem /usr/local/include<br> -fno-strict-aliasing -isystem /usr/local/include -Wno-deprecated<br> -fvisibility-inlines-hidden -fopenmp=libomp -DNDEBUG -DUSE_KINETO<br> -DLIBKINETO_NOCUPTI -DSYMBOLICATE_MOBILE_DEBUG_HANDLE<br> -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra<br> -Werror=return-type -Werror=non-virtual-dtor<br> -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds<br> -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter<br> -Wno-unused-function -Wno-unused-result -Wno-strict-overflow<br> -Wno-strict-aliasing -Wno-error=deprecated-declarations<br> -Wvla-extension -Wno-range-loop-analysis -Wno-pass-failed<br> -Wno-error=pedantic -Wno-error=redundant-decls<br> -Wno-error=old-style-cast -Wconstant-conversion<br> -Wno-invalid-partial-specialization -Wno-typedef-redefinition<br> -Wno-unused-private-field -Wno-inconsistent-missing-override<br> -Wno-c++14-extensions -Wno-constexpr-not-const -Wno-missing-braces<br> -Wunused-lambda-capture -Wunused-local-typedef -Qunused-arguments<br> -fcolor-diagnostics -fdiagnostics-color=always<br> -Wno-unused-but-set-variable -fno-math-errno -fno-trapping-math<br> -Werror=format -Werror=cast-function-type -DHAVE_AVX512_CPU_DEFINITION<br> -DHAVE_AVX2_CPU_DEFINITION -O2 -pipe -fstack-protector-strong -isystem<br> /usr/local/include -fno-strict-aliasing -isystem /usr/local/include<br> -DNDEBUG -DNDEBUG -std=gnu++14 -fPIC -DTH_HAVE_THREAD -Wall -Wextra<br> -Wno-unused-parameter -Wno-unused-function -Wno-unused-result<br> -Wno-missing-field-initializers -Wno-write-strings<br> -Wno-unknown-pragmas -Wno-type-limits -Wno-array-bounds<br> -Wno-sign-compare -Wno-strict-overflow -Wno-strict-aliasing<br> -Wno-error=deprecated-declarations -Wno-missing-braces<br> -Wno-range-loop-analysis -fvisibility=hidden -O2 -fopenmp=libomp<br> -DCAFFE2_BUILD_MAIN_LIB -pthread -MD -MT<br> caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/input-archive.cpp.o<br> -MF caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/input-archive.cpp.o.d<br> -o caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/input-archive.cpp.o<br> -c /usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc/api/src/serialize/input-archive.cpp<br> [ 80% 1035/1283] /usr/bin/c++ -DAT_PER_OPERATOR_HEADERS<br> -DCPUINFO_SUPPORTED_PLATFORM=0 -DFMT_HEADER_ONLY=1<br> -DHAVE_MALLOC_USABLE_SIZE=1 -DHAVE_MMAP=1 -DHAVE_SHM_OPEN=1<br> -DHAVE_SHM_UNLINK=1 -DMINIZ_DISABLE_ZIP_READER_CRC32_CHECKS<br> -DONNXIFI_ENABLE_EXT=1 -DONNX_ML=1 -DONNX_NAMESPACE=onnx<br> -DUSE_EXTERNAL_MZCRC -D_FILE_OFFSET_BITS=64 -Dtorch_cpu_EXPORTS<br> -I/usr/ports/misc/pytorch/work/.build/aten/src<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src<br> -I/usr/ports/misc/pytorch/work/.build<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/foxi<br> -I/usr/ports/misc/pytorch/work/.build/third_party/foxi<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc/api<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc/api/include<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/caffe2/aten/src/TH<br> -I/usr/ports/misc/pytorch/work/.build/caffe2/aten/src/TH<br> -I/usr/ports/misc/pytorch/work/.build/caffe2/aten/src<br> -I/usr/ports/misc/pytorch/work/.build/caffe2/../aten/src<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/miniz-2.1.0<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/kineto/libkineto/include<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/kineto/libkineto/src<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/../third_party/catch/single_include<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/aten/src/ATen/..<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/c10/..<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/cpuinfo/include<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/FP16/include<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/fmt/include<br> -I/usr/ports/misc/pytorch/work/pytorch-v1.13.1/third_party/flatbuffers/include<br> -isystem /usr/ports/misc/pytorch/work/pytorch-v1.13.1/cmake/../third_party/eigen<br> -isystem /usr/ports/misc/pytorch/work/pytorch-v1.13.1/caffe2 -O2 -pipe<br> -fstack-protector-strong -isystem /usr/local/include<br> -fno-strict-aliasing -isystem /usr/local/include -Wno-deprecated<br> -fvisibility-inlines-hidden -fopenmp=libomp -DNDEBUG -DUSE_KINETO<br> -DLIBKINETO_NOCUPTI -DSYMBOLICATE_MOBILE_DEBUG_HANDLE<br> -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra<br> -Werror=return-type -Werror=non-virtual-dtor<br> -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds<br> -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter<br> -Wno-unused-function -Wno-unused-result -Wno-strict-overflow<br> -Wno-strict-aliasing -Wno-error=deprecated-declarations<br> -Wvla-extension -Wno-range-loop-analysis -Wno-pass-failed<br> -Wno-error=pedantic -Wno-error=redundant-decls<br> -Wno-error=old-style-cast -Wconstant-conversion<br> -Wno-invalid-partial-specialization -Wno-typedef-redefinition<br> -Wno-unused-private-field -Wno-inconsistent-missing-override<br> -Wno-c++14-extensions -Wno-constexpr-not-const -Wno-missing-braces<br> -Wunused-lambda-capture -Wunused-local-typedef -Qunused-arguments<br> -fcolor-diagnostics -fdiagnostics-color=always<br> -Wno-unused-but-set-variable -fno-math-errno -fno-trapping-math<br> -Werror=format -Werror=cast-function-type -DHAVE_AVX512_CPU_DEFINITION<br> -DHAVE_AVX2_CPU_DEFINITION -O2 -pipe -fstack-protector-strong -isystem<br> /usr/local/include -fno-strict-aliasing -isystem /usr/local/include<br> -DNDEBUG -DNDEBUG -std=gnu++14 -fPIC -DTH_HAVE_THREAD -Wall -Wextra<br> -Wno-unused-parameter -Wno-unused-function -Wno-unused-result<br> -Wno-missing-field-initializers -Wno-write-strings<br> -Wno-unknown-pragmas -Wno-type-limits -Wno-array-bounds<br> -Wno-sign-compare -Wno-strict-overflow -Wno-strict-aliasing<br> -Wno-error=deprecated-declarations -Wno-missing-braces<br> -Wno-range-loop-analysis -fvisibility=hidden -O2 -fopenmp=libomp<br> -DCAFFE2_BUILD_MAIN_LIB -pthread -MD -MT<br> caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/output-archive.cpp.o<br> -MF caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/output-archive.cpp.o.d<br> -o caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/api/src/serialize/output-archive.cpp.o<br> -c /usr/ports/misc/pytorch/work/pytorch-v1.13.1/torch/csrc/api/src/serialize/output-archive.cpp<br> ninja: build stopped: subcommand failed.<br> ===> Compilation failed unexpectedly.<br> Try to set MAKE_JOBS_UNSAFE=yes and rebuild before reporting the failure to<br> the maintainer.<br> *** Error code 1<br> <br> Stop.<br> make: stopped in /usr/ports/misc/pytorch<br> <br> ><br> > Il gio 20 apr 2023, 17:51 Aryeh Friedman <<a href="mailto:aryeh.friedman@gmail.com" target="_blank">aryeh.friedman@gmail.com</a>> ha scritto:<br> >><br> >> On Thu, Apr 20, 2023 at 7:52 AM Thierry Thomas <<a href="mailto:thierry@freebsd.org" target="_blank">thierry@freebsd.org</a>> wrote:<br> >> ><br> >> > Le jeu. 20 avr. 23 à 12:53:05 +0200, Aryeh Friedman <<a href="mailto:aryeh.friedman@gmail.com" target="_blank">aryeh.friedman@gmail.com</a>><br> >> > écrivait :<br> >> ><br> >> > > Running without GPU (for now) on a bhyve vm (3 CPU, 2 GB RAM and 100<br> >> > > GB of disk) which I intend for determining if it is worse going out<br> >> > > and getting the hardware to do GPU. The problem I had was getting<br> >> > > pytorch to work since it appears I have to build it from source and it<br> >> > > blows up in that build.<br> >> ><br> >> > Have you seen<br> >> > <<a href="https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=269739" rel="noreferrer" target="_blank">https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=269739</a>> ?<br> >><br> >> This seems to be true for all OS's I guess I will have to find an<br> >> intel machine... this is as bad as the motivation that led me to do<br> >> petitecloud in the first place (openstack not running on AMD period).<br> >> Is there just no way to run a ANN in pytorch data format in any other<br> >> way that is not python (like Java?!!?) note the tensorflow port<br> >> required pytorch<br> >><br> >><br> >> --<br> >> Aryeh M. Friedman, Lead Developer, <a href="http://www.PetiteCloud.org" rel="noreferrer" target="_blank">http://www.PetiteCloud.org</a><br>; >><br> <br> <br> -- <br> Aryeh M. Friedman, Lead Developer, <a href="http://www.PetiteCloud.org" rel="noreferrer" target="_blank">http://www.PetiteCloud.org</a><br>; </blockquote></div><br clear="all"><br><span class="gmail_signature_prefix">-- </span><br><div dir="ltr" class="gmail_signature">Mario.<br></div>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CA%2B1FSijsSSpCFeKeaOt4gR36BAZ4J8j4QSJRJa-VF-a=J9e2uw>