How to know which gcc version to choose when compiling pytorch from source?
I am working on installing PyTorch from source but am unsure about the specific dependency versions to use for the version of PyTorch I want to install.
In particular, I noticed performance variations depending on the gcc version I use to compile PyTorch. Which compiler should I be using to get the best PyTorch performance?
Tensorflow doc provides such useful information. They call it "Tested build configurations": https://www.tensorflow.org/install/source#tested_build_configurations.
The README.md has instructions to build from source.
If you are installing from source, you will need a C++14 compiler. Also, we highly recommend installing an Anaconda environment. You will get a high-quality BLAS library (MKL) and you get controlled dependency versions regardless of your Linux distro.
Once you have Anaconda installed, here are the instructions.
If you want to compile with CUDA support, install
- NVIDIA CUDA 9 or above
- NVIDIA cuDNN v7 or above
- Compiler compatible with CUDA
If you want to disable CUDA support, export environment variable USE_CUDA=0. Other potentially useful environment variables may be found in setup.py.
If you are building for NVIDIA's Jetson platforms (Jetson Nano, TX1, TX2, AGX Xavier), Instructions to install PyTorch for Jetson Nano are available here
You can found Latest, official Compiler requirements here
- C++ #include and #import difference
- Is there a way to output the assembly of a single function in isolation?
- ld64 -s strip output flag - "ignored" but not really? OS X
- GCC switches to enable analysis for warnings
- Trying to no-op an instruction
- How to manually throw a compiler error in GCC and Xcode
- Initializing a dynamically-sized Variable-Length Array (VLA) to 0
- GCC complaining about a duplicate definition (where only one exists)
- What is the difference between make and gcc?
- What is _GLIBCXX_USE_NANOSLEEP all about?
- how to implement AES128 encryption/decryption using AES-NI instructions and GCC
- GBA C programming compiler gcc
- Adding compiler to code::blocks
- Creating linux cross compiler running on mac 10.6
- Variable name 'xor' results in internal compiler error 'segmentation fault'!
- modify GCC ".md" machine definition template
- Best way to strip off symbols
- Forward-declaration of enum member of template-specialization - fails with GCC
- Structure of SSE vectorization calls for summing vector of floats
- Invalid operands to binary
- likely/unlikely equivalent for MSVC
- Update GCC on OSX
- Dealing with "C compiler cannot create executables" in Cygwin
- Why is a function that only returns a stateful lambda compiled down to any assembly at all?
- Compiling Sundance-4.73
- Why is Premake5 producing incorrect file separator on Windows?
- How do I compile the asm generated by GCC?
- How to enforce the struct bit order with the GCC compiler?
- Is this unstable behavior of execve after vfork a gcc bug?
- UCRT vs. MSVCRT for GCC 14.x