lseongjoo

sudo apt-get update
sudo apt-get install cmake ninja-build gfortran

pip install --upgrade pip
pip install cython
pip install numpy==1.23.* --no-binary=:all: --force-reinstall --no-cache -v

pip install typing-extensions pytest hypothesis
python -c "import numpy; numpy.test()"

lseongjoo

./configure --enable-cblas --enable-threading=openmp auto
make
make check
sudo make install

lseongjoo

Creating additional virtual hard disk (VHDX) for WSL2 with ext4 filesystem

Lines starting with # mean the commands have to be executed by root user under Linux shell (WSL distro). All other commands have to be executed under Windows-PowerShell as Administrator.

Make VHDX disk file and mount it under Windows

New-VHD support.vhdx -SizeBytes 25GB -Dynamic -BlockSizeBytes 1MB
Write-Output "\\.\PhysicalDrive$((Mount-VHD -Path support.vhdx -PassThru | Get-Disk).Number)"

lseongjoo

name: pydeep
channels:
  - defaults
  - conda-forge
  - apple
dependencies:
  - python=3.9
  - pip
  - pandas
  - matplotlib

lseongjoo

name: pydeep
channels:
  - defaults
  - conda-forge
  - nvidia
dependencies:
  - python=3.9
  - pip
  - pandas
  - matplotlib

lseongjoo

How to install numpy on M1 Max, with the most accelerated performance (Apple's vecLib)? Here's the answer as of Dec 6 2021.

---

Steps

I. Install miniforge

So that your Python is run natively on arm64, not translated via Rosseta.

1. Download Miniforge3-MacOSX-arm64.sh, then
2. Run the script, then open another shell

$ bash Miniforge3-MacOSX-arm64.sh

lseongjoo

# 파이썬 환경 생성 및 활성화
conda create --name tf python=3.9
conda activate tf

# CUDA+cuDNN 설치 및 설정
conda install -c conda-forge cudatoolkit=11.2.2 cudnn=8.1.0
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CONDA_PREFIX/lib/
mkdir -p $CONDA_PREFIX/etc/conda/activate.d
echo 'export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CONDA_PREFIX/lib/' > $CONDA_PREFIX/etc/conda/activate.d/env_vars.sh

lseongjoo

Build from source.

pip install cython pybind11
pip install --no-binary :all: numpy

lseongjoo

Codebasic (c) 2023

다음 문서는 아래 플랫폼별 딥러닝 기반 자연어 소프트웨어 라이브러리 설치 절차를 안내합니다.

플랫폼

1. Windows (x86-64bit)
2. 유닉스 계열 (Unix-Like)
   1. Mac (Apple Silicon/Intel x86-64bit)
   2. Linux (x86-64bit)

lseongjoo

NumPy

BLAS(Basic Linear Algebra Subprograms)와 LAPACK(Linear Algebra Package)

https://numpy.org/install/#numpy-packages--accelerated-linear-algebra-libraries

소스 빌드

BLAS/LAPACK를 활용한 가속화 사용을 위해 NumPy 소스 코드르 직접 빌드할 때, 연결되는 라이브러리를 지정할 수 있습니다.