設(shè)置Xcode

打開終端并執(zhí)行

sudo xcode－select －－install

安裝HomeBrew（原生Apple Silicon M1）

打開終端，逐個寫入

這將為M1芯片安裝最新的Brew／bin／bash －c ＂＄（curl －fsSL https：／／raw．githubusercontent．com／Homebrew／install／HEAD／install．sh）＂
echo “export PATH＝／opt／homebrew／bin：＄PATH” ＞＞ ～／．zshrc
／／Restart The Terminal

brew install gcc

brew install cmake

brew install wget

安裝Miniforge，設(shè)置Conda環(huán)境點(diǎn)擊下面的鏈接下載（Apple Silicon）版本https：／／github．com／conda－forge／miniforge

打開終端并執(zhí)行以下操作／／ If the Downloaded File Stored in Download
cd Downloads
bash Miniforge3－MacOSX－arm64．sh
／／After Installation Completes Restart Terminal
／／Creating Conda Environment named ml You can use any name in place           of ＂ml＂
conda create －－name ml
conda install －y python＝＝3．8．6
conda install －y pandas matplotlib scikit－learn jupyterlab
安裝Tensorflow單擊下面的鏈接并下載文件https：／／github．com／apple／tensorflow＿macos／releasesAm在2021年3月3日為M1使用最新的TF alpha 2版本。

／／if Download Directory is Downloads
cd Downloads
tar xvf tensorflow＿macos－0．1alpha2．tar．gz
cd tensorflow＿macos／arm64
／／Dont Forget To Activate Conda Environment
conda activate ml
／／ Install specific pip version and some other base packages
pip install －－force pip＝＝20．2．4 wheel setuptools cached－property six
／／ Install all the packages provided by Apple but TensorFlow
pip install －－upgrade －－no－dependencies －－force numpy－1．18．5－cp38－cp38－macosx＿11＿0＿arm64．whl grpcio－1．33．2－cp38－cp38－macosx＿11＿0＿arm64．whl h5py－2．10．0－cp38－cp38－macosx＿11＿0＿arm64．whl tensorflow＿addons＿macos－0．1a2－cp38－cp38－macosx＿11＿0＿arm64．whl
／／ Install additional packages
pip install absl－py astunparse flatbuffers gast google＿pasta keras＿preprocessing opt＿einsum protobuf tensorflow＿estimator termcolor typing＿extensions wrapt wheel tensorboard typeguard
／／ Install TensorFlow
pip install －－upgrade －－force －－no－dependencies tensorflow＿macos－0．1a2－cp38－cp38－macosx＿11＿0＿arm64．whl
安裝額外的包pip install matplotlib
conda install －c conda－forge scikit－learn
pip install keras
pip install notebook
編譯和安裝OpenCV／／I Suggest To Do all this Inside miniforge3 dir for that
／／  cd miniforge3
wget －O opencv．zip https：／／github．com／opencv／opencv／archive／4．5．0．zip
wget －O opencv＿contrib．zip https：／／github．com／opencv／opencv＿contrib／archive／4．5．0．zip
unzip opencv．zip
unzip opencv＿contrib．zip
cd opencv－4．5．0
mkdir build ＆＆ cd build
／／Here Take Care Of Paths of OPENCV＿EXTRA＿M(jìn)ODULES＿PATH and  
／／    PYTHON3＿EXECUTABLE If you＇re Beginner watch the YouTube  video
／／And If Inside miniforge3 just place your ＜username＞．
cmake
－DCMAKE＿SYSTEM＿PROCESSOR＝arm64
－DCMAKE＿OSX＿ARCHITECTURES＝arm64
－DWITH＿OPENJPEG＝OFF
－DWITH＿IPP＝OFF
－D CMAKE＿BUILD＿TYPE＝RELEASE
－D CMAKE＿INSTALL＿PREFIX＝／usr／local
－D OPENCV＿EXTRA＿M(jìn)ODULES＿PATH＝／Users／＜username＞／miniforge3／opencv＿contrib－4．5．0／modules
－D PYTHON3＿EXECUTABLE＝／Users／＜username＞／miniforge3／envs／ml／bin／python3
－D BUILD＿opencv＿python2＝OFF
－D BUILD＿opencv＿python3＝ON
－D INSTALL＿PYTHON＿EXAMPLES＝ON
－D INSTALL＿C＿EXAMPLES＝OFF
－D OPENCV＿ENABLE＿NONFREE＝ON
－D BUILD＿EXAMPLES＝ON ．．
make －j8
／／＂8＂ is the number of cores To be used（This Step Takes Time）
sudo make install
／／Linking OpenCV To Conda Environment
mdfind cv2．cpython
／／From the output Copy the Path similar to the below one
＂／usr／local／lib／python3．8／site－packages／cv2／python－3．8／cv2．cpython－38－darwin．so cv2．so＂
cd
cd miniforge3／envs／dev／lib／python3．8／site－packages
ln －s PasteYourCopiedPathHere