Article directory
The first step is to install the Pytorch environment on the M1 chip.
Install Miniforge
Open the terminal on Mac (How to start the terminal on a Mac computer? Open the launch pad and search for the terminal)
and install the package management tool Homebrew
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
If you encounter an error
curl: (7) Failed to connect to raw.githubusercontent.com port 443: Connection refused
You need to run the following command and then install Homebrew. The reason is that it is blocked and the port needs to be configured.
export https_proxy=http://127.0.0.1:7890 http_proxy=http://127.0.0.1:7890 all_proxy=socks5://127.0.0.1:789
After successfully installing Homebrew, restart the terminal (how to start the terminal on a Mac? Open the launch pad and search for the terminal) and install miniforge directly.
brew install miniforge
After the installation is complete, initialize conda for the terminal:
conda init zsh
Create a virtual environment
Create a virtual environment named pytorch_env using python 3.8
conda create --name pytorch_env python=3.8
Activate virtual environment
conda activate pytorch_env
After activation is successful, you will see the following interface
Install PyTorch
Use the conda command to install pytorch and torchvision
conda install -c pytorch pytorch torchvision
The second step is to download the Li Mu Jupyter file.
Enter the course homepage: https://courses.d2l.ai/zh-v2/Click
on the Jupyter Notepad in the upper right corner to download.
The third step is to configure vscode
Search vscode and enter the vscode official website: https://code.visualstudio.com/
to download and install the mac version of vscode
Configure vscode: Click the virtual environment below the settings button on the right side of the picture below: pytorch_env
If it is a newly downloaded vscode, you need to manually open the following panel
Activate the virtual environment in the lower panel
conda activate pytorch_env
Then use conda to install Li Mu’s special d2l package
conda install d2l
As shown in the figure below
, then open a downloaded jupyter file and run it
reference
【1】https://zhuanlan.zhihu.com/p/394514049