![]() ![]() Then, install the tensorflow-macosbase along with the tensorflow-metal plugin: $ conda install -c apple tensorflow-deps $ pip install tensorflow-macos $ pip install tensorflow-metalĪfterward, install the Jupyter notebook or Jupyter lab: $ conda install -c conda-forge jupyter jupyterlabĮxecute the following code snippet and check if everything went as expected. Don’t forget to activate it: $ conda create -name tensorflow_silicon python=3.9 $ conda activate tensorflow_silicon ![]() According to the official guideline, we should use at least python version 3.9.0. Now, let’s create a new conda environment in MiniForge, say tensorflow_silicon. This article describes a trick that can help you manage many conda distributions simultaneously! Step 3: Setup conda environment and install MiniForge In case you already have a pre-existing conda installation, there is no need to uninstall it to use MiniForge. Anaconda or MiniConda can be “functional” at a given time. Note 2: It is worth mentioning that only one conda distribution, e.g. You don’t have to involve Homebrew or custom download scripts (which are usually buggy). Note 1: This is the most straightforward installation process. To install it, execute: $ bash Miniforge3-MacOSX-arm64.sh Image 1: The correct conda Miniforge package for Apple Silicon ( Source) To download it, go to this page and choose the installer for Apple Silicon as shown in the following picture: Also, MiniForge is Apple-friendly, making it an ideal candidate for MacOS, including the M1 devices. To do this, execute the following: $ xcode-select -install Step 2: Install MiniForgeįor those who don’t know, MiniForge is a conda installer, similar to MiniConda. Afterward, you should also install the Xcode Command Line Tools. Alternatively, you can easily download it from the App Store. Nowadays, most Macbooks have Xcode preinstalled.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |