Install packages within a virtual environment without affecting the host system When the virtual environment is active, your shell prompt is prefixed with (venv). venvĪctivate the virtual environment using a shell-specific command: Python virtual environments are used to isolate package installation from the system.Ĭreate a new virtual environment by choosing a Python interpreter and making a Create a virtual environment (recommended)
This ensures that you upgrade and use theĢ. If not in a virtual environment, use python3 -m pip for Other curl -o get-pip.py python get-pip.py Caution: Upgrading the system pip can cause Install the 64-bit Python 3 release for Windows (select pip as an optional feature). Make sure long paths are enabled on Windows.
Install using the Homebrew package manager: /usr/bin/ruby -e "$(curl -fsSL )" export PATH="/usr/local/opt/python/libexec/bin:$PATH" # if you are on macOS 10.12 (Sierra) use `export PATH="/usr/local/bin:/usr/local/sbin:$PATH"` brew update brew install python # Python 3 Windows If you need those libraries, you will have to use TensorFlow Some TensorFlowīinaries (specifically, ones with custom C++ extensions like TensorFlowĭecision Forests, object detection, and TFX) are not compiled against M1 Ubuntu sudo apt update sudo apt install python3-dev python3-pip python3-venv macOS Note:įor users of Apple M1 computers, to get native performance, you'll wantĬonda has shown to have the smoothest install.
If these packages are already installed, skip to the next step. Requires Python 3.7-3.10, and pip >= 19.0 python3 -version pip3 -version Install the Python development environment on your systemĬheck if your Python environment is already configured: Read the GPU support guide to set up a CUDA®-enabled GPU card on Ubuntu or Windows.ġ.Starting with TensorFlow 1.6, binaries use AVX instructions which may not run on older CPUs.Note: Installing TensorFlow 2 requires a newer version of pip. GPU support requires a CUDA®-enabled card (Ubuntu and Windows).Redistributable for Visual Studio 2015, 20 macOS 10.12.6 (Sierra) or later (64-bit) (no GPU support).pip 19.0 or later (requires manylinux2010 support).Python 3.8 support requires TensorFlow 2.2 or later.Python 3.9 support requires TensorFlow 2.5 or later.
Python 3.10 support requires TensorFlow 2.8 or later.tensorflow-gpu=1.15 -Release with GPU support (Ubuntu and Windows).Ubuntu and Windows include GPU support.įor TensorFlow 1.x, CPU and GPU packages are separate: tensorflow -Latest stable release with CPU and GPU support (Ubuntu and Windows).