TensorFlow对Python 3.7.4的支持问题及新版本发布时间咨询

Hey there! Let's break down your issue and options clearly:

First, you’ve nailed the root cause of that error—back when you tried installing TensorFlow, the official Windows packages only supported Python 3.5 and 3.6, so pip couldn’t find a matching build for your Python 3.7.4 environment. That’s exactly why you saw those "no matching distribution" messages.

About the release date for Python 3.7 support: #

At the time of your question, TensorFlow’s team didn’t share exact future release dates publicly. But I can confirm that TensorFlow 2.0 (officially launched in September 2019) added full support for Python 3.7 on Windows, covering both CPU and GPU builds. Before that, the best way to track progress would’ve been checking TensorFlow’s GitHub milestones or official release notes for early hints.

If you needed to use TensorFlow with Python 3.7 right away, here are practical workarounds: #

• Downgrade Python to 3.6.x: This is the most stable path—you’ll get access to fully tested, official TensorFlow builds without extra setup.
• Install the nightly preview build: Run pip install tf-nightly to grab the development version. Nightly builds often add support for newer Python versions ahead of stable releases, though they might have untested bugs or incomplete features.
• Compile from source: If you’re comfortable with build environments, you could compile TensorFlow directly from its GitHub source code to make it compatible with Python 3.7. This is more advanced, but it lets you stick with your current Python version immediately.

内容的提问来源于stack exchange，提问作者Pepv