If I do an ML project copying code heavily from Tensorflow's tutorials, am I still allowed to share it on Github (or other repositories)?

Question

(I asked this on Reddit and didn't get a good reaction. I wasn't sure if this was the right forum, but I wanted to ask somewhere related.)

I'm fairly new to machine learning (and programming in general).

The Tensorflow website provides a lot of tutorials on how to use their tools, and general ML tutorials too.

I am interested in building a machine translator using a transformer. Tensorflow has a tutorial to teach exactly that. I'll be using a different dataset, but I will be heavily copying the code provided in the tutorial and then sharing the finished project on Github. Of course, I do plan giving credit to the Tensorflow tutorial.

Am I allowed to share this project on Github? Are there any legal issues that could come from this? I don't plan on commercializing this project.

Answer 1

At the bottom of the linked page, it notes that content "is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License". Assuming you're not also copying chunks of expository text, your use of the code is governed by the Apache licence.

That licence grants you the right to make and distribute derivative works (which is what you will be doing, in copyright terms, when you use the code to make your program, and then when you share that program on (eg) GitHub), subject to certain obligations.

We have many good summaries of those obligations, such as Bart's here, so I won't repeat them yet again, but as long as you honour those then what you propose sounds fine to me.