Hi,
I am trying to run the last assignment of course 2 about Tensorflow on my personal M1pro Mac. But I find it very difficult due to version conflicts. Many methods have been used from Tensoflow version 2.3.0 is not compatible with other versions and I cannot successfully install v2.3.0.
I am wondering if you have an updated version of the assignments in the best case or any suggestions about how to run these codes on our personal computers?
Thanks
Actually, different versions of TF are used in Coursera environment. In addition, lots of additional components which have version dependencies to other components are used. Sometimes different version of Python are required. So, it is not a matter only of TF. But, I also understand that in the case of M1 Mac, the most problematic component is tensorflow, since the version of tensorflow-macos is 2.5 or higher.
I have multiple local environment on conda to make all assignments work on my INTEL Mac. In my case, most frustration is eGPU support by Apple. 
In any cases, the versions of required components are fixed in the Grader environment. But, as a fact, some of assignments work even on TensorFlow 2.9. So, just try it. And, if it does not work, then, you may have some choices… Modify assignment itself to run on tensorflow-macos 2.8, which may not work on the grader environment, or, just focus on the grader environment… 