When trying to load the libraries in google colab and kaggle it is not getting load correctly.

[mahii6991]

This is the error it is showing.

Though I have tried loading the kernel multiple times, but still I am getting the same error.

[ptn77]

Did you get this resolved? I'm also getting similar errors. Thanks!

[PlatypusProgrammer]

I am also not able to execute the example notebooks. E.g.: https://www.tensorflow.org/federated/tutorials/federated_learning_for_image_classification Pip is not able to install the packages.

[mahii6991]

yes, for me it is not resolved till now !!!

[michaelreneer]

@mahii6991

It is somewhat expected (though unfortunate) that you are getting error in Google colab. The reason for this is because the default runtime that you are executing the notebook in contains python packages that are not compatible with tensorflow-federated. However, these errors should not prevent you from continuing to execute the cells in the notebook.

1. Do you get an error when you run the next cell in the notebook?
2. If you run !pip freeze after running !pip install --quiet --upgrade tensorflow-federated, do you see that tensorflow-federated is installed?

If you want to get rid of the errors, then you can:

1. Create your own runtime or connect to an existing runtime that contains a set of packages that does not conflict with tensorflow-federated.
2. You could uninstall all the python packages that conflict with tensorflow-federated before trying to install tensorflow-federated, !pip uninstall --yes <foo>.
3. You could try to uninstall all the python packages before trying to install tensorflow-federated, !pip freeze | xargs uninstall --yes.
4. You could redirect the errors to 2>/dev/null.