[Python] pip keep collecting many Tensorflow versions when installing alibi==0.5.5

I'm using a python project and installing its dependencies. The following are all the packages I need to install.

torch==1.3.0
numpy==1.15.0
pandas==0.25.2
scikit-learn==0.23.1
pyod==0.8.2
tqdm==4.48.2
prettytable==0.7.2
shap==0.36.0
lime==0.2.0.1
alibi==0.5.5


I put them in a requirements.txt file. Then I run pip install -r requirements.txt.

The problem happens that, I can install everything successfully before the last one, alibi==0.5.5. When installing it, it seems require a tensorflow >= 2.0. That's okay. But when the installation starts, it keeps collecting all tensorflow versions >= 2.0, as is shown in the figure. Why does this happen? How can I deal with it?

Thank you all for helping me!!!



Continue reading...