install into new 3.7 environment, Keras requires TensorFlow 2.2 or higher

[shawngraham]

Hiya,

Just trying to follow the minimal example. I've got a new python 3.7 environment (made with conda), I pip install dvt then give python3 -m dvt video-viz target.mp4 a whirl:

ModuleNotFoundError: No module named 'tensorflow.keras.layers.experimental.preprocessing'

and then

Keras requires TensorFlow 2.2 or higher.

But if I installed tf2.2, I get dvt 0.3.3 requires tensorflow==1.15.2, but you have tensorflow 2.2.0 which is incompatible.

so I'm assuming that I've got something borked on my machine.

[JoannaBy]

Hi, I have the same issue, installed in a virtualenv with all the dependencies as in the proposed setup but I seem to get a conflict between the tf that dvt and keras require:
when I run the minimal demo I get

ImportError: Keras requires TensorFlow 2.2 or higher. Install TensorFlow via `pip install tensorflow`

and again after I upgrade tf

dvt 0.3.3 requires tensorflow==1.15.2, but you have tensorflow 2.4.1 which is incompatible.

I will be grateful for any tips on how to bypass this issue!

[taylor-arnold]

Hello! We've been working on getting a major re-write of the API finished and just pushed a new version to GitHub and PyPI (0.4.0). It has the same core functionality at the older version, but is a lot leaner and should be easier to install. Perhaps you all could take a look at the Demo through the Colab link on the main page. There is also an updated INSTALL.md that should work for Linux and macOS.

Please let us know if you continue to have any issues. I am sure there will still be some things to fix as we work through this new version, but expect the newer API to be relatively stable for the coming months.

[Goli-SF]

Hi there,

I'm trying to use the online tutorial on Google colab, but when I run the first command block under heading #9 (Detecting and Identifying Faces), I get the following error:

cannot import name 'get_file' from 'keras.utils' (/usr/local/lib/python3.7/dist-packages/keras/utils/init.py)

I also seemed to have an issue with tensorflow on my pc, as I tried to use dvt to analyze clips on my hard disc. Apparently I do not have the right version. I tried to pip install the lowest version available (2.2.0), but without any luck, because I get the following errors:

pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
tensorflow-gpu 2.3.0 requires numpy<1.19.0,>=1.16.0, but you have numpy 1.19.5 which is incompatible.
tensorflow-gpu 2.3.0 requires tensorboard<3,>=2.3.0, but you have tensorboard 2.2.2 which is incompatible.

I tried installing the mentioned versions of numpy and tensorboard, but they also don't install because of incompatibility of versions with other dependancies.

I would be very thankful for help and tips on how to resolve these issues!