I am sure the main reason is the default of @GoogleColab is not TF2 yet.
Every time I use Colab, the first step is always this:
!pip install tensorflow==2.0.0
PyPI downloads for TensorFlow (and its closest competitor, added for scale). Notice how it starts jumping after the release of TF 2.0 late last year (the short gap afterwards is the holiday break)
Up and to the right 
pic.twitter.com/Tlltw40hMO
-
-
-
You can actually select TF 2 at startup: https://colab.research.google.com/notebooks/tensorflow_version.ipynb … If installing from pypi I would suggest using the latest version (2.1 right now): !pip install tensorflow --quiet --upgrade
- Show replies
New conversation -
-
-
Do you host a PyPI-like mirror for internal TF deployments/CI? If no, these numbers might be biased
-
Software at Google is always built from source, it is never downloaded from an external source.
- Show replies
New conversation -
-
-
The jump could be just trying out tf2 from a lot of people who had a long break to other frameworks (like pytorch). Personally, i just tried tf2 and then pip uninstalled it.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
People upgrading because of breaking incompatibilities?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Nice, but... Using this metric doesn't incentivize to keep the dependency hell at bay. First setup of a new stack is currently painful (many tries --> many downloads)
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Hello François, can I know what software/web app you used to get these pypi stats?
-
Get the latest pypi stats out with BigQuery https://towardsdatascience.com/python-pypi-stats-in-bigquery-reclustered-d80e583e1bfe … (There's a official repo, but in this copy your queries will cost way less)
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.