Tfx installing problem


Hi guys,
I hope you are doing well.
Guys, I need help if someone can .
As you see in the given picture , I could import tfx library although I have installed it .
it is the problem
ImportError: cannot import name ‘WKBWriter’ from ‘shapely.geos’ (C:\ProgramData\Anaconda3\envs\tensorflow-developer\lib\site-packages\shapely\geos.py)

1 Like

Tfx seems to be brittle, even via pypi. Please file a ticket on the github repo. or reply to this

Try docker:

docker pull tensorflow/tfx:1.14.0
docker run -it --entrypoint bash tensorflow/tfx:1.14.0

Inside the container:

root@bb3fadbd1807:/tfx/src# python

Python 3.8.18 (default, Aug 25 2023, 13:20:30) 
[GCC 11.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from tfx import v1 as tfx
2024-02-27 07:34:58.922814: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
>>> 
1 Like

bro , isn’t it possible that installing in jupyter notebook

1 Like

Sorry but I’m not a tfx maintainer. Please comment on github for better assistance.

Hi @Diyorbek yes tfx can be installed but at the moment it’s not stable with some of the latest Python, and pip versions due to dependencies issues. This is from my personal experience when using tfx for my projects. At the moment, using tfx on Docker is still better, especially if you want to utilize GPU for training.