TFX Standard Components Walkthrough Reviews

You have to use an old version of debian with python 3.7

Filippo M. · Reviewed over 2 years ago

it does not work, i can't download the correct version of tfx, ...

Filippo M. · Reviewed over 2 years ago

very poor

Arun M. · Reviewed over 2 years ago

Andrea A. · Reviewed over 2 years ago

Amal A. · Reviewed over 2 years ago

Impossibile to complete: task 4 leads to incompatibilities with library version within JupyterLab

Mauro C. · Reviewed over 2 years ago

Could not run the lab, problems with install.sh script

Giovanni Daián R. · Reviewed over 2 years ago

This lab is very outdated. I spent 75% of the time finding fixes to get the notebook to run, and could hardly read through the explanations. The first 2 tasks ran fine. Issues started at task 3: - The tensorflow2.11 notebook VM has python3.10, which is incompatible with tensorflow-data-validation. So tfdv could not be imported or installed. Quick fix was to switch to tensorflow2.6 on Debian10 which has python3.7. - The notebook required tensorflow2.3.1 (not 2.11 as specified). Installing this led to compatibility issues. - I had to install markupsafe==2.0.1 due to an import error "cannot import name 'soft_unicode' from 'markupsafe'". - I had to install apache-beam==2.26.0 so that the tfx components could run. While debugging I found out from StackOverflow that the error was due to incompatibility, and the suggested fix was to try an older version. I randomly tried a few and 2.26.0 worked for me. - Still, there were many warnings and non-critical errors. But I managed to get to the end.

Timothy A. · Reviewed over 2 years ago

Joyce C. · Reviewed over 2 years ago

Vemal M. · Reviewed over 2 years ago

Joyce C. · Reviewed over 2 years ago

Joyce C. · Reviewed over 2 years ago

Giovanni Daián R. · Reviewed over 2 years ago

Literarily no cell in the notebook works!

Kasra Z. · Reviewed over 2 years ago

Nikolaus M. · Reviewed over 2 years ago

TFX components installation was not successful after kernel restart also

Samsuddin S. · Reviewed over 2 years ago

Rajkumar T. · Reviewed over 2 years ago

tenserflow enterprise 2.3 is not available

Ranjit B. · Reviewed over 2 years ago

fully broken lab please fix :( *nothing imports :/ no option for TF2.3

Ray M. · Reviewed over 2 years ago

Juan Jose G. · Reviewed over 2 years ago

TensorFlow 2.3 not available at the UI. The notebook does not run with 2.6. I had to execute through cloud shell: gcloud notebooks instances create tensorflow-notebook --vm-image-project=deeplearning-platform-release --vm-image-family=tf-2-3-cpu --machine-type=n1-standard-1 --location=us-central1-a

Jordi V. · Reviewed over 2 years ago

Erron

Jia Jun G. · Reviewed over 2 years ago

this lab was obsolete

23-Muhammad D. · Reviewed over 2 years ago

Volodymyr S. · Reviewed over 2 years ago

very poor

Arun M. · Reviewed over 2 years ago