Having NoneType error in multiple pipeline build using TFX and Airflow #29123
Unanswered
Engineer1999
asked this question in
Q&A
Replies: 1 comment 1 reply
-
I'm not sure that we could do something with it based on the traceback the actual error happen not in Airflow. And looks like TFX officially not supported Airflow, see their Issue Tracker |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I am working with TFX and AIrflow to build the image classification and segmentation pipeline. I am replicating the repo by Jan Marcel Kezmann (Link). While executing the classification_dag.py in dags/classification_pipeline and segmentation_dag.py in dags/segmentation_pipeline, I am getting a runtime error. The error message is mentioned below.
Airflow Dag is as below Segmentation Dag
I tried to check multiple versions of Airflow and TFX. I believe this error is due to incompatible versions as the error is coming from the executor of the transform component and airflow. Before the dag execution, I compiled the individual module and have not found any errors in them.
Versions TFX:- 1.12.0 Tensorflow:- 2.11.0 Airflow:- 2.3.0
Beta Was this translation helpful? Give feedback.
All reactions