Jiayi Zhao
Jiayi Zhao
Thanks for the advice!! 1. TFX itself currently doesn't support retry, and outside retry might mess up TFX mlmd status (e.g., 2 output artifacts instead of 1, this is because...
for the 60s, can we confirm if that is taken by MLMD? There are some costs for e.g., container setup if running on Cloud. Which DagRunner are you using?
Could you try 1.6.1? Sanjay, do you know why 1.5 test didn't catch this?
Is retry a functionality provided by airflow? seems partial code is retried
“Seems MLMD is reusing context and registering 2 output artifacts for the same component run instead of overwriting it on the 2nd run” this is expected, MLMD won't do overwriting....
My current guess is, TFX uses MLMD to check the status, if previous execution failed before publishing results to MLMD (previous artifact is pending), retry shouldn't cause any issue but...
For long queries, there are multiple limitations, e.g., mlmd exec properties field size, tfx ir size etc. Consider create a custom component that instead of passing in the query, it...
what information are you trying to pass through component_config?
Hi, InteractiveContext is mainly for local debugging, you can follow https://www.tensorflow.org/tfx/tutorials/tfx/cloud-ai-platform-pipelines
thanks, will do