Jiayi Zhao
Jiayi Zhao
I see, them I misunderstand. For Vertex, before our PR is in, you can try add that code to KubeflowV2DagRunner.run, and retry?
next release about a month
Could you open a separate issue under [tft](https://github.com/tensorflow/transform)? Thanks!
[This](https://github.com/tensorflow/tfx/pull/4620) should fix it somehow cluster_spec is set differently when there are multiple worker vs single worker
The fix is in and will be in next release
For the single worker case, you can use Tuner (run in the same worker as component) instead of CloudTuner (a remote job)
Here is an [example](https://github.com/tensorflow/tfx/blob/master/tfx/examples/chicago_taxi_pipeline/taxi_pipeline_kubeflow_gcp.py#L273) of custom config
Some context, needs TFMA release to work https://github.com/tensorflow/tfx/pull/4533
> Seems the CI job is failing due to pickling issues - https://github.com/tensorflow/tfx-addons/runs/5340812545?check_suite_focus=true I think it needs TFMA head / next release to work
> > I think it needs TFMA head / next release to work > > You can modify requirements.txt to point to TFMA HEAD so we can make sure its...