Dynamic Train_Run_ID in inference pipeline

[kyle-musser]

Hello,

I am using the solutions accelerator (AutoML) to run a train and inference pipeline that i am hoping to publish.
Is there a way to pass the most recent training run ID as a parameter to the inference pipeline?
Right now as far as i can tell i have to register the pipeline with a static train_run_id value and cant add a pipeline parameter than i can change dynamically from inside Azure Synapse when the pipeline is called on a daily basis

[bszewczyk]

I also want to first train data and then run inference pipeline. I try to find solution how to pass latest run id to it but I found nothing. Did you figure out something?