Ok, so you circumvented this issue by using custom containers for performing batch inference with your model. Good to know you found a way out. I feel like Vertex-AI batch prediction using foundational non Google models is not too well documented. Fo...

Hi there, were you able to make progress on this issue?