google cloud platform – What is the best scheduler for running deep learning training on GCP? Any recommendations?

I recently joined a new company and have been given access to a VM on GCP with 16 A100 GPUs. I want to efficiently use these resources for running my code. What would be the best way to go about this?
At my previous company, we had our own servers, so I used Conda and Slurm. However, at my current company, I can’t use Conda, so I’ve been creating Docker containers, installing the necessary packages, and running simple training jobs.
What tools or methods do you typically use for this kind of setup?