[Mernit]

You should look into beam.cloud (I'm the founder, but it's pretty great)

It lets you quickly run long-running jobs on the cloud by adding a simple decorator to your Python code:

  from beam import function

  # Some long training function
  @function(gpu="A100-80")
  def handler():
    return {}

  if __name__ == "__main__":
    # Runs on the cloud
    handler.remote()

Fully open-source also.