Run an Experiment on Hybrid Mode

Run NNI on hybrid mode means that NNI will run trials jobs in multiple kinds of training platforms. For example, NNI could submit trial jobs to remote machine and AML simultaneously.

Setup environment

NNI has supported local, remote, PAI, and AML for hybrid training service. Before starting an experiment using these mode, users should setup the corresponding environment for the platforms. More details about the environment setup could be found in the corresponding docs.

Run an experiment

Use examples/trials/mnist-tfv1 as an example. The NNI config YAML file’s content is like:

experimentName: MNIST
searchSpaceFile: search_space.json
trialCommand: python3
trialCodeDirectory: .
trialConcurrency: 2
trialGpuNumber: 0
maxExperimentDuration: 24h
maxTrialNumber: 100
  name: TPE
    optimize_mode: maximize
  - platform: remote
      - host:
        user: bob
        password: bob
  - platform: local

To use hybrid training services, users should set training service configurations as a list in trainingService field. Currently, hybrid support setting local, remote, pai and aml training services.