PBT update config in custom explore function

I am new to ray and tune. I want to use PBT with a dynamic config. For example,

def explore(config):
    # here config['varA'] = None not something as in TrainMNIST._setup
    return config

class TrainMNIST(Trainable):
    def _setup(self, config):
        config['varA'] = something
        self.config.update(config)

pbt = PopulationBasedTraining(
        ...,        
        custom_explore_fn=explore)

tune.run(
        TrainMNIST,
        name="exp",
        scheduler=pbt,
        stop={
            "test_acc": 0.99,
            "training_iteration": 100,
        },
        resources_per_trial={
            "cpu": 2,
            "gpu": 0.25,
        },
        **{"config":
            { "args": vars(args),
              "varA": None,
              }
        }

How can I explore with config[‘varA’] = something?

[heurainbow]

I found a way to work aound this by adding trial.config.update(result[‘config’]) in on_trial_result. I use this to make the explore function adapt to each model’s performance. Say, for better performance I use small step for hyperparams.

Still, I think synchronize config is essential for robustness.

    def on_trial_result(self, trial_runner, trial, result):
        trial.config.update(result['config'])
        ...

[heurainbow]

update config when getting results