`neuralforecast` should allow to passthrough tune `runConfig`

[mcanouil]

Description

Currently, the neuralforecast uses default values for runConfig() in the tuning function.
Unfortunately, the default of Ray is bad as the logging is set to write outside the working directory, in user's home. I don't believe there is any good reason to store these logging there, but regardless of my opinion, users should have control over this when using AutoNHITS, etc.

• neuralforecast/neuralforecast/common/_base_auto.py

  Lines 246 to 257 in fbb012c

  	tuner = tune.Tuner(
  	tune.with_resources(train_fn_with_parameters, device_dict),
  	run_config=air.RunConfig(callbacks=self.callbacks, verbose=verbose),
  	tune_config=tune.TuneConfig(
  	metric="loss",
  	mode="min",
  	num_samples=num_samples,
  	search_alg=search_alg,
  	trial_dirname_creator=trial_dirname_creator,
  	),
  	param_space=config,
  	)

neuralforecast should at least allow to change the storage path and, better, set a default that is localised within the working directory.

• https://docs.ray.io/en/latest/tune/tutorials/tune-output.html#how-to-configure-logging-in-tune
• https://docs.ray.io/en/latest/tune/api/doc/ray.tune.RunConfig.html#ray.tune.RunConfig

Use case

No response