RuntimeError: Cannot find new config in BaseSearcher, even after 100 trials

I am getting RuntimeError: Cannot find new config in BaseSearcher, even after 100 trials when using medium_quality_faster_inference as preset in ImagePredictor.

Colab notebook https://colab.research.google.com/drive/1ekpD-r29DqtDiPXCy9zHHA8LTKLeTDrN?usp=sharing

Error log

<ipython-input-5-b3c4aba9b9da> in <module>()
      1 predictor = ImagePredictor(verbosity=1)
      2 # since the original dataset does not provide validation split, the `fit` function splits it randomly with 90/10 ratio
----> 3 predictor.fit(train_dataset,tuning_data=test_dataset, presets=['medium_quality_faster_inference'],hyperparameters={'epochs': 1})  # you can trust the default config, we reduce the # epoch to save some build time

1 frames
/usr/local/lib/python3.7/dist-packages/autogluon/vision/configs/presets_configs.py in _call(*args, **kwargs)
     13         def _call(*args, **kwargs):
     14             gargs, gkwargs = set_presets(preset_name, *args, **kwargs)
---> 15             return f(*gargs, **gkwargs)
     16         return _call
     17     return _unpack_inner

/usr/local/lib/python3.7/dist-packages/autogluon/vision/predictor/predictor.py in fit(self, train_data, tuning_data, time_limit, presets, hyperparameters, **kwargs)
    375                 self._classifier = task.fit(train_data, tuning_data, 1 - holdout_frac, random_state)
    376             if err.exc_value is not None:
--> 377                 raise RuntimeError(err.exc_value + err.hint)
    378         self._classifier._logger.setLevel(log_level)
    379         self._classifier._logger.propagate = True

RuntimeError: Cannot find new config in BaseSearcher, even after 100 trials

[talhaanwarch]

Btw thanks @Innixma you are doing great. AG is quite an amazing package.

[talhaanwarch]

Please check this notebook, https://colab.research.google.com/drive/1lo9FIkiubR28ZOxTAkT7qMcB4-ECU5iU?usp=sharing The error is quite weird. By default preset is ['medium_quality_faster_train'], if I define it explicitly, I got the error. Otherwise, I did not