Autogluon taking too long for sanity checking

Question

from autogluon.multimodal import MultiModalPredictor
predictor = MultiModalPredictor(label='finallikleyTOAST')
predictor.fit(train_data) 

I am using above code for multimodal through autogluon (structured + unstructured)

Number of features: 1258 (structured columns)
Text columns: 2

When I use the above code for multi-modal, the sanity checking takes too long, 1 hour in and the progress bar doesn't show the estimated time.

Is there a way to skip sanity checking? or any optimizations needed in the code?