TabularModel.fit( ) creates a directory saved_models in the project folder which stores .ckpt

[pranavvp16]

I want to know if there any is way to avoid this behaviour . I don't want the tabular_model.fit( ) to create the saved_models directory by default in my project folder

[manujosephv]

I thought I had answered this. Maybe I missed.
So by default, PyTorch Tabular enables a checkpoint save callback, which saves to saved_models

You can configure this through the TrainerConfig with the parameter checkpoints_path