fix:Improve scaler handling in predict and train methods

[AlejandroOrozcoValero]

Fix scaler handling in predict method

This PR fixes the scaler loading issue that was causing errors in the inference process.

Changes:

• Uncommented scaler path assignment: The line that sets the scaler file path is now active
• Uncommented scaler loading: The code that loads the scaler from disk is now executed
• Fixed conditional logic: Updated the if statement to properly handle scaler loading

Impact:

These changes resolve the errors that were occurring when running LFP_developing_brain.py, allowing the prediction process to complete successfully.