Refactor data classes

[jhj0517]

Summarize Changes

1. Renamed files whisper_base.py to base_transcription_pipeline.py and whisper_parameter.py to data_classes.py
2. Several functions / classes are renamed along with change 1
3. Separate data classes by features
4. Use to_gradio_inputs() for each data class for gradio inputs
5. Use WhisperImpl enum for each whisper implementation
6. Handle None values with constants that cannot be displayed as None in the gradio UI ( See : None type initial value is not possible in gr.Textbox() gradio-app/gradio#8723)