- [ ] Setting up the best optimized CPU strategy for a faster inference. - [ ] Improving the windowing strategy in terms of token length and number of windows. - [ ] Improving the subtoken length decision engine strategy for a smarter and faster usage. - [ ] Setting up a better strategy for already semantically good identifiers for a shorter decision timeline span.