[fix] High chunk_overlap causes last chunk to be entirely redundant

[bhavnicksm]

This pull request introduces changes to the src/chonkie/chunker/token.py file to improve the chunking process by refactoring how token groups are generated. The main change involves moving the _token_group_generator method to a more appropriate location within the file and integrating its usage more consistently across the codebase.

Key changes include:

• Refactoring _token_group_generator:
  • Moved _token_group_generator method to a more appropriate location within the file.
  • Updated the chunk method to use _token_group_generator for calculating token groups.
  • Modified the _process_text_batch method to use _token_group_generator for generating token groups.