Openvino Implementation

[Nisyhaal]

Added openvino implementation - an AI inferencing accelerator. Can be used with "--openvino" flag

[what-the-diff]

PR Summary

• Modified the .gitignore file:
  We've updated the file with additional items, ensuring we don't track certain types of files in our version control system. This helps to keep our codebase clean and focused on the essential code elements.

• Introduced a pyproject.toml file:
  This newly added file holds information about project dependencies, build system configurations and details about the project itself. This file essentially streamlines project management and setup process by centralizing this metadata.

• Removed the setup.py file:
  The traditional installation file (setup.py) has been eliminated because we've modernized our setup approach via the new pyproject.toml file.

• Updated the cli.py file with new functions and imports:
  New functionalities have been added to this file, notably enabling usage of the 'OpenVINO' inference pipeline. This translates to improved capabilities of our application to make sophisticated predictions.

  • Notable Imports: Imported functionalities from optimum.intel.openvino and transformers, enhancing efficiency and the range of operations.

  • Added Functions: Functions for handling OpenVINO pipeline written transcripts have been added. This means our program is now more versatile in dealing with different file types.

  • OpenVINO Integration: The application can now use OpenVINO when running pipeline for inferences, which would lead to faster and more efficient processing.

• Devised two new functions write_vtt_openvino and write_srt_openvino in the utils.py file:
  These add-on functions boost the program's capacity to handle OpenVINO-generated transcripts. They ensure that our code can effectively translate and store the predictions in the mentioned file formats, widening its utility.

[Nisyhaal]

Added toml file for packaging