Develop autotagging datalake processing

[SoniaGrh]

As a Data Scientist,
I want to develop a processing tool that automatically tags images in the datalake using a Visual Language Model (VLM) based on a predefined list of tags
So that we can efficiently categorize and search through the images in the datalake.

Acceptance Criteria:

• The tool takes as input either the entire datalake or a specified list of images from the datalake.
• The tool accepts a list of tags (e.g., "day", "night", "autumn", "winter", "summer", "spring").
• The tool runs the VLM on each image with the appropriate prompts to assign the most relevant tag from the provided list.
• The assigned tags are accurately associated with each image in the datalake.

Definition of Done:

• All acceptance criteria are met.
• Code is reviewed and approved.
• The tool is tested and validated with a variety of images and tag lists.
• Documentation is updated to reflect the new tagging process.