[FEA]: Add nightly accuracy testing for text, image, table, and chart extraction

Is this a new feature, an improvement, or a change to existing functionality?

New Feature

How would you describe the priority of this feature request

Currently preventing usage

Please provide a clear description of problem this feature solves

Implement a nightly testing pipeline to evaluate the accuracy and performance of our extraction processes for various document types. The testing should focus on extracting text, images, tables, and charts, and should cover multiple formats including jpeg, png, svg, jpg, docx, pptx, txt, pdf, and subsets of HTML.

Describe the feature, and optionally a solution or implementation and any alternatives

• Capture Performance Metrics

  • Measure and capture time to completion for each document type.
  • Record pages per second throughput for each document type.

• Generate Accuracy Metrics

  • Calculate F1 scores for text, image, table, and chart extraction for each document type.
  • Calculate Recall for text, image, table, and chart extraction for each document type.
  • Calculate Precision for text, image, table, and chart extraction for each document type.

Additional context

No response

No files, tests, or entry points are named in the issue. Start by locating the existing extraction and testing infrastructure, then determine how the listed document formats and extraction types can be evaluated nightly; done means the pipeline records throughput, completion time, F1, recall, and precision.