Skip to content

v0.3.8

Choose a tag to compare

@rstrahan rstrahan released this 18 Jul 22:10
· 2742 commits to main since this release

[0.3.8]

Added

  • Lending Package Configuration Support for Pattern-2

    • Added new lending-package-sample configuration to Pattern-2, providing comprehensive support for lending and financial document processing workflows
    • New default configuration for Pattern-2 stack deployments, optimized for loan applications, mortgage processing, and financial verification documents
    • Previous rvl-cdip-sample configuration remains available by selecting rvl-cdip-package-sample for the Pattern2Configuration parameter when deploying or updating stacks
  • Text Confidence View for Document Pages

    • Added support for displaying OCR text confidence data through new TextConfidenceUri field
    • New "Text Confidence View" option in the UI pages panel alongside existing Markdown and Text views
    • Fixed issues with view persistence - Text Confidence View button now always visible with appropriate messaging when content unavailable
    • Fixed view toggle behavior - switching between views no longer closes the viewer window
    • Reordered view buttons to: Markdown View, Text Confidence View, Text View for better user experience
  • Enhanced OCR DPI Configuration for PDF files

    • DPI for PDF image conversion is now configurable in the configuration editor under OCR image processing settings
    • Default DPI improved from 96 to 150 DPI for better default quality and OCR accuracy
    • Configurable through Web UI without requiring code changes or redeployment

Changed

  • Converted text confidence data format from JSON to markdown table for improved readability and reduced token usage

    • Removed unnecessary "page_count" field
    • Changed "text_blocks" array to "text" field containing a markdown table with Text and Confidence columns
    • Reduces prompt size for assessment service while improving UI readability
    • OCR confidence values now rounded to 1 decimal point (e.g., 99.1, 87.3) for cleaner display
    • Markdown table headers now explicitly left-aligned using |:-----|:-----------| format for consistent appearance
  • Simplified OCR Service Initialization

    • OCR service now accepts a single config dictionary parameter for cleaner, more consistent API
    • Aligned with classification service pattern for better consistency across IDP services
    • Backward compatibility maintained - old parameter pattern still supported with deprecation warning
    • Updated all lambda functions and notebooks to use new simplified pattern
  • Removed fixed image target_height and target_width from default configurations, so images are processed in original resolution by default.

  • Updated Default Configuration for Pattern1 and Pattern2

    • Changed default configuration for new stacks from "default" to "lending-package-sample" for both Pattern1 and Pattern2
    • Maintains backward compatibility for stack updates by keeping the parameter value "default" mapped to the rvl-cdip-sample for pattern-2.
  • Reduce assessment step costs

    • Default model for granular assessment is now us.amazon.nova-lite-v1:0 - experimentation recommended
    • Improved placement of <> tags in assessment prompt to improve utilization of prompt caching

Fixed

  • Fixed Image Resizing Behavior for High-Resolution Documents
    • Fixed issue where empty strings in image configuration were incorrectly resizing images to default 951x1268 pixels instead of preserving original resolution
    • Empty strings ("") in target_width and target_height configuration now preserve original document resolution for maximum processing accuracy
  • Fixed issue where PNG files were being unnecessarily converted to JPEG format and resized to lower resolution with lost quality
  • Fixed issue where PNG and JPG image files were not rendering inline in the Document Details page
  • Fixed issue where PDF files were being downloaded instead of displayed inline
  • Fixed pricing data for cacheWrite tokens for Amazon Nova models to resolve innacurate cost estimation in UI.