Skip to content

v1.6.0

Choose a tag to compare

@docling-ops docling-ops released this 04 Jul 16:11
· 126 commits to main since this release

Features

Dense Extraction Pipeline
  • Add autonomous skeleton-then-flesh contract and wire dense into backend, CLI & Python API (2475a85)
  • Add series & conditional schema primitives and refine skeleton prompt for granularity and global & singleton entities (cc476d5)
  • Add opt in barren-branch pruning (9d0669f)
Deterministic Data Grounding
  • Implement deterministic data grounding and document lineage tracking (c21efc0)

Bug Fixes

Orchestration & Key Resolution
  • Resolve canonical key fallback divergence to prevent edge-attachment collapse (eef791a)
  • Resolve sparse root-id drift, enforce specific instance labels, and add direct-mode hints (186386e)
Pipeline Telemetry
  • Resolve telemetry edge-cases, enhance entity-resolution, and export chunks (32e7569)
  • Optimize docling graph extraction resilience and token economics (e2bd17a)
Linting & Validation
  • Clean up dead type guards and resolve mypy warnings (1ecc28f)
Corporate Compliance
  • Rename WatsonX to watsonx per official branding, closes #72 (2a1c375)

Refactoring

Extraction Modernization
  • Remove delta and staged extraction code base and references (25b7e3c)
  • Collapse dense extraction surface to intent primitives and fix core telemetry leaks (bb18f0c)
  • Remove dead code, fix dense config propagation and improve performance (7e85841)

Documentation

Schema & System Guidance
  • Document chunk schemas, advanced reporting, and extraction templates (fa907d7)
  • Document data grounding architecture and provenance schema contracts (ef69c6b)
  • Optimize pydantic templates for pipeline mechanisms and overhaul best practices (4c734ea)