feat: [WIP] Integration of SmolDocling#708

Draft

maxmnemonic wants to merge 14 commits intomainfrom mly/smol-docling-integration

+846-3

Commits on Jan 16, 2025

Skeleton for SmolDocling model and VLM Pipeline

cau-git
authored and
Maksym Lysak
committed
wip smolDocling inference and vlm pipeline
Maksym Lysak
committed
WIP, first working code for inference of SmolDocling, and vlm pipeline assembly code, example included.
Maksym Lysak
committed
Fixes to preserve page image and demo export to html
Maksym Lysak
committed
Enabled figure support in vlm_pipeline
Maksym Lysak
committed
Fix for table span compute in vlm_pipeline
Maksym Lysak
committed
Properly propagating image data per page, together with predicted tags in VLM pipeline. This enables correct figure extraction and page numbers in provenances
Maksym Lysak
committed
Cleaned up logs, added pages to vlm_pipeline, basic timing per page measurement in smol_docling models
Maksym Lysak
committed
Replaced hardcoded otsl tokens with the ones from docling-core tokens.py enum
Maksym Lysak
committed
Added tokens/sec measurement, improved example
Maksym Lysak
committed
Added capability for vlm_pipeline to grab text from preconfigured backend
Maksym Lysak
committed
Exposed "force_backend_text" as pipeline parameter
Maksym Lysak
committed
Flipped keep_backend to True for vlm_pipeline assembly to work
Maksym Lysak
committed

Commits on Jan 17, 2025

Updated vlm pipeline assembly and smol docling model code to support updated doctags
Maksym Lysak
committed