Skip to content

v0.2.0

Choose a tag to compare

View all tags
@FWao FWao released this 11 Aug 09:07
· 22 commits to main since this release
v0.2.0
2bcdfce

Added

  • VISION_OCR_DIALECT selects the vision-OCR model family (unlimited_ocr,
    chandra, plain) and supplies the prompt, token and body defaults — see
    OCR engines.

  • VISION_OCR_PROFILES offers several OCR models at once, one pickable per
    document in the advanced settings — see
    OCR engines.

  • compose.chandra.yml runs a datalab chandra GPU OCR sidecar, as
    compose.unlimited-ocr.yml does for Unlimited-OCR — see
    OCR engines.

Changed

  • .env.example is now a short setup worksheet that ships with the LLM
    detector enabled: fill in OPENAI_API_BASE/LLM_MODEL before first start
    (or set DETECTORS=rules for a look without an LLM) — see
    Configuration.

  • LLM_REQUEST_TIMEOUT_SECONDS now defaults to 600 instead of 120 — see
    Configuration.

Full Changelog: v0.1.3...v0.2.0

Assets 2
Loading