myvouvis

Herbarijní DwC databot: celý arch → LM2 detekce primárního štítku → vision LLM (e-infra) → Darwin Core JSON do databots.databot_results (stejný kontrakt jako coco-bbox-detector).

Rychlý start

Python 3.13 recommended:

python3.13 -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
export PYTHONPATH=src

nebo Poetry:

poetry env use python3.13
poetry install
export OPENAI_API_KEY=...
export OPENAI_BASE_URL=https://llm.ai.e-infra.cz/v1/
# LM2 váhy jsou v vendor/lm2/weights/best.pt (git lfs pull)
PYTHONPATH=src poetry run python -m cli test_image/test.png

Viz docs/getting-started.md.

Struktura

Cesta	Účel
src/pipeline/sheet.py	Jádro process_sheet()
src/bots/implementations/herbarium_dwc_databot.py	Batch databot adapter
src/api/app.py	HTTP /v1/transcribe-full, /v1/transcribe-dwc, /v1/transcribe-bbox
src/vendor/lm2.py	Jediný most k LM2 YOLOv5
vendor/lm2/	component_detector/ + weights/best.pt (Git LFS)

Výstup do DB: result_data.dwc + result_data.validation + LM2 bboxy (detections, primary_label).

LM2 detektor je vendored v vendor/lm2/ (viz NOTICE.md).