Aman Nalakath
a0211155fa
- Slim core deps: move ML stack to optional extras (paddle/tables/figures/scientific/full) - Lazy settings proxy with config search paths (env var, cwd, user dir) - New commands: init, demo, setup [basic|full], first-run guard on run/watch - Engine-aware preprocessing: Paddle gets original image (fixes dark mode 0.83->0.95) - Results table shows Skipped count; lazy run-dir creation - kg_ocr marked experimental with extra, Docker defaults with OCR_PIPELINE_CONFIG - 25/25 tests, ruff clean
106 lines
2.3 KiB
YAML
106 lines
2.3 KiB
YAML
# Full-stack config used for this repo's development. New users: run
|
|
# `ocr-pipeline init` to generate a minimal config instead of copying this file.
|
|
input:
|
|
paths:
|
|
- "~/Pictures"
|
|
patterns:
|
|
- "SCR-*.png"
|
|
- "*.jpg"
|
|
- "*.jpeg"
|
|
- "*.tiff"
|
|
- "*.bmp"
|
|
recursive: true
|
|
exclude_patterns:
|
|
- "*.photoslibrary/*"
|
|
|
|
ocr:
|
|
engine: "paddleocr"
|
|
languages: ["en", "latin"]
|
|
use_gpu: false
|
|
use_angle_cls: true
|
|
det_db_thresh: 0.3
|
|
det_db_box_thresh: 0.6
|
|
det_db_unclip_ratio: 1.5
|
|
rec_batch_num: 6
|
|
cpu_threads: 4
|
|
|
|
processing:
|
|
workers: 4
|
|
batch_size: 10
|
|
retry_attempts: 2
|
|
timeout_per_image: 120
|
|
skip_existing: true
|
|
|
|
detectors:
|
|
figures:
|
|
enabled: true
|
|
model: "lp://PubLayNet/faster_rcnn_R_50_FPN_3x/config"
|
|
confidence_threshold: 0.7
|
|
tables:
|
|
enabled: true
|
|
model: "microsoft/table-transformer-detection"
|
|
confidence_threshold: 0.7
|
|
captions:
|
|
enabled: true
|
|
model: "lp://PubLayNet/faster_rcnn_R_50_FPN_3x/config"
|
|
confidence_threshold: 0.5
|
|
|
|
entities:
|
|
enabled: true
|
|
model: "en_core_sci_lg"
|
|
types:
|
|
- "GENE"
|
|
- "PROTEIN"
|
|
- "CHEMICAL"
|
|
- "SPECIES"
|
|
- "DISEASE"
|
|
- "CELL_LINE"
|
|
- "ORGANISM"
|
|
- "CELL_TYPE"
|
|
merge_entities: true
|
|
|
|
citations:
|
|
regex_enabled: true
|
|
grobid_enabled: false
|
|
grobid_url: "http://localhost:8070"
|
|
grobid_timeout: 30
|
|
|
|
chunking:
|
|
chunk_size: 1000
|
|
chunk_overlap: 200
|
|
separators: ["\n\n", "\n", ". ", " ", ""]
|
|
keep_separator: true
|
|
|
|
output:
|
|
write_consolidated: true
|
|
consolidated_filename: "all_ocr.md"
|
|
# Uses platformdirs default: ~/.local/share/ocr-pipeline (Linux), ~/Library/Application Support/ocr-pipeline (macOS), %LOCALAPPDATA%\ocr-pipeline (Windows)
|
|
format: "markdown"
|
|
organize_by: "date_run"
|
|
frontmatter:
|
|
fields:
|
|
- "source_path"
|
|
- "source_hash"
|
|
- "timestamp"
|
|
- "ocr_engine"
|
|
- "ocr_confidence_mean"
|
|
- "language"
|
|
- "detected_entities"
|
|
- "entity_extraction_backend"
|
|
- "has_figures"
|
|
- "has_tables"
|
|
- "citations_found"
|
|
- "chunk_index"
|
|
- "total_chunks"
|
|
include_raw_text: false
|
|
|
|
watch:
|
|
enabled: true
|
|
debounce_seconds: 5
|
|
# Uses platform-aware defaults from config.py
|
|
# ignore_patterns:
|
|
# - ".DS_Store"
|
|
# - "*.tmp"
|
|
# - "*.partial"
|
|
# - "*.crdownload"
|
|
poll_interval: 1.0 |