WilData CLI Reference

Complete command-line interface reference for WilData.

Main Commands

import-dataset

Import dataset from various formats.

wildata import-dataset [SOURCE] [OPTIONS]

Options: - -f, --format TEXT: Source format (coco/yolo/ls) - -n, --name TEXT: Dataset name - -c, --config PATH: Config file path - --root PATH: Data root directory - --split TEXT: Split name (train/val/test) - --enable-tiling: Enable image tiling - --tile-size INTEGER: Tile size - -v, --verbose: Verbose output

Examples:

# From config
wildata import-dataset --config configs/import-config.yaml

# Direct arguments
wildata import-dataset data.json --format coco --name dataset1

bulk-import-datasets

Bulk import multiple datasets.

wildata bulk-import-datasets [OPTIONS]

Options: - -c, --config PATH: Bulk import config - -n, --num-workers INTEGER: Number of workers

create-roi-dataset

Create ROI classification dataset from detection annotations.

wildata create-roi-dataset [OPTIONS]

Options: - -c, --config PATH: ROI config file - --roi-size INTEGER: ROI box size - --random-count INTEGER: Background samples per image

dataset list

List all datasets.

wildata dataset list [--root PATH]

dataset export

Export dataset to format.

wildata dataset export <name> [OPTIONS]

Options: - --format TEXT: Target format (coco/yolo) - --output PATH: Output directory

visualize-dataset

Launch dataset visualization.

wildata visualize-dataset --dataset <name> --split <split>

update-gps-from-csv

Update image GPS metadata from CSV.

wildata update-gps-from-csv [OPTIONS]

Options: - -c, --config PATH: GPS update config - --image-folder PATH: Image folder - --csv PATH: CSV file path - --output PATH: Output directory

---

For detailed script documentation, see WilData Scripts.