badc.cli.main

Primary Typer CLI for the Bird Acoustic Data Compiler project.

Functions

chunk_manifest(file[, chunk_duration, ...])	Create a manifest CSV describing fixed-duration chunks.
chunk_orchestrate([dataset, pattern, ...])	Plan chunk jobs across an entire dataset without touching audio files.
chunk_probe(file[, initial_duration, ...])	Estimate chunk duration feasibility for a single audio file.
chunk_run(file, chunk_duration[, overlap, ...])	Write chunk WAVs (optional) and a manifest for downstream inference.
chunk_split(file[, chunk_duration])	List placeholder chunk identifiers for an audio file.
data_connect(name[, path, url, method, ...])	Clone or update a DataLad dataset and record its metadata.
data_disconnect(name[, drop_content, path, ...])	Mark a dataset as disconnected and optionally drop its contents.
data_status([details, show_siblings])	Report all datasets tracked in ~/.config/badc/data.toml.
infer_aggregate(detections_dir[, output, ...])	Aggregate per-chunk detection JSON files into a summary CSV.
infer_monitor([log_path, tail, follow, interval])	Render GPU/telemetry summaries for a run-specific log.
infer_orchestrate([dataset, manifest_dir, ...])	Plan inference runs across manifests without executing them.
infer_run(manifest[, max_gpus, output_dir, ...])	Run HawkEars (or a custom runner) for every chunk in a manifest.
infer_run_config(config[, print_datalad_run])	Execute badc infer run using a TOML configuration file.
list_gpus()	Display GPU inventory as reported by nvidia-smi.
main()	Entrypoint invoked by the badc console script.
pipeline_run([dataset, chunk_plan, ...])	Run chunking + inference (+ optional bundle) in a single command.
report_aggregate_dir([aggregate_dir, limit, ...])	Summarize detections across every Parquet bundle in an aggregate directory.
report_bundle(parquet[, output_dir, ...])	Produce quicklook, parquet, and DuckDB artifacts in one pass.
report_duckdb(parquet[, database, ...])	Materialize detections into DuckDB and print ready-to-review summaries.
report_parquet(parquet[, top_labels, ...])	Summarize canonical detections via DuckDB with optional exports.
report_quicklook(parquet[, top_labels, ...])	Generate quicklook tables/plots for canonical detections.
report_summary(parquet[, group_by, output, ...])	Summarize detections via DuckDB (group counts + avg confidence).
telemetry_monitor([log_path])	Print the most recent telemetry records for quick inspection.
version()	Display the current BADC version banner.