dspk run

Run a pipeline spec (.json or .toml) over its inputs, or inputs given here

dspk run [OPTIONS] <SPEC>

Arguments

• <SPEC>
  Pipeline spec, .json or .toml; paths in it are relative to its folder

Options

• -i, --input <PATH>...
  Input recordings, replacing the spec’s: files or globs (quote a glob to let dspk expand it, sorted); several inputs run one after the other
• --merge <PATH> <PATH>...
  Recordings read as one input, channels side by side (same rate and length), replacing the spec’s inputs
• -o, --output <PATH>
  Output path, replacing the spec’s; with several inputs it needs {stem} or {index}
• --mode <MODE>
  Execution mode, replacing the spec’s
  • auto: Stream where the stages allow it, else independent windows
  • stream: Windows in order with filter state carried (equal to the whole recording at once)
  • windows: Independent windows with halos of context
• --chunk-samples <N>
  Samples per channel of each window, replacing the spec’s
• -n, --dry-run
  Show what would run (inputs, outputs, mode, windows, bytes) and stop
• --json
  Print the plan (with –dry-run) or the reports as JSON on stdout, nothing else