ls models

ls-models

List the models specifications (identifiers) known to the tools and which
can generally be used as an argument to the -model option of tools such as
train and evaluate. By default, the identifier, type and specification
details (e.g., javascript) are shown.
Usage : ... [-v | model-spec (model-spec ...)  ] 
Options:
  -v        : Print untrained Java model class detail.
  -q        : Only list the model identifiers.
Usage:
  ...              # With no args, shows all model specifications
  ... -q           # Lists only the model identifiers
  ... -v           # Adds the untrained Java class detail for models
  ... ensemble gmm # Shows only data for the listed model identifiers
  ... -v gmm       # Shows Java class details for the gmm model

capture

Captures, labels and stores sound clips.
Sounds may be stored locally and labeled in a metadata.csv file and/or
they can be stored in a named server under a specific name.
Options:
  -label name=value: specifies a name/value pair to stored as labels for all
               sounds captured during the run. More than one of this options
               can be provided to apply multiple labels to the sounds. If the
               -name option is not used, then the first label value will be
               used as the base name for stored files and files stored in the
               server. 
               NOTE: double-quotes must surround name=value on Windows.
  -dir dir : a directory in which to store sounds and metadata.csv.
               If a metadata.csv file already exists there, referenced sounds
               will not be overwritten and metadata.csv will be appended to.
  -n count : the number of sounds recordings to capture. Default is unlimited.
  -name name : the name used to store sounds in the server or the base name of 
               the file used to stores sounds under.  If not provided then the
               value of the first label is used.
Usage: 
  ... -label status=normal -dir .
  ... -label status=normal -label source=fan -dir .