swirec_acoustic_adapt_suppress_acoustic_model_update

Stops self-learning adaptation of recognition models.

Value	Boolean 0—Recognizer performs updates of acoustic models. 1—Recognizer collects recognition statistics, but does not perform updates.
Default	0 (updates enabled) By default, Recognizer updates models every 24 hours at the time specified by swirec_acoustic_adapt_model_update_time.
How to set	In Management Station set on the Nuance recognition service. If not using Management Station, set in a Recognizer configuration file (User-nrsxx.xml).
Usage	Seldom used. Use this parameter when you have multiple Recognizer hosts and require identical Recognizerinstallations on each machine. Using this parameter ensures that the Recognizer’s acoustic models remain unchanged until a single, simultaneous update for all machines is performed.

This parameter does not stop the collection of self-learning statistics. To do that, see swirec_acoustic_adapt_adapt_model and swirec_acoustic_adapt_suppress_adaptation

Without this parameter, the acoustic models adapt independently on each machine, and recognition accuracy of some machines improves faster on some machines.

For an overview of adaptation, see Controlling the self-learning recognizer.

Related parameters:

• swirec_acoustic_adapt_adapt_model
• swirec_acoustic_adapt_min_num_utts
• swirec_acoustic_adapt_model_update_time
• swirec_acoustic_adapt_num_archive
• swirec_acoustic_adapt_rate
• swirec_acoustic_adapt_root
• swirec_acoustic_adapt_suppress_adaptation