The Lang parameter sets either the name of the language pack used for operations in the module (speech-to-text, transcript alignment, and so on), or the name of the base language identification feature extraction pack(for language identification tasks). Specify the name of the configuration section in the IDOL Speech Server configuration file where the settings for the language pack or feature extraction model are configured.
stt module), set the parameter to the name of the language pack, for example ENUK. For the stt module, you must set the Lang configuration parameter using the corresponding action parameter.For transcript alignment (using the align module), set this name to the language pack to use for alignment.
For text normalization (using the textnorm module), set this name to the language pack to use for normalization. For the textnorm module, you must set the Lang configuration parameter using the corresponding action parameter.
For language model building (using the lmbuild module), set this name to the base language pack to use to create the custom language model. For the lmbuild module, you must set the Lang configuration parameter using the corresponding action parameter.
For language identification (using the lidfeature module), set this name to the feature extraction models used for language identification.
Note: A base language identification feature extraction pack, Syls, is provided with Speech Server and is configured for use with 16k broadband data. You should not need to change this setting unless you need to use Speech Server with 8k telephony data, in which case you must set Lang Syls-tel.
| Type: | String |
|---|---|
| Default: | None |
| Required: | Yes |
| Configuration Section: | align, lidfeature, lmbuild, splitspeech, stt, textnorm, and textsegment modules |
| Example: | Lang=ENUK
|
| See Also: | Lang (action parameter) |
|
|