SpeechLanguageResources
A comma-separated list of sections in the configuration file that contain information about the speech language resources to load when Media Server starts.
In the sections that you specify, use the following parameters to define the resource to load:
- LanguagePack must be set.
- ModelVersion is optional, but the default value is
legacy
. - Set CustomLanguageModel to define the custom language model(s) to include in the resource (legacy models only).
- Set CustomWordDatabase if you want to include a custom word database (legacy models only).
Ensure that the values of these parameters match those you intend to use in your speech-to-text analysis task. For information about why you might want to pre-load language resources, see Pre-Load Language Resources.
Type: | String (comma-separated list) |
Default: | |
Required: | No |
Configuration Section: | PersistentData |
Example: |
The following example configures Media Server to load two language resources when it starts. The second resource includes a custom language model. [PersistentData] SpeechLanguageResources=BritishEnglish,LegacyWithCustomLM [BritishEnglish] LanguagePack=ENUK ModelVersion=small [LegacyWithCustomLM] LanguagePack=ENUK CustomLanguageModel=MedicalTerms:0.1 |
See Also: |