Skip to content

Latest commit

 

History

History
105 lines (62 loc) · 6.2 KB

speech-container-configuration.md

File metadata and controls

105 lines (62 loc) · 6.2 KB
Raw
title titleSuffix description author manager ms.service ms.topic ms.date ms.author
Configure Speech containers
Azure AI services
Speech service provides each container with a common configuration framework, so that you can easily configure and manage storage, logging, and security settings for your containers.
eric-urban
nitinme
azure-ai-speech
how-to
1/21/2024
eur

Configure Speech service containers

Speech containers enable customers to build one speech application architecture that is optimized to take advantage of both robust cloud capabilities and edge locality.

The Speech container runtime environment is configured using the docker run command arguments. This container has some required and optional settings. The container-specific settings are the billing settings.

Configuration settings

[!INCLUDE Container shared configuration settings table]

Important

The ApiKey, Billing, and Eula settings are used together, and you must provide valid values for all three of them; otherwise your container won't start. For more information about using these configuration settings to instantiate a container, see Billing.

ApiKey configuration setting

The ApiKey setting specifies the Azure resource key used to track billing information for the container. You must specify a value for the ApiKey and the value must be a valid key for the Speech resource specified for the Billing configuration setting.

This setting can be found in the following place:

  • Azure portal: Speech Resource Management, under Keys

ApplicationInsights setting

[!INCLUDE Container shared configuration ApplicationInsights settings]

Billing configuration setting

The Billing setting specifies the endpoint URI of the Speech resource on Azure used to meter billing information for the container. You must specify a value for this configuration setting, and the value must be a valid endpoint URI for a Speech resource on Azure. The container reports usage about every 10 to 15 minutes.

This setting can be found in the following place:

  • Azure portal: Labeled Endpoint on the Speech overview page
Required Name Data type Description
Yes Billing String Billing endpoint URI. For more information on obtaining the billing URI, see billing. For more information and a complete list of regional endpoints, see Custom subdomain names for Azure AI services.

Eula setting

[!INCLUDE Container shared configuration eula settings]

Fluentd settings

[!INCLUDE Container shared configuration fluentd settings]

HTTP proxy credentials settings

[!INCLUDE Container shared HTTP proxy settings]

Logging settings

[!INCLUDE Container shared configuration logging settings]

Mount settings

Use bind mounts to read and write data to and from the container. You can specify an input mount or output mount by specifying the --mount option in the docker run command.

The Standard Speech containers don't use input or output mounts to store training or service data. However, custom speech containers rely on volume mounts.

The exact syntax of the host mount location varies depending on the host operating system. Additionally, the host computer's mount location might not be accessible due to a conflict between permissions used by the docker service account and the host mount location permissions.

Optional Name Data type Description
Not allowed Input String Standard Speech containers don't use this. Custom speech containers use volume mounts.
Optional Output String The target of the output mount. The default value is /output. This is the location of the logs. This includes container logs.

Example:
--mount type=bind,src=c:\output,target=/output

Volume mount settings

The custom speech containers use volume mounts to persist custom models. You can specify a volume mount by adding the -v (or --volume) option to the docker run command.

Note

The volume mount settings are only applicable for custom speech to text containers.

Custom models are downloaded the first time that a new model is ingested as part of the custom speech container docker run command. Sequential runs of the same ModelId for a custom speech container uses the previously downloaded model. If the volume mount isn't provided, custom models can't be persisted.

The volume mount setting consists of three color : separated fields:

  1. The first field is the name of the volume on the host machine, for example C:\input.
  2. The second field is the directory in the container, for example /usr/local/models.
  3. The third field (optional) is a comma-separated list of options, for more information, see use volumes.

Here's a volume mount example that mounts the host machine C:\input directory to the containers /usr/local/models directory.

-v C:\input:/usr/local/models

Next steps