Configuration Script
Following article describes the configure-speech-platform.sh script which is
included in Virtual Appliance.
Introduction
This script serves the purpose of configuring Phonexia Speech Platform 4 so that
you don't have to configure the values.yaml file yourself. More specifically,
the script enables licensed technologies and enables GPU use. For more
information, see the flag table below.
Usage
This section describes how to use the configure-speech-platform.sh script.
Calling the script
The script is located in the /root/scripts/ directory. If you are not using
root user, please first run the command, and after you are prompted, enter
password for the root user:
su root
After you are logged in as the root user, you can use the script by calling the command:
/root/scripts/configure-speech-platform.sh <extra flags>
Flags
| Name | Shortcut | Description |
|---|---|---|
‑‑auto‑configure | -a | Configures the system automatically based on licenses and models in /data/ directory. If GPU is present also configures licensed technologies to run on GPU |
‑‑disable | -d | Disables a technology. |
‑‑enable | -e | Enables a technology. |
‑‑filebrowser | Used only during automatic runs inside the Filebrowser pod. Changes how GPU presence is checked. Do not use when running manually! | |
‑‑gpu | If possible, runs technologies on the GPU. | |
‑‑help | -h | Prints out help page. |
‑‑microservices | -m | List of technologies to be configured. |
‑‑model‑names | Technologies model names. Can be specified only with one technology. Must match the number of model versions parameter arguments. When used with speech-to-text-phoneixa, audio-quality-estimation and time-analysis, multiple models can be specified | |
‑‑model‑versions | Technologies model versions. Can be specified only with one technology. Must match the number of model names parameter arguments. When used with speech-to-text-phonexia, audio-quality-estimation and time-analysis multiple models can be specified. | |
‑‑no‑gpu | Runs technologies without GPU. | |
‑‑no‑on‑demand | Runs technologies as a stable replica. | |
‑‑on‑demand | If possible, runs technologies on demand. | |
‑‑replica‑count | Replica count of technologies. You need to specify the exact model for which the replica count applies, | |
‑‑technologies | -t | List of technologies, which will be configured. |