Golem can provide a computation backend for your applications.
In this task, you will learn how to utilize Golem to offload your system.
Build a web application that will accept mp3 files, (split them into smaller chunks), perform STT conversion, and display the transcription.
You can use an image prepared by Golem with whisper
.
Please observe how long it takes to complete the technical part of the tasks. We will ask you for such information in the feedback form.
-
Create a web page that will offer UI to upload wav files and display the output text after STT conversion.
-
Create a server that will accept the wav files and will do STT conversion on the Golem Network.
-
Add changes in the
use-Golem-as-backend
directory. -
Fill out the feedback form and publish changes in your repository fork.
-
The image's
hash
withwhisper
ise8a84adfcc19d1df939a626004b64b50fa0729664e40856933d708ed
. -
The image had VOLUME set as
/golem/work
. -
The
whisper
tool usage is:whisper audio.wav --model tiny --language en
. -
Usage of
Golem Network
is demonstrated in REST API example with Express.js. -
You can always ask on the Discord
#js-discussion
channel in the SDK section.