Price: 0
Number of applications: 0
16.09.25 (inclusive)
Payment by card
Idea
R&D Tasks
Media sphere
Dubbing
marked up json files
We are fine-tuning the whisper model in order to teach it how to identify code switching between languages well. For this purpose, we need high-quality dialogues that need to be transcribed and timed.
We expect to receive well-marked json files suitable for use in the fine-tuning process of the whisper model.
Mutyshev Iskander
Purpose and description of task (project)
As part of this task, you need to find 200 videos on youtube in which there are several participants in the dialogue and the dialogue is in a mixed language (Kazakh-Russian).
Whatsapp
