# ASR on multilingual audio data (code-switching)

**URL:** <https://discuss.huggingface.co/t/asr-on-multilingual-audio-data-code-switching/68624>\
**Category:** Intermediate\
**Created:** [January 10, 2024, 9:49am UTC](https://discuss.huggingface.co/t/asr-on-multilingual-audio-data-code-switching/68624 "2024-01-10T09:49:54Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![sisi](https://avatars.discourse-cdn.com/v4/letter/s/e9a140/32.png) [@sisi](https://discuss.huggingface.co/u/sisi)\
**Post date:** [January 10, 2024, 9:49am UTC](https://discuss.huggingface.co/t/asr-on-multilingual-audio-data-code-switching/68624/1 "2024-01-10T09:49:54Z")

</div>

Hello , i want to finetune whisper model on audio files of an arabic dialect however the audios can have foreign words in french or english (code-switching) how can i handle these words to be transcribed correctly?