-
-
Notifications
You must be signed in to change notification settings - Fork 24
language_translation
Cyber Official edited this page Sep 22, 2025
·
5 revisions
These arguments control language selection and translation features.
| Flag | Description |
|---|---|
--language |
Source language of the audio content for improved transcription accuracy. Use language codes like 'en', 'es', 'fr', 'de', 'ja', etc. |
--target_language |
Target language for translation output when using --translate mode. Must be used in combination with --translate flag. |
--translate |
Automatically translate transcribed text to English using built-in translation capabilities. Works with both live microphone input and file processing. |
--transcribe |
Force transcription mode instead of real-time processing. Automatically enabled for file inputs, but can be explicitly set for microphone input. |
--condition_on_previous_text |
Use previous output as prompt for next window (reduces repetition). |
Set the source and target languages. Example:
python synthalingua.py --language ja --target_language enEnable translation or transcription to a target language. Example:
python synthalingua.py --translate --language ja
python synthalingua.py --transcribe --target_language esHelps prevent repeated or similar outputs by conditioning on previous text.