# Command Line usage

## Command-line usage 

### 1). The following command will <mark style="color:red;">Transcribe</mark> speech in audio files, using the <mark style="color:red;">medium model</mark>:

<mark style="color:red;">!whisper</mark> audio.flac audio.mp3 audio.wav <mark style="color:red;background-color:red;">--model medium</mark>&#x20;

The default setting (which selects the small model) works well for transcribing English. *<mark style="color:blue;">(The above will transcribe an</mark> <mark style="color:blue;"></mark><mark style="color:blue;">**English**</mark> <mark style="color:blue;"></mark><mark style="color:blue;">language file into</mark> <mark style="color:blue;"></mark><mark style="color:blue;">**English text script**</mark><mark style="color:blue;">)</mark>*&#x20;

### 2) The following command will <mark style="color:red;">Transcribe</mark> speech in an Urdu or another language audio file to English script using the:

!whisper "/content/sample\_data/<mark style="color:red;">sample\_urdu1.mp3</mark>" <mark style="color:red;background-color:red;">--model medium</mark>  <mark style="background-color:red;">--language en</mark>

*<mark style="color:blue;">(The above will transcribe an</mark> <mark style="color:blue;"></mark><mark style="color:blue;">**Urdu**</mark> <mark style="color:blue;"></mark><mark style="color:blue;">language input</mark>* *<mark style="color:blue;">file into</mark> <mark style="color:blue;"></mark><mark style="color:blue;">**English text script**</mark><mark style="color:blue;">)</mark>*&#x20;

### 3) The following command will <mark style="color:red;">Transcribe</mark> speech in an <mark style="color:blue;">Urdu Audio file</mark>, using the <mark style="color:blue;">medium model</mark>:&#x20;

!whisper "/content/sample\_data/<mark style="color:red;">sampleurdufile1.mp3</mark>" <mark style="color:red;background-color:red;">--model medium</mark> <mark style="color:red;background-color:red;">--language ur</mark>

*<mark style="color:blue;">(The above will transcribe an</mark> <mark style="color:blue;"></mark><mark style="color:blue;">**Urdu**</mark> <mark style="color:blue;"></mark><mark style="color:blue;">input file, into an</mark> <mark style="color:blue;"></mark><mark style="color:blue;">**Urdu script**</mark><mark style="color:blue;">)</mark>*&#x20;

### The following command will <mark style="color:red;">Transcribe</mark> speech in audio files, using the <mark style="color:red;">medium model</mark>:

<mark style="color:red;">!whisper</mark> audio.flac audio.mp3 audio.wav <mark style="color:red;">--model medium</mark>&#x20;

The default setting (which selects the small model) works well for transcribing English.&#x20;

### To transcribe an audio file containing non-English speech, you can specify the language using the --language option:

<mark style="color:red;">!whisper</mark> filename.wav <mark style="color:red;">--language Japanese</mark>&#x20;

### To translate an Audio file from one language to another language use the --translate option

<mark style="color:red;">!whisper</mark> filename.wav --<mark style="color:orange;">task translate</mark> <mark style="color:red;">--language en</mark> (will translate the speech into English)

### Run the following command to view all available options:

<mark style="color:red;">!whisper --help</mark>