Transcribe Youtube Video to text with OpenAI Whisper 3 lines of code in Python #nlp

In this video I explain and demonstrate using a google collab notebook 1) How to download audio stream of a youtube video using pytube python library 2) How to use OpenAI Whisper speech recognition model to convert audio to text transcript

If you like such content please subscribe to the channel here:

Relevant Links:

[idanplaut]

How can I transcribe full youtube channel automaticly, and to get it Organized in separate files for each video?

[ewoutlagendijk]

Thank you for this video ! Some remarks:
Even after setup&install I kept getting the error that streams was not an attribute of StreamsQuery. The error is a bit odd, because when I kept tweaking the code I got it to work. The code I ended up using:

from pytube import YouTube


try:
yt = YouTube(link)
yt2 =
stream = yt.streams.get_by_itag(139)
stream.download('', 'GoogleImagen.mp4')
except:
print("Connection error")

!whisper "GoogleImagen.mp4" --model medium --language zh

[_kishorjoshi]

Do you know this kind of any good way to transcribe youtube video which is in some other language, but we need it to be transcribed in English (hindi audio to english transcription)

[kabijoshi]

Why dont you provide all the alg. for it?

[shaad]

hey, if i want to build app with whisper, i don't need any whisper api or any other things limited credit, then i have to pay to increase credit or other things, exactly like we need to do api integration when we make dalle2 or chatgpt app ?
is it mean i can put unlimited request, unlimited loads on it?
and how can i use it in java script?

[cryptocoeus]

Pyhton code at Colab does not work. It throws an error. Any idea what's wrong with it?