Best Speech-to-Text Project in 3 Lines of Python Code | OpenAI Whisper 2022

[@yereloyun]

Hey! Can you make a video on how to install the StoryToolKitAI for DaVinci Resolve on Windows and how with this tool we can convert audio to text with Whisper in DaVinci?

[@kritsaphongphuthibpaphaisi1509]

Sir, your tutorial is amazing

[@lobsangtenzin9169]

Bro what is the best Text to Speech apps?

[@ansjaved3484]

do you know any other opensource speech-to-text conversion model? It should be more accurate than this.

[@HubertRozmarynowski]

I've had some cases where Whisper struggles to properly transcribe names of fictional characters, places, etc. Can you somehow provide the model with "special" words to expect?

[@1littlecoder]

I think that's where the fine-tuning aspect might come into play.

[@antarus6338]

He can translate other languages? Or only English

[@1littlecoder]

Any language to English. But not the other way

[@ansjaved3484]

will this model only accept .mp3 audio format?

[@jurandfantom]

Question (level noob): how to transfer such colab and use it offline on our own pc ? Would be great to just open python, point to file and get text from it - idea for tutorial maybe ?

[@1littlecoder]

You can download the ipynb file from github and just run it on local machine (importing into your Jupyter Notebook)

[@1littlecoder]

Does it help?

[@RandomGuy-hi2jm]

can i use this in raspberry pi??

[@1littlecoder]

I just recently bought one, I can check it

[@ifeanyiidiaye1889]

Can you share the link to the open ai whisper library?

[@1littlecoder]

It's in the first line of jupyter notebook in the KZbin description

[@hibayounis7860]

How can I use this model for Arabic language?

[@soumendrak]

Is there any way we can modify the whisper methods to do real-time speech-to-text transcription?

[@1littlecoder]

When you say real-time, could you please elaborate? I'm working on near-real-time

[@simunaqv]

@@1littlecoder I suppose it mean that instead of using an audio file, it takes the audio from computer's microphone.

[@soumendrak]

@@1littlecoder Yes, I mean near real-time; record on the browser and transcribe. Thanks for working on that; waiting for that video.

[@timniyo1085]

@@soumendrak wondering about this as well, is there a similar !wget function to directly import audio files (as opposed to downloading from a website link). Thanks for the video btw I was able to install the model collab notebook even though im stuck on this step now

[@1littlecoder]

Do you mean upload audio file?

[@prostabkundu8105]

Can you please share some cutting edge projects so that we can build portfolio?

[@1littlecoder]

Thank you will try!

[@prostabkundu8105]

@@1littlecoder thanks sir.

[@arivarasand4641]

How do i generate subtitles for a 1.5 hour video. How to generate subtitles in tamil ?

[@1littlecoder]

If you have powerful enough GPU, you can . Try my colab app and see if it works

[@maelleamand]

Dear Little coder, I was wondering if an offline audio could be uploaded to whisper? I tried (using colab) but did not seem to be able to find the right code to use it

[@1littlecoder]

It could definitely be used. Offline video means, from your local machine?

[@himenatika]

I think you should open folder icon and drag and drop your file into this space ( below "sample_data" folder ), then you should be abe to use your uploaded files. Please consider to download the result file (txt or str), because all data is lost after notebook reload.

[@jawadchaudhary3528]

Ti to follow in Toby's foot steps if ya know what I an.

[@1littlecoder]

The transcription?

[@zunnen4347]

"As a native English speaker" 👀👀👀

[@1littlecoder]

I think meant As a non native English speaker.

[@zunnen4347]

@@1littlecoder yeah the video was pretty good tho, clear instructions and all.

[@zunnen4347]

@@1littlecoder also, could you explain to me how I can download this on to my own pc? I have tried but the pip install thing doesn't work. What programs do I need and what do I have to do?

[@ltamiChan]

I was specifically looking for this comment. 🤣