Integration with NodeJS

[santisoifer]

Hi there, I was looking foward to make a web app with Whisper, but when I started seraching for information about how could I integrate NodeJs and Whisper and I didn't find anyone who had the same question, so there wasn't an answer. Is there any way to make that posible? Or I have to integrate Python in my web? Thank you.

[mayeaux]

You can check out my app here https://github.com/mayeaux/generate-subtitles

It's built using Nodejs and Whisper

[santisoifer]

Thank you very much! 😁

[Illyism]

I made a simple version as an OpenAI wrapper https://github.com/Illyism/whisper-docker

[chilarai]

I have written a small tutorial for beginners to integrate node.js with OpenAI Whisper along with a Github library.
Blog: https://techpro.ninja/whisper-openai-tutorial-with-python-nodejs/
Library: https://github.com/Techpro-ninja/openai-whisper-with-python-and-nodejs

[Goddard]

If it is possible to load the model with onnx then it should be some what trivial to get it working.

For example : https://github.com/Fhrozen/whisper_onnx
And : https://www.w3.org/2020/06/machine-learning-workshop/talks/onnx_js_a_javascript_library_to_run_onnx_models_in_browsers_and_node_js.html

[FerLuisxd]

I'm currently looking for a solution too, other users pointed to use the CLI or their API.
But I think it's best to be able to call the model on memory rather than invoking it every time you want to do a new transcript (which takes a lot of time)

[juzarantricreole]

Did you find any way to integrate with nodejs?

[santisoifer]

Not yet. I founded many people doing it using the CLI but I don't know how to do it. However I think some answers here may help you. Let me know if I'm right!

[FerLuisxd]

Hey guys! I think one solution could be python binding (haven't tried it yet but it looks really promising!).
https://github.com/hmenyus/node-calls-python

[ChetanXpro]

hey you could try this npm package https://www.npmjs.com/package/nodejs-whisper , nodejs binding for model,
i am using this to get word level timestamps for one of my project

[ayoadeoye1]

it's full of bug. may work fine for your personal purpose

[juzarantricreole]

I think we can use nodejs opnai library. it provide transcription mechanism

…

On Thu, Jun 15, 2023 at 4:21 PM Santi ***@***.***> wrote: Not yet. I founded many people doing it using the CLI but I don't know how to do it. However I think some answers here may help you. Let me know if I'm right! — Reply to this email directly, view it on GitHub <#637 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BAKWQTYIZYGSVVFOV7QB7JDXLLSJXANCNFSM6AAAAAASTULISY> . You are receiving this because you commented.Message ID: ***@***.***>

-- Regards -- [image: user] Juzar Antri JR. SOFTWARE ENGINEER [image: msg] ***@***.***> [image: linkedin] <https://www.linkedin.com/in/juzar-antri-10537b212/> [image: github] <https://github.com/juzarantri> [image: logo] <https://www.creolestudios.com/> [image: web-apps] <https://www.creolestudios.com/our-services/ecommerce-web-development/> Web Apps <https://www.creolestudios.com/our-services/ecommerce-web-development/> <https://www.creolestudios.com/our-services/ecommerce-web-development/> [image: cloud-apps] <https://www.creolestudios.com/our-services/developing-cloud-based-systems-apps/> Cloud Apps <https://www.creolestudios.com/our-services/developing-cloud-based-systems-apps/> <https://www.creolestudios.com/our-services/developing-cloud-based-systems-apps/> [image: mobile-apps] <https://www.creolestudios.com/our-services/mobile-app-development/> Mobile Apps <https://www.creolestudios.com/our-services/mobile-app-development/> <https://www.creolestudios.com/our-services/mobile-app-development/> [image: resources] <https://www.creolestudios.com/> Resources for Hire <https://www.creolestudios.com/our-services/dedicated-resource-hiring/> <https://www.creolestudios.com/our-services/dedicated-resource-hiring/> *Disclaimer:* Information transmitted/written in this e-mail is proprietary to the sender and is intended for exclusive use by the individual or entity to which it is addressed by the sender (subject to his duties/delegation power) and may contain information that is privileged, confidential or exempt from disclosure under applicable law and shall not attach any liability on the originator. Any use, distribution, transmission, forwarding, printing, copying or dissemination of this information in any way or in any manner is strictly prohibited. All dispute if any are subject to the jurisdiction at the place where the sender resides. If you have reason to believe that you are not the intended recipient of this communication, please contact the sender immediately. No responsibility whatsoever it may be for loss or damage arising from the use of the information transmitted by this email including damage from virus on the sender.

[ChetanXpro]

hey you could try this npm package https://www.npmjs.com/package/nodejs-whisper , nodejs binding for model,
i am using this to get word level timestamps for one of my project