Wątek: using the whisper library to transcribe text

Strona: 1 z 1

1 z 8: rudolf

Hi, some people are using the whisper library in python to transcribe text. My questions, how much resources is taking that, and is there a python application in which I can select an audio file to transcribe, and it does it's job?
#MCMacy4admin2024
12.04.2024 22:36

2 z 8: bomberman29

there are some whisper tools you can find for free. One is whisperjax (on huggingface), free model. Another option is to use github.com/ggerganov/whisper.cpp and it's python bindings
my github: https://github.com/denizsincar29
15.04.2024 08:07

3 z 8: rudolf

right now taking a look at the releases. mostly interested at the server, but I've to get the model file somehow. they're not included in the release. I need to get the english and german ones.
#MCMacy4admin2024
15.04.2024 12:25

4 z 8: rudolf

It seems a bit confusing, isn't there a nice app with a cool web or normal interface and without that I have to set 95 tings up for myself?
#MCMacy4admin2024
15.04.2024 14:08

5 z 8: Seedy-Threepio

If you have Intel hardware, you can use Audacity with the OpenVino modules.

15.04.2024 14:09

6 z 8: rudolf

I have an intel I 5. Also, where can I find these plugins and how to use them? Sorry, I don't use audacity usually.
#MCMacy4admin2024
15.04.2024 15:07

7 z 8: rudolf

Hmm, anyone? It really would be cool to be able to use it.
#MCMacy4admin2024
27.04.2024 14:41

8 z 8: rudolf

Nevermind, I now figured it out. I am now using the server.exe with the base.en model, though I wonder if:
1. I can use a better one with my stats?
2. How exactly can I send a file through that thing now?
#MCMacy4admin2024
27.04.2024 16:18