Whisper Web (w/webGPU)

Forked to add a few conveniences:

use webGPU for transcription by default
use the whisper-large-v3-turbo model by default
name downloaded transcript with the same name as the input file (instead of transcript.txt)
automatically download a text file transcript when transcription is complete
downloaded text files will have a timestamp for each transcribed chunk of audio data (instead of a massive single paragraph of text)
verify that the models get cached
add drag and drop for input file uploads
add more performance metrics for transcription
refactor UI components to make them easier (imho) to understand, a-la clean code's recommendations for functions

Running locally

Clone the repo and install dependencies:

git clone https://github.com/shola/whisper-web.git
cd whisper-web
pnpm install  #optional, `npm` will work just fine

Run the development server:
```
pnpm run dev
```
Firefox users need to change the dom.workers.modules.enabled setting in about:config to true to enable Web Workers. Check out this issue for more details.
Open the link (e.g., http://localhost:5173/) in your browser.