Transcriber

because LRCs should make themselves

I use a Jellyfin media server, and finding .lrc lyric files for some songs can be tough.
So I made this — a simple tool that uses Faster Whisper to generate them automatically.

---

Setup

1. Configure .env

Edit your .env file with these options:

• INPATH — folder where songs are read from
• OUTPATH — folder where .lrc files will be saved
• BEAMS — number of simultaneous transcriptions
• MODEL_SIZE — choose speed vs accuracy:

Size	Parameters	English-only	Multilingual	VRAM	Speed
tiny	39M	tiny.en	tiny	~1GB	~10x
base	74M	base.en	base	~1GB	~7x
small	244M	small.en	small	~2GB	~4x
medium	769M	medium.en	medium	~5GB	~2x
large	1550M	N/A	large	~10GB	1x
turbo	809M	N/A	turbo	~6GB	~8x

---

2. Add Music

Place your audio files in the input folder (INPATH).

---

3. Run It

On Linux:

# Make sure Python is installed (tested on 3.13.7)
sudo chmod +x ./start.sh
./start.sh

On Windows:

# Make sure Python is installed and added to PATH
start.bat

(Note: the Windows script is untested for now.)