Jarian 0da74d39a0 init: media-transcriber pipeline ...

- whisper.cpp base.en transcription on RTX 5060 Ti (GPU 1)
- 25,439 videos indexed from /mnt/mediaserver/
- pipeline: ffmpeg extract -> whisper-server infer -> save SRT
- thermal cycle: 45min work / 15min rest
- fixes: curl quotes for spaces, --convert + WorkingDirectory=/tmp

2026-07-07 04:47:18 +00:00

src

init: media-transcriber pipeline

2026-07-07 04:47:18 +00:00

systemd

init: media-transcriber pipeline

2026-07-07 04:47:18 +00:00

.gitignore

init: media-transcriber pipeline

2026-07-07 04:47:18 +00:00

mt.py

init: media-transcriber pipeline

2026-07-07 04:47:18 +00:00

setup.sh

init: media-transcriber pipeline

2026-07-07 04:47:18 +00:00

Description

Local Whisper-based media transcription service — speech-to-text pipeline with systemd deployment and test suite.

32 KiB

Languages

Python 89.4%

Shell 10.6%