Python · MIT
alexkroman/tiny-audio
A speech-to-text system you can train for $25. A frozen speech encoder and a Qwen LLM joined by a small trained projector, with word timestamps, speaker diarization, a batched HTTP server, a live demo, and a free course.
★ 169
Stars
18 forks · 0 open issues
—
Growth · stale
We stopped seeing this repository in its topic's top results, so its star count is frozen at the date it was last read. Subtracting from a frozen number would read as zero growth, which would be false, so no rate is shown.
19.3
Momentum / 100
see the breakdown below
13mo ago
Created
last push 29d ago
Star history
How this score was built
| Component | Weight | Earned | Why |
|---|---|---|---|
| Velocity | 55 | 0.0 | — — no measurable rate |
| Acceleration | 15 | 0.0 | needs 7-day and 30-day rates side by side |
| Freshness | 15 | 12.8 | last push 29d ago; decays to zero at 180 days |
| Scale | 15 | 6.5 | 169 stars, log-scaled against 150k — stops a 3-star repo topping the board |
| Momentum | 100 | 19.3 | capped at 35 because growth is unmeasured |
Detail
| Topics | asraudio-llmforced-alignmenthuggingfacellmloramultimodalpipecatpytorchqwenrunpodspeaker-diarizationspeech-llmspeech-recognitionspeech-to-texttransformersturn-detectiontutorialvoice-aiword-timestamps |
| Matched on | Learning & Roadmaps (tutorial) |
| Short window | no snapshot 5–10 days old yet |
| Long window | +9 stars over 26 days = 0.3/day |
| Lifetime average | 0.44 stars/day over 387 days |
| First seen here | 2026-10-11 |
| Last read | 25d ago |