Keysat
713cd09cc2
Adds a new pipeline for diarized transcription that any client (recap-relay,
ad-hoc curl, future Mac-side tools) can call. Pure data pipeline, no LLM
or UI included — name resolution / analysis happen downstream where prompts
and rendering are configurable.
Architecture:
Spark 2 / parakeet-asr container:
+ /opt/parakeet/app/diarizer.py (new: SortformerDiarizer class)
+ /opt/parakeet/app/main.py (patched: loads diarizer, adds
/v1/audio/diarize endpoint)
Model: nvidia/diar_sortformer_4spk-v1 (~150 MB, ungated, NeMo native)
Spark Control:
+ POST /api/audio/transcribe-with-speakers
Body: multipart file
Returns: {
duration, language, speakers_detected,
segments: [{start_ms, end_ms, speaker, text}, ...],
models: {transcription, diarization}
}
Runs Parakeet ASR + Sortformer in parallel, merges words to speaker
turns by timestamp, groups into speaker-change blocks (breaks also
on >1.5s silence gaps).
+ If Parakeet 500s mid-pipeline, kicks deep-health probe and returns
503/Retry-After: 60 — same wedge-recovery pattern as v0.9.0:2.
Apply Sortformer patches to the running Parakeet container with:
bash image/parakeet_patches/apply.sh <spark2-host> <ssh-user>
Patches are reversible — apply.sh backs up the original main.py inside the
container at main.py.pre-sortformer before overwriting. Restore by copying
that file back and removing diarizer.py, then docker restart.
v0.11 follow-up: dashboard "Speech Models" panel to swap/update model
versions from the UI instead of needing to re-run apply.sh.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
14 lines
1.1 KiB
TypeScript
14 lines
1.1 KiB
TypeScript
import { VersionInfo, IMPOSSIBLE } from '@start9labs/start-sdk'
|
|
|
|
export const v0_1_0 = VersionInfo.of({
|
|
version: '0.10.0:0',
|
|
releaseNotes: {
|
|
en_US:
|
|
'v0.10.0 — Speaker diarization. Spark Control now offers a merged transcription + diarization endpoint at POST /api/audio/transcribe-with-speakers. Returns the spoken text broken into blocks with anonymous speaker labels (Speaker_0, Speaker_1, ...) and millisecond timestamps — designed as input for downstream apps (recap-relay, custom UIs) that handle speaker→name mapping and LLM analysis with their own configurable prompts. Diarization runs via NVIDIA NeMo Sortformer (nvidia/diar_sortformer_4spk-v1), loaded alongside Parakeet ASR inside the existing parakeet-asr container on Spark 2 — no new infrastructure, ~150 MB model addition. A new /v1/audio/diarize endpoint is also exposed on Parakeet directly for clients that just want speaker turns. Apply Sortformer patches via image/parakeet_patches/apply.sh after install. v0.11 will add a Speech Models dashboard panel for in-UI model swap/update.',
|
|
},
|
|
migrations: {
|
|
up: async ({ effects }) => {},
|
|
down: IMPOSSIBLE,
|
|
},
|
|
})
|