grant/spark-control
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 3 Wiki Activity

v0.13.0:0 - revert WhisperX migration; back to Parakeet + Sortformer

Browse Source 
After five hotfix iterations on the WhisperX install (v0.12.0:0–:4) we
never got a working docker build. The fundamental constraint isn't
patchable from outside NVIDIA: NGC PyTorch on ARM64 (the only base that
runs on Spark 2's GB10 Blackwell) ships a custom-versioned torch
2.10.0a0+b558c98 that has no pre-built torchaudio match anywhere.
WhisperX → pyannote → torchaudio is a hard dependency chain we couldn't
satisfy without rebuilding torchaudio against torch 2.10's alpha API.
Walking away cleanly is better than another night of chasing.

Removed from the codebase:
  - image/whisperx_container/* (Dockerfile + requirements + app/main.py)
  - image/app/whisperx_install.py (install manager + SSH ship-context logic)
  - image/Dockerfile COPY whisperx_container
  - WHISPERX_* config keys in config.py
  - whisperx service entry in services.py
  - WhisperX-preferred branch in audio_proxy.py
  - /api/whisperx/* endpoints in server.py
  - install banner + progress dialog in index.html
  - render + handlers in app.js
  - .whisperx-install styles in style.css

Spark 2 cleaned in tandem (user-authorized): container removed,
~/whisperx-build/ removed, 5.4 GB of dangling image layers + 1.3 GB of
builder cache reclaimed. parakeet-asr and magpie-tts unaffected and
healthy throughout.

The audio path is back to exactly what shipped in v0.11.0:3:
  POST /api/audio/transcribe-with-speakers
    → Parakeet (transcription) + Sortformer (diarization) in parallel
    → merged by timestamp into speaker-labeled blocks

v0.13.0:1+ will add the actually-needed fixes that the WhisperX detour
was meant to address:
  1. memory cap on the parakeet-asr container so a long-audio crash
     can't swap-thrash Spark 2 again
  2. a chunking proxy in /api/audio/transcribe-with-speakers that
     splits inputs >10 min before Sortformer

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
Keysat
2026-05-19 08:03:19 -05:00
parent a24610ad2a
commit 95524f4983
14 changed files with 14 additions and 1086 deletions
Download Patch File Download Diff File Expand all files Collapse all files
image/app/static/index.html
-40
View File
@@ -103,46 +103,6 @@
<div class="tab-content" id="tab-audio" role="tabpanel" aria-labelledby="tab-audio-trigger">
<section id="whisperx-install-card" class="whisperx-install hidden">
<div class="wx-install-body">
<div class="wx-install-title">
<strong>Add WhisperX</strong>
<span class="tag ok">recommended</span>
</div>
<p class="muted small">
WhisperX is a single-container speech pipeline (faster-whisper for transcription + pyannote 3.1 for diarization)
designed to handle long audio cleanly. Replaces the Parakeet + Sortformer combo we patched together,
which crashed on a 90-min meeting. Pulled and built directly on Spark 2 (~10–15 min first time;
you only do this once).
</p>
<p class="muted small">
Requires a Hugging Face token at <code>~/.cache/huggingface/token</code> on Spark 2 (already set up).
</p>
<div class="wx-install-actions">
<button id="wx-install" class="btn primary">Install WhisperX</button>
</div>
</div>
</section>
<dialog id="whisperx-progress-dialog" class="modal">
<form method="dialog" class="modal-form">
<h3 id="wx-prog-title">Installing WhisperX…</h3>
<div class="phase-row">
<span class="spinner"></span>
<div class="phase" id="wx-prog-phase">Starting…</div>
<span class="spacer"></span>
<span class="timer" id="wx-prog-elapsed">0:00</span>
</div>
<details open>
<summary class="muted small">Build log</summary>
<pre id="wx-prog-log" class="log"></pre>
</details>
<div class="modal-actions">
<button type="button" id="wx-prog-close" class="btn">Close</button>
</div>
</form>
</dialog>
<section id="services-panel" class="services hidden">
<div class="section-header">
<h2 class="section-title">Always-on services</h2>