STT CUDA failure mid-transcription CPU fallback fix
Cluster 105295. Data as of
Nothing open on GitHub. Fix merged in #112167 on 16 Sep 2026. Check it covers your case.
The short answer
- Known?
- Yes. 3 related items grouped here, 1 issue already closed.
- Fix event
- A fix was merged to main in PR #112167. Check it covers your case. No release data is tracked.
- Tracker
- The maintainers' tracker marks this
resolved, deciding PR #112167. That is separate from the fix event. - Still open
- Nothing open on GitHub.
These items address a bug where local speech-to-text using Whisper did not fall back to CPU when CUDA errors (such as a missing libcublas library) occurred during segment iteration rather than at startup. The work adds retry logic to handle this mid-transcription failure case.
Open issues0
Open PRs0
Fixes on record
-
#112167 fix(stt): local whisper falls back to CPU when the CUDA runtime fails mid-transcribe (#111929, #105295, salvage #103848)
files:
contributors/emails/[redacted:email],tests/tools/test_transcription_tools.py,tools/transcription_tools.py
1 PR closed without merging
From the close comments: merged via another change 1
-
#105336 fix(stt): retry CPU fallback when CUDA fails during segment iteration · merged via another change →
a1cbed3,#112167
Closed issues · 1
- Closed#105295Local STT CUDA fallback misses errors raised during segment iteration (libcublas.so.12 missing)
How this cluster was grouped
-
Link evidence (
structural links throughout): every member is connected by structural evidence: shared changed files, or a closing/superseded reference. -
Pairwise link density
0.67: the share of item pairs that carry a link, counting a structural link in full and a relevance-only link at 0.36.1.00means every pair is linked.
Pick up a Lab task
Check whether a fix covers the open issues here, with your own agent. Bring back evidence for maintainers.
Open the LabAt a glance
- Merged PRs
- #112167
- Closed issues
- 1
- Attempts
- 1 PR closed without merging
- Fix link checked
- 16 Sep 2026