I-IAL9000/shrinkerr v0.9.21

on GitHub

Fixed

• Audio detection runs in a killable subprocess so it can't hang the app. v0.9.19's timeout let a wedged whisper transcribe keep running in an un-cancellable thread; across a batch these piled up, pegged every core, and starved the event loop (the app went fully unresponsive — proven by health-checks that never returned). Detection now runs in a subprocess that's killed on timeout, freeing the CPU. Trade-off: the model is loaded per detection, so bulk runs with base/small are slower — use tiny for large batches, or accept the cost.

Docker images

Pinned to this release:

docker pull ghcr.io/i-ial9000/shrinkerr:0.9.21         # portable (amd64 + arm64, CPU)
docker pull ghcr.io/i-ial9000/shrinkerr:0.9.21-nvenc   # x86_64 + NVIDIA GPU (NVENC)
docker pull ghcr.io/i-ial9000/shrinkerr:0.9.21-edge    # as above but ffmpeg master
docker pull ghcr.io/i-ial9000/shrinkerr:0.9.21-edge-nvenc

Floating tags (:latest, :edge, :nvenc, :edge-nvenc) track the most recent release of each lineage.