taco/microWakeWord-Trainer-Nvidia-Docker
1
0
Fork 0
mirror of https://github.com/TaterTotterson/microWakeWord-Trainer-Nvidia-Docker.git synced 2026-08-12 07:55:33 -06:00
Code Issues Packages Projects Releases Wiki Activity

Release NVIDIA WakeWord Trainer v19

Browse Source
This commit is contained in:
MasterPhooey
2026-07-26 11:23:35 -05:00
parent 6ee228e8d3
commit 518df63161
4 changed files with 108 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
WHATS_NEW.md
View File

@@ -1,3 +1,3 @@
- Added Parakeet ONNX as a second local Auto Training STT engine alongside Faster Whisper.
- Replaced manual model, device, and compute fields with a simple engine selector and managed language-aware models.
- Added CUDA-enabled ONNX Runtime with CPU fallback, runtime reporting, and model-cache cleanup when switching engines.
- Fixed first-run Parakeet ONNX setup failing when its empty model directory was mistaken for a complete offline model.
- Parakeet now downloads or resumes the required INT8 snapshot before loading through ONNX ASR.
- Complete local snapshots are reused without Hub access, preserving offline startup after the initial download.