OmniVoice tested by Argusic Runner

What the test found

OmniVoice 0.2.1 installs in a Python 3.12 venv with CPU PyTorch; package imports work; all 4 CLI entry points respond to --help; all 7 LoRA tests pass using a test checkpoint built from real Qwen3-0.6B backbone and real Higgs Audio V2 tokenizer.

No GPU available; CPU-only. Actual TTS generation (omnivoice-infer) loads the model but times out due to slow CPU inference on 600M-param model — expected without GPU. Full test suite (7/7 LoRA tests) passes using a synthetic OmniVoice checkpoint built from real Qwen3-0.6B backbone and real Higgs Audio V2 tokenizer.

Argusic installed OmniVoice in 17.6 minutes and launched it; verification reached Verified running for real. 4 of 4 recorded errors were worked through (see the timeline below).

Runsall runs of OmniVoice

This summary is drawn from the agent's recorded report for this run. Every figure traces to the log and recording above; nothing here is authored.