speech.cpp tested by Argusic Runner

What the test found

The speech.cpp 0.10.1 build from source produces a working speech executable and libspeech.so.4 C API shared library on Linux x86_64 CPU. The binary enumerates devices, lists all models in the catalog with their tasks/languages/sizes, downloads models from Hugging Face on demand (verified with silero-vad), runs Silero VAD inference on a WAV file, and validates all subcommand arguments correctly.

Built from source on Ubuntu 24.04 x86_64 with g++ 13.3.0 and CMake 3.28.3. The speech binary, libspeech.so.4 shared library, and libspeech-static.a all built successfully. Verified speech --version (0.10.1, C API 4.1), speech --help, speech models list, speech serve argument validation, error messages on missing files. Ran speech devices which detected CPU only. Downloaded Silero VAD model from Hugging Face and ran speech vad which processed a tone WAV correctly (0 regions of speech, RTFx 329). speech update --check confirmed latest release. All subcommands parse correctly and produce appropriate error messages. Could not run stage checks or smoke tests which require model GGUF files and reference dumps.

Argusic installed speech.cpp in 12 minutes and launched it; verification reached Verified running for real. 1 of 1 recorded error was worked through (see the timeline below).

Runsall runs of speech.cpp

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.