tiny-voca
A 4.5M-parameter two-voice English text-to-speech model, trained from scratch
Runssource: GitHubhomepagePythonApache-2.0commit 62e2adef158b
Python, Apache-2.0 licensed. The project labels itself: onnx, text to speech and tts.
tiny-voca runs. An Argusic agent installed it in 15 minutes and hit 3 errors and fixed every one without help on a clean machine with no GPU, and the whole session was recorded.
tiny-voca 0.2.0 installed in a venv with PyTorch CPU 2.15.0; model weights downloaded; dizione phonemizer configured; both F2 and M2 voices produce valid 24 kHz WAV audio via the CLI; the full test suite passes (78/78, 3 skipped for missing espeak-ng or librosa); ONNX export and inference with onnxruntime works.
What the agent ended up with on a clean machine, in its own words. How this is measured
Measured by Argusic on a fresh machine every time. Every number links to its evidence. Argusic Score 100 of 100 (the mean of this project's run scores).
At a glance
Subject data from GitHub, linked at the top of this page, refreshed . Test data by Argusic (CC BY 4.0); every number links to a run page with the full log, the recording, and their sha256 hashes.
What was measured
From the run that decided this verdict. Every line is a recorded value, not an estimate.

What went wrong, in the agent's own words
The agent's own words, worst first, never rewritten into advice.
- system Python externally managed (PEP 668) — pip refused to install globally2 minutes
- dizione phonemizer could not find data (FileNotFoundError)0.5 minutes
- ONNX export test failed — module 'onnx' not installed (only onnxruntime was present)0.5 minutes
- Install time
- 15 minutes
- Cold machine to finish
- 5 minutes
- Errors hit and fixed
- 3 hit, 3 fixed with no human help
- How the result was proved
- python -m tiny_voca.infer --checkpoint weights/tiny-voca.pt --voice F2 --text "Hello world." --output /tmp/test.wav printed phonemes and 'Saved 1.17s in 1 utterance; peak=0.8626'; ffprobe confirmed a valid 1.17s WAV at 24 kHz. Same with voice M2 (2.77s, peak=0.634). Full test suite: 78 passed, 3 skipped.
- Model tokens used
- 55,415
- Exact commit tested
- 62e2adef158b
- Container image
- sha256:cdd920bce783
What was not measured
The other half, so you can see exactly how far this evidence goes. A gap is printed as a gap and never filled in.
- Peak GPU memory
- this ran on a CPU machine, so there was no GPU to sample
- Behaviour over a long run
- each attempt uses the smallest documented settings and a fixed time budget, so this measures whether it runs, not how it holds up under load
The end of the recording
The last lines of the terminal session that decided this verdict, exactly as they were recorded.
tokens used 55,415 All 78 tests pass (3 skipped). Here's the final report. **Summary of what happened:** - Created a virtual environment (system Python is externally managed) - Installed torch (CPU), numpy, soundfile, huggingface-hub, onnx+onnxruntime - Installed tiny-voca with `pip install -e ".[own]"` which pulled in the dizione phonemizer - Downloaded weights from Hugging Face to `weights/` - Set `DIZIONE_DATA=weights/phonemizer` so dizione finds its data - Ran the full test suite: **78 tests passed, 3 skipped** (2 need espeak-ng binary, 1 needs librosa… - Verified real TTS end-to-end: `tiny-voca-speak` generated valid 24 kHz WAV audio for both F2 and … - Verified export/ONNX round-trip works (test passes after installing `onnx`)
Replay the whole session, every command from a clean machine to this point.
How it was tested
One agent, one container, no human help. It installs the project the way its own documentation says to, runs it, and fixes what breaks. Everything below is recorded as it happened: the terminal session, the log and the exact commit. The full method.
Strengths and limits
Measured facts, not opinions. How this is written.
What went well
- Reached a running state on a clean machine, with the session recorded.
- Installed in 15 minutes.
- Recovered from all 3 errors without a human stepping in, which says the failures are documented well enough to solve.
- Ran without a GPU, so it does not need one to start.
- Licensed Apache-2.0, as reported by its host.
What did not, or is not known
- Measured once so far. A single attempt shows it can run, not that it runs reliably.
What it is a reasonable choice for
- Trying it on a laptop or a small server: it reached a running state without a GPU.
Also tested, in the same area
Every one of these was installed and run by Argusic on a clean machine. Nothing appears here that was not tested.
Run history
Topics (from GitHub)
Embed the badge
Markdown for the project README. It links back here; terms on the terms page.
[](https://argusic.com/subject/tiny-voca)Questions
- Does tiny-voca run?
- Yes. tiny-voca runs. Argusic installed and launched it on a clean machine in 15 minutes, hitting 3 errors on the way, and recorded the session.
- How did Argusic test tiny-voca?
- On a fresh, disposable machine, with every command recorded and the repository pinned at commit 62e2adef158b. One attempt is recorded, and the full method is on the methodology page.
- What did it cost to test tiny-voca?
- The run that produced this verdict cost $0.07: real compute and model cost, not a list price per million tokens. It is recorded on that run's page.
- How long does tiny-voca take to install?
- 15 minutes in the run that decided this verdict, from a clean machine to a running state. That is one measured attempt, not an average.
- Does tiny-voca need a GPU?
- Not to start. Argusic reached a running state on a machine with no GPU.
- What goes wrong when installing tiny-voca?
- 3 things broke in the recorded run, and 3 were fixed without human help. Each one, and the time it cost, is listed on this page.
- Where is the evidence for tiny-voca?
- The recorded run is on this page, each linking to its full log and terminal recording, stored with a sha256 fingerprint so it cannot be quietly altered.