ownvoice-cli 0.1.1

Train a LoRA voice adapter for pocket-tts and own the resulting model, no API lock-in.

OwnVoice

Train a LoRA voice adapter for pocket-tts and keep the result: a file on your own disk, not an API subscription.

Why this exists

pocket-tts is a genuinely good, MIT-licensed, CPU-capable local text-to-speech model from Kyutai. Its own maintainers have been clear that fine-tuning code isn't coming any time soon: on issue #30, maintainer @vvolhejn wrote "We are not planning to release fine-tuning code for our TTS and STT models in the near future," and 18 people reacted to that thread asking for exactly this. OwnVoice is a small, standalone CLI that fills that specific gap: point it at a handful of your own voice recordings, and it trains a LoRA adapter you keep and run yourself.

It is not a hosted service, it has no billing, and it does not track usage. It is a training script, an inference script, and a scoring script, wired together behind three CLI commands.

What OwnVoice is not

pocket-tts already ships zero-shot voice cloning out of the box: pass a .wav file to --voice (or call get_state_for_audio_prompt() from Python) and it clones that voice with no training step at all. If that is all you need, use pocket-tts directly, it is simpler and faster.

OwnVoice exists for a narrower case: baking a voice permanently into trained weights, so generation no longer depends on distributing or re-processing a reference audio clip at runtime, with (based on the training objective, not yet independently benchmarked at scale) more consistent output across many generations than a single-clip zero-shot embedding tends to produce. That is the specific gap the 18 reactors on issue #30 were describing, and it is the only thing OwnVoice adds on top of what pocket-tts already does well.

Install

Requires Python 3.11 or newer.

pip install ownvoice

npx / agent-native environments: OwnVoice is a Python/PyTorch CLI, so the npm package is a thin wrapper, not a Node reimplementation. It bootstraps into the real CLI via uv or pipx, whichever is already on PATH -- useful for coding-agent sandboxes and CI runners that default to a Node toolchain.

npx ownvoice check

Not functional yet: the npm wrapper is published and installable, but it delegates to uvx ownvoice, and OwnVoice itself is not yet published to PyPI (pip install git+... above is the only working install path today). npx ownvoice will fail with a clear "package not found" error from uv until the PyPI release ships.

Torch and CUDA: ownvoice check (see below) needs no GPU at all and runs on CPU, matching pocket-tts's own CPU-capable design. Training a real adapter is much faster on an NVIDIA GPU. If you have one, install the CUDA build of PyTorch first by following pytorch.org/get-started/locally, then install OwnVoice on top of it, so pip does not silently pull the CPU-only wheel instead. On Apple Silicon or a CPU-only machine, the default pip install of torch is fine: ownvoice check and ownvoice infer will run normally, ownvoice train will just take longer per epoch.

Quickstart

1. ownvoice check, the free Day-0 validation

Before recording anything or renting a GPU, confirm that PEFT's LoRA injection actually works against pocket-tts's real model structure. This is entirely free: CPU only, no training, no GPU.

$ ownvoice check
[ownvoice check] PASS: PEFT LoRA injection succeeded against pocket-tts's flow_lm module (target_modules="all-linear").

If it fails, OwnVoice prints the model's real module tree instead of a raw stack trace, so you can see exactly what did not match and report it precisely:

$ ownvoice check
[ownvoice check] FAIL: PEFT LoRA injection failed against pocket-tts's flow_lm module structure: <error detail>. Please post an honest blocker (this error plus the module tree above) as a comment on https://github.com/kyutai-labs/pocket-tts/issues/30 rather than working around it silently, that issue is exactly where this gap needs to be visible.

Module tree (for debugging / for the issue #30 blocker post):
<root>: FlowLMModel
input_linear: Linear
transformer: StreamingTransformer
transformer.layers.0.self_attn.in_proj: Linear
transformer.layers.0.self_attn.out_proj: Linear
...

2. ownvoice train

Record 5 to 10 minutes of clean audio of the voice you want to train (your own voice, with your own consent, see Consent and misuse below), split into a few .wav clips in one directory, then point OwnVoice at it:

$ ownvoice train --voice-clips ./my-voice-clips
[ownvoice train] USABLE ADAPTER
Usable adapter (similarity 0.812 >= 0.75). Try it now:
  ownvoice infer --adapter ownvoice-adapter/adapter.safetensors --text "This is my own voice, trained with OwnVoice."

Only --voice-clips is required. Every other flag has a sensible default: --out (./ownvoice-adapter/), --epochs (10), --lora-rank (8), --lora-alpha (16), --lora-dropout (0.05), --learning-rate (1e-4), and --eval-text (the sentence synthesized to compute the similarity score).

A run that finishes but does not clear the similarity bar still exits 0. It is a labeled result with a concrete next step, not a crash:

$ ownvoice train --voice-clips ./my-voice-clips
[ownvoice train] BELOW THRESHOLD
Below threshold (similarity 0.612 < 0.75). The adapter was still saved, try more/cleaner voice clips, more epochs, or a higher --lora-rank, then re-run. You can still listen to it:
  ownvoice infer --adapter ownvoice-adapter/adapter.safetensors --text "This is my own voice, trained with OwnVoice."

Only a data-loading problem (no usable clips) or a caught PEFT-injection failure exits non-zero. Every successful run writes adapter.safetensors and metadata.json (training config, the similarity score, a timestamp) to the output directory: two files you keep, with no server round-trip needed to use them again.

3. ownvoice infer

$ ownvoice infer --adapter ownvoice-adapter/adapter.safetensors --text "Hello, this is my own voice."
[ownvoice infer] Wrote ownvoice-output.wav

Every subcommand also supports --json for a structured, machine-parseable output mode, useful if a script or an agent is calling ownvoice programmatically instead of a person reading the terminal:

$ ownvoice check --json
{"success": true, "message": "PEFT LoRA injection succeeded against pocket-tts's flow_lm module (target_modules=\"all-linear\").", "module_tree": null}

How it works

voice clips (wav)
      |
      v
  data.py   --validate format/duration-->  clean clip set
      |
      v
  train.py  --PEFT LoRA (target_modules="all-linear")--> adapter.safetensors + metadata.json
      |
      v
  infer.py  --generate test utterance--> synthesized audio
      |
      v
  score.py  --resample to 16kHz mono--> Resemblyzer cosine similarity
      |
      v
  CLI report (>= 0.75 = usable adapter, below triggers a labeled next-step message)

ownvoice/data.py loads and validates the voice-clip directory. ownvoice/train.py loads pocket-tts's frozen base model, injects a LoRA adapter into its flow_lm transformer with PEFT (target_modules="all-linear"), runs the training loop, and saves the adapter plus a manifest. ownvoice/infer.py loads a saved adapter back onto the base model and generates speech. ownvoice/score.py resamples audio to 16kHz mono with torchaudio.transforms.Resample and scores speaker similarity with Resemblyzer.

OwnVoice is intentionally single-model: it wraps pocket-tts only, with no abstraction layer for a second base model, since none is in scope.

Consent and misuse

This tool clones a voice from audio you have the right to use. Do not clone someone else's voice, or a public figure's voice, without their explicit consent. OwnVoice ships no bulk-generation or auto-scaling feature in this version, keeping the blast radius of any single misuse case small.

Setup-time benchmark vs comparable tools

Tool	Time to first working setup	Notable design choice	Source
kokoro-tts	under 2 minutes	pip install git+..., instant CLI synthesis, no fine-tuning	kokoro-tts README
Unsloth	under 1 minute to start a run	one-command training start (uv pip install)	Unsloth docs
pocket-tts	seconds	--voice <wav> zero-shot cloning, no training available	pocket-tts README
OwnVoice	under 2 minutes to a confirmed-working training environment	ownvoice check: free, instant, CPU-only PEFT-compatibility validation before spending anything on a GPU	this repo

OwnVoice's own training run is real GPU time, honestly labeled and not hidden behind a fake progress bar, the same category norm Unsloth uses. What OwnVoice compresses to under two minutes is everything before that: confirming your environment actually works.

Implementation status

This is a young v0.1. ownvoice check, the CLI argument parsing, voice-clip validation, the similarity scoring math, and the adapter/manifest save and load path are implemented and covered by the test suite (pytest). LoRA injection was verified structurally against pocket-tts's real source and then confirmed for real: ownvoice check was run against pocket-tts's actual downloaded weights, on CPU, and PEFT's target_modules="all-linear" injection genuinely succeeded. The full training and generation path has since been verified end to end for real too: a real 2-epoch LoRA training run against loaded pocket-tts weights produced a finite, non-NaN flow-matching loss, and the resulting adapter produced a real, non-silent generated .wav file via ownvoice infer. That validation surfaced two real gaps in the naive approach and fixed them: (1) pocket-tts's published, inference-only PyPI package does not actually expose a way to compute the training loss through FlowLMModel.forward() despite its own docstring claiming otherwise, so OwnVoice computes the flow-matching loss directly from flow_lm's real submodules instead; (2) swapping base_model.flow_lm to the PEFT-wrapped model before calling generate_audio() breaks pocket-tts's internal KV-cache state lookup -- no swap is needed at all, since PEFT's LoRA injection already mutates base_model.flow_lm in place. One real, external limitation to know about: the publicly downloadable pocket-tts weights (kyutai/pocket-tts-without-voice-cloning) refuse a raw reference-clip path/URL outright; OwnVoice works around this by pre-loading and resampling the clip itself, but voice-cloning fidelity from that checkpoint is a known limitation of the base model, not an OwnVoice bug -- for kyutai's best-quality cloning weights, request gated access at huggingface.co/kyutai/pocket-tts. Run ownvoice check yourself and read the source before trusting any of it further, that is the right amount of skepticism for a v0.1.

Contributing

Issues and PRs welcome, MIT licensed throughout. If you want to help close the actual gap this project targets, the most useful contribution is upstream: a lightweight LoRA-adapter training script contributed back to kyutai-labs/pocket-tts itself, discussed on issue #30.

License

MIT. See LICENSE.