OVOS Model2Vec Intent Pipeline

An intent matching pipeline for OpenVoiceOS (OVOS), powered by the Model2Vec model for intent classification.

This plugin uses a pretrained Model2Vec model to classify natural language utterances into intent labels registered with the system (Adapt, Padatious, and plugin-specific labels). It only considers intents from loaded skills and ignores any labels from unregistered intents. This pipeline is ideal for use cases where other deterministic engines fail to provide a high-confidence match.

✨ Features

• ✅ Powered by Model2Vec for high-quality intent classification
• ✅ Plug-and-play integration with OVOS pipelines
• ✅ Model2Vec trained on GitLocalize exports
• ✅ English models in various sizes, distilled from Potion
• ✅ Multilingual model, distilled from LaBSE
• ✅ Syncs Adapt and Padatious intents dynamically at runtime
• ✅ Only considers intents from loaded skills, ignoring unregistered labels

💡 english models size ranges from 8MB to 150MB, the multilingual model (default) is over 500MB

📦 Installation

You can install the plugin via pip:

pip install ovos-m2v-pipeline

⚙️ Configuration

In your mycroft.conf:

{
  "intents": {
    "ovos-m2v-pipeline": {
      "model": "Jarbas/ovos-model2vec-intents-LaBSE",
      "conf_high": 0.7,
      "conf_medium": 0.5,
      "conf_low": 0.15,
      "ignore_intents": []
    }
  }
}

• model: Path to your pretrained Model2Vec model or huggingface repo.
• conf_xxx: Minimum confidence threshold for intent matching.
• ignore_intents: List of intents to ignore during matching.
• prototype_strategy: Scoring strategy for prototype mode ("max_over_all" default — back-compatible). See docs/strategies.md.
• prototype_top_k: Top-k cosines averaged by the top_k_mean strategy (default 3).
• prototype_tau: Softmax temperature for the softmax_weighted strategy (default 0.1).

⚠️ The Model2Vec model is pretrained based on GitLocalize exports and cannot learn new skills dynamically.

🧩 Which entrypoint do I want?

This plugin ships two opm.pipeline entrypoints, backed by the same Model2VecIntentPipeline class but running in different modes:

• ovos-m2v-pipeline (Model2VecIntentPipeline, mode: "classifier", the default) loads a pretrained, frozen classification head with a fixed label set baked in at training time. It is fast and needs no runtime fitting, but it can only ever return the labels it was trained on. It still tracks OVOS-INTENT-4 ovos.intent.register.template registrations from skills so it can gate/allowlist a trained label, but registering a new intent that was not part of training does not teach it to that skill — it will never be matched.
• ovos-m2v-prototype-pipeline (Model2VecPrototypePipeline, mode: "prototype") loads a bare embedding model with no classification head and builds its label set entirely at runtime, from the example utterances supplied by Adapt/Padatious registrations and OVOS-INTENT-4 template registrations. Use this entrypoint whenever skills need to register new intents (including custom/dynamically-created skills) that must actually be matched.

Both entrypoints can be enabled together — configure each independently under its own intents.<entrypoint-name> key (see Model2VecPrototypePipeline docstring for an example) — so a deployment can keep the fast frozen classifier for its core trained intents while the prototype matcher picks up everything else.

🧠 Usage

The Model2VecIntentPipeline class integrates with the OVOS intent system. It:

1. Receives an utterance (text).
2. Predicts intent labels using the pretrained Model2Vec model.
3. Filters out intents that are not part of the loaded skills.
4. Returns a match for the highest-confidence intent from the list of valid intents.

🧪 Tips

• Tune min_conf to control the confidence threshold for intent matching.
• Use the ignore_intents list to filter out specific problematic intent from predictions.
• Syncing of Adapt and Padatious intents is done automatically at runtime via the OVOS message bus.

💡 pre-trained models available in this huggingface collection ovos-model2vec-intents

🛡 License

This project is licensed under the Apache 2.0 License.

Credits

The model2vec intent pipeline was first prototyped by TigreGótico under the ILENIA project for OpenVoiceOS and substantially extended — an embeddings-only mode and new models — through the NGI0 Commons Fund.

This project was funded by the Ministerio para la Transformación Digital y de la Función Pública and Plan de Recuperación, Transformación y Resiliencia - Funded by EU – NextGenerationEU within the framework of the project ILENIA with reference 2022/TL22/00215337

This project was funded through the NGI0 Commons Fund, a fund established by NLnet with financial support from the European Commission's Next Generation Internet programme, under the aegis of DG Communications Networks, Content and Technology under grant agreement No 101135429.