| license: cc-by-4.0 | |
| language: | |
| - en | |
| tags: | |
| - speech | |
| - asr | |
| - coreml | |
| - parakeet | |
| - transducer | |
| - ios | |
| base_model: nvidia/parakeet-tdt-0.6b-v2 | |
| # Parakeet TDT v3 — CoreML INT8 (iOS, 5s) | |
| CoreML INT8 conversion of [NVIDIA Parakeet-TDT 0.6B v2](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v2) for iOS, with encoder shape optimized for audio segments up to 5 seconds. Smaller and faster than the variable-length variant for short-form audio. | |
| ## Models | |
| | Model | Description | Compute | Quantization | | |
| |-------|-------------|---------|-------------| | |
| | `encoder.mlmodelc` | FastConformer encoder (5s max) | Neural Engine | INT8 palettized | | |
| | `decoder.mlmodelc` | LSTM prediction network | Neural Engine | FP16 | | |
| ## Usage | |
| Used by [speech-swift](https://github.com/soniqo/speech-swift) `ParakeetASR` module: | |
| ```swift | |
| let model = try await ParakeetASRModel.fromPretrained() | |
| let text = try model.transcribeAudio(samples, sampleRate: 16000) | |
| ``` | |
| --- | |
| - **Guide**: [soniqo.audio/guides/parakeet](https://soniqo.audio/guides/parakeet) | |
| - **Docs**: [soniqo.audio](https://soniqo.audio) | |
| - **GitHub**: [soniqo/speech-swift](https://github.com/soniqo/speech-swift) | |