Whisper Large-v3 Turbo
यह Soniqo पेज स्थानीय speech-swift / speech-core implementation में Whisper Large-v3 Turbo को दस्तावेज़ करता है। Hugging Face bundle links integration notes के बाद दिए गए हैं।
पहले आंतरिक पेज
Landing cards और docs menus पहले इसी पेज पर आते हैं; source model और bundle links यहीं उपलब्ध रहते हैं।
सारांश
| मॉडल | Whisper Large-v3 Turbo |
|---|---|
| भूमिका | General multilingual speech-to-text |
| Backend | CoreML fp16 on CPU and Neural Engine |
| Output | Text transcript |
| भाषाएँ | Whisper multilingual set, about 100 languages |
| लाइसेंस | MIT weights from OpenAI |
| स्थिति | Ready through speech transcribe --engine whisper and the WhisperASR Swift product |
| Source | openai/whisper-large-v3-turbo |
| Swift product | WhisperASR |
| CLI / runtime | speech transcribe --engine whisper |
उपयोग
नीचे का snippet मौजूदा speech-swift API या command से मेल खाता है।
# Transcribe with the native CoreML Whisper runtime.
.build/release/speech transcribe recording.wav --engine whisper
.build/release/speech transcribe recording.wav --engine whisper --language enमॉडल लिंक
implementation notes
- LibriSpeech test-clean slice on an M5 Pro: 1.40% WER, mean RTF 0.089, 6.1 s model load, 384 MB peak RSS — versus 1.53% / 0.085 / 100.2 s / 507 MB for a direct WhisperKit run of the same model.
- The bundle splits into four CoreML models — mel spectrogram, audio encoder, decoder context prefill, and a KV-cached text decoder — with the encoder and decoder on the Neural Engine.
- Greedy no-timestamp decoding with automatic language detection or a --language hint; audio is processed in 30 s chunks with a repeated-word guard against greedy hallucination loops.
- Word timestamps, temperature fallback, and VAD-guided long-form seeking are not implemented yet; --model accepts turbo, large-v3-turbo, or any Hugging Face repo ID.