Results for “parakeet”

语音转文字 (ASR) (语音转文字 (ASR / 语音识别 / Speech-to-Text) 应用工程 (从业者视角) — 选型、集成、优化语音转文字能力，尤其面向移动端、低成本、快速识别的场景。覆盖: (a) 引擎/API 地图 — 云端 API(OpenAI gpt-4o-transcribe / Whisper API、Deepgram、AssemblyAI、Google / Azure / AWS Transcribe、讯飞、字节火山、阿里、腾讯、百度) vs 开源模型(Whisper / faster-whisper / whisper.cpp / distil-whisper、NVIDIA NeMo Parakeet / Canary、阿里 FunASR / Paraformer / SenseVoice、Moonshine、Vosk、Kaldi / k2 / icefall) vs 端侧·移动 SDK(whisper.cpp + CoreML / Metal、iOS Speech framework、Android SpeechRecognizer、Picovoice、SenseVoice 端侧); (b) 准确率(WER / CER) × 延迟(RTF / 流式) × 成本 三角权衡与选型决策树; (c) 成本优化 playbook — 端侧免费 / 批量折扣 / VAD 裁静音 / 量化(int8 / ggml) / 蒸馏 / 自托管 break-even; (d) 移动端集成 — 端侧 vs 云、流式 vs 批量、断点检测(endpointing)、隐私 / 离线; (e) 后处理 — 标点 / 数字规整(ITN) / 说话人分离(diarization) / 时间戳。学派分歧: 云 API vs 端侧自托管、通用大模型(Whisper) vs 专用流式(RNN-T / Conformer)、闭源 API vs 开源、准确率派 vs 成本派、英文优先 vs 中文 ASR(FunASR / SenseVoice / 讯飞)。不含: 文字转语音(TTS / 语音合成，方向相反)、声纹识别 / 说话人验证为主业、语音 agent / 对话式 AI、ASR 模型训练科研深水区。) Master OS — automated mastery of 语音转文字 (ASR / 语音识别 / Speech-to-Text) 应用工程 (从业者视角) — 选型、集成、优化语音转文字能力，尤其面向移动端、低成本、快速识别的场景。覆盖: (a) 引擎/API 地图 — 云端 API(OpenAI gpt-4o-transcribe / Whisper API、Deepgram、AssemblyAI、Google / Azure / AWS Transcribe、讯飞、字节火山、阿里、腾讯、百度) vs 开源模型(Whisper / faster-whisper / whisper.cpp / distil-whisper、NVIDIA NeMo Parakeet / Canary、阿里 FunASR / Paraformer / SenseVoice、Moonshine、Vosk、Kaldi / k2 / icefall) vs 端侧·移动 SDK(whisper.cpp + CoreML / Metal、iOS Speech framework、Android SpeechRecognizer、Picovoice、SenseVoice 端侧); (b) 准确率(WER / CER) × 延迟(RTF / 流式) × 成本 三角权衡与选型决策树; (c) 成本优化 playbook — 端侧免费 / 批量折扣 / VAD 裁静音 / 量化(int8 / ggml) / 蒸馏 / 自托管 break-even; (d) 移动端集成 — 端侧 vs 云、流式 vs 批量、断点检测(endpointing)、隐私 / 离线; (e) 后处理 — 标点 / 数字规整(ITN) / 说话人分离(diarization) / 时间戳。学派分歧: 云 API vs 端侧自托管、通用大模型(Whisper) vs 专用流式(RNN-T / Conformer)、闭源 API vs 开源、准确率派 vs 成本派、英文优先 vs 中文 ASR(FunASR / SenseVoice / 讯飞)。不含: 文字转语音(TTS / 语音合成，方向相反)、声纹识别 / 说话人验证为主业、语音 agent / 对话式 AI、ASR 模型训练科研深水区。: top builders' mental models, tool stack, current workflows, jargon, and where to keep up. Trigger this skill when the user works on 语音转文字 (ASR / 语音识别 / Speech-to-Text) 应用工程 (从业者视角) — 选型、集成、优化语音转文字能力，尤其面向移动端、低成本、快速识别的场景。覆盖: (a) 引擎/API 地图 — 云端 API(OpenAI gpt-4o-transcribe / Whisper API、Deepgram、AssemblyAI、Google / Azure / AWS Transcribe、讯飞、字节火山、阿里、腾讯、百度) vs 开源模型(Whisper / faster-whisper / whisper.cpp / distil-whisper、NVIDIA NeMo Parakeet / Canary、阿里 FunASR / Paraformer / SenseVoice、Moonshine、Vosk、Kaldi / k2 / icefall) vs 端侧·移动 SDK(whisper.cpp + CoreML / Metal、iOS Speech framework、Android SpeechRecognizer、Picovoice、SenseVoice 端侧); (b) 准确率(WER / CER) × 延迟(RTF / 流式) × 成本 三角权衡与选型决策树; (c) 成本优化 playbook — 端侧免费 / 批量折扣 / VAD 裁静音 / 量化(int8 / ggml) / 蒸馏 / 自托管 break-even; (d) 移动端集成 — 端侧 vs 云、流式 vs 批量、断点检测(endpointing)、隐私 / 离线; (e) 后处理 — 标点 / 数字规整(ITN) / 说话人分离(diarization) / 时间戳。学派分歧: 云 API vs 端侧自托管、通用大模型(Whisper) vs 专用流式(RNN-T / Conformer)、闭源 API vs 开源、准确率派 vs 成本派、英文优先 vs 中文 ASR(FunASR / SenseVoice / 讯飞)。不含: 文字转语音(TTS / 语音合成，方向相反)、声纹识别 / 说话人验证为主业、语音 agent / 对话式 AI、ASR 模型训练科研深水区。 problems and wants industry-grade thinking, tool selection, or workflow guidance. 触发词：「语音转文字」「语音识别」「asr」「speech to text」「stt」

The English summary is being prepared. Showing the original description.

Local speech-to-text on Apple Silicon with two cross-checking MLX engines: NVIDIA Parakeet TDT v2 (parakeet-mlx) and Qwen3-ASR 1.7B (mlx-audio, hotword biasing). Batch JSONL with word timestamps, confidence, provenance, and an agreement verdict. No NeMo or PyTorch. Use when transcribing audio files, identifying game or legacy voice clips by content, telling speech from SFX, serving an OpenAI-compatible STT endpoint, or dictating. For push-to-talk into any text field, use the Handy app.

The English summary is being prepared. Showing the original description.

Stage 4 of the Clinical ASR Flywheel. Use when priority KER is above 0.3 to run stock NeMo SFT on Parakeet TDT v2 and offline cycle N+1 re-eval. NOT for generic word boosting (use /finetune-asr).

The English summary is being prepared. Showing the original description.

Routes NVIDIA Nemotron Speech (Formerly Riva) NIM tasks — deploys, runs, and tests ASR, TTS, and NMT NIMs on build.nvidia.com or self-hosted.

The English summary is being prepared. Showing the original description.

Operational guide for the NV oOS Dictation desktop companion app (Tauri v2 + Rust, examples/nvoos-dictate) — build/run on Windows/MSVC, the local-first engine track (parakeet/whisper/fastconformer/moonshine/silero-vad), the embedded remote-STT fallback, the Stage-2 polish provider track (disabled/openai_compat/nvoos plus local LM Studio and Ollama), the file/media decode seam (Symphonia 0.5), the live camera + local object-detection/tracking track (Proposals 073 + 075) with an opt-in VLM label-polish layer (Proposal 076), the CDP browser-control agent (feature agent) + FastConformer engine + NV oOS multi-SPA hub, model registry + verification harnesses, the plugins.updater boot-crash config rule, and the full test/clippy gate matrix. Use when building, extending, or debugging the dictation app, adding an engine/model, a polish provider, the browser agent, file/video transcription, camera/vision detection, or VLM relabeling.

The English summary is being prepared. Showing the original description.