{{ '推荐路径' if language == 'zh' else 'Recommended path' }}
OPEN SOURCE / SPEECH INTELLIGENCE
{{ '让语音,成为可用的信息。' if language == 'zh' else 'Turn speech into useful information.' }}
{{ '语音识别、时间戳、说话人分离。从一行 Python,到你自己的生产服务。' if language == 'zh' else 'Recognition, timestamps, and diarization. From a line of Python to your own production service.' }}
欢迎大家来体验达摩院推出的语音识别模型。
{{ '一个工具箱,多种可能' if language == 'zh' else 'ONE TOOLKIT. MANY POSSIBILITIES.' }}
{{ '开发者文档' if language == 'zh' else 'DEVELOPER DOCUMENTATION' }}
{{ '部署选择器' if language == 'zh' else 'Deployment selector' }}
{{ '从本地 CPU 到 GPU 集群,每条路径都有安装步骤、测试记录和已知限制。' if language == 'zh' else 'From a local CPU to GPU infrastructure, with setup steps, test records, and known limitations.' }}
{{ '推荐路径' if language == 'zh' else 'Recommended path' }}
{{ '完整矩阵' if language == 'zh' else 'Full matrix' }}
| {{ '方案' if language == 'zh' else 'Path' }} | {{ '工作负载' if language == 'zh' else 'Workload' }} | {{ '硬件' if language == 'zh' else 'Hardware' }} | {{ '接口' if language == 'zh' else 'Interface' }} | {{ '状态' if language == 'zh' else 'Status' }} |
|---|---|---|---|---|
| {{ item.translations[language].name }} | {{ item.workloads | join(' / ') }} | {{ item.hardware | join(' / ') }} | {{ item.interfaces[:2] | join(' / ') }} | {{ item.translations[language].status_label }} |
{{ '可验证接口' if language == 'zh' else 'Inspectable API contract' }}
{{ '健康、模型身份和真实转写是三个独立检查。' if language == 'zh' else 'Health, model identity, and real transcription are three separate checks.' }}
curl -fsS http://localhost:8000/health
# {"status":"ok"}
curl -fsS http://localhost:8000/v1/models
# {"data":[{"id":"sensevoice"}]}
curl -fsS http://localhost:8000/v1/audio/transcriptions \
-F file=@meeting.wav -F model=sensevoice
{{ '生产边界' if language == 'zh' else 'Production boundary' }}
{{ '模型加载、转写接口、时间戳、VAD、流式协议和运行时日志。' if language == 'zh' else 'Model loading, transcription, timestamps, VAD, streaming protocols, and runtime logs.' }}
{{ 'TLS、认证、租户配额、限流、审计、上传策略、监控和滚动发布。' if language == 'zh' else 'TLS, authentication, tenant quotas, rate limits, audit, upload policy, monitoring, and rollout.' }}
{{ '目标硬件的容量测试、已知音频 smoke test、版本清单和可执行回滚。' if language == 'zh' else 'Target-hardware capacity tests, known-audio smoke tests, version manifests, and executable rollback.' }}
{{ '生态仓库' if language == 'zh' else 'Ecosystem repositories' }}
{{ '按目标进入专用仓库,核对模型、部署边界与可复现示例。' if language == 'zh' else 'Open the focused repository for its models, deployment boundaries, and reproducible examples.' }}
{{ '工业语音工具箱' if language == 'zh' else 'Production speech toolkit' }}
{{ '统一语音识别、VAD、标点、说话人分离与工业部署的开源工具箱。' if language == 'zh' else 'The open toolkit unifying ASR, VAD, punctuation, diarization, and production deployment.' }}
{{ '查看 FunASR' if language == 'zh' else 'Open FunASR on GitHub' }}{{ '高精度转写' if language == 'zh' else 'High-accuracy ASR' }}
{{ '端到端语音识别模型家族,覆盖旗舰与 31 语言 MLT 检查点。' if language == 'zh' else 'A family of end-to-end speech recognition models spanning flagship and 31-language MLT checkpoints.' }}
{{ '查看 Fun-ASR' if language == 'zh' else 'Open Fun-ASR on GitHub' }}{{ '多语种与情感' if language == 'zh' else 'Languages and emotion' }}
{{ '覆盖语音识别、语言、情感与音频事件的多语言理解模型。' if language == 'zh' else 'Multilingual understanding across speech recognition, language, emotion, and audio events.' }}
{{ '查看 SenseVoice' if language == 'zh' else 'Open SenseVoice on GitHub' }}{{ '智能视频剪辑' if language == 'zh' else 'AI video editing' }}
{{ '以语音与文本定位为核心的视频理解与智能剪辑工作流。' if language == 'zh' else 'Video understanding and intelligent clipping workflows driven by speech and text localization.' }}
{{ '查看 FunClip' if language == 'zh' else 'Open FunClip on GitHub' }}{{ '下一步' if language == 'zh' else 'Next step' }}