Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.
Python asraudiochineseemotion-recognitionfunasrmcp-servermultilingual-asropenai-compatible-apiparaformerpunctuationpytorchreal-time-asrspeaker-diarizationspeech-recognitionspeech-to-textstreaming-asrtranscriptionvllmvoice-activity-detectionwhisper-alternative
Commits on main
Merge branch 'vllm' - vLLM inference engine + streaming WebSocket service
a23bdcd 游雁 committed 5 months ago
fix: Fun-ASR-Nano batch_size check - use len(data_in) instead of kwargs batch_size_ms
d008cbf xiaoyunchong.xyc committed 5 months ago
feat: add dynamic silence threshold to fsmn-vad model (default on)
b9bcf9d xiaoyunchong.xyc committed 6 months ago
refactor: move DynamicStreamingVAD to funasr/models/fsmn_vad_streaming/
7be505f xiaoyunchong.xyc committed 6 months ago
feat: extract DynamicStreamingVAD as reusable component
ca352d9 xiaoyunchong.xyc committed 6 months ago
docs: add detailed inference logic section to realtime_demo.md
84d24ab xiaoyunchong.xyc committed 6 months ago
docs: sync vllm_streaming.md with code - fix params, add AutoModelVLLM comparison
6e3663c xiaoyunchong.xyc committed 6 months ago
docs: fix vLLM version requirement, add AutoModelVLLM usage to vllm_inference.md
55c035f xiaoyunchong.xyc committed 6 months ago
docs: add WebSocket streaming server section to vllm_guide.md
6ee9620 xiaoyunchong.xyc committed 6 months ago
docs: update realtime_demo.md with vLLM engine details
753a0bb xiaoyunchong.xyc committed 6 months ago
feat: switch streaming ASR to vLLM inference via AutoModelVLLM
81e2fc0 xiaoyunchong.xyc committed 6 months ago