Spaces:

Luigi
/

Qwen3-ASR-Micro-Streaming

Running

App Files Files Community

Qwen3-ASR-Micro-Streaming

26.3 kB

Ctrl+K

Ctrl+K

1 contributor

History: 63 commits

Luigi's picture

v9: VAD-segment both models per utterance + token-aligned diff view (no real-time)

cde7551 verified about 7 hours ago

.gitattributes

1.57 kB
restore Silero VAD in rs-asr-stream (reuse engine SileroVadModel; vendored gguf, no torch) about 17 hours ago
Dockerfile

809 Bytes
fetch 0.6B from csukuangfj2 HF repo (reliable); drop Dockerfile GitHub fetch about 8 hours ago
README.md

1.86 kB
migrate Gradio -> FastAPI + WebSocket + AudioWorklet (real duplex streaming) about 17 hours ago
app.py

5.33 kB
v9: VAD-segment both models per utterance + token-aligned diff view (no real-time) about 7 hours ago
index.html

16 kB
v9: VAD-segment both models per utterance + token-aligned diff view (no real-time) about 7 hours ago
requirements.txt

88 Bytes
greedy decode + max_new_tokens=200 + OpenCC s2twp (consistent zh-TW Traditional) about 12 hours ago
worklet.js

634 Bytes
migrate Gradio -> FastAPI + WebSocket + AudioWorklet (real duplex streaming) about 17 hours ago