794ecbedc9
config: back to sherpa TTS, Qwen3 0.6B still too slow on 3090
912b54547a
config: switch to Qwen3-TTS 0.6B for faster inference
2ec0dcf401
config: switch back to sherpa TTS for better latency
d5eb5d6a38
fix: revert to single GPU, multi-GPU breaks TTS
17923f3bdc
feat: TTS on 2 GPUs (cuda:2,cuda:3) for faster inference
b75e813c03
feat: enable flash_attention_2 for Qwen3-TTS
78bc3f71c0
fix: disable end_prompt, filter more thinking patterns from antaf
fb53c4a6b7
fix: replace 蚂蚁集团/蚂蚁/支付宝/健康是福 in antaf responses
d3fd9cc391
feat: add Qwen3-TTS CustomVoice GPU provider, switch TTS
51fa106d7d
feat: replace 蚂蚁阿福/阿福 with 泰小虎 in antaf responses
21998c0777
fix: filter markdown tables, status text, residual formatting from antaf
49ae06ae45
fix: strip markdown links, URLs, bold/italic from antaf response
6707351540
config: remove all system prompts, let antaf use its own persona
8e371b827f
fix: remove appended hint, only send pure user text to antaf
58c67de338
config: use Debian public IP for antaf bridge
56042fddf7
config: switch LLM to antaf (text bridge)
c55a7a010b
config: rename 阿福 to 小虎, wake word 小虎小虎