fix: Play produces no sound — switch to streaming audio + system deps

fix: Play produces no sound on HF Space

refactor: consolidate to Base 1.7B only, remove CustomVoice

refactor: runtime config, Q&A flow module, storage paths, and cleanup

feat: enable FlashAttention-2 for all models via runtime install + SDPA fallback

perf: reduce TTS latency - shorter chunks + sub-segment streaming

revert: voice cloning back to GPU - user will use larger GPU

fix: voice cloning OOM on T4 - load 1.7B Base model on CPU

fix: remove custom dtype/attn params to fix CUDA assert on T4

fix: CUDA assert on T4 GPU - use eager attention, disable torch.compile

fix: replace Supertonic with Qwen3-TTS 0.6B CustomVoice, fix Play button

feat: persist voice profiles to Voice_Profile/ with auto-load on startup

perf: add TTS latency optimizations

feat: implement real voice cloning with Qwen3-TTS (sprint step 3)