shorten desc

Merge pull request #13 from MomsVoiceAI/fix/readme-hf-trag

add preload_from_hub

Merge pull request #12 from MomsVoiceAI/fix/readme-hf-trag

update tag in readme

add youtube link in readme

fix README demo img

fix README syntax err

update README link

update README team members

fix duplicated demo step

Merge pull request #11 from MomsVoiceAI/fix/README-spec

update README demo

Merge pull request #10 from MomsVoiceAI/fix/lfm-model-param

update report md

fix prompt for short answer

fix to ui to visible

Fix record button

perf: improve LFM2.5-Audio inference quality and reliability

Merge pull request #8 from MomsVoiceAI/fix/answer-audio

perf: warmup generation, async LFM loading, reduce max_new_tokens

perf: add torch.compile for LFM model on HF Space (20-60% faster inference)

Merge pull request #9 from MomsVoiceAI/fix/sox-pkg

fix: restore packages.txt with sox and ffmpeg dependencies for HF Spaces

fix: proper Mimi decode on HF Space — only disable dynamo on Windows, explicit device handling

fix: use Mimi streaming decoder for proper speech audio (not LFM detokenizer)

feat: restore story cover PNGs via Xet storage + migration docs

remove packages.txt - ffmpeg already in base image

fix: packages.txt use only ffmpeg

fix: remove invalid libsox-fmt-all package

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

feat: upload pre-generated audio and copy to runtime cache on startup

fix: Play produces no sound on HF Space

fix: correct LFM audio sample rate to 48kHz and normalize volume

refactor: consolidate to Base 1.7B only, remove CustomVoice

fix: audio decode normalization + warmup first 5 paragraphs

feat: show paragraph progress during Play + warmup first paragraphs at startup

ui: hide chunk/queue processing details from user during Play

feat: auto-start mic recording on Ask click (100ms user gesture window)

fix: add stop_recording JS fallback to auto-submit Q&A

fix: dedup only on auto-submit, manual Get Answer always works

fix: convert Path to str for wave.open (Python 3.13 compat)

feat: bounded pre-gen, cancellation, disk cache pruning, GPU lock

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

perf: 5 latency optimizations for TTS playback

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

rename: app.py -> app_cascade.py, app_lfm.py -> app.py

feat: add LFM2.5-Audio end-to-end Q&A variant

fix: auto-submit - JS clicks submit 500ms after stop_recording + .change backup

fix: auto-submit Q&A - stop_recording + change + JS fallback click