Install ffmpeg for torchcodec runtime support

ASR lazy-load: per-repo cache under /tmp; writable caches; disable lifespan; Farmlingua branding

Reduce storage: move HF caches to /tmp; store model ckpt in /tmp/models; ASR snapshot_download to /tmp with no symlinks

Docker: remove ASR predownload; rely on runtime snapshot_download

Uvicorn: disable lifespan to avoid CancelledError on startup in Spaces

ASR+Ask integration, single /speak, scipy WAV, build-time ASR cache, cleanup

Fix FileResponse background task; add StreamingResponse cleanup; add sounddevice and system libs

TTS: robust ckpt handling; download to /tmp then move; make /code writable

Dockerfile: avoid repo models/ dependency; skip large ckpt; install yarngpt if missing

Dockerfile: fix HF login RUN block; remove heredoc ; parsing issue

Sync: add API, Dockerfile, assets (no large ckpt)

Update Dockerfile

Update Dockerfile

push

push

push

push

push

push

push

push